diff --git a/Documentation/00-INDEX b/Documentation/00-INDEX index a82a113b4a4b1d3aa6aaf2f9815076ab4f5f2627..1977fab386566e23f501bea233f8f487d16f5c38 100644 --- a/Documentation/00-INDEX +++ b/Documentation/00-INDEX @@ -329,8 +329,6 @@ sgi-visws.txt - short blurb on the SGI Visual Workstations. sh/ - directory with info on porting Linux to a new architecture. -smart-config.txt - - description of the Smart Config makefile feature. sound/ - directory with info on sound card support. sparc/ diff --git a/Documentation/i386/boot.txt b/Documentation/i386/boot.txt index 2eb16100bb3fef7658799f5f7f4d259be644b537..0fac3465f2e38e5612a31cfa83e83cdcad95e83e 100644 --- a/Documentation/i386/boot.txt +++ b/Documentation/i386/boot.txt @@ -42,6 +42,8 @@ Protocol 2.05: (Kernel 2.6.20) Make protected mode kernel relocatable. Protocol 2.06: (Kernel 2.6.22) Added a field that contains the size of the boot command line +Protocol 2.09: (kernel 2.6.26) Added a field of 64-bit physical + pointer to single linked list of struct setup_data. **** MEMORY LAYOUT @@ -172,6 +174,8 @@ Offset Proto Name Meaning 0240/8 2.07+ hardware_subarch_data Subarchitecture-specific data 0248/4 2.08+ payload_offset Offset of kernel payload 024C/4 2.08+ payload_length Length of kernel payload +0250/8 2.09+ setup_data 64-bit physical pointer to linked list + of struct setup_data (1) For backwards compatibility, if the setup_sects field contains 0, the real value is 4. @@ -572,6 +576,28 @@ command line is entered using the following protocol: covered by setup_move_size, so you may need to adjust this field. +Field name: setup_data +Type: write (obligatory) +Offset/size: 0x250/8 +Protocol: 2.09+ + + The 64-bit physical pointer to NULL terminated single linked list of + struct setup_data. This is used to define a more extensible boot + parameters passing mechanism. The definition of struct setup_data is + as follow: + + struct setup_data { + u64 next; + u32 type; + u32 len; + u8 data[0]; + }; + + Where, the next is a 64-bit physical pointer to the next node of + linked list, the next field of the last node is 0; the type is used + to identify the contents of data; the len is the length of data + field; the data holds the real payload. + **** MEMORY LAYOUT OF THE REAL-MODE CODE diff --git a/Documentation/ia64/kvm.txt b/Documentation/ia64/kvm.txt new file mode 100644 index 0000000000000000000000000000000000000000..bec9d815da33a0495c7ea4d2eebecf7f4c1ab2fc --- /dev/null +++ b/Documentation/ia64/kvm.txt @@ -0,0 +1,82 @@ +Currently, kvm module in EXPERIMENTAL stage on IA64. This means that +interfaces are not stable enough to use. So, plase had better don't run +critical applications in virtual machine. We will try our best to make it +strong in future versions! + Guide: How to boot up guests on kvm/ia64 + +This guide is to describe how to enable kvm support for IA-64 systems. + +1. Get the kvm source from git.kernel.org. + Userspace source: + git clone git://git.kernel.org/pub/scm/virt/kvm/kvm-userspace.git + Kernel Source: + git clone git://git.kernel.org/pub/scm/linux/kernel/git/xiantao/kvm-ia64.git + +2. Compile the source code. + 2.1 Compile userspace code: + (1)cd ./kvm-userspace + (2)./configure + (3)cd kernel + (4)make sync LINUX= $kernel_dir (kernel_dir is the directory of kernel source.) + (5)cd .. + (6)make qemu + (7)cd qemu; make install + + 2.2 Compile kernel source code: + (1) cd ./$kernel_dir + (2) Make menuconfig + (3) Enter into virtualization option, and choose kvm. + (4) make + (5) Once (4) done, make modules_install + (6) Make initrd, and use new kernel to reboot up host machine. + (7) Once (6) done, cd $kernel_dir/arch/ia64/kvm + (8) insmod kvm.ko; insmod kvm-intel.ko + +Note: For step 2, please make sure that host page size == TARGET_PAGE_SIZE of qemu, otherwise, may fail. + +3. Get Guest Firmware named as Flash.fd, and put it under right place: + (1) If you have the guest firmware (binary) released by Intel Corp for Xen, use it directly. + + (2) If you have no firmware at hand, Please download its source from + hg clone http://xenbits.xensource.com/ext/efi-vfirmware.hg + you can get the firmware's binary in the directory of efi-vfirmware.hg/binaries. + + (3) Rename the firware you owned to Flash.fd, and copy it to /usr/local/share/qemu + +4. Boot up Linux or Windows guests: + 4.1 Create or install a image for guest boot. If you have xen experience, it should be easy. + + 4.2 Boot up guests use the following command. + /usr/local/bin/qemu-system-ia64 -smp xx -m 512 -hda $your_image + (xx is the number of virtual processors for the guest, now the maximum value is 4) + +5. Known possibile issue on some platforms with old Firmware. + +If meet strange host crashe issues, try to solve it through either of the following ways: + +(1): Upgrade your Firmware to the latest one. + +(2): Applying the below patch to kernel source. +diff --git a/arch/ia64/kernel/pal.S b/arch/ia64/kernel/pal.S +index 0b53344..f02b0f7 100644 +--- a/arch/ia64/kernel/pal.S ++++ b/arch/ia64/kernel/pal.S +@@ -84,7 +84,8 @@ GLOBAL_ENTRY(ia64_pal_call_static) + mov ar.pfs = loc1 + mov rp = loc0 + ;; +- srlz.d // seralize restoration of psr.l ++ srlz.i // seralize restoration of psr.l ++ ;; + br.ret.sptk.many b0 + END(ia64_pal_call_static) + +6. Bug report: + If you found any issues when use kvm/ia64, Please post the bug info to kvm-ia64-devel mailing list. + https://lists.sourceforge.net/lists/listinfo/kvm-ia64-devel/ + +Thanks for your interest! Let's work together, and make kvm/ia64 stronger and stronger! + + + Xiantao Zhang + 2008.3.10 diff --git a/Documentation/ide/ide-tape.txt b/Documentation/ide/ide-tape.txt index 658f271a373fca8c57dc50a32dc04ad494250fcf..3f348a0b21d83e4f9354d90d47ca61b221543fad 100644 --- a/Documentation/ide/ide-tape.txt +++ b/Documentation/ide/ide-tape.txt @@ -1,146 +1,65 @@ -/* - * IDE ATAPI streaming tape driver. - * - * This driver is a part of the Linux ide driver. - * - * The driver, in co-operation with ide.c, basically traverses the - * request-list for the block device interface. The character device - * interface, on the other hand, creates new requests, adds them - * to the request-list of the block device, and waits for their completion. - * - * Pipelined operation mode is now supported on both reads and writes. - * - * The block device major and minor numbers are determined from the - * tape's relative position in the ide interfaces, as explained in ide.c. - * - * The character device interface consists of the following devices: - * - * ht0 major 37, minor 0 first IDE tape, rewind on close. - * ht1 major 37, minor 1 second IDE tape, rewind on close. - * ... - * nht0 major 37, minor 128 first IDE tape, no rewind on close. - * nht1 major 37, minor 129 second IDE tape, no rewind on close. - * ... - * - * The general magnetic tape commands compatible interface, as defined by - * include/linux/mtio.h, is accessible through the character device. - * - * General ide driver configuration options, such as the interrupt-unmask - * flag, can be configured by issuing an ioctl to the block device interface, - * as any other ide device. - * - * Our own ide-tape ioctl's can be issued to either the block device or - * the character device interface. - * - * Maximal throughput with minimal bus load will usually be achieved in the - * following scenario: - * - * 1. ide-tape is operating in the pipelined operation mode. - * 2. No buffering is performed by the user backup program. - * - * Testing was done with a 2 GB CONNER CTMA 4000 IDE ATAPI Streaming Tape Drive. - * - * Here are some words from the first releases of hd.c, which are quoted - * in ide.c and apply here as well: - * - * | Special care is recommended. Have Fun! - * - * - * An overview of the pipelined operation mode. - * - * In the pipelined write mode, we will usually just add requests to our - * pipeline and return immediately, before we even start to service them. The - * user program will then have enough time to prepare the next request while - * we are still busy servicing previous requests. In the pipelined read mode, - * the situation is similar - we add read-ahead requests into the pipeline, - * before the user even requested them. - * - * The pipeline can be viewed as a "safety net" which will be activated when - * the system load is high and prevents the user backup program from keeping up - * with the current tape speed. At this point, the pipeline will get - * shorter and shorter but the tape will still be streaming at the same speed. - * Assuming we have enough pipeline stages, the system load will hopefully - * decrease before the pipeline is completely empty, and the backup program - * will be able to "catch up" and refill the pipeline again. - * - * When using the pipelined mode, it would be best to disable any type of - * buffering done by the user program, as ide-tape already provides all the - * benefits in the kernel, where it can be done in a more efficient way. - * As we will usually not block the user program on a request, the most - * efficient user code will then be a simple read-write-read-... cycle. - * Any additional logic will usually just slow down the backup process. - * - * Using the pipelined mode, I get a constant over 400 KBps throughput, - * which seems to be the maximum throughput supported by my tape. - * - * However, there are some downfalls: - * - * 1. We use memory (for data buffers) in proportional to the number - * of pipeline stages (each stage is about 26 KB with my tape). - * 2. In the pipelined write mode, we cheat and postpone error codes - * to the user task. In read mode, the actual tape position - * will be a bit further than the last requested block. - * - * Concerning (1): - * - * 1. We allocate stages dynamically only when we need them. When - * we don't need them, we don't consume additional memory. In - * case we can't allocate stages, we just manage without them - * (at the expense of decreased throughput) so when Linux is - * tight in memory, we will not pose additional difficulties. - * - * 2. The maximum number of stages (which is, in fact, the maximum - * amount of memory) which we allocate is limited by the compile - * time parameter IDETAPE_MAX_PIPELINE_STAGES. - * - * 3. The maximum number of stages is a controlled parameter - We - * don't start from the user defined maximum number of stages - * but from the lower IDETAPE_MIN_PIPELINE_STAGES (again, we - * will not even allocate this amount of stages if the user - * program can't handle the speed). We then implement a feedback - * loop which checks if the pipeline is empty, and if it is, we - * increase the maximum number of stages as necessary until we - * reach the optimum value which just manages to keep the tape - * busy with minimum allocated memory or until we reach - * IDETAPE_MAX_PIPELINE_STAGES. - * - * Concerning (2): - * - * In pipelined write mode, ide-tape can not return accurate error codes - * to the user program since we usually just add the request to the - * pipeline without waiting for it to be serviced. In case an error - * occurs, I will report it on the next user request. - * - * In the pipelined read mode, subsequent read requests or forward - * filemark spacing will perform correctly, as we preserve all blocks - * and filemarks which we encountered during our excess read-ahead. - * - * For accurate tape positioning and error reporting, disabling - * pipelined mode might be the best option. - * - * You can enable/disable/tune the pipelined operation mode by adjusting - * the compile time parameters below. - * - * - * Possible improvements. - * - * 1. Support for the ATAPI overlap protocol. - * - * In order to maximize bus throughput, we currently use the DSC - * overlap method which enables ide.c to service requests from the - * other device while the tape is busy executing a command. The - * DSC overlap method involves polling the tape's status register - * for the DSC bit, and servicing the other device while the tape - * isn't ready. - * - * In the current QIC development standard (December 1995), - * it is recommended that new tape drives will *in addition* - * implement the ATAPI overlap protocol, which is used for the - * same purpose - efficient use of the IDE bus, but is interrupt - * driven and thus has much less CPU overhead. - * - * ATAPI overlap is likely to be supported in most new ATAPI - * devices, including new ATAPI cdroms, and thus provides us - * a method by which we can achieve higher throughput when - * sharing a (fast) ATA-2 disk with any (slow) new ATAPI device. - */ +IDE ATAPI streaming tape driver. + +This driver is a part of the Linux ide driver. + +The driver, in co-operation with ide.c, basically traverses the +request-list for the block device interface. The character device +interface, on the other hand, creates new requests, adds them +to the request-list of the block device, and waits for their completion. + +The block device major and minor numbers are determined from the +tape's relative position in the ide interfaces, as explained in ide.c. + +The character device interface consists of the following devices: + +ht0 major 37, minor 0 first IDE tape, rewind on close. +ht1 major 37, minor 1 second IDE tape, rewind on close. +... +nht0 major 37, minor 128 first IDE tape, no rewind on close. +nht1 major 37, minor 129 second IDE tape, no rewind on close. +... + +The general magnetic tape commands compatible interface, as defined by +include/linux/mtio.h, is accessible through the character device. + +General ide driver configuration options, such as the interrupt-unmask +flag, can be configured by issuing an ioctl to the block device interface, +as any other ide device. + +Our own ide-tape ioctl's can be issued to either the block device or +the character device interface. + +Maximal throughput with minimal bus load will usually be achieved in the +following scenario: + + 1. ide-tape is operating in the pipelined operation mode. + 2. No buffering is performed by the user backup program. + +Testing was done with a 2 GB CONNER CTMA 4000 IDE ATAPI Streaming Tape Drive. + +Here are some words from the first releases of hd.c, which are quoted +in ide.c and apply here as well: + +| Special care is recommended. Have Fun! + +Possible improvements: + +1. Support for the ATAPI overlap protocol. + +In order to maximize bus throughput, we currently use the DSC +overlap method which enables ide.c to service requests from the +other device while the tape is busy executing a command. The +DSC overlap method involves polling the tape's status register +for the DSC bit, and servicing the other device while the tape +isn't ready. + +In the current QIC development standard (December 1995), +it is recommended that new tape drives will *in addition* +implement the ATAPI overlap protocol, which is used for the +same purpose - efficient use of the IDE bus, but is interrupt +driven and thus has much less CPU overhead. + +ATAPI overlap is likely to be supported in most new ATAPI +devices, including new ATAPI cdroms, and thus provides us +a method by which we can achieve higher throughput when +sharing a (fast) ATA-2 disk with any (slow) new ATAPI device. diff --git a/Documentation/ide/ide.txt b/Documentation/ide/ide.txt index 486c699f4aea4722bb822a7af7ad12053f5c321c..0c78f4b1d9d9f1df5298e9147c058a32aee90353 100644 --- a/Documentation/ide/ide.txt +++ b/Documentation/ide/ide.txt @@ -82,27 +82,26 @@ Drives are normally found by auto-probing and/or examining the CMOS/BIOS data. For really weird situations, the apparent (fdisk) geometry can also be specified on the kernel "command line" using LILO. The format of such lines is: - hdx=cyls,heads,sects -or hdx=cdrom + ide_core.chs=[interface_number.device_number]:cyls,heads,sects +or ide_core.cdrom=[interface_number.device_number] -where hdx can be any of hda through hdh, Three values are required -(cyls,heads,sects). For example: +For example: - hdc=1050,32,64 hdd=cdrom + ide_core.chs=1.0:1050,32,64 ide_core.cdrom=1.1 -either {hda,hdb} or {hdc,hdd}. The results of successful auto-probing may -override the physical geometry/irq specified, though the "original" geometry -may be retained as the "logical" geometry for partitioning purposes (fdisk). +The results of successful auto-probing may override the physical geometry/irq +specified, though the "original" geometry may be retained as the "logical" +geometry for partitioning purposes (fdisk). If the auto-probing during boot time confuses a drive (ie. the drive works with hd.c but not with ide.c), then an command line option may be specified for each drive for which you'd like the drive to skip the hardware probe/identification sequence. For example: - hdb=noprobe + ide_core.noprobe=0.1 or - hdc=768,16,32 - hdc=noprobe + ide_core.chs=1.0:768,16,32 + ide_core.noprobe=1.0 Note that when only one IDE device is attached to an interface, it should be jumpered as "single" or "master", *not* "slave". Many folks have had @@ -118,9 +117,9 @@ If for some reason your cdrom drive is *not* found at boot time, you can force the probe to look harder by supplying a kernel command line parameter via LILO, such as: - hdc=cdrom /* hdc = "master" on second interface */ + ide_core.cdrom=1.0 /* "master" on second interface (hdc) */ or - hdd=cdrom /* hdd = "slave" on second interface */ + ide_core.cdrom=1.1 /* "slave" on second interface (hdd) */ For example, a GW2000 system might have a hard drive on the primary interface (/dev/hda) and an IDE cdrom drive on the secondary interface @@ -174,9 +173,7 @@ to /etc/modprobe.conf. When ide.c is used as a module, you can pass command line parameters to the driver using the "options=" keyword to insmod, while replacing any ',' with -';'. For example: - - insmod ide.o options="hda=nodma hdb=nodma" +';'. ================================================================================ @@ -184,57 +181,6 @@ driver using the "options=" keyword to insmod, while replacing any ',' with Summary of ide driver parameters for kernel command line -------------------------------------------------------- - "hdx=" is recognized for all "x" from "a" to "u", such as "hdc". - - "idex=" is recognized for all "x" from "0" to "9", such as "ide1". - - "hdx=noprobe" : drive may be present, but do not probe for it - - "hdx=none" : drive is NOT present, ignore cmos and do not probe - - "hdx=nowerr" : ignore the WRERR_STAT bit on this drive - - "hdx=cdrom" : drive is present, and is a cdrom drive - - "hdx=cyl,head,sect" : disk drive is present, with specified geometry - - "hdx=autotune" : driver will attempt to tune interface speed - to the fastest PIO mode supported, - if possible for this drive only. - Not fully supported by all chipset types, - and quite likely to cause trouble with - older/odd IDE drives. - - "hdx=nodma" : disallow DMA - - "idebus=xx" : inform IDE driver of VESA/PCI bus speed in MHz, - where "xx" is between 20 and 66 inclusive, - used when tuning chipset PIO modes. - For PCI bus, 25 is correct for a P75 system, - 30 is correct for P90,P120,P180 systems, - and 33 is used for P100,P133,P166 systems. - If in doubt, use idebus=33 for PCI. - As for VLB, it is safest to not specify it. - Bigger values are safer than smaller ones. - - "idex=serialize" : do not overlap operations on idex. Please note - that you will have to specify this option for - both the respective primary and secondary channel - to take effect. - - "idex=reset" : reset interface after probe - - "idex=ata66" : informs the interface that it has an 80c cable - for chipsets that are ATA-66 capable, but the - ability to bit test for detection is currently - unknown. - - "ide=doubler" : probe/support IDE doublers on Amiga - -There may be more options than shown -- use the source, Luke! - -Everything else is rejected with a "BAD OPTION" message. - For legacy IDE VLB host drivers (ali14xx/dtc2278/ht6560b/qd65xx/umc8672) you need to explicitly enable probing by using "probe" kernel parameter, i.e. to enable probing for ALI M14xx chipsets (ali14xx host driver) use: @@ -251,6 +197,33 @@ are detected automatically). You also need to use "probe" kernel parameter for ide-4drives driver (support for IDE generic chipset with four drives on one port). +To enable support for IDE doublers on Amiga use "doubler" kernel parameter +for gayle host driver (i.e. "gayle.doubler" if the driver is built-in). + +To force ignoring cable detection (this should be needed only if you're using +short 40-wires cable which cannot be automatically detected - if this is not +a case please report it as a bug instead) use "ignore_cable" kernel parameter: + +* "ide_core.ignore_cable=[interface_number]" boot option if IDE is built-in + (i.e. "ide_core.ignore_cable=1" to force ignoring cable for "ide1") + +* "ignore_cable=[interface_number]" module parameter (for ide_core module) + if IDE is compiled as module + +Other kernel parameters for ide_core are: + +* "nodma=[interface_number.device_number]" to disallow DMA for a device + +* "noflush=[interface_number.device_number]" to disable flush requests + +* "noprobe=[interface_number.device_number]" to skip probing + +* "nowerr=[interface_number.device_number]" to ignore the WRERR_STAT bit + +* "cdrom=[interface_number.device_number]" to force device as a CD-ROM + +* "chs=[interface_number.device_number]" to force device as a disk (using CHS) + ================================================================================ Some Terminology diff --git a/Documentation/ioctl-number.txt b/Documentation/ioctl-number.txt index c18363bd8d11b1f8b3c725c7c12639568033a115..240ce7a56c40c9edfd141468ef2f2f1cf40dfab5 100644 --- a/Documentation/ioctl-number.txt +++ b/Documentation/ioctl-number.txt @@ -183,6 +183,8 @@ Code Seq# Include File Comments 0xAC 00-1F linux/raw.h 0xAD 00 Netfilter device in development: +0xAE all linux/kvm.h Kernel-based Virtual Machine + 0xB0 all RATIO devices in development: 0xB1 00-1F PPPoX diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index bf6303ec0bde4784b71370457eb8510c6198ed05..e5f3d918316f5d421cafc88da6b9baf005c929ac 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -772,10 +772,6 @@ and is between 256 and 4096 characters. It is defined in the file Format: ide=nodma or ide=doubler See Documentation/ide/ide.txt. - ide?= [HW] (E)IDE subsystem - Format: ide?=ata66 or chipset specific parameters. - See Documentation/ide/ide.txt. - idebus= [HW] (E)IDE subsystem - VLB/PCI bus speed See Documentation/ide/ide.txt. diff --git a/Documentation/mips/AU1xxx_IDE.README b/Documentation/mips/AU1xxx_IDE.README index 5c8334123f4f7b6f29d41c6129ecd511a7e03d10..25a6ed1aaa5b1e3e7aa2e774a7beab34f78001f9 100644 --- a/Documentation/mips/AU1xxx_IDE.README +++ b/Documentation/mips/AU1xxx_IDE.README @@ -46,8 +46,6 @@ Two files are introduced: a) 'include/asm-mips/mach-au1x00/au1xxx_ide.h' containes : struct _auide_hwif - struct drive_list_entry dma_white_list - struct drive_list_entry dma_black_list timing parameters for PIO mode 0/1/2/3/4 timing parameters for MWDMA 0/1/2 @@ -63,12 +61,6 @@ Four configs variables are introduced: CONFIG_BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ - maximum transfer size per descriptor -If MWDMA is enabled and the connected hard disc is not on the white list, the -kernel switches to a "safe mwdma mode" at boot time. In this mode the IDE -performance is substantial slower then in full speed mwdma. In this case -please add your hard disc to the white list (follow instruction from 'ADD NEW -HARD DISC TO WHITE OR BLACK LIST' section). - SUPPORTED IDE MODES ------------------- @@ -120,44 +112,6 @@ CONFIG_IDEDMA_AUTO=y Also undefine 'IDE_AU1XXX_BURSTMODE' in 'drivers/ide/mips/au1xxx-ide.c' to disable the burst support on DBDMA controller. -ADD NEW HARD DISC TO WHITE OR BLACK LIST ----------------------------------------- - -Step 1 : detect the model name of your hard disc - - a) connect your hard disc to the AU1XXX - - b) boot your kernel and get the hard disc model. - - Example boot log: - - --snipped-- - Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2 - ide: Assuming 50MHz system bus speed for PIO modes; override with idebus=xx - Au1xxx IDE(builtin) configured for MWDMA2 - Probing IDE interface ide0... - hda: Maxtor 6E040L0, ATA DISK drive - ide0 at 0xac800000-0xac800007,0xac8001c0 on irq 64 - hda: max request size: 64KiB - hda: 80293248 sectors (41110 MB) w/2048KiB Cache, CHS=65535/16/63, (U)DMA - --snipped-- - - In this example 'Maxtor 6E040L0'. - -Step 2 : edit 'include/asm-mips/mach-au1x00/au1xxx_ide.h' - - Add your hard disc to the dma_white_list or dma_black_list structur. - -Step 3 : Recompile the kernel - - Enable MWDMA support in the kernel configuration. Recompile the kernel and - reboot. - -Step 4 : Tests - - If you have add a hard disc to the white list, please run some stress tests - for verification. - ACKNOWLEDGMENTS --------------- diff --git a/Documentation/powerpc/kvm_440.txt b/Documentation/powerpc/kvm_440.txt new file mode 100644 index 0000000000000000000000000000000000000000..c02a003fa03a8b9580dbabf60feda4fb41952fa3 --- /dev/null +++ b/Documentation/powerpc/kvm_440.txt @@ -0,0 +1,41 @@ +Hollis Blanchard +15 Apr 2008 + +Various notes on the implementation of KVM for PowerPC 440: + +To enforce isolation, host userspace, guest kernel, and guest userspace all +run at user privilege level. Only the host kernel runs in supervisor mode. +Executing privileged instructions in the guest traps into KVM (in the host +kernel), where we decode and emulate them. Through this technique, unmodified +440 Linux kernels can be run (slowly) as guests. Future performance work will +focus on reducing the overhead and frequency of these traps. + +The usual code flow is started from userspace invoking an "run" ioctl, which +causes KVM to switch into guest context. We use IVPR to hijack the host +interrupt vectors while running the guest, which allows us to direct all +interrupts to kvmppc_handle_interrupt(). At this point, we could either +- handle the interrupt completely (e.g. emulate "mtspr SPRG0"), or +- let the host interrupt handler run (e.g. when the decrementer fires), or +- return to host userspace (e.g. when the guest performs device MMIO) + +Address spaces: We take advantage of the fact that Linux doesn't use the AS=1 +address space (in host or guest), which gives us virtual address space to use +for guest mappings. While the guest is running, the host kernel remains mapped +in AS=0, but the guest can only use AS=1 mappings. + +TLB entries: The TLB entries covering the host linear mapping remain +present while running the guest. This reduces the overhead of lightweight +exits, which are handled by KVM running in the host kernel. We keep three +copies of the TLB: + - guest TLB: contents of the TLB as the guest sees it + - shadow TLB: the TLB that is actually in hardware while guest is running + - host TLB: to restore TLB state when context switching guest -> host +When a TLB miss occurs because a mapping was not present in the shadow TLB, +but was present in the guest TLB, KVM handles the fault without invoking the +guest. Large guest pages are backed by multiple 4KB shadow pages through this +mechanism. + +IO: MMIO and DCR accesses are emulated by userspace. We use virtio for network +and block IO, so those drivers must be enabled in the guest. It's possible +that some qemu device emulation (e.g. e1000 or rtl8139) may also work with +little effort. diff --git a/Documentation/s390/kvm.txt b/Documentation/s390/kvm.txt new file mode 100644 index 0000000000000000000000000000000000000000..6f5ceb0f09fca03fd19e3fd061e05ca86aba6cd3 --- /dev/null +++ b/Documentation/s390/kvm.txt @@ -0,0 +1,125 @@ +*** BIG FAT WARNING *** +The kvm module is currently in EXPERIMENTAL state for s390. This means that +the interface to the module is not yet considered to remain stable. Thus, be +prepared that we keep breaking your userspace application and guest +compatibility over and over again until we feel happy with the result. Make sure +your guest kernel, your host kernel, and your userspace launcher are in a +consistent state. + +This Documentation describes the unique ioctl calls to /dev/kvm, the resulting +kvm-vm file descriptors, and the kvm-vcpu file descriptors that differ from x86. + +1. ioctl calls to /dev/kvm +KVM does support the following ioctls on s390 that are common with other +architectures and do behave the same: +KVM_GET_API_VERSION +KVM_CREATE_VM (*) see note +KVM_CHECK_EXTENSION +KVM_GET_VCPU_MMAP_SIZE + +Notes: +* KVM_CREATE_VM may fail on s390, if the calling process has multiple +threads and has not called KVM_S390_ENABLE_SIE before. + +In addition, on s390 the following architecture specific ioctls are supported: +ioctl: KVM_S390_ENABLE_SIE +args: none +see also: include/linux/kvm.h +This call causes the kernel to switch on PGSTE in the user page table. This +operation is needed in order to run a virtual machine, and it requires the +calling process to be single-threaded. Note that the first call to KVM_CREATE_VM +will implicitly try to switch on PGSTE if the user process has not called +KVM_S390_ENABLE_SIE before. User processes that want to launch multiple threads +before creating a virtual machine have to call KVM_S390_ENABLE_SIE, or will +observe an error calling KVM_CREATE_VM. Switching on PGSTE is a one-time +operation, is not reversible, and will persist over the entire lifetime of +the calling process. It does not have any user-visible effect other than a small +performance penalty. + +2. ioctl calls to the kvm-vm file descriptor +KVM does support the following ioctls on s390 that are common with other +architectures and do behave the same: +KVM_CREATE_VCPU +KVM_SET_USER_MEMORY_REGION (*) see note +KVM_GET_DIRTY_LOG (**) see note + +Notes: +* kvm does only allow exactly one memory slot on s390, which has to start + at guest absolute address zero and at a user address that is aligned on any + page boundary. This hardware "limitation" allows us to have a few unique + optimizations. The memory slot doesn't have to be filled + with memory actually, it may contain sparse holes. That said, with different + user memory layout this does still allow a large flexibility when + doing the guest memory setup. +** KVM_GET_DIRTY_LOG doesn't work properly yet. The user will receive an empty +log. This ioctl call is only needed for guest migration, and we intend to +implement this one in the future. + +In addition, on s390 the following architecture specific ioctls for the kvm-vm +file descriptor are supported: +ioctl: KVM_S390_INTERRUPT +args: struct kvm_s390_interrupt * +see also: include/linux/kvm.h +This ioctl is used to submit a floating interrupt for a virtual machine. +Floating interrupts may be delivered to any virtual cpu in the configuration. +Only some interrupt types defined in include/linux/kvm.h make sense when +submitted as floating interrupts. The following interrupts are not considered +to be useful as floating interrupts, and a call to inject them will result in +-EINVAL error code: program interrupts and interprocessor signals. Valid +floating interrupts are: +KVM_S390_INT_VIRTIO +KVM_S390_INT_SERVICE + +3. ioctl calls to the kvm-vcpu file descriptor +KVM does support the following ioctls on s390 that are common with other +architectures and do behave the same: +KVM_RUN +KVM_GET_REGS +KVM_SET_REGS +KVM_GET_SREGS +KVM_SET_SREGS +KVM_GET_FPU +KVM_SET_FPU + +In addition, on s390 the following architecture specific ioctls for the +kvm-vcpu file descriptor are supported: +ioctl: KVM_S390_INTERRUPT +args: struct kvm_s390_interrupt * +see also: include/linux/kvm.h +This ioctl is used to submit an interrupt for a specific virtual cpu. +Only some interrupt types defined in include/linux/kvm.h make sense when +submitted for a specific cpu. The following interrupts are not considered +to be useful, and a call to inject them will result in -EINVAL error code: +service processor calls and virtio interrupts. Valid interrupt types are: +KVM_S390_PROGRAM_INT +KVM_S390_SIGP_STOP +KVM_S390_RESTART +KVM_S390_SIGP_SET_PREFIX +KVM_S390_INT_EMERGENCY + +ioctl: KVM_S390_STORE_STATUS +args: unsigned long +see also: include/linux/kvm.h +This ioctl stores the state of the cpu at the guest real address given as +argument, unless one of the following values defined in include/linux/kvm.h +is given as arguement: +KVM_S390_STORE_STATUS_NOADDR - the CPU stores its status to the save area in +absolute lowcore as defined by the principles of operation +KVM_S390_STORE_STATUS_PREFIXED - the CPU stores its status to the save area in +its prefix page just like the dump tool that comes with zipl. This is useful +to create a system dump for use with lkcdutils or crash. + +ioctl: KVM_S390_SET_INITIAL_PSW +args: struct kvm_s390_psw * +see also: include/linux/kvm.h +This ioctl can be used to set the processor status word (psw) of a stopped cpu +prior to running it with KVM_RUN. Note that this call is not required to modify +the psw during sie intercepts that fall back to userspace because struct kvm_run +does contain the psw, and this value is evaluated during reentry of KVM_RUN +after the intercept exit was recognized. + +ioctl: KVM_S390_INITIAL_RESET +args: none +see also: include/linux/kvm.h +This ioctl can be used to perform an initial cpu reset as defined by the +principles of operation. The target cpu has to be in stopped state. diff --git a/Documentation/smart-config.txt b/Documentation/smart-config.txt deleted file mode 100644 index 8467447b5a87adbd0252dfb31f166471cbacf3f0..0000000000000000000000000000000000000000 --- a/Documentation/smart-config.txt +++ /dev/null @@ -1,98 +0,0 @@ -Smart CONFIG_* Dependencies -1 August 1999 - -Michael Chastain -Werner Almesberger -Martin von Loewis - -Here is the problem: - - Suppose that drivers/net/foo.c has the following lines: - - #include - - ... - - #ifdef CONFIG_FOO_AUTOFROB - /* Code for auto-frobbing */ - #else - /* Manual frobbing only */ - #endif - - ... - - #ifdef CONFIG_FOO_MODEL_TWO - /* Code for model two */ - #endif - - Now suppose the user (the person building kernels) reconfigures the - kernel to change some unrelated setting. This will regenerate the - file include/linux/autoconf.h, which will cause include/linux/config.h - to be out of date, which will cause drivers/net/foo.c to be recompiled. - - Most kernel sources, perhaps 80% of them, have at least one CONFIG_* - dependency somewhere. So changing _any_ CONFIG_* setting requires - almost _all_ of the kernel to be recompiled. - -Here is the solution: - - We've made the dependency generator, mkdep.c, smarter. Instead of - generating this dependency: - - drivers/net/foo.c: include/linux/config.h - - It now generates these dependencies: - - drivers/net/foo.c: \ - include/config/foo/autofrob.h \ - include/config/foo/model/two.h - - So drivers/net/foo.c depends only on the CONFIG_* lines that - it actually uses. - - A new program, split-include.c, runs at the beginning of - compilation (make bzImage or make zImage). split-include reads - include/linux/autoconf.h and updates the include/config/ tree, - writing one file per option. It updates only the files for options - that have changed. - -Flag Dependencies - - Martin Von Loewis contributed another feature to this patch: - 'flag dependencies'. The idea is that a .o file depends on - the compilation flags used to build it. The file foo.o has - its flags stored in .flags.foo.o. - - Suppose the user changes the foo driver from resident to modular. - 'make' will notice that the current foo.o was not compiled with - -DMODULE and will recompile foo.c. - - All .o files made from C source have flag dependencies. So do .o - files made with ld, and .a files made with ar. However, .o files - made from assembly source do not have flag dependencies (nobody - needs this yet, but it would be good to fix). - -Per-source-file Flags - - Flag dependencies also work with per-source-file flags. - You can specify compilation flags for individual source files - like this: - - CFLAGS_foo.o = -DSPECIAL_FOO_DEFINE - - This helps clean up drivers/net/Makefile, drivers/scsi/Makefile, - and several other Makefiles. - -Credit - - Werner Almesberger had the original idea and wrote the first - version of this patch. - - Michael Chastain picked it up and continued development. He is - now the principal author and maintainer. Please report any bugs - to him. - - Martin von Loewis wrote flag dependencies, with some modifications - by Michael Chastain. - - Thanks to all of the beta testers. diff --git a/MAINTAINERS b/MAINTAINERS index a942f385249901d21e679a5d77478bd3e4f49174..c1dd1ae7b1335d9de927e6a7731d72066424660d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2329,6 +2329,13 @@ L: kvm-devel@lists.sourceforge.net W: kvm.sourceforge.net S: Supported +KERNEL VIRTUAL MACHINE (KVM) FOR POWERPC +P: Hollis Blanchard +M: hollisb@us.ibm.com +L: kvm-ppc-devel@lists.sourceforge.net +W: kvm.sourceforge.net +S: Supported + KERNEL VIRTUAL MACHINE For Itanium(KVM/IA64) P: Anthony Xu M: anthony.xu@intel.com @@ -2338,6 +2345,16 @@ L: kvm-ia64-devel@lists.sourceforge.net W: kvm.sourceforge.net S: Supported +KERNEL VIRTUAL MACHINE for s390 (KVM/s390) +P: Carsten Otte +M: cotte@de.ibm.com +P: Christian Borntraeger +M: borntraeger@de.ibm.com +M: linux390@de.ibm.com +L: linux-s390@vger.kernel.org +W: http://www.ibm.com/developerworks/linux/linux390/ +S: Supported + KEXEC P: Eric Biederman M: ebiederm@xmission.com diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index cd13e138bd03800edd9ab5cc935d8e8c5418bb0e..3aa6c821449a90fd43d0169fb3bbcf410140602d 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -19,6 +19,7 @@ config IA64 select HAVE_OPROFILE select HAVE_KPROBES select HAVE_KRETPROBES + select HAVE_KVM default y help The Itanium Processor Family is Intel's 64-bit successor to @@ -589,6 +590,8 @@ config MSPEC source "fs/Kconfig" +source "arch/ia64/kvm/Kconfig" + source "lib/Kconfig" # diff --git a/arch/ia64/Makefile b/arch/ia64/Makefile index f1645c4f70393c5294ecc9148300c8368178d5f5..ec4cca477f491c31cf74b79f184e1fce812cf4dc 100644 --- a/arch/ia64/Makefile +++ b/arch/ia64/Makefile @@ -57,6 +57,7 @@ core-$(CONFIG_IA64_GENERIC) += arch/ia64/dig/ core-$(CONFIG_IA64_HP_ZX1) += arch/ia64/dig/ core-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/dig/ core-$(CONFIG_IA64_SGI_SN2) += arch/ia64/sn/ +core-$(CONFIG_KVM) += arch/ia64/kvm/ drivers-$(CONFIG_PCI) += arch/ia64/pci/ drivers-$(CONFIG_IA64_HP_SIM) += arch/ia64/hp/sim/ diff --git a/arch/ia64/kvm/Kconfig b/arch/ia64/kvm/Kconfig new file mode 100644 index 0000000000000000000000000000000000000000..7914e4828504595c161aee52025ed9e5f0cdaba0 --- /dev/null +++ b/arch/ia64/kvm/Kconfig @@ -0,0 +1,49 @@ +# +# KVM configuration +# +config HAVE_KVM + bool + +menuconfig VIRTUALIZATION + bool "Virtualization" + depends on HAVE_KVM || IA64 + default y + ---help--- + Say Y here to get to see options for using your Linux host to run other + operating systems inside virtual machines (guests). + This option alone does not add any kernel code. + + If you say N, all options in this submenu will be skipped and disabled. + +if VIRTUALIZATION + +config KVM + tristate "Kernel-based Virtual Machine (KVM) support" + depends on HAVE_KVM && EXPERIMENTAL + select PREEMPT_NOTIFIERS + select ANON_INODES + ---help--- + Support hosting fully virtualized guest machines using hardware + virtualization extensions. You will need a fairly recent + processor equipped with virtualization extensions. You will also + need to select one or more of the processor modules below. + + This module provides access to the hardware capabilities through + a character device node named /dev/kvm. + + To compile this as a module, choose M here: the module + will be called kvm. + + If unsure, say N. + +config KVM_INTEL + tristate "KVM for Intel Itanium 2 processors support" + depends on KVM && m + ---help--- + Provides support for KVM on Itanium 2 processors equipped with the VT + extensions. + +config KVM_TRACE + bool + +endif # VIRTUALIZATION diff --git a/arch/ia64/kvm/Makefile b/arch/ia64/kvm/Makefile new file mode 100644 index 0000000000000000000000000000000000000000..41b034ffa73bf9311d4815548008582ac2e69ef4 --- /dev/null +++ b/arch/ia64/kvm/Makefile @@ -0,0 +1,61 @@ +#This Make file is to generate asm-offsets.h and build source. +# + +#Generate asm-offsets.h for vmm module build +offsets-file := asm-offsets.h + +always := $(offsets-file) +targets := $(offsets-file) +targets += arch/ia64/kvm/asm-offsets.s +clean-files := $(addprefix $(objtree)/,$(targets) $(obj)/memcpy.S $(obj)/memset.S) + +# Default sed regexp - multiline due to syntax constraints +define sed-y + "/^->/{s:^->\([^ ]*\) [\$$#]*\([^ ]*\) \(.*\):#define \1 \2 /* \3 */:; s:->::; p;}" +endef + +quiet_cmd_offsets = GEN $@ +define cmd_offsets + (set -e; \ + echo "#ifndef __ASM_KVM_OFFSETS_H__"; \ + echo "#define __ASM_KVM_OFFSETS_H__"; \ + echo "/*"; \ + echo " * DO NOT MODIFY."; \ + echo " *"; \ + echo " * This file was generated by Makefile"; \ + echo " *"; \ + echo " */"; \ + echo ""; \ + sed -ne $(sed-y) $<; \ + echo ""; \ + echo "#endif" ) > $@ +endef +# We use internal rules to avoid the "is up to date" message from make +arch/ia64/kvm/asm-offsets.s: arch/ia64/kvm/asm-offsets.c + $(call if_changed_dep,cc_s_c) + +$(obj)/$(offsets-file): arch/ia64/kvm/asm-offsets.s + $(call cmd,offsets) + +# +# Makefile for Kernel-based Virtual Machine module +# + +EXTRA_CFLAGS += -Ivirt/kvm -Iarch/ia64/kvm/ + +$(addprefix $(objtree)/,$(obj)/memcpy.S $(obj)/memset.S): + $(shell ln -snf ../lib/memcpy.S $(src)/memcpy.S) + $(shell ln -snf ../lib/memset.S $(src)/memset.S) + +common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o ioapic.o) + +kvm-objs := $(common-objs) kvm-ia64.o kvm_fw.o +obj-$(CONFIG_KVM) += kvm.o + +FORCE : $(obj)/$(offsets-file) +EXTRA_CFLAGS_vcpu.o += -mfixed-range=f2-f5,f12-f127 +kvm-intel-objs = vmm.o vmm_ivt.o trampoline.o vcpu.o optvfault.o mmio.o \ + vtlb.o process.o +#Add link memcpy and memset to avoid possible structure assignment error +kvm-intel-objs += memset.o memcpy.o +obj-$(CONFIG_KVM_INTEL) += kvm-intel.o diff --git a/arch/ia64/kvm/asm-offsets.c b/arch/ia64/kvm/asm-offsets.c new file mode 100644 index 0000000000000000000000000000000000000000..4e3dc13a619c6782f94337a4a2873a4a208058c1 --- /dev/null +++ b/arch/ia64/kvm/asm-offsets.c @@ -0,0 +1,251 @@ +/* + * asm-offsets.c Generate definitions needed by assembly language modules. + * This code generates raw asm output which is post-processed + * to extract and format the required data. + * + * Anthony Xu + * Xiantao Zhang + * Copyright (c) 2007 Intel Corporation KVM support. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +#include +#include + +#include "vcpu.h" + +#define task_struct kvm_vcpu + +#define DEFINE(sym, val) \ + asm volatile("\n->" #sym " (%0) " #val : : "i" (val)) + +#define BLANK() asm volatile("\n->" : :) + +#define OFFSET(_sym, _str, _mem) \ + DEFINE(_sym, offsetof(_str, _mem)); + +void foo(void) +{ + DEFINE(VMM_TASK_SIZE, sizeof(struct kvm_vcpu)); + DEFINE(VMM_PT_REGS_SIZE, sizeof(struct kvm_pt_regs)); + + BLANK(); + + DEFINE(VMM_VCPU_META_RR0_OFFSET, + offsetof(struct kvm_vcpu, arch.metaphysical_rr0)); + DEFINE(VMM_VCPU_META_SAVED_RR0_OFFSET, + offsetof(struct kvm_vcpu, + arch.metaphysical_saved_rr0)); + DEFINE(VMM_VCPU_VRR0_OFFSET, + offsetof(struct kvm_vcpu, arch.vrr[0])); + DEFINE(VMM_VPD_IRR0_OFFSET, + offsetof(struct vpd, irr[0])); + DEFINE(VMM_VCPU_ITC_CHECK_OFFSET, + offsetof(struct kvm_vcpu, arch.itc_check)); + DEFINE(VMM_VCPU_IRQ_CHECK_OFFSET, + offsetof(struct kvm_vcpu, arch.irq_check)); + DEFINE(VMM_VPD_VHPI_OFFSET, + offsetof(struct vpd, vhpi)); + DEFINE(VMM_VCPU_VSA_BASE_OFFSET, + offsetof(struct kvm_vcpu, arch.vsa_base)); + DEFINE(VMM_VCPU_VPD_OFFSET, + offsetof(struct kvm_vcpu, arch.vpd)); + DEFINE(VMM_VCPU_IRQ_CHECK, + offsetof(struct kvm_vcpu, arch.irq_check)); + DEFINE(VMM_VCPU_TIMER_PENDING, + offsetof(struct kvm_vcpu, arch.timer_pending)); + DEFINE(VMM_VCPU_META_SAVED_RR0_OFFSET, + offsetof(struct kvm_vcpu, arch.metaphysical_saved_rr0)); + DEFINE(VMM_VCPU_MODE_FLAGS_OFFSET, + offsetof(struct kvm_vcpu, arch.mode_flags)); + DEFINE(VMM_VCPU_ITC_OFS_OFFSET, + offsetof(struct kvm_vcpu, arch.itc_offset)); + DEFINE(VMM_VCPU_LAST_ITC_OFFSET, + offsetof(struct kvm_vcpu, arch.last_itc)); + DEFINE(VMM_VCPU_SAVED_GP_OFFSET, + offsetof(struct kvm_vcpu, arch.saved_gp)); + + BLANK(); + + DEFINE(VMM_PT_REGS_B6_OFFSET, + offsetof(struct kvm_pt_regs, b6)); + DEFINE(VMM_PT_REGS_B7_OFFSET, + offsetof(struct kvm_pt_regs, b7)); + DEFINE(VMM_PT_REGS_AR_CSD_OFFSET, + offsetof(struct kvm_pt_regs, ar_csd)); + DEFINE(VMM_PT_REGS_AR_SSD_OFFSET, + offsetof(struct kvm_pt_regs, ar_ssd)); + DEFINE(VMM_PT_REGS_R8_OFFSET, + offsetof(struct kvm_pt_regs, r8)); + DEFINE(VMM_PT_REGS_R9_OFFSET, + offsetof(struct kvm_pt_regs, r9)); + DEFINE(VMM_PT_REGS_R10_OFFSET, + offsetof(struct kvm_pt_regs, r10)); + DEFINE(VMM_PT_REGS_R11_OFFSET, + offsetof(struct kvm_pt_regs, r11)); + DEFINE(VMM_PT_REGS_CR_IPSR_OFFSET, + offsetof(struct kvm_pt_regs, cr_ipsr)); + DEFINE(VMM_PT_REGS_CR_IIP_OFFSET, + offsetof(struct kvm_pt_regs, cr_iip)); + DEFINE(VMM_PT_REGS_CR_IFS_OFFSET, + offsetof(struct kvm_pt_regs, cr_ifs)); + DEFINE(VMM_PT_REGS_AR_UNAT_OFFSET, + offsetof(struct kvm_pt_regs, ar_unat)); + DEFINE(VMM_PT_REGS_AR_PFS_OFFSET, + offsetof(struct kvm_pt_regs, ar_pfs)); + DEFINE(VMM_PT_REGS_AR_RSC_OFFSET, + offsetof(struct kvm_pt_regs, ar_rsc)); + DEFINE(VMM_PT_REGS_AR_RNAT_OFFSET, + offsetof(struct kvm_pt_regs, ar_rnat)); + + DEFINE(VMM_PT_REGS_AR_BSPSTORE_OFFSET, + offsetof(struct kvm_pt_regs, ar_bspstore)); + DEFINE(VMM_PT_REGS_PR_OFFSET, + offsetof(struct kvm_pt_regs, pr)); + DEFINE(VMM_PT_REGS_B0_OFFSET, + offsetof(struct kvm_pt_regs, b0)); + DEFINE(VMM_PT_REGS_LOADRS_OFFSET, + offsetof(struct kvm_pt_regs, loadrs)); + DEFINE(VMM_PT_REGS_R1_OFFSET, + offsetof(struct kvm_pt_regs, r1)); + DEFINE(VMM_PT_REGS_R12_OFFSET, + offsetof(struct kvm_pt_regs, r12)); + DEFINE(VMM_PT_REGS_R13_OFFSET, + offsetof(struct kvm_pt_regs, r13)); + DEFINE(VMM_PT_REGS_AR_FPSR_OFFSET, + offsetof(struct kvm_pt_regs, ar_fpsr)); + DEFINE(VMM_PT_REGS_R15_OFFSET, + offsetof(struct kvm_pt_regs, r15)); + DEFINE(VMM_PT_REGS_R14_OFFSET, + offsetof(struct kvm_pt_regs, r14)); + DEFINE(VMM_PT_REGS_R2_OFFSET, + offsetof(struct kvm_pt_regs, r2)); + DEFINE(VMM_PT_REGS_R3_OFFSET, + offsetof(struct kvm_pt_regs, r3)); + DEFINE(VMM_PT_REGS_R16_OFFSET, + offsetof(struct kvm_pt_regs, r16)); + DEFINE(VMM_PT_REGS_R17_OFFSET, + offsetof(struct kvm_pt_regs, r17)); + DEFINE(VMM_PT_REGS_R18_OFFSET, + offsetof(struct kvm_pt_regs, r18)); + DEFINE(VMM_PT_REGS_R19_OFFSET, + offsetof(struct kvm_pt_regs, r19)); + DEFINE(VMM_PT_REGS_R20_OFFSET, + offsetof(struct kvm_pt_regs, r20)); + DEFINE(VMM_PT_REGS_R21_OFFSET, + offsetof(struct kvm_pt_regs, r21)); + DEFINE(VMM_PT_REGS_R22_OFFSET, + offsetof(struct kvm_pt_regs, r22)); + DEFINE(VMM_PT_REGS_R23_OFFSET, + offsetof(struct kvm_pt_regs, r23)); + DEFINE(VMM_PT_REGS_R24_OFFSET, + offsetof(struct kvm_pt_regs, r24)); + DEFINE(VMM_PT_REGS_R25_OFFSET, + offsetof(struct kvm_pt_regs, r25)); + DEFINE(VMM_PT_REGS_R26_OFFSET, + offsetof(struct kvm_pt_regs, r26)); + DEFINE(VMM_PT_REGS_R27_OFFSET, + offsetof(struct kvm_pt_regs, r27)); + DEFINE(VMM_PT_REGS_R28_OFFSET, + offsetof(struct kvm_pt_regs, r28)); + DEFINE(VMM_PT_REGS_R29_OFFSET, + offsetof(struct kvm_pt_regs, r29)); + DEFINE(VMM_PT_REGS_R30_OFFSET, + offsetof(struct kvm_pt_regs, r30)); + DEFINE(VMM_PT_REGS_R31_OFFSET, + offsetof(struct kvm_pt_regs, r31)); + DEFINE(VMM_PT_REGS_AR_CCV_OFFSET, + offsetof(struct kvm_pt_regs, ar_ccv)); + DEFINE(VMM_PT_REGS_F6_OFFSET, + offsetof(struct kvm_pt_regs, f6)); + DEFINE(VMM_PT_REGS_F7_OFFSET, + offsetof(struct kvm_pt_regs, f7)); + DEFINE(VMM_PT_REGS_F8_OFFSET, + offsetof(struct kvm_pt_regs, f8)); + DEFINE(VMM_PT_REGS_F9_OFFSET, + offsetof(struct kvm_pt_regs, f9)); + DEFINE(VMM_PT_REGS_F10_OFFSET, + offsetof(struct kvm_pt_regs, f10)); + DEFINE(VMM_PT_REGS_F11_OFFSET, + offsetof(struct kvm_pt_regs, f11)); + DEFINE(VMM_PT_REGS_R4_OFFSET, + offsetof(struct kvm_pt_regs, r4)); + DEFINE(VMM_PT_REGS_R5_OFFSET, + offsetof(struct kvm_pt_regs, r5)); + DEFINE(VMM_PT_REGS_R6_OFFSET, + offsetof(struct kvm_pt_regs, r6)); + DEFINE(VMM_PT_REGS_R7_OFFSET, + offsetof(struct kvm_pt_regs, r7)); + DEFINE(VMM_PT_REGS_EML_UNAT_OFFSET, + offsetof(struct kvm_pt_regs, eml_unat)); + DEFINE(VMM_VCPU_IIPA_OFFSET, + offsetof(struct kvm_vcpu, arch.cr_iipa)); + DEFINE(VMM_VCPU_OPCODE_OFFSET, + offsetof(struct kvm_vcpu, arch.opcode)); + DEFINE(VMM_VCPU_CAUSE_OFFSET, offsetof(struct kvm_vcpu, arch.cause)); + DEFINE(VMM_VCPU_ISR_OFFSET, + offsetof(struct kvm_vcpu, arch.cr_isr)); + DEFINE(VMM_PT_REGS_R16_SLOT, + (((offsetof(struct kvm_pt_regs, r16) + - sizeof(struct kvm_pt_regs)) >> 3) & 0x3f)); + DEFINE(VMM_VCPU_MODE_FLAGS_OFFSET, + offsetof(struct kvm_vcpu, arch.mode_flags)); + DEFINE(VMM_VCPU_GP_OFFSET, offsetof(struct kvm_vcpu, arch.__gp)); + BLANK(); + + DEFINE(VMM_VPD_BASE_OFFSET, offsetof(struct kvm_vcpu, arch.vpd)); + DEFINE(VMM_VPD_VIFS_OFFSET, offsetof(struct vpd, ifs)); + DEFINE(VMM_VLSAPIC_INSVC_BASE_OFFSET, + offsetof(struct kvm_vcpu, arch.insvc[0])); + DEFINE(VMM_VPD_VPTA_OFFSET, offsetof(struct vpd, pta)); + DEFINE(VMM_VPD_VPSR_OFFSET, offsetof(struct vpd, vpsr)); + + DEFINE(VMM_CTX_R4_OFFSET, offsetof(union context, gr[4])); + DEFINE(VMM_CTX_R5_OFFSET, offsetof(union context, gr[5])); + DEFINE(VMM_CTX_R12_OFFSET, offsetof(union context, gr[12])); + DEFINE(VMM_CTX_R13_OFFSET, offsetof(union context, gr[13])); + DEFINE(VMM_CTX_KR0_OFFSET, offsetof(union context, ar[0])); + DEFINE(VMM_CTX_KR1_OFFSET, offsetof(union context, ar[1])); + DEFINE(VMM_CTX_B0_OFFSET, offsetof(union context, br[0])); + DEFINE(VMM_CTX_B1_OFFSET, offsetof(union context, br[1])); + DEFINE(VMM_CTX_B2_OFFSET, offsetof(union context, br[2])); + DEFINE(VMM_CTX_RR0_OFFSET, offsetof(union context, rr[0])); + DEFINE(VMM_CTX_RSC_OFFSET, offsetof(union context, ar[16])); + DEFINE(VMM_CTX_BSPSTORE_OFFSET, offsetof(union context, ar[18])); + DEFINE(VMM_CTX_RNAT_OFFSET, offsetof(union context, ar[19])); + DEFINE(VMM_CTX_FCR_OFFSET, offsetof(union context, ar[21])); + DEFINE(VMM_CTX_EFLAG_OFFSET, offsetof(union context, ar[24])); + DEFINE(VMM_CTX_CFLG_OFFSET, offsetof(union context, ar[27])); + DEFINE(VMM_CTX_FSR_OFFSET, offsetof(union context, ar[28])); + DEFINE(VMM_CTX_FIR_OFFSET, offsetof(union context, ar[29])); + DEFINE(VMM_CTX_FDR_OFFSET, offsetof(union context, ar[30])); + DEFINE(VMM_CTX_UNAT_OFFSET, offsetof(union context, ar[36])); + DEFINE(VMM_CTX_FPSR_OFFSET, offsetof(union context, ar[40])); + DEFINE(VMM_CTX_PFS_OFFSET, offsetof(union context, ar[64])); + DEFINE(VMM_CTX_LC_OFFSET, offsetof(union context, ar[65])); + DEFINE(VMM_CTX_DCR_OFFSET, offsetof(union context, cr[0])); + DEFINE(VMM_CTX_IVA_OFFSET, offsetof(union context, cr[2])); + DEFINE(VMM_CTX_PTA_OFFSET, offsetof(union context, cr[8])); + DEFINE(VMM_CTX_IBR0_OFFSET, offsetof(union context, ibr[0])); + DEFINE(VMM_CTX_DBR0_OFFSET, offsetof(union context, dbr[0])); + DEFINE(VMM_CTX_F2_OFFSET, offsetof(union context, fr[2])); + DEFINE(VMM_CTX_F3_OFFSET, offsetof(union context, fr[3])); + DEFINE(VMM_CTX_F32_OFFSET, offsetof(union context, fr[32])); + DEFINE(VMM_CTX_F33_OFFSET, offsetof(union context, fr[33])); + DEFINE(VMM_CTX_PKR0_OFFSET, offsetof(union context, pkr[0])); + DEFINE(VMM_CTX_PSR_OFFSET, offsetof(union context, psr)); + BLANK(); +} diff --git a/arch/ia64/kvm/kvm-ia64.c b/arch/ia64/kvm/kvm-ia64.c new file mode 100644 index 0000000000000000000000000000000000000000..6df0732401353f56524d1ee1d60ef2a9358bae8d --- /dev/null +++ b/arch/ia64/kvm/kvm-ia64.c @@ -0,0 +1,1806 @@ + +/* + * kvm_ia64.c: Basic KVM suppport On Itanium series processors + * + * + * Copyright (C) 2007, Intel Corporation. + * Xiantao Zhang (xiantao.zhang@intel.com) + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include "misc.h" +#include "vti.h" +#include "iodev.h" +#include "ioapic.h" +#include "lapic.h" + +static unsigned long kvm_vmm_base; +static unsigned long kvm_vsa_base; +static unsigned long kvm_vm_buffer; +static unsigned long kvm_vm_buffer_size; +unsigned long kvm_vmm_gp; + +static long vp_env_info; + +static struct kvm_vmm_info *kvm_vmm_info; + +static DEFINE_PER_CPU(struct kvm_vcpu *, last_vcpu); + +struct kvm_stats_debugfs_item debugfs_entries[] = { + { NULL } +}; + + +struct fdesc{ + unsigned long ip; + unsigned long gp; +}; + +static void kvm_flush_icache(unsigned long start, unsigned long len) +{ + int l; + + for (l = 0; l < (len + 32); l += 32) + ia64_fc(start + l); + + ia64_sync_i(); + ia64_srlz_i(); +} + +static void kvm_flush_tlb_all(void) +{ + unsigned long i, j, count0, count1, stride0, stride1, addr; + long flags; + + addr = local_cpu_data->ptce_base; + count0 = local_cpu_data->ptce_count[0]; + count1 = local_cpu_data->ptce_count[1]; + stride0 = local_cpu_data->ptce_stride[0]; + stride1 = local_cpu_data->ptce_stride[1]; + + local_irq_save(flags); + for (i = 0; i < count0; ++i) { + for (j = 0; j < count1; ++j) { + ia64_ptce(addr); + addr += stride1; + } + addr += stride0; + } + local_irq_restore(flags); + ia64_srlz_i(); /* srlz.i implies srlz.d */ +} + +long ia64_pal_vp_create(u64 *vpd, u64 *host_iva, u64 *opt_handler) +{ + struct ia64_pal_retval iprv; + + PAL_CALL_STK(iprv, PAL_VP_CREATE, (u64)vpd, (u64)host_iva, + (u64)opt_handler); + + return iprv.status; +} + +static DEFINE_SPINLOCK(vp_lock); + +void kvm_arch_hardware_enable(void *garbage) +{ + long status; + long tmp_base; + unsigned long pte; + unsigned long saved_psr; + int slot; + + pte = pte_val(mk_pte_phys(__pa(kvm_vmm_base), + PAGE_KERNEL)); + local_irq_save(saved_psr); + slot = ia64_itr_entry(0x3, KVM_VMM_BASE, pte, KVM_VMM_SHIFT); + if (slot < 0) + return; + local_irq_restore(saved_psr); + + spin_lock(&vp_lock); + status = ia64_pal_vp_init_env(kvm_vsa_base ? + VP_INIT_ENV : VP_INIT_ENV_INITALIZE, + __pa(kvm_vm_buffer), KVM_VM_BUFFER_BASE, &tmp_base); + if (status != 0) { + printk(KERN_WARNING"kvm: Failed to Enable VT Support!!!!\n"); + return ; + } + + if (!kvm_vsa_base) { + kvm_vsa_base = tmp_base; + printk(KERN_INFO"kvm: kvm_vsa_base:0x%lx\n", kvm_vsa_base); + } + spin_unlock(&vp_lock); + ia64_ptr_entry(0x3, slot); +} + +void kvm_arch_hardware_disable(void *garbage) +{ + + long status; + int slot; + unsigned long pte; + unsigned long saved_psr; + unsigned long host_iva = ia64_getreg(_IA64_REG_CR_IVA); + + pte = pte_val(mk_pte_phys(__pa(kvm_vmm_base), + PAGE_KERNEL)); + + local_irq_save(saved_psr); + slot = ia64_itr_entry(0x3, KVM_VMM_BASE, pte, KVM_VMM_SHIFT); + if (slot < 0) + return; + local_irq_restore(saved_psr); + + status = ia64_pal_vp_exit_env(host_iva); + if (status) + printk(KERN_DEBUG"kvm: Failed to disable VT support! :%ld\n", + status); + ia64_ptr_entry(0x3, slot); +} + +void kvm_arch_check_processor_compat(void *rtn) +{ + *(int *)rtn = 0; +} + +int kvm_dev_ioctl_check_extension(long ext) +{ + + int r; + + switch (ext) { + case KVM_CAP_IRQCHIP: + case KVM_CAP_USER_MEMORY: + + r = 1; + break; + default: + r = 0; + } + return r; + +} + +static struct kvm_io_device *vcpu_find_mmio_dev(struct kvm_vcpu *vcpu, + gpa_t addr) +{ + struct kvm_io_device *dev; + + dev = kvm_io_bus_find_dev(&vcpu->kvm->mmio_bus, addr); + + return dev; +} + +static int handle_vm_error(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + kvm_run->exit_reason = KVM_EXIT_UNKNOWN; + kvm_run->hw.hardware_exit_reason = 1; + return 0; +} + +static int handle_mmio(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + struct kvm_mmio_req *p; + struct kvm_io_device *mmio_dev; + + p = kvm_get_vcpu_ioreq(vcpu); + + if ((p->addr & PAGE_MASK) == IOAPIC_DEFAULT_BASE_ADDRESS) + goto mmio; + vcpu->mmio_needed = 1; + vcpu->mmio_phys_addr = kvm_run->mmio.phys_addr = p->addr; + vcpu->mmio_size = kvm_run->mmio.len = p->size; + vcpu->mmio_is_write = kvm_run->mmio.is_write = !p->dir; + + if (vcpu->mmio_is_write) + memcpy(vcpu->mmio_data, &p->data, p->size); + memcpy(kvm_run->mmio.data, &p->data, p->size); + kvm_run->exit_reason = KVM_EXIT_MMIO; + return 0; +mmio: + mmio_dev = vcpu_find_mmio_dev(vcpu, p->addr); + if (mmio_dev) { + if (!p->dir) + kvm_iodevice_write(mmio_dev, p->addr, p->size, + &p->data); + else + kvm_iodevice_read(mmio_dev, p->addr, p->size, + &p->data); + + } else + printk(KERN_ERR"kvm: No iodevice found! addr:%lx\n", p->addr); + p->state = STATE_IORESP_READY; + + return 1; +} + +static int handle_pal_call(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + struct exit_ctl_data *p; + + p = kvm_get_exit_data(vcpu); + + if (p->exit_reason == EXIT_REASON_PAL_CALL) + return kvm_pal_emul(vcpu, kvm_run); + else { + kvm_run->exit_reason = KVM_EXIT_UNKNOWN; + kvm_run->hw.hardware_exit_reason = 2; + return 0; + } +} + +static int handle_sal_call(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + struct exit_ctl_data *p; + + p = kvm_get_exit_data(vcpu); + + if (p->exit_reason == EXIT_REASON_SAL_CALL) { + kvm_sal_emul(vcpu); + return 1; + } else { + kvm_run->exit_reason = KVM_EXIT_UNKNOWN; + kvm_run->hw.hardware_exit_reason = 3; + return 0; + } + +} + +/* + * offset: address offset to IPI space. + * value: deliver value. + */ +static void vcpu_deliver_ipi(struct kvm_vcpu *vcpu, uint64_t dm, + uint64_t vector) +{ + switch (dm) { + case SAPIC_FIXED: + kvm_apic_set_irq(vcpu, vector, 0); + break; + case SAPIC_NMI: + kvm_apic_set_irq(vcpu, 2, 0); + break; + case SAPIC_EXTINT: + kvm_apic_set_irq(vcpu, 0, 0); + break; + case SAPIC_INIT: + case SAPIC_PMI: + default: + printk(KERN_ERR"kvm: Unimplemented Deliver reserved IPI!\n"); + break; + } +} + +static struct kvm_vcpu *lid_to_vcpu(struct kvm *kvm, unsigned long id, + unsigned long eid) +{ + union ia64_lid lid; + int i; + + for (i = 0; i < KVM_MAX_VCPUS; i++) { + if (kvm->vcpus[i]) { + lid.val = VCPU_LID(kvm->vcpus[i]); + if (lid.id == id && lid.eid == eid) + return kvm->vcpus[i]; + } + } + + return NULL; +} + +static int handle_ipi(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + struct exit_ctl_data *p = kvm_get_exit_data(vcpu); + struct kvm_vcpu *target_vcpu; + struct kvm_pt_regs *regs; + union ia64_ipi_a addr = p->u.ipi_data.addr; + union ia64_ipi_d data = p->u.ipi_data.data; + + target_vcpu = lid_to_vcpu(vcpu->kvm, addr.id, addr.eid); + if (!target_vcpu) + return handle_vm_error(vcpu, kvm_run); + + if (!target_vcpu->arch.launched) { + regs = vcpu_regs(target_vcpu); + + regs->cr_iip = vcpu->kvm->arch.rdv_sal_data.boot_ip; + regs->r1 = vcpu->kvm->arch.rdv_sal_data.boot_gp; + + target_vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE; + if (waitqueue_active(&target_vcpu->wq)) + wake_up_interruptible(&target_vcpu->wq); + } else { + vcpu_deliver_ipi(target_vcpu, data.dm, data.vector); + if (target_vcpu != vcpu) + kvm_vcpu_kick(target_vcpu); + } + + return 1; +} + +struct call_data { + struct kvm_ptc_g ptc_g_data; + struct kvm_vcpu *vcpu; +}; + +static void vcpu_global_purge(void *info) +{ + struct call_data *p = (struct call_data *)info; + struct kvm_vcpu *vcpu = p->vcpu; + + if (test_bit(KVM_REQ_TLB_FLUSH, &vcpu->requests)) + return; + + set_bit(KVM_REQ_PTC_G, &vcpu->requests); + if (vcpu->arch.ptc_g_count < MAX_PTC_G_NUM) { + vcpu->arch.ptc_g_data[vcpu->arch.ptc_g_count++] = + p->ptc_g_data; + } else { + clear_bit(KVM_REQ_PTC_G, &vcpu->requests); + vcpu->arch.ptc_g_count = 0; + set_bit(KVM_REQ_TLB_FLUSH, &vcpu->requests); + } +} + +static int handle_global_purge(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + struct exit_ctl_data *p = kvm_get_exit_data(vcpu); + struct kvm *kvm = vcpu->kvm; + struct call_data call_data; + int i; + call_data.ptc_g_data = p->u.ptc_g_data; + + for (i = 0; i < KVM_MAX_VCPUS; i++) { + if (!kvm->vcpus[i] || kvm->vcpus[i]->arch.mp_state == + KVM_MP_STATE_UNINITIALIZED || + vcpu == kvm->vcpus[i]) + continue; + + if (waitqueue_active(&kvm->vcpus[i]->wq)) + wake_up_interruptible(&kvm->vcpus[i]->wq); + + if (kvm->vcpus[i]->cpu != -1) { + call_data.vcpu = kvm->vcpus[i]; + smp_call_function_single(kvm->vcpus[i]->cpu, + vcpu_global_purge, &call_data, 0, 1); + } else + printk(KERN_WARNING"kvm: Uninit vcpu received ipi!\n"); + + } + return 1; +} + +static int handle_switch_rr6(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + return 1; +} + +int kvm_emulate_halt(struct kvm_vcpu *vcpu) +{ + + ktime_t kt; + long itc_diff; + unsigned long vcpu_now_itc; + + unsigned long expires; + struct hrtimer *p_ht = &vcpu->arch.hlt_timer; + unsigned long cyc_per_usec = local_cpu_data->cyc_per_usec; + struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd); + + vcpu_now_itc = ia64_getreg(_IA64_REG_AR_ITC) + vcpu->arch.itc_offset; + + if (time_after(vcpu_now_itc, vpd->itm)) { + vcpu->arch.timer_check = 1; + return 1; + } + itc_diff = vpd->itm - vcpu_now_itc; + if (itc_diff < 0) + itc_diff = -itc_diff; + + expires = div64_64(itc_diff, cyc_per_usec); + kt = ktime_set(0, 1000 * expires); + vcpu->arch.ht_active = 1; + hrtimer_start(p_ht, kt, HRTIMER_MODE_ABS); + + if (irqchip_in_kernel(vcpu->kvm)) { + vcpu->arch.mp_state = KVM_MP_STATE_HALTED; + kvm_vcpu_block(vcpu); + hrtimer_cancel(p_ht); + vcpu->arch.ht_active = 0; + + if (vcpu->arch.mp_state != KVM_MP_STATE_RUNNABLE) + return -EINTR; + return 1; + } else { + printk(KERN_ERR"kvm: Unsupported userspace halt!"); + return 0; + } +} + +static int handle_vm_shutdown(struct kvm_vcpu *vcpu, + struct kvm_run *kvm_run) +{ + kvm_run->exit_reason = KVM_EXIT_SHUTDOWN; + return 0; +} + +static int handle_external_interrupt(struct kvm_vcpu *vcpu, + struct kvm_run *kvm_run) +{ + return 1; +} + +static int (*kvm_vti_exit_handlers[])(struct kvm_vcpu *vcpu, + struct kvm_run *kvm_run) = { + [EXIT_REASON_VM_PANIC] = handle_vm_error, + [EXIT_REASON_MMIO_INSTRUCTION] = handle_mmio, + [EXIT_REASON_PAL_CALL] = handle_pal_call, + [EXIT_REASON_SAL_CALL] = handle_sal_call, + [EXIT_REASON_SWITCH_RR6] = handle_switch_rr6, + [EXIT_REASON_VM_DESTROY] = handle_vm_shutdown, + [EXIT_REASON_EXTERNAL_INTERRUPT] = handle_external_interrupt, + [EXIT_REASON_IPI] = handle_ipi, + [EXIT_REASON_PTC_G] = handle_global_purge, + +}; + +static const int kvm_vti_max_exit_handlers = + sizeof(kvm_vti_exit_handlers)/sizeof(*kvm_vti_exit_handlers); + +static void kvm_prepare_guest_switch(struct kvm_vcpu *vcpu) +{ +} + +static uint32_t kvm_get_exit_reason(struct kvm_vcpu *vcpu) +{ + struct exit_ctl_data *p_exit_data; + + p_exit_data = kvm_get_exit_data(vcpu); + return p_exit_data->exit_reason; +} + +/* + * The guest has exited. See if we can fix it or if we need userspace + * assistance. + */ +static int kvm_handle_exit(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu) +{ + u32 exit_reason = kvm_get_exit_reason(vcpu); + vcpu->arch.last_exit = exit_reason; + + if (exit_reason < kvm_vti_max_exit_handlers + && kvm_vti_exit_handlers[exit_reason]) + return kvm_vti_exit_handlers[exit_reason](vcpu, kvm_run); + else { + kvm_run->exit_reason = KVM_EXIT_UNKNOWN; + kvm_run->hw.hardware_exit_reason = exit_reason; + } + return 0; +} + +static inline void vti_set_rr6(unsigned long rr6) +{ + ia64_set_rr(RR6, rr6); + ia64_srlz_i(); +} + +static int kvm_insert_vmm_mapping(struct kvm_vcpu *vcpu) +{ + unsigned long pte; + struct kvm *kvm = vcpu->kvm; + int r; + + /*Insert a pair of tr to map vmm*/ + pte = pte_val(mk_pte_phys(__pa(kvm_vmm_base), PAGE_KERNEL)); + r = ia64_itr_entry(0x3, KVM_VMM_BASE, pte, KVM_VMM_SHIFT); + if (r < 0) + goto out; + vcpu->arch.vmm_tr_slot = r; + /*Insert a pairt of tr to map data of vm*/ + pte = pte_val(mk_pte_phys(__pa(kvm->arch.vm_base), PAGE_KERNEL)); + r = ia64_itr_entry(0x3, KVM_VM_DATA_BASE, + pte, KVM_VM_DATA_SHIFT); + if (r < 0) + goto out; + vcpu->arch.vm_tr_slot = r; + r = 0; +out: + return r; + +} + +static void kvm_purge_vmm_mapping(struct kvm_vcpu *vcpu) +{ + + ia64_ptr_entry(0x3, vcpu->arch.vmm_tr_slot); + ia64_ptr_entry(0x3, vcpu->arch.vm_tr_slot); + +} + +static int kvm_vcpu_pre_transition(struct kvm_vcpu *vcpu) +{ + int cpu = smp_processor_id(); + + if (vcpu->arch.last_run_cpu != cpu || + per_cpu(last_vcpu, cpu) != vcpu) { + per_cpu(last_vcpu, cpu) = vcpu; + vcpu->arch.last_run_cpu = cpu; + kvm_flush_tlb_all(); + } + + vcpu->arch.host_rr6 = ia64_get_rr(RR6); + vti_set_rr6(vcpu->arch.vmm_rr); + return kvm_insert_vmm_mapping(vcpu); +} +static void kvm_vcpu_post_transition(struct kvm_vcpu *vcpu) +{ + kvm_purge_vmm_mapping(vcpu); + vti_set_rr6(vcpu->arch.host_rr6); +} + +static int vti_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + union context *host_ctx, *guest_ctx; + int r; + + /*Get host and guest context with guest address space.*/ + host_ctx = kvm_get_host_context(vcpu); + guest_ctx = kvm_get_guest_context(vcpu); + + r = kvm_vcpu_pre_transition(vcpu); + if (r < 0) + goto out; + kvm_vmm_info->tramp_entry(host_ctx, guest_ctx); + kvm_vcpu_post_transition(vcpu); + r = 0; +out: + return r; +} + +static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + int r; + +again: + preempt_disable(); + + kvm_prepare_guest_switch(vcpu); + local_irq_disable(); + + if (signal_pending(current)) { + local_irq_enable(); + preempt_enable(); + r = -EINTR; + kvm_run->exit_reason = KVM_EXIT_INTR; + goto out; + } + + vcpu->guest_mode = 1; + kvm_guest_enter(); + + r = vti_vcpu_run(vcpu, kvm_run); + if (r < 0) { + local_irq_enable(); + preempt_enable(); + kvm_run->exit_reason = KVM_EXIT_FAIL_ENTRY; + goto out; + } + + vcpu->arch.launched = 1; + vcpu->guest_mode = 0; + local_irq_enable(); + + /* + * We must have an instruction between local_irq_enable() and + * kvm_guest_exit(), so the timer interrupt isn't delayed by + * the interrupt shadow. The stat.exits increment will do nicely. + * But we need to prevent reordering, hence this barrier(): + */ + barrier(); + + kvm_guest_exit(); + + preempt_enable(); + + r = kvm_handle_exit(kvm_run, vcpu); + + if (r > 0) { + if (!need_resched()) + goto again; + } + +out: + if (r > 0) { + kvm_resched(vcpu); + goto again; + } + + return r; +} + +static void kvm_set_mmio_data(struct kvm_vcpu *vcpu) +{ + struct kvm_mmio_req *p = kvm_get_vcpu_ioreq(vcpu); + + if (!vcpu->mmio_is_write) + memcpy(&p->data, vcpu->mmio_data, 8); + p->state = STATE_IORESP_READY; +} + +int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + int r; + sigset_t sigsaved; + + vcpu_load(vcpu); + + if (unlikely(vcpu->arch.mp_state == KVM_MP_STATE_UNINITIALIZED)) { + kvm_vcpu_block(vcpu); + vcpu_put(vcpu); + return -EAGAIN; + } + + if (vcpu->sigset_active) + sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved); + + if (vcpu->mmio_needed) { + memcpy(vcpu->mmio_data, kvm_run->mmio.data, 8); + kvm_set_mmio_data(vcpu); + vcpu->mmio_read_completed = 1; + vcpu->mmio_needed = 0; + } + r = __vcpu_run(vcpu, kvm_run); + + if (vcpu->sigset_active) + sigprocmask(SIG_SETMASK, &sigsaved, NULL); + + vcpu_put(vcpu); + return r; +} + +/* + * Allocate 16M memory for every vm to hold its specific data. + * Its memory map is defined in kvm_host.h. + */ +static struct kvm *kvm_alloc_kvm(void) +{ + + struct kvm *kvm; + uint64_t vm_base; + + vm_base = __get_free_pages(GFP_KERNEL, get_order(KVM_VM_DATA_SIZE)); + + if (!vm_base) + return ERR_PTR(-ENOMEM); + printk(KERN_DEBUG"kvm: VM data's base Address:0x%lx\n", vm_base); + + /* Zero all pages before use! */ + memset((void *)vm_base, 0, KVM_VM_DATA_SIZE); + + kvm = (struct kvm *)(vm_base + KVM_VM_OFS); + kvm->arch.vm_base = vm_base; + + return kvm; +} + +struct kvm_io_range { + unsigned long start; + unsigned long size; + unsigned long type; +}; + +static const struct kvm_io_range io_ranges[] = { + {VGA_IO_START, VGA_IO_SIZE, GPFN_FRAME_BUFFER}, + {MMIO_START, MMIO_SIZE, GPFN_LOW_MMIO}, + {LEGACY_IO_START, LEGACY_IO_SIZE, GPFN_LEGACY_IO}, + {IO_SAPIC_START, IO_SAPIC_SIZE, GPFN_IOSAPIC}, + {PIB_START, PIB_SIZE, GPFN_PIB}, +}; + +static void kvm_build_io_pmt(struct kvm *kvm) +{ + unsigned long i, j; + + /* Mark I/O ranges */ + for (i = 0; i < (sizeof(io_ranges) / sizeof(struct kvm_io_range)); + i++) { + for (j = io_ranges[i].start; + j < io_ranges[i].start + io_ranges[i].size; + j += PAGE_SIZE) + kvm_set_pmt_entry(kvm, j >> PAGE_SHIFT, + io_ranges[i].type, 0); + } + +} + +/*Use unused rids to virtualize guest rid.*/ +#define GUEST_PHYSICAL_RR0 0x1739 +#define GUEST_PHYSICAL_RR4 0x2739 +#define VMM_INIT_RR 0x1660 + +static void kvm_init_vm(struct kvm *kvm) +{ + long vm_base; + + BUG_ON(!kvm); + + kvm->arch.metaphysical_rr0 = GUEST_PHYSICAL_RR0; + kvm->arch.metaphysical_rr4 = GUEST_PHYSICAL_RR4; + kvm->arch.vmm_init_rr = VMM_INIT_RR; + + vm_base = kvm->arch.vm_base; + if (vm_base) { + kvm->arch.vhpt_base = vm_base + KVM_VHPT_OFS; + kvm->arch.vtlb_base = vm_base + KVM_VTLB_OFS; + kvm->arch.vpd_base = vm_base + KVM_VPD_OFS; + } + + /* + *Fill P2M entries for MMIO/IO ranges + */ + kvm_build_io_pmt(kvm); + +} + +struct kvm *kvm_arch_create_vm(void) +{ + struct kvm *kvm = kvm_alloc_kvm(); + + if (IS_ERR(kvm)) + return ERR_PTR(-ENOMEM); + kvm_init_vm(kvm); + + return kvm; + +} + +static int kvm_vm_ioctl_get_irqchip(struct kvm *kvm, + struct kvm_irqchip *chip) +{ + int r; + + r = 0; + switch (chip->chip_id) { + case KVM_IRQCHIP_IOAPIC: + memcpy(&chip->chip.ioapic, ioapic_irqchip(kvm), + sizeof(struct kvm_ioapic_state)); + break; + default: + r = -EINVAL; + break; + } + return r; +} + +static int kvm_vm_ioctl_set_irqchip(struct kvm *kvm, struct kvm_irqchip *chip) +{ + int r; + + r = 0; + switch (chip->chip_id) { + case KVM_IRQCHIP_IOAPIC: + memcpy(ioapic_irqchip(kvm), + &chip->chip.ioapic, + sizeof(struct kvm_ioapic_state)); + break; + default: + r = -EINVAL; + break; + } + return r; +} + +#define RESTORE_REGS(_x) vcpu->arch._x = regs->_x + +int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +{ + int i; + struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd); + int r; + + vcpu_load(vcpu); + + for (i = 0; i < 16; i++) { + vpd->vgr[i] = regs->vpd.vgr[i]; + vpd->vbgr[i] = regs->vpd.vbgr[i]; + } + for (i = 0; i < 128; i++) + vpd->vcr[i] = regs->vpd.vcr[i]; + vpd->vhpi = regs->vpd.vhpi; + vpd->vnat = regs->vpd.vnat; + vpd->vbnat = regs->vpd.vbnat; + vpd->vpsr = regs->vpd.vpsr; + + vpd->vpr = regs->vpd.vpr; + + r = -EFAULT; + r = copy_from_user(&vcpu->arch.guest, regs->saved_guest, + sizeof(union context)); + if (r) + goto out; + r = copy_from_user(vcpu + 1, regs->saved_stack + + sizeof(struct kvm_vcpu), + IA64_STK_OFFSET - sizeof(struct kvm_vcpu)); + if (r) + goto out; + vcpu->arch.exit_data = + ((struct kvm_vcpu *)(regs->saved_stack))->arch.exit_data; + + RESTORE_REGS(mp_state); + RESTORE_REGS(vmm_rr); + memcpy(vcpu->arch.itrs, regs->itrs, sizeof(struct thash_data) * NITRS); + memcpy(vcpu->arch.dtrs, regs->dtrs, sizeof(struct thash_data) * NDTRS); + RESTORE_REGS(itr_regions); + RESTORE_REGS(dtr_regions); + RESTORE_REGS(tc_regions); + RESTORE_REGS(irq_check); + RESTORE_REGS(itc_check); + RESTORE_REGS(timer_check); + RESTORE_REGS(timer_pending); + RESTORE_REGS(last_itc); + for (i = 0; i < 8; i++) { + vcpu->arch.vrr[i] = regs->vrr[i]; + vcpu->arch.ibr[i] = regs->ibr[i]; + vcpu->arch.dbr[i] = regs->dbr[i]; + } + for (i = 0; i < 4; i++) + vcpu->arch.insvc[i] = regs->insvc[i]; + RESTORE_REGS(xtp); + RESTORE_REGS(metaphysical_rr0); + RESTORE_REGS(metaphysical_rr4); + RESTORE_REGS(metaphysical_saved_rr0); + RESTORE_REGS(metaphysical_saved_rr4); + RESTORE_REGS(fp_psr); + RESTORE_REGS(saved_gp); + + vcpu->arch.irq_new_pending = 1; + vcpu->arch.itc_offset = regs->saved_itc - ia64_getreg(_IA64_REG_AR_ITC); + set_bit(KVM_REQ_RESUME, &vcpu->requests); + + vcpu_put(vcpu); + r = 0; +out: + return r; +} + +long kvm_arch_vm_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + struct kvm *kvm = filp->private_data; + void __user *argp = (void __user *)arg; + int r = -EINVAL; + + switch (ioctl) { + case KVM_SET_MEMORY_REGION: { + struct kvm_memory_region kvm_mem; + struct kvm_userspace_memory_region kvm_userspace_mem; + + r = -EFAULT; + if (copy_from_user(&kvm_mem, argp, sizeof kvm_mem)) + goto out; + kvm_userspace_mem.slot = kvm_mem.slot; + kvm_userspace_mem.flags = kvm_mem.flags; + kvm_userspace_mem.guest_phys_addr = + kvm_mem.guest_phys_addr; + kvm_userspace_mem.memory_size = kvm_mem.memory_size; + r = kvm_vm_ioctl_set_memory_region(kvm, + &kvm_userspace_mem, 0); + if (r) + goto out; + break; + } + case KVM_CREATE_IRQCHIP: + r = -EFAULT; + r = kvm_ioapic_init(kvm); + if (r) + goto out; + break; + case KVM_IRQ_LINE: { + struct kvm_irq_level irq_event; + + r = -EFAULT; + if (copy_from_user(&irq_event, argp, sizeof irq_event)) + goto out; + if (irqchip_in_kernel(kvm)) { + mutex_lock(&kvm->lock); + kvm_ioapic_set_irq(kvm->arch.vioapic, + irq_event.irq, + irq_event.level); + mutex_unlock(&kvm->lock); + r = 0; + } + break; + } + case KVM_GET_IRQCHIP: { + /* 0: PIC master, 1: PIC slave, 2: IOAPIC */ + struct kvm_irqchip chip; + + r = -EFAULT; + if (copy_from_user(&chip, argp, sizeof chip)) + goto out; + r = -ENXIO; + if (!irqchip_in_kernel(kvm)) + goto out; + r = kvm_vm_ioctl_get_irqchip(kvm, &chip); + if (r) + goto out; + r = -EFAULT; + if (copy_to_user(argp, &chip, sizeof chip)) + goto out; + r = 0; + break; + } + case KVM_SET_IRQCHIP: { + /* 0: PIC master, 1: PIC slave, 2: IOAPIC */ + struct kvm_irqchip chip; + + r = -EFAULT; + if (copy_from_user(&chip, argp, sizeof chip)) + goto out; + r = -ENXIO; + if (!irqchip_in_kernel(kvm)) + goto out; + r = kvm_vm_ioctl_set_irqchip(kvm, &chip); + if (r) + goto out; + r = 0; + break; + } + default: + ; + } +out: + return r; +} + +int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu, + struct kvm_sregs *sregs) +{ + return -EINVAL; +} + +int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu, + struct kvm_sregs *sregs) +{ + return -EINVAL; + +} +int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu, + struct kvm_translation *tr) +{ + + return -EINVAL; +} + +static int kvm_alloc_vmm_area(void) +{ + if (!kvm_vmm_base && (kvm_vm_buffer_size < KVM_VM_BUFFER_SIZE)) { + kvm_vmm_base = __get_free_pages(GFP_KERNEL, + get_order(KVM_VMM_SIZE)); + if (!kvm_vmm_base) + return -ENOMEM; + + memset((void *)kvm_vmm_base, 0, KVM_VMM_SIZE); + kvm_vm_buffer = kvm_vmm_base + VMM_SIZE; + + printk(KERN_DEBUG"kvm:VMM's Base Addr:0x%lx, vm_buffer:0x%lx\n", + kvm_vmm_base, kvm_vm_buffer); + } + + return 0; +} + +static void kvm_free_vmm_area(void) +{ + if (kvm_vmm_base) { + /*Zero this area before free to avoid bits leak!!*/ + memset((void *)kvm_vmm_base, 0, KVM_VMM_SIZE); + free_pages(kvm_vmm_base, get_order(KVM_VMM_SIZE)); + kvm_vmm_base = 0; + kvm_vm_buffer = 0; + kvm_vsa_base = 0; + } +} + +/* + * Make sure that a cpu that is being hot-unplugged does not have any vcpus + * cached on it. Leave it as blank for IA64. + */ +void decache_vcpus_on_cpu(int cpu) +{ +} + +static void vti_vcpu_load(struct kvm_vcpu *vcpu, int cpu) +{ +} + +static int vti_init_vpd(struct kvm_vcpu *vcpu) +{ + int i; + union cpuid3_t cpuid3; + struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd); + + if (IS_ERR(vpd)) + return PTR_ERR(vpd); + + /* CPUID init */ + for (i = 0; i < 5; i++) + vpd->vcpuid[i] = ia64_get_cpuid(i); + + /* Limit the CPUID number to 5 */ + cpuid3.value = vpd->vcpuid[3]; + cpuid3.number = 4; /* 5 - 1 */ + vpd->vcpuid[3] = cpuid3.value; + + /*Set vac and vdc fields*/ + vpd->vac.a_from_int_cr = 1; + vpd->vac.a_to_int_cr = 1; + vpd->vac.a_from_psr = 1; + vpd->vac.a_from_cpuid = 1; + vpd->vac.a_cover = 1; + vpd->vac.a_bsw = 1; + vpd->vac.a_int = 1; + vpd->vdc.d_vmsw = 1; + + /*Set virtual buffer*/ + vpd->virt_env_vaddr = KVM_VM_BUFFER_BASE; + + return 0; +} + +static int vti_create_vp(struct kvm_vcpu *vcpu) +{ + long ret; + struct vpd *vpd = vcpu->arch.vpd; + unsigned long vmm_ivt; + + vmm_ivt = kvm_vmm_info->vmm_ivt; + + printk(KERN_DEBUG "kvm: vcpu:%p,ivt: 0x%lx\n", vcpu, vmm_ivt); + + ret = ia64_pal_vp_create((u64 *)vpd, (u64 *)vmm_ivt, 0); + + if (ret) { + printk(KERN_ERR"kvm: ia64_pal_vp_create failed!\n"); + return -EINVAL; + } + return 0; +} + +static void init_ptce_info(struct kvm_vcpu *vcpu) +{ + ia64_ptce_info_t ptce = {0}; + + ia64_get_ptce(&ptce); + vcpu->arch.ptce_base = ptce.base; + vcpu->arch.ptce_count[0] = ptce.count[0]; + vcpu->arch.ptce_count[1] = ptce.count[1]; + vcpu->arch.ptce_stride[0] = ptce.stride[0]; + vcpu->arch.ptce_stride[1] = ptce.stride[1]; +} + +static void kvm_migrate_hlt_timer(struct kvm_vcpu *vcpu) +{ + struct hrtimer *p_ht = &vcpu->arch.hlt_timer; + + if (hrtimer_cancel(p_ht)) + hrtimer_start(p_ht, p_ht->expires, HRTIMER_MODE_ABS); +} + +static enum hrtimer_restart hlt_timer_fn(struct hrtimer *data) +{ + struct kvm_vcpu *vcpu; + wait_queue_head_t *q; + + vcpu = container_of(data, struct kvm_vcpu, arch.hlt_timer); + if (vcpu->arch.mp_state != KVM_MP_STATE_HALTED) + goto out; + + q = &vcpu->wq; + if (waitqueue_active(q)) { + vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE; + wake_up_interruptible(q); + } +out: + vcpu->arch.timer_check = 1; + return HRTIMER_NORESTART; +} + +#define PALE_RESET_ENTRY 0x80000000ffffffb0UL + +int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) +{ + struct kvm_vcpu *v; + int r; + int i; + long itc_offset; + struct kvm *kvm = vcpu->kvm; + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + union context *p_ctx = &vcpu->arch.guest; + struct kvm_vcpu *vmm_vcpu = to_guest(vcpu->kvm, vcpu); + + /*Init vcpu context for first run.*/ + if (IS_ERR(vmm_vcpu)) + return PTR_ERR(vmm_vcpu); + + if (vcpu->vcpu_id == 0) { + vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE; + + /*Set entry address for first run.*/ + regs->cr_iip = PALE_RESET_ENTRY; + + /*Initilize itc offset for vcpus*/ + itc_offset = 0UL - ia64_getreg(_IA64_REG_AR_ITC); + for (i = 0; i < MAX_VCPU_NUM; i++) { + v = (struct kvm_vcpu *)((char *)vcpu + VCPU_SIZE * i); + v->arch.itc_offset = itc_offset; + v->arch.last_itc = 0; + } + } else + vcpu->arch.mp_state = KVM_MP_STATE_UNINITIALIZED; + + r = -ENOMEM; + vcpu->arch.apic = kzalloc(sizeof(struct kvm_lapic), GFP_KERNEL); + if (!vcpu->arch.apic) + goto out; + vcpu->arch.apic->vcpu = vcpu; + + p_ctx->gr[1] = 0; + p_ctx->gr[12] = (unsigned long)((char *)vmm_vcpu + IA64_STK_OFFSET); + p_ctx->gr[13] = (unsigned long)vmm_vcpu; + p_ctx->psr = 0x1008522000UL; + p_ctx->ar[40] = FPSR_DEFAULT; /*fpsr*/ + p_ctx->caller_unat = 0; + p_ctx->pr = 0x0; + p_ctx->ar[36] = 0x0; /*unat*/ + p_ctx->ar[19] = 0x0; /*rnat*/ + p_ctx->ar[18] = (unsigned long)vmm_vcpu + + ((sizeof(struct kvm_vcpu)+15) & ~15); + p_ctx->ar[64] = 0x0; /*pfs*/ + p_ctx->cr[0] = 0x7e04UL; + p_ctx->cr[2] = (unsigned long)kvm_vmm_info->vmm_ivt; + p_ctx->cr[8] = 0x3c; + + /*Initilize region register*/ + p_ctx->rr[0] = 0x30; + p_ctx->rr[1] = 0x30; + p_ctx->rr[2] = 0x30; + p_ctx->rr[3] = 0x30; + p_ctx->rr[4] = 0x30; + p_ctx->rr[5] = 0x30; + p_ctx->rr[7] = 0x30; + + /*Initilize branch register 0*/ + p_ctx->br[0] = *(unsigned long *)kvm_vmm_info->vmm_entry; + + vcpu->arch.vmm_rr = kvm->arch.vmm_init_rr; + vcpu->arch.metaphysical_rr0 = kvm->arch.metaphysical_rr0; + vcpu->arch.metaphysical_rr4 = kvm->arch.metaphysical_rr4; + + hrtimer_init(&vcpu->arch.hlt_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + vcpu->arch.hlt_timer.function = hlt_timer_fn; + + vcpu->arch.last_run_cpu = -1; + vcpu->arch.vpd = (struct vpd *)VPD_ADDR(vcpu->vcpu_id); + vcpu->arch.vsa_base = kvm_vsa_base; + vcpu->arch.__gp = kvm_vmm_gp; + vcpu->arch.dirty_log_lock_pa = __pa(&kvm->arch.dirty_log_lock); + vcpu->arch.vhpt.hash = (struct thash_data *)VHPT_ADDR(vcpu->vcpu_id); + vcpu->arch.vtlb.hash = (struct thash_data *)VTLB_ADDR(vcpu->vcpu_id); + init_ptce_info(vcpu); + + r = 0; +out: + return r; +} + +static int vti_vcpu_setup(struct kvm_vcpu *vcpu, int id) +{ + unsigned long psr; + int r; + + local_irq_save(psr); + r = kvm_insert_vmm_mapping(vcpu); + if (r) + goto fail; + r = kvm_vcpu_init(vcpu, vcpu->kvm, id); + if (r) + goto fail; + + r = vti_init_vpd(vcpu); + if (r) { + printk(KERN_DEBUG"kvm: vpd init error!!\n"); + goto uninit; + } + + r = vti_create_vp(vcpu); + if (r) + goto uninit; + + kvm_purge_vmm_mapping(vcpu); + local_irq_restore(psr); + + return 0; +uninit: + kvm_vcpu_uninit(vcpu); +fail: + return r; +} + +struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, + unsigned int id) +{ + struct kvm_vcpu *vcpu; + unsigned long vm_base = kvm->arch.vm_base; + int r; + int cpu; + + r = -ENOMEM; + if (!vm_base) { + printk(KERN_ERR"kvm: Create vcpu[%d] error!\n", id); + goto fail; + } + vcpu = (struct kvm_vcpu *)(vm_base + KVM_VCPU_OFS + VCPU_SIZE * id); + vcpu->kvm = kvm; + + cpu = get_cpu(); + vti_vcpu_load(vcpu, cpu); + r = vti_vcpu_setup(vcpu, id); + put_cpu(); + + if (r) { + printk(KERN_DEBUG"kvm: vcpu_setup error!!\n"); + goto fail; + } + + return vcpu; +fail: + return ERR_PTR(r); +} + +int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) +{ + return 0; +} + +int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu) +{ + return -EINVAL; +} + +int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu) +{ + return -EINVAL; +} + +int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu, + struct kvm_debug_guest *dbg) +{ + return -EINVAL; +} + +static void free_kvm(struct kvm *kvm) +{ + unsigned long vm_base = kvm->arch.vm_base; + + if (vm_base) { + memset((void *)vm_base, 0, KVM_VM_DATA_SIZE); + free_pages(vm_base, get_order(KVM_VM_DATA_SIZE)); + } + +} + +static void kvm_release_vm_pages(struct kvm *kvm) +{ + struct kvm_memory_slot *memslot; + int i, j; + unsigned long base_gfn; + + for (i = 0; i < kvm->nmemslots; i++) { + memslot = &kvm->memslots[i]; + base_gfn = memslot->base_gfn; + + for (j = 0; j < memslot->npages; j++) { + if (memslot->rmap[j]) + put_page((struct page *)memslot->rmap[j]); + } + } +} + +void kvm_arch_destroy_vm(struct kvm *kvm) +{ + kfree(kvm->arch.vioapic); + kvm_release_vm_pages(kvm); + kvm_free_physmem(kvm); + free_kvm(kvm); +} + +void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) +{ +} + +void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) +{ + if (cpu != vcpu->cpu) { + vcpu->cpu = cpu; + if (vcpu->arch.ht_active) + kvm_migrate_hlt_timer(vcpu); + } +} + +#define SAVE_REGS(_x) regs->_x = vcpu->arch._x + +int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +{ + int i; + int r; + struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd); + vcpu_load(vcpu); + + for (i = 0; i < 16; i++) { + regs->vpd.vgr[i] = vpd->vgr[i]; + regs->vpd.vbgr[i] = vpd->vbgr[i]; + } + for (i = 0; i < 128; i++) + regs->vpd.vcr[i] = vpd->vcr[i]; + regs->vpd.vhpi = vpd->vhpi; + regs->vpd.vnat = vpd->vnat; + regs->vpd.vbnat = vpd->vbnat; + regs->vpd.vpsr = vpd->vpsr; + regs->vpd.vpr = vpd->vpr; + + r = -EFAULT; + r = copy_to_user(regs->saved_guest, &vcpu->arch.guest, + sizeof(union context)); + if (r) + goto out; + r = copy_to_user(regs->saved_stack, (void *)vcpu, IA64_STK_OFFSET); + if (r) + goto out; + SAVE_REGS(mp_state); + SAVE_REGS(vmm_rr); + memcpy(regs->itrs, vcpu->arch.itrs, sizeof(struct thash_data) * NITRS); + memcpy(regs->dtrs, vcpu->arch.dtrs, sizeof(struct thash_data) * NDTRS); + SAVE_REGS(itr_regions); + SAVE_REGS(dtr_regions); + SAVE_REGS(tc_regions); + SAVE_REGS(irq_check); + SAVE_REGS(itc_check); + SAVE_REGS(timer_check); + SAVE_REGS(timer_pending); + SAVE_REGS(last_itc); + for (i = 0; i < 8; i++) { + regs->vrr[i] = vcpu->arch.vrr[i]; + regs->ibr[i] = vcpu->arch.ibr[i]; + regs->dbr[i] = vcpu->arch.dbr[i]; + } + for (i = 0; i < 4; i++) + regs->insvc[i] = vcpu->arch.insvc[i]; + regs->saved_itc = vcpu->arch.itc_offset + ia64_getreg(_IA64_REG_AR_ITC); + SAVE_REGS(xtp); + SAVE_REGS(metaphysical_rr0); + SAVE_REGS(metaphysical_rr4); + SAVE_REGS(metaphysical_saved_rr0); + SAVE_REGS(metaphysical_saved_rr4); + SAVE_REGS(fp_psr); + SAVE_REGS(saved_gp); + vcpu_put(vcpu); + r = 0; +out: + return r; +} + +void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) +{ + + hrtimer_cancel(&vcpu->arch.hlt_timer); + kfree(vcpu->arch.apic); +} + + +long kvm_arch_vcpu_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + return -EINVAL; +} + +int kvm_arch_set_memory_region(struct kvm *kvm, + struct kvm_userspace_memory_region *mem, + struct kvm_memory_slot old, + int user_alloc) +{ + unsigned long i; + struct page *page; + int npages = mem->memory_size >> PAGE_SHIFT; + struct kvm_memory_slot *memslot = &kvm->memslots[mem->slot]; + unsigned long base_gfn = memslot->base_gfn; + + for (i = 0; i < npages; i++) { + page = gfn_to_page(kvm, base_gfn + i); + kvm_set_pmt_entry(kvm, base_gfn + i, + page_to_pfn(page) << PAGE_SHIFT, + _PAGE_AR_RWX|_PAGE_MA_WB); + memslot->rmap[i] = (unsigned long)page; + } + + return 0; +} + + +long kvm_arch_dev_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + return -EINVAL; +} + +void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) +{ + kvm_vcpu_uninit(vcpu); +} + +static int vti_cpu_has_kvm_support(void) +{ + long avail = 1, status = 1, control = 1; + long ret; + + ret = ia64_pal_proc_get_features(&avail, &status, &control, 0); + if (ret) + goto out; + + if (!(avail & PAL_PROC_VM_BIT)) + goto out; + + printk(KERN_DEBUG"kvm: Hardware Supports VT\n"); + + ret = ia64_pal_vp_env_info(&kvm_vm_buffer_size, &vp_env_info); + if (ret) + goto out; + printk(KERN_DEBUG"kvm: VM Buffer Size:0x%lx\n", kvm_vm_buffer_size); + + if (!(vp_env_info & VP_OPCODE)) { + printk(KERN_WARNING"kvm: No opcode ability on hardware, " + "vm_env_info:0x%lx\n", vp_env_info); + } + + return 1; +out: + return 0; +} + +static int kvm_relocate_vmm(struct kvm_vmm_info *vmm_info, + struct module *module) +{ + unsigned long module_base; + unsigned long vmm_size; + + unsigned long vmm_offset, func_offset, fdesc_offset; + struct fdesc *p_fdesc; + + BUG_ON(!module); + + if (!kvm_vmm_base) { + printk("kvm: kvm area hasn't been initilized yet!!\n"); + return -EFAULT; + } + + /*Calculate new position of relocated vmm module.*/ + module_base = (unsigned long)module->module_core; + vmm_size = module->core_size; + if (unlikely(vmm_size > KVM_VMM_SIZE)) + return -EFAULT; + + memcpy((void *)kvm_vmm_base, (void *)module_base, vmm_size); + kvm_flush_icache(kvm_vmm_base, vmm_size); + + /*Recalculate kvm_vmm_info based on new VMM*/ + vmm_offset = vmm_info->vmm_ivt - module_base; + kvm_vmm_info->vmm_ivt = KVM_VMM_BASE + vmm_offset; + printk(KERN_DEBUG"kvm: Relocated VMM's IVT Base Addr:%lx\n", + kvm_vmm_info->vmm_ivt); + + fdesc_offset = (unsigned long)vmm_info->vmm_entry - module_base; + kvm_vmm_info->vmm_entry = (kvm_vmm_entry *)(KVM_VMM_BASE + + fdesc_offset); + func_offset = *(unsigned long *)vmm_info->vmm_entry - module_base; + p_fdesc = (struct fdesc *)(kvm_vmm_base + fdesc_offset); + p_fdesc->ip = KVM_VMM_BASE + func_offset; + p_fdesc->gp = KVM_VMM_BASE+(p_fdesc->gp - module_base); + + printk(KERN_DEBUG"kvm: Relocated VMM's Init Entry Addr:%lx\n", + KVM_VMM_BASE+func_offset); + + fdesc_offset = (unsigned long)vmm_info->tramp_entry - module_base; + kvm_vmm_info->tramp_entry = (kvm_tramp_entry *)(KVM_VMM_BASE + + fdesc_offset); + func_offset = *(unsigned long *)vmm_info->tramp_entry - module_base; + p_fdesc = (struct fdesc *)(kvm_vmm_base + fdesc_offset); + p_fdesc->ip = KVM_VMM_BASE + func_offset; + p_fdesc->gp = KVM_VMM_BASE + (p_fdesc->gp - module_base); + + kvm_vmm_gp = p_fdesc->gp; + + printk(KERN_DEBUG"kvm: Relocated VMM's Entry IP:%p\n", + kvm_vmm_info->vmm_entry); + printk(KERN_DEBUG"kvm: Relocated VMM's Trampoline Entry IP:0x%lx\n", + KVM_VMM_BASE + func_offset); + + return 0; +} + +int kvm_arch_init(void *opaque) +{ + int r; + struct kvm_vmm_info *vmm_info = (struct kvm_vmm_info *)opaque; + + if (!vti_cpu_has_kvm_support()) { + printk(KERN_ERR "kvm: No Hardware Virtualization Support!\n"); + r = -EOPNOTSUPP; + goto out; + } + + if (kvm_vmm_info) { + printk(KERN_ERR "kvm: Already loaded VMM module!\n"); + r = -EEXIST; + goto out; + } + + r = -ENOMEM; + kvm_vmm_info = kzalloc(sizeof(struct kvm_vmm_info), GFP_KERNEL); + if (!kvm_vmm_info) + goto out; + + if (kvm_alloc_vmm_area()) + goto out_free0; + + r = kvm_relocate_vmm(vmm_info, vmm_info->module); + if (r) + goto out_free1; + + return 0; + +out_free1: + kvm_free_vmm_area(); +out_free0: + kfree(kvm_vmm_info); +out: + return r; +} + +void kvm_arch_exit(void) +{ + kvm_free_vmm_area(); + kfree(kvm_vmm_info); + kvm_vmm_info = NULL; +} + +static int kvm_ia64_sync_dirty_log(struct kvm *kvm, + struct kvm_dirty_log *log) +{ + struct kvm_memory_slot *memslot; + int r, i; + long n, base; + unsigned long *dirty_bitmap = (unsigned long *)((void *)kvm - KVM_VM_OFS + + KVM_MEM_DIRTY_LOG_OFS); + + r = -EINVAL; + if (log->slot >= KVM_MEMORY_SLOTS) + goto out; + + memslot = &kvm->memslots[log->slot]; + r = -ENOENT; + if (!memslot->dirty_bitmap) + goto out; + + n = ALIGN(memslot->npages, BITS_PER_LONG) / 8; + base = memslot->base_gfn / BITS_PER_LONG; + + for (i = 0; i < n/sizeof(long); ++i) { + memslot->dirty_bitmap[i] = dirty_bitmap[base + i]; + dirty_bitmap[base + i] = 0; + } + r = 0; +out: + return r; +} + +int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm, + struct kvm_dirty_log *log) +{ + int r; + int n; + struct kvm_memory_slot *memslot; + int is_dirty = 0; + + spin_lock(&kvm->arch.dirty_log_lock); + + r = kvm_ia64_sync_dirty_log(kvm, log); + if (r) + goto out; + + r = kvm_get_dirty_log(kvm, log, &is_dirty); + if (r) + goto out; + + /* If nothing is dirty, don't bother messing with page tables. */ + if (is_dirty) { + kvm_flush_remote_tlbs(kvm); + memslot = &kvm->memslots[log->slot]; + n = ALIGN(memslot->npages, BITS_PER_LONG) / 8; + memset(memslot->dirty_bitmap, 0, n); + } + r = 0; +out: + spin_unlock(&kvm->arch.dirty_log_lock); + return r; +} + +int kvm_arch_hardware_setup(void) +{ + return 0; +} + +void kvm_arch_hardware_unsetup(void) +{ +} + +static void vcpu_kick_intr(void *info) +{ +#ifdef DEBUG + struct kvm_vcpu *vcpu = (struct kvm_vcpu *)info; + printk(KERN_DEBUG"vcpu_kick_intr %p \n", vcpu); +#endif +} + +void kvm_vcpu_kick(struct kvm_vcpu *vcpu) +{ + int ipi_pcpu = vcpu->cpu; + + if (waitqueue_active(&vcpu->wq)) + wake_up_interruptible(&vcpu->wq); + + if (vcpu->guest_mode) + smp_call_function_single(ipi_pcpu, vcpu_kick_intr, vcpu, 0, 0); +} + +int kvm_apic_set_irq(struct kvm_vcpu *vcpu, u8 vec, u8 trig) +{ + + struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd); + + if (!test_and_set_bit(vec, &vpd->irr[0])) { + vcpu->arch.irq_new_pending = 1; + if (vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE) + kvm_vcpu_kick(vcpu); + else if (vcpu->arch.mp_state == KVM_MP_STATE_HALTED) { + vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE; + if (waitqueue_active(&vcpu->wq)) + wake_up_interruptible(&vcpu->wq); + } + return 1; + } + return 0; +} + +int kvm_apic_match_physical_addr(struct kvm_lapic *apic, u16 dest) +{ + return apic->vcpu->vcpu_id == dest; +} + +int kvm_apic_match_logical_addr(struct kvm_lapic *apic, u8 mda) +{ + return 0; +} + +struct kvm_vcpu *kvm_get_lowest_prio_vcpu(struct kvm *kvm, u8 vector, + unsigned long bitmap) +{ + struct kvm_vcpu *lvcpu = kvm->vcpus[0]; + int i; + + for (i = 1; i < KVM_MAX_VCPUS; i++) { + if (!kvm->vcpus[i]) + continue; + if (lvcpu->arch.xtp > kvm->vcpus[i]->arch.xtp) + lvcpu = kvm->vcpus[i]; + } + + return lvcpu; +} + +static int find_highest_bits(int *dat) +{ + u32 bits, bitnum; + int i; + + /* loop for all 256 bits */ + for (i = 7; i >= 0 ; i--) { + bits = dat[i]; + if (bits) { + bitnum = fls(bits); + return i * 32 + bitnum - 1; + } + } + + return -1; +} + +int kvm_highest_pending_irq(struct kvm_vcpu *vcpu) +{ + struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd); + + if (vpd->irr[0] & (1UL << NMI_VECTOR)) + return NMI_VECTOR; + if (vpd->irr[0] & (1UL << ExtINT_VECTOR)) + return ExtINT_VECTOR; + + return find_highest_bits((int *)&vpd->irr[0]); +} + +int kvm_cpu_has_interrupt(struct kvm_vcpu *vcpu) +{ + if (kvm_highest_pending_irq(vcpu) != -1) + return 1; + return 0; +} + +int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu) +{ + return 0; +} + +gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn) +{ + return gfn; +} + +int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu) +{ + return vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE; +} + +int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu, + struct kvm_mp_state *mp_state) +{ + return -EINVAL; +} + +int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu, + struct kvm_mp_state *mp_state) +{ + return -EINVAL; +} diff --git a/arch/ia64/kvm/kvm_fw.c b/arch/ia64/kvm/kvm_fw.c new file mode 100644 index 0000000000000000000000000000000000000000..091f936c44853d83dab753d37d1d930eaee6740d --- /dev/null +++ b/arch/ia64/kvm/kvm_fw.c @@ -0,0 +1,500 @@ +/* + * PAL/SAL call delegation + * + * Copyright (c) 2004 Li Susie + * Copyright (c) 2005 Yu Ke + * Copyright (c) 2007 Xiantao Zhang + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + */ + +#include +#include + +#include "vti.h" +#include "misc.h" + +#include +#include +#include + +/* + * Handy macros to make sure that the PAL return values start out + * as something meaningful. + */ +#define INIT_PAL_STATUS_UNIMPLEMENTED(x) \ + { \ + x.status = PAL_STATUS_UNIMPLEMENTED; \ + x.v0 = 0; \ + x.v1 = 0; \ + x.v2 = 0; \ + } + +#define INIT_PAL_STATUS_SUCCESS(x) \ + { \ + x.status = PAL_STATUS_SUCCESS; \ + x.v0 = 0; \ + x.v1 = 0; \ + x.v2 = 0; \ + } + +static void kvm_get_pal_call_data(struct kvm_vcpu *vcpu, + u64 *gr28, u64 *gr29, u64 *gr30, u64 *gr31) { + struct exit_ctl_data *p; + + if (vcpu) { + p = &vcpu->arch.exit_data; + if (p->exit_reason == EXIT_REASON_PAL_CALL) { + *gr28 = p->u.pal_data.gr28; + *gr29 = p->u.pal_data.gr29; + *gr30 = p->u.pal_data.gr30; + *gr31 = p->u.pal_data.gr31; + return ; + } + } + printk(KERN_DEBUG"Failed to get vcpu pal data!!!\n"); +} + +static void set_pal_result(struct kvm_vcpu *vcpu, + struct ia64_pal_retval result) { + + struct exit_ctl_data *p; + + p = kvm_get_exit_data(vcpu); + if (p && p->exit_reason == EXIT_REASON_PAL_CALL) { + p->u.pal_data.ret = result; + return ; + } + INIT_PAL_STATUS_UNIMPLEMENTED(p->u.pal_data.ret); +} + +static void set_sal_result(struct kvm_vcpu *vcpu, + struct sal_ret_values result) { + struct exit_ctl_data *p; + + p = kvm_get_exit_data(vcpu); + if (p && p->exit_reason == EXIT_REASON_SAL_CALL) { + p->u.sal_data.ret = result; + return ; + } + printk(KERN_WARNING"Failed to set sal result!!\n"); +} + +struct cache_flush_args { + u64 cache_type; + u64 operation; + u64 progress; + long status; +}; + +cpumask_t cpu_cache_coherent_map; + +static void remote_pal_cache_flush(void *data) +{ + struct cache_flush_args *args = data; + long status; + u64 progress = args->progress; + + status = ia64_pal_cache_flush(args->cache_type, args->operation, + &progress, NULL); + if (status != 0) + args->status = status; +} + +static struct ia64_pal_retval pal_cache_flush(struct kvm_vcpu *vcpu) +{ + u64 gr28, gr29, gr30, gr31; + struct ia64_pal_retval result = {0, 0, 0, 0}; + struct cache_flush_args args = {0, 0, 0, 0}; + long psr; + + gr28 = gr29 = gr30 = gr31 = 0; + kvm_get_pal_call_data(vcpu, &gr28, &gr29, &gr30, &gr31); + + if (gr31 != 0) + printk(KERN_ERR"vcpu:%p called cache_flush error!\n", vcpu); + + /* Always call Host Pal in int=1 */ + gr30 &= ~PAL_CACHE_FLUSH_CHK_INTRS; + args.cache_type = gr29; + args.operation = gr30; + smp_call_function(remote_pal_cache_flush, + (void *)&args, 1, 1); + if (args.status != 0) + printk(KERN_ERR"pal_cache_flush error!," + "status:0x%lx\n", args.status); + /* + * Call Host PAL cache flush + * Clear psr.ic when call PAL_CACHE_FLUSH + */ + local_irq_save(psr); + result.status = ia64_pal_cache_flush(gr29, gr30, &result.v1, + &result.v0); + local_irq_restore(psr); + if (result.status != 0) + printk(KERN_ERR"vcpu:%p crashed due to cache_flush err:%ld" + "in1:%lx,in2:%lx\n", + vcpu, result.status, gr29, gr30); + +#if 0 + if (gr29 == PAL_CACHE_TYPE_COHERENT) { + cpus_setall(vcpu->arch.cache_coherent_map); + cpu_clear(vcpu->cpu, vcpu->arch.cache_coherent_map); + cpus_setall(cpu_cache_coherent_map); + cpu_clear(vcpu->cpu, cpu_cache_coherent_map); + } +#endif + return result; +} + +struct ia64_pal_retval pal_cache_summary(struct kvm_vcpu *vcpu) +{ + + struct ia64_pal_retval result; + + PAL_CALL(result, PAL_CACHE_SUMMARY, 0, 0, 0); + return result; +} + +static struct ia64_pal_retval pal_freq_base(struct kvm_vcpu *vcpu) +{ + + struct ia64_pal_retval result; + + PAL_CALL(result, PAL_FREQ_BASE, 0, 0, 0); + + /* + * PAL_FREQ_BASE may not be implemented in some platforms, + * call SAL instead. + */ + if (result.v0 == 0) { + result.status = ia64_sal_freq_base(SAL_FREQ_BASE_PLATFORM, + &result.v0, + &result.v1); + result.v2 = 0; + } + + return result; +} + +static struct ia64_pal_retval pal_freq_ratios(struct kvm_vcpu *vcpu) +{ + + struct ia64_pal_retval result; + + PAL_CALL(result, PAL_FREQ_RATIOS, 0, 0, 0); + return result; +} + +static struct ia64_pal_retval pal_logical_to_physica(struct kvm_vcpu *vcpu) +{ + struct ia64_pal_retval result; + + INIT_PAL_STATUS_UNIMPLEMENTED(result); + return result; +} + +static struct ia64_pal_retval pal_platform_addr(struct kvm_vcpu *vcpu) +{ + + struct ia64_pal_retval result; + + INIT_PAL_STATUS_SUCCESS(result); + return result; +} + +static struct ia64_pal_retval pal_proc_get_features(struct kvm_vcpu *vcpu) +{ + + struct ia64_pal_retval result = {0, 0, 0, 0}; + long in0, in1, in2, in3; + + kvm_get_pal_call_data(vcpu, &in0, &in1, &in2, &in3); + result.status = ia64_pal_proc_get_features(&result.v0, &result.v1, + &result.v2, in2); + + return result; +} + +static struct ia64_pal_retval pal_cache_info(struct kvm_vcpu *vcpu) +{ + + pal_cache_config_info_t ci; + long status; + unsigned long in0, in1, in2, in3, r9, r10; + + kvm_get_pal_call_data(vcpu, &in0, &in1, &in2, &in3); + status = ia64_pal_cache_config_info(in1, in2, &ci); + r9 = ci.pcci_info_1.pcci1_data; + r10 = ci.pcci_info_2.pcci2_data; + return ((struct ia64_pal_retval){status, r9, r10, 0}); +} + +#define GUEST_IMPL_VA_MSB 59 +#define GUEST_RID_BITS 18 + +static struct ia64_pal_retval pal_vm_summary(struct kvm_vcpu *vcpu) +{ + + pal_vm_info_1_u_t vminfo1; + pal_vm_info_2_u_t vminfo2; + struct ia64_pal_retval result; + + PAL_CALL(result, PAL_VM_SUMMARY, 0, 0, 0); + if (!result.status) { + vminfo1.pvi1_val = result.v0; + vminfo1.pal_vm_info_1_s.max_itr_entry = 8; + vminfo1.pal_vm_info_1_s.max_dtr_entry = 8; + result.v0 = vminfo1.pvi1_val; + vminfo2.pal_vm_info_2_s.impl_va_msb = GUEST_IMPL_VA_MSB; + vminfo2.pal_vm_info_2_s.rid_size = GUEST_RID_BITS; + result.v1 = vminfo2.pvi2_val; + } + + return result; +} + +static struct ia64_pal_retval pal_vm_info(struct kvm_vcpu *vcpu) +{ + struct ia64_pal_retval result; + + INIT_PAL_STATUS_UNIMPLEMENTED(result); + + return result; +} + +static u64 kvm_get_pal_call_index(struct kvm_vcpu *vcpu) +{ + u64 index = 0; + struct exit_ctl_data *p; + + p = kvm_get_exit_data(vcpu); + if (p && (p->exit_reason == EXIT_REASON_PAL_CALL)) + index = p->u.pal_data.gr28; + + return index; +} + +int kvm_pal_emul(struct kvm_vcpu *vcpu, struct kvm_run *run) +{ + + u64 gr28; + struct ia64_pal_retval result; + int ret = 1; + + gr28 = kvm_get_pal_call_index(vcpu); + /*printk("pal_call index:%lx\n",gr28);*/ + switch (gr28) { + case PAL_CACHE_FLUSH: + result = pal_cache_flush(vcpu); + break; + case PAL_CACHE_SUMMARY: + result = pal_cache_summary(vcpu); + break; + case PAL_HALT_LIGHT: + { + vcpu->arch.timer_pending = 1; + INIT_PAL_STATUS_SUCCESS(result); + if (kvm_highest_pending_irq(vcpu) == -1) + ret = kvm_emulate_halt(vcpu); + + } + break; + + case PAL_FREQ_RATIOS: + result = pal_freq_ratios(vcpu); + break; + + case PAL_FREQ_BASE: + result = pal_freq_base(vcpu); + break; + + case PAL_LOGICAL_TO_PHYSICAL : + result = pal_logical_to_physica(vcpu); + break; + + case PAL_VM_SUMMARY : + result = pal_vm_summary(vcpu); + break; + + case PAL_VM_INFO : + result = pal_vm_info(vcpu); + break; + case PAL_PLATFORM_ADDR : + result = pal_platform_addr(vcpu); + break; + case PAL_CACHE_INFO: + result = pal_cache_info(vcpu); + break; + case PAL_PTCE_INFO: + INIT_PAL_STATUS_SUCCESS(result); + result.v1 = (1L << 32) | 1L; + break; + case PAL_VM_PAGE_SIZE: + result.status = ia64_pal_vm_page_size(&result.v0, + &result.v1); + break; + case PAL_RSE_INFO: + result.status = ia64_pal_rse_info(&result.v0, + (pal_hints_u_t *)&result.v1); + break; + case PAL_PROC_GET_FEATURES: + result = pal_proc_get_features(vcpu); + break; + case PAL_DEBUG_INFO: + result.status = ia64_pal_debug_info(&result.v0, + &result.v1); + break; + case PAL_VERSION: + result.status = ia64_pal_version( + (pal_version_u_t *)&result.v0, + (pal_version_u_t *)&result.v1); + + break; + case PAL_FIXED_ADDR: + result.status = PAL_STATUS_SUCCESS; + result.v0 = vcpu->vcpu_id; + break; + default: + INIT_PAL_STATUS_UNIMPLEMENTED(result); + printk(KERN_WARNING"kvm: Unsupported pal call," + " index:0x%lx\n", gr28); + } + set_pal_result(vcpu, result); + return ret; +} + +static struct sal_ret_values sal_emulator(struct kvm *kvm, + long index, unsigned long in1, + unsigned long in2, unsigned long in3, + unsigned long in4, unsigned long in5, + unsigned long in6, unsigned long in7) +{ + unsigned long r9 = 0; + unsigned long r10 = 0; + long r11 = 0; + long status; + + status = 0; + switch (index) { + case SAL_FREQ_BASE: + status = ia64_sal_freq_base(in1, &r9, &r10); + break; + case SAL_PCI_CONFIG_READ: + printk(KERN_WARNING"kvm: Not allowed to call here!" + " SAL_PCI_CONFIG_READ\n"); + break; + case SAL_PCI_CONFIG_WRITE: + printk(KERN_WARNING"kvm: Not allowed to call here!" + " SAL_PCI_CONFIG_WRITE\n"); + break; + case SAL_SET_VECTORS: + if (in1 == SAL_VECTOR_OS_BOOT_RENDEZ) { + if (in4 != 0 || in5 != 0 || in6 != 0 || in7 != 0) { + status = -2; + } else { + kvm->arch.rdv_sal_data.boot_ip = in2; + kvm->arch.rdv_sal_data.boot_gp = in3; + } + printk("Rendvous called! iip:%lx\n\n", in2); + } else + printk(KERN_WARNING"kvm: CALLED SAL_SET_VECTORS %lu." + "ignored...\n", in1); + break; + case SAL_GET_STATE_INFO: + /* No more info. */ + status = -5; + r9 = 0; + break; + case SAL_GET_STATE_INFO_SIZE: + /* Return a dummy size. */ + status = 0; + r9 = 128; + break; + case SAL_CLEAR_STATE_INFO: + /* Noop. */ + break; + case SAL_MC_RENDEZ: + printk(KERN_WARNING + "kvm: called SAL_MC_RENDEZ. ignored...\n"); + break; + case SAL_MC_SET_PARAMS: + printk(KERN_WARNING + "kvm: called SAL_MC_SET_PARAMS.ignored!\n"); + break; + case SAL_CACHE_FLUSH: + if (1) { + /*Flush using SAL. + This method is faster but has a side + effect on other vcpu running on + this cpu. */ + status = ia64_sal_cache_flush(in1); + } else { + /*Maybe need to implement the method + without side effect!*/ + status = 0; + } + break; + case SAL_CACHE_INIT: + printk(KERN_WARNING + "kvm: called SAL_CACHE_INIT. ignored...\n"); + break; + case SAL_UPDATE_PAL: + printk(KERN_WARNING + "kvm: CALLED SAL_UPDATE_PAL. ignored...\n"); + break; + default: + printk(KERN_WARNING"kvm: called SAL_CALL with unknown index." + " index:%ld\n", index); + status = -1; + break; + } + return ((struct sal_ret_values) {status, r9, r10, r11}); +} + +static void kvm_get_sal_call_data(struct kvm_vcpu *vcpu, u64 *in0, u64 *in1, + u64 *in2, u64 *in3, u64 *in4, u64 *in5, u64 *in6, u64 *in7){ + + struct exit_ctl_data *p; + + p = kvm_get_exit_data(vcpu); + + if (p) { + if (p->exit_reason == EXIT_REASON_SAL_CALL) { + *in0 = p->u.sal_data.in0; + *in1 = p->u.sal_data.in1; + *in2 = p->u.sal_data.in2; + *in3 = p->u.sal_data.in3; + *in4 = p->u.sal_data.in4; + *in5 = p->u.sal_data.in5; + *in6 = p->u.sal_data.in6; + *in7 = p->u.sal_data.in7; + return ; + } + } + *in0 = 0; +} + +void kvm_sal_emul(struct kvm_vcpu *vcpu) +{ + + struct sal_ret_values result; + u64 index, in1, in2, in3, in4, in5, in6, in7; + + kvm_get_sal_call_data(vcpu, &index, &in1, &in2, + &in3, &in4, &in5, &in6, &in7); + result = sal_emulator(vcpu->kvm, index, in1, in2, in3, + in4, in5, in6, in7); + set_sal_result(vcpu, result); +} diff --git a/arch/ia64/kvm/kvm_minstate.h b/arch/ia64/kvm/kvm_minstate.h new file mode 100644 index 0000000000000000000000000000000000000000..13980d9b8bcf1542ea29756990e2ce4a3b941d9e --- /dev/null +++ b/arch/ia64/kvm/kvm_minstate.h @@ -0,0 +1,273 @@ +/* + * kvm_minstate.h: min save macros + * Copyright (c) 2007, Intel Corporation. + * + * Xuefei Xu (Anthony Xu) (Anthony.xu@intel.com) + * Xiantao Zhang (xiantao.zhang@intel.com) + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + + +#include +#include +#include +#include "asm-offsets.h" + +#define KVM_MINSTATE_START_SAVE_MIN \ + mov ar.rsc = 0;/* set enforced lazy mode, pl 0, little-endian, loadrs=0 */\ + ;; \ + mov.m r28 = ar.rnat; \ + addl r22 = VMM_RBS_OFFSET,r1; /* compute base of RBS */ \ + ;; \ + lfetch.fault.excl.nt1 [r22]; \ + addl r1 = IA64_STK_OFFSET-VMM_PT_REGS_SIZE,r1; /* compute base of memory stack */ \ + mov r23 = ar.bspstore; /* save ar.bspstore */ \ + ;; \ + mov ar.bspstore = r22; /* switch to kernel RBS */\ + ;; \ + mov r18 = ar.bsp; \ + mov ar.rsc = 0x3; /* set eager mode, pl 0, little-endian, loadrs=0 */ + + + +#define KVM_MINSTATE_END_SAVE_MIN \ + bsw.1; /* switch back to bank 1 (must be last in insn group) */\ + ;; + + +#define PAL_VSA_SYNC_READ \ + /* begin to call pal vps sync_read */ \ + add r25 = VMM_VPD_BASE_OFFSET, r21; \ + adds r20 = VMM_VCPU_VSA_BASE_OFFSET, r21; /* entry point */ \ + ;; \ + ld8 r25 = [r25]; /* read vpd base */ \ + ld8 r20 = [r20]; \ + ;; \ + add r20 = PAL_VPS_SYNC_READ,r20; \ + ;; \ +{ .mii; \ + nop 0x0; \ + mov r24 = ip; \ + mov b0 = r20; \ + ;; \ +}; \ +{ .mmb; \ + add r24 = 0x20, r24; \ + nop 0x0; \ + br.cond.sptk b0; /* call the service */ \ + ;; \ +}; + + + +#define KVM_MINSTATE_GET_CURRENT(reg) mov reg=r21 + +/* + * KVM_DO_SAVE_MIN switches to the kernel stacks (if necessary) and saves + * the minimum state necessary that allows us to turn psr.ic back + * on. + * + * Assumed state upon entry: + * psr.ic: off + * r31: contains saved predicates (pr) + * + * Upon exit, the state is as follows: + * psr.ic: off + * r2 = points to &pt_regs.r16 + * r8 = contents of ar.ccv + * r9 = contents of ar.csd + * r10 = contents of ar.ssd + * r11 = FPSR_DEFAULT + * r12 = kernel sp (kernel virtual address) + * r13 = points to current task_struct (kernel virtual address) + * p15 = TRUE if psr.i is set in cr.ipsr + * predicate registers (other than p2, p3, and p15), b6, r3, r14, r15: + * preserved + * + * Note that psr.ic is NOT turned on by this macro. This is so that + * we can pass interruption state as arguments to a handler. + */ + + +#define PT(f) (VMM_PT_REGS_##f##_OFFSET) + +#define KVM_DO_SAVE_MIN(COVER,SAVE_IFS,EXTRA) \ + KVM_MINSTATE_GET_CURRENT(r16); /* M (or M;;I) */ \ + mov r27 = ar.rsc; /* M */ \ + mov r20 = r1; /* A */ \ + mov r25 = ar.unat; /* M */ \ + mov r29 = cr.ipsr; /* M */ \ + mov r26 = ar.pfs; /* I */ \ + mov r18 = cr.isr; \ + COVER; /* B;; (or nothing) */ \ + ;; \ + tbit.z p0,p15 = r29,IA64_PSR_I_BIT; \ + mov r1 = r16; \ +/* mov r21=r16; */ \ + /* switch from user to kernel RBS: */ \ + ;; \ + invala; /* M */ \ + SAVE_IFS; \ + ;; \ + KVM_MINSTATE_START_SAVE_MIN \ + adds r17 = 2*L1_CACHE_BYTES,r1;/* cache-line size */ \ + adds r16 = PT(CR_IPSR),r1; \ + ;; \ + lfetch.fault.excl.nt1 [r17],L1_CACHE_BYTES; \ + st8 [r16] = r29; /* save cr.ipsr */ \ + ;; \ + lfetch.fault.excl.nt1 [r17]; \ + tbit.nz p15,p0 = r29,IA64_PSR_I_BIT; \ + mov r29 = b0 \ + ;; \ + adds r16 = PT(R8),r1; /* initialize first base pointer */\ + adds r17 = PT(R9),r1; /* initialize second base pointer */\ + ;; \ +.mem.offset 0,0; st8.spill [r16] = r8,16; \ +.mem.offset 8,0; st8.spill [r17] = r9,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r16] = r10,24; \ +.mem.offset 8,0; st8.spill [r17] = r11,24; \ + ;; \ + mov r9 = cr.iip; /* M */ \ + mov r10 = ar.fpsr; /* M */ \ + ;; \ + st8 [r16] = r9,16; /* save cr.iip */ \ + st8 [r17] = r30,16; /* save cr.ifs */ \ + sub r18 = r18,r22; /* r18=RSE.ndirty*8 */ \ + ;; \ + st8 [r16] = r25,16; /* save ar.unat */ \ + st8 [r17] = r26,16; /* save ar.pfs */ \ + shl r18 = r18,16; /* calu ar.rsc used for "loadrs" */\ + ;; \ + st8 [r16] = r27,16; /* save ar.rsc */ \ + st8 [r17] = r28,16; /* save ar.rnat */ \ + ;; /* avoid RAW on r16 & r17 */ \ + st8 [r16] = r23,16; /* save ar.bspstore */ \ + st8 [r17] = r31,16; /* save predicates */ \ + ;; \ + st8 [r16] = r29,16; /* save b0 */ \ + st8 [r17] = r18,16; /* save ar.rsc value for "loadrs" */\ + ;; \ +.mem.offset 0,0; st8.spill [r16] = r20,16;/* save original r1 */ \ +.mem.offset 8,0; st8.spill [r17] = r12,16; \ + adds r12 = -16,r1; /* switch to kernel memory stack */ \ + ;; \ +.mem.offset 0,0; st8.spill [r16] = r13,16; \ +.mem.offset 8,0; st8.spill [r17] = r10,16; /* save ar.fpsr */\ + mov r13 = r21; /* establish `current' */ \ + ;; \ +.mem.offset 0,0; st8.spill [r16] = r15,16; \ +.mem.offset 8,0; st8.spill [r17] = r14,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r16] = r2,16; \ +.mem.offset 8,0; st8.spill [r17] = r3,16; \ + adds r2 = VMM_PT_REGS_R16_OFFSET,r1; \ + ;; \ + adds r16 = VMM_VCPU_IIPA_OFFSET,r13; \ + adds r17 = VMM_VCPU_ISR_OFFSET,r13; \ + mov r26 = cr.iipa; \ + mov r27 = cr.isr; \ + ;; \ + st8 [r16] = r26; \ + st8 [r17] = r27; \ + ;; \ + EXTRA; \ + mov r8 = ar.ccv; \ + mov r9 = ar.csd; \ + mov r10 = ar.ssd; \ + movl r11 = FPSR_DEFAULT; /* L-unit */ \ + adds r17 = VMM_VCPU_GP_OFFSET,r13; \ + ;; \ + ld8 r1 = [r17];/* establish kernel global pointer */ \ + ;; \ + PAL_VSA_SYNC_READ \ + KVM_MINSTATE_END_SAVE_MIN + +/* + * SAVE_REST saves the remainder of pt_regs (with psr.ic on). + * + * Assumed state upon entry: + * psr.ic: on + * r2: points to &pt_regs.f6 + * r3: points to &pt_regs.f7 + * r8: contents of ar.ccv + * r9: contents of ar.csd + * r10: contents of ar.ssd + * r11: FPSR_DEFAULT + * + * Registers r14 and r15 are guaranteed not to be touched by SAVE_REST. + */ +#define KVM_SAVE_REST \ +.mem.offset 0,0; st8.spill [r2] = r16,16; \ +.mem.offset 8,0; st8.spill [r3] = r17,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r18,16; \ +.mem.offset 8,0; st8.spill [r3] = r19,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r20,16; \ +.mem.offset 8,0; st8.spill [r3] = r21,16; \ + mov r18=b6; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r22,16; \ +.mem.offset 8,0; st8.spill [r3] = r23,16; \ + mov r19 = b7; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r24,16; \ +.mem.offset 8,0; st8.spill [r3] = r25,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r26,16; \ +.mem.offset 8,0; st8.spill [r3] = r27,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r28,16; \ +.mem.offset 8,0; st8.spill [r3] = r29,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r30,16; \ +.mem.offset 8,0; st8.spill [r3] = r31,32; \ + ;; \ + mov ar.fpsr = r11; \ + st8 [r2] = r8,8; \ + adds r24 = PT(B6)-PT(F7),r3; \ + adds r25 = PT(B7)-PT(F7),r3; \ + ;; \ + st8 [r24] = r18,16; /* b6 */ \ + st8 [r25] = r19,16; /* b7 */ \ + adds r2 = PT(R4)-PT(F6),r2; \ + adds r3 = PT(R5)-PT(F7),r3; \ + ;; \ + st8 [r24] = r9; /* ar.csd */ \ + st8 [r25] = r10; /* ar.ssd */ \ + ;; \ + mov r18 = ar.unat; \ + adds r19 = PT(EML_UNAT)-PT(R4),r2; \ + ;; \ + st8 [r19] = r18; /* eml_unat */ \ + + +#define KVM_SAVE_EXTRA \ +.mem.offset 0,0; st8.spill [r2] = r4,16; \ +.mem.offset 8,0; st8.spill [r3] = r5,16; \ + ;; \ +.mem.offset 0,0; st8.spill [r2] = r6,16; \ +.mem.offset 8,0; st8.spill [r3] = r7; \ + ;; \ + mov r26 = ar.unat; \ + ;; \ + st8 [r2] = r26;/* eml_unat */ \ + +#define KVM_SAVE_MIN_WITH_COVER KVM_DO_SAVE_MIN(cover, mov r30 = cr.ifs,) +#define KVM_SAVE_MIN_WITH_COVER_R19 KVM_DO_SAVE_MIN(cover, mov r30 = cr.ifs, mov r15 = r19) +#define KVM_SAVE_MIN KVM_DO_SAVE_MIN( , mov r30 = r0, ) diff --git a/arch/ia64/kvm/lapic.h b/arch/ia64/kvm/lapic.h new file mode 100644 index 0000000000000000000000000000000000000000..6d6cbcb14893c281952f82365779cf925cf7f0e0 --- /dev/null +++ b/arch/ia64/kvm/lapic.h @@ -0,0 +1,25 @@ +#ifndef __KVM_IA64_LAPIC_H +#define __KVM_IA64_LAPIC_H + +#include + +/* + * vlsapic + */ +struct kvm_lapic{ + struct kvm_vcpu *vcpu; + uint64_t insvc[4]; + uint64_t vhpi; + uint8_t xtp; + uint8_t pal_init_pending; + uint8_t pad[2]; +}; + +int kvm_create_lapic(struct kvm_vcpu *vcpu); +void kvm_free_lapic(struct kvm_vcpu *vcpu); + +int kvm_apic_match_physical_addr(struct kvm_lapic *apic, u16 dest); +int kvm_apic_match_logical_addr(struct kvm_lapic *apic, u8 mda); +int kvm_apic_set_irq(struct kvm_vcpu *vcpu, u8 vec, u8 trig); + +#endif diff --git a/arch/ia64/kvm/misc.h b/arch/ia64/kvm/misc.h new file mode 100644 index 0000000000000000000000000000000000000000..e585c46073443de4cdd4687301876255e7fab87a --- /dev/null +++ b/arch/ia64/kvm/misc.h @@ -0,0 +1,93 @@ +#ifndef __KVM_IA64_MISC_H +#define __KVM_IA64_MISC_H + +#include +/* + * misc.h + * Copyright (C) 2007, Intel Corporation. + * Xiantao Zhang (xiantao.zhang@intel.com) + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +/* + *Return p2m base address at host side! + */ +static inline uint64_t *kvm_host_get_pmt(struct kvm *kvm) +{ + return (uint64_t *)(kvm->arch.vm_base + KVM_P2M_OFS); +} + +static inline void kvm_set_pmt_entry(struct kvm *kvm, gfn_t gfn, + u64 paddr, u64 mem_flags) +{ + uint64_t *pmt_base = kvm_host_get_pmt(kvm); + unsigned long pte; + + pte = PAGE_ALIGN(paddr) | mem_flags; + pmt_base[gfn] = pte; +} + +/*Function for translating host address to guest address*/ + +static inline void *to_guest(struct kvm *kvm, void *addr) +{ + return (void *)((unsigned long)(addr) - kvm->arch.vm_base + + KVM_VM_DATA_BASE); +} + +/*Function for translating guest address to host address*/ + +static inline void *to_host(struct kvm *kvm, void *addr) +{ + return (void *)((unsigned long)addr - KVM_VM_DATA_BASE + + kvm->arch.vm_base); +} + +/* Get host context of the vcpu */ +static inline union context *kvm_get_host_context(struct kvm_vcpu *vcpu) +{ + union context *ctx = &vcpu->arch.host; + return to_guest(vcpu->kvm, ctx); +} + +/* Get guest context of the vcpu */ +static inline union context *kvm_get_guest_context(struct kvm_vcpu *vcpu) +{ + union context *ctx = &vcpu->arch.guest; + return to_guest(vcpu->kvm, ctx); +} + +/* kvm get exit data from gvmm! */ +static inline struct exit_ctl_data *kvm_get_exit_data(struct kvm_vcpu *vcpu) +{ + return &vcpu->arch.exit_data; +} + +/*kvm get vcpu ioreq for kvm module!*/ +static inline struct kvm_mmio_req *kvm_get_vcpu_ioreq(struct kvm_vcpu *vcpu) +{ + struct exit_ctl_data *p_ctl_data; + + if (vcpu) { + p_ctl_data = kvm_get_exit_data(vcpu); + if (p_ctl_data->exit_reason == EXIT_REASON_MMIO_INSTRUCTION) + return &p_ctl_data->u.ioreq; + } + + return NULL; +} + +#endif diff --git a/arch/ia64/kvm/mmio.c b/arch/ia64/kvm/mmio.c new file mode 100644 index 0000000000000000000000000000000000000000..351bf70da46333d5c8eb67b28a3a976680658c4a --- /dev/null +++ b/arch/ia64/kvm/mmio.c @@ -0,0 +1,341 @@ +/* + * mmio.c: MMIO emulation components. + * Copyright (c) 2004, Intel Corporation. + * Yaozu Dong (Eddie Dong) (Eddie.dong@intel.com) + * Kun Tian (Kevin Tian) (Kevin.tian@intel.com) + * + * Copyright (c) 2007 Intel Corporation KVM support. + * Xuefei Xu (Anthony Xu) (anthony.xu@intel.com) + * Xiantao Zhang (xiantao.zhang@intel.com) + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +#include + +#include "vcpu.h" + +static void vlsapic_write_xtp(struct kvm_vcpu *v, uint8_t val) +{ + VLSAPIC_XTP(v) = val; +} + +/* + * LSAPIC OFFSET + */ +#define PIB_LOW_HALF(ofst) !(ofst & (1 << 20)) +#define PIB_OFST_INTA 0x1E0000 +#define PIB_OFST_XTP 0x1E0008 + +/* + * execute write IPI op. + */ +static void vlsapic_write_ipi(struct kvm_vcpu *vcpu, + uint64_t addr, uint64_t data) +{ + struct exit_ctl_data *p = ¤t_vcpu->arch.exit_data; + unsigned long psr; + + local_irq_save(psr); + + p->exit_reason = EXIT_REASON_IPI; + p->u.ipi_data.addr.val = addr; + p->u.ipi_data.data.val = data; + vmm_transition(current_vcpu); + + local_irq_restore(psr); + +} + +void lsapic_write(struct kvm_vcpu *v, unsigned long addr, + unsigned long length, unsigned long val) +{ + addr &= (PIB_SIZE - 1); + + switch (addr) { + case PIB_OFST_INTA: + /*panic_domain(NULL, "Undefined write on PIB INTA\n");*/ + panic_vm(v); + break; + case PIB_OFST_XTP: + if (length == 1) { + vlsapic_write_xtp(v, val); + } else { + /*panic_domain(NULL, + "Undefined write on PIB XTP\n");*/ + panic_vm(v); + } + break; + default: + if (PIB_LOW_HALF(addr)) { + /*lower half */ + if (length != 8) + /*panic_domain(NULL, + "Can't LHF write with size %ld!\n", + length);*/ + panic_vm(v); + else + vlsapic_write_ipi(v, addr, val); + } else { /* upper half + printk("IPI-UHF write %lx\n",addr);*/ + panic_vm(v); + } + break; + } +} + +unsigned long lsapic_read(struct kvm_vcpu *v, unsigned long addr, + unsigned long length) +{ + uint64_t result = 0; + + addr &= (PIB_SIZE - 1); + + switch (addr) { + case PIB_OFST_INTA: + if (length == 1) /* 1 byte load */ + ; /* There is no i8259, there is no INTA access*/ + else + /*panic_domain(NULL,"Undefined read on PIB INTA\n"); */ + panic_vm(v); + + break; + case PIB_OFST_XTP: + if (length == 1) { + result = VLSAPIC_XTP(v); + /* printk("read xtp %lx\n", result); */ + } else { + /*panic_domain(NULL, + "Undefined read on PIB XTP\n");*/ + panic_vm(v); + } + break; + default: + panic_vm(v); + break; + } + return result; +} + +static void mmio_access(struct kvm_vcpu *vcpu, u64 src_pa, u64 *dest, + u16 s, int ma, int dir) +{ + unsigned long iot; + struct exit_ctl_data *p = &vcpu->arch.exit_data; + unsigned long psr; + + iot = __gpfn_is_io(src_pa >> PAGE_SHIFT); + + local_irq_save(psr); + + /*Intercept the acces for PIB range*/ + if (iot == GPFN_PIB) { + if (!dir) + lsapic_write(vcpu, src_pa, s, *dest); + else + *dest = lsapic_read(vcpu, src_pa, s); + goto out; + } + p->exit_reason = EXIT_REASON_MMIO_INSTRUCTION; + p->u.ioreq.addr = src_pa; + p->u.ioreq.size = s; + p->u.ioreq.dir = dir; + if (dir == IOREQ_WRITE) + p->u.ioreq.data = *dest; + p->u.ioreq.state = STATE_IOREQ_READY; + vmm_transition(vcpu); + + if (p->u.ioreq.state == STATE_IORESP_READY) { + if (dir == IOREQ_READ) + *dest = p->u.ioreq.data; + } else + panic_vm(vcpu); +out: + local_irq_restore(psr); + return ; +} + +/* + dir 1: read 0:write + inst_type 0:integer 1:floating point + */ +#define SL_INTEGER 0 /* store/load interger*/ +#define SL_FLOATING 1 /* store/load floating*/ + +void emulate_io_inst(struct kvm_vcpu *vcpu, u64 padr, u64 ma) +{ + struct kvm_pt_regs *regs; + IA64_BUNDLE bundle; + int slot, dir = 0; + int inst_type = -1; + u16 size = 0; + u64 data, slot1a, slot1b, temp, update_reg; + s32 imm; + INST64 inst; + + regs = vcpu_regs(vcpu); + + if (fetch_code(vcpu, regs->cr_iip, &bundle)) { + /* if fetch code fail, return and try again */ + return; + } + slot = ((struct ia64_psr *)&(regs->cr_ipsr))->ri; + if (!slot) + inst.inst = bundle.slot0; + else if (slot == 1) { + slot1a = bundle.slot1a; + slot1b = bundle.slot1b; + inst.inst = slot1a + (slot1b << 18); + } else if (slot == 2) + inst.inst = bundle.slot2; + + /* Integer Load/Store */ + if (inst.M1.major == 4 && inst.M1.m == 0 && inst.M1.x == 0) { + inst_type = SL_INTEGER; + size = (inst.M1.x6 & 0x3); + if ((inst.M1.x6 >> 2) > 0xb) { + /*write*/ + dir = IOREQ_WRITE; + data = vcpu_get_gr(vcpu, inst.M4.r2); + } else if ((inst.M1.x6 >> 2) < 0xb) { + /*read*/ + dir = IOREQ_READ; + } + } else if (inst.M2.major == 4 && inst.M2.m == 1 && inst.M2.x == 0) { + /* Integer Load + Reg update */ + inst_type = SL_INTEGER; + dir = IOREQ_READ; + size = (inst.M2.x6 & 0x3); + temp = vcpu_get_gr(vcpu, inst.M2.r3); + update_reg = vcpu_get_gr(vcpu, inst.M2.r2); + temp += update_reg; + vcpu_set_gr(vcpu, inst.M2.r3, temp, 0); + } else if (inst.M3.major == 5) { + /*Integer Load/Store + Imm update*/ + inst_type = SL_INTEGER; + size = (inst.M3.x6&0x3); + if ((inst.M5.x6 >> 2) > 0xb) { + /*write*/ + dir = IOREQ_WRITE; + data = vcpu_get_gr(vcpu, inst.M5.r2); + temp = vcpu_get_gr(vcpu, inst.M5.r3); + imm = (inst.M5.s << 31) | (inst.M5.i << 30) | + (inst.M5.imm7 << 23); + temp += imm >> 23; + vcpu_set_gr(vcpu, inst.M5.r3, temp, 0); + + } else if ((inst.M3.x6 >> 2) < 0xb) { + /*read*/ + dir = IOREQ_READ; + temp = vcpu_get_gr(vcpu, inst.M3.r3); + imm = (inst.M3.s << 31) | (inst.M3.i << 30) | + (inst.M3.imm7 << 23); + temp += imm >> 23; + vcpu_set_gr(vcpu, inst.M3.r3, temp, 0); + + } + } else if (inst.M9.major == 6 && inst.M9.x6 == 0x3B + && inst.M9.m == 0 && inst.M9.x == 0) { + /* Floating-point spill*/ + struct ia64_fpreg v; + + inst_type = SL_FLOATING; + dir = IOREQ_WRITE; + vcpu_get_fpreg(vcpu, inst.M9.f2, &v); + /* Write high word. FIXME: this is a kludge! */ + v.u.bits[1] &= 0x3ffff; + mmio_access(vcpu, padr + 8, &v.u.bits[1], 8, ma, IOREQ_WRITE); + data = v.u.bits[0]; + size = 3; + } else if (inst.M10.major == 7 && inst.M10.x6 == 0x3B) { + /* Floating-point spill + Imm update */ + struct ia64_fpreg v; + + inst_type = SL_FLOATING; + dir = IOREQ_WRITE; + vcpu_get_fpreg(vcpu, inst.M10.f2, &v); + temp = vcpu_get_gr(vcpu, inst.M10.r3); + imm = (inst.M10.s << 31) | (inst.M10.i << 30) | + (inst.M10.imm7 << 23); + temp += imm >> 23; + vcpu_set_gr(vcpu, inst.M10.r3, temp, 0); + + /* Write high word.FIXME: this is a kludge! */ + v.u.bits[1] &= 0x3ffff; + mmio_access(vcpu, padr + 8, &v.u.bits[1], 8, ma, IOREQ_WRITE); + data = v.u.bits[0]; + size = 3; + } else if (inst.M10.major == 7 && inst.M10.x6 == 0x31) { + /* Floating-point stf8 + Imm update */ + struct ia64_fpreg v; + inst_type = SL_FLOATING; + dir = IOREQ_WRITE; + size = 3; + vcpu_get_fpreg(vcpu, inst.M10.f2, &v); + data = v.u.bits[0]; /* Significand. */ + temp = vcpu_get_gr(vcpu, inst.M10.r3); + imm = (inst.M10.s << 31) | (inst.M10.i << 30) | + (inst.M10.imm7 << 23); + temp += imm >> 23; + vcpu_set_gr(vcpu, inst.M10.r3, temp, 0); + } else if (inst.M15.major == 7 && inst.M15.x6 >= 0x2c + && inst.M15.x6 <= 0x2f) { + temp = vcpu_get_gr(vcpu, inst.M15.r3); + imm = (inst.M15.s << 31) | (inst.M15.i << 30) | + (inst.M15.imm7 << 23); + temp += imm >> 23; + vcpu_set_gr(vcpu, inst.M15.r3, temp, 0); + + vcpu_increment_iip(vcpu); + return; + } else if (inst.M12.major == 6 && inst.M12.m == 1 + && inst.M12.x == 1 && inst.M12.x6 == 1) { + /* Floating-point Load Pair + Imm ldfp8 M12*/ + struct ia64_fpreg v; + + inst_type = SL_FLOATING; + dir = IOREQ_READ; + size = 8; /*ldfd*/ + mmio_access(vcpu, padr, &data, size, ma, dir); + v.u.bits[0] = data; + v.u.bits[1] = 0x1003E; + vcpu_set_fpreg(vcpu, inst.M12.f1, &v); + padr += 8; + mmio_access(vcpu, padr, &data, size, ma, dir); + v.u.bits[0] = data; + v.u.bits[1] = 0x1003E; + vcpu_set_fpreg(vcpu, inst.M12.f2, &v); + padr += 8; + vcpu_set_gr(vcpu, inst.M12.r3, padr, 0); + vcpu_increment_iip(vcpu); + return; + } else { + inst_type = -1; + panic_vm(vcpu); + } + + size = 1 << size; + if (dir == IOREQ_WRITE) { + mmio_access(vcpu, padr, &data, size, ma, dir); + } else { + mmio_access(vcpu, padr, &data, size, ma, dir); + if (inst_type == SL_INTEGER) + vcpu_set_gr(vcpu, inst.M1.r1, data, 0); + else + panic_vm(vcpu); + + } + vcpu_increment_iip(vcpu); +} diff --git a/arch/ia64/kvm/optvfault.S b/arch/ia64/kvm/optvfault.S new file mode 100644 index 0000000000000000000000000000000000000000..e4f15d641b22dfbc5c21ea85f4c2e2db1d54a36b --- /dev/null +++ b/arch/ia64/kvm/optvfault.S @@ -0,0 +1,918 @@ +/* + * arch/ia64/vmx/optvfault.S + * optimize virtualization fault handler + * + * Copyright (C) 2006 Intel Co + * Xuefei Xu (Anthony Xu) + */ + +#include +#include + +#include "vti.h" +#include "asm-offsets.h" + +#define ACCE_MOV_FROM_AR +#define ACCE_MOV_FROM_RR +#define ACCE_MOV_TO_RR +#define ACCE_RSM +#define ACCE_SSM +#define ACCE_MOV_TO_PSR +#define ACCE_THASH + +//mov r1=ar3 +GLOBAL_ENTRY(kvm_asm_mov_from_ar) +#ifndef ACCE_MOV_FROM_AR + br.many kvm_virtualization_fault_back +#endif + add r18=VMM_VCPU_ITC_OFS_OFFSET, r21 + add r16=VMM_VCPU_LAST_ITC_OFFSET,r21 + extr.u r17=r25,6,7 + ;; + ld8 r18=[r18] + mov r19=ar.itc + mov r24=b0 + ;; + add r19=r19,r18 + addl r20=@gprel(asm_mov_to_reg),gp + ;; + st8 [r16] = r19 + adds r30=kvm_resume_to_guest-asm_mov_to_reg,r20 + shladd r17=r17,4,r20 + ;; + mov b0=r17 + br.sptk.few b0 + ;; +END(kvm_asm_mov_from_ar) + + +// mov r1=rr[r3] +GLOBAL_ENTRY(kvm_asm_mov_from_rr) +#ifndef ACCE_MOV_FROM_RR + br.many kvm_virtualization_fault_back +#endif + extr.u r16=r25,20,7 + extr.u r17=r25,6,7 + addl r20=@gprel(asm_mov_from_reg),gp + ;; + adds r30=kvm_asm_mov_from_rr_back_1-asm_mov_from_reg,r20 + shladd r16=r16,4,r20 + mov r24=b0 + ;; + add r27=VMM_VCPU_VRR0_OFFSET,r21 + mov b0=r16 + br.many b0 + ;; +kvm_asm_mov_from_rr_back_1: + adds r30=kvm_resume_to_guest-asm_mov_from_reg,r20 + adds r22=asm_mov_to_reg-asm_mov_from_reg,r20 + shr.u r26=r19,61 + ;; + shladd r17=r17,4,r22 + shladd r27=r26,3,r27 + ;; + ld8 r19=[r27] + mov b0=r17 + br.many b0 +END(kvm_asm_mov_from_rr) + + +// mov rr[r3]=r2 +GLOBAL_ENTRY(kvm_asm_mov_to_rr) +#ifndef ACCE_MOV_TO_RR + br.many kvm_virtualization_fault_back +#endif + extr.u r16=r25,20,7 + extr.u r17=r25,13,7 + addl r20=@gprel(asm_mov_from_reg),gp + ;; + adds r30=kvm_asm_mov_to_rr_back_1-asm_mov_from_reg,r20 + shladd r16=r16,4,r20 + mov r22=b0 + ;; + add r27=VMM_VCPU_VRR0_OFFSET,r21 + mov b0=r16 + br.many b0 + ;; +kvm_asm_mov_to_rr_back_1: + adds r30=kvm_asm_mov_to_rr_back_2-asm_mov_from_reg,r20 + shr.u r23=r19,61 + shladd r17=r17,4,r20 + ;; + //if rr6, go back + cmp.eq p6,p0=6,r23 + mov b0=r22 + (p6) br.cond.dpnt.many kvm_virtualization_fault_back + ;; + mov r28=r19 + mov b0=r17 + br.many b0 +kvm_asm_mov_to_rr_back_2: + adds r30=kvm_resume_to_guest-asm_mov_from_reg,r20 + shladd r27=r23,3,r27 + ;; // vrr.rid<<4 |0xe + st8 [r27]=r19 + mov b0=r30 + ;; + extr.u r16=r19,8,26 + extr.u r18 =r19,2,6 + mov r17 =0xe + ;; + shladd r16 = r16, 4, r17 + extr.u r19 =r19,0,8 + ;; + shl r16 = r16,8 + ;; + add r19 = r19, r16 + ;; //set ve 1 + dep r19=-1,r19,0,1 + cmp.lt p6,p0=14,r18 + ;; + (p6) mov r18=14 + ;; + (p6) dep r19=r18,r19,2,6 + ;; + cmp.eq p6,p0=0,r23 + ;; + cmp.eq.or p6,p0=4,r23 + ;; + adds r16=VMM_VCPU_MODE_FLAGS_OFFSET,r21 + (p6) adds r17=VMM_VCPU_META_SAVED_RR0_OFFSET,r21 + ;; + ld4 r16=[r16] + cmp.eq p7,p0=r0,r0 + (p6) shladd r17=r23,1,r17 + ;; + (p6) st8 [r17]=r19 + (p6) tbit.nz p6,p7=r16,0 + ;; + (p7) mov rr[r28]=r19 + mov r24=r22 + br.many b0 +END(kvm_asm_mov_to_rr) + + +//rsm +GLOBAL_ENTRY(kvm_asm_rsm) +#ifndef ACCE_RSM + br.many kvm_virtualization_fault_back +#endif + add r16=VMM_VPD_BASE_OFFSET,r21 + extr.u r26=r25,6,21 + extr.u r27=r25,31,2 + ;; + ld8 r16=[r16] + extr.u r28=r25,36,1 + dep r26=r27,r26,21,2 + ;; + add r17=VPD_VPSR_START_OFFSET,r16 + add r22=VMM_VCPU_MODE_FLAGS_OFFSET,r21 + //r26 is imm24 + dep r26=r28,r26,23,1 + ;; + ld8 r18=[r17] + movl r28=IA64_PSR_IC+IA64_PSR_I+IA64_PSR_DT+IA64_PSR_SI + ld4 r23=[r22] + sub r27=-1,r26 + mov r24=b0 + ;; + mov r20=cr.ipsr + or r28=r27,r28 + and r19=r18,r27 + ;; + st8 [r17]=r19 + and r20=r20,r28 + /* Comment it out due to short of fp lazy alorgithm support + adds r27=IA64_VCPU_FP_PSR_OFFSET,r21 + ;; + ld8 r27=[r27] + ;; + tbit.nz p8,p0= r27,IA64_PSR_DFH_BIT + ;; + (p8) dep r20=-1,r20,IA64_PSR_DFH_BIT,1 + */ + ;; + mov cr.ipsr=r20 + tbit.nz p6,p0=r23,0 + ;; + tbit.z.or p6,p0=r26,IA64_PSR_DT_BIT + (p6) br.dptk kvm_resume_to_guest + ;; + add r26=VMM_VCPU_META_RR0_OFFSET,r21 + add r27=VMM_VCPU_META_RR0_OFFSET+8,r21 + dep r23=-1,r23,0,1 + ;; + ld8 r26=[r26] + ld8 r27=[r27] + st4 [r22]=r23 + dep.z r28=4,61,3 + ;; + mov rr[r0]=r26 + ;; + mov rr[r28]=r27 + ;; + srlz.d + br.many kvm_resume_to_guest +END(kvm_asm_rsm) + + +//ssm +GLOBAL_ENTRY(kvm_asm_ssm) +#ifndef ACCE_SSM + br.many kvm_virtualization_fault_back +#endif + add r16=VMM_VPD_BASE_OFFSET,r21 + extr.u r26=r25,6,21 + extr.u r27=r25,31,2 + ;; + ld8 r16=[r16] + extr.u r28=r25,36,1 + dep r26=r27,r26,21,2 + ;; //r26 is imm24 + add r27=VPD_VPSR_START_OFFSET,r16 + dep r26=r28,r26,23,1 + ;; //r19 vpsr + ld8 r29=[r27] + mov r24=b0 + ;; + add r22=VMM_VCPU_MODE_FLAGS_OFFSET,r21 + mov r20=cr.ipsr + or r19=r29,r26 + ;; + ld4 r23=[r22] + st8 [r27]=r19 + or r20=r20,r26 + ;; + mov cr.ipsr=r20 + movl r28=IA64_PSR_DT+IA64_PSR_RT+IA64_PSR_IT + ;; + and r19=r28,r19 + tbit.z p6,p0=r23,0 + ;; + cmp.ne.or p6,p0=r28,r19 + (p6) br.dptk kvm_asm_ssm_1 + ;; + add r26=VMM_VCPU_META_SAVED_RR0_OFFSET,r21 + add r27=VMM_VCPU_META_SAVED_RR0_OFFSET+8,r21 + dep r23=0,r23,0,1 + ;; + ld8 r26=[r26] + ld8 r27=[r27] + st4 [r22]=r23 + dep.z r28=4,61,3 + ;; + mov rr[r0]=r26 + ;; + mov rr[r28]=r27 + ;; + srlz.d + ;; +kvm_asm_ssm_1: + tbit.nz p6,p0=r29,IA64_PSR_I_BIT + ;; + tbit.z.or p6,p0=r19,IA64_PSR_I_BIT + (p6) br.dptk kvm_resume_to_guest + ;; + add r29=VPD_VTPR_START_OFFSET,r16 + add r30=VPD_VHPI_START_OFFSET,r16 + ;; + ld8 r29=[r29] + ld8 r30=[r30] + ;; + extr.u r17=r29,4,4 + extr.u r18=r29,16,1 + ;; + dep r17=r18,r17,4,1 + ;; + cmp.gt p6,p0=r30,r17 + (p6) br.dpnt.few kvm_asm_dispatch_vexirq + br.many kvm_resume_to_guest +END(kvm_asm_ssm) + + +//mov psr.l=r2 +GLOBAL_ENTRY(kvm_asm_mov_to_psr) +#ifndef ACCE_MOV_TO_PSR + br.many kvm_virtualization_fault_back +#endif + add r16=VMM_VPD_BASE_OFFSET,r21 + extr.u r26=r25,13,7 //r2 + ;; + ld8 r16=[r16] + addl r20=@gprel(asm_mov_from_reg),gp + ;; + adds r30=kvm_asm_mov_to_psr_back-asm_mov_from_reg,r20 + shladd r26=r26,4,r20 + mov r24=b0 + ;; + add r27=VPD_VPSR_START_OFFSET,r16 + mov b0=r26 + br.many b0 + ;; +kvm_asm_mov_to_psr_back: + ld8 r17=[r27] + add r22=VMM_VCPU_MODE_FLAGS_OFFSET,r21 + dep r19=0,r19,32,32 + ;; + ld4 r23=[r22] + dep r18=0,r17,0,32 + ;; + add r30=r18,r19 + movl r28=IA64_PSR_DT+IA64_PSR_RT+IA64_PSR_IT + ;; + st8 [r27]=r30 + and r27=r28,r30 + and r29=r28,r17 + ;; + cmp.eq p5,p0=r29,r27 + cmp.eq p6,p7=r28,r27 + (p5) br.many kvm_asm_mov_to_psr_1 + ;; + //virtual to physical + (p7) add r26=VMM_VCPU_META_RR0_OFFSET,r21 + (p7) add r27=VMM_VCPU_META_RR0_OFFSET+8,r21 + (p7) dep r23=-1,r23,0,1 + ;; + //physical to virtual + (p6) add r26=VMM_VCPU_META_SAVED_RR0_OFFSET,r21 + (p6) add r27=VMM_VCPU_META_SAVED_RR0_OFFSET+8,r21 + (p6) dep r23=0,r23,0,1 + ;; + ld8 r26=[r26] + ld8 r27=[r27] + st4 [r22]=r23 + dep.z r28=4,61,3 + ;; + mov rr[r0]=r26 + ;; + mov rr[r28]=r27 + ;; + srlz.d + ;; +kvm_asm_mov_to_psr_1: + mov r20=cr.ipsr + movl r28=IA64_PSR_IC+IA64_PSR_I+IA64_PSR_DT+IA64_PSR_SI+IA64_PSR_RT + ;; + or r19=r19,r28 + dep r20=0,r20,0,32 + ;; + add r20=r19,r20 + mov b0=r24 + ;; + /* Comment it out due to short of fp lazy algorithm support + adds r27=IA64_VCPU_FP_PSR_OFFSET,r21 + ;; + ld8 r27=[r27] + ;; + tbit.nz p8,p0=r27,IA64_PSR_DFH_BIT + ;; + (p8) dep r20=-1,r20,IA64_PSR_DFH_BIT,1 + ;; + */ + mov cr.ipsr=r20 + cmp.ne p6,p0=r0,r0 + ;; + tbit.nz.or p6,p0=r17,IA64_PSR_I_BIT + tbit.z.or p6,p0=r30,IA64_PSR_I_BIT + (p6) br.dpnt.few kvm_resume_to_guest + ;; + add r29=VPD_VTPR_START_OFFSET,r16 + add r30=VPD_VHPI_START_OFFSET,r16 + ;; + ld8 r29=[r29] + ld8 r30=[r30] + ;; + extr.u r17=r29,4,4 + extr.u r18=r29,16,1 + ;; + dep r17=r18,r17,4,1 + ;; + cmp.gt p6,p0=r30,r17 + (p6) br.dpnt.few kvm_asm_dispatch_vexirq + br.many kvm_resume_to_guest +END(kvm_asm_mov_to_psr) + + +ENTRY(kvm_asm_dispatch_vexirq) +//increment iip + mov r16=cr.ipsr + ;; + extr.u r17=r16,IA64_PSR_RI_BIT,2 + tbit.nz p6,p7=r16,IA64_PSR_RI_BIT+1 + ;; + (p6) mov r18=cr.iip + (p6) mov r17=r0 + (p7) add r17=1,r17 + ;; + (p6) add r18=0x10,r18 + dep r16=r17,r16,IA64_PSR_RI_BIT,2 + ;; + (p6) mov cr.iip=r18 + mov cr.ipsr=r16 + mov r30 =1 + br.many kvm_dispatch_vexirq +END(kvm_asm_dispatch_vexirq) + +// thash +// TODO: add support when pta.vf = 1 +GLOBAL_ENTRY(kvm_asm_thash) +#ifndef ACCE_THASH + br.many kvm_virtualization_fault_back +#endif + extr.u r17=r25,20,7 // get r3 from opcode in r25 + extr.u r18=r25,6,7 // get r1 from opcode in r25 + addl r20=@gprel(asm_mov_from_reg),gp + ;; + adds r30=kvm_asm_thash_back1-asm_mov_from_reg,r20 + shladd r17=r17,4,r20 // get addr of MOVE_FROM_REG(r17) + adds r16=VMM_VPD_BASE_OFFSET,r21 // get vcpu.arch.priveregs + ;; + mov r24=b0 + ;; + ld8 r16=[r16] // get VPD addr + mov b0=r17 + br.many b0 // r19 return value + ;; +kvm_asm_thash_back1: + shr.u r23=r19,61 // get RR number + adds r25=VMM_VCPU_VRR0_OFFSET,r21 // get vcpu->arch.vrr[0]'s addr + adds r16=VMM_VPD_VPTA_OFFSET,r16 // get vpta + ;; + shladd r27=r23,3,r25 // get vcpu->arch.vrr[r23]'s addr + ld8 r17=[r16] // get PTA + mov r26=1 + ;; + extr.u r29=r17,2,6 // get pta.size + ld8 r25=[r27] // get vcpu->arch.vrr[r23]'s value + ;; + extr.u r25=r25,2,6 // get rr.ps + shl r22=r26,r29 // 1UL << pta.size + ;; + shr.u r23=r19,r25 // vaddr >> rr.ps + adds r26=3,r29 // pta.size + 3 + shl r27=r17,3 // pta << 3 + ;; + shl r23=r23,3 // (vaddr >> rr.ps) << 3 + shr.u r27=r27,r26 // (pta << 3) >> (pta.size+3) + movl r16=7<<61 + ;; + adds r22=-1,r22 // (1UL << pta.size) - 1 + shl r27=r27,r29 // ((pta<<3)>>(pta.size+3))<>(pta.size + 3))< + * Xiaoyan Feng (Fleming Feng) + * Xuefei Xu (Anthony Xu) (Anthony.xu@intel.com) + * Xiantao Zhang (xiantao.zhang@intel.com) + */ +#include "vcpu.h" + +#include +#include +#include +#include +#include + +fpswa_interface_t *vmm_fpswa_interface; + +#define IA64_VHPT_TRANS_VECTOR 0x0000 +#define IA64_INST_TLB_VECTOR 0x0400 +#define IA64_DATA_TLB_VECTOR 0x0800 +#define IA64_ALT_INST_TLB_VECTOR 0x0c00 +#define IA64_ALT_DATA_TLB_VECTOR 0x1000 +#define IA64_DATA_NESTED_TLB_VECTOR 0x1400 +#define IA64_INST_KEY_MISS_VECTOR 0x1800 +#define IA64_DATA_KEY_MISS_VECTOR 0x1c00 +#define IA64_DIRTY_BIT_VECTOR 0x2000 +#define IA64_INST_ACCESS_BIT_VECTOR 0x2400 +#define IA64_DATA_ACCESS_BIT_VECTOR 0x2800 +#define IA64_BREAK_VECTOR 0x2c00 +#define IA64_EXTINT_VECTOR 0x3000 +#define IA64_PAGE_NOT_PRESENT_VECTOR 0x5000 +#define IA64_KEY_PERMISSION_VECTOR 0x5100 +#define IA64_INST_ACCESS_RIGHTS_VECTOR 0x5200 +#define IA64_DATA_ACCESS_RIGHTS_VECTOR 0x5300 +#define IA64_GENEX_VECTOR 0x5400 +#define IA64_DISABLED_FPREG_VECTOR 0x5500 +#define IA64_NAT_CONSUMPTION_VECTOR 0x5600 +#define IA64_SPECULATION_VECTOR 0x5700 /* UNUSED */ +#define IA64_DEBUG_VECTOR 0x5900 +#define IA64_UNALIGNED_REF_VECTOR 0x5a00 +#define IA64_UNSUPPORTED_DATA_REF_VECTOR 0x5b00 +#define IA64_FP_FAULT_VECTOR 0x5c00 +#define IA64_FP_TRAP_VECTOR 0x5d00 +#define IA64_LOWERPRIV_TRANSFER_TRAP_VECTOR 0x5e00 +#define IA64_TAKEN_BRANCH_TRAP_VECTOR 0x5f00 +#define IA64_SINGLE_STEP_TRAP_VECTOR 0x6000 + +/* SDM vol2 5.5 - IVA based interruption handling */ +#define INITIAL_PSR_VALUE_AT_INTERRUPTION (IA64_PSR_UP | IA64_PSR_MFL |\ + IA64_PSR_MFH | IA64_PSR_PK | IA64_PSR_DT | \ + IA64_PSR_RT | IA64_PSR_MC|IA64_PSR_IT) + +#define DOMN_PAL_REQUEST 0x110000 +#define DOMN_SAL_REQUEST 0x110001 + +static u64 vec2off[68] = {0x0, 0x400, 0x800, 0xc00, 0x1000, 0x1400, 0x1800, + 0x1c00, 0x2000, 0x2400, 0x2800, 0x2c00, 0x3000, 0x3400, 0x3800, 0x3c00, + 0x4000, 0x4400, 0x4800, 0x4c00, 0x5000, 0x5100, 0x5200, 0x5300, 0x5400, + 0x5500, 0x5600, 0x5700, 0x5800, 0x5900, 0x5a00, 0x5b00, 0x5c00, 0x5d00, + 0x5e00, 0x5f00, 0x6000, 0x6100, 0x6200, 0x6300, 0x6400, 0x6500, 0x6600, + 0x6700, 0x6800, 0x6900, 0x6a00, 0x6b00, 0x6c00, 0x6d00, 0x6e00, 0x6f00, + 0x7000, 0x7100, 0x7200, 0x7300, 0x7400, 0x7500, 0x7600, 0x7700, 0x7800, + 0x7900, 0x7a00, 0x7b00, 0x7c00, 0x7d00, 0x7e00, 0x7f00 +}; + +static void collect_interruption(struct kvm_vcpu *vcpu) +{ + u64 ipsr; + u64 vdcr; + u64 vifs; + unsigned long vpsr; + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + vpsr = vcpu_get_psr(vcpu); + vcpu_bsw0(vcpu); + if (vpsr & IA64_PSR_IC) { + + /* Sync mpsr id/da/dd/ss/ed bits to vipsr + * since after guest do rfi, we still want these bits on in + * mpsr + */ + + ipsr = regs->cr_ipsr; + vpsr = vpsr | (ipsr & (IA64_PSR_ID | IA64_PSR_DA + | IA64_PSR_DD | IA64_PSR_SS + | IA64_PSR_ED)); + vcpu_set_ipsr(vcpu, vpsr); + + /* Currently, for trap, we do not advance IIP to next + * instruction. That's because we assume caller already + * set up IIP correctly + */ + + vcpu_set_iip(vcpu , regs->cr_iip); + + /* set vifs.v to zero */ + vifs = VCPU(vcpu, ifs); + vifs &= ~IA64_IFS_V; + vcpu_set_ifs(vcpu, vifs); + + vcpu_set_iipa(vcpu, VMX(vcpu, cr_iipa)); + } + + vdcr = VCPU(vcpu, dcr); + + /* Set guest psr + * up/mfl/mfh/pk/dt/rt/mc/it keeps unchanged + * be: set to the value of dcr.be + * pp: set to the value of dcr.pp + */ + vpsr &= INITIAL_PSR_VALUE_AT_INTERRUPTION; + vpsr |= (vdcr & IA64_DCR_BE); + + /* VDCR pp bit position is different from VPSR pp bit */ + if (vdcr & IA64_DCR_PP) { + vpsr |= IA64_PSR_PP; + } else { + vpsr &= ~IA64_PSR_PP;; + } + + vcpu_set_psr(vcpu, vpsr); + +} + +void inject_guest_interruption(struct kvm_vcpu *vcpu, u64 vec) +{ + u64 viva; + struct kvm_pt_regs *regs; + union ia64_isr pt_isr; + + regs = vcpu_regs(vcpu); + + /* clear cr.isr.ir (incomplete register frame)*/ + pt_isr.val = VMX(vcpu, cr_isr); + pt_isr.ir = 0; + VMX(vcpu, cr_isr) = pt_isr.val; + + collect_interruption(vcpu); + + viva = vcpu_get_iva(vcpu); + regs->cr_iip = viva + vec; +} + +static u64 vcpu_get_itir_on_fault(struct kvm_vcpu *vcpu, u64 ifa) +{ + union ia64_rr rr, rr1; + + rr.val = vcpu_get_rr(vcpu, ifa); + rr1.val = 0; + rr1.ps = rr.ps; + rr1.rid = rr.rid; + return (rr1.val); +} + + +/* + * Set vIFA & vITIR & vIHA, when vPSR.ic =1 + * Parameter: + * set_ifa: if true, set vIFA + * set_itir: if true, set vITIR + * set_iha: if true, set vIHA + */ +void set_ifa_itir_iha(struct kvm_vcpu *vcpu, u64 vadr, + int set_ifa, int set_itir, int set_iha) +{ + long vpsr; + u64 value; + + vpsr = VCPU(vcpu, vpsr); + /* Vol2, Table 8-1 */ + if (vpsr & IA64_PSR_IC) { + if (set_ifa) + vcpu_set_ifa(vcpu, vadr); + if (set_itir) { + value = vcpu_get_itir_on_fault(vcpu, vadr); + vcpu_set_itir(vcpu, value); + } + + if (set_iha) { + value = vcpu_thash(vcpu, vadr); + vcpu_set_iha(vcpu, value); + } + } +} + +/* + * Data TLB Fault + * @ Data TLB vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void dtlb_fault(struct kvm_vcpu *vcpu, u64 vadr) +{ + /* If vPSR.ic, IFA, ITIR, IHA */ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 1); + inject_guest_interruption(vcpu, IA64_DATA_TLB_VECTOR); +} + +/* + * Instruction TLB Fault + * @ Instruction TLB vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void itlb_fault(struct kvm_vcpu *vcpu, u64 vadr) +{ + /* If vPSR.ic, IFA, ITIR, IHA */ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 1); + inject_guest_interruption(vcpu, IA64_INST_TLB_VECTOR); +} + + + +/* + * Data Nested TLB Fault + * @ Data Nested TLB Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void nested_dtlb(struct kvm_vcpu *vcpu) +{ + inject_guest_interruption(vcpu, IA64_DATA_NESTED_TLB_VECTOR); +} + +/* + * Alternate Data TLB Fault + * @ Alternate Data TLB vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void alt_dtlb(struct kvm_vcpu *vcpu, u64 vadr) +{ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 0); + inject_guest_interruption(vcpu, IA64_ALT_DATA_TLB_VECTOR); +} + + +/* + * Data TLB Fault + * @ Data TLB vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void alt_itlb(struct kvm_vcpu *vcpu, u64 vadr) +{ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 0); + inject_guest_interruption(vcpu, IA64_ALT_INST_TLB_VECTOR); +} + +/* Deal with: + * VHPT Translation Vector + */ +static void _vhpt_fault(struct kvm_vcpu *vcpu, u64 vadr) +{ + /* If vPSR.ic, IFA, ITIR, IHA*/ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 1); + inject_guest_interruption(vcpu, IA64_VHPT_TRANS_VECTOR); + + +} + +/* + * VHPT Instruction Fault + * @ VHPT Translation vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void ivhpt_fault(struct kvm_vcpu *vcpu, u64 vadr) +{ + _vhpt_fault(vcpu, vadr); +} + + +/* + * VHPT Data Fault + * @ VHPT Translation vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void dvhpt_fault(struct kvm_vcpu *vcpu, u64 vadr) +{ + _vhpt_fault(vcpu, vadr); +} + + + +/* + * Deal with: + * General Exception vector + */ +void _general_exception(struct kvm_vcpu *vcpu) +{ + inject_guest_interruption(vcpu, IA64_GENEX_VECTOR); +} + + +/* + * Illegal Operation Fault + * @ General Exception Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void illegal_op(struct kvm_vcpu *vcpu) +{ + _general_exception(vcpu); +} + +/* + * Illegal Dependency Fault + * @ General Exception Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void illegal_dep(struct kvm_vcpu *vcpu) +{ + _general_exception(vcpu); +} + +/* + * Reserved Register/Field Fault + * @ General Exception Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void rsv_reg_field(struct kvm_vcpu *vcpu) +{ + _general_exception(vcpu); +} +/* + * Privileged Operation Fault + * @ General Exception Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ + +void privilege_op(struct kvm_vcpu *vcpu) +{ + _general_exception(vcpu); +} + +/* + * Unimplement Data Address Fault + * @ General Exception Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void unimpl_daddr(struct kvm_vcpu *vcpu) +{ + _general_exception(vcpu); +} + +/* + * Privileged Register Fault + * @ General Exception Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void privilege_reg(struct kvm_vcpu *vcpu) +{ + _general_exception(vcpu); +} + +/* Deal with + * Nat consumption vector + * Parameter: + * vaddr: Optional, if t == REGISTER + */ +static void _nat_consumption_fault(struct kvm_vcpu *vcpu, u64 vadr, + enum tlb_miss_type t) +{ + /* If vPSR.ic && t == DATA/INST, IFA */ + if (t == DATA || t == INSTRUCTION) { + /* IFA */ + set_ifa_itir_iha(vcpu, vadr, 1, 0, 0); + } + + inject_guest_interruption(vcpu, IA64_NAT_CONSUMPTION_VECTOR); +} + +/* + * Instruction Nat Page Consumption Fault + * @ Nat Consumption Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void inat_page_consumption(struct kvm_vcpu *vcpu, u64 vadr) +{ + _nat_consumption_fault(vcpu, vadr, INSTRUCTION); +} + +/* + * Register Nat Consumption Fault + * @ Nat Consumption Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void rnat_consumption(struct kvm_vcpu *vcpu) +{ + _nat_consumption_fault(vcpu, 0, REGISTER); +} + +/* + * Data Nat Page Consumption Fault + * @ Nat Consumption Vector + * Refer to SDM Vol2 Table 5-6 & 8-1 + */ +void dnat_page_consumption(struct kvm_vcpu *vcpu, u64 vadr) +{ + _nat_consumption_fault(vcpu, vadr, DATA); +} + +/* Deal with + * Page not present vector + */ +static void __page_not_present(struct kvm_vcpu *vcpu, u64 vadr) +{ + /* If vPSR.ic, IFA, ITIR */ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 0); + inject_guest_interruption(vcpu, IA64_PAGE_NOT_PRESENT_VECTOR); +} + + +void data_page_not_present(struct kvm_vcpu *vcpu, u64 vadr) +{ + __page_not_present(vcpu, vadr); +} + + +void inst_page_not_present(struct kvm_vcpu *vcpu, u64 vadr) +{ + __page_not_present(vcpu, vadr); +} + + +/* Deal with + * Data access rights vector + */ +void data_access_rights(struct kvm_vcpu *vcpu, u64 vadr) +{ + /* If vPSR.ic, IFA, ITIR */ + set_ifa_itir_iha(vcpu, vadr, 1, 1, 0); + inject_guest_interruption(vcpu, IA64_DATA_ACCESS_RIGHTS_VECTOR); +} + +fpswa_ret_t vmm_fp_emulate(int fp_fault, void *bundle, unsigned long *ipsr, + unsigned long *fpsr, unsigned long *isr, unsigned long *pr, + unsigned long *ifs, struct kvm_pt_regs *regs) +{ + fp_state_t fp_state; + fpswa_ret_t ret; + struct kvm_vcpu *vcpu = current_vcpu; + + uint64_t old_rr7 = ia64_get_rr(7UL<<61); + + if (!vmm_fpswa_interface) + return (fpswa_ret_t) {-1, 0, 0, 0}; + + /* + * Just let fpswa driver to use hardware fp registers. + * No fp register is valid in memory. + */ + memset(&fp_state, 0, sizeof(fp_state_t)); + + /* + * unsigned long (*EFI_FPSWA) ( + * unsigned long trap_type, + * void *Bundle, + * unsigned long *pipsr, + * unsigned long *pfsr, + * unsigned long *pisr, + * unsigned long *ppreds, + * unsigned long *pifs, + * void *fp_state); + */ + /*Call host fpswa interface directly to virtualize + *guest fpswa request! + */ + ia64_set_rr(7UL << 61, vcpu->arch.host.rr[7]); + ia64_srlz_d(); + + ret = (*vmm_fpswa_interface->fpswa) (fp_fault, bundle, + ipsr, fpsr, isr, pr, ifs, &fp_state); + ia64_set_rr(7UL << 61, old_rr7); + ia64_srlz_d(); + return ret; +} + +/* + * Handle floating-point assist faults and traps for domain. + */ +unsigned long vmm_handle_fpu_swa(int fp_fault, struct kvm_pt_regs *regs, + unsigned long isr) +{ + struct kvm_vcpu *v = current_vcpu; + IA64_BUNDLE bundle; + unsigned long fault_ip; + fpswa_ret_t ret; + + fault_ip = regs->cr_iip; + /* + * When the FP trap occurs, the trapping instruction is completed. + * If ipsr.ri == 0, there is the trapping instruction in previous + * bundle. + */ + if (!fp_fault && (ia64_psr(regs)->ri == 0)) + fault_ip -= 16; + + if (fetch_code(v, fault_ip, &bundle)) + return -EAGAIN; + + if (!bundle.i64[0] && !bundle.i64[1]) + return -EACCES; + + ret = vmm_fp_emulate(fp_fault, &bundle, ®s->cr_ipsr, ®s->ar_fpsr, + &isr, ®s->pr, ®s->cr_ifs, regs); + return ret.status; +} + +void reflect_interruption(u64 ifa, u64 isr, u64 iim, + u64 vec, struct kvm_pt_regs *regs) +{ + u64 vector; + int status ; + struct kvm_vcpu *vcpu = current_vcpu; + u64 vpsr = VCPU(vcpu, vpsr); + + vector = vec2off[vec]; + + if (!(vpsr & IA64_PSR_IC) && (vector != IA64_DATA_NESTED_TLB_VECTOR)) { + panic_vm(vcpu); + return; + } + + switch (vec) { + case 32: /*IA64_FP_FAULT_VECTOR*/ + status = vmm_handle_fpu_swa(1, regs, isr); + if (!status) { + vcpu_increment_iip(vcpu); + return; + } else if (-EAGAIN == status) + return; + break; + case 33: /*IA64_FP_TRAP_VECTOR*/ + status = vmm_handle_fpu_swa(0, regs, isr); + if (!status) + return ; + else if (-EAGAIN == status) { + vcpu_decrement_iip(vcpu); + return ; + } + break; + } + + VCPU(vcpu, isr) = isr; + VCPU(vcpu, iipa) = regs->cr_iip; + if (vector == IA64_BREAK_VECTOR || vector == IA64_SPECULATION_VECTOR) + VCPU(vcpu, iim) = iim; + else + set_ifa_itir_iha(vcpu, ifa, 1, 1, 1); + + inject_guest_interruption(vcpu, vector); +} + +static void set_pal_call_data(struct kvm_vcpu *vcpu) +{ + struct exit_ctl_data *p = &vcpu->arch.exit_data; + + /*FIXME:For static and stacked convention, firmware + * has put the parameters in gr28-gr31 before + * break to vmm !!*/ + + p->u.pal_data.gr28 = vcpu_get_gr(vcpu, 28); + p->u.pal_data.gr29 = vcpu_get_gr(vcpu, 29); + p->u.pal_data.gr30 = vcpu_get_gr(vcpu, 30); + p->u.pal_data.gr31 = vcpu_get_gr(vcpu, 31); + p->exit_reason = EXIT_REASON_PAL_CALL; +} + +static void set_pal_call_result(struct kvm_vcpu *vcpu) +{ + struct exit_ctl_data *p = &vcpu->arch.exit_data; + + if (p->exit_reason == EXIT_REASON_PAL_CALL) { + vcpu_set_gr(vcpu, 8, p->u.pal_data.ret.status, 0); + vcpu_set_gr(vcpu, 9, p->u.pal_data.ret.v0, 0); + vcpu_set_gr(vcpu, 10, p->u.pal_data.ret.v1, 0); + vcpu_set_gr(vcpu, 11, p->u.pal_data.ret.v2, 0); + } else + panic_vm(vcpu); +} + +static void set_sal_call_data(struct kvm_vcpu *vcpu) +{ + struct exit_ctl_data *p = &vcpu->arch.exit_data; + + p->u.sal_data.in0 = vcpu_get_gr(vcpu, 32); + p->u.sal_data.in1 = vcpu_get_gr(vcpu, 33); + p->u.sal_data.in2 = vcpu_get_gr(vcpu, 34); + p->u.sal_data.in3 = vcpu_get_gr(vcpu, 35); + p->u.sal_data.in4 = vcpu_get_gr(vcpu, 36); + p->u.sal_data.in5 = vcpu_get_gr(vcpu, 37); + p->u.sal_data.in6 = vcpu_get_gr(vcpu, 38); + p->u.sal_data.in7 = vcpu_get_gr(vcpu, 39); + p->exit_reason = EXIT_REASON_SAL_CALL; +} + +static void set_sal_call_result(struct kvm_vcpu *vcpu) +{ + struct exit_ctl_data *p = &vcpu->arch.exit_data; + + if (p->exit_reason == EXIT_REASON_SAL_CALL) { + vcpu_set_gr(vcpu, 8, p->u.sal_data.ret.r8, 0); + vcpu_set_gr(vcpu, 9, p->u.sal_data.ret.r9, 0); + vcpu_set_gr(vcpu, 10, p->u.sal_data.ret.r10, 0); + vcpu_set_gr(vcpu, 11, p->u.sal_data.ret.r11, 0); + } else + panic_vm(vcpu); +} + +void kvm_ia64_handle_break(unsigned long ifa, struct kvm_pt_regs *regs, + unsigned long isr, unsigned long iim) +{ + struct kvm_vcpu *v = current_vcpu; + + if (ia64_psr(regs)->cpl == 0) { + /* Allow hypercalls only when cpl = 0. */ + if (iim == DOMN_PAL_REQUEST) { + set_pal_call_data(v); + vmm_transition(v); + set_pal_call_result(v); + vcpu_increment_iip(v); + return; + } else if (iim == DOMN_SAL_REQUEST) { + set_sal_call_data(v); + vmm_transition(v); + set_sal_call_result(v); + vcpu_increment_iip(v); + return; + } + } + reflect_interruption(ifa, isr, iim, 11, regs); +} + +void check_pending_irq(struct kvm_vcpu *vcpu) +{ + int mask, h_pending, h_inservice; + u64 isr; + unsigned long vpsr; + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + h_pending = highest_pending_irq(vcpu); + if (h_pending == NULL_VECTOR) { + update_vhpi(vcpu, NULL_VECTOR); + return; + } + h_inservice = highest_inservice_irq(vcpu); + + vpsr = VCPU(vcpu, vpsr); + mask = irq_masked(vcpu, h_pending, h_inservice); + if ((vpsr & IA64_PSR_I) && IRQ_NO_MASKED == mask) { + isr = vpsr & IA64_PSR_RI; + update_vhpi(vcpu, h_pending); + reflect_interruption(0, isr, 0, 12, regs); /* EXT IRQ */ + } else if (mask == IRQ_MASKED_BY_INSVC) { + if (VCPU(vcpu, vhpi)) + update_vhpi(vcpu, NULL_VECTOR); + } else { + /* masked by vpsr.i or vtpr.*/ + update_vhpi(vcpu, h_pending); + } +} + +static void generate_exirq(struct kvm_vcpu *vcpu) +{ + unsigned vpsr; + uint64_t isr; + + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + vpsr = VCPU(vcpu, vpsr); + isr = vpsr & IA64_PSR_RI; + if (!(vpsr & IA64_PSR_IC)) + panic_vm(vcpu); + reflect_interruption(0, isr, 0, 12, regs); /* EXT IRQ */ +} + +void vhpi_detection(struct kvm_vcpu *vcpu) +{ + uint64_t threshold, vhpi; + union ia64_tpr vtpr; + struct ia64_psr vpsr; + + vpsr = *(struct ia64_psr *)&VCPU(vcpu, vpsr); + vtpr.val = VCPU(vcpu, tpr); + + threshold = ((!vpsr.i) << 5) | (vtpr.mmi << 4) | vtpr.mic; + vhpi = VCPU(vcpu, vhpi); + if (vhpi > threshold) { + /* interrupt actived*/ + generate_exirq(vcpu); + } +} + + +void leave_hypervisor_tail(void) +{ + struct kvm_vcpu *v = current_vcpu; + + if (VMX(v, timer_check)) { + VMX(v, timer_check) = 0; + if (VMX(v, itc_check)) { + if (vcpu_get_itc(v) > VCPU(v, itm)) { + if (!(VCPU(v, itv) & (1 << 16))) { + vcpu_pend_interrupt(v, VCPU(v, itv) + & 0xff); + VMX(v, itc_check) = 0; + } else { + v->arch.timer_pending = 1; + } + VMX(v, last_itc) = VCPU(v, itm) + 1; + } + } + } + + rmb(); + if (v->arch.irq_new_pending) { + v->arch.irq_new_pending = 0; + VMX(v, irq_check) = 0; + check_pending_irq(v); + return; + } + if (VMX(v, irq_check)) { + VMX(v, irq_check) = 0; + vhpi_detection(v); + } +} + + +static inline void handle_lds(struct kvm_pt_regs *regs) +{ + regs->cr_ipsr |= IA64_PSR_ED; +} + +void physical_tlb_miss(struct kvm_vcpu *vcpu, unsigned long vadr, int type) +{ + unsigned long pte; + union ia64_rr rr; + + rr.val = ia64_get_rr(vadr); + pte = vadr & _PAGE_PPN_MASK; + pte = pte | PHY_PAGE_WB; + thash_vhpt_insert(vcpu, pte, (u64)(rr.ps << 2), vadr, type); + return; +} + +void kvm_page_fault(u64 vadr , u64 vec, struct kvm_pt_regs *regs) +{ + unsigned long vpsr; + int type; + + u64 vhpt_adr, gppa, pteval, rr, itir; + union ia64_isr misr; + union ia64_pta vpta; + struct thash_data *data; + struct kvm_vcpu *v = current_vcpu; + + vpsr = VCPU(v, vpsr); + misr.val = VMX(v, cr_isr); + + type = vec; + + if (is_physical_mode(v) && (!(vadr << 1 >> 62))) { + if (vec == 2) { + if (__gpfn_is_io((vadr << 1) >> (PAGE_SHIFT + 1))) { + emulate_io_inst(v, ((vadr << 1) >> 1), 4); + return; + } + } + physical_tlb_miss(v, vadr, type); + return; + } + data = vtlb_lookup(v, vadr, type); + if (data != 0) { + if (type == D_TLB) { + gppa = (vadr & ((1UL << data->ps) - 1)) + + (data->ppn >> (data->ps - 12) << data->ps); + if (__gpfn_is_io(gppa >> PAGE_SHIFT)) { + if (data->pl >= ((regs->cr_ipsr >> + IA64_PSR_CPL0_BIT) & 3)) + emulate_io_inst(v, gppa, data->ma); + else { + vcpu_set_isr(v, misr.val); + data_access_rights(v, vadr); + } + return ; + } + } + thash_vhpt_insert(v, data->page_flags, data->itir, vadr, type); + + } else if (type == D_TLB) { + if (misr.sp) { + handle_lds(regs); + return; + } + + rr = vcpu_get_rr(v, vadr); + itir = rr & (RR_RID_MASK | RR_PS_MASK); + + if (!vhpt_enabled(v, vadr, misr.rs ? RSE_REF : DATA_REF)) { + if (vpsr & IA64_PSR_IC) { + vcpu_set_isr(v, misr.val); + alt_dtlb(v, vadr); + } else { + nested_dtlb(v); + } + return ; + } + + vpta.val = vcpu_get_pta(v); + /* avoid recursively walking (short format) VHPT */ + + vhpt_adr = vcpu_thash(v, vadr); + if (!guest_vhpt_lookup(vhpt_adr, &pteval)) { + /* VHPT successfully read. */ + if (!(pteval & _PAGE_P)) { + if (vpsr & IA64_PSR_IC) { + vcpu_set_isr(v, misr.val); + dtlb_fault(v, vadr); + } else { + nested_dtlb(v); + } + } else if ((pteval & _PAGE_MA_MASK) != _PAGE_MA_ST) { + thash_purge_and_insert(v, pteval, itir, + vadr, D_TLB); + } else if (vpsr & IA64_PSR_IC) { + vcpu_set_isr(v, misr.val); + dtlb_fault(v, vadr); + } else { + nested_dtlb(v); + } + } else { + /* Can't read VHPT. */ + if (vpsr & IA64_PSR_IC) { + vcpu_set_isr(v, misr.val); + dvhpt_fault(v, vadr); + } else { + nested_dtlb(v); + } + } + } else if (type == I_TLB) { + if (!(vpsr & IA64_PSR_IC)) + misr.ni = 1; + if (!vhpt_enabled(v, vadr, INST_REF)) { + vcpu_set_isr(v, misr.val); + alt_itlb(v, vadr); + return; + } + + vpta.val = vcpu_get_pta(v); + + vhpt_adr = vcpu_thash(v, vadr); + if (!guest_vhpt_lookup(vhpt_adr, &pteval)) { + /* VHPT successfully read. */ + if (pteval & _PAGE_P) { + if ((pteval & _PAGE_MA_MASK) == _PAGE_MA_ST) { + vcpu_set_isr(v, misr.val); + itlb_fault(v, vadr); + return ; + } + rr = vcpu_get_rr(v, vadr); + itir = rr & (RR_RID_MASK | RR_PS_MASK); + thash_purge_and_insert(v, pteval, itir, + vadr, I_TLB); + } else { + vcpu_set_isr(v, misr.val); + inst_page_not_present(v, vadr); + } + } else { + vcpu_set_isr(v, misr.val); + ivhpt_fault(v, vadr); + } + } +} + +void kvm_vexirq(struct kvm_vcpu *vcpu) +{ + u64 vpsr, isr; + struct kvm_pt_regs *regs; + + regs = vcpu_regs(vcpu); + vpsr = VCPU(vcpu, vpsr); + isr = vpsr & IA64_PSR_RI; + reflect_interruption(0, isr, 0, 12, regs); /*EXT IRQ*/ +} + +void kvm_ia64_handle_irq(struct kvm_vcpu *v) +{ + struct exit_ctl_data *p = &v->arch.exit_data; + long psr; + + local_irq_save(psr); + p->exit_reason = EXIT_REASON_EXTERNAL_INTERRUPT; + vmm_transition(v); + local_irq_restore(psr); + + VMX(v, timer_check) = 1; + +} + +static void ptc_ga_remote_func(struct kvm_vcpu *v, int pos) +{ + u64 oldrid, moldrid, oldpsbits, vaddr; + struct kvm_ptc_g *p = &v->arch.ptc_g_data[pos]; + vaddr = p->vaddr; + + oldrid = VMX(v, vrr[0]); + VMX(v, vrr[0]) = p->rr; + oldpsbits = VMX(v, psbits[0]); + VMX(v, psbits[0]) = VMX(v, psbits[REGION_NUMBER(vaddr)]); + moldrid = ia64_get_rr(0x0); + ia64_set_rr(0x0, vrrtomrr(p->rr)); + ia64_srlz_d(); + + vaddr = PAGEALIGN(vaddr, p->ps); + thash_purge_entries_remote(v, vaddr, p->ps); + + VMX(v, vrr[0]) = oldrid; + VMX(v, psbits[0]) = oldpsbits; + ia64_set_rr(0x0, moldrid); + ia64_dv_serialize_data(); +} + +static void vcpu_do_resume(struct kvm_vcpu *vcpu) +{ + /*Re-init VHPT and VTLB once from resume*/ + vcpu->arch.vhpt.num = VHPT_NUM_ENTRIES; + thash_init(&vcpu->arch.vhpt, VHPT_SHIFT); + vcpu->arch.vtlb.num = VTLB_NUM_ENTRIES; + thash_init(&vcpu->arch.vtlb, VTLB_SHIFT); + + ia64_set_pta(vcpu->arch.vhpt.pta.val); +} + +static void kvm_do_resume_op(struct kvm_vcpu *vcpu) +{ + if (test_and_clear_bit(KVM_REQ_RESUME, &vcpu->requests)) { + vcpu_do_resume(vcpu); + return; + } + + if (unlikely(test_and_clear_bit(KVM_REQ_TLB_FLUSH, &vcpu->requests))) { + thash_purge_all(vcpu); + return; + } + + if (test_and_clear_bit(KVM_REQ_PTC_G, &vcpu->requests)) { + while (vcpu->arch.ptc_g_count > 0) + ptc_ga_remote_func(vcpu, --vcpu->arch.ptc_g_count); + } +} + +void vmm_transition(struct kvm_vcpu *vcpu) +{ + ia64_call_vsa(PAL_VPS_SAVE, (unsigned long)vcpu->arch.vpd, + 0, 0, 0, 0, 0, 0); + vmm_trampoline(&vcpu->arch.guest, &vcpu->arch.host); + ia64_call_vsa(PAL_VPS_RESTORE, (unsigned long)vcpu->arch.vpd, + 0, 0, 0, 0, 0, 0); + kvm_do_resume_op(vcpu); +} diff --git a/arch/ia64/kvm/trampoline.S b/arch/ia64/kvm/trampoline.S new file mode 100644 index 0000000000000000000000000000000000000000..30897d44d61eacce620eaea515bcb6faaa0ccbab --- /dev/null +++ b/arch/ia64/kvm/trampoline.S @@ -0,0 +1,1038 @@ +/* Save all processor states + * + * Copyright (c) 2007 Fleming Feng + * Copyright (c) 2007 Anthony Xu + */ + +#include +#include "asm-offsets.h" + + +#define CTX(name) VMM_CTX_##name##_OFFSET + + /* + * r32: context_t base address + */ +#define SAVE_BRANCH_REGS \ + add r2 = CTX(B0),r32; \ + add r3 = CTX(B1),r32; \ + mov r16 = b0; \ + mov r17 = b1; \ + ;; \ + st8 [r2]=r16,16; \ + st8 [r3]=r17,16; \ + ;; \ + mov r16 = b2; \ + mov r17 = b3; \ + ;; \ + st8 [r2]=r16,16; \ + st8 [r3]=r17,16; \ + ;; \ + mov r16 = b4; \ + mov r17 = b5; \ + ;; \ + st8 [r2]=r16; \ + st8 [r3]=r17; \ + ;; + + /* + * r33: context_t base address + */ +#define RESTORE_BRANCH_REGS \ + add r2 = CTX(B0),r33; \ + add r3 = CTX(B1),r33; \ + ;; \ + ld8 r16=[r2],16; \ + ld8 r17=[r3],16; \ + ;; \ + mov b0 = r16; \ + mov b1 = r17; \ + ;; \ + ld8 r16=[r2],16; \ + ld8 r17=[r3],16; \ + ;; \ + mov b2 = r16; \ + mov b3 = r17; \ + ;; \ + ld8 r16=[r2]; \ + ld8 r17=[r3]; \ + ;; \ + mov b4=r16; \ + mov b5=r17; \ + ;; + + + /* + * r32: context_t base address + * bsw == 1 + * Save all bank1 general registers, r4 ~ r7 + */ +#define SAVE_GENERAL_REGS \ + add r2=CTX(R4),r32; \ + add r3=CTX(R5),r32; \ + ;; \ +.mem.offset 0,0; \ + st8.spill [r2]=r4,16; \ +.mem.offset 8,0; \ + st8.spill [r3]=r5,16; \ + ;; \ +.mem.offset 0,0; \ + st8.spill [r2]=r6,48; \ +.mem.offset 8,0; \ + st8.spill [r3]=r7,48; \ + ;; \ +.mem.offset 0,0; \ + st8.spill [r2]=r12; \ +.mem.offset 8,0; \ + st8.spill [r3]=r13; \ + ;; + + /* + * r33: context_t base address + * bsw == 1 + */ +#define RESTORE_GENERAL_REGS \ + add r2=CTX(R4),r33; \ + add r3=CTX(R5),r33; \ + ;; \ + ld8.fill r4=[r2],16; \ + ld8.fill r5=[r3],16; \ + ;; \ + ld8.fill r6=[r2],48; \ + ld8.fill r7=[r3],48; \ + ;; \ + ld8.fill r12=[r2]; \ + ld8.fill r13 =[r3]; \ + ;; + + + + + /* + * r32: context_t base address + */ +#define SAVE_KERNEL_REGS \ + add r2 = CTX(KR0),r32; \ + add r3 = CTX(KR1),r32; \ + mov r16 = ar.k0; \ + mov r17 = ar.k1; \ + ;; \ + st8 [r2] = r16,16; \ + st8 [r3] = r17,16; \ + ;; \ + mov r16 = ar.k2; \ + mov r17 = ar.k3; \ + ;; \ + st8 [r2] = r16,16; \ + st8 [r3] = r17,16; \ + ;; \ + mov r16 = ar.k4; \ + mov r17 = ar.k5; \ + ;; \ + st8 [r2] = r16,16; \ + st8 [r3] = r17,16; \ + ;; \ + mov r16 = ar.k6; \ + mov r17 = ar.k7; \ + ;; \ + st8 [r2] = r16; \ + st8 [r3] = r17; \ + ;; + + + + /* + * r33: context_t base address + */ +#define RESTORE_KERNEL_REGS \ + add r2 = CTX(KR0),r33; \ + add r3 = CTX(KR1),r33; \ + ;; \ + ld8 r16=[r2],16; \ + ld8 r17=[r3],16; \ + ;; \ + mov ar.k0=r16; \ + mov ar.k1=r17; \ + ;; \ + ld8 r16=[r2],16; \ + ld8 r17=[r3],16; \ + ;; \ + mov ar.k2=r16; \ + mov ar.k3=r17; \ + ;; \ + ld8 r16=[r2],16; \ + ld8 r17=[r3],16; \ + ;; \ + mov ar.k4=r16; \ + mov ar.k5=r17; \ + ;; \ + ld8 r16=[r2],16; \ + ld8 r17=[r3],16; \ + ;; \ + mov ar.k6=r16; \ + mov ar.k7=r17; \ + ;; + + + + /* + * r32: context_t base address + */ +#define SAVE_APP_REGS \ + add r2 = CTX(BSPSTORE),r32; \ + mov r16 = ar.bspstore; \ + ;; \ + st8 [r2] = r16,CTX(RNAT)-CTX(BSPSTORE);\ + mov r16 = ar.rnat; \ + ;; \ + st8 [r2] = r16,CTX(FCR)-CTX(RNAT); \ + mov r16 = ar.fcr; \ + ;; \ + st8 [r2] = r16,CTX(EFLAG)-CTX(FCR); \ + mov r16 = ar.eflag; \ + ;; \ + st8 [r2] = r16,CTX(CFLG)-CTX(EFLAG); \ + mov r16 = ar.cflg; \ + ;; \ + st8 [r2] = r16,CTX(FSR)-CTX(CFLG); \ + mov r16 = ar.fsr; \ + ;; \ + st8 [r2] = r16,CTX(FIR)-CTX(FSR); \ + mov r16 = ar.fir; \ + ;; \ + st8 [r2] = r16,CTX(FDR)-CTX(FIR); \ + mov r16 = ar.fdr; \ + ;; \ + st8 [r2] = r16,CTX(UNAT)-CTX(FDR); \ + mov r16 = ar.unat; \ + ;; \ + st8 [r2] = r16,CTX(FPSR)-CTX(UNAT); \ + mov r16 = ar.fpsr; \ + ;; \ + st8 [r2] = r16,CTX(PFS)-CTX(FPSR); \ + mov r16 = ar.pfs; \ + ;; \ + st8 [r2] = r16,CTX(LC)-CTX(PFS); \ + mov r16 = ar.lc; \ + ;; \ + st8 [r2] = r16; \ + ;; + + /* + * r33: context_t base address + */ +#define RESTORE_APP_REGS \ + add r2=CTX(BSPSTORE),r33; \ + ;; \ + ld8 r16=[r2],CTX(RNAT)-CTX(BSPSTORE); \ + ;; \ + mov ar.bspstore=r16; \ + ld8 r16=[r2],CTX(FCR)-CTX(RNAT); \ + ;; \ + mov ar.rnat=r16; \ + ld8 r16=[r2],CTX(EFLAG)-CTX(FCR); \ + ;; \ + mov ar.fcr=r16; \ + ld8 r16=[r2],CTX(CFLG)-CTX(EFLAG); \ + ;; \ + mov ar.eflag=r16; \ + ld8 r16=[r2],CTX(FSR)-CTX(CFLG); \ + ;; \ + mov ar.cflg=r16; \ + ld8 r16=[r2],CTX(FIR)-CTX(FSR); \ + ;; \ + mov ar.fsr=r16; \ + ld8 r16=[r2],CTX(FDR)-CTX(FIR); \ + ;; \ + mov ar.fir=r16; \ + ld8 r16=[r2],CTX(UNAT)-CTX(FDR); \ + ;; \ + mov ar.fdr=r16; \ + ld8 r16=[r2],CTX(FPSR)-CTX(UNAT); \ + ;; \ + mov ar.unat=r16; \ + ld8 r16=[r2],CTX(PFS)-CTX(FPSR); \ + ;; \ + mov ar.fpsr=r16; \ + ld8 r16=[r2],CTX(LC)-CTX(PFS); \ + ;; \ + mov ar.pfs=r16; \ + ld8 r16=[r2]; \ + ;; \ + mov ar.lc=r16; \ + ;; + + /* + * r32: context_t base address + */ +#define SAVE_CTL_REGS \ + add r2 = CTX(DCR),r32; \ + mov r16 = cr.dcr; \ + ;; \ + st8 [r2] = r16,CTX(IVA)-CTX(DCR); \ + ;; \ + mov r16 = cr.iva; \ + ;; \ + st8 [r2] = r16,CTX(PTA)-CTX(IVA); \ + ;; \ + mov r16 = cr.pta; \ + ;; \ + st8 [r2] = r16 ; \ + ;; + + /* + * r33: context_t base address + */ +#define RESTORE_CTL_REGS \ + add r2 = CTX(DCR),r33; \ + ;; \ + ld8 r16 = [r2],CTX(IVA)-CTX(DCR); \ + ;; \ + mov cr.dcr = r16; \ + dv_serialize_data; \ + ;; \ + ld8 r16 = [r2],CTX(PTA)-CTX(IVA); \ + ;; \ + mov cr.iva = r16; \ + dv_serialize_data; \ + ;; \ + ld8 r16 = [r2]; \ + ;; \ + mov cr.pta = r16; \ + dv_serialize_data; \ + ;; + + + /* + * r32: context_t base address + */ +#define SAVE_REGION_REGS \ + add r2=CTX(RR0),r32; \ + mov r16=rr[r0]; \ + dep.z r18=1,61,3; \ + ;; \ + st8 [r2]=r16,8; \ + mov r17=rr[r18]; \ + dep.z r18=2,61,3; \ + ;; \ + st8 [r2]=r17,8; \ + mov r16=rr[r18]; \ + dep.z r18=3,61,3; \ + ;; \ + st8 [r2]=r16,8; \ + mov r17=rr[r18]; \ + dep.z r18=4,61,3; \ + ;; \ + st8 [r2]=r17,8; \ + mov r16=rr[r18]; \ + dep.z r18=5,61,3; \ + ;; \ + st8 [r2]=r16,8; \ + mov r17=rr[r18]; \ + dep.z r18=7,61,3; \ + ;; \ + st8 [r2]=r17,16; \ + mov r16=rr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + ;; + + /* + * r33:context_t base address + */ +#define RESTORE_REGION_REGS \ + add r2=CTX(RR0),r33;\ + mov r18=r0; \ + ;; \ + ld8 r20=[r2],8; \ + ;; /* rr0 */ \ + ld8 r21=[r2],8; \ + ;; /* rr1 */ \ + ld8 r22=[r2],8; \ + ;; /* rr2 */ \ + ld8 r23=[r2],8; \ + ;; /* rr3 */ \ + ld8 r24=[r2],8; \ + ;; /* rr4 */ \ + ld8 r25=[r2],16; \ + ;; /* rr5 */ \ + ld8 r27=[r2]; \ + ;; /* rr7 */ \ + mov rr[r18]=r20; \ + dep.z r18=1,61,3; \ + ;; /* rr1 */ \ + mov rr[r18]=r21; \ + dep.z r18=2,61,3; \ + ;; /* rr2 */ \ + mov rr[r18]=r22; \ + dep.z r18=3,61,3; \ + ;; /* rr3 */ \ + mov rr[r18]=r23; \ + dep.z r18=4,61,3; \ + ;; /* rr4 */ \ + mov rr[r18]=r24; \ + dep.z r18=5,61,3; \ + ;; /* rr5 */ \ + mov rr[r18]=r25; \ + dep.z r18=7,61,3; \ + ;; /* rr7 */ \ + mov rr[r18]=r27; \ + ;; \ + srlz.i; \ + ;; + + + + /* + * r32: context_t base address + * r36~r39:scratch registers + */ +#define SAVE_DEBUG_REGS \ + add r2=CTX(IBR0),r32; \ + add r3=CTX(DBR0),r32; \ + mov r16=ibr[r0]; \ + mov r17=dbr[r0]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=1,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=2,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=2,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=3,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=4,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=5,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=6,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + add r18=7,r0; \ + ;; \ + mov r16=ibr[r18]; \ + mov r17=dbr[r18]; \ + ;; \ + st8 [r2]=r16,8; \ + st8 [r3]=r17,8; \ + ;; + + +/* + * r33: point to context_t structure + * ar.lc are corrupted. + */ +#define RESTORE_DEBUG_REGS \ + add r2=CTX(IBR0),r33; \ + add r3=CTX(DBR0),r33; \ + mov r16=7; \ + mov r17=r0; \ + ;; \ + mov ar.lc = r16; \ + ;; \ +1: \ + ld8 r18=[r2],8; \ + ld8 r19=[r3],8; \ + ;; \ + mov ibr[r17]=r18; \ + mov dbr[r17]=r19; \ + ;; \ + srlz.i; \ + ;; \ + add r17=1,r17; \ + br.cloop.sptk 1b; \ + ;; + + + /* + * r32: context_t base address + */ +#define SAVE_FPU_LOW \ + add r2=CTX(F2),r32; \ + add r3=CTX(F3),r32; \ + ;; \ + stf.spill.nta [r2]=f2,32; \ + stf.spill.nta [r3]=f3,32; \ + ;; \ + stf.spill.nta [r2]=f4,32; \ + stf.spill.nta [r3]=f5,32; \ + ;; \ + stf.spill.nta [r2]=f6,32; \ + stf.spill.nta [r3]=f7,32; \ + ;; \ + stf.spill.nta [r2]=f8,32; \ + stf.spill.nta [r3]=f9,32; \ + ;; \ + stf.spill.nta [r2]=f10,32; \ + stf.spill.nta [r3]=f11,32; \ + ;; \ + stf.spill.nta [r2]=f12,32; \ + stf.spill.nta [r3]=f13,32; \ + ;; \ + stf.spill.nta [r2]=f14,32; \ + stf.spill.nta [r3]=f15,32; \ + ;; \ + stf.spill.nta [r2]=f16,32; \ + stf.spill.nta [r3]=f17,32; \ + ;; \ + stf.spill.nta [r2]=f18,32; \ + stf.spill.nta [r3]=f19,32; \ + ;; \ + stf.spill.nta [r2]=f20,32; \ + stf.spill.nta [r3]=f21,32; \ + ;; \ + stf.spill.nta [r2]=f22,32; \ + stf.spill.nta [r3]=f23,32; \ + ;; \ + stf.spill.nta [r2]=f24,32; \ + stf.spill.nta [r3]=f25,32; \ + ;; \ + stf.spill.nta [r2]=f26,32; \ + stf.spill.nta [r3]=f27,32; \ + ;; \ + stf.spill.nta [r2]=f28,32; \ + stf.spill.nta [r3]=f29,32; \ + ;; \ + stf.spill.nta [r2]=f30; \ + stf.spill.nta [r3]=f31; \ + ;; + + /* + * r32: context_t base address + */ +#define SAVE_FPU_HIGH \ + add r2=CTX(F32),r32; \ + add r3=CTX(F33),r32; \ + ;; \ + stf.spill.nta [r2]=f32,32; \ + stf.spill.nta [r3]=f33,32; \ + ;; \ + stf.spill.nta [r2]=f34,32; \ + stf.spill.nta [r3]=f35,32; \ + ;; \ + stf.spill.nta [r2]=f36,32; \ + stf.spill.nta [r3]=f37,32; \ + ;; \ + stf.spill.nta [r2]=f38,32; \ + stf.spill.nta [r3]=f39,32; \ + ;; \ + stf.spill.nta [r2]=f40,32; \ + stf.spill.nta [r3]=f41,32; \ + ;; \ + stf.spill.nta [r2]=f42,32; \ + stf.spill.nta [r3]=f43,32; \ + ;; \ + stf.spill.nta [r2]=f44,32; \ + stf.spill.nta [r3]=f45,32; \ + ;; \ + stf.spill.nta [r2]=f46,32; \ + stf.spill.nta [r3]=f47,32; \ + ;; \ + stf.spill.nta [r2]=f48,32; \ + stf.spill.nta [r3]=f49,32; \ + ;; \ + stf.spill.nta [r2]=f50,32; \ + stf.spill.nta [r3]=f51,32; \ + ;; \ + stf.spill.nta [r2]=f52,32; \ + stf.spill.nta [r3]=f53,32; \ + ;; \ + stf.spill.nta [r2]=f54,32; \ + stf.spill.nta [r3]=f55,32; \ + ;; \ + stf.spill.nta [r2]=f56,32; \ + stf.spill.nta [r3]=f57,32; \ + ;; \ + stf.spill.nta [r2]=f58,32; \ + stf.spill.nta [r3]=f59,32; \ + ;; \ + stf.spill.nta [r2]=f60,32; \ + stf.spill.nta [r3]=f61,32; \ + ;; \ + stf.spill.nta [r2]=f62,32; \ + stf.spill.nta [r3]=f63,32; \ + ;; \ + stf.spill.nta [r2]=f64,32; \ + stf.spill.nta [r3]=f65,32; \ + ;; \ + stf.spill.nta [r2]=f66,32; \ + stf.spill.nta [r3]=f67,32; \ + ;; \ + stf.spill.nta [r2]=f68,32; \ + stf.spill.nta [r3]=f69,32; \ + ;; \ + stf.spill.nta [r2]=f70,32; \ + stf.spill.nta [r3]=f71,32; \ + ;; \ + stf.spill.nta [r2]=f72,32; \ + stf.spill.nta [r3]=f73,32; \ + ;; \ + stf.spill.nta [r2]=f74,32; \ + stf.spill.nta [r3]=f75,32; \ + ;; \ + stf.spill.nta [r2]=f76,32; \ + stf.spill.nta [r3]=f77,32; \ + ;; \ + stf.spill.nta [r2]=f78,32; \ + stf.spill.nta [r3]=f79,32; \ + ;; \ + stf.spill.nta [r2]=f80,32; \ + stf.spill.nta [r3]=f81,32; \ + ;; \ + stf.spill.nta [r2]=f82,32; \ + stf.spill.nta [r3]=f83,32; \ + ;; \ + stf.spill.nta [r2]=f84,32; \ + stf.spill.nta [r3]=f85,32; \ + ;; \ + stf.spill.nta [r2]=f86,32; \ + stf.spill.nta [r3]=f87,32; \ + ;; \ + stf.spill.nta [r2]=f88,32; \ + stf.spill.nta [r3]=f89,32; \ + ;; \ + stf.spill.nta [r2]=f90,32; \ + stf.spill.nta [r3]=f91,32; \ + ;; \ + stf.spill.nta [r2]=f92,32; \ + stf.spill.nta [r3]=f93,32; \ + ;; \ + stf.spill.nta [r2]=f94,32; \ + stf.spill.nta [r3]=f95,32; \ + ;; \ + stf.spill.nta [r2]=f96,32; \ + stf.spill.nta [r3]=f97,32; \ + ;; \ + stf.spill.nta [r2]=f98,32; \ + stf.spill.nta [r3]=f99,32; \ + ;; \ + stf.spill.nta [r2]=f100,32; \ + stf.spill.nta [r3]=f101,32; \ + ;; \ + stf.spill.nta [r2]=f102,32; \ + stf.spill.nta [r3]=f103,32; \ + ;; \ + stf.spill.nta [r2]=f104,32; \ + stf.spill.nta [r3]=f105,32; \ + ;; \ + stf.spill.nta [r2]=f106,32; \ + stf.spill.nta [r3]=f107,32; \ + ;; \ + stf.spill.nta [r2]=f108,32; \ + stf.spill.nta [r3]=f109,32; \ + ;; \ + stf.spill.nta [r2]=f110,32; \ + stf.spill.nta [r3]=f111,32; \ + ;; \ + stf.spill.nta [r2]=f112,32; \ + stf.spill.nta [r3]=f113,32; \ + ;; \ + stf.spill.nta [r2]=f114,32; \ + stf.spill.nta [r3]=f115,32; \ + ;; \ + stf.spill.nta [r2]=f116,32; \ + stf.spill.nta [r3]=f117,32; \ + ;; \ + stf.spill.nta [r2]=f118,32; \ + stf.spill.nta [r3]=f119,32; \ + ;; \ + stf.spill.nta [r2]=f120,32; \ + stf.spill.nta [r3]=f121,32; \ + ;; \ + stf.spill.nta [r2]=f122,32; \ + stf.spill.nta [r3]=f123,32; \ + ;; \ + stf.spill.nta [r2]=f124,32; \ + stf.spill.nta [r3]=f125,32; \ + ;; \ + stf.spill.nta [r2]=f126; \ + stf.spill.nta [r3]=f127; \ + ;; + + /* + * r33: point to context_t structure + */ +#define RESTORE_FPU_LOW \ + add r2 = CTX(F2), r33; \ + add r3 = CTX(F3), r33; \ + ;; \ + ldf.fill.nta f2 = [r2], 32; \ + ldf.fill.nta f3 = [r3], 32; \ + ;; \ + ldf.fill.nta f4 = [r2], 32; \ + ldf.fill.nta f5 = [r3], 32; \ + ;; \ + ldf.fill.nta f6 = [r2], 32; \ + ldf.fill.nta f7 = [r3], 32; \ + ;; \ + ldf.fill.nta f8 = [r2], 32; \ + ldf.fill.nta f9 = [r3], 32; \ + ;; \ + ldf.fill.nta f10 = [r2], 32; \ + ldf.fill.nta f11 = [r3], 32; \ + ;; \ + ldf.fill.nta f12 = [r2], 32; \ + ldf.fill.nta f13 = [r3], 32; \ + ;; \ + ldf.fill.nta f14 = [r2], 32; \ + ldf.fill.nta f15 = [r3], 32; \ + ;; \ + ldf.fill.nta f16 = [r2], 32; \ + ldf.fill.nta f17 = [r3], 32; \ + ;; \ + ldf.fill.nta f18 = [r2], 32; \ + ldf.fill.nta f19 = [r3], 32; \ + ;; \ + ldf.fill.nta f20 = [r2], 32; \ + ldf.fill.nta f21 = [r3], 32; \ + ;; \ + ldf.fill.nta f22 = [r2], 32; \ + ldf.fill.nta f23 = [r3], 32; \ + ;; \ + ldf.fill.nta f24 = [r2], 32; \ + ldf.fill.nta f25 = [r3], 32; \ + ;; \ + ldf.fill.nta f26 = [r2], 32; \ + ldf.fill.nta f27 = [r3], 32; \ + ;; \ + ldf.fill.nta f28 = [r2], 32; \ + ldf.fill.nta f29 = [r3], 32; \ + ;; \ + ldf.fill.nta f30 = [r2], 32; \ + ldf.fill.nta f31 = [r3], 32; \ + ;; + + + + /* + * r33: point to context_t structure + */ +#define RESTORE_FPU_HIGH \ + add r2 = CTX(F32), r33; \ + add r3 = CTX(F33), r33; \ + ;; \ + ldf.fill.nta f32 = [r2], 32; \ + ldf.fill.nta f33 = [r3], 32; \ + ;; \ + ldf.fill.nta f34 = [r2], 32; \ + ldf.fill.nta f35 = [r3], 32; \ + ;; \ + ldf.fill.nta f36 = [r2], 32; \ + ldf.fill.nta f37 = [r3], 32; \ + ;; \ + ldf.fill.nta f38 = [r2], 32; \ + ldf.fill.nta f39 = [r3], 32; \ + ;; \ + ldf.fill.nta f40 = [r2], 32; \ + ldf.fill.nta f41 = [r3], 32; \ + ;; \ + ldf.fill.nta f42 = [r2], 32; \ + ldf.fill.nta f43 = [r3], 32; \ + ;; \ + ldf.fill.nta f44 = [r2], 32; \ + ldf.fill.nta f45 = [r3], 32; \ + ;; \ + ldf.fill.nta f46 = [r2], 32; \ + ldf.fill.nta f47 = [r3], 32; \ + ;; \ + ldf.fill.nta f48 = [r2], 32; \ + ldf.fill.nta f49 = [r3], 32; \ + ;; \ + ldf.fill.nta f50 = [r2], 32; \ + ldf.fill.nta f51 = [r3], 32; \ + ;; \ + ldf.fill.nta f52 = [r2], 32; \ + ldf.fill.nta f53 = [r3], 32; \ + ;; \ + ldf.fill.nta f54 = [r2], 32; \ + ldf.fill.nta f55 = [r3], 32; \ + ;; \ + ldf.fill.nta f56 = [r2], 32; \ + ldf.fill.nta f57 = [r3], 32; \ + ;; \ + ldf.fill.nta f58 = [r2], 32; \ + ldf.fill.nta f59 = [r3], 32; \ + ;; \ + ldf.fill.nta f60 = [r2], 32; \ + ldf.fill.nta f61 = [r3], 32; \ + ;; \ + ldf.fill.nta f62 = [r2], 32; \ + ldf.fill.nta f63 = [r3], 32; \ + ;; \ + ldf.fill.nta f64 = [r2], 32; \ + ldf.fill.nta f65 = [r3], 32; \ + ;; \ + ldf.fill.nta f66 = [r2], 32; \ + ldf.fill.nta f67 = [r3], 32; \ + ;; \ + ldf.fill.nta f68 = [r2], 32; \ + ldf.fill.nta f69 = [r3], 32; \ + ;; \ + ldf.fill.nta f70 = [r2], 32; \ + ldf.fill.nta f71 = [r3], 32; \ + ;; \ + ldf.fill.nta f72 = [r2], 32; \ + ldf.fill.nta f73 = [r3], 32; \ + ;; \ + ldf.fill.nta f74 = [r2], 32; \ + ldf.fill.nta f75 = [r3], 32; \ + ;; \ + ldf.fill.nta f76 = [r2], 32; \ + ldf.fill.nta f77 = [r3], 32; \ + ;; \ + ldf.fill.nta f78 = [r2], 32; \ + ldf.fill.nta f79 = [r3], 32; \ + ;; \ + ldf.fill.nta f80 = [r2], 32; \ + ldf.fill.nta f81 = [r3], 32; \ + ;; \ + ldf.fill.nta f82 = [r2], 32; \ + ldf.fill.nta f83 = [r3], 32; \ + ;; \ + ldf.fill.nta f84 = [r2], 32; \ + ldf.fill.nta f85 = [r3], 32; \ + ;; \ + ldf.fill.nta f86 = [r2], 32; \ + ldf.fill.nta f87 = [r3], 32; \ + ;; \ + ldf.fill.nta f88 = [r2], 32; \ + ldf.fill.nta f89 = [r3], 32; \ + ;; \ + ldf.fill.nta f90 = [r2], 32; \ + ldf.fill.nta f91 = [r3], 32; \ + ;; \ + ldf.fill.nta f92 = [r2], 32; \ + ldf.fill.nta f93 = [r3], 32; \ + ;; \ + ldf.fill.nta f94 = [r2], 32; \ + ldf.fill.nta f95 = [r3], 32; \ + ;; \ + ldf.fill.nta f96 = [r2], 32; \ + ldf.fill.nta f97 = [r3], 32; \ + ;; \ + ldf.fill.nta f98 = [r2], 32; \ + ldf.fill.nta f99 = [r3], 32; \ + ;; \ + ldf.fill.nta f100 = [r2], 32; \ + ldf.fill.nta f101 = [r3], 32; \ + ;; \ + ldf.fill.nta f102 = [r2], 32; \ + ldf.fill.nta f103 = [r3], 32; \ + ;; \ + ldf.fill.nta f104 = [r2], 32; \ + ldf.fill.nta f105 = [r3], 32; \ + ;; \ + ldf.fill.nta f106 = [r2], 32; \ + ldf.fill.nta f107 = [r3], 32; \ + ;; \ + ldf.fill.nta f108 = [r2], 32; \ + ldf.fill.nta f109 = [r3], 32; \ + ;; \ + ldf.fill.nta f110 = [r2], 32; \ + ldf.fill.nta f111 = [r3], 32; \ + ;; \ + ldf.fill.nta f112 = [r2], 32; \ + ldf.fill.nta f113 = [r3], 32; \ + ;; \ + ldf.fill.nta f114 = [r2], 32; \ + ldf.fill.nta f115 = [r3], 32; \ + ;; \ + ldf.fill.nta f116 = [r2], 32; \ + ldf.fill.nta f117 = [r3], 32; \ + ;; \ + ldf.fill.nta f118 = [r2], 32; \ + ldf.fill.nta f119 = [r3], 32; \ + ;; \ + ldf.fill.nta f120 = [r2], 32; \ + ldf.fill.nta f121 = [r3], 32; \ + ;; \ + ldf.fill.nta f122 = [r2], 32; \ + ldf.fill.nta f123 = [r3], 32; \ + ;; \ + ldf.fill.nta f124 = [r2], 32; \ + ldf.fill.nta f125 = [r3], 32; \ + ;; \ + ldf.fill.nta f126 = [r2], 32; \ + ldf.fill.nta f127 = [r3], 32; \ + ;; + + /* + * r32: context_t base address + */ +#define SAVE_PTK_REGS \ + add r2=CTX(PKR0), r32; \ + mov r16=7; \ + ;; \ + mov ar.lc=r16; \ + mov r17=r0; \ + ;; \ +1: \ + mov r18=pkr[r17]; \ + ;; \ + srlz.i; \ + ;; \ + st8 [r2]=r18, 8; \ + ;; \ + add r17 =1,r17; \ + ;; \ + br.cloop.sptk 1b; \ + ;; + +/* + * r33: point to context_t structure + * ar.lc are corrupted. + */ +#define RESTORE_PTK_REGS \ + add r2=CTX(PKR0), r33; \ + mov r16=7; \ + ;; \ + mov ar.lc=r16; \ + mov r17=r0; \ + ;; \ +1: \ + ld8 r18=[r2], 8; \ + ;; \ + mov pkr[r17]=r18; \ + ;; \ + srlz.i; \ + ;; \ + add r17 =1,r17; \ + ;; \ + br.cloop.sptk 1b; \ + ;; + + +/* + * void vmm_trampoline( context_t * from, + * context_t * to) + * + * from: r32 + * to: r33 + * note: interrupt disabled before call this function. + */ +GLOBAL_ENTRY(vmm_trampoline) + mov r16 = psr + adds r2 = CTX(PSR), r32 + ;; + st8 [r2] = r16, 8 // psr + mov r17 = pr + ;; + st8 [r2] = r17, 8 // pr + mov r18 = ar.unat + ;; + st8 [r2] = r18 + mov r17 = ar.rsc + ;; + adds r2 = CTX(RSC),r32 + ;; + st8 [r2]= r17 + mov ar.rsc =0 + flushrs + ;; + SAVE_GENERAL_REGS + ;; + SAVE_KERNEL_REGS + ;; + SAVE_APP_REGS + ;; + SAVE_BRANCH_REGS + ;; + SAVE_CTL_REGS + ;; + SAVE_REGION_REGS + ;; + //SAVE_DEBUG_REGS + ;; + rsm psr.dfl + ;; + srlz.d + ;; + SAVE_FPU_LOW + ;; + rsm psr.dfh + ;; + srlz.d + ;; + SAVE_FPU_HIGH + ;; + SAVE_PTK_REGS + ;; + RESTORE_PTK_REGS + ;; + RESTORE_FPU_HIGH + ;; + RESTORE_FPU_LOW + ;; + //RESTORE_DEBUG_REGS + ;; + RESTORE_REGION_REGS + ;; + RESTORE_CTL_REGS + ;; + RESTORE_BRANCH_REGS + ;; + RESTORE_APP_REGS + ;; + RESTORE_KERNEL_REGS + ;; + RESTORE_GENERAL_REGS + ;; + adds r2=CTX(PSR), r33 + ;; + ld8 r16=[r2], 8 // psr + ;; + mov psr.l=r16 + ;; + srlz.d + ;; + ld8 r16=[r2], 8 // pr + ;; + mov pr =r16,-1 + ld8 r16=[r2] // unat + ;; + mov ar.unat=r16 + ;; + adds r2=CTX(RSC),r33 + ;; + ld8 r16 =[r2] + ;; + mov ar.rsc = r16 + ;; + br.ret.sptk.few b0 +END(vmm_trampoline) diff --git a/arch/ia64/kvm/vcpu.c b/arch/ia64/kvm/vcpu.c new file mode 100644 index 0000000000000000000000000000000000000000..e44027ce5667e553b89abab2abb95012b98f8048 --- /dev/null +++ b/arch/ia64/kvm/vcpu.c @@ -0,0 +1,2163 @@ +/* + * kvm_vcpu.c: handling all virtual cpu related thing. + * Copyright (c) 2005, Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + * Shaofan Li (Susue Li) + * Yaozu Dong (Eddie Dong) (Eddie.dong@intel.com) + * Xuefei Xu (Anthony Xu) (Anthony.xu@intel.com) + * Xiantao Zhang + */ + +#include +#include + +#include +#include +#include +#include +#include +#include + +#include "asm-offsets.h" +#include "vcpu.h" + +/* + * Special notes: + * - Index by it/dt/rt sequence + * - Only existing mode transitions are allowed in this table + * - RSE is placed at lazy mode when emulating guest partial mode + * - If gva happens to be rr0 and rr4, only allowed case is identity + * mapping (gva=gpa), or panic! (How?) + */ +int mm_switch_table[8][8] = { + /* 2004/09/12(Kevin): Allow switch to self */ + /* + * (it,dt,rt): (0,0,0) -> (1,1,1) + * This kind of transition usually occurs in the very early + * stage of Linux boot up procedure. Another case is in efi + * and pal calls. (see "arch/ia64/kernel/head.S") + * + * (it,dt,rt): (0,0,0) -> (0,1,1) + * This kind of transition is found when OSYa exits efi boot + * service. Due to gva = gpa in this case (Same region), + * data access can be satisfied though itlb entry for physical + * emulation is hit. + */ + {SW_SELF, 0, 0, SW_NOP, 0, 0, 0, SW_P2V}, + {0, 0, 0, 0, 0, 0, 0, 0}, + {0, 0, 0, 0, 0, 0, 0, 0}, + /* + * (it,dt,rt): (0,1,1) -> (1,1,1) + * This kind of transition is found in OSYa. + * + * (it,dt,rt): (0,1,1) -> (0,0,0) + * This kind of transition is found in OSYa + */ + {SW_NOP, 0, 0, SW_SELF, 0, 0, 0, SW_P2V}, + /* (1,0,0)->(1,1,1) */ + {0, 0, 0, 0, 0, 0, 0, SW_P2V}, + /* + * (it,dt,rt): (1,0,1) -> (1,1,1) + * This kind of transition usually occurs when Linux returns + * from the low level TLB miss handlers. + * (see "arch/ia64/kernel/ivt.S") + */ + {0, 0, 0, 0, 0, SW_SELF, 0, SW_P2V}, + {0, 0, 0, 0, 0, 0, 0, 0}, + /* + * (it,dt,rt): (1,1,1) -> (1,0,1) + * This kind of transition usually occurs in Linux low level + * TLB miss handler. (see "arch/ia64/kernel/ivt.S") + * + * (it,dt,rt): (1,1,1) -> (0,0,0) + * This kind of transition usually occurs in pal and efi calls, + * which requires running in physical mode. + * (see "arch/ia64/kernel/head.S") + * (1,1,1)->(1,0,0) + */ + + {SW_V2P, 0, 0, 0, SW_V2P, SW_V2P, 0, SW_SELF}, +}; + +void physical_mode_init(struct kvm_vcpu *vcpu) +{ + vcpu->arch.mode_flags = GUEST_IN_PHY; +} + +void switch_to_physical_rid(struct kvm_vcpu *vcpu) +{ + unsigned long psr; + + /* Save original virtual mode rr[0] and rr[4] */ + psr = ia64_clear_ic(); + ia64_set_rr(VRN0<arch.metaphysical_rr0); + ia64_srlz_d(); + ia64_set_rr(VRN4<arch.metaphysical_rr4); + ia64_srlz_d(); + + ia64_set_psr(psr); + return; +} + + +void switch_to_virtual_rid(struct kvm_vcpu *vcpu) +{ + unsigned long psr; + + psr = ia64_clear_ic(); + ia64_set_rr(VRN0 << VRN_SHIFT, vcpu->arch.metaphysical_saved_rr0); + ia64_srlz_d(); + ia64_set_rr(VRN4 << VRN_SHIFT, vcpu->arch.metaphysical_saved_rr4); + ia64_srlz_d(); + ia64_set_psr(psr); + return; +} + +static int mm_switch_action(struct ia64_psr opsr, struct ia64_psr npsr) +{ + return mm_switch_table[MODE_IND(opsr)][MODE_IND(npsr)]; +} + +void switch_mm_mode(struct kvm_vcpu *vcpu, struct ia64_psr old_psr, + struct ia64_psr new_psr) +{ + int act; + act = mm_switch_action(old_psr, new_psr); + switch (act) { + case SW_V2P: + /*printk("V -> P mode transition: (0x%lx -> 0x%lx)\n", + old_psr.val, new_psr.val);*/ + switch_to_physical_rid(vcpu); + /* + * Set rse to enforced lazy, to prevent active rse + *save/restor when guest physical mode. + */ + vcpu->arch.mode_flags |= GUEST_IN_PHY; + break; + case SW_P2V: + switch_to_virtual_rid(vcpu); + /* + * recover old mode which is saved when entering + * guest physical mode + */ + vcpu->arch.mode_flags &= ~GUEST_IN_PHY; + break; + case SW_SELF: + break; + case SW_NOP: + break; + default: + /* Sanity check */ + break; + } + return; +} + + + +/* + * In physical mode, insert tc/tr for region 0 and 4 uses + * RID[0] and RID[4] which is for physical mode emulation. + * However what those inserted tc/tr wants is rid for + * virtual mode. So original virtual rid needs to be restored + * before insert. + * + * Operations which required such switch include: + * - insertions (itc.*, itr.*) + * - purges (ptc.* and ptr.*) + * - tpa + * - tak + * - thash?, ttag? + * All above needs actual virtual rid for destination entry. + */ + +void check_mm_mode_switch(struct kvm_vcpu *vcpu, struct ia64_psr old_psr, + struct ia64_psr new_psr) +{ + + if ((old_psr.dt != new_psr.dt) + || (old_psr.it != new_psr.it) + || (old_psr.rt != new_psr.rt)) + switch_mm_mode(vcpu, old_psr, new_psr); + + return; +} + + +/* + * In physical mode, insert tc/tr for region 0 and 4 uses + * RID[0] and RID[4] which is for physical mode emulation. + * However what those inserted tc/tr wants is rid for + * virtual mode. So original virtual rid needs to be restored + * before insert. + * + * Operations which required such switch include: + * - insertions (itc.*, itr.*) + * - purges (ptc.* and ptr.*) + * - tpa + * - tak + * - thash?, ttag? + * All above needs actual virtual rid for destination entry. + */ + +void prepare_if_physical_mode(struct kvm_vcpu *vcpu) +{ + if (is_physical_mode(vcpu)) { + vcpu->arch.mode_flags |= GUEST_PHY_EMUL; + switch_to_virtual_rid(vcpu); + } + return; +} + +/* Recover always follows prepare */ +void recover_if_physical_mode(struct kvm_vcpu *vcpu) +{ + if (is_physical_mode(vcpu)) + switch_to_physical_rid(vcpu); + vcpu->arch.mode_flags &= ~GUEST_PHY_EMUL; + return; +} + +#define RPT(x) ((u16) &((struct kvm_pt_regs *)0)->x) + +static u16 gr_info[32] = { + 0, /* r0 is read-only : WE SHOULD NEVER GET THIS */ + RPT(r1), RPT(r2), RPT(r3), + RPT(r4), RPT(r5), RPT(r6), RPT(r7), + RPT(r8), RPT(r9), RPT(r10), RPT(r11), + RPT(r12), RPT(r13), RPT(r14), RPT(r15), + RPT(r16), RPT(r17), RPT(r18), RPT(r19), + RPT(r20), RPT(r21), RPT(r22), RPT(r23), + RPT(r24), RPT(r25), RPT(r26), RPT(r27), + RPT(r28), RPT(r29), RPT(r30), RPT(r31) +}; + +#define IA64_FIRST_STACKED_GR 32 +#define IA64_FIRST_ROTATING_FR 32 + +static inline unsigned long +rotate_reg(unsigned long sor, unsigned long rrb, unsigned long reg) +{ + reg += rrb; + if (reg >= sor) + reg -= sor; + return reg; +} + +/* + * Return the (rotated) index for floating point register + * be in the REGNUM (REGNUM must range from 32-127, + * result is in the range from 0-95. + */ +static inline unsigned long fph_index(struct kvm_pt_regs *regs, + long regnum) +{ + unsigned long rrb_fr = (regs->cr_ifs >> 25) & 0x7f; + return rotate_reg(96, rrb_fr, (regnum - IA64_FIRST_ROTATING_FR)); +} + + +/* + * The inverse of the above: given bspstore and the number of + * registers, calculate ar.bsp. + */ +static inline unsigned long *kvm_rse_skip_regs(unsigned long *addr, + long num_regs) +{ + long delta = ia64_rse_slot_num(addr) + num_regs; + int i = 0; + + if (num_regs < 0) + delta -= 0x3e; + if (delta < 0) { + while (delta <= -0x3f) { + i--; + delta += 0x3f; + } + } else { + while (delta >= 0x3f) { + i++; + delta -= 0x3f; + } + } + + return addr + num_regs + i; +} + +static void get_rse_reg(struct kvm_pt_regs *regs, unsigned long r1, + unsigned long *val, int *nat) +{ + unsigned long *bsp, *addr, *rnat_addr, *bspstore; + unsigned long *kbs = (void *) current_vcpu + VMM_RBS_OFFSET; + unsigned long nat_mask; + unsigned long old_rsc, new_rsc; + long sof = (regs->cr_ifs) & 0x7f; + long sor = (((regs->cr_ifs >> 14) & 0xf) << 3); + long rrb_gr = (regs->cr_ifs >> 18) & 0x7f; + long ridx = r1 - 32; + + if (ridx < sor) + ridx = rotate_reg(sor, rrb_gr, ridx); + + old_rsc = ia64_getreg(_IA64_REG_AR_RSC); + new_rsc = old_rsc&(~(0x3)); + ia64_setreg(_IA64_REG_AR_RSC, new_rsc); + + bspstore = (unsigned long *)ia64_getreg(_IA64_REG_AR_BSPSTORE); + bsp = kbs + (regs->loadrs >> 19); + + addr = kvm_rse_skip_regs(bsp, -sof + ridx); + nat_mask = 1UL << ia64_rse_slot_num(addr); + rnat_addr = ia64_rse_rnat_addr(addr); + + if (addr >= bspstore) { + ia64_flushrs(); + ia64_mf(); + bspstore = (unsigned long *)ia64_getreg(_IA64_REG_AR_BSPSTORE); + } + *val = *addr; + if (nat) { + if (bspstore < rnat_addr) + *nat = (int)!!(ia64_getreg(_IA64_REG_AR_RNAT) + & nat_mask); + else + *nat = (int)!!((*rnat_addr) & nat_mask); + ia64_setreg(_IA64_REG_AR_RSC, old_rsc); + } +} + +void set_rse_reg(struct kvm_pt_regs *regs, unsigned long r1, + unsigned long val, unsigned long nat) +{ + unsigned long *bsp, *bspstore, *addr, *rnat_addr; + unsigned long *kbs = (void *) current_vcpu + VMM_RBS_OFFSET; + unsigned long nat_mask; + unsigned long old_rsc, new_rsc, psr; + unsigned long rnat; + long sof = (regs->cr_ifs) & 0x7f; + long sor = (((regs->cr_ifs >> 14) & 0xf) << 3); + long rrb_gr = (regs->cr_ifs >> 18) & 0x7f; + long ridx = r1 - 32; + + if (ridx < sor) + ridx = rotate_reg(sor, rrb_gr, ridx); + + old_rsc = ia64_getreg(_IA64_REG_AR_RSC); + /* put RSC to lazy mode, and set loadrs 0 */ + new_rsc = old_rsc & (~0x3fff0003); + ia64_setreg(_IA64_REG_AR_RSC, new_rsc); + bsp = kbs + (regs->loadrs >> 19); /* 16 + 3 */ + + addr = kvm_rse_skip_regs(bsp, -sof + ridx); + nat_mask = 1UL << ia64_rse_slot_num(addr); + rnat_addr = ia64_rse_rnat_addr(addr); + + local_irq_save(psr); + bspstore = (unsigned long *)ia64_getreg(_IA64_REG_AR_BSPSTORE); + if (addr >= bspstore) { + + ia64_flushrs(); + ia64_mf(); + *addr = val; + bspstore = (unsigned long *)ia64_getreg(_IA64_REG_AR_BSPSTORE); + rnat = ia64_getreg(_IA64_REG_AR_RNAT); + if (bspstore < rnat_addr) + rnat = rnat & (~nat_mask); + else + *rnat_addr = (*rnat_addr)&(~nat_mask); + + ia64_mf(); + ia64_loadrs(); + ia64_setreg(_IA64_REG_AR_RNAT, rnat); + } else { + rnat = ia64_getreg(_IA64_REG_AR_RNAT); + *addr = val; + if (bspstore < rnat_addr) + rnat = rnat&(~nat_mask); + else + *rnat_addr = (*rnat_addr) & (~nat_mask); + + ia64_setreg(_IA64_REG_AR_BSPSTORE, bspstore); + ia64_setreg(_IA64_REG_AR_RNAT, rnat); + } + local_irq_restore(psr); + ia64_setreg(_IA64_REG_AR_RSC, old_rsc); +} + +void getreg(unsigned long regnum, unsigned long *val, + int *nat, struct kvm_pt_regs *regs) +{ + unsigned long addr, *unat; + if (regnum >= IA64_FIRST_STACKED_GR) { + get_rse_reg(regs, regnum, val, nat); + return; + } + + /* + * Now look at registers in [0-31] range and init correct UNAT + */ + addr = (unsigned long)regs; + unat = ®s->eml_unat;; + + addr += gr_info[regnum]; + + *val = *(unsigned long *)addr; + /* + * do it only when requested + */ + if (nat) + *nat = (*unat >> ((addr >> 3) & 0x3f)) & 0x1UL; +} + +void setreg(unsigned long regnum, unsigned long val, + int nat, struct kvm_pt_regs *regs) +{ + unsigned long addr; + unsigned long bitmask; + unsigned long *unat; + + /* + * First takes care of stacked registers + */ + if (regnum >= IA64_FIRST_STACKED_GR) { + set_rse_reg(regs, regnum, val, nat); + return; + } + + /* + * Now look at registers in [0-31] range and init correct UNAT + */ + addr = (unsigned long)regs; + unat = ®s->eml_unat; + /* + * add offset from base of struct + * and do it ! + */ + addr += gr_info[regnum]; + + *(unsigned long *)addr = val; + + /* + * We need to clear the corresponding UNAT bit to fully emulate the load + * UNAT bit_pos = GR[r3]{8:3} form EAS-2.4 + */ + bitmask = 1UL << ((addr >> 3) & 0x3f); + if (nat) + *unat |= bitmask; + else + *unat &= ~bitmask; + +} + +u64 vcpu_get_gr(struct kvm_vcpu *vcpu, unsigned long reg) +{ + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + u64 val; + + if (!reg) + return 0; + getreg(reg, &val, 0, regs); + return val; +} + +void vcpu_set_gr(struct kvm_vcpu *vcpu, u64 reg, u64 value, int nat) +{ + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + long sof = (regs->cr_ifs) & 0x7f; + + if (!reg) + return; + if (reg >= sof + 32) + return; + setreg(reg, value, nat, regs); /* FIXME: handle NATs later*/ +} + +void getfpreg(unsigned long regnum, struct ia64_fpreg *fpval, + struct kvm_pt_regs *regs) +{ + /* Take floating register rotation into consideration*/ + if (regnum >= IA64_FIRST_ROTATING_FR) + regnum = IA64_FIRST_ROTATING_FR + fph_index(regs, regnum); +#define CASE_FIXED_FP(reg) \ + case (reg) : \ + ia64_stf_spill(fpval, reg); \ + break + + switch (regnum) { + CASE_FIXED_FP(0); + CASE_FIXED_FP(1); + CASE_FIXED_FP(2); + CASE_FIXED_FP(3); + CASE_FIXED_FP(4); + CASE_FIXED_FP(5); + + CASE_FIXED_FP(6); + CASE_FIXED_FP(7); + CASE_FIXED_FP(8); + CASE_FIXED_FP(9); + CASE_FIXED_FP(10); + CASE_FIXED_FP(11); + + CASE_FIXED_FP(12); + CASE_FIXED_FP(13); + CASE_FIXED_FP(14); + CASE_FIXED_FP(15); + CASE_FIXED_FP(16); + CASE_FIXED_FP(17); + CASE_FIXED_FP(18); + CASE_FIXED_FP(19); + CASE_FIXED_FP(20); + CASE_FIXED_FP(21); + CASE_FIXED_FP(22); + CASE_FIXED_FP(23); + CASE_FIXED_FP(24); + CASE_FIXED_FP(25); + CASE_FIXED_FP(26); + CASE_FIXED_FP(27); + CASE_FIXED_FP(28); + CASE_FIXED_FP(29); + CASE_FIXED_FP(30); + CASE_FIXED_FP(31); + CASE_FIXED_FP(32); + CASE_FIXED_FP(33); + CASE_FIXED_FP(34); + CASE_FIXED_FP(35); + CASE_FIXED_FP(36); + CASE_FIXED_FP(37); + CASE_FIXED_FP(38); + CASE_FIXED_FP(39); + CASE_FIXED_FP(40); + CASE_FIXED_FP(41); + CASE_FIXED_FP(42); + CASE_FIXED_FP(43); + CASE_FIXED_FP(44); + CASE_FIXED_FP(45); + CASE_FIXED_FP(46); + CASE_FIXED_FP(47); + CASE_FIXED_FP(48); + CASE_FIXED_FP(49); + CASE_FIXED_FP(50); + CASE_FIXED_FP(51); + CASE_FIXED_FP(52); + CASE_FIXED_FP(53); + CASE_FIXED_FP(54); + CASE_FIXED_FP(55); + CASE_FIXED_FP(56); + CASE_FIXED_FP(57); + CASE_FIXED_FP(58); + CASE_FIXED_FP(59); + CASE_FIXED_FP(60); + CASE_FIXED_FP(61); + CASE_FIXED_FP(62); + CASE_FIXED_FP(63); + CASE_FIXED_FP(64); + CASE_FIXED_FP(65); + CASE_FIXED_FP(66); + CASE_FIXED_FP(67); + CASE_FIXED_FP(68); + CASE_FIXED_FP(69); + CASE_FIXED_FP(70); + CASE_FIXED_FP(71); + CASE_FIXED_FP(72); + CASE_FIXED_FP(73); + CASE_FIXED_FP(74); + CASE_FIXED_FP(75); + CASE_FIXED_FP(76); + CASE_FIXED_FP(77); + CASE_FIXED_FP(78); + CASE_FIXED_FP(79); + CASE_FIXED_FP(80); + CASE_FIXED_FP(81); + CASE_FIXED_FP(82); + CASE_FIXED_FP(83); + CASE_FIXED_FP(84); + CASE_FIXED_FP(85); + CASE_FIXED_FP(86); + CASE_FIXED_FP(87); + CASE_FIXED_FP(88); + CASE_FIXED_FP(89); + CASE_FIXED_FP(90); + CASE_FIXED_FP(91); + CASE_FIXED_FP(92); + CASE_FIXED_FP(93); + CASE_FIXED_FP(94); + CASE_FIXED_FP(95); + CASE_FIXED_FP(96); + CASE_FIXED_FP(97); + CASE_FIXED_FP(98); + CASE_FIXED_FP(99); + CASE_FIXED_FP(100); + CASE_FIXED_FP(101); + CASE_FIXED_FP(102); + CASE_FIXED_FP(103); + CASE_FIXED_FP(104); + CASE_FIXED_FP(105); + CASE_FIXED_FP(106); + CASE_FIXED_FP(107); + CASE_FIXED_FP(108); + CASE_FIXED_FP(109); + CASE_FIXED_FP(110); + CASE_FIXED_FP(111); + CASE_FIXED_FP(112); + CASE_FIXED_FP(113); + CASE_FIXED_FP(114); + CASE_FIXED_FP(115); + CASE_FIXED_FP(116); + CASE_FIXED_FP(117); + CASE_FIXED_FP(118); + CASE_FIXED_FP(119); + CASE_FIXED_FP(120); + CASE_FIXED_FP(121); + CASE_FIXED_FP(122); + CASE_FIXED_FP(123); + CASE_FIXED_FP(124); + CASE_FIXED_FP(125); + CASE_FIXED_FP(126); + CASE_FIXED_FP(127); + } +#undef CASE_FIXED_FP +} + +void setfpreg(unsigned long regnum, struct ia64_fpreg *fpval, + struct kvm_pt_regs *regs) +{ + /* Take floating register rotation into consideration*/ + if (regnum >= IA64_FIRST_ROTATING_FR) + regnum = IA64_FIRST_ROTATING_FR + fph_index(regs, regnum); + +#define CASE_FIXED_FP(reg) \ + case (reg) : \ + ia64_ldf_fill(reg, fpval); \ + break + + switch (regnum) { + CASE_FIXED_FP(2); + CASE_FIXED_FP(3); + CASE_FIXED_FP(4); + CASE_FIXED_FP(5); + + CASE_FIXED_FP(6); + CASE_FIXED_FP(7); + CASE_FIXED_FP(8); + CASE_FIXED_FP(9); + CASE_FIXED_FP(10); + CASE_FIXED_FP(11); + + CASE_FIXED_FP(12); + CASE_FIXED_FP(13); + CASE_FIXED_FP(14); + CASE_FIXED_FP(15); + CASE_FIXED_FP(16); + CASE_FIXED_FP(17); + CASE_FIXED_FP(18); + CASE_FIXED_FP(19); + CASE_FIXED_FP(20); + CASE_FIXED_FP(21); + CASE_FIXED_FP(22); + CASE_FIXED_FP(23); + CASE_FIXED_FP(24); + CASE_FIXED_FP(25); + CASE_FIXED_FP(26); + CASE_FIXED_FP(27); + CASE_FIXED_FP(28); + CASE_FIXED_FP(29); + CASE_FIXED_FP(30); + CASE_FIXED_FP(31); + CASE_FIXED_FP(32); + CASE_FIXED_FP(33); + CASE_FIXED_FP(34); + CASE_FIXED_FP(35); + CASE_FIXED_FP(36); + CASE_FIXED_FP(37); + CASE_FIXED_FP(38); + CASE_FIXED_FP(39); + CASE_FIXED_FP(40); + CASE_FIXED_FP(41); + CASE_FIXED_FP(42); + CASE_FIXED_FP(43); + CASE_FIXED_FP(44); + CASE_FIXED_FP(45); + CASE_FIXED_FP(46); + CASE_FIXED_FP(47); + CASE_FIXED_FP(48); + CASE_FIXED_FP(49); + CASE_FIXED_FP(50); + CASE_FIXED_FP(51); + CASE_FIXED_FP(52); + CASE_FIXED_FP(53); + CASE_FIXED_FP(54); + CASE_FIXED_FP(55); + CASE_FIXED_FP(56); + CASE_FIXED_FP(57); + CASE_FIXED_FP(58); + CASE_FIXED_FP(59); + CASE_FIXED_FP(60); + CASE_FIXED_FP(61); + CASE_FIXED_FP(62); + CASE_FIXED_FP(63); + CASE_FIXED_FP(64); + CASE_FIXED_FP(65); + CASE_FIXED_FP(66); + CASE_FIXED_FP(67); + CASE_FIXED_FP(68); + CASE_FIXED_FP(69); + CASE_FIXED_FP(70); + CASE_FIXED_FP(71); + CASE_FIXED_FP(72); + CASE_FIXED_FP(73); + CASE_FIXED_FP(74); + CASE_FIXED_FP(75); + CASE_FIXED_FP(76); + CASE_FIXED_FP(77); + CASE_FIXED_FP(78); + CASE_FIXED_FP(79); + CASE_FIXED_FP(80); + CASE_FIXED_FP(81); + CASE_FIXED_FP(82); + CASE_FIXED_FP(83); + CASE_FIXED_FP(84); + CASE_FIXED_FP(85); + CASE_FIXED_FP(86); + CASE_FIXED_FP(87); + CASE_FIXED_FP(88); + CASE_FIXED_FP(89); + CASE_FIXED_FP(90); + CASE_FIXED_FP(91); + CASE_FIXED_FP(92); + CASE_FIXED_FP(93); + CASE_FIXED_FP(94); + CASE_FIXED_FP(95); + CASE_FIXED_FP(96); + CASE_FIXED_FP(97); + CASE_FIXED_FP(98); + CASE_FIXED_FP(99); + CASE_FIXED_FP(100); + CASE_FIXED_FP(101); + CASE_FIXED_FP(102); + CASE_FIXED_FP(103); + CASE_FIXED_FP(104); + CASE_FIXED_FP(105); + CASE_FIXED_FP(106); + CASE_FIXED_FP(107); + CASE_FIXED_FP(108); + CASE_FIXED_FP(109); + CASE_FIXED_FP(110); + CASE_FIXED_FP(111); + CASE_FIXED_FP(112); + CASE_FIXED_FP(113); + CASE_FIXED_FP(114); + CASE_FIXED_FP(115); + CASE_FIXED_FP(116); + CASE_FIXED_FP(117); + CASE_FIXED_FP(118); + CASE_FIXED_FP(119); + CASE_FIXED_FP(120); + CASE_FIXED_FP(121); + CASE_FIXED_FP(122); + CASE_FIXED_FP(123); + CASE_FIXED_FP(124); + CASE_FIXED_FP(125); + CASE_FIXED_FP(126); + CASE_FIXED_FP(127); + } +} + +void vcpu_get_fpreg(struct kvm_vcpu *vcpu, unsigned long reg, + struct ia64_fpreg *val) +{ + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + getfpreg(reg, val, regs); /* FIXME: handle NATs later*/ +} + +void vcpu_set_fpreg(struct kvm_vcpu *vcpu, unsigned long reg, + struct ia64_fpreg *val) +{ + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + if (reg > 1) + setfpreg(reg, val, regs); /* FIXME: handle NATs later*/ +} + +/************************************************************************ + * lsapic timer + ***********************************************************************/ +u64 vcpu_get_itc(struct kvm_vcpu *vcpu) +{ + unsigned long guest_itc; + guest_itc = VMX(vcpu, itc_offset) + ia64_getreg(_IA64_REG_AR_ITC); + + if (guest_itc >= VMX(vcpu, last_itc)) { + VMX(vcpu, last_itc) = guest_itc; + return guest_itc; + } else + return VMX(vcpu, last_itc); +} + +static inline void vcpu_set_itm(struct kvm_vcpu *vcpu, u64 val); +static void vcpu_set_itc(struct kvm_vcpu *vcpu, u64 val) +{ + struct kvm_vcpu *v; + int i; + long itc_offset = val - ia64_getreg(_IA64_REG_AR_ITC); + unsigned long vitv = VCPU(vcpu, itv); + + if (vcpu->vcpu_id == 0) { + for (i = 0; i < MAX_VCPU_NUM; i++) { + v = (struct kvm_vcpu *)((char *)vcpu + VCPU_SIZE * i); + VMX(v, itc_offset) = itc_offset; + VMX(v, last_itc) = 0; + } + } + VMX(vcpu, last_itc) = 0; + if (VCPU(vcpu, itm) <= val) { + VMX(vcpu, itc_check) = 0; + vcpu_unpend_interrupt(vcpu, vitv); + } else { + VMX(vcpu, itc_check) = 1; + vcpu_set_itm(vcpu, VCPU(vcpu, itm)); + } + +} + +static inline u64 vcpu_get_itm(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, itm)); +} + +static inline void vcpu_set_itm(struct kvm_vcpu *vcpu, u64 val) +{ + unsigned long vitv = VCPU(vcpu, itv); + VCPU(vcpu, itm) = val; + + if (val > vcpu_get_itc(vcpu)) { + VMX(vcpu, itc_check) = 1; + vcpu_unpend_interrupt(vcpu, vitv); + VMX(vcpu, timer_pending) = 0; + } else + VMX(vcpu, itc_check) = 0; +} + +#define ITV_VECTOR(itv) (itv&0xff) +#define ITV_IRQ_MASK(itv) (itv&(1<<16)) + +static inline void vcpu_set_itv(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, itv) = val; + if (!ITV_IRQ_MASK(val) && vcpu->arch.timer_pending) { + vcpu_pend_interrupt(vcpu, ITV_VECTOR(val)); + vcpu->arch.timer_pending = 0; + } +} + +static inline void vcpu_set_eoi(struct kvm_vcpu *vcpu, u64 val) +{ + int vec; + + vec = highest_inservice_irq(vcpu); + if (vec == NULL_VECTOR) + return; + VMX(vcpu, insvc[vec >> 6]) &= ~(1UL << (vec & 63)); + VCPU(vcpu, eoi) = 0; + vcpu->arch.irq_new_pending = 1; + +} + +/* See Table 5-8 in SDM vol2 for the definition */ +int irq_masked(struct kvm_vcpu *vcpu, int h_pending, int h_inservice) +{ + union ia64_tpr vtpr; + + vtpr.val = VCPU(vcpu, tpr); + + if (h_inservice == NMI_VECTOR) + return IRQ_MASKED_BY_INSVC; + + if (h_pending == NMI_VECTOR) { + /* Non Maskable Interrupt */ + return IRQ_NO_MASKED; + } + + if (h_inservice == ExtINT_VECTOR) + return IRQ_MASKED_BY_INSVC; + + if (h_pending == ExtINT_VECTOR) { + if (vtpr.mmi) { + /* mask all external IRQ */ + return IRQ_MASKED_BY_VTPR; + } else + return IRQ_NO_MASKED; + } + + if (is_higher_irq(h_pending, h_inservice)) { + if (is_higher_class(h_pending, vtpr.mic + (vtpr.mmi << 4))) + return IRQ_NO_MASKED; + else + return IRQ_MASKED_BY_VTPR; + } else { + return IRQ_MASKED_BY_INSVC; + } +} + +void vcpu_pend_interrupt(struct kvm_vcpu *vcpu, u8 vec) +{ + long spsr; + int ret; + + local_irq_save(spsr); + ret = test_and_set_bit(vec, &VCPU(vcpu, irr[0])); + local_irq_restore(spsr); + + vcpu->arch.irq_new_pending = 1; +} + +void vcpu_unpend_interrupt(struct kvm_vcpu *vcpu, u8 vec) +{ + long spsr; + int ret; + + local_irq_save(spsr); + ret = test_and_clear_bit(vec, &VCPU(vcpu, irr[0])); + local_irq_restore(spsr); + if (ret) { + vcpu->arch.irq_new_pending = 1; + wmb(); + } +} + +void update_vhpi(struct kvm_vcpu *vcpu, int vec) +{ + u64 vhpi; + + if (vec == NULL_VECTOR) + vhpi = 0; + else if (vec == NMI_VECTOR) + vhpi = 32; + else if (vec == ExtINT_VECTOR) + vhpi = 16; + else + vhpi = vec >> 4; + + VCPU(vcpu, vhpi) = vhpi; + if (VCPU(vcpu, vac).a_int) + ia64_call_vsa(PAL_VPS_SET_PENDING_INTERRUPT, + (u64)vcpu->arch.vpd, 0, 0, 0, 0, 0, 0); +} + +u64 vcpu_get_ivr(struct kvm_vcpu *vcpu) +{ + int vec, h_inservice, mask; + + vec = highest_pending_irq(vcpu); + h_inservice = highest_inservice_irq(vcpu); + mask = irq_masked(vcpu, vec, h_inservice); + if (vec == NULL_VECTOR || mask == IRQ_MASKED_BY_INSVC) { + if (VCPU(vcpu, vhpi)) + update_vhpi(vcpu, NULL_VECTOR); + return IA64_SPURIOUS_INT_VECTOR; + } + if (mask == IRQ_MASKED_BY_VTPR) { + update_vhpi(vcpu, vec); + return IA64_SPURIOUS_INT_VECTOR; + } + VMX(vcpu, insvc[vec >> 6]) |= (1UL << (vec & 63)); + vcpu_unpend_interrupt(vcpu, vec); + return (u64)vec; +} + +/************************************************************************** + Privileged operation emulation routines + **************************************************************************/ +u64 vcpu_thash(struct kvm_vcpu *vcpu, u64 vadr) +{ + union ia64_pta vpta; + union ia64_rr vrr; + u64 pval; + u64 vhpt_offset; + + vpta.val = vcpu_get_pta(vcpu); + vrr.val = vcpu_get_rr(vcpu, vadr); + vhpt_offset = ((vadr >> vrr.ps) << 3) & ((1UL << (vpta.size)) - 1); + if (vpta.vf) { + pval = ia64_call_vsa(PAL_VPS_THASH, vadr, vrr.val, + vpta.val, 0, 0, 0, 0); + } else { + pval = (vadr & VRN_MASK) | vhpt_offset | + (vpta.val << 3 >> (vpta.size + 3) << (vpta.size)); + } + return pval; +} + +u64 vcpu_ttag(struct kvm_vcpu *vcpu, u64 vadr) +{ + union ia64_rr vrr; + union ia64_pta vpta; + u64 pval; + + vpta.val = vcpu_get_pta(vcpu); + vrr.val = vcpu_get_rr(vcpu, vadr); + if (vpta.vf) { + pval = ia64_call_vsa(PAL_VPS_TTAG, vadr, vrr.val, + 0, 0, 0, 0, 0); + } else + pval = 1; + + return pval; +} + +u64 vcpu_tak(struct kvm_vcpu *vcpu, u64 vadr) +{ + struct thash_data *data; + union ia64_pta vpta; + u64 key; + + vpta.val = vcpu_get_pta(vcpu); + if (vpta.vf == 0) { + key = 1; + return key; + } + data = vtlb_lookup(vcpu, vadr, D_TLB); + if (!data || !data->p) + key = 1; + else + key = data->key; + + return key; +} + + + +void kvm_thash(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long thash, vadr; + + vadr = vcpu_get_gr(vcpu, inst.M46.r3); + thash = vcpu_thash(vcpu, vadr); + vcpu_set_gr(vcpu, inst.M46.r1, thash, 0); +} + + +void kvm_ttag(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long tag, vadr; + + vadr = vcpu_get_gr(vcpu, inst.M46.r3); + tag = vcpu_ttag(vcpu, vadr); + vcpu_set_gr(vcpu, inst.M46.r1, tag, 0); +} + +int vcpu_tpa(struct kvm_vcpu *vcpu, u64 vadr, u64 *padr) +{ + struct thash_data *data; + union ia64_isr visr, pt_isr; + struct kvm_pt_regs *regs; + struct ia64_psr vpsr; + + regs = vcpu_regs(vcpu); + pt_isr.val = VMX(vcpu, cr_isr); + visr.val = 0; + visr.ei = pt_isr.ei; + visr.ir = pt_isr.ir; + vpsr = *(struct ia64_psr *)&VCPU(vcpu, vpsr); + visr.na = 1; + + data = vhpt_lookup(vadr); + if (data) { + if (data->p == 0) { + vcpu_set_isr(vcpu, visr.val); + data_page_not_present(vcpu, vadr); + return IA64_FAULT; + } else if (data->ma == VA_MATTR_NATPAGE) { + vcpu_set_isr(vcpu, visr.val); + dnat_page_consumption(vcpu, vadr); + return IA64_FAULT; + } else { + *padr = (data->gpaddr >> data->ps << data->ps) | + (vadr & (PSIZE(data->ps) - 1)); + return IA64_NO_FAULT; + } + } + + data = vtlb_lookup(vcpu, vadr, D_TLB); + if (data) { + if (data->p == 0) { + vcpu_set_isr(vcpu, visr.val); + data_page_not_present(vcpu, vadr); + return IA64_FAULT; + } else if (data->ma == VA_MATTR_NATPAGE) { + vcpu_set_isr(vcpu, visr.val); + dnat_page_consumption(vcpu, vadr); + return IA64_FAULT; + } else{ + *padr = ((data->ppn >> (data->ps - 12)) << data->ps) + | (vadr & (PSIZE(data->ps) - 1)); + return IA64_NO_FAULT; + } + } + if (!vhpt_enabled(vcpu, vadr, NA_REF)) { + if (vpsr.ic) { + vcpu_set_isr(vcpu, visr.val); + alt_dtlb(vcpu, vadr); + return IA64_FAULT; + } else { + nested_dtlb(vcpu); + return IA64_FAULT; + } + } else { + if (vpsr.ic) { + vcpu_set_isr(vcpu, visr.val); + dvhpt_fault(vcpu, vadr); + return IA64_FAULT; + } else{ + nested_dtlb(vcpu); + return IA64_FAULT; + } + } + + return IA64_NO_FAULT; +} + + +int kvm_tpa(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r1, r3; + + r3 = vcpu_get_gr(vcpu, inst.M46.r3); + + if (vcpu_tpa(vcpu, r3, &r1)) + return IA64_FAULT; + + vcpu_set_gr(vcpu, inst.M46.r1, r1, 0); + return(IA64_NO_FAULT); +} + +void kvm_tak(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r1, r3; + + r3 = vcpu_get_gr(vcpu, inst.M46.r3); + r1 = vcpu_tak(vcpu, r3); + vcpu_set_gr(vcpu, inst.M46.r1, r1, 0); +} + + +/************************************ + * Insert/Purge translation register/cache + ************************************/ +void vcpu_itc_i(struct kvm_vcpu *vcpu, u64 pte, u64 itir, u64 ifa) +{ + thash_purge_and_insert(vcpu, pte, itir, ifa, I_TLB); +} + +void vcpu_itc_d(struct kvm_vcpu *vcpu, u64 pte, u64 itir, u64 ifa) +{ + thash_purge_and_insert(vcpu, pte, itir, ifa, D_TLB); +} + +void vcpu_itr_i(struct kvm_vcpu *vcpu, u64 slot, u64 pte, u64 itir, u64 ifa) +{ + u64 ps, va, rid; + struct thash_data *p_itr; + + ps = itir_ps(itir); + va = PAGEALIGN(ifa, ps); + pte &= ~PAGE_FLAGS_RV_MASK; + rid = vcpu_get_rr(vcpu, ifa); + rid = rid & RR_RID_MASK; + p_itr = (struct thash_data *)&vcpu->arch.itrs[slot]; + vcpu_set_tr(p_itr, pte, itir, va, rid); + vcpu_quick_region_set(VMX(vcpu, itr_regions), va); +} + + +void vcpu_itr_d(struct kvm_vcpu *vcpu, u64 slot, u64 pte, u64 itir, u64 ifa) +{ + u64 gpfn; + u64 ps, va, rid; + struct thash_data *p_dtr; + + ps = itir_ps(itir); + va = PAGEALIGN(ifa, ps); + pte &= ~PAGE_FLAGS_RV_MASK; + + if (ps != _PAGE_SIZE_16M) + thash_purge_entries(vcpu, va, ps); + gpfn = (pte & _PAGE_PPN_MASK) >> PAGE_SHIFT; + if (__gpfn_is_io(gpfn)) + pte |= VTLB_PTE_IO; + rid = vcpu_get_rr(vcpu, va); + rid = rid & RR_RID_MASK; + p_dtr = (struct thash_data *)&vcpu->arch.dtrs[slot]; + vcpu_set_tr((struct thash_data *)&vcpu->arch.dtrs[slot], + pte, itir, va, rid); + vcpu_quick_region_set(VMX(vcpu, dtr_regions), va); +} + +void vcpu_ptr_d(struct kvm_vcpu *vcpu, u64 ifa, u64 ps) +{ + int index; + u64 va; + + va = PAGEALIGN(ifa, ps); + while ((index = vtr_find_overlap(vcpu, va, ps, D_TLB)) >= 0) + vcpu->arch.dtrs[index].page_flags = 0; + + thash_purge_entries(vcpu, va, ps); +} + +void vcpu_ptr_i(struct kvm_vcpu *vcpu, u64 ifa, u64 ps) +{ + int index; + u64 va; + + va = PAGEALIGN(ifa, ps); + while ((index = vtr_find_overlap(vcpu, va, ps, I_TLB)) >= 0) + vcpu->arch.itrs[index].page_flags = 0; + + thash_purge_entries(vcpu, va, ps); +} + +void vcpu_ptc_l(struct kvm_vcpu *vcpu, u64 va, u64 ps) +{ + va = PAGEALIGN(va, ps); + thash_purge_entries(vcpu, va, ps); +} + +void vcpu_ptc_e(struct kvm_vcpu *vcpu, u64 va) +{ + thash_purge_all(vcpu); +} + +void vcpu_ptc_ga(struct kvm_vcpu *vcpu, u64 va, u64 ps) +{ + struct exit_ctl_data *p = &vcpu->arch.exit_data; + long psr; + local_irq_save(psr); + p->exit_reason = EXIT_REASON_PTC_G; + + p->u.ptc_g_data.rr = vcpu_get_rr(vcpu, va); + p->u.ptc_g_data.vaddr = va; + p->u.ptc_g_data.ps = ps; + vmm_transition(vcpu); + /* Do Local Purge Here*/ + vcpu_ptc_l(vcpu, va, ps); + local_irq_restore(psr); +} + + +void vcpu_ptc_g(struct kvm_vcpu *vcpu, u64 va, u64 ps) +{ + vcpu_ptc_ga(vcpu, va, ps); +} + +void kvm_ptc_e(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long ifa; + + ifa = vcpu_get_gr(vcpu, inst.M45.r3); + vcpu_ptc_e(vcpu, ifa); +} + +void kvm_ptc_g(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long ifa, itir; + + ifa = vcpu_get_gr(vcpu, inst.M45.r3); + itir = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_ptc_g(vcpu, ifa, itir_ps(itir)); +} + +void kvm_ptc_ga(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long ifa, itir; + + ifa = vcpu_get_gr(vcpu, inst.M45.r3); + itir = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_ptc_ga(vcpu, ifa, itir_ps(itir)); +} + +void kvm_ptc_l(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long ifa, itir; + + ifa = vcpu_get_gr(vcpu, inst.M45.r3); + itir = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_ptc_l(vcpu, ifa, itir_ps(itir)); +} + +void kvm_ptr_d(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long ifa, itir; + + ifa = vcpu_get_gr(vcpu, inst.M45.r3); + itir = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_ptr_d(vcpu, ifa, itir_ps(itir)); +} + +void kvm_ptr_i(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long ifa, itir; + + ifa = vcpu_get_gr(vcpu, inst.M45.r3); + itir = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_ptr_i(vcpu, ifa, itir_ps(itir)); +} + +void kvm_itr_d(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long itir, ifa, pte, slot; + + slot = vcpu_get_gr(vcpu, inst.M45.r3); + pte = vcpu_get_gr(vcpu, inst.M45.r2); + itir = vcpu_get_itir(vcpu); + ifa = vcpu_get_ifa(vcpu); + vcpu_itr_d(vcpu, slot, pte, itir, ifa); +} + + + +void kvm_itr_i(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long itir, ifa, pte, slot; + + slot = vcpu_get_gr(vcpu, inst.M45.r3); + pte = vcpu_get_gr(vcpu, inst.M45.r2); + itir = vcpu_get_itir(vcpu); + ifa = vcpu_get_ifa(vcpu); + vcpu_itr_i(vcpu, slot, pte, itir, ifa); +} + +void kvm_itc_d(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long itir, ifa, pte; + + itir = vcpu_get_itir(vcpu); + ifa = vcpu_get_ifa(vcpu); + pte = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_itc_d(vcpu, pte, itir, ifa); +} + +void kvm_itc_i(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long itir, ifa, pte; + + itir = vcpu_get_itir(vcpu); + ifa = vcpu_get_ifa(vcpu); + pte = vcpu_get_gr(vcpu, inst.M45.r2); + vcpu_itc_i(vcpu, pte, itir, ifa); +} + +/************************************* + * Moves to semi-privileged registers + *************************************/ + +void kvm_mov_to_ar_imm(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long imm; + + if (inst.M30.s) + imm = -inst.M30.imm; + else + imm = inst.M30.imm; + + vcpu_set_itc(vcpu, imm); +} + +void kvm_mov_to_ar_reg(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r2; + + r2 = vcpu_get_gr(vcpu, inst.M29.r2); + vcpu_set_itc(vcpu, r2); +} + + +void kvm_mov_from_ar_reg(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r1; + + r1 = vcpu_get_itc(vcpu); + vcpu_set_gr(vcpu, inst.M31.r1, r1, 0); +} +/************************************************************************** + struct kvm_vcpu*protection key register access routines + **************************************************************************/ + +unsigned long vcpu_get_pkr(struct kvm_vcpu *vcpu, unsigned long reg) +{ + return ((unsigned long)ia64_get_pkr(reg)); +} + +void vcpu_set_pkr(struct kvm_vcpu *vcpu, unsigned long reg, unsigned long val) +{ + ia64_set_pkr(reg, val); +} + + +unsigned long vcpu_get_itir_on_fault(struct kvm_vcpu *vcpu, unsigned long ifa) +{ + union ia64_rr rr, rr1; + + rr.val = vcpu_get_rr(vcpu, ifa); + rr1.val = 0; + rr1.ps = rr.ps; + rr1.rid = rr.rid; + return (rr1.val); +} + + + +/******************************** + * Moves to privileged registers + ********************************/ +unsigned long vcpu_set_rr(struct kvm_vcpu *vcpu, unsigned long reg, + unsigned long val) +{ + union ia64_rr oldrr, newrr; + unsigned long rrval; + struct exit_ctl_data *p = &vcpu->arch.exit_data; + unsigned long psr; + + oldrr.val = vcpu_get_rr(vcpu, reg); + newrr.val = val; + vcpu->arch.vrr[reg >> VRN_SHIFT] = val; + + switch ((unsigned long)(reg >> VRN_SHIFT)) { + case VRN6: + vcpu->arch.vmm_rr = vrrtomrr(val); + local_irq_save(psr); + p->exit_reason = EXIT_REASON_SWITCH_RR6; + vmm_transition(vcpu); + local_irq_restore(psr); + break; + case VRN4: + rrval = vrrtomrr(val); + vcpu->arch.metaphysical_saved_rr4 = rrval; + if (!is_physical_mode(vcpu)) + ia64_set_rr(reg, rrval); + break; + case VRN0: + rrval = vrrtomrr(val); + vcpu->arch.metaphysical_saved_rr0 = rrval; + if (!is_physical_mode(vcpu)) + ia64_set_rr(reg, rrval); + break; + default: + ia64_set_rr(reg, vrrtomrr(val)); + break; + } + + return (IA64_NO_FAULT); +} + + + +void kvm_mov_to_rr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r2; + + r3 = vcpu_get_gr(vcpu, inst.M42.r3); + r2 = vcpu_get_gr(vcpu, inst.M42.r2); + vcpu_set_rr(vcpu, r3, r2); +} + +void kvm_mov_to_dbr(struct kvm_vcpu *vcpu, INST64 inst) +{ +} + +void kvm_mov_to_ibr(struct kvm_vcpu *vcpu, INST64 inst) +{ +} + +void kvm_mov_to_pmc(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r2; + + r3 = vcpu_get_gr(vcpu, inst.M42.r3); + r2 = vcpu_get_gr(vcpu, inst.M42.r2); + vcpu_set_pmc(vcpu, r3, r2); +} + +void kvm_mov_to_pmd(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r2; + + r3 = vcpu_get_gr(vcpu, inst.M42.r3); + r2 = vcpu_get_gr(vcpu, inst.M42.r2); + vcpu_set_pmd(vcpu, r3, r2); +} + +void kvm_mov_to_pkr(struct kvm_vcpu *vcpu, INST64 inst) +{ + u64 r3, r2; + + r3 = vcpu_get_gr(vcpu, inst.M42.r3); + r2 = vcpu_get_gr(vcpu, inst.M42.r2); + vcpu_set_pkr(vcpu, r3, r2); +} + + + +void kvm_mov_from_rr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r1; + + r3 = vcpu_get_gr(vcpu, inst.M43.r3); + r1 = vcpu_get_rr(vcpu, r3); + vcpu_set_gr(vcpu, inst.M43.r1, r1, 0); +} + +void kvm_mov_from_pkr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r1; + + r3 = vcpu_get_gr(vcpu, inst.M43.r3); + r1 = vcpu_get_pkr(vcpu, r3); + vcpu_set_gr(vcpu, inst.M43.r1, r1, 0); +} + +void kvm_mov_from_dbr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r1; + + r3 = vcpu_get_gr(vcpu, inst.M43.r3); + r1 = vcpu_get_dbr(vcpu, r3); + vcpu_set_gr(vcpu, inst.M43.r1, r1, 0); +} + +void kvm_mov_from_ibr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r1; + + r3 = vcpu_get_gr(vcpu, inst.M43.r3); + r1 = vcpu_get_ibr(vcpu, r3); + vcpu_set_gr(vcpu, inst.M43.r1, r1, 0); +} + +void kvm_mov_from_pmc(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r1; + + r3 = vcpu_get_gr(vcpu, inst.M43.r3); + r1 = vcpu_get_pmc(vcpu, r3); + vcpu_set_gr(vcpu, inst.M43.r1, r1, 0); +} + + +unsigned long vcpu_get_cpuid(struct kvm_vcpu *vcpu, unsigned long reg) +{ + /* FIXME: This could get called as a result of a rsvd-reg fault */ + if (reg > (ia64_get_cpuid(3) & 0xff)) + return 0; + else + return ia64_get_cpuid(reg); +} + +void kvm_mov_from_cpuid(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r3, r1; + + r3 = vcpu_get_gr(vcpu, inst.M43.r3); + r1 = vcpu_get_cpuid(vcpu, r3); + vcpu_set_gr(vcpu, inst.M43.r1, r1, 0); +} + +void vcpu_set_tpr(struct kvm_vcpu *vcpu, unsigned long val) +{ + VCPU(vcpu, tpr) = val; + vcpu->arch.irq_check = 1; +} + +unsigned long kvm_mov_to_cr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long r2; + + r2 = vcpu_get_gr(vcpu, inst.M32.r2); + VCPU(vcpu, vcr[inst.M32.cr3]) = r2; + + switch (inst.M32.cr3) { + case 0: + vcpu_set_dcr(vcpu, r2); + break; + case 1: + vcpu_set_itm(vcpu, r2); + break; + case 66: + vcpu_set_tpr(vcpu, r2); + break; + case 67: + vcpu_set_eoi(vcpu, r2); + break; + default: + break; + } + + return 0; +} + + +unsigned long kvm_mov_from_cr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long tgt = inst.M33.r1; + unsigned long val; + + switch (inst.M33.cr3) { + case 65: + val = vcpu_get_ivr(vcpu); + vcpu_set_gr(vcpu, tgt, val, 0); + break; + + case 67: + vcpu_set_gr(vcpu, tgt, 0L, 0); + break; + default: + val = VCPU(vcpu, vcr[inst.M33.cr3]); + vcpu_set_gr(vcpu, tgt, val, 0); + break; + } + + return 0; +} + + + +void vcpu_set_psr(struct kvm_vcpu *vcpu, unsigned long val) +{ + + unsigned long mask; + struct kvm_pt_regs *regs; + struct ia64_psr old_psr, new_psr; + + old_psr = *(struct ia64_psr *)&VCPU(vcpu, vpsr); + + regs = vcpu_regs(vcpu); + /* We only support guest as: + * vpsr.pk = 0 + * vpsr.is = 0 + * Otherwise panic + */ + if (val & (IA64_PSR_PK | IA64_PSR_IS | IA64_PSR_VM)) + panic_vm(vcpu); + + /* + * For those IA64_PSR bits: id/da/dd/ss/ed/ia + * Since these bits will become 0, after success execution of each + * instruction, we will change set them to mIA64_PSR + */ + VCPU(vcpu, vpsr) = val + & (~(IA64_PSR_ID | IA64_PSR_DA | IA64_PSR_DD | + IA64_PSR_SS | IA64_PSR_ED | IA64_PSR_IA)); + + if (!old_psr.i && (val & IA64_PSR_I)) { + /* vpsr.i 0->1 */ + vcpu->arch.irq_check = 1; + } + new_psr = *(struct ia64_psr *)&VCPU(vcpu, vpsr); + + /* + * All vIA64_PSR bits shall go to mPSR (v->tf->tf_special.psr) + * , except for the following bits: + * ic/i/dt/si/rt/mc/it/bn/vm + */ + mask = IA64_PSR_IC + IA64_PSR_I + IA64_PSR_DT + IA64_PSR_SI + + IA64_PSR_RT + IA64_PSR_MC + IA64_PSR_IT + IA64_PSR_BN + + IA64_PSR_VM; + + regs->cr_ipsr = (regs->cr_ipsr & mask) | (val & (~mask)); + + check_mm_mode_switch(vcpu, old_psr, new_psr); + + return ; +} + +unsigned long vcpu_cover(struct kvm_vcpu *vcpu) +{ + struct ia64_psr vpsr; + + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + vpsr = *(struct ia64_psr *)&VCPU(vcpu, vpsr); + + if (!vpsr.ic) + VCPU(vcpu, ifs) = regs->cr_ifs; + regs->cr_ifs = IA64_IFS_V; + return (IA64_NO_FAULT); +} + + + +/************************************************************************** + VCPU banked general register access routines + **************************************************************************/ +#define vcpu_bsw0_unat(i, b0unat, b1unat, runat, VMM_PT_REGS_R16_SLOT) \ + do { \ + __asm__ __volatile__ ( \ + ";;extr.u %0 = %3,%6,16;;\n" \ + "dep %1 = %0, %1, 0, 16;;\n" \ + "st8 [%4] = %1\n" \ + "extr.u %0 = %2, 16, 16;;\n" \ + "dep %3 = %0, %3, %6, 16;;\n" \ + "st8 [%5] = %3\n" \ + ::"r"(i), "r"(*b1unat), "r"(*b0unat), \ + "r"(*runat), "r"(b1unat), "r"(runat), \ + "i"(VMM_PT_REGS_R16_SLOT) : "memory"); \ + } while (0) + +void vcpu_bsw0(struct kvm_vcpu *vcpu) +{ + unsigned long i; + + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + unsigned long *r = ®s->r16; + unsigned long *b0 = &VCPU(vcpu, vbgr[0]); + unsigned long *b1 = &VCPU(vcpu, vgr[0]); + unsigned long *runat = ®s->eml_unat; + unsigned long *b0unat = &VCPU(vcpu, vbnat); + unsigned long *b1unat = &VCPU(vcpu, vnat); + + + if (VCPU(vcpu, vpsr) & IA64_PSR_BN) { + for (i = 0; i < 16; i++) { + *b1++ = *r; + *r++ = *b0++; + } + vcpu_bsw0_unat(i, b0unat, b1unat, runat, + VMM_PT_REGS_R16_SLOT); + VCPU(vcpu, vpsr) &= ~IA64_PSR_BN; + } +} + +#define vcpu_bsw1_unat(i, b0unat, b1unat, runat, VMM_PT_REGS_R16_SLOT) \ + do { \ + __asm__ __volatile__ (";;extr.u %0 = %3, %6, 16;;\n" \ + "dep %1 = %0, %1, 16, 16;;\n" \ + "st8 [%4] = %1\n" \ + "extr.u %0 = %2, 0, 16;;\n" \ + "dep %3 = %0, %3, %6, 16;;\n" \ + "st8 [%5] = %3\n" \ + ::"r"(i), "r"(*b0unat), "r"(*b1unat), \ + "r"(*runat), "r"(b0unat), "r"(runat), \ + "i"(VMM_PT_REGS_R16_SLOT) : "memory"); \ + } while (0) + +void vcpu_bsw1(struct kvm_vcpu *vcpu) +{ + unsigned long i; + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + unsigned long *r = ®s->r16; + unsigned long *b0 = &VCPU(vcpu, vbgr[0]); + unsigned long *b1 = &VCPU(vcpu, vgr[0]); + unsigned long *runat = ®s->eml_unat; + unsigned long *b0unat = &VCPU(vcpu, vbnat); + unsigned long *b1unat = &VCPU(vcpu, vnat); + + if (!(VCPU(vcpu, vpsr) & IA64_PSR_BN)) { + for (i = 0; i < 16; i++) { + *b0++ = *r; + *r++ = *b1++; + } + vcpu_bsw1_unat(i, b0unat, b1unat, runat, + VMM_PT_REGS_R16_SLOT); + VCPU(vcpu, vpsr) |= IA64_PSR_BN; + } +} + + + + +void vcpu_rfi(struct kvm_vcpu *vcpu) +{ + unsigned long ifs, psr; + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + psr = VCPU(vcpu, ipsr); + if (psr & IA64_PSR_BN) + vcpu_bsw1(vcpu); + else + vcpu_bsw0(vcpu); + vcpu_set_psr(vcpu, psr); + ifs = VCPU(vcpu, ifs); + if (ifs >> 63) + regs->cr_ifs = ifs; + regs->cr_iip = VCPU(vcpu, iip); +} + + +/* + VPSR can't keep track of below bits of guest PSR + This function gets guest PSR + */ + +unsigned long vcpu_get_psr(struct kvm_vcpu *vcpu) +{ + unsigned long mask; + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + + mask = IA64_PSR_BE | IA64_PSR_UP | IA64_PSR_AC | IA64_PSR_MFL | + IA64_PSR_MFH | IA64_PSR_CPL | IA64_PSR_RI; + return (VCPU(vcpu, vpsr) & ~mask) | (regs->cr_ipsr & mask); +} + +void kvm_rsm(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long vpsr; + unsigned long imm24 = (inst.M44.i<<23) | (inst.M44.i2<<21) + | inst.M44.imm; + + vpsr = vcpu_get_psr(vcpu); + vpsr &= (~imm24); + vcpu_set_psr(vcpu, vpsr); +} + +void kvm_ssm(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long vpsr; + unsigned long imm24 = (inst.M44.i << 23) | (inst.M44.i2 << 21) + | inst.M44.imm; + + vpsr = vcpu_get_psr(vcpu); + vpsr |= imm24; + vcpu_set_psr(vcpu, vpsr); +} + +/* Generate Mask + * Parameter: + * bit -- starting bit + * len -- how many bits + */ +#define MASK(bit,len) \ +({ \ + __u64 ret; \ + \ + __asm __volatile("dep %0=-1, r0, %1, %2"\ + : "=r" (ret): \ + "M" (bit), \ + "M" (len)); \ + ret; \ +}) + +void vcpu_set_psr_l(struct kvm_vcpu *vcpu, unsigned long val) +{ + val = (val & MASK(0, 32)) | (vcpu_get_psr(vcpu) & MASK(32, 32)); + vcpu_set_psr(vcpu, val); +} + +void kvm_mov_to_psr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long val; + + val = vcpu_get_gr(vcpu, inst.M35.r2); + vcpu_set_psr_l(vcpu, val); +} + +void kvm_mov_from_psr(struct kvm_vcpu *vcpu, INST64 inst) +{ + unsigned long val; + + val = vcpu_get_psr(vcpu); + val = (val & MASK(0, 32)) | (val & MASK(35, 2)); + vcpu_set_gr(vcpu, inst.M33.r1, val, 0); +} + +void vcpu_increment_iip(struct kvm_vcpu *vcpu) +{ + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + struct ia64_psr *ipsr = (struct ia64_psr *)®s->cr_ipsr; + if (ipsr->ri == 2) { + ipsr->ri = 0; + regs->cr_iip += 16; + } else + ipsr->ri++; +} + +void vcpu_decrement_iip(struct kvm_vcpu *vcpu) +{ + struct kvm_pt_regs *regs = vcpu_regs(vcpu); + struct ia64_psr *ipsr = (struct ia64_psr *)®s->cr_ipsr; + + if (ipsr->ri == 0) { + ipsr->ri = 2; + regs->cr_iip -= 16; + } else + ipsr->ri--; +} + +/** Emulate a privileged operation. + * + * + * @param vcpu virtual cpu + * @cause the reason cause virtualization fault + * @opcode the instruction code which cause virtualization fault + */ + +void kvm_emulate(struct kvm_vcpu *vcpu, struct kvm_pt_regs *regs) +{ + unsigned long status, cause, opcode ; + INST64 inst; + + status = IA64_NO_FAULT; + cause = VMX(vcpu, cause); + opcode = VMX(vcpu, opcode); + inst.inst = opcode; + /* + * Switch to actual virtual rid in rr0 and rr4, + * which is required by some tlb related instructions. + */ + prepare_if_physical_mode(vcpu); + + switch (cause) { + case EVENT_RSM: + kvm_rsm(vcpu, inst); + break; + case EVENT_SSM: + kvm_ssm(vcpu, inst); + break; + case EVENT_MOV_TO_PSR: + kvm_mov_to_psr(vcpu, inst); + break; + case EVENT_MOV_FROM_PSR: + kvm_mov_from_psr(vcpu, inst); + break; + case EVENT_MOV_FROM_CR: + kvm_mov_from_cr(vcpu, inst); + break; + case EVENT_MOV_TO_CR: + kvm_mov_to_cr(vcpu, inst); + break; + case EVENT_BSW_0: + vcpu_bsw0(vcpu); + break; + case EVENT_BSW_1: + vcpu_bsw1(vcpu); + break; + case EVENT_COVER: + vcpu_cover(vcpu); + break; + case EVENT_RFI: + vcpu_rfi(vcpu); + break; + case EVENT_ITR_D: + kvm_itr_d(vcpu, inst); + break; + case EVENT_ITR_I: + kvm_itr_i(vcpu, inst); + break; + case EVENT_PTR_D: + kvm_ptr_d(vcpu, inst); + break; + case EVENT_PTR_I: + kvm_ptr_i(vcpu, inst); + break; + case EVENT_ITC_D: + kvm_itc_d(vcpu, inst); + break; + case EVENT_ITC_I: + kvm_itc_i(vcpu, inst); + break; + case EVENT_PTC_L: + kvm_ptc_l(vcpu, inst); + break; + case EVENT_PTC_G: + kvm_ptc_g(vcpu, inst); + break; + case EVENT_PTC_GA: + kvm_ptc_ga(vcpu, inst); + break; + case EVENT_PTC_E: + kvm_ptc_e(vcpu, inst); + break; + case EVENT_MOV_TO_RR: + kvm_mov_to_rr(vcpu, inst); + break; + case EVENT_MOV_FROM_RR: + kvm_mov_from_rr(vcpu, inst); + break; + case EVENT_THASH: + kvm_thash(vcpu, inst); + break; + case EVENT_TTAG: + kvm_ttag(vcpu, inst); + break; + case EVENT_TPA: + status = kvm_tpa(vcpu, inst); + break; + case EVENT_TAK: + kvm_tak(vcpu, inst); + break; + case EVENT_MOV_TO_AR_IMM: + kvm_mov_to_ar_imm(vcpu, inst); + break; + case EVENT_MOV_TO_AR: + kvm_mov_to_ar_reg(vcpu, inst); + break; + case EVENT_MOV_FROM_AR: + kvm_mov_from_ar_reg(vcpu, inst); + break; + case EVENT_MOV_TO_DBR: + kvm_mov_to_dbr(vcpu, inst); + break; + case EVENT_MOV_TO_IBR: + kvm_mov_to_ibr(vcpu, inst); + break; + case EVENT_MOV_TO_PMC: + kvm_mov_to_pmc(vcpu, inst); + break; + case EVENT_MOV_TO_PMD: + kvm_mov_to_pmd(vcpu, inst); + break; + case EVENT_MOV_TO_PKR: + kvm_mov_to_pkr(vcpu, inst); + break; + case EVENT_MOV_FROM_DBR: + kvm_mov_from_dbr(vcpu, inst); + break; + case EVENT_MOV_FROM_IBR: + kvm_mov_from_ibr(vcpu, inst); + break; + case EVENT_MOV_FROM_PMC: + kvm_mov_from_pmc(vcpu, inst); + break; + case EVENT_MOV_FROM_PKR: + kvm_mov_from_pkr(vcpu, inst); + break; + case EVENT_MOV_FROM_CPUID: + kvm_mov_from_cpuid(vcpu, inst); + break; + case EVENT_VMSW: + status = IA64_FAULT; + break; + default: + break; + }; + /*Assume all status is NO_FAULT ?*/ + if (status == IA64_NO_FAULT && cause != EVENT_RFI) + vcpu_increment_iip(vcpu); + + recover_if_physical_mode(vcpu); +} + +void init_vcpu(struct kvm_vcpu *vcpu) +{ + int i; + + vcpu->arch.mode_flags = GUEST_IN_PHY; + VMX(vcpu, vrr[0]) = 0x38; + VMX(vcpu, vrr[1]) = 0x38; + VMX(vcpu, vrr[2]) = 0x38; + VMX(vcpu, vrr[3]) = 0x38; + VMX(vcpu, vrr[4]) = 0x38; + VMX(vcpu, vrr[5]) = 0x38; + VMX(vcpu, vrr[6]) = 0x38; + VMX(vcpu, vrr[7]) = 0x38; + VCPU(vcpu, vpsr) = IA64_PSR_BN; + VCPU(vcpu, dcr) = 0; + /* pta.size must not be 0. The minimum is 15 (32k) */ + VCPU(vcpu, pta) = 15 << 2; + VCPU(vcpu, itv) = 0x10000; + VCPU(vcpu, itm) = 0; + VMX(vcpu, last_itc) = 0; + + VCPU(vcpu, lid) = VCPU_LID(vcpu); + VCPU(vcpu, ivr) = 0; + VCPU(vcpu, tpr) = 0x10000; + VCPU(vcpu, eoi) = 0; + VCPU(vcpu, irr[0]) = 0; + VCPU(vcpu, irr[1]) = 0; + VCPU(vcpu, irr[2]) = 0; + VCPU(vcpu, irr[3]) = 0; + VCPU(vcpu, pmv) = 0x10000; + VCPU(vcpu, cmcv) = 0x10000; + VCPU(vcpu, lrr0) = 0x10000; /* default reset value? */ + VCPU(vcpu, lrr1) = 0x10000; /* default reset value? */ + update_vhpi(vcpu, NULL_VECTOR); + VLSAPIC_XTP(vcpu) = 0x80; /* disabled */ + + for (i = 0; i < 4; i++) + VLSAPIC_INSVC(vcpu, i) = 0; +} + +void kvm_init_all_rr(struct kvm_vcpu *vcpu) +{ + unsigned long psr; + + local_irq_save(psr); + + /* WARNING: not allow co-exist of both virtual mode and physical + * mode in same region + */ + + vcpu->arch.metaphysical_saved_rr0 = vrrtomrr(VMX(vcpu, vrr[VRN0])); + vcpu->arch.metaphysical_saved_rr4 = vrrtomrr(VMX(vcpu, vrr[VRN4])); + + if (is_physical_mode(vcpu)) { + if (vcpu->arch.mode_flags & GUEST_PHY_EMUL) + panic_vm(vcpu); + + ia64_set_rr((VRN0 << VRN_SHIFT), vcpu->arch.metaphysical_rr0); + ia64_dv_serialize_data(); + ia64_set_rr((VRN4 << VRN_SHIFT), vcpu->arch.metaphysical_rr4); + ia64_dv_serialize_data(); + } else { + ia64_set_rr((VRN0 << VRN_SHIFT), + vcpu->arch.metaphysical_saved_rr0); + ia64_dv_serialize_data(); + ia64_set_rr((VRN4 << VRN_SHIFT), + vcpu->arch.metaphysical_saved_rr4); + ia64_dv_serialize_data(); + } + ia64_set_rr((VRN1 << VRN_SHIFT), + vrrtomrr(VMX(vcpu, vrr[VRN1]))); + ia64_dv_serialize_data(); + ia64_set_rr((VRN2 << VRN_SHIFT), + vrrtomrr(VMX(vcpu, vrr[VRN2]))); + ia64_dv_serialize_data(); + ia64_set_rr((VRN3 << VRN_SHIFT), + vrrtomrr(VMX(vcpu, vrr[VRN3]))); + ia64_dv_serialize_data(); + ia64_set_rr((VRN5 << VRN_SHIFT), + vrrtomrr(VMX(vcpu, vrr[VRN5]))); + ia64_dv_serialize_data(); + ia64_set_rr((VRN7 << VRN_SHIFT), + vrrtomrr(VMX(vcpu, vrr[VRN7]))); + ia64_dv_serialize_data(); + ia64_srlz_d(); + ia64_set_psr(psr); +} + +int vmm_entry(void) +{ + struct kvm_vcpu *v; + v = current_vcpu; + + ia64_call_vsa(PAL_VPS_RESTORE, (unsigned long)v->arch.vpd, + 0, 0, 0, 0, 0, 0); + kvm_init_vtlb(v); + kvm_init_vhpt(v); + init_vcpu(v); + kvm_init_all_rr(v); + vmm_reset_entry(); + + return 0; +} + +void panic_vm(struct kvm_vcpu *v) +{ + struct exit_ctl_data *p = &v->arch.exit_data; + + p->exit_reason = EXIT_REASON_VM_PANIC; + vmm_transition(v); + /*Never to return*/ + while (1); +} diff --git a/arch/ia64/kvm/vcpu.h b/arch/ia64/kvm/vcpu.h new file mode 100644 index 0000000000000000000000000000000000000000..b0fcfb62c49e8a660464be9fca38c2e555ab20cc --- /dev/null +++ b/arch/ia64/kvm/vcpu.h @@ -0,0 +1,740 @@ +/* + * vcpu.h: vcpu routines + * Copyright (c) 2005, Intel Corporation. + * Xuefei Xu (Anthony Xu) (Anthony.xu@intel.com) + * Yaozu Dong (Eddie Dong) (Eddie.dong@intel.com) + * + * Copyright (c) 2007, Intel Corporation. + * Xuefei Xu (Anthony Xu) (Anthony.xu@intel.com) + * Xiantao Zhang (xiantao.zhang@intel.com) + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + + +#ifndef __KVM_VCPU_H__ +#define __KVM_VCPU_H__ + +#include +#include +#include + +#ifndef __ASSEMBLY__ +#include "vti.h" + +#include +#include + +typedef unsigned long IA64_INST; + +typedef union U_IA64_BUNDLE { + unsigned long i64[2]; + struct { unsigned long template:5, slot0:41, slot1a:18, + slot1b:23, slot2:41; }; + /* NOTE: following doesn't work because bitfields can't cross natural + size boundaries + struct { unsigned long template:5, slot0:41, slot1:41, slot2:41; }; */ +} IA64_BUNDLE; + +typedef union U_INST64_A5 { + IA64_INST inst; + struct { unsigned long qp:6, r1:7, imm7b:7, r3:2, imm5c:5, + imm9d:9, s:1, major:4; }; +} INST64_A5; + +typedef union U_INST64_B4 { + IA64_INST inst; + struct { unsigned long qp:6, btype:3, un3:3, p:1, b2:3, un11:11, x6:6, + wh:2, d:1, un1:1, major:4; }; +} INST64_B4; + +typedef union U_INST64_B8 { + IA64_INST inst; + struct { unsigned long qp:6, un21:21, x6:6, un4:4, major:4; }; +} INST64_B8; + +typedef union U_INST64_B9 { + IA64_INST inst; + struct { unsigned long qp:6, imm20:20, :1, x6:6, :3, i:1, major:4; }; +} INST64_B9; + +typedef union U_INST64_I19 { + IA64_INST inst; + struct { unsigned long qp:6, imm20:20, :1, x6:6, x3:3, i:1, major:4; }; +} INST64_I19; + +typedef union U_INST64_I26 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, ar3:7, x6:6, x3:3, :1, major:4; }; +} INST64_I26; + +typedef union U_INST64_I27 { + IA64_INST inst; + struct { unsigned long qp:6, :7, imm:7, ar3:7, x6:6, x3:3, s:1, major:4; }; +} INST64_I27; + +typedef union U_INST64_I28 { /* not privileged (mov from AR) */ + IA64_INST inst; + struct { unsigned long qp:6, r1:7, :7, ar3:7, x6:6, x3:3, :1, major:4; }; +} INST64_I28; + +typedef union U_INST64_M28 { + IA64_INST inst; + struct { unsigned long qp:6, :14, r3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M28; + +typedef union U_INST64_M29 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, ar3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M29; + +typedef union U_INST64_M30 { + IA64_INST inst; + struct { unsigned long qp:6, :7, imm:7, ar3:7, x4:4, x2:2, + x3:3, s:1, major:4; }; +} INST64_M30; + +typedef union U_INST64_M31 { + IA64_INST inst; + struct { unsigned long qp:6, r1:7, :7, ar3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M31; + +typedef union U_INST64_M32 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, cr3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M32; + +typedef union U_INST64_M33 { + IA64_INST inst; + struct { unsigned long qp:6, r1:7, :7, cr3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M33; + +typedef union U_INST64_M35 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, :7, x6:6, x3:3, :1, major:4; }; + +} INST64_M35; + +typedef union U_INST64_M36 { + IA64_INST inst; + struct { unsigned long qp:6, r1:7, :14, x6:6, x3:3, :1, major:4; }; +} INST64_M36; + +typedef union U_INST64_M37 { + IA64_INST inst; + struct { unsigned long qp:6, imm20a:20, :1, x4:4, x2:2, x3:3, + i:1, major:4; }; +} INST64_M37; + +typedef union U_INST64_M41 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, :7, x6:6, x3:3, :1, major:4; }; +} INST64_M41; + +typedef union U_INST64_M42 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, r3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M42; + +typedef union U_INST64_M43 { + IA64_INST inst; + struct { unsigned long qp:6, r1:7, :7, r3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M43; + +typedef union U_INST64_M44 { + IA64_INST inst; + struct { unsigned long qp:6, imm:21, x4:4, i2:2, x3:3, i:1, major:4; }; +} INST64_M44; + +typedef union U_INST64_M45 { + IA64_INST inst; + struct { unsigned long qp:6, :7, r2:7, r3:7, x6:6, x3:3, :1, major:4; }; +} INST64_M45; + +typedef union U_INST64_M46 { + IA64_INST inst; + struct { unsigned long qp:6, r1:7, un7:7, r3:7, x6:6, + x3:3, un1:1, major:4; }; +} INST64_M46; + +typedef union U_INST64_M47 { + IA64_INST inst; + struct { unsigned long qp:6, un14:14, r3:7, x6:6, x3:3, un1:1, major:4; }; +} INST64_M47; + +typedef union U_INST64_M1{ + IA64_INST inst; + struct { unsigned long qp:6, r1:7, un7:7, r3:7, x:1, hint:2, + x6:6, m:1, major:4; }; +} INST64_M1; + +typedef union U_INST64_M2{ + IA64_INST inst; + struct { unsigned long qp:6, r1:7, r2:7, r3:7, x:1, hint:2, + x6:6, m:1, major:4; }; +} INST64_M2; + +typedef union U_INST64_M3{ + IA64_INST inst; + struct { unsigned long qp:6, r1:7, imm7:7, r3:7, i:1, hint:2, + x6:6, s:1, major:4; }; +} INST64_M3; + +typedef union U_INST64_M4 { + IA64_INST inst; + struct { unsigned long qp:6, un7:7, r2:7, r3:7, x:1, hint:2, + x6:6, m:1, major:4; }; +} INST64_M4; + +typedef union U_INST64_M5 { + IA64_INST inst; + struct { unsigned long qp:6, imm7:7, r2:7, r3:7, i:1, hint:2, + x6:6, s:1, major:4; }; +} INST64_M5; + +typedef union U_INST64_M6 { + IA64_INST inst; + struct { unsigned long qp:6, f1:7, un7:7, r3:7, x:1, hint:2, + x6:6, m:1, major:4; }; +} INST64_M6; + +typedef union U_INST64_M9 { + IA64_INST inst; + struct { unsigned long qp:6, :7, f2:7, r3:7, x:1, hint:2, + x6:6, m:1, major:4; }; +} INST64_M9; + +typedef union U_INST64_M10 { + IA64_INST inst; + struct { unsigned long qp:6, imm7:7, f2:7, r3:7, i:1, hint:2, + x6:6, s:1, major:4; }; +} INST64_M10; + +typedef union U_INST64_M12 { + IA64_INST inst; + struct { unsigned long qp:6, f1:7, f2:7, r3:7, x:1, hint:2, + x6:6, m:1, major:4; }; +} INST64_M12; + +typedef union U_INST64_M15 { + IA64_INST inst; + struct { unsigned long qp:6, :7, imm7:7, r3:7, i:1, hint:2, + x6:6, s:1, major:4; }; +} INST64_M15; + +typedef union U_INST64 { + IA64_INST inst; + struct { unsigned long :37, major:4; } generic; + INST64_A5 A5; /* used in build_hypercall_bundle only */ + INST64_B4 B4; /* used in build_hypercall_bundle only */ + INST64_B8 B8; /* rfi, bsw.[01] */ + INST64_B9 B9; /* break.b */ + INST64_I19 I19; /* used in build_hypercall_bundle only */ + INST64_I26 I26; /* mov register to ar (I unit) */ + INST64_I27 I27; /* mov immediate to ar (I unit) */ + INST64_I28 I28; /* mov from ar (I unit) */ + INST64_M1 M1; /* ld integer */ + INST64_M2 M2; + INST64_M3 M3; + INST64_M4 M4; /* st integer */ + INST64_M5 M5; + INST64_M6 M6; /* ldfd floating pointer */ + INST64_M9 M9; /* stfd floating pointer */ + INST64_M10 M10; /* stfd floating pointer */ + INST64_M12 M12; /* ldfd pair floating pointer */ + INST64_M15 M15; /* lfetch + imm update */ + INST64_M28 M28; /* purge translation cache entry */ + INST64_M29 M29; /* mov register to ar (M unit) */ + INST64_M30 M30; /* mov immediate to ar (M unit) */ + INST64_M31 M31; /* mov from ar (M unit) */ + INST64_M32 M32; /* mov reg to cr */ + INST64_M33 M33; /* mov from cr */ + INST64_M35 M35; /* mov to psr */ + INST64_M36 M36; /* mov from psr */ + INST64_M37 M37; /* break.m */ + INST64_M41 M41; /* translation cache insert */ + INST64_M42 M42; /* mov to indirect reg/translation reg insert*/ + INST64_M43 M43; /* mov from indirect reg */ + INST64_M44 M44; /* set/reset system mask */ + INST64_M45 M45; /* translation purge */ + INST64_M46 M46; /* translation access (tpa,tak) */ + INST64_M47 M47; /* purge translation entry */ +} INST64; + +#define MASK_41 ((unsigned long)0x1ffffffffff) + +/* Virtual address memory attributes encoding */ +#define VA_MATTR_WB 0x0 +#define VA_MATTR_UC 0x4 +#define VA_MATTR_UCE 0x5 +#define VA_MATTR_WC 0x6 +#define VA_MATTR_NATPAGE 0x7 + +#define PMASK(size) (~((size) - 1)) +#define PSIZE(size) (1UL<<(size)) +#define CLEARLSB(ppn, nbits) (((ppn) >> (nbits)) << (nbits)) +#define PAGEALIGN(va, ps) CLEARLSB(va, ps) +#define PAGE_FLAGS_RV_MASK (0x2|(0x3UL<<50)|(((1UL<<11)-1)<<53)) +#define _PAGE_MA_ST (0x1 << 2) /* is reserved for software use */ + +#define ARCH_PAGE_SHIFT 12 + +#define INVALID_TI_TAG (1UL << 63) + +#define VTLB_PTE_P_BIT 0 +#define VTLB_PTE_IO_BIT 60 +#define VTLB_PTE_IO (1UL<> 61))) + +#define vcpu_quick_region_set(_tr_regions,_ifa) \ + do {_tr_regions |= (1 << ((unsigned long)_ifa >> 61)); } while (0) + +static inline void vcpu_set_tr(struct thash_data *trp, u64 pte, u64 itir, + u64 va, u64 rid) +{ + trp->page_flags = pte; + trp->itir = itir; + trp->vadr = va; + trp->rid = rid; +} + +extern u64 kvm_lookup_mpa(u64 gpfn); +extern u64 kvm_gpa_to_mpa(u64 gpa); + +/* Return I/O type if trye */ +#define __gpfn_is_io(gpfn) \ + ({ \ + u64 pte, ret = 0; \ + pte = kvm_lookup_mpa(gpfn); \ + if (!(pte & GPFN_INV_MASK)) \ + ret = pte & GPFN_IO_MASK; \ + ret; \ + }) + +#endif + +#define IA64_NO_FAULT 0 +#define IA64_FAULT 1 + +#define VMM_RBS_OFFSET ((VMM_TASK_SIZE + 15) & ~15) + +#define SW_BAD 0 /* Bad mode transitition */ +#define SW_V2P 1 /* Physical emulatino is activated */ +#define SW_P2V 2 /* Exit physical mode emulation */ +#define SW_SELF 3 /* No mode transition */ +#define SW_NOP 4 /* Mode transition, but without action required */ + +#define GUEST_IN_PHY 0x1 +#define GUEST_PHY_EMUL 0x2 + +#define current_vcpu ((struct kvm_vcpu *) ia64_getreg(_IA64_REG_TP)) + +#define VRN_SHIFT 61 +#define VRN_MASK 0xe000000000000000 +#define VRN0 0x0UL +#define VRN1 0x1UL +#define VRN2 0x2UL +#define VRN3 0x3UL +#define VRN4 0x4UL +#define VRN5 0x5UL +#define VRN6 0x6UL +#define VRN7 0x7UL + +#define IRQ_NO_MASKED 0 +#define IRQ_MASKED_BY_VTPR 1 +#define IRQ_MASKED_BY_INSVC 2 /* masked by inservice IRQ */ + +#define PTA_BASE_SHIFT 15 + +#define IA64_PSR_VM_BIT 46 +#define IA64_PSR_VM (__IA64_UL(1) << IA64_PSR_VM_BIT) + +/* Interruption Function State */ +#define IA64_IFS_V_BIT 63 +#define IA64_IFS_V (__IA64_UL(1) << IA64_IFS_V_BIT) + +#define PHY_PAGE_UC (_PAGE_A|_PAGE_D|_PAGE_P|_PAGE_MA_UC|_PAGE_AR_RWX) +#define PHY_PAGE_WB (_PAGE_A|_PAGE_D|_PAGE_P|_PAGE_MA_WB|_PAGE_AR_RWX) + +#ifndef __ASSEMBLY__ + +#include + +#define is_physical_mode(v) \ + ((v->arch.mode_flags) & GUEST_IN_PHY) + +#define is_virtual_mode(v) \ + (!is_physical_mode(v)) + +#define MODE_IND(psr) \ + (((psr).it << 2) + ((psr).dt << 1) + (psr).rt) + +#define _vmm_raw_spin_lock(x) \ + do { \ + __u32 *ia64_spinlock_ptr = (__u32 *) (x); \ + __u64 ia64_spinlock_val; \ + ia64_spinlock_val = ia64_cmpxchg4_acq(ia64_spinlock_ptr, 1, 0);\ + if (unlikely(ia64_spinlock_val)) { \ + do { \ + while (*ia64_spinlock_ptr) \ + ia64_barrier(); \ + ia64_spinlock_val = \ + ia64_cmpxchg4_acq(ia64_spinlock_ptr, 1, 0);\ + } while (ia64_spinlock_val); \ + } \ + } while (0) + +#define _vmm_raw_spin_unlock(x) \ + do { barrier(); \ + ((spinlock_t *)x)->raw_lock.lock = 0; } \ +while (0) + +void vmm_spin_lock(spinlock_t *lock); +void vmm_spin_unlock(spinlock_t *lock); +enum { + I_TLB = 1, + D_TLB = 2 +}; + +union kvm_va { + struct { + unsigned long off : 60; /* intra-region offset */ + unsigned long reg : 4; /* region number */ + } f; + unsigned long l; + void *p; +}; + +#define __kvm_pa(x) ({union kvm_va _v; _v.l = (long) (x); \ + _v.f.reg = 0; _v.l; }) +#define __kvm_va(x) ({union kvm_va _v; _v.l = (long) (x); \ + _v.f.reg = -1; _v.p; }) + +#define _REGION_ID(x) ({union ia64_rr _v; _v.val = (long)(x); \ + _v.rid; }) +#define _REGION_PAGE_SIZE(x) ({union ia64_rr _v; _v.val = (long)(x); \ + _v.ps; }) +#define _REGION_HW_WALKER(x) ({union ia64_rr _v; _v.val = (long)(x); \ + _v.ve; }) + +enum vhpt_ref{ DATA_REF, NA_REF, INST_REF, RSE_REF }; +enum tlb_miss_type { INSTRUCTION, DATA, REGISTER }; + +#define VCPU(_v, _x) ((_v)->arch.vpd->_x) +#define VMX(_v, _x) ((_v)->arch._x) + +#define VLSAPIC_INSVC(vcpu, i) ((vcpu)->arch.insvc[i]) +#define VLSAPIC_XTP(_v) VMX(_v, xtp) + +static inline unsigned long itir_ps(unsigned long itir) +{ + return ((itir >> 2) & 0x3f); +} + + +/************************************************************************** + VCPU control register access routines + **************************************************************************/ + +static inline u64 vcpu_get_itir(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, itir)); +} + +static inline void vcpu_set_itir(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, itir) = val; +} + +static inline u64 vcpu_get_ifa(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, ifa)); +} + +static inline void vcpu_set_ifa(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, ifa) = val; +} + +static inline u64 vcpu_get_iva(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, iva)); +} + +static inline u64 vcpu_get_pta(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, pta)); +} + +static inline u64 vcpu_get_lid(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, lid)); +} + +static inline u64 vcpu_get_tpr(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, tpr)); +} + +static inline u64 vcpu_get_eoi(struct kvm_vcpu *vcpu) +{ + return (0UL); /*reads of eoi always return 0 */ +} + +static inline u64 vcpu_get_irr0(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, irr[0])); +} + +static inline u64 vcpu_get_irr1(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, irr[1])); +} + +static inline u64 vcpu_get_irr2(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, irr[2])); +} + +static inline u64 vcpu_get_irr3(struct kvm_vcpu *vcpu) +{ + return ((u64)VCPU(vcpu, irr[3])); +} + +static inline void vcpu_set_dcr(struct kvm_vcpu *vcpu, u64 val) +{ + ia64_setreg(_IA64_REG_CR_DCR, val); +} + +static inline void vcpu_set_isr(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, isr) = val; +} + +static inline void vcpu_set_lid(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, lid) = val; +} + +static inline void vcpu_set_ipsr(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, ipsr) = val; +} + +static inline void vcpu_set_iip(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, iip) = val; +} + +static inline void vcpu_set_ifs(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, ifs) = val; +} + +static inline void vcpu_set_iipa(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, iipa) = val; +} + +static inline void vcpu_set_iha(struct kvm_vcpu *vcpu, u64 val) +{ + VCPU(vcpu, iha) = val; +} + + +static inline u64 vcpu_get_rr(struct kvm_vcpu *vcpu, u64 reg) +{ + return vcpu->arch.vrr[reg>>61]; +} + +/************************************************************************** + VCPU debug breakpoint register access routines + **************************************************************************/ + +static inline void vcpu_set_dbr(struct kvm_vcpu *vcpu, u64 reg, u64 val) +{ + __ia64_set_dbr(reg, val); +} + +static inline void vcpu_set_ibr(struct kvm_vcpu *vcpu, u64 reg, u64 val) +{ + ia64_set_ibr(reg, val); +} + +static inline u64 vcpu_get_dbr(struct kvm_vcpu *vcpu, u64 reg) +{ + return ((u64)__ia64_get_dbr(reg)); +} + +static inline u64 vcpu_get_ibr(struct kvm_vcpu *vcpu, u64 reg) +{ + return ((u64)ia64_get_ibr(reg)); +} + +/************************************************************************** + VCPU performance monitor register access routines + **************************************************************************/ +static inline void vcpu_set_pmc(struct kvm_vcpu *vcpu, u64 reg, u64 val) +{ + /* NOTE: Writes to unimplemented PMC registers are discarded */ + ia64_set_pmc(reg, val); +} + +static inline void vcpu_set_pmd(struct kvm_vcpu *vcpu, u64 reg, u64 val) +{ + /* NOTE: Writes to unimplemented PMD registers are discarded */ + ia64_set_pmd(reg, val); +} + +static inline u64 vcpu_get_pmc(struct kvm_vcpu *vcpu, u64 reg) +{ + /* NOTE: Reads from unimplemented PMC registers return zero */ + return ((u64)ia64_get_pmc(reg)); +} + +static inline u64 vcpu_get_pmd(struct kvm_vcpu *vcpu, u64 reg) +{ + /* NOTE: Reads from unimplemented PMD registers return zero */ + return ((u64)ia64_get_pmd(reg)); +} + +static inline unsigned long vrrtomrr(unsigned long val) +{ + union ia64_rr rr; + rr.val = val; + rr.rid = (rr.rid << 4) | 0xe; + if (rr.ps > PAGE_SHIFT) + rr.ps = PAGE_SHIFT; + rr.ve = 1; + return rr.val; +} + + +static inline int highest_bits(int *dat) +{ + u32 bits, bitnum; + int i; + + /* loop for all 256 bits */ + for (i = 7; i >= 0 ; i--) { + bits = dat[i]; + if (bits) { + bitnum = fls(bits); + return i * 32 + bitnum - 1; + } + } + return NULL_VECTOR; +} + +/* + * The pending irq is higher than the inservice one. + * + */ +static inline int is_higher_irq(int pending, int inservice) +{ + return ((pending > inservice) + || ((pending != NULL_VECTOR) + && (inservice == NULL_VECTOR))); +} + +static inline int is_higher_class(int pending, int mic) +{ + return ((pending >> 4) > mic); +} + +/* + * Return 0-255 for pending irq. + * NULL_VECTOR: when no pending. + */ +static inline int highest_pending_irq(struct kvm_vcpu *vcpu) +{ + if (VCPU(vcpu, irr[0]) & (1UL< +#include + +#include "vcpu.h" + +MODULE_AUTHOR("Intel"); +MODULE_LICENSE("GPL"); + +extern char kvm_ia64_ivt; +extern fpswa_interface_t *vmm_fpswa_interface; + +struct kvm_vmm_info vmm_info = { + .module = THIS_MODULE, + .vmm_entry = vmm_entry, + .tramp_entry = vmm_trampoline, + .vmm_ivt = (unsigned long)&kvm_ia64_ivt, +}; + +static int __init kvm_vmm_init(void) +{ + + vmm_fpswa_interface = fpswa_interface; + + /*Register vmm data to kvm side*/ + return kvm_init(&vmm_info, 1024, THIS_MODULE); +} + +static void __exit kvm_vmm_exit(void) +{ + kvm_exit(); + return ; +} + +void vmm_spin_lock(spinlock_t *lock) +{ + _vmm_raw_spin_lock(lock); +} + +void vmm_spin_unlock(spinlock_t *lock) +{ + _vmm_raw_spin_unlock(lock); +} +module_init(kvm_vmm_init) +module_exit(kvm_vmm_exit) diff --git a/arch/ia64/kvm/vmm_ivt.S b/arch/ia64/kvm/vmm_ivt.S new file mode 100644 index 0000000000000000000000000000000000000000..3ee5f481c06dccbd62f88940cacb8f2662567c71 --- /dev/null +++ b/arch/ia64/kvm/vmm_ivt.S @@ -0,0 +1,1424 @@ +/* + * /ia64/kvm_ivt.S + * + * Copyright (C) 1998-2001, 2003 Hewlett-Packard Co + * Stephane Eranian + * David Mosberger + * Copyright (C) 2000, 2002-2003 Intel Co + * Asit Mallick + * Suresh Siddha + * Kenneth Chen + * Fenghua Yu + * + * + * 00/08/23 Asit Mallick TLB handling + * for SMP + * 00/12/20 David Mosberger-Tang DTLB/ITLB + * handler now uses virtual PT. + * + * 07/6/20 Xuefei Xu (Anthony Xu) (anthony.xu@intel.com) + * Supporting Intel virtualization architecture + * + */ + +/* + * This file defines the interruption vector table used by the CPU. + * It does not include one entry per possible cause of interruption. + * + * The first 20 entries of the table contain 64 bundles each while the + * remaining 48 entries contain only 16 bundles each. + * + * The 64 bundles are used to allow inlining the whole handler for + * critical + * interruptions like TLB misses. + * + * For each entry, the comment is as follows: + * + * // 0x1c00 Entry 7 (size 64 bundles) Data Key Miss + * (12,51) + * entry offset ----/ / / / + * / + * entry number ---------/ / / + * / + * size of the entry -------------/ / + * / + * vector name -------------------------------------/ + * / + * interruptions triggering this vector + * ----------------------/ + * + * The table is 32KB in size and must be aligned on 32KB + * boundary. + * (The CPU ignores the 15 lower bits of the address) + * + * Table is based upon EAS2.6 (Oct 1999) + */ + + +#include +#include +#include + +#include "asm-offsets.h" +#include "vcpu.h" +#include "kvm_minstate.h" +#include "vti.h" + +#if 1 +# define PSR_DEFAULT_BITS psr.ac +#else +# define PSR_DEFAULT_BITS 0 +#endif + + +#define KVM_FAULT(n) \ + kvm_fault_##n:; \ + mov r19=n;; \ + br.sptk.many kvm_fault_##n; \ + ;; \ + + +#define KVM_REFLECT(n) \ + mov r31=pr; \ + mov r19=n; /* prepare to save predicates */ \ + mov r29=cr.ipsr; \ + ;; \ + tbit.z p6,p7=r29,IA64_PSR_VM_BIT; \ +(p7)br.sptk.many kvm_dispatch_reflection; \ + br.sptk.many kvm_panic; \ + + +GLOBAL_ENTRY(kvm_panic) + br.sptk.many kvm_panic + ;; +END(kvm_panic) + + + + + + .section .text.ivt,"ax" + + .align 32768 // align on 32KB boundary + .global kvm_ia64_ivt +kvm_ia64_ivt: +/////////////////////////////////////////////////////////////// +// 0x0000 Entry 0 (size 64 bundles) VHPT Translation (8,20,47) +ENTRY(kvm_vhpt_miss) + KVM_FAULT(0) +END(kvm_vhpt_miss) + + + .org kvm_ia64_ivt+0x400 +//////////////////////////////////////////////////////////////// +// 0x0400 Entry 1 (size 64 bundles) ITLB (21) +ENTRY(kvm_itlb_miss) + mov r31 = pr + mov r29=cr.ipsr; + ;; + tbit.z p6,p7=r29,IA64_PSR_VM_BIT; + (p6) br.sptk kvm_alt_itlb_miss + mov r19 = 1 + br.sptk kvm_itlb_miss_dispatch + KVM_FAULT(1); +END(kvm_itlb_miss) + + .org kvm_ia64_ivt+0x0800 +////////////////////////////////////////////////////////////////// +// 0x0800 Entry 2 (size 64 bundles) DTLB (9,48) +ENTRY(kvm_dtlb_miss) + mov r31 = pr + mov r29=cr.ipsr; + ;; + tbit.z p6,p7=r29,IA64_PSR_VM_BIT; +(p6)br.sptk kvm_alt_dtlb_miss + br.sptk kvm_dtlb_miss_dispatch +END(kvm_dtlb_miss) + + .org kvm_ia64_ivt+0x0c00 +//////////////////////////////////////////////////////////////////// +// 0x0c00 Entry 3 (size 64 bundles) Alt ITLB (19) +ENTRY(kvm_alt_itlb_miss) + mov r16=cr.ifa // get address that caused the TLB miss + ;; + movl r17=PAGE_KERNEL + mov r24=cr.ipsr + movl r19=(((1 << IA64_MAX_PHYS_BITS) - 1) & ~0xfff) + ;; + and r19=r19,r16 // clear ed, reserved bits, and PTE control bits + ;; + or r19=r17,r19 // insert PTE control bits into r19 + ;; + movl r20=IA64_GRANULE_SHIFT<<2 + ;; + mov cr.itir=r20 + ;; + itc.i r19 // insert the TLB entry + mov pr=r31,-1 + rfi +END(kvm_alt_itlb_miss) + + .org kvm_ia64_ivt+0x1000 +///////////////////////////////////////////////////////////////////// +// 0x1000 Entry 4 (size 64 bundles) Alt DTLB (7,46) +ENTRY(kvm_alt_dtlb_miss) + mov r16=cr.ifa // get address that caused the TLB miss + ;; + movl r17=PAGE_KERNEL + movl r19=(((1 << IA64_MAX_PHYS_BITS) - 1) & ~0xfff) + mov r24=cr.ipsr + ;; + and r19=r19,r16 // clear ed, reserved bits, and PTE control bits + ;; + or r19=r19,r17 // insert PTE control bits into r19 + ;; + movl r20=IA64_GRANULE_SHIFT<<2 + ;; + mov cr.itir=r20 + ;; + itc.d r19 // insert the TLB entry + mov pr=r31,-1 + rfi +END(kvm_alt_dtlb_miss) + + .org kvm_ia64_ivt+0x1400 +////////////////////////////////////////////////////////////////////// +// 0x1400 Entry 5 (size 64 bundles) Data nested TLB (6,45) +ENTRY(kvm_nested_dtlb_miss) + KVM_FAULT(5) +END(kvm_nested_dtlb_miss) + + .org kvm_ia64_ivt+0x1800 +///////////////////////////////////////////////////////////////////// +// 0x1800 Entry 6 (size 64 bundles) Instruction Key Miss (24) +ENTRY(kvm_ikey_miss) + KVM_REFLECT(6) +END(kvm_ikey_miss) + + .org kvm_ia64_ivt+0x1c00 +///////////////////////////////////////////////////////////////////// +// 0x1c00 Entry 7 (size 64 bundles) Data Key Miss (12,51) +ENTRY(kvm_dkey_miss) + KVM_REFLECT(7) +END(kvm_dkey_miss) + + .org kvm_ia64_ivt+0x2000 +//////////////////////////////////////////////////////////////////// +// 0x2000 Entry 8 (size 64 bundles) Dirty-bit (54) +ENTRY(kvm_dirty_bit) + KVM_REFLECT(8) +END(kvm_dirty_bit) + + .org kvm_ia64_ivt+0x2400 +//////////////////////////////////////////////////////////////////// +// 0x2400 Entry 9 (size 64 bundles) Instruction Access-bit (27) +ENTRY(kvm_iaccess_bit) + KVM_REFLECT(9) +END(kvm_iaccess_bit) + + .org kvm_ia64_ivt+0x2800 +/////////////////////////////////////////////////////////////////// +// 0x2800 Entry 10 (size 64 bundles) Data Access-bit (15,55) +ENTRY(kvm_daccess_bit) + KVM_REFLECT(10) +END(kvm_daccess_bit) + + .org kvm_ia64_ivt+0x2c00 +///////////////////////////////////////////////////////////////// +// 0x2c00 Entry 11 (size 64 bundles) Break instruction (33) +ENTRY(kvm_break_fault) + mov r31=pr + mov r19=11 + mov r29=cr.ipsr + ;; + KVM_SAVE_MIN_WITH_COVER_R19 + ;; + alloc r14=ar.pfs,0,0,4,0 // now it's safe (must be first in insn group!) + mov out0=cr.ifa + mov out2=cr.isr // FIXME: pity to make this slow access twice + mov out3=cr.iim // FIXME: pity to make this slow access twice + adds r3=8,r2 // set up second base pointer + ;; + ssm psr.ic + ;; + srlz.i // guarantee that interruption collection is on + ;; + //(p15)ssm psr.i // restore psr.i + addl r14=@gprel(ia64_leave_hypervisor),gp + ;; + KVM_SAVE_REST + mov rp=r14 + ;; + adds out1=16,sp + br.call.sptk.many b6=kvm_ia64_handle_break + ;; +END(kvm_break_fault) + + .org kvm_ia64_ivt+0x3000 +///////////////////////////////////////////////////////////////// +// 0x3000 Entry 12 (size 64 bundles) External Interrupt (4) +ENTRY(kvm_interrupt) + mov r31=pr // prepare to save predicates + mov r19=12 + mov r29=cr.ipsr + ;; + tbit.z p6,p7=r29,IA64_PSR_VM_BIT + tbit.z p0,p15=r29,IA64_PSR_I_BIT + ;; +(p7) br.sptk kvm_dispatch_interrupt + ;; + mov r27=ar.rsc /* M */ + mov r20=r1 /* A */ + mov r25=ar.unat /* M */ + mov r26=ar.pfs /* I */ + mov r28=cr.iip /* M */ + cover /* B (or nothing) */ + ;; + mov r1=sp + ;; + invala /* M */ + mov r30=cr.ifs + ;; + addl r1=-VMM_PT_REGS_SIZE,r1 + ;; + adds r17=2*L1_CACHE_BYTES,r1 /* really: biggest cache-line size */ + adds r16=PT(CR_IPSR),r1 + ;; + lfetch.fault.excl.nt1 [r17],L1_CACHE_BYTES + st8 [r16]=r29 /* save cr.ipsr */ + ;; + lfetch.fault.excl.nt1 [r17] + mov r29=b0 + ;; + adds r16=PT(R8),r1 /* initialize first base pointer */ + adds r17=PT(R9),r1 /* initialize second base pointer */ + mov r18=r0 /* make sure r18 isn't NaT */ + ;; +.mem.offset 0,0; st8.spill [r16]=r8,16 +.mem.offset 8,0; st8.spill [r17]=r9,16 + ;; +.mem.offset 0,0; st8.spill [r16]=r10,24 +.mem.offset 8,0; st8.spill [r17]=r11,24 + ;; + st8 [r16]=r28,16 /* save cr.iip */ + st8 [r17]=r30,16 /* save cr.ifs */ + mov r8=ar.fpsr /* M */ + mov r9=ar.csd + mov r10=ar.ssd + movl r11=FPSR_DEFAULT /* L-unit */ + ;; + st8 [r16]=r25,16 /* save ar.unat */ + st8 [r17]=r26,16 /* save ar.pfs */ + shl r18=r18,16 /* compute ar.rsc to be used for "loadrs" */ + ;; + st8 [r16]=r27,16 /* save ar.rsc */ + adds r17=16,r17 /* skip over ar_rnat field */ + ;; + st8 [r17]=r31,16 /* save predicates */ + adds r16=16,r16 /* skip over ar_bspstore field */ + ;; + st8 [r16]=r29,16 /* save b0 */ + st8 [r17]=r18,16 /* save ar.rsc value for "loadrs" */ + ;; +.mem.offset 0,0; st8.spill [r16]=r20,16 /* save original r1 */ +.mem.offset 8,0; st8.spill [r17]=r12,16 + adds r12=-16,r1 + /* switch to kernel memory stack (with 16 bytes of scratch) */ + ;; +.mem.offset 0,0; st8.spill [r16]=r13,16 +.mem.offset 8,0; st8.spill [r17]=r8,16 /* save ar.fpsr */ + ;; +.mem.offset 0,0; st8.spill [r16]=r15,16 +.mem.offset 8,0; st8.spill [r17]=r14,16 + dep r14=-1,r0,60,4 + ;; +.mem.offset 0,0; st8.spill [r16]=r2,16 +.mem.offset 8,0; st8.spill [r17]=r3,16 + adds r2=VMM_PT_REGS_R16_OFFSET,r1 + adds r14 = VMM_VCPU_GP_OFFSET,r13 + ;; + mov r8=ar.ccv + ld8 r14 = [r14] + ;; + mov r1=r14 /* establish kernel global pointer */ + ;; \ + bsw.1 + ;; + alloc r14=ar.pfs,0,0,1,0 // must be first in an insn group + mov out0=r13 + ;; + ssm psr.ic + ;; + srlz.i + ;; + //(p15) ssm psr.i + adds r3=8,r2 // set up second base pointer for SAVE_REST + srlz.i // ensure everybody knows psr.ic is back on + ;; +.mem.offset 0,0; st8.spill [r2]=r16,16 +.mem.offset 8,0; st8.spill [r3]=r17,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r18,16 +.mem.offset 8,0; st8.spill [r3]=r19,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r20,16 +.mem.offset 8,0; st8.spill [r3]=r21,16 + mov r18=b6 + ;; +.mem.offset 0,0; st8.spill [r2]=r22,16 +.mem.offset 8,0; st8.spill [r3]=r23,16 + mov r19=b7 + ;; +.mem.offset 0,0; st8.spill [r2]=r24,16 +.mem.offset 8,0; st8.spill [r3]=r25,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r26,16 +.mem.offset 8,0; st8.spill [r3]=r27,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r28,16 +.mem.offset 8,0; st8.spill [r3]=r29,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r30,16 +.mem.offset 8,0; st8.spill [r3]=r31,32 + ;; + mov ar.fpsr=r11 /* M-unit */ + st8 [r2]=r8,8 /* ar.ccv */ + adds r24=PT(B6)-PT(F7),r3 + ;; + stf.spill [r2]=f6,32 + stf.spill [r3]=f7,32 + ;; + stf.spill [r2]=f8,32 + stf.spill [r3]=f9,32 + ;; + stf.spill [r2]=f10 + stf.spill [r3]=f11 + adds r25=PT(B7)-PT(F11),r3 + ;; + st8 [r24]=r18,16 /* b6 */ + st8 [r25]=r19,16 /* b7 */ + ;; + st8 [r24]=r9 /* ar.csd */ + st8 [r25]=r10 /* ar.ssd */ + ;; + srlz.d // make sure we see the effect of cr.ivr + addl r14=@gprel(ia64_leave_nested),gp + ;; + mov rp=r14 + br.call.sptk.many b6=kvm_ia64_handle_irq + ;; +END(kvm_interrupt) + + .global kvm_dispatch_vexirq + .org kvm_ia64_ivt+0x3400 +////////////////////////////////////////////////////////////////////// +// 0x3400 Entry 13 (size 64 bundles) Reserved +ENTRY(kvm_virtual_exirq) + mov r31=pr + mov r19=13 + mov r30 =r0 + ;; +kvm_dispatch_vexirq: + cmp.eq p6,p0 = 1,r30 + ;; +(p6)add r29 = VMM_VCPU_SAVED_GP_OFFSET,r21 + ;; +(p6)ld8 r1 = [r29] + ;; + KVM_SAVE_MIN_WITH_COVER_R19 + alloc r14=ar.pfs,0,0,1,0 + mov out0=r13 + + ssm psr.ic + ;; + srlz.i // guarantee that interruption collection is on + ;; + //(p15) ssm psr.i // restore psr.i + adds r3=8,r2 // set up second base pointer + ;; + KVM_SAVE_REST + addl r14=@gprel(ia64_leave_hypervisor),gp + ;; + mov rp=r14 + br.call.sptk.many b6=kvm_vexirq +END(kvm_virtual_exirq) + + .org kvm_ia64_ivt+0x3800 +///////////////////////////////////////////////////////////////////// +// 0x3800 Entry 14 (size 64 bundles) Reserved + KVM_FAULT(14) + // this code segment is from 2.6.16.13 + + + .org kvm_ia64_ivt+0x3c00 +/////////////////////////////////////////////////////////////////////// +// 0x3c00 Entry 15 (size 64 bundles) Reserved + KVM_FAULT(15) + + + .org kvm_ia64_ivt+0x4000 +/////////////////////////////////////////////////////////////////////// +// 0x4000 Entry 16 (size 64 bundles) Reserved + KVM_FAULT(16) + + .org kvm_ia64_ivt+0x4400 +////////////////////////////////////////////////////////////////////// +// 0x4400 Entry 17 (size 64 bundles) Reserved + KVM_FAULT(17) + + .org kvm_ia64_ivt+0x4800 +////////////////////////////////////////////////////////////////////// +// 0x4800 Entry 18 (size 64 bundles) Reserved + KVM_FAULT(18) + + .org kvm_ia64_ivt+0x4c00 +////////////////////////////////////////////////////////////////////// +// 0x4c00 Entry 19 (size 64 bundles) Reserved + KVM_FAULT(19) + + .org kvm_ia64_ivt+0x5000 +////////////////////////////////////////////////////////////////////// +// 0x5000 Entry 20 (size 16 bundles) Page Not Present +ENTRY(kvm_page_not_present) + KVM_REFLECT(20) +END(kvm_page_not_present) + + .org kvm_ia64_ivt+0x5100 +/////////////////////////////////////////////////////////////////////// +// 0x5100 Entry 21 (size 16 bundles) Key Permission vector +ENTRY(kvm_key_permission) + KVM_REFLECT(21) +END(kvm_key_permission) + + .org kvm_ia64_ivt+0x5200 +////////////////////////////////////////////////////////////////////// +// 0x5200 Entry 22 (size 16 bundles) Instruction Access Rights (26) +ENTRY(kvm_iaccess_rights) + KVM_REFLECT(22) +END(kvm_iaccess_rights) + + .org kvm_ia64_ivt+0x5300 +////////////////////////////////////////////////////////////////////// +// 0x5300 Entry 23 (size 16 bundles) Data Access Rights (14,53) +ENTRY(kvm_daccess_rights) + KVM_REFLECT(23) +END(kvm_daccess_rights) + + .org kvm_ia64_ivt+0x5400 +///////////////////////////////////////////////////////////////////// +// 0x5400 Entry 24 (size 16 bundles) General Exception (5,32,34,36,38,39) +ENTRY(kvm_general_exception) + KVM_REFLECT(24) + KVM_FAULT(24) +END(kvm_general_exception) + + .org kvm_ia64_ivt+0x5500 +////////////////////////////////////////////////////////////////////// +// 0x5500 Entry 25 (size 16 bundles) Disabled FP-Register (35) +ENTRY(kvm_disabled_fp_reg) + KVM_REFLECT(25) +END(kvm_disabled_fp_reg) + + .org kvm_ia64_ivt+0x5600 +//////////////////////////////////////////////////////////////////// +// 0x5600 Entry 26 (size 16 bundles) Nat Consumption (11,23,37,50) +ENTRY(kvm_nat_consumption) + KVM_REFLECT(26) +END(kvm_nat_consumption) + + .org kvm_ia64_ivt+0x5700 +///////////////////////////////////////////////////////////////////// +// 0x5700 Entry 27 (size 16 bundles) Speculation (40) +ENTRY(kvm_speculation_vector) + KVM_REFLECT(27) +END(kvm_speculation_vector) + + .org kvm_ia64_ivt+0x5800 +///////////////////////////////////////////////////////////////////// +// 0x5800 Entry 28 (size 16 bundles) Reserved + KVM_FAULT(28) + + .org kvm_ia64_ivt+0x5900 +/////////////////////////////////////////////////////////////////// +// 0x5900 Entry 29 (size 16 bundles) Debug (16,28,56) +ENTRY(kvm_debug_vector) + KVM_FAULT(29) +END(kvm_debug_vector) + + .org kvm_ia64_ivt+0x5a00 +/////////////////////////////////////////////////////////////// +// 0x5a00 Entry 30 (size 16 bundles) Unaligned Reference (57) +ENTRY(kvm_unaligned_access) + KVM_REFLECT(30) +END(kvm_unaligned_access) + + .org kvm_ia64_ivt+0x5b00 +////////////////////////////////////////////////////////////////////// +// 0x5b00 Entry 31 (size 16 bundles) Unsupported Data Reference (57) +ENTRY(kvm_unsupported_data_reference) + KVM_REFLECT(31) +END(kvm_unsupported_data_reference) + + .org kvm_ia64_ivt+0x5c00 +//////////////////////////////////////////////////////////////////// +// 0x5c00 Entry 32 (size 16 bundles) Floating Point FAULT (65) +ENTRY(kvm_floating_point_fault) + KVM_REFLECT(32) +END(kvm_floating_point_fault) + + .org kvm_ia64_ivt+0x5d00 +///////////////////////////////////////////////////////////////////// +// 0x5d00 Entry 33 (size 16 bundles) Floating Point Trap (66) +ENTRY(kvm_floating_point_trap) + KVM_REFLECT(33) +END(kvm_floating_point_trap) + + .org kvm_ia64_ivt+0x5e00 +////////////////////////////////////////////////////////////////////// +// 0x5e00 Entry 34 (size 16 bundles) Lower Privilege Transfer Trap (66) +ENTRY(kvm_lower_privilege_trap) + KVM_REFLECT(34) +END(kvm_lower_privilege_trap) + + .org kvm_ia64_ivt+0x5f00 +////////////////////////////////////////////////////////////////////// +// 0x5f00 Entry 35 (size 16 bundles) Taken Branch Trap (68) +ENTRY(kvm_taken_branch_trap) + KVM_REFLECT(35) +END(kvm_taken_branch_trap) + + .org kvm_ia64_ivt+0x6000 +//////////////////////////////////////////////////////////////////// +// 0x6000 Entry 36 (size 16 bundles) Single Step Trap (69) +ENTRY(kvm_single_step_trap) + KVM_REFLECT(36) +END(kvm_single_step_trap) + .global kvm_virtualization_fault_back + .org kvm_ia64_ivt+0x6100 +///////////////////////////////////////////////////////////////////// +// 0x6100 Entry 37 (size 16 bundles) Virtualization Fault +ENTRY(kvm_virtualization_fault) + mov r31=pr + adds r16 = VMM_VCPU_SAVED_GP_OFFSET,r21 + ;; + st8 [r16] = r1 + adds r17 = VMM_VCPU_GP_OFFSET, r21 + ;; + ld8 r1 = [r17] + cmp.eq p6,p0=EVENT_MOV_FROM_AR,r24 + cmp.eq p7,p0=EVENT_MOV_FROM_RR,r24 + cmp.eq p8,p0=EVENT_MOV_TO_RR,r24 + cmp.eq p9,p0=EVENT_RSM,r24 + cmp.eq p10,p0=EVENT_SSM,r24 + cmp.eq p11,p0=EVENT_MOV_TO_PSR,r24 + cmp.eq p12,p0=EVENT_THASH,r24 + (p6) br.dptk.many kvm_asm_mov_from_ar + (p7) br.dptk.many kvm_asm_mov_from_rr + (p8) br.dptk.many kvm_asm_mov_to_rr + (p9) br.dptk.many kvm_asm_rsm + (p10) br.dptk.many kvm_asm_ssm + (p11) br.dptk.many kvm_asm_mov_to_psr + (p12) br.dptk.many kvm_asm_thash + ;; +kvm_virtualization_fault_back: + adds r16 = VMM_VCPU_SAVED_GP_OFFSET,r21 + ;; + ld8 r1 = [r16] + ;; + mov r19=37 + adds r16 = VMM_VCPU_CAUSE_OFFSET,r21 + adds r17 = VMM_VCPU_OPCODE_OFFSET,r21 + ;; + st8 [r16] = r24 + st8 [r17] = r25 + ;; + cmp.ne p6,p0=EVENT_RFI, r24 + (p6) br.sptk kvm_dispatch_virtualization_fault + ;; + adds r18=VMM_VPD_BASE_OFFSET,r21 + ;; + ld8 r18=[r18] + ;; + adds r18=VMM_VPD_VIFS_OFFSET,r18 + ;; + ld8 r18=[r18] + ;; + tbit.z p6,p0=r18,63 + (p6) br.sptk kvm_dispatch_virtualization_fault + ;; + //if vifs.v=1 desert current register frame + alloc r18=ar.pfs,0,0,0,0 + br.sptk kvm_dispatch_virtualization_fault +END(kvm_virtualization_fault) + + .org kvm_ia64_ivt+0x6200 +////////////////////////////////////////////////////////////// +// 0x6200 Entry 38 (size 16 bundles) Reserved + KVM_FAULT(38) + + .org kvm_ia64_ivt+0x6300 +///////////////////////////////////////////////////////////////// +// 0x6300 Entry 39 (size 16 bundles) Reserved + KVM_FAULT(39) + + .org kvm_ia64_ivt+0x6400 +///////////////////////////////////////////////////////////////// +// 0x6400 Entry 40 (size 16 bundles) Reserved + KVM_FAULT(40) + + .org kvm_ia64_ivt+0x6500 +////////////////////////////////////////////////////////////////// +// 0x6500 Entry 41 (size 16 bundles) Reserved + KVM_FAULT(41) + + .org kvm_ia64_ivt+0x6600 +////////////////////////////////////////////////////////////////// +// 0x6600 Entry 42 (size 16 bundles) Reserved + KVM_FAULT(42) + + .org kvm_ia64_ivt+0x6700 +////////////////////////////////////////////////////////////////// +// 0x6700 Entry 43 (size 16 bundles) Reserved + KVM_FAULT(43) + + .org kvm_ia64_ivt+0x6800 +////////////////////////////////////////////////////////////////// +// 0x6800 Entry 44 (size 16 bundles) Reserved + KVM_FAULT(44) + + .org kvm_ia64_ivt+0x6900 +/////////////////////////////////////////////////////////////////// +// 0x6900 Entry 45 (size 16 bundles) IA-32 Exeception +//(17,18,29,41,42,43,44,58,60,61,62,72,73,75,76,77) +ENTRY(kvm_ia32_exception) + KVM_FAULT(45) +END(kvm_ia32_exception) + + .org kvm_ia64_ivt+0x6a00 +//////////////////////////////////////////////////////////////////// +// 0x6a00 Entry 46 (size 16 bundles) IA-32 Intercept (30,31,59,70,71) +ENTRY(kvm_ia32_intercept) + KVM_FAULT(47) +END(kvm_ia32_intercept) + + .org kvm_ia64_ivt+0x6c00 +///////////////////////////////////////////////////////////////////// +// 0x6c00 Entry 48 (size 16 bundles) Reserved + KVM_FAULT(48) + + .org kvm_ia64_ivt+0x6d00 +////////////////////////////////////////////////////////////////////// +// 0x6d00 Entry 49 (size 16 bundles) Reserved + KVM_FAULT(49) + + .org kvm_ia64_ivt+0x6e00 +////////////////////////////////////////////////////////////////////// +// 0x6e00 Entry 50 (size 16 bundles) Reserved + KVM_FAULT(50) + + .org kvm_ia64_ivt+0x6f00 +///////////////////////////////////////////////////////////////////// +// 0x6f00 Entry 51 (size 16 bundles) Reserved + KVM_FAULT(52) + + .org kvm_ia64_ivt+0x7100 +//////////////////////////////////////////////////////////////////// +// 0x7100 Entry 53 (size 16 bundles) Reserved + KVM_FAULT(53) + + .org kvm_ia64_ivt+0x7200 +///////////////////////////////////////////////////////////////////// +// 0x7200 Entry 54 (size 16 bundles) Reserved + KVM_FAULT(54) + + .org kvm_ia64_ivt+0x7300 +//////////////////////////////////////////////////////////////////// +// 0x7300 Entry 55 (size 16 bundles) Reserved + KVM_FAULT(55) + + .org kvm_ia64_ivt+0x7400 +//////////////////////////////////////////////////////////////////// +// 0x7400 Entry 56 (size 16 bundles) Reserved + KVM_FAULT(56) + + .org kvm_ia64_ivt+0x7500 +///////////////////////////////////////////////////////////////////// +// 0x7500 Entry 57 (size 16 bundles) Reserved + KVM_FAULT(57) + + .org kvm_ia64_ivt+0x7600 +///////////////////////////////////////////////////////////////////// +// 0x7600 Entry 58 (size 16 bundles) Reserved + KVM_FAULT(58) + + .org kvm_ia64_ivt+0x7700 +//////////////////////////////////////////////////////////////////// +// 0x7700 Entry 59 (size 16 bundles) Reserved + KVM_FAULT(59) + + .org kvm_ia64_ivt+0x7800 +//////////////////////////////////////////////////////////////////// +// 0x7800 Entry 60 (size 16 bundles) Reserved + KVM_FAULT(60) + + .org kvm_ia64_ivt+0x7900 +///////////////////////////////////////////////////////////////////// +// 0x7900 Entry 61 (size 16 bundles) Reserved + KVM_FAULT(61) + + .org kvm_ia64_ivt+0x7a00 +///////////////////////////////////////////////////////////////////// +// 0x7a00 Entry 62 (size 16 bundles) Reserved + KVM_FAULT(62) + + .org kvm_ia64_ivt+0x7b00 +///////////////////////////////////////////////////////////////////// +// 0x7b00 Entry 63 (size 16 bundles) Reserved + KVM_FAULT(63) + + .org kvm_ia64_ivt+0x7c00 +//////////////////////////////////////////////////////////////////// +// 0x7c00 Entry 64 (size 16 bundles) Reserved + KVM_FAULT(64) + + .org kvm_ia64_ivt+0x7d00 +///////////////////////////////////////////////////////////////////// +// 0x7d00 Entry 65 (size 16 bundles) Reserved + KVM_FAULT(65) + + .org kvm_ia64_ivt+0x7e00 +///////////////////////////////////////////////////////////////////// +// 0x7e00 Entry 66 (size 16 bundles) Reserved + KVM_FAULT(66) + + .org kvm_ia64_ivt+0x7f00 +//////////////////////////////////////////////////////////////////// +// 0x7f00 Entry 67 (size 16 bundles) Reserved + KVM_FAULT(67) + + .org kvm_ia64_ivt+0x8000 +// There is no particular reason for this code to be here, other than that +// there happens to be space here that would go unused otherwise. If this +// fault ever gets "unreserved", simply moved the following code to a more +// suitable spot... + + +ENTRY(kvm_dtlb_miss_dispatch) + mov r19 = 2 + KVM_SAVE_MIN_WITH_COVER_R19 + alloc r14=ar.pfs,0,0,3,0 + mov out0=cr.ifa + mov out1=r15 + adds r3=8,r2 // set up second base pointer + ;; + ssm psr.ic + ;; + srlz.i // guarantee that interruption collection is on + ;; + //(p15) ssm psr.i // restore psr.i + addl r14=@gprel(ia64_leave_hypervisor_prepare),gp + ;; + KVM_SAVE_REST + KVM_SAVE_EXTRA + mov rp=r14 + ;; + adds out2=16,r12 + br.call.sptk.many b6=kvm_page_fault +END(kvm_dtlb_miss_dispatch) + +ENTRY(kvm_itlb_miss_dispatch) + + KVM_SAVE_MIN_WITH_COVER_R19 + alloc r14=ar.pfs,0,0,3,0 + mov out0=cr.ifa + mov out1=r15 + adds r3=8,r2 // set up second base pointer + ;; + ssm psr.ic + ;; + srlz.i // guarantee that interruption collection is on + ;; + //(p15) ssm psr.i // restore psr.i + addl r14=@gprel(ia64_leave_hypervisor),gp + ;; + KVM_SAVE_REST + mov rp=r14 + ;; + adds out2=16,r12 + br.call.sptk.many b6=kvm_page_fault +END(kvm_itlb_miss_dispatch) + +ENTRY(kvm_dispatch_reflection) + /* + * Input: + * psr.ic: off + * r19: intr type (offset into ivt, see ia64_int.h) + * r31: contains saved predicates (pr) + */ + KVM_SAVE_MIN_WITH_COVER_R19 + alloc r14=ar.pfs,0,0,5,0 + mov out0=cr.ifa + mov out1=cr.isr + mov out2=cr.iim + mov out3=r15 + adds r3=8,r2 // set up second base pointer + ;; + ssm psr.ic + ;; + srlz.i // guarantee that interruption collection is on + ;; + //(p15) ssm psr.i // restore psr.i + addl r14=@gprel(ia64_leave_hypervisor),gp + ;; + KVM_SAVE_REST + mov rp=r14 + ;; + adds out4=16,r12 + br.call.sptk.many b6=reflect_interruption +END(kvm_dispatch_reflection) + +ENTRY(kvm_dispatch_virtualization_fault) + adds r16 = VMM_VCPU_CAUSE_OFFSET,r21 + adds r17 = VMM_VCPU_OPCODE_OFFSET,r21 + ;; + st8 [r16] = r24 + st8 [r17] = r25 + ;; + KVM_SAVE_MIN_WITH_COVER_R19 + ;; + alloc r14=ar.pfs,0,0,2,0 // now it's safe (must be first in insn group!) + mov out0=r13 //vcpu + adds r3=8,r2 // set up second base pointer + ;; + ssm psr.ic + ;; + srlz.i // guarantee that interruption collection is on + ;; + //(p15) ssm psr.i // restore psr.i + addl r14=@gprel(ia64_leave_hypervisor_prepare),gp + ;; + KVM_SAVE_REST + KVM_SAVE_EXTRA + mov rp=r14 + ;; + adds out1=16,sp //regs + br.call.sptk.many b6=kvm_emulate +END(kvm_dispatch_virtualization_fault) + + +ENTRY(kvm_dispatch_interrupt) + KVM_SAVE_MIN_WITH_COVER_R19 // uses r31; defines r2 and r3 + ;; + alloc r14=ar.pfs,0,0,1,0 // must be first in an insn group + //mov out0=cr.ivr // pass cr.ivr as first arg + adds r3=8,r2 // set up second base pointer for SAVE_REST + ;; + ssm psr.ic + ;; + srlz.i + ;; + //(p15) ssm psr.i + addl r14=@gprel(ia64_leave_hypervisor),gp + ;; + KVM_SAVE_REST + mov rp=r14 + ;; + mov out0=r13 // pass pointer to pt_regs as second arg + br.call.sptk.many b6=kvm_ia64_handle_irq +END(kvm_dispatch_interrupt) + + + + +GLOBAL_ENTRY(ia64_leave_nested) + rsm psr.i + ;; + adds r21=PT(PR)+16,r12 + ;; + lfetch [r21],PT(CR_IPSR)-PT(PR) + adds r2=PT(B6)+16,r12 + adds r3=PT(R16)+16,r12 + ;; + lfetch [r21] + ld8 r28=[r2],8 // load b6 + adds r29=PT(R24)+16,r12 + + ld8.fill r16=[r3] + adds r3=PT(AR_CSD)-PT(R16),r3 + adds r30=PT(AR_CCV)+16,r12 + ;; + ld8.fill r24=[r29] + ld8 r15=[r30] // load ar.ccv + ;; + ld8 r29=[r2],16 // load b7 + ld8 r30=[r3],16 // load ar.csd + ;; + ld8 r31=[r2],16 // load ar.ssd + ld8.fill r8=[r3],16 + ;; + ld8.fill r9=[r2],16 + ld8.fill r10=[r3],PT(R17)-PT(R10) + ;; + ld8.fill r11=[r2],PT(R18)-PT(R11) + ld8.fill r17=[r3],16 + ;; + ld8.fill r18=[r2],16 + ld8.fill r19=[r3],16 + ;; + ld8.fill r20=[r2],16 + ld8.fill r21=[r3],16 + mov ar.csd=r30 + mov ar.ssd=r31 + ;; + rsm psr.i | psr.ic + // initiate turning off of interrupt and interruption collection + invala // invalidate ALAT + ;; + srlz.i + ;; + ld8.fill r22=[r2],24 + ld8.fill r23=[r3],24 + mov b6=r28 + ;; + ld8.fill r25=[r2],16 + ld8.fill r26=[r3],16 + mov b7=r29 + ;; + ld8.fill r27=[r2],16 + ld8.fill r28=[r3],16 + ;; + ld8.fill r29=[r2],16 + ld8.fill r30=[r3],24 + ;; + ld8.fill r31=[r2],PT(F9)-PT(R31) + adds r3=PT(F10)-PT(F6),r3 + ;; + ldf.fill f9=[r2],PT(F6)-PT(F9) + ldf.fill f10=[r3],PT(F8)-PT(F10) + ;; + ldf.fill f6=[r2],PT(F7)-PT(F6) + ;; + ldf.fill f7=[r2],PT(F11)-PT(F7) + ldf.fill f8=[r3],32 + ;; + srlz.i // ensure interruption collection is off + mov ar.ccv=r15 + ;; + bsw.0 // switch back to bank 0 (no stop bit required beforehand...) + ;; + ldf.fill f11=[r2] +// mov r18=r13 +// mov r21=r13 + adds r16=PT(CR_IPSR)+16,r12 + adds r17=PT(CR_IIP)+16,r12 + ;; + ld8 r29=[r16],16 // load cr.ipsr + ld8 r28=[r17],16 // load cr.iip + ;; + ld8 r30=[r16],16 // load cr.ifs + ld8 r25=[r17],16 // load ar.unat + ;; + ld8 r26=[r16],16 // load ar.pfs + ld8 r27=[r17],16 // load ar.rsc + cmp.eq p9,p0=r0,r0 + // set p9 to indicate that we should restore cr.ifs + ;; + ld8 r24=[r16],16 // load ar.rnat (may be garbage) + ld8 r23=[r17],16// load ar.bspstore (may be garbage) + ;; + ld8 r31=[r16],16 // load predicates + ld8 r22=[r17],16 // load b0 + ;; + ld8 r19=[r16],16 // load ar.rsc value for "loadrs" + ld8.fill r1=[r17],16 // load r1 + ;; + ld8.fill r12=[r16],16 + ld8.fill r13=[r17],16 + ;; + ld8 r20=[r16],16 // ar.fpsr + ld8.fill r15=[r17],16 + ;; + ld8.fill r14=[r16],16 + ld8.fill r2=[r17] + ;; + ld8.fill r3=[r16] + ;; + mov r16=ar.bsp // get existing backing store pointer + ;; + mov b0=r22 + mov ar.pfs=r26 + mov cr.ifs=r30 + mov cr.ipsr=r29 + mov ar.fpsr=r20 + mov cr.iip=r28 + ;; + mov ar.rsc=r27 + mov ar.unat=r25 + mov pr=r31,-1 + rfi +END(ia64_leave_nested) + + + +GLOBAL_ENTRY(ia64_leave_hypervisor_prepare) + /* + * work.need_resched etc. mustn't get changed + *by this CPU before it returns to + ;; + * user- or fsys-mode, hence we disable interrupts early on: + */ + adds r2 = PT(R4)+16,r12 + adds r3 = PT(R5)+16,r12 + adds r8 = PT(EML_UNAT)+16,r12 + ;; + ld8 r8 = [r8] + ;; + mov ar.unat=r8 + ;; + ld8.fill r4=[r2],16 //load r4 + ld8.fill r5=[r3],16 //load r5 + ;; + ld8.fill r6=[r2] //load r6 + ld8.fill r7=[r3] //load r7 + ;; +END(ia64_leave_hypervisor_prepare) +//fall through +GLOBAL_ENTRY(ia64_leave_hypervisor) + rsm psr.i + ;; + br.call.sptk.many b0=leave_hypervisor_tail + ;; + adds r20=PT(PR)+16,r12 + adds r8=PT(EML_UNAT)+16,r12 + ;; + ld8 r8=[r8] + ;; + mov ar.unat=r8 + ;; + lfetch [r20],PT(CR_IPSR)-PT(PR) + adds r2 = PT(B6)+16,r12 + adds r3 = PT(B7)+16,r12 + ;; + lfetch [r20] + ;; + ld8 r24=[r2],16 /* B6 */ + ld8 r25=[r3],16 /* B7 */ + ;; + ld8 r26=[r2],16 /* ar_csd */ + ld8 r27=[r3],16 /* ar_ssd */ + mov b6 = r24 + ;; + ld8.fill r8=[r2],16 + ld8.fill r9=[r3],16 + mov b7 = r25 + ;; + mov ar.csd = r26 + mov ar.ssd = r27 + ;; + ld8.fill r10=[r2],PT(R15)-PT(R10) + ld8.fill r11=[r3],PT(R14)-PT(R11) + ;; + ld8.fill r15=[r2],PT(R16)-PT(R15) + ld8.fill r14=[r3],PT(R17)-PT(R14) + ;; + ld8.fill r16=[r2],16 + ld8.fill r17=[r3],16 + ;; + ld8.fill r18=[r2],16 + ld8.fill r19=[r3],16 + ;; + ld8.fill r20=[r2],16 + ld8.fill r21=[r3],16 + ;; + ld8.fill r22=[r2],16 + ld8.fill r23=[r3],16 + ;; + ld8.fill r24=[r2],16 + ld8.fill r25=[r3],16 + ;; + ld8.fill r26=[r2],16 + ld8.fill r27=[r3],16 + ;; + ld8.fill r28=[r2],16 + ld8.fill r29=[r3],16 + ;; + ld8.fill r30=[r2],PT(F6)-PT(R30) + ld8.fill r31=[r3],PT(F7)-PT(R31) + ;; + rsm psr.i | psr.ic + // initiate turning off of interrupt and interruption collection + invala // invalidate ALAT + ;; + srlz.i // ensure interruption collection is off + ;; + bsw.0 + ;; + adds r16 = PT(CR_IPSR)+16,r12 + adds r17 = PT(CR_IIP)+16,r12 + mov r21=r13 // get current + ;; + ld8 r31=[r16],16 // load cr.ipsr + ld8 r30=[r17],16 // load cr.iip + ;; + ld8 r29=[r16],16 // load cr.ifs + ld8 r28=[r17],16 // load ar.unat + ;; + ld8 r27=[r16],16 // load ar.pfs + ld8 r26=[r17],16 // load ar.rsc + ;; + ld8 r25=[r16],16 // load ar.rnat + ld8 r24=[r17],16 // load ar.bspstore + ;; + ld8 r23=[r16],16 // load predicates + ld8 r22=[r17],16 // load b0 + ;; + ld8 r20=[r16],16 // load ar.rsc value for "loadrs" + ld8.fill r1=[r17],16 //load r1 + ;; + ld8.fill r12=[r16],16 //load r12 + ld8.fill r13=[r17],PT(R2)-PT(R13) //load r13 + ;; + ld8 r19=[r16],PT(R3)-PT(AR_FPSR) //load ar_fpsr + ld8.fill r2=[r17],PT(AR_CCV)-PT(R2) //load r2 + ;; + ld8.fill r3=[r16] //load r3 + ld8 r18=[r17] //load ar_ccv + ;; + mov ar.fpsr=r19 + mov ar.ccv=r18 + shr.u r18=r20,16 + ;; +kvm_rbs_switch: + mov r19=96 + +kvm_dont_preserve_current_frame: +/* + * To prevent leaking bits between the hypervisor and guest domain, + * we must clear the stacked registers in the "invalid" partition here. + * 5 registers/cycle on McKinley). + */ +# define pRecurse p6 +# define pReturn p7 +# define Nregs 14 + + alloc loc0=ar.pfs,2,Nregs-2,2,0 + shr.u loc1=r18,9 // RNaTslots <= floor(dirtySize / (64*8)) + sub r19=r19,r18 // r19 = (physStackedSize + 8) - dirtySize + ;; + mov ar.rsc=r20 // load ar.rsc to be used for "loadrs" + shladd in0=loc1,3,r19 + mov in1=0 + ;; + TEXT_ALIGN(32) +kvm_rse_clear_invalid: + alloc loc0=ar.pfs,2,Nregs-2,2,0 + cmp.lt pRecurse,p0=Nregs*8,in0 + // if more than Nregs regs left to clear, (re)curse + add out0=-Nregs*8,in0 + add out1=1,in1 // increment recursion count + mov loc1=0 + mov loc2=0 + ;; + mov loc3=0 + mov loc4=0 + mov loc5=0 + mov loc6=0 + mov loc7=0 +(pRecurse) br.call.dptk.few b0=kvm_rse_clear_invalid + ;; + mov loc8=0 + mov loc9=0 + cmp.ne pReturn,p0=r0,in1 + // if recursion count != 0, we need to do a br.ret + mov loc10=0 + mov loc11=0 +(pReturn) br.ret.dptk.many b0 + +# undef pRecurse +# undef pReturn + +// loadrs has already been shifted + alloc r16=ar.pfs,0,0,0,0 // drop current register frame + ;; + loadrs + ;; + mov ar.bspstore=r24 + ;; + mov ar.unat=r28 + mov ar.rnat=r25 + mov ar.rsc=r26 + ;; + mov cr.ipsr=r31 + mov cr.iip=r30 + mov cr.ifs=r29 + mov ar.pfs=r27 + adds r18=VMM_VPD_BASE_OFFSET,r21 + ;; + ld8 r18=[r18] //vpd + adds r17=VMM_VCPU_ISR_OFFSET,r21 + ;; + ld8 r17=[r17] + adds r19=VMM_VPD_VPSR_OFFSET,r18 + ;; + ld8 r19=[r19] //vpsr + adds r20=VMM_VCPU_VSA_BASE_OFFSET,r21 + ;; + ld8 r20=[r20] + ;; +//vsa_sync_write_start + mov r25=r18 + adds r16= VMM_VCPU_GP_OFFSET,r21 + ;; + ld8 r16= [r16] // Put gp in r24 + movl r24=@gprel(ia64_vmm_entry) // calculate return address + ;; + add r24=r24,r16 + ;; + add r16=PAL_VPS_SYNC_WRITE,r20 + ;; + mov b0=r16 + br.cond.sptk b0 // call the service + ;; +END(ia64_leave_hypervisor) +// fall through +GLOBAL_ENTRY(ia64_vmm_entry) +/* + * must be at bank 0 + * parameter: + * r17:cr.isr + * r18:vpd + * r19:vpsr + * r20:__vsa_base + * r22:b0 + * r23:predicate + */ + mov r24=r22 + mov r25=r18 + tbit.nz p1,p2 = r19,IA64_PSR_IC_BIT // p1=vpsr.ic + ;; + (p1) add r29=PAL_VPS_RESUME_NORMAL,r20 + (p1) br.sptk.many ia64_vmm_entry_out + ;; + tbit.nz p1,p2 = r17,IA64_ISR_IR_BIT //p1=cr.isr.ir + ;; + (p1) add r29=PAL_VPS_RESUME_NORMAL,r20 + (p2) add r29=PAL_VPS_RESUME_HANDLER,r20 + (p2) ld8 r26=[r25] + ;; +ia64_vmm_entry_out: + mov pr=r23,-2 + mov b0=r29 + ;; + br.cond.sptk b0 // call pal service +END(ia64_vmm_entry) + + + +/* + * extern u64 ia64_call_vsa(u64 proc, u64 arg1, u64 arg2, + * u64 arg3, u64 arg4, u64 arg5, + * u64 arg6, u64 arg7); + * + * XXX: The currently defined services use only 4 args at the max. The + * rest are not consumed. + */ +GLOBAL_ENTRY(ia64_call_vsa) + .regstk 4,4,0,0 + +rpsave = loc0 +pfssave = loc1 +psrsave = loc2 +entry = loc3 +hostret = r24 + + alloc pfssave=ar.pfs,4,4,0,0 + mov rpsave=rp + adds entry=VMM_VCPU_VSA_BASE_OFFSET, r13 + ;; + ld8 entry=[entry] +1: mov hostret=ip + mov r25=in1 // copy arguments + mov r26=in2 + mov r27=in3 + mov psrsave=psr + ;; + tbit.nz p6,p0=psrsave,14 // IA64_PSR_I + tbit.nz p7,p0=psrsave,13 // IA64_PSR_IC + ;; + add hostret=2f-1b,hostret // calculate return address + add entry=entry,in0 + ;; + rsm psr.i | psr.ic + ;; + srlz.i + mov b6=entry + br.cond.sptk b6 // call the service +2: + // Architectural sequence for enabling interrupts if necessary +(p7) ssm psr.ic + ;; +(p7) srlz.i + ;; +//(p6) ssm psr.i + ;; + mov rp=rpsave + mov ar.pfs=pfssave + mov r8=r31 + ;; + srlz.d + br.ret.sptk rp + +END(ia64_call_vsa) + +#define INIT_BSPSTORE ((4<<30)-(12<<20)-0x100) + +GLOBAL_ENTRY(vmm_reset_entry) + //set up ipsr, iip, vpd.vpsr, dcr + // For IPSR: it/dt/rt=1, i/ic=1, si=1, vm/bn=1 + // For DCR: all bits 0 + adds r14=-VMM_PT_REGS_SIZE, r12 + ;; + movl r6=0x501008826000 // IPSR dt/rt/it:1;i/ic:1, si:1, vm/bn:1 + movl r10=0x8000000000000000 + adds r16=PT(CR_IIP), r14 + adds r20=PT(R1), r14 + ;; + rsm psr.ic | psr.i + ;; + srlz.i + ;; + bsw.0 + ;; + mov r21 =r13 + ;; + bsw.1 + ;; + mov ar.rsc = 0 + ;; + flushrs + ;; + mov ar.bspstore = 0 + // clear BSPSTORE + ;; + mov cr.ipsr=r6 + mov cr.ifs=r10 + ld8 r4 = [r16] // Set init iip for first run. + ld8 r1 = [r20] + ;; + mov cr.iip=r4 + ;; + adds r16=VMM_VPD_BASE_OFFSET,r13 + adds r20=VMM_VCPU_VSA_BASE_OFFSET,r13 + ;; + ld8 r18=[r16] + ld8 r20=[r20] + ;; + adds r19=VMM_VPD_VPSR_OFFSET,r18 + ;; + ld8 r19=[r19] + mov r17=r0 + mov r22=r0 + mov r23=r0 + br.cond.sptk ia64_vmm_entry + br.ret.sptk b0 +END(vmm_reset_entry) diff --git a/arch/ia64/kvm/vti.h b/arch/ia64/kvm/vti.h new file mode 100644 index 0000000000000000000000000000000000000000..f6c5617e16af85dd246e4fde9cd403898a89cdb2 --- /dev/null +++ b/arch/ia64/kvm/vti.h @@ -0,0 +1,290 @@ +/* + * vti.h: prototype for generial vt related interface + * Copyright (c) 2004, Intel Corporation. + * + * Xuefei Xu (Anthony Xu) (anthony.xu@intel.com) + * Fred Yang (fred.yang@intel.com) + * Kun Tian (Kevin Tian) (kevin.tian@intel.com) + * + * Copyright (c) 2007, Intel Corporation. + * Zhang xiantao + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + */ +#ifndef _KVM_VT_I_H +#define _KVM_VT_I_H + +#ifndef __ASSEMBLY__ +#include + +#include + +/* define itr.i and itr.d in ia64_itr function */ +#define ITR 0x01 +#define DTR 0x02 +#define IaDTR 0x03 + +#define IA64_TR_VMM 6 /*itr6, dtr6 : maps vmm code, vmbuffer*/ +#define IA64_TR_VM_DATA 7 /*dtr7 : maps current vm data*/ + +#define RR6 (6UL<<61) +#define RR7 (7UL<<61) + + +/* config_options in pal_vp_init_env */ +#define VP_INITIALIZE 1UL +#define VP_FR_PMC 1UL<<1 +#define VP_OPCODE 1UL<<8 +#define VP_CAUSE 1UL<<9 +#define VP_FW_ACC 1UL<<63 + +/* init vp env with initializing vm_buffer */ +#define VP_INIT_ENV_INITALIZE (VP_INITIALIZE | VP_FR_PMC |\ + VP_OPCODE | VP_CAUSE | VP_FW_ACC) +/* init vp env without initializing vm_buffer */ +#define VP_INIT_ENV VP_FR_PMC | VP_OPCODE | VP_CAUSE | VP_FW_ACC + +#define PAL_VP_CREATE 265 +/* Stacked Virt. Initializes a new VPD for the operation of + * a new virtual processor in the virtual environment. + */ +#define PAL_VP_ENV_INFO 266 +/*Stacked Virt. Returns the parameters needed to enter a virtual environment.*/ +#define PAL_VP_EXIT_ENV 267 +/*Stacked Virt. Allows a logical processor to exit a virtual environment.*/ +#define PAL_VP_INIT_ENV 268 +/*Stacked Virt. Allows a logical processor to enter a virtual environment.*/ +#define PAL_VP_REGISTER 269 +/*Stacked Virt. Register a different host IVT for the virtual processor.*/ +#define PAL_VP_RESUME 270 +/* Renamed from PAL_VP_RESUME */ +#define PAL_VP_RESTORE 270 +/*Stacked Virt. Resumes virtual processor operation on the logical processor.*/ +#define PAL_VP_SUSPEND 271 +/* Renamed from PAL_VP_SUSPEND */ +#define PAL_VP_SAVE 271 +/* Stacked Virt. Suspends operation for the specified virtual processor on + * the logical processor. + */ +#define PAL_VP_TERMINATE 272 +/* Stacked Virt. Terminates operation for the specified virtual processor.*/ + +union vac { + unsigned long value; + struct { + int a_int:1; + int a_from_int_cr:1; + int a_to_int_cr:1; + int a_from_psr:1; + int a_from_cpuid:1; + int a_cover:1; + int a_bsw:1; + long reserved:57; + }; +}; + +union vdc { + unsigned long value; + struct { + int d_vmsw:1; + int d_extint:1; + int d_ibr_dbr:1; + int d_pmc:1; + int d_to_pmd:1; + int d_itm:1; + long reserved:58; + }; +}; + +struct vpd { + union vac vac; + union vdc vdc; + unsigned long virt_env_vaddr; + unsigned long reserved1[29]; + unsigned long vhpi; + unsigned long reserved2[95]; + unsigned long vgr[16]; + unsigned long vbgr[16]; + unsigned long vnat; + unsigned long vbnat; + unsigned long vcpuid[5]; + unsigned long reserved3[11]; + unsigned long vpsr; + unsigned long vpr; + unsigned long reserved4[76]; + union { + unsigned long vcr[128]; + struct { + unsigned long dcr; + unsigned long itm; + unsigned long iva; + unsigned long rsv1[5]; + unsigned long pta; + unsigned long rsv2[7]; + unsigned long ipsr; + unsigned long isr; + unsigned long rsv3; + unsigned long iip; + unsigned long ifa; + unsigned long itir; + unsigned long iipa; + unsigned long ifs; + unsigned long iim; + unsigned long iha; + unsigned long rsv4[38]; + unsigned long lid; + unsigned long ivr; + unsigned long tpr; + unsigned long eoi; + unsigned long irr[4]; + unsigned long itv; + unsigned long pmv; + unsigned long cmcv; + unsigned long rsv5[5]; + unsigned long lrr0; + unsigned long lrr1; + unsigned long rsv6[46]; + }; + }; + unsigned long reserved5[128]; + unsigned long reserved6[3456]; + unsigned long vmm_avail[128]; + unsigned long reserved7[4096]; +}; + +#define PAL_PROC_VM_BIT (1UL << 40) +#define PAL_PROC_VMSW_BIT (1UL << 54) + +static inline s64 ia64_pal_vp_env_info(u64 *buffer_size, + u64 *vp_env_info) +{ + struct ia64_pal_retval iprv; + PAL_CALL_STK(iprv, PAL_VP_ENV_INFO, 0, 0, 0); + *buffer_size = iprv.v0; + *vp_env_info = iprv.v1; + return iprv.status; +} + +static inline s64 ia64_pal_vp_exit_env(u64 iva) +{ + struct ia64_pal_retval iprv; + + PAL_CALL_STK(iprv, PAL_VP_EXIT_ENV, (u64)iva, 0, 0); + return iprv.status; +} + +static inline s64 ia64_pal_vp_init_env(u64 config_options, u64 pbase_addr, + u64 vbase_addr, u64 *vsa_base) +{ + struct ia64_pal_retval iprv; + + PAL_CALL_STK(iprv, PAL_VP_INIT_ENV, config_options, pbase_addr, + vbase_addr); + *vsa_base = iprv.v0; + + return iprv.status; +} + +static inline s64 ia64_pal_vp_restore(u64 *vpd, u64 pal_proc_vector) +{ + struct ia64_pal_retval iprv; + + PAL_CALL_STK(iprv, PAL_VP_RESTORE, (u64)vpd, pal_proc_vector, 0); + + return iprv.status; +} + +static inline s64 ia64_pal_vp_save(u64 *vpd, u64 pal_proc_vector) +{ + struct ia64_pal_retval iprv; + + PAL_CALL_STK(iprv, PAL_VP_SAVE, (u64)vpd, pal_proc_vector, 0); + + return iprv.status; +} + +#endif + +/*VPD field offset*/ +#define VPD_VAC_START_OFFSET 0 +#define VPD_VDC_START_OFFSET 8 +#define VPD_VHPI_START_OFFSET 256 +#define VPD_VGR_START_OFFSET 1024 +#define VPD_VBGR_START_OFFSET 1152 +#define VPD_VNAT_START_OFFSET 1280 +#define VPD_VBNAT_START_OFFSET 1288 +#define VPD_VCPUID_START_OFFSET 1296 +#define VPD_VPSR_START_OFFSET 1424 +#define VPD_VPR_START_OFFSET 1432 +#define VPD_VRSE_CFLE_START_OFFSET 1440 +#define VPD_VCR_START_OFFSET 2048 +#define VPD_VTPR_START_OFFSET 2576 +#define VPD_VRR_START_OFFSET 3072 +#define VPD_VMM_VAIL_START_OFFSET 31744 + +/*Virtualization faults*/ + +#define EVENT_MOV_TO_AR 1 +#define EVENT_MOV_TO_AR_IMM 2 +#define EVENT_MOV_FROM_AR 3 +#define EVENT_MOV_TO_CR 4 +#define EVENT_MOV_FROM_CR 5 +#define EVENT_MOV_TO_PSR 6 +#define EVENT_MOV_FROM_PSR 7 +#define EVENT_ITC_D 8 +#define EVENT_ITC_I 9 +#define EVENT_MOV_TO_RR 10 +#define EVENT_MOV_TO_DBR 11 +#define EVENT_MOV_TO_IBR 12 +#define EVENT_MOV_TO_PKR 13 +#define EVENT_MOV_TO_PMC 14 +#define EVENT_MOV_TO_PMD 15 +#define EVENT_ITR_D 16 +#define EVENT_ITR_I 17 +#define EVENT_MOV_FROM_RR 18 +#define EVENT_MOV_FROM_DBR 19 +#define EVENT_MOV_FROM_IBR 20 +#define EVENT_MOV_FROM_PKR 21 +#define EVENT_MOV_FROM_PMC 22 +#define EVENT_MOV_FROM_CPUID 23 +#define EVENT_SSM 24 +#define EVENT_RSM 25 +#define EVENT_PTC_L 26 +#define EVENT_PTC_G 27 +#define EVENT_PTC_GA 28 +#define EVENT_PTR_D 29 +#define EVENT_PTR_I 30 +#define EVENT_THASH 31 +#define EVENT_TTAG 32 +#define EVENT_TPA 33 +#define EVENT_TAK 34 +#define EVENT_PTC_E 35 +#define EVENT_COVER 36 +#define EVENT_RFI 37 +#define EVENT_BSW_0 38 +#define EVENT_BSW_1 39 +#define EVENT_VMSW 40 + +/**PAL virtual services offsets */ +#define PAL_VPS_RESUME_NORMAL 0x0000 +#define PAL_VPS_RESUME_HANDLER 0x0400 +#define PAL_VPS_SYNC_READ 0x0800 +#define PAL_VPS_SYNC_WRITE 0x0c00 +#define PAL_VPS_SET_PENDING_INTERRUPT 0x1000 +#define PAL_VPS_THASH 0x1400 +#define PAL_VPS_TTAG 0x1800 +#define PAL_VPS_RESTORE 0x1c00 +#define PAL_VPS_SAVE 0x2000 + +#endif/* _VT_I_H*/ diff --git a/arch/ia64/kvm/vtlb.c b/arch/ia64/kvm/vtlb.c new file mode 100644 index 0000000000000000000000000000000000000000..def4576d22b1b9d3524319d5c7dffe0fa43c24ba --- /dev/null +++ b/arch/ia64/kvm/vtlb.c @@ -0,0 +1,636 @@ +/* + * vtlb.c: guest virtual tlb handling module. + * Copyright (c) 2004, Intel Corporation. + * Yaozu Dong (Eddie Dong) + * Xuefei Xu (Anthony Xu) + * + * Copyright (c) 2007, Intel Corporation. + * Xuefei Xu (Anthony Xu) + * Xiantao Zhang + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 Temple + * Place - Suite 330, Boston, MA 02111-1307 USA. + * + */ + +#include "vcpu.h" + +#include + +#include + +/* + * Check to see if the address rid:va is translated by the TLB + */ + +static int __is_tr_translated(struct thash_data *trp, u64 rid, u64 va) +{ + return ((trp->p) && (trp->rid == rid) + && ((va-trp->vadr) < PSIZE(trp->ps))); +} + +/* + * Only for GUEST TR format. + */ +static int __is_tr_overlap(struct thash_data *trp, u64 rid, u64 sva, u64 eva) +{ + u64 sa1, ea1; + + if (!trp->p || trp->rid != rid) + return 0; + + sa1 = trp->vadr; + ea1 = sa1 + PSIZE(trp->ps) - 1; + eva -= 1; + if ((sva > ea1) || (sa1 > eva)) + return 0; + else + return 1; + +} + +void machine_tlb_purge(u64 va, u64 ps) +{ + ia64_ptcl(va, ps << 2); +} + +void local_flush_tlb_all(void) +{ + int i, j; + unsigned long flags, count0, count1; + unsigned long stride0, stride1, addr; + + addr = current_vcpu->arch.ptce_base; + count0 = current_vcpu->arch.ptce_count[0]; + count1 = current_vcpu->arch.ptce_count[1]; + stride0 = current_vcpu->arch.ptce_stride[0]; + stride1 = current_vcpu->arch.ptce_stride[1]; + + local_irq_save(flags); + for (i = 0; i < count0; ++i) { + for (j = 0; j < count1; ++j) { + ia64_ptce(addr); + addr += stride1; + } + addr += stride0; + } + local_irq_restore(flags); + ia64_srlz_i(); /* srlz.i implies srlz.d */ +} + +int vhpt_enabled(struct kvm_vcpu *vcpu, u64 vadr, enum vhpt_ref ref) +{ + union ia64_rr vrr; + union ia64_pta vpta; + struct ia64_psr vpsr; + + vpsr = *(struct ia64_psr *)&VCPU(vcpu, vpsr); + vrr.val = vcpu_get_rr(vcpu, vadr); + vpta.val = vcpu_get_pta(vcpu); + + if (vrr.ve & vpta.ve) { + switch (ref) { + case DATA_REF: + case NA_REF: + return vpsr.dt; + case INST_REF: + return vpsr.dt && vpsr.it && vpsr.ic; + case RSE_REF: + return vpsr.dt && vpsr.rt; + + } + } + return 0; +} + +struct thash_data *vsa_thash(union ia64_pta vpta, u64 va, u64 vrr, u64 *tag) +{ + u64 index, pfn, rid, pfn_bits; + + pfn_bits = vpta.size - 5 - 8; + pfn = REGION_OFFSET(va) >> _REGION_PAGE_SIZE(vrr); + rid = _REGION_ID(vrr); + index = ((rid & 0xff) << pfn_bits)|(pfn & ((1UL << pfn_bits) - 1)); + *tag = ((rid >> 8) & 0xffff) | ((pfn >> pfn_bits) << 16); + + return (struct thash_data *)((vpta.base << PTA_BASE_SHIFT) + + (index << 5)); +} + +struct thash_data *__vtr_lookup(struct kvm_vcpu *vcpu, u64 va, int type) +{ + + struct thash_data *trp; + int i; + u64 rid; + + rid = vcpu_get_rr(vcpu, va); + rid = rid & RR_RID_MASK;; + if (type == D_TLB) { + if (vcpu_quick_region_check(vcpu->arch.dtr_regions, va)) { + for (trp = (struct thash_data *)&vcpu->arch.dtrs, i = 0; + i < NDTRS; i++, trp++) { + if (__is_tr_translated(trp, rid, va)) + return trp; + } + } + } else { + if (vcpu_quick_region_check(vcpu->arch.itr_regions, va)) { + for (trp = (struct thash_data *)&vcpu->arch.itrs, i = 0; + i < NITRS; i++, trp++) { + if (__is_tr_translated(trp, rid, va)) + return trp; + } + } + } + + return NULL; +} + +static void vhpt_insert(u64 pte, u64 itir, u64 ifa, u64 gpte) +{ + union ia64_rr rr; + struct thash_data *head; + unsigned long ps, gpaddr; + + ps = itir_ps(itir); + + gpaddr = ((gpte & _PAGE_PPN_MASK) >> ps << ps) | + (ifa & ((1UL << ps) - 1)); + + rr.val = ia64_get_rr(ifa); + head = (struct thash_data *)ia64_thash(ifa); + head->etag = INVALID_TI_TAG; + ia64_mf(); + head->page_flags = pte & ~PAGE_FLAGS_RV_MASK; + head->itir = rr.ps << 2; + head->etag = ia64_ttag(ifa); + head->gpaddr = gpaddr; +} + +void mark_pages_dirty(struct kvm_vcpu *v, u64 pte, u64 ps) +{ + u64 i, dirty_pages = 1; + u64 base_gfn = (pte&_PAGE_PPN_MASK) >> PAGE_SHIFT; + spinlock_t *lock = __kvm_va(v->arch.dirty_log_lock_pa); + void *dirty_bitmap = (void *)v - (KVM_VCPU_OFS + v->vcpu_id * VCPU_SIZE) + + KVM_MEM_DIRTY_LOG_OFS; + dirty_pages <<= ps <= PAGE_SHIFT ? 0 : ps - PAGE_SHIFT; + + vmm_spin_lock(lock); + for (i = 0; i < dirty_pages; i++) { + /* avoid RMW */ + if (!test_bit(base_gfn + i, dirty_bitmap)) + set_bit(base_gfn + i , dirty_bitmap); + } + vmm_spin_unlock(lock); +} + +void thash_vhpt_insert(struct kvm_vcpu *v, u64 pte, u64 itir, u64 va, int type) +{ + u64 phy_pte, psr; + union ia64_rr mrr; + + mrr.val = ia64_get_rr(va); + phy_pte = translate_phy_pte(&pte, itir, va); + + if (itir_ps(itir) >= mrr.ps) { + vhpt_insert(phy_pte, itir, va, pte); + } else { + phy_pte &= ~PAGE_FLAGS_RV_MASK; + psr = ia64_clear_ic(); + ia64_itc(type, va, phy_pte, itir_ps(itir)); + ia64_set_psr(psr); + } + + if (!(pte&VTLB_PTE_IO)) + mark_pages_dirty(v, pte, itir_ps(itir)); +} + +/* + * vhpt lookup + */ +struct thash_data *vhpt_lookup(u64 va) +{ + struct thash_data *head; + u64 tag; + + head = (struct thash_data *)ia64_thash(va); + tag = ia64_ttag(va); + if (head->etag == tag) + return head; + return NULL; +} + +u64 guest_vhpt_lookup(u64 iha, u64 *pte) +{ + u64 ret; + struct thash_data *data; + + data = __vtr_lookup(current_vcpu, iha, D_TLB); + if (data != NULL) + thash_vhpt_insert(current_vcpu, data->page_flags, + data->itir, iha, D_TLB); + + asm volatile ("rsm psr.ic|psr.i;;" + "srlz.d;;" + "ld8.s r9=[%1];;" + "tnat.nz p6,p7=r9;;" + "(p6) mov %0=1;" + "(p6) mov r9=r0;" + "(p7) extr.u r9=r9,0,53;;" + "(p7) mov %0=r0;" + "(p7) st8 [%2]=r9;;" + "ssm psr.ic;;" + "srlz.d;;" + /* "ssm psr.i;;" Once interrupts in vmm open, need fix*/ + : "=r"(ret) : "r"(iha), "r"(pte):"memory"); + + return ret; +} + +/* + * purge software guest tlb + */ + +static void vtlb_purge(struct kvm_vcpu *v, u64 va, u64 ps) +{ + struct thash_data *cur; + u64 start, curadr, size, psbits, tag, rr_ps, num; + union ia64_rr vrr; + struct thash_cb *hcb = &v->arch.vtlb; + + vrr.val = vcpu_get_rr(v, va); + psbits = VMX(v, psbits[(va >> 61)]); + start = va & ~((1UL << ps) - 1); + while (psbits) { + curadr = start; + rr_ps = __ffs(psbits); + psbits &= ~(1UL << rr_ps); + num = 1UL << ((ps < rr_ps) ? 0 : (ps - rr_ps)); + size = PSIZE(rr_ps); + vrr.ps = rr_ps; + while (num) { + cur = vsa_thash(hcb->pta, curadr, vrr.val, &tag); + if (cur->etag == tag && cur->ps == rr_ps) + cur->etag = INVALID_TI_TAG; + curadr += size; + num--; + } + } +} + + +/* + * purge VHPT and machine TLB + */ +static void vhpt_purge(struct kvm_vcpu *v, u64 va, u64 ps) +{ + struct thash_data *cur; + u64 start, size, tag, num; + union ia64_rr rr; + + start = va & ~((1UL << ps) - 1); + rr.val = ia64_get_rr(va); + size = PSIZE(rr.ps); + num = 1UL << ((ps < rr.ps) ? 0 : (ps - rr.ps)); + while (num) { + cur = (struct thash_data *)ia64_thash(start); + tag = ia64_ttag(start); + if (cur->etag == tag) + cur->etag = INVALID_TI_TAG; + start += size; + num--; + } + machine_tlb_purge(va, ps); +} + +/* + * Insert an entry into hash TLB or VHPT. + * NOTES: + * 1: When inserting VHPT to thash, "va" is a must covered + * address by the inserted machine VHPT entry. + * 2: The format of entry is always in TLB. + * 3: The caller need to make sure the new entry will not overlap + * with any existed entry. + */ +void vtlb_insert(struct kvm_vcpu *v, u64 pte, u64 itir, u64 va) +{ + struct thash_data *head; + union ia64_rr vrr; + u64 tag; + struct thash_cb *hcb = &v->arch.vtlb; + + vrr.val = vcpu_get_rr(v, va); + vrr.ps = itir_ps(itir); + VMX(v, psbits[va >> 61]) |= (1UL << vrr.ps); + head = vsa_thash(hcb->pta, va, vrr.val, &tag); + head->page_flags = pte; + head->itir = itir; + head->etag = tag; +} + +int vtr_find_overlap(struct kvm_vcpu *vcpu, u64 va, u64 ps, int type) +{ + struct thash_data *trp; + int i; + u64 end, rid; + + rid = vcpu_get_rr(vcpu, va); + rid = rid & RR_RID_MASK; + end = va + PSIZE(ps); + if (type == D_TLB) { + if (vcpu_quick_region_check(vcpu->arch.dtr_regions, va)) { + for (trp = (struct thash_data *)&vcpu->arch.dtrs, i = 0; + i < NDTRS; i++, trp++) { + if (__is_tr_overlap(trp, rid, va, end)) + return i; + } + } + } else { + if (vcpu_quick_region_check(vcpu->arch.itr_regions, va)) { + for (trp = (struct thash_data *)&vcpu->arch.itrs, i = 0; + i < NITRS; i++, trp++) { + if (__is_tr_overlap(trp, rid, va, end)) + return i; + } + } + } + return -1; +} + +/* + * Purge entries in VTLB and VHPT + */ +void thash_purge_entries(struct kvm_vcpu *v, u64 va, u64 ps) +{ + if (vcpu_quick_region_check(v->arch.tc_regions, va)) + vtlb_purge(v, va, ps); + vhpt_purge(v, va, ps); +} + +void thash_purge_entries_remote(struct kvm_vcpu *v, u64 va, u64 ps) +{ + u64 old_va = va; + va = REGION_OFFSET(va); + if (vcpu_quick_region_check(v->arch.tc_regions, old_va)) + vtlb_purge(v, va, ps); + vhpt_purge(v, va, ps); +} + +u64 translate_phy_pte(u64 *pte, u64 itir, u64 va) +{ + u64 ps, ps_mask, paddr, maddr; + union pte_flags phy_pte; + + ps = itir_ps(itir); + ps_mask = ~((1UL << ps) - 1); + phy_pte.val = *pte; + paddr = *pte; + paddr = ((paddr & _PAGE_PPN_MASK) & ps_mask) | (va & ~ps_mask); + maddr = kvm_lookup_mpa(paddr >> PAGE_SHIFT); + if (maddr & GPFN_IO_MASK) { + *pte |= VTLB_PTE_IO; + return -1; + } + maddr = ((maddr & _PAGE_PPN_MASK) & PAGE_MASK) | + (paddr & ~PAGE_MASK); + phy_pte.ppn = maddr >> ARCH_PAGE_SHIFT; + return phy_pte.val; +} + +/* + * Purge overlap TCs and then insert the new entry to emulate itc ops. + * Notes: Only TC entry can purge and insert. + * 1 indicates this is MMIO + */ +int thash_purge_and_insert(struct kvm_vcpu *v, u64 pte, u64 itir, + u64 ifa, int type) +{ + u64 ps; + u64 phy_pte; + union ia64_rr vrr, mrr; + int ret = 0; + + ps = itir_ps(itir); + vrr.val = vcpu_get_rr(v, ifa); + mrr.val = ia64_get_rr(ifa); + + phy_pte = translate_phy_pte(&pte, itir, ifa); + + /* Ensure WB attribute if pte is related to a normal mem page, + * which is required by vga acceleration since qemu maps shared + * vram buffer with WB. + */ + if (!(pte & VTLB_PTE_IO) && ((pte & _PAGE_MA_MASK) != _PAGE_MA_NAT)) { + pte &= ~_PAGE_MA_MASK; + phy_pte &= ~_PAGE_MA_MASK; + } + + if (pte & VTLB_PTE_IO) + ret = 1; + + vtlb_purge(v, ifa, ps); + vhpt_purge(v, ifa, ps); + + if (ps == mrr.ps) { + if (!(pte&VTLB_PTE_IO)) { + vhpt_insert(phy_pte, itir, ifa, pte); + } else { + vtlb_insert(v, pte, itir, ifa); + vcpu_quick_region_set(VMX(v, tc_regions), ifa); + } + } else if (ps > mrr.ps) { + vtlb_insert(v, pte, itir, ifa); + vcpu_quick_region_set(VMX(v, tc_regions), ifa); + if (!(pte&VTLB_PTE_IO)) + vhpt_insert(phy_pte, itir, ifa, pte); + } else { + u64 psr; + phy_pte &= ~PAGE_FLAGS_RV_MASK; + psr = ia64_clear_ic(); + ia64_itc(type, ifa, phy_pte, ps); + ia64_set_psr(psr); + } + if (!(pte&VTLB_PTE_IO)) + mark_pages_dirty(v, pte, ps); + + return ret; +} + +/* + * Purge all TCs or VHPT entries including those in Hash table. + * + */ + +void thash_purge_all(struct kvm_vcpu *v) +{ + int i; + struct thash_data *head; + struct thash_cb *vtlb, *vhpt; + vtlb = &v->arch.vtlb; + vhpt = &v->arch.vhpt; + + for (i = 0; i < 8; i++) + VMX(v, psbits[i]) = 0; + + head = vtlb->hash; + for (i = 0; i < vtlb->num; i++) { + head->page_flags = 0; + head->etag = INVALID_TI_TAG; + head->itir = 0; + head->next = 0; + head++; + }; + + head = vhpt->hash; + for (i = 0; i < vhpt->num; i++) { + head->page_flags = 0; + head->etag = INVALID_TI_TAG; + head->itir = 0; + head->next = 0; + head++; + }; + + local_flush_tlb_all(); +} + + +/* + * Lookup the hash table and its collision chain to find an entry + * covering this address rid:va or the entry. + * + * INPUT: + * in: TLB format for both VHPT & TLB. + */ + +struct thash_data *vtlb_lookup(struct kvm_vcpu *v, u64 va, int is_data) +{ + struct thash_data *cch; + u64 psbits, ps, tag; + union ia64_rr vrr; + + struct thash_cb *hcb = &v->arch.vtlb; + + cch = __vtr_lookup(v, va, is_data);; + if (cch) + return cch; + + if (vcpu_quick_region_check(v->arch.tc_regions, va) == 0) + return NULL; + + psbits = VMX(v, psbits[(va >> 61)]); + vrr.val = vcpu_get_rr(v, va); + while (psbits) { + ps = __ffs(psbits); + psbits &= ~(1UL << ps); + vrr.ps = ps; + cch = vsa_thash(hcb->pta, va, vrr.val, &tag); + if (cch->etag == tag && cch->ps == ps) + return cch; + } + + return NULL; +} + + +/* + * Initialize internal control data before service. + */ +void thash_init(struct thash_cb *hcb, u64 sz) +{ + int i; + struct thash_data *head; + + hcb->pta.val = (unsigned long)hcb->hash; + hcb->pta.vf = 1; + hcb->pta.ve = 1; + hcb->pta.size = sz; + head = hcb->hash; + for (i = 0; i < hcb->num; i++) { + head->page_flags = 0; + head->itir = 0; + head->etag = INVALID_TI_TAG; + head->next = 0; + head++; + } +} + +u64 kvm_lookup_mpa(u64 gpfn) +{ + u64 *base = (u64 *) KVM_P2M_BASE; + return *(base + gpfn); +} + +u64 kvm_gpa_to_mpa(u64 gpa) +{ + u64 pte = kvm_lookup_mpa(gpa >> PAGE_SHIFT); + return (pte >> PAGE_SHIFT << PAGE_SHIFT) | (gpa & ~PAGE_MASK); +} + + +/* + * Fetch guest bundle code. + * INPUT: + * gip: guest ip + * pbundle: used to return fetched bundle. + */ +int fetch_code(struct kvm_vcpu *vcpu, u64 gip, IA64_BUNDLE *pbundle) +{ + u64 gpip = 0; /* guest physical IP*/ + u64 *vpa; + struct thash_data *tlb; + u64 maddr; + + if (!(VCPU(vcpu, vpsr) & IA64_PSR_IT)) { + /* I-side physical mode */ + gpip = gip; + } else { + tlb = vtlb_lookup(vcpu, gip, I_TLB); + if (tlb) + gpip = (tlb->ppn >> (tlb->ps - 12) << tlb->ps) | + (gip & (PSIZE(tlb->ps) - 1)); + } + if (gpip) { + maddr = kvm_gpa_to_mpa(gpip); + } else { + tlb = vhpt_lookup(gip); + if (tlb == NULL) { + ia64_ptcl(gip, ARCH_PAGE_SHIFT << 2); + return IA64_FAULT; + } + maddr = (tlb->ppn >> (tlb->ps - 12) << tlb->ps) + | (gip & (PSIZE(tlb->ps) - 1)); + } + vpa = (u64 *)__kvm_va(maddr); + + pbundle->i64[0] = *vpa++; + pbundle->i64[1] = *vpa; + + return IA64_NO_FAULT; +} + + +void kvm_init_vhpt(struct kvm_vcpu *v) +{ + v->arch.vhpt.num = VHPT_NUM_ENTRIES; + thash_init(&v->arch.vhpt, VHPT_SHIFT); + ia64_set_pta(v->arch.vhpt.pta.val); + /*Enable VHPT here?*/ +} + +void kvm_init_vtlb(struct kvm_vcpu *v) +{ + v->arch.vtlb.num = VTLB_NUM_ENTRIES; + thash_init(&v->arch.vtlb, VTLB_SHIFT); +} diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index 20f45a8b87e3e56c7e12615610f3706497e26003..4e40c122bf26568db0fdfe36c3f55d89f526f9e1 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -803,3 +803,4 @@ config PPC_CLOCK config PPC_LIB_RHEAP bool +source "arch/powerpc/kvm/Kconfig" diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug index a86d8d853214a37c6e807c3e5515c7cf21ef36c6..807a2dce626381997dcd65a6ac0cd9c0931a91d6 100644 --- a/arch/powerpc/Kconfig.debug +++ b/arch/powerpc/Kconfig.debug @@ -151,6 +151,9 @@ config BOOTX_TEXT config PPC_EARLY_DEBUG bool "Early debugging (dangerous)" + # PPC_EARLY_DEBUG on 440 leaves AS=1 mappings above the TLB high water + # mark, which doesn't work with current 440 KVM. + depends on !KVM help Say Y to enable some early debugging facilities that may be available for your processor/board combination. Those facilities are hacks diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile index e2ec4a91ccefee1b68b22a4b2c834ebc578983f7..9dcdc036cdf7ff13a6e09ad02d61b01c888e7d6f 100644 --- a/arch/powerpc/Makefile +++ b/arch/powerpc/Makefile @@ -145,6 +145,7 @@ core-y += arch/powerpc/kernel/ \ arch/powerpc/platforms/ core-$(CONFIG_MATH_EMULATION) += arch/powerpc/math-emu/ core-$(CONFIG_XMON) += arch/powerpc/xmon/ +core-$(CONFIG_KVM) += arch/powerpc/kvm/ drivers-$(CONFIG_OPROFILE) += arch/powerpc/oprofile/ diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c index adf1d09d726f9102199ca87ec0aeb01828349128..62134845af081e529f88ea80820ea3d286a2f0f5 100644 --- a/arch/powerpc/kernel/asm-offsets.c +++ b/arch/powerpc/kernel/asm-offsets.c @@ -23,6 +23,9 @@ #include #include #include +#ifdef CONFIG_KVM +#include +#endif #ifdef CONFIG_PPC64 #include #include @@ -324,5 +327,30 @@ int main(void) DEFINE(PGD_TABLE_SIZE, PGD_TABLE_SIZE); +#ifdef CONFIG_KVM + DEFINE(TLBE_BYTES, sizeof(struct tlbe)); + + DEFINE(VCPU_HOST_STACK, offsetof(struct kvm_vcpu, arch.host_stack)); + DEFINE(VCPU_HOST_PID, offsetof(struct kvm_vcpu, arch.host_pid)); + DEFINE(VCPU_HOST_TLB, offsetof(struct kvm_vcpu, arch.host_tlb)); + DEFINE(VCPU_SHADOW_TLB, offsetof(struct kvm_vcpu, arch.shadow_tlb)); + DEFINE(VCPU_GPRS, offsetof(struct kvm_vcpu, arch.gpr)); + DEFINE(VCPU_LR, offsetof(struct kvm_vcpu, arch.lr)); + DEFINE(VCPU_CR, offsetof(struct kvm_vcpu, arch.cr)); + DEFINE(VCPU_XER, offsetof(struct kvm_vcpu, arch.xer)); + DEFINE(VCPU_CTR, offsetof(struct kvm_vcpu, arch.ctr)); + DEFINE(VCPU_PC, offsetof(struct kvm_vcpu, arch.pc)); + DEFINE(VCPU_MSR, offsetof(struct kvm_vcpu, arch.msr)); + DEFINE(VCPU_SPRG4, offsetof(struct kvm_vcpu, arch.sprg4)); + DEFINE(VCPU_SPRG5, offsetof(struct kvm_vcpu, arch.sprg5)); + DEFINE(VCPU_SPRG6, offsetof(struct kvm_vcpu, arch.sprg6)); + DEFINE(VCPU_SPRG7, offsetof(struct kvm_vcpu, arch.sprg7)); + DEFINE(VCPU_PID, offsetof(struct kvm_vcpu, arch.pid)); + + DEFINE(VCPU_LAST_INST, offsetof(struct kvm_vcpu, arch.last_inst)); + DEFINE(VCPU_FAULT_DEAR, offsetof(struct kvm_vcpu, arch.fault_dear)); + DEFINE(VCPU_FAULT_ESR, offsetof(struct kvm_vcpu, arch.fault_esr)); +#endif + return 0; } diff --git a/arch/powerpc/kvm/44x_tlb.c b/arch/powerpc/kvm/44x_tlb.c new file mode 100644 index 0000000000000000000000000000000000000000..f5d7a5eab96e7e31405ce77df9b30a10302ec2d0 --- /dev/null +++ b/arch/powerpc/kvm/44x_tlb.c @@ -0,0 +1,224 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License, version 2, as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2007 + * + * Authors: Hollis Blanchard + */ + +#include +#include +#include +#include +#include +#include + +#include "44x_tlb.h" + +#define PPC44x_TLB_USER_PERM_MASK (PPC44x_TLB_UX|PPC44x_TLB_UR|PPC44x_TLB_UW) +#define PPC44x_TLB_SUPER_PERM_MASK (PPC44x_TLB_SX|PPC44x_TLB_SR|PPC44x_TLB_SW) + +static unsigned int kvmppc_tlb_44x_pos; + +static u32 kvmppc_44x_tlb_shadow_attrib(u32 attrib, int usermode) +{ + /* Mask off reserved bits. */ + attrib &= PPC44x_TLB_PERM_MASK|PPC44x_TLB_ATTR_MASK; + + if (!usermode) { + /* Guest is in supervisor mode, so we need to translate guest + * supervisor permissions into user permissions. */ + attrib &= ~PPC44x_TLB_USER_PERM_MASK; + attrib |= (attrib & PPC44x_TLB_SUPER_PERM_MASK) << 3; + } + + /* Make sure host can always access this memory. */ + attrib |= PPC44x_TLB_SX|PPC44x_TLB_SR|PPC44x_TLB_SW; + + return attrib; +} + +/* Search the guest TLB for a matching entry. */ +int kvmppc_44x_tlb_index(struct kvm_vcpu *vcpu, gva_t eaddr, unsigned int pid, + unsigned int as) +{ + int i; + + /* XXX Replace loop with fancy data structures. */ + for (i = 0; i < PPC44x_TLB_SIZE; i++) { + struct tlbe *tlbe = &vcpu->arch.guest_tlb[i]; + unsigned int tid; + + if (eaddr < get_tlb_eaddr(tlbe)) + continue; + + if (eaddr > get_tlb_end(tlbe)) + continue; + + tid = get_tlb_tid(tlbe); + if (tid && (tid != pid)) + continue; + + if (!get_tlb_v(tlbe)) + continue; + + if (get_tlb_ts(tlbe) != as) + continue; + + return i; + } + + return -1; +} + +struct tlbe *kvmppc_44x_itlb_search(struct kvm_vcpu *vcpu, gva_t eaddr) +{ + unsigned int as = !!(vcpu->arch.msr & MSR_IS); + unsigned int index; + + index = kvmppc_44x_tlb_index(vcpu, eaddr, vcpu->arch.pid, as); + if (index == -1) + return NULL; + return &vcpu->arch.guest_tlb[index]; +} + +struct tlbe *kvmppc_44x_dtlb_search(struct kvm_vcpu *vcpu, gva_t eaddr) +{ + unsigned int as = !!(vcpu->arch.msr & MSR_DS); + unsigned int index; + + index = kvmppc_44x_tlb_index(vcpu, eaddr, vcpu->arch.pid, as); + if (index == -1) + return NULL; + return &vcpu->arch.guest_tlb[index]; +} + +static int kvmppc_44x_tlbe_is_writable(struct tlbe *tlbe) +{ + return tlbe->word2 & (PPC44x_TLB_SW|PPC44x_TLB_UW); +} + +/* Must be called with mmap_sem locked for writing. */ +static void kvmppc_44x_shadow_release(struct kvm_vcpu *vcpu, + unsigned int index) +{ + struct tlbe *stlbe = &vcpu->arch.shadow_tlb[index]; + struct page *page = vcpu->arch.shadow_pages[index]; + + kunmap(vcpu->arch.shadow_pages[index]); + + if (get_tlb_v(stlbe)) { + if (kvmppc_44x_tlbe_is_writable(stlbe)) + kvm_release_page_dirty(page); + else + kvm_release_page_clean(page); + } +} + +/* Caller must ensure that the specified guest TLB entry is safe to insert into + * the shadow TLB. */ +void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid, + u32 flags) +{ + struct page *new_page; + struct tlbe *stlbe; + hpa_t hpaddr; + unsigned int victim; + + /* Future optimization: don't overwrite the TLB entry containing the + * current PC (or stack?). */ + victim = kvmppc_tlb_44x_pos++; + if (kvmppc_tlb_44x_pos > tlb_44x_hwater) + kvmppc_tlb_44x_pos = 0; + stlbe = &vcpu->arch.shadow_tlb[victim]; + + /* Get reference to new page. */ + down_write(¤t->mm->mmap_sem); + new_page = gfn_to_page(vcpu->kvm, gfn); + if (is_error_page(new_page)) { + printk(KERN_ERR "Couldn't get guest page!\n"); + kvm_release_page_clean(new_page); + return; + } + hpaddr = page_to_phys(new_page); + + /* Drop reference to old page. */ + kvmppc_44x_shadow_release(vcpu, victim); + up_write(¤t->mm->mmap_sem); + + vcpu->arch.shadow_pages[victim] = new_page; + + /* XXX Make sure (va, size) doesn't overlap any other + * entries. 440x6 user manual says the result would be + * "undefined." */ + + /* XXX what about AS? */ + + stlbe->tid = asid & 0xff; + + /* Force TS=1 for all guest mappings. */ + /* For now we hardcode 4KB mappings, but it will be important to + * use host large pages in the future. */ + stlbe->word0 = (gvaddr & PAGE_MASK) | PPC44x_TLB_VALID | PPC44x_TLB_TS + | PPC44x_TLB_4K; + + stlbe->word1 = (hpaddr & 0xfffffc00) | ((hpaddr >> 32) & 0xf); + stlbe->word2 = kvmppc_44x_tlb_shadow_attrib(flags, + vcpu->arch.msr & MSR_PR); +} + +void kvmppc_mmu_invalidate(struct kvm_vcpu *vcpu, u64 eaddr, u64 asid) +{ + unsigned int pid = asid & 0xff; + int i; + + /* XXX Replace loop with fancy data structures. */ + down_write(¤t->mm->mmap_sem); + for (i = 0; i <= tlb_44x_hwater; i++) { + struct tlbe *stlbe = &vcpu->arch.shadow_tlb[i]; + unsigned int tid; + + if (!get_tlb_v(stlbe)) + continue; + + if (eaddr < get_tlb_eaddr(stlbe)) + continue; + + if (eaddr > get_tlb_end(stlbe)) + continue; + + tid = get_tlb_tid(stlbe); + if (tid && (tid != pid)) + continue; + + kvmppc_44x_shadow_release(vcpu, i); + stlbe->word0 = 0; + } + up_write(¤t->mm->mmap_sem); +} + +/* Invalidate all mappings, so that when they fault back in they will get the + * proper permission bits. */ +void kvmppc_mmu_priv_switch(struct kvm_vcpu *vcpu, int usermode) +{ + int i; + + /* XXX Replace loop with fancy data structures. */ + down_write(¤t->mm->mmap_sem); + for (i = 0; i <= tlb_44x_hwater; i++) { + kvmppc_44x_shadow_release(vcpu, i); + vcpu->arch.shadow_tlb[i].word0 = 0; + } + up_write(¤t->mm->mmap_sem); +} diff --git a/arch/powerpc/kvm/44x_tlb.h b/arch/powerpc/kvm/44x_tlb.h new file mode 100644 index 0000000000000000000000000000000000000000..2ccd46b6f6b762474a63a89fe2e4a8ec63e62b8e --- /dev/null +++ b/arch/powerpc/kvm/44x_tlb.h @@ -0,0 +1,91 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License, version 2, as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2007 + * + * Authors: Hollis Blanchard + */ + +#ifndef __KVM_POWERPC_TLB_H__ +#define __KVM_POWERPC_TLB_H__ + +#include +#include + +extern int kvmppc_44x_tlb_index(struct kvm_vcpu *vcpu, gva_t eaddr, + unsigned int pid, unsigned int as); +extern struct tlbe *kvmppc_44x_dtlb_search(struct kvm_vcpu *vcpu, gva_t eaddr); +extern struct tlbe *kvmppc_44x_itlb_search(struct kvm_vcpu *vcpu, gva_t eaddr); + +/* TLB helper functions */ +static inline unsigned int get_tlb_size(const struct tlbe *tlbe) +{ + return (tlbe->word0 >> 4) & 0xf; +} + +static inline gva_t get_tlb_eaddr(const struct tlbe *tlbe) +{ + return tlbe->word0 & 0xfffffc00; +} + +static inline gva_t get_tlb_bytes(const struct tlbe *tlbe) +{ + unsigned int pgsize = get_tlb_size(tlbe); + return 1 << 10 << (pgsize << 1); +} + +static inline gva_t get_tlb_end(const struct tlbe *tlbe) +{ + return get_tlb_eaddr(tlbe) + get_tlb_bytes(tlbe) - 1; +} + +static inline u64 get_tlb_raddr(const struct tlbe *tlbe) +{ + u64 word1 = tlbe->word1; + return ((word1 & 0xf) << 32) | (word1 & 0xfffffc00); +} + +static inline unsigned int get_tlb_tid(const struct tlbe *tlbe) +{ + return tlbe->tid & 0xff; +} + +static inline unsigned int get_tlb_ts(const struct tlbe *tlbe) +{ + return (tlbe->word0 >> 8) & 0x1; +} + +static inline unsigned int get_tlb_v(const struct tlbe *tlbe) +{ + return (tlbe->word0 >> 9) & 0x1; +} + +static inline unsigned int get_mmucr_stid(const struct kvm_vcpu *vcpu) +{ + return vcpu->arch.mmucr & 0xff; +} + +static inline unsigned int get_mmucr_sts(const struct kvm_vcpu *vcpu) +{ + return (vcpu->arch.mmucr >> 16) & 0x1; +} + +static inline gpa_t tlb_xlate(struct tlbe *tlbe, gva_t eaddr) +{ + unsigned int pgmask = get_tlb_bytes(tlbe) - 1; + + return get_tlb_raddr(tlbe) | (eaddr & pgmask); +} + +#endif /* __KVM_POWERPC_TLB_H__ */ diff --git a/arch/powerpc/kvm/Kconfig b/arch/powerpc/kvm/Kconfig new file mode 100644 index 0000000000000000000000000000000000000000..6b076010213ba28248a9c81813ca0920bef43e08 --- /dev/null +++ b/arch/powerpc/kvm/Kconfig @@ -0,0 +1,42 @@ +# +# KVM configuration +# + +menuconfig VIRTUALIZATION + bool "Virtualization" + ---help--- + Say Y here to get to see options for using your Linux host to run + other operating systems inside virtual machines (guests). + This option alone does not add any kernel code. + + If you say N, all options in this submenu will be skipped and + disabled. + +if VIRTUALIZATION + +config KVM + bool "Kernel-based Virtual Machine (KVM) support" + depends on 44x && EXPERIMENTAL + select PREEMPT_NOTIFIERS + select ANON_INODES + # We can only run on Book E hosts so far + select KVM_BOOKE_HOST + ---help--- + Support hosting virtualized guest machines. You will also + need to select one or more of the processor modules below. + + This module provides access to the hardware capabilities through + a character device node named /dev/kvm. + + If unsure, say N. + +config KVM_BOOKE_HOST + bool "KVM host support for Book E PowerPC processors" + depends on KVM && 44x + ---help--- + Provides host support for KVM on Book E PowerPC processors. Currently + this works on 440 processors only. + +source drivers/virtio/Kconfig + +endif # VIRTUALIZATION diff --git a/arch/powerpc/kvm/Makefile b/arch/powerpc/kvm/Makefile new file mode 100644 index 0000000000000000000000000000000000000000..d0d358d367ecaa359ffed9a351971a5748620fa0 --- /dev/null +++ b/arch/powerpc/kvm/Makefile @@ -0,0 +1,15 @@ +# +# Makefile for Kernel-based Virtual Machine module +# + +EXTRA_CFLAGS += -Ivirt/kvm -Iarch/powerpc/kvm + +common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o) + +kvm-objs := $(common-objs) powerpc.o emulate.o booke_guest.o +obj-$(CONFIG_KVM) += kvm.o + +AFLAGS_booke_interrupts.o := -I$(obj) + +kvm-booke-host-objs := booke_host.o booke_interrupts.o 44x_tlb.o +obj-$(CONFIG_KVM_BOOKE_HOST) += kvm-booke-host.o diff --git a/arch/powerpc/kvm/booke_guest.c b/arch/powerpc/kvm/booke_guest.c new file mode 100644 index 0000000000000000000000000000000000000000..6d9884a6884aa2379f68992738ec40baf141e0bc --- /dev/null +++ b/arch/powerpc/kvm/booke_guest.c @@ -0,0 +1,615 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License, version 2, as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2007 + * + * Authors: Hollis Blanchard + * Christian Ehrhardt + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "44x_tlb.h" + +#define VM_STAT(x) offsetof(struct kvm, stat.x), KVM_STAT_VM +#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU + +struct kvm_stats_debugfs_item debugfs_entries[] = { + { "exits", VCPU_STAT(sum_exits) }, + { "mmio", VCPU_STAT(mmio_exits) }, + { "dcr", VCPU_STAT(dcr_exits) }, + { "sig", VCPU_STAT(signal_exits) }, + { "light", VCPU_STAT(light_exits) }, + { "itlb_r", VCPU_STAT(itlb_real_miss_exits) }, + { "itlb_v", VCPU_STAT(itlb_virt_miss_exits) }, + { "dtlb_r", VCPU_STAT(dtlb_real_miss_exits) }, + { "dtlb_v", VCPU_STAT(dtlb_virt_miss_exits) }, + { "sysc", VCPU_STAT(syscall_exits) }, + { "isi", VCPU_STAT(isi_exits) }, + { "dsi", VCPU_STAT(dsi_exits) }, + { "inst_emu", VCPU_STAT(emulated_inst_exits) }, + { "dec", VCPU_STAT(dec_exits) }, + { "ext_intr", VCPU_STAT(ext_intr_exits) }, + { NULL } +}; + +static const u32 interrupt_msr_mask[16] = { + [BOOKE_INTERRUPT_CRITICAL] = MSR_ME, + [BOOKE_INTERRUPT_MACHINE_CHECK] = 0, + [BOOKE_INTERRUPT_DATA_STORAGE] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_INST_STORAGE] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_EXTERNAL] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_ALIGNMENT] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_PROGRAM] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_FP_UNAVAIL] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_SYSCALL] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_AP_UNAVAIL] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_DECREMENTER] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_FIT] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_WATCHDOG] = MSR_ME, + [BOOKE_INTERRUPT_DTLB_MISS] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_ITLB_MISS] = MSR_CE|MSR_ME|MSR_DE, + [BOOKE_INTERRUPT_DEBUG] = MSR_ME, +}; + +const unsigned char exception_priority[] = { + [BOOKE_INTERRUPT_DATA_STORAGE] = 0, + [BOOKE_INTERRUPT_INST_STORAGE] = 1, + [BOOKE_INTERRUPT_ALIGNMENT] = 2, + [BOOKE_INTERRUPT_PROGRAM] = 3, + [BOOKE_INTERRUPT_FP_UNAVAIL] = 4, + [BOOKE_INTERRUPT_SYSCALL] = 5, + [BOOKE_INTERRUPT_AP_UNAVAIL] = 6, + [BOOKE_INTERRUPT_DTLB_MISS] = 7, + [BOOKE_INTERRUPT_ITLB_MISS] = 8, + [BOOKE_INTERRUPT_MACHINE_CHECK] = 9, + [BOOKE_INTERRUPT_DEBUG] = 10, + [BOOKE_INTERRUPT_CRITICAL] = 11, + [BOOKE_INTERRUPT_WATCHDOG] = 12, + [BOOKE_INTERRUPT_EXTERNAL] = 13, + [BOOKE_INTERRUPT_FIT] = 14, + [BOOKE_INTERRUPT_DECREMENTER] = 15, +}; + +const unsigned char priority_exception[] = { + BOOKE_INTERRUPT_DATA_STORAGE, + BOOKE_INTERRUPT_INST_STORAGE, + BOOKE_INTERRUPT_ALIGNMENT, + BOOKE_INTERRUPT_PROGRAM, + BOOKE_INTERRUPT_FP_UNAVAIL, + BOOKE_INTERRUPT_SYSCALL, + BOOKE_INTERRUPT_AP_UNAVAIL, + BOOKE_INTERRUPT_DTLB_MISS, + BOOKE_INTERRUPT_ITLB_MISS, + BOOKE_INTERRUPT_MACHINE_CHECK, + BOOKE_INTERRUPT_DEBUG, + BOOKE_INTERRUPT_CRITICAL, + BOOKE_INTERRUPT_WATCHDOG, + BOOKE_INTERRUPT_EXTERNAL, + BOOKE_INTERRUPT_FIT, + BOOKE_INTERRUPT_DECREMENTER, +}; + + +void kvmppc_dump_tlbs(struct kvm_vcpu *vcpu) +{ + struct tlbe *tlbe; + int i; + + printk("vcpu %d TLB dump:\n", vcpu->vcpu_id); + printk("| %2s | %3s | %8s | %8s | %8s |\n", + "nr", "tid", "word0", "word1", "word2"); + + for (i = 0; i < PPC44x_TLB_SIZE; i++) { + tlbe = &vcpu->arch.guest_tlb[i]; + if (tlbe->word0 & PPC44x_TLB_VALID) + printk(" G%2d | %02X | %08X | %08X | %08X |\n", + i, tlbe->tid, tlbe->word0, tlbe->word1, + tlbe->word2); + } + + for (i = 0; i < PPC44x_TLB_SIZE; i++) { + tlbe = &vcpu->arch.shadow_tlb[i]; + if (tlbe->word0 & PPC44x_TLB_VALID) + printk(" S%2d | %02X | %08X | %08X | %08X |\n", + i, tlbe->tid, tlbe->word0, tlbe->word1, + tlbe->word2); + } +} + +/* TODO: use vcpu_printf() */ +void kvmppc_dump_vcpu(struct kvm_vcpu *vcpu) +{ + int i; + + printk("pc: %08x msr: %08x\n", vcpu->arch.pc, vcpu->arch.msr); + printk("lr: %08x ctr: %08x\n", vcpu->arch.lr, vcpu->arch.ctr); + printk("srr0: %08x srr1: %08x\n", vcpu->arch.srr0, vcpu->arch.srr1); + + printk("exceptions: %08lx\n", vcpu->arch.pending_exceptions); + + for (i = 0; i < 32; i += 4) { + printk("gpr%02d: %08x %08x %08x %08x\n", i, + vcpu->arch.gpr[i], + vcpu->arch.gpr[i+1], + vcpu->arch.gpr[i+2], + vcpu->arch.gpr[i+3]); + } +} + +/* Check if we are ready to deliver the interrupt */ +static int kvmppc_can_deliver_interrupt(struct kvm_vcpu *vcpu, int interrupt) +{ + int r; + + switch (interrupt) { + case BOOKE_INTERRUPT_CRITICAL: + r = vcpu->arch.msr & MSR_CE; + break; + case BOOKE_INTERRUPT_MACHINE_CHECK: + r = vcpu->arch.msr & MSR_ME; + break; + case BOOKE_INTERRUPT_EXTERNAL: + r = vcpu->arch.msr & MSR_EE; + break; + case BOOKE_INTERRUPT_DECREMENTER: + r = vcpu->arch.msr & MSR_EE; + break; + case BOOKE_INTERRUPT_FIT: + r = vcpu->arch.msr & MSR_EE; + break; + case BOOKE_INTERRUPT_WATCHDOG: + r = vcpu->arch.msr & MSR_CE; + break; + case BOOKE_INTERRUPT_DEBUG: + r = vcpu->arch.msr & MSR_DE; + break; + default: + r = 1; + } + + return r; +} + +static void kvmppc_deliver_interrupt(struct kvm_vcpu *vcpu, int interrupt) +{ + switch (interrupt) { + case BOOKE_INTERRUPT_DECREMENTER: + vcpu->arch.tsr |= TSR_DIS; + break; + } + + vcpu->arch.srr0 = vcpu->arch.pc; + vcpu->arch.srr1 = vcpu->arch.msr; + vcpu->arch.pc = vcpu->arch.ivpr | vcpu->arch.ivor[interrupt]; + kvmppc_set_msr(vcpu, vcpu->arch.msr & interrupt_msr_mask[interrupt]); +} + +/* Check pending exceptions and deliver one, if possible. */ +void kvmppc_check_and_deliver_interrupts(struct kvm_vcpu *vcpu) +{ + unsigned long *pending = &vcpu->arch.pending_exceptions; + unsigned int exception; + unsigned int priority; + + priority = find_first_bit(pending, BITS_PER_BYTE * sizeof(*pending)); + while (priority <= BOOKE_MAX_INTERRUPT) { + exception = priority_exception[priority]; + if (kvmppc_can_deliver_interrupt(vcpu, exception)) { + kvmppc_clear_exception(vcpu, exception); + kvmppc_deliver_interrupt(vcpu, exception); + break; + } + + priority = find_next_bit(pending, + BITS_PER_BYTE * sizeof(*pending), + priority + 1); + } +} + +static int kvmppc_emulate_mmio(struct kvm_run *run, struct kvm_vcpu *vcpu) +{ + enum emulation_result er; + int r; + + er = kvmppc_emulate_instruction(run, vcpu); + switch (er) { + case EMULATE_DONE: + /* Future optimization: only reload non-volatiles if they were + * actually modified. */ + r = RESUME_GUEST_NV; + break; + case EMULATE_DO_MMIO: + run->exit_reason = KVM_EXIT_MMIO; + /* We must reload nonvolatiles because "update" load/store + * instructions modify register state. */ + /* Future optimization: only reload non-volatiles if they were + * actually modified. */ + r = RESUME_HOST_NV; + break; + case EMULATE_FAIL: + /* XXX Deliver Program interrupt to guest. */ + printk(KERN_EMERG "%s: emulation failed (%08x)\n", __func__, + vcpu->arch.last_inst); + r = RESUME_HOST; + break; + default: + BUG(); + } + + return r; +} + +/** + * kvmppc_handle_exit + * + * Return value is in the form (errcode<<2 | RESUME_FLAG_HOST | RESUME_FLAG_NV) + */ +int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu, + unsigned int exit_nr) +{ + enum emulation_result er; + int r = RESUME_HOST; + + local_irq_enable(); + + run->exit_reason = KVM_EXIT_UNKNOWN; + run->ready_for_interrupt_injection = 1; + + switch (exit_nr) { + case BOOKE_INTERRUPT_MACHINE_CHECK: + printk("MACHINE CHECK: %lx\n", mfspr(SPRN_MCSR)); + kvmppc_dump_vcpu(vcpu); + r = RESUME_HOST; + break; + + case BOOKE_INTERRUPT_EXTERNAL: + case BOOKE_INTERRUPT_DECREMENTER: + /* Since we switched IVPR back to the host's value, the host + * handled this interrupt the moment we enabled interrupts. + * Now we just offer it a chance to reschedule the guest. */ + + /* XXX At this point the TLB still holds our shadow TLB, so if + * we do reschedule the host will fault over it. Perhaps we + * should politely restore the host's entries to minimize + * misses before ceding control. */ + if (need_resched()) + cond_resched(); + if (exit_nr == BOOKE_INTERRUPT_DECREMENTER) + vcpu->stat.dec_exits++; + else + vcpu->stat.ext_intr_exits++; + r = RESUME_GUEST; + break; + + case BOOKE_INTERRUPT_PROGRAM: + if (vcpu->arch.msr & MSR_PR) { + /* Program traps generated by user-level software must be handled + * by the guest kernel. */ + vcpu->arch.esr = vcpu->arch.fault_esr; + kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_PROGRAM); + r = RESUME_GUEST; + break; + } + + er = kvmppc_emulate_instruction(run, vcpu); + switch (er) { + case EMULATE_DONE: + /* Future optimization: only reload non-volatiles if + * they were actually modified by emulation. */ + vcpu->stat.emulated_inst_exits++; + r = RESUME_GUEST_NV; + break; + case EMULATE_DO_DCR: + run->exit_reason = KVM_EXIT_DCR; + r = RESUME_HOST; + break; + case EMULATE_FAIL: + /* XXX Deliver Program interrupt to guest. */ + printk(KERN_CRIT "%s: emulation at %x failed (%08x)\n", + __func__, vcpu->arch.pc, vcpu->arch.last_inst); + /* For debugging, encode the failing instruction and + * report it to userspace. */ + run->hw.hardware_exit_reason = ~0ULL << 32; + run->hw.hardware_exit_reason |= vcpu->arch.last_inst; + r = RESUME_HOST; + break; + default: + BUG(); + } + break; + + case BOOKE_INTERRUPT_DATA_STORAGE: + vcpu->arch.dear = vcpu->arch.fault_dear; + vcpu->arch.esr = vcpu->arch.fault_esr; + kvmppc_queue_exception(vcpu, exit_nr); + vcpu->stat.dsi_exits++; + r = RESUME_GUEST; + break; + + case BOOKE_INTERRUPT_INST_STORAGE: + vcpu->arch.esr = vcpu->arch.fault_esr; + kvmppc_queue_exception(vcpu, exit_nr); + vcpu->stat.isi_exits++; + r = RESUME_GUEST; + break; + + case BOOKE_INTERRUPT_SYSCALL: + kvmppc_queue_exception(vcpu, exit_nr); + vcpu->stat.syscall_exits++; + r = RESUME_GUEST; + break; + + case BOOKE_INTERRUPT_DTLB_MISS: { + struct tlbe *gtlbe; + unsigned long eaddr = vcpu->arch.fault_dear; + gfn_t gfn; + + /* Check the guest TLB. */ + gtlbe = kvmppc_44x_dtlb_search(vcpu, eaddr); + if (!gtlbe) { + /* The guest didn't have a mapping for it. */ + kvmppc_queue_exception(vcpu, exit_nr); + vcpu->arch.dear = vcpu->arch.fault_dear; + vcpu->arch.esr = vcpu->arch.fault_esr; + vcpu->stat.dtlb_real_miss_exits++; + r = RESUME_GUEST; + break; + } + + vcpu->arch.paddr_accessed = tlb_xlate(gtlbe, eaddr); + gfn = vcpu->arch.paddr_accessed >> PAGE_SHIFT; + + if (kvm_is_visible_gfn(vcpu->kvm, gfn)) { + /* The guest TLB had a mapping, but the shadow TLB + * didn't, and it is RAM. This could be because: + * a) the entry is mapping the host kernel, or + * b) the guest used a large mapping which we're faking + * Either way, we need to satisfy the fault without + * invoking the guest. */ + kvmppc_mmu_map(vcpu, eaddr, gfn, gtlbe->tid, + gtlbe->word2); + vcpu->stat.dtlb_virt_miss_exits++; + r = RESUME_GUEST; + } else { + /* Guest has mapped and accessed a page which is not + * actually RAM. */ + r = kvmppc_emulate_mmio(run, vcpu); + } + + break; + } + + case BOOKE_INTERRUPT_ITLB_MISS: { + struct tlbe *gtlbe; + unsigned long eaddr = vcpu->arch.pc; + gfn_t gfn; + + r = RESUME_GUEST; + + /* Check the guest TLB. */ + gtlbe = kvmppc_44x_itlb_search(vcpu, eaddr); + if (!gtlbe) { + /* The guest didn't have a mapping for it. */ + kvmppc_queue_exception(vcpu, exit_nr); + vcpu->stat.itlb_real_miss_exits++; + break; + } + + vcpu->stat.itlb_virt_miss_exits++; + + gfn = tlb_xlate(gtlbe, eaddr) >> PAGE_SHIFT; + + if (kvm_is_visible_gfn(vcpu->kvm, gfn)) { + /* The guest TLB had a mapping, but the shadow TLB + * didn't. This could be because: + * a) the entry is mapping the host kernel, or + * b) the guest used a large mapping which we're faking + * Either way, we need to satisfy the fault without + * invoking the guest. */ + kvmppc_mmu_map(vcpu, eaddr, gfn, gtlbe->tid, + gtlbe->word2); + } else { + /* Guest mapped and leaped at non-RAM! */ + kvmppc_queue_exception(vcpu, + BOOKE_INTERRUPT_MACHINE_CHECK); + } + + break; + } + + default: + printk(KERN_EMERG "exit_nr %d\n", exit_nr); + BUG(); + } + + local_irq_disable(); + + kvmppc_check_and_deliver_interrupts(vcpu); + + /* Do some exit accounting. */ + vcpu->stat.sum_exits++; + if (!(r & RESUME_HOST)) { + /* To avoid clobbering exit_reason, only check for signals if + * we aren't already exiting to userspace for some other + * reason. */ + if (signal_pending(current)) { + run->exit_reason = KVM_EXIT_INTR; + r = (-EINTR << 2) | RESUME_HOST | (r & RESUME_FLAG_NV); + + vcpu->stat.signal_exits++; + } else { + vcpu->stat.light_exits++; + } + } else { + switch (run->exit_reason) { + case KVM_EXIT_MMIO: + vcpu->stat.mmio_exits++; + break; + case KVM_EXIT_DCR: + vcpu->stat.dcr_exits++; + break; + case KVM_EXIT_INTR: + vcpu->stat.signal_exits++; + break; + } + } + + return r; +} + +/* Initial guest state: 16MB mapping 0 -> 0, PC = 0, MSR = 0, R1 = 16MB */ +int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) +{ + struct tlbe *tlbe = &vcpu->arch.guest_tlb[0]; + + tlbe->tid = 0; + tlbe->word0 = PPC44x_TLB_16M | PPC44x_TLB_VALID; + tlbe->word1 = 0; + tlbe->word2 = PPC44x_TLB_SX | PPC44x_TLB_SW | PPC44x_TLB_SR; + + tlbe++; + tlbe->tid = 0; + tlbe->word0 = 0xef600000 | PPC44x_TLB_4K | PPC44x_TLB_VALID; + tlbe->word1 = 0xef600000; + tlbe->word2 = PPC44x_TLB_SX | PPC44x_TLB_SW | PPC44x_TLB_SR + | PPC44x_TLB_I | PPC44x_TLB_G; + + vcpu->arch.pc = 0; + vcpu->arch.msr = 0; + vcpu->arch.gpr[1] = (16<<20) - 8; /* -8 for the callee-save LR slot */ + + /* Eye-catching number so we know if the guest takes an interrupt + * before it's programmed its own IVPR. */ + vcpu->arch.ivpr = 0x55550000; + + /* Since the guest can directly access the timebase, it must know the + * real timebase frequency. Accordingly, it must see the state of + * CCR1[TCS]. */ + vcpu->arch.ccr1 = mfspr(SPRN_CCR1); + + return 0; +} + +int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +{ + int i; + + regs->pc = vcpu->arch.pc; + regs->cr = vcpu->arch.cr; + regs->ctr = vcpu->arch.ctr; + regs->lr = vcpu->arch.lr; + regs->xer = vcpu->arch.xer; + regs->msr = vcpu->arch.msr; + regs->srr0 = vcpu->arch.srr0; + regs->srr1 = vcpu->arch.srr1; + regs->pid = vcpu->arch.pid; + regs->sprg0 = vcpu->arch.sprg0; + regs->sprg1 = vcpu->arch.sprg1; + regs->sprg2 = vcpu->arch.sprg2; + regs->sprg3 = vcpu->arch.sprg3; + regs->sprg5 = vcpu->arch.sprg4; + regs->sprg6 = vcpu->arch.sprg5; + regs->sprg7 = vcpu->arch.sprg6; + + for (i = 0; i < ARRAY_SIZE(regs->gpr); i++) + regs->gpr[i] = vcpu->arch.gpr[i]; + + return 0; +} + +int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +{ + int i; + + vcpu->arch.pc = regs->pc; + vcpu->arch.cr = regs->cr; + vcpu->arch.ctr = regs->ctr; + vcpu->arch.lr = regs->lr; + vcpu->arch.xer = regs->xer; + vcpu->arch.msr = regs->msr; + vcpu->arch.srr0 = regs->srr0; + vcpu->arch.srr1 = regs->srr1; + vcpu->arch.sprg0 = regs->sprg0; + vcpu->arch.sprg1 = regs->sprg1; + vcpu->arch.sprg2 = regs->sprg2; + vcpu->arch.sprg3 = regs->sprg3; + vcpu->arch.sprg5 = regs->sprg4; + vcpu->arch.sprg6 = regs->sprg5; + vcpu->arch.sprg7 = regs->sprg6; + + for (i = 0; i < ARRAY_SIZE(vcpu->arch.gpr); i++) + vcpu->arch.gpr[i] = regs->gpr[i]; + + return 0; +} + +int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu, + struct kvm_sregs *sregs) +{ + return -ENOTSUPP; +} + +int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu, + struct kvm_sregs *sregs) +{ + return -ENOTSUPP; +} + +int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu) +{ + return -ENOTSUPP; +} + +int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu) +{ + return -ENOTSUPP; +} + +/* 'linear_address' is actually an encoding of AS|PID|EADDR . */ +int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu, + struct kvm_translation *tr) +{ + struct tlbe *gtlbe; + int index; + gva_t eaddr; + u8 pid; + u8 as; + + eaddr = tr->linear_address; + pid = (tr->linear_address >> 32) & 0xff; + as = (tr->linear_address >> 40) & 0x1; + + index = kvmppc_44x_tlb_index(vcpu, eaddr, pid, as); + if (index == -1) { + tr->valid = 0; + return 0; + } + + gtlbe = &vcpu->arch.guest_tlb[index]; + + tr->physical_address = tlb_xlate(gtlbe, eaddr); + /* XXX what does "writeable" and "usermode" even mean? */ + tr->valid = 1; + + return 0; +} diff --git a/arch/powerpc/kvm/booke_host.c b/arch/powerpc/kvm/booke_host.c new file mode 100644 index 0000000000000000000000000000000000000000..b480341bc31eb2cf01b40e52cf275f6f960202ff --- /dev/null +++ b/arch/powerpc/kvm/booke_host.c @@ -0,0 +1,83 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License, version 2, as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2008 + * + * Authors: Hollis Blanchard + */ + +#include +#include +#include +#include +#include + +unsigned long kvmppc_booke_handlers; + +static int kvmppc_booke_init(void) +{ + unsigned long ivor[16]; + unsigned long max_ivor = 0; + int i; + + /* We install our own exception handlers by hijacking IVPR. IVPR must + * be 16-bit aligned, so we need a 64KB allocation. */ + kvmppc_booke_handlers = __get_free_pages(GFP_KERNEL | __GFP_ZERO, + VCPU_SIZE_ORDER); + if (!kvmppc_booke_handlers) + return -ENOMEM; + + /* XXX make sure our handlers are smaller than Linux's */ + + /* Copy our interrupt handlers to match host IVORs. That way we don't + * have to swap the IVORs on every guest/host transition. */ + ivor[0] = mfspr(SPRN_IVOR0); + ivor[1] = mfspr(SPRN_IVOR1); + ivor[2] = mfspr(SPRN_IVOR2); + ivor[3] = mfspr(SPRN_IVOR3); + ivor[4] = mfspr(SPRN_IVOR4); + ivor[5] = mfspr(SPRN_IVOR5); + ivor[6] = mfspr(SPRN_IVOR6); + ivor[7] = mfspr(SPRN_IVOR7); + ivor[8] = mfspr(SPRN_IVOR8); + ivor[9] = mfspr(SPRN_IVOR9); + ivor[10] = mfspr(SPRN_IVOR10); + ivor[11] = mfspr(SPRN_IVOR11); + ivor[12] = mfspr(SPRN_IVOR12); + ivor[13] = mfspr(SPRN_IVOR13); + ivor[14] = mfspr(SPRN_IVOR14); + ivor[15] = mfspr(SPRN_IVOR15); + + for (i = 0; i < 16; i++) { + if (ivor[i] > max_ivor) + max_ivor = ivor[i]; + + memcpy((void *)kvmppc_booke_handlers + ivor[i], + kvmppc_handlers_start + i * kvmppc_handler_len, + kvmppc_handler_len); + } + flush_icache_range(kvmppc_booke_handlers, + kvmppc_booke_handlers + max_ivor + kvmppc_handler_len); + + return kvm_init(NULL, sizeof(struct kvm_vcpu), THIS_MODULE); +} + +static void __exit kvmppc_booke_exit(void) +{ + free_pages(kvmppc_booke_handlers, VCPU_SIZE_ORDER); + kvm_exit(); +} + +module_init(kvmppc_booke_init) +module_exit(kvmppc_booke_exit) diff --git a/arch/powerpc/kvm/booke_interrupts.S b/arch/powerpc/kvm/booke_interrupts.S new file mode 100644 index 0000000000000000000000000000000000000000..3b653b5309b8ecc2dad5b5137163a9465408ca22 --- /dev/null +++ b/arch/powerpc/kvm/booke_interrupts.S @@ -0,0 +1,436 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License, version 2, as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2007 + * + * Authors: Hollis Blanchard + */ + +#include +#include +#include +#include +#include +#include + +#define KVMPPC_MSR_MASK (MSR_CE|MSR_EE|MSR_PR|MSR_DE|MSR_ME|MSR_IS|MSR_DS) + +#define VCPU_GPR(n) (VCPU_GPRS + (n * 4)) + +/* The host stack layout: */ +#define HOST_R1 0 /* Implied by stwu. */ +#define HOST_CALLEE_LR 4 +#define HOST_RUN 8 +/* r2 is special: it holds 'current', and it made nonvolatile in the + * kernel with the -ffixed-r2 gcc option. */ +#define HOST_R2 12 +#define HOST_NV_GPRS 16 +#define HOST_NV_GPR(n) (HOST_NV_GPRS + ((n - 14) * 4)) +#define HOST_MIN_STACK_SIZE (HOST_NV_GPR(31) + 4) +#define HOST_STACK_SIZE (((HOST_MIN_STACK_SIZE + 15) / 16) * 16) /* Align. */ +#define HOST_STACK_LR (HOST_STACK_SIZE + 4) /* In caller stack frame. */ + +#define NEED_INST_MASK ((1< + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include "44x_tlb.h" + +/* Instruction decoding */ +static inline unsigned int get_op(u32 inst) +{ + return inst >> 26; +} + +static inline unsigned int get_xop(u32 inst) +{ + return (inst >> 1) & 0x3ff; +} + +static inline unsigned int get_sprn(u32 inst) +{ + return ((inst >> 16) & 0x1f) | ((inst >> 6) & 0x3e0); +} + +static inline unsigned int get_dcrn(u32 inst) +{ + return ((inst >> 16) & 0x1f) | ((inst >> 6) & 0x3e0); +} + +static inline unsigned int get_rt(u32 inst) +{ + return (inst >> 21) & 0x1f; +} + +static inline unsigned int get_rs(u32 inst) +{ + return (inst >> 21) & 0x1f; +} + +static inline unsigned int get_ra(u32 inst) +{ + return (inst >> 16) & 0x1f; +} + +static inline unsigned int get_rb(u32 inst) +{ + return (inst >> 11) & 0x1f; +} + +static inline unsigned int get_rc(u32 inst) +{ + return inst & 0x1; +} + +static inline unsigned int get_ws(u32 inst) +{ + return (inst >> 11) & 0x1f; +} + +static inline unsigned int get_d(u32 inst) +{ + return inst & 0xffff; +} + +static int tlbe_is_host_safe(const struct kvm_vcpu *vcpu, + const struct tlbe *tlbe) +{ + gpa_t gpa; + + if (!get_tlb_v(tlbe)) + return 0; + + /* Does it match current guest AS? */ + /* XXX what about IS != DS? */ + if (get_tlb_ts(tlbe) != !!(vcpu->arch.msr & MSR_IS)) + return 0; + + gpa = get_tlb_raddr(tlbe); + if (!gfn_to_memslot(vcpu->kvm, gpa >> PAGE_SHIFT)) + /* Mapping is not for RAM. */ + return 0; + + return 1; +} + +static int kvmppc_emul_tlbwe(struct kvm_vcpu *vcpu, u32 inst) +{ + u64 eaddr; + u64 raddr; + u64 asid; + u32 flags; + struct tlbe *tlbe; + unsigned int ra; + unsigned int rs; + unsigned int ws; + unsigned int index; + + ra = get_ra(inst); + rs = get_rs(inst); + ws = get_ws(inst); + + index = vcpu->arch.gpr[ra]; + if (index > PPC44x_TLB_SIZE) { + printk("%s: index %d\n", __func__, index); + kvmppc_dump_vcpu(vcpu); + return EMULATE_FAIL; + } + + tlbe = &vcpu->arch.guest_tlb[index]; + + /* Invalidate shadow mappings for the about-to-be-clobbered TLBE. */ + if (tlbe->word0 & PPC44x_TLB_VALID) { + eaddr = get_tlb_eaddr(tlbe); + asid = (tlbe->word0 & PPC44x_TLB_TS) | tlbe->tid; + kvmppc_mmu_invalidate(vcpu, eaddr, asid); + } + + switch (ws) { + case PPC44x_TLB_PAGEID: + tlbe->tid = vcpu->arch.mmucr & 0xff; + tlbe->word0 = vcpu->arch.gpr[rs]; + break; + + case PPC44x_TLB_XLAT: + tlbe->word1 = vcpu->arch.gpr[rs]; + break; + + case PPC44x_TLB_ATTRIB: + tlbe->word2 = vcpu->arch.gpr[rs]; + break; + + default: + return EMULATE_FAIL; + } + + if (tlbe_is_host_safe(vcpu, tlbe)) { + eaddr = get_tlb_eaddr(tlbe); + raddr = get_tlb_raddr(tlbe); + asid = (tlbe->word0 & PPC44x_TLB_TS) | tlbe->tid; + flags = tlbe->word2 & 0xffff; + + /* Create a 4KB mapping on the host. If the guest wanted a + * large page, only the first 4KB is mapped here and the rest + * are mapped on the fly. */ + kvmppc_mmu_map(vcpu, eaddr, raddr >> PAGE_SHIFT, asid, flags); + } + + return EMULATE_DONE; +} + +static void kvmppc_emulate_dec(struct kvm_vcpu *vcpu) +{ + if (vcpu->arch.tcr & TCR_DIE) { + /* The decrementer ticks at the same rate as the timebase, so + * that's how we convert the guest DEC value to the number of + * host ticks. */ + unsigned long nr_jiffies; + + nr_jiffies = vcpu->arch.dec / tb_ticks_per_jiffy; + mod_timer(&vcpu->arch.dec_timer, + get_jiffies_64() + nr_jiffies); + } else { + del_timer(&vcpu->arch.dec_timer); + } +} + +static void kvmppc_emul_rfi(struct kvm_vcpu *vcpu) +{ + vcpu->arch.pc = vcpu->arch.srr0; + kvmppc_set_msr(vcpu, vcpu->arch.srr1); +} + +/* XXX to do: + * lhax + * lhaux + * lswx + * lswi + * stswx + * stswi + * lha + * lhau + * lmw + * stmw + * + * XXX is_bigendian should depend on MMU mapping or MSR[LE] + */ +int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu) +{ + u32 inst = vcpu->arch.last_inst; + u32 ea; + int ra; + int rb; + int rc; + int rs; + int rt; + int sprn; + int dcrn; + enum emulation_result emulated = EMULATE_DONE; + int advance = 1; + + switch (get_op(inst)) { + case 3: /* trap */ + printk("trap!\n"); + kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_PROGRAM); + advance = 0; + break; + + case 19: + switch (get_xop(inst)) { + case 50: /* rfi */ + kvmppc_emul_rfi(vcpu); + advance = 0; + break; + + default: + emulated = EMULATE_FAIL; + break; + } + break; + + case 31: + switch (get_xop(inst)) { + + case 83: /* mfmsr */ + rt = get_rt(inst); + vcpu->arch.gpr[rt] = vcpu->arch.msr; + break; + + case 87: /* lbzx */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 1, 1); + break; + + case 131: /* wrtee */ + rs = get_rs(inst); + vcpu->arch.msr = (vcpu->arch.msr & ~MSR_EE) + | (vcpu->arch.gpr[rs] & MSR_EE); + break; + + case 146: /* mtmsr */ + rs = get_rs(inst); + kvmppc_set_msr(vcpu, vcpu->arch.gpr[rs]); + break; + + case 163: /* wrteei */ + vcpu->arch.msr = (vcpu->arch.msr & ~MSR_EE) + | (inst & MSR_EE); + break; + + case 215: /* stbx */ + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, + vcpu->arch.gpr[rs], + 1, 1); + break; + + case 247: /* stbux */ + rs = get_rs(inst); + ra = get_ra(inst); + rb = get_rb(inst); + + ea = vcpu->arch.gpr[rb]; + if (ra) + ea += vcpu->arch.gpr[ra]; + + emulated = kvmppc_handle_store(run, vcpu, + vcpu->arch.gpr[rs], + 1, 1); + vcpu->arch.gpr[rs] = ea; + break; + + case 279: /* lhzx */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 2, 1); + break; + + case 311: /* lhzux */ + rt = get_rt(inst); + ra = get_ra(inst); + rb = get_rb(inst); + + ea = vcpu->arch.gpr[rb]; + if (ra) + ea += vcpu->arch.gpr[ra]; + + emulated = kvmppc_handle_load(run, vcpu, rt, 2, 1); + vcpu->arch.gpr[ra] = ea; + break; + + case 323: /* mfdcr */ + dcrn = get_dcrn(inst); + rt = get_rt(inst); + + /* The guest may access CPR0 registers to determine the timebase + * frequency, and it must know the real host frequency because it + * can directly access the timebase registers. + * + * It would be possible to emulate those accesses in userspace, + * but userspace can really only figure out the end frequency. + * We could decompose that into the factors that compute it, but + * that's tricky math, and it's easier to just report the real + * CPR0 values. + */ + switch (dcrn) { + case DCRN_CPR0_CONFIG_ADDR: + vcpu->arch.gpr[rt] = vcpu->arch.cpr0_cfgaddr; + break; + case DCRN_CPR0_CONFIG_DATA: + local_irq_disable(); + mtdcr(DCRN_CPR0_CONFIG_ADDR, + vcpu->arch.cpr0_cfgaddr); + vcpu->arch.gpr[rt] = mfdcr(DCRN_CPR0_CONFIG_DATA); + local_irq_enable(); + break; + default: + run->dcr.dcrn = dcrn; + run->dcr.data = 0; + run->dcr.is_write = 0; + vcpu->arch.io_gpr = rt; + vcpu->arch.dcr_needed = 1; + emulated = EMULATE_DO_DCR; + } + + break; + + case 339: /* mfspr */ + sprn = get_sprn(inst); + rt = get_rt(inst); + + switch (sprn) { + case SPRN_SRR0: + vcpu->arch.gpr[rt] = vcpu->arch.srr0; break; + case SPRN_SRR1: + vcpu->arch.gpr[rt] = vcpu->arch.srr1; break; + case SPRN_MMUCR: + vcpu->arch.gpr[rt] = vcpu->arch.mmucr; break; + case SPRN_PID: + vcpu->arch.gpr[rt] = vcpu->arch.pid; break; + case SPRN_IVPR: + vcpu->arch.gpr[rt] = vcpu->arch.ivpr; break; + case SPRN_CCR0: + vcpu->arch.gpr[rt] = vcpu->arch.ccr0; break; + case SPRN_CCR1: + vcpu->arch.gpr[rt] = vcpu->arch.ccr1; break; + case SPRN_PVR: + vcpu->arch.gpr[rt] = vcpu->arch.pvr; break; + case SPRN_DEAR: + vcpu->arch.gpr[rt] = vcpu->arch.dear; break; + case SPRN_ESR: + vcpu->arch.gpr[rt] = vcpu->arch.esr; break; + case SPRN_DBCR0: + vcpu->arch.gpr[rt] = vcpu->arch.dbcr0; break; + case SPRN_DBCR1: + vcpu->arch.gpr[rt] = vcpu->arch.dbcr1; break; + + /* Note: mftb and TBRL/TBWL are user-accessible, so + * the guest can always access the real TB anyways. + * In fact, we probably will never see these traps. */ + case SPRN_TBWL: + vcpu->arch.gpr[rt] = mftbl(); break; + case SPRN_TBWU: + vcpu->arch.gpr[rt] = mftbu(); break; + + case SPRN_SPRG0: + vcpu->arch.gpr[rt] = vcpu->arch.sprg0; break; + case SPRN_SPRG1: + vcpu->arch.gpr[rt] = vcpu->arch.sprg1; break; + case SPRN_SPRG2: + vcpu->arch.gpr[rt] = vcpu->arch.sprg2; break; + case SPRN_SPRG3: + vcpu->arch.gpr[rt] = vcpu->arch.sprg3; break; + /* Note: SPRG4-7 are user-readable, so we don't get + * a trap. */ + + case SPRN_IVOR0: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[0]; break; + case SPRN_IVOR1: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[1]; break; + case SPRN_IVOR2: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[2]; break; + case SPRN_IVOR3: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[3]; break; + case SPRN_IVOR4: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[4]; break; + case SPRN_IVOR5: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[5]; break; + case SPRN_IVOR6: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[6]; break; + case SPRN_IVOR7: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[7]; break; + case SPRN_IVOR8: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[8]; break; + case SPRN_IVOR9: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[9]; break; + case SPRN_IVOR10: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[10]; break; + case SPRN_IVOR11: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[11]; break; + case SPRN_IVOR12: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[12]; break; + case SPRN_IVOR13: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[13]; break; + case SPRN_IVOR14: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[14]; break; + case SPRN_IVOR15: + vcpu->arch.gpr[rt] = vcpu->arch.ivor[15]; break; + + default: + printk("mfspr: unknown spr %x\n", sprn); + vcpu->arch.gpr[rt] = 0; + break; + } + break; + + case 407: /* sthx */ + rs = get_rs(inst); + ra = get_ra(inst); + rb = get_rb(inst); + + emulated = kvmppc_handle_store(run, vcpu, + vcpu->arch.gpr[rs], + 2, 1); + break; + + case 439: /* sthux */ + rs = get_rs(inst); + ra = get_ra(inst); + rb = get_rb(inst); + + ea = vcpu->arch.gpr[rb]; + if (ra) + ea += vcpu->arch.gpr[ra]; + + emulated = kvmppc_handle_store(run, vcpu, + vcpu->arch.gpr[rs], + 2, 1); + vcpu->arch.gpr[ra] = ea; + break; + + case 451: /* mtdcr */ + dcrn = get_dcrn(inst); + rs = get_rs(inst); + + /* emulate some access in kernel */ + switch (dcrn) { + case DCRN_CPR0_CONFIG_ADDR: + vcpu->arch.cpr0_cfgaddr = vcpu->arch.gpr[rs]; + break; + default: + run->dcr.dcrn = dcrn; + run->dcr.data = vcpu->arch.gpr[rs]; + run->dcr.is_write = 1; + vcpu->arch.dcr_needed = 1; + emulated = EMULATE_DO_DCR; + } + + break; + + case 467: /* mtspr */ + sprn = get_sprn(inst); + rs = get_rs(inst); + switch (sprn) { + case SPRN_SRR0: + vcpu->arch.srr0 = vcpu->arch.gpr[rs]; break; + case SPRN_SRR1: + vcpu->arch.srr1 = vcpu->arch.gpr[rs]; break; + case SPRN_MMUCR: + vcpu->arch.mmucr = vcpu->arch.gpr[rs]; break; + case SPRN_PID: + vcpu->arch.pid = vcpu->arch.gpr[rs]; break; + case SPRN_CCR0: + vcpu->arch.ccr0 = vcpu->arch.gpr[rs]; break; + case SPRN_CCR1: + vcpu->arch.ccr1 = vcpu->arch.gpr[rs]; break; + case SPRN_DEAR: + vcpu->arch.dear = vcpu->arch.gpr[rs]; break; + case SPRN_ESR: + vcpu->arch.esr = vcpu->arch.gpr[rs]; break; + case SPRN_DBCR0: + vcpu->arch.dbcr0 = vcpu->arch.gpr[rs]; break; + case SPRN_DBCR1: + vcpu->arch.dbcr1 = vcpu->arch.gpr[rs]; break; + + /* XXX We need to context-switch the timebase for + * watchdog and FIT. */ + case SPRN_TBWL: break; + case SPRN_TBWU: break; + + case SPRN_DEC: + vcpu->arch.dec = vcpu->arch.gpr[rs]; + kvmppc_emulate_dec(vcpu); + break; + + case SPRN_TSR: + vcpu->arch.tsr &= ~vcpu->arch.gpr[rs]; break; + + case SPRN_TCR: + vcpu->arch.tcr = vcpu->arch.gpr[rs]; + kvmppc_emulate_dec(vcpu); + break; + + case SPRN_SPRG0: + vcpu->arch.sprg0 = vcpu->arch.gpr[rs]; break; + case SPRN_SPRG1: + vcpu->arch.sprg1 = vcpu->arch.gpr[rs]; break; + case SPRN_SPRG2: + vcpu->arch.sprg2 = vcpu->arch.gpr[rs]; break; + case SPRN_SPRG3: + vcpu->arch.sprg3 = vcpu->arch.gpr[rs]; break; + + /* Note: SPRG4-7 are user-readable. These values are + * loaded into the real SPRGs when resuming the + * guest. */ + case SPRN_SPRG4: + vcpu->arch.sprg4 = vcpu->arch.gpr[rs]; break; + case SPRN_SPRG5: + vcpu->arch.sprg5 = vcpu->arch.gpr[rs]; break; + case SPRN_SPRG6: + vcpu->arch.sprg6 = vcpu->arch.gpr[rs]; break; + case SPRN_SPRG7: + vcpu->arch.sprg7 = vcpu->arch.gpr[rs]; break; + + case SPRN_IVPR: + vcpu->arch.ivpr = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR0: + vcpu->arch.ivor[0] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR1: + vcpu->arch.ivor[1] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR2: + vcpu->arch.ivor[2] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR3: + vcpu->arch.ivor[3] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR4: + vcpu->arch.ivor[4] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR5: + vcpu->arch.ivor[5] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR6: + vcpu->arch.ivor[6] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR7: + vcpu->arch.ivor[7] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR8: + vcpu->arch.ivor[8] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR9: + vcpu->arch.ivor[9] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR10: + vcpu->arch.ivor[10] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR11: + vcpu->arch.ivor[11] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR12: + vcpu->arch.ivor[12] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR13: + vcpu->arch.ivor[13] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR14: + vcpu->arch.ivor[14] = vcpu->arch.gpr[rs]; break; + case SPRN_IVOR15: + vcpu->arch.ivor[15] = vcpu->arch.gpr[rs]; break; + + default: + printk("mtspr: unknown spr %x\n", sprn); + emulated = EMULATE_FAIL; + break; + } + break; + + case 470: /* dcbi */ + /* Do nothing. The guest is performing dcbi because + * hardware DMA is not snooped by the dcache, but + * emulated DMA either goes through the dcache as + * normal writes, or the host kernel has handled dcache + * coherence. */ + break; + + case 534: /* lwbrx */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 4, 0); + break; + + case 566: /* tlbsync */ + break; + + case 662: /* stwbrx */ + rs = get_rs(inst); + ra = get_ra(inst); + rb = get_rb(inst); + + emulated = kvmppc_handle_store(run, vcpu, + vcpu->arch.gpr[rs], + 4, 0); + break; + + case 978: /* tlbwe */ + emulated = kvmppc_emul_tlbwe(vcpu, inst); + break; + + case 914: { /* tlbsx */ + int index; + unsigned int as = get_mmucr_sts(vcpu); + unsigned int pid = get_mmucr_stid(vcpu); + + rt = get_rt(inst); + ra = get_ra(inst); + rb = get_rb(inst); + rc = get_rc(inst); + + ea = vcpu->arch.gpr[rb]; + if (ra) + ea += vcpu->arch.gpr[ra]; + + index = kvmppc_44x_tlb_index(vcpu, ea, pid, as); + if (rc) { + if (index < 0) + vcpu->arch.cr &= ~0x20000000; + else + vcpu->arch.cr |= 0x20000000; + } + vcpu->arch.gpr[rt] = index; + + } + break; + + case 790: /* lhbrx */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 2, 0); + break; + + case 918: /* sthbrx */ + rs = get_rs(inst); + ra = get_ra(inst); + rb = get_rb(inst); + + emulated = kvmppc_handle_store(run, vcpu, + vcpu->arch.gpr[rs], + 2, 0); + break; + + case 966: /* iccci */ + break; + + default: + printk("unknown: op %d xop %d\n", get_op(inst), + get_xop(inst)); + emulated = EMULATE_FAIL; + break; + } + break; + + case 32: /* lwz */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 4, 1); + break; + + case 33: /* lwzu */ + ra = get_ra(inst); + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 4, 1); + vcpu->arch.gpr[ra] = vcpu->arch.paddr_accessed; + break; + + case 34: /* lbz */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 1, 1); + break; + + case 35: /* lbzu */ + ra = get_ra(inst); + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 1, 1); + vcpu->arch.gpr[ra] = vcpu->arch.paddr_accessed; + break; + + case 36: /* stw */ + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, vcpu->arch.gpr[rs], + 4, 1); + break; + + case 37: /* stwu */ + ra = get_ra(inst); + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, vcpu->arch.gpr[rs], + 4, 1); + vcpu->arch.gpr[ra] = vcpu->arch.paddr_accessed; + break; + + case 38: /* stb */ + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, vcpu->arch.gpr[rs], + 1, 1); + break; + + case 39: /* stbu */ + ra = get_ra(inst); + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, vcpu->arch.gpr[rs], + 1, 1); + vcpu->arch.gpr[ra] = vcpu->arch.paddr_accessed; + break; + + case 40: /* lhz */ + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 2, 1); + break; + + case 41: /* lhzu */ + ra = get_ra(inst); + rt = get_rt(inst); + emulated = kvmppc_handle_load(run, vcpu, rt, 2, 1); + vcpu->arch.gpr[ra] = vcpu->arch.paddr_accessed; + break; + + case 44: /* sth */ + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, vcpu->arch.gpr[rs], + 2, 1); + break; + + case 45: /* sthu */ + ra = get_ra(inst); + rs = get_rs(inst); + emulated = kvmppc_handle_store(run, vcpu, vcpu->arch.gpr[rs], + 2, 1); + vcpu->arch.gpr[ra] = vcpu->arch.paddr_accessed; + break; + + default: + printk("unknown op %d\n", get_op(inst)); + emulated = EMULATE_FAIL; + break; + } + + if (advance) + vcpu->arch.pc += 4; /* Advance past emulated instruction. */ + + return emulated; +} diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c new file mode 100644 index 0000000000000000000000000000000000000000..bad40bd2d3ac5e966926487cb10df89513e2acc6 --- /dev/null +++ b/arch/powerpc/kvm/powerpc.c @@ -0,0 +1,436 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License, version 2, as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright IBM Corp. 2007 + * + * Authors: Hollis Blanchard + * Christian Ehrhardt + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + + +gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn) +{ + return gfn; +} + +int kvm_cpu_has_interrupt(struct kvm_vcpu *v) +{ + /* XXX implement me */ + return 0; +} + +int kvm_arch_vcpu_runnable(struct kvm_vcpu *v) +{ + return 1; +} + + +int kvmppc_emulate_mmio(struct kvm_run *run, struct kvm_vcpu *vcpu) +{ + enum emulation_result er; + int r; + + er = kvmppc_emulate_instruction(run, vcpu); + switch (er) { + case EMULATE_DONE: + /* Future optimization: only reload non-volatiles if they were + * actually modified. */ + r = RESUME_GUEST_NV; + break; + case EMULATE_DO_MMIO: + run->exit_reason = KVM_EXIT_MMIO; + /* We must reload nonvolatiles because "update" load/store + * instructions modify register state. */ + /* Future optimization: only reload non-volatiles if they were + * actually modified. */ + r = RESUME_HOST_NV; + break; + case EMULATE_FAIL: + /* XXX Deliver Program interrupt to guest. */ + printk(KERN_EMERG "%s: emulation failed (%08x)\n", __func__, + vcpu->arch.last_inst); + r = RESUME_HOST; + break; + default: + BUG(); + } + + return r; +} + +void kvm_arch_hardware_enable(void *garbage) +{ +} + +void kvm_arch_hardware_disable(void *garbage) +{ +} + +int kvm_arch_hardware_setup(void) +{ + return 0; +} + +void kvm_arch_hardware_unsetup(void) +{ +} + +void kvm_arch_check_processor_compat(void *rtn) +{ + int r; + + if (strcmp(cur_cpu_spec->platform, "ppc440") == 0) + r = 0; + else + r = -ENOTSUPP; + + *(int *)rtn = r; +} + +struct kvm *kvm_arch_create_vm(void) +{ + struct kvm *kvm; + + kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL); + if (!kvm) + return ERR_PTR(-ENOMEM); + + return kvm; +} + +static void kvmppc_free_vcpus(struct kvm *kvm) +{ + unsigned int i; + + for (i = 0; i < KVM_MAX_VCPUS; ++i) { + if (kvm->vcpus[i]) { + kvm_arch_vcpu_free(kvm->vcpus[i]); + kvm->vcpus[i] = NULL; + } + } +} + +void kvm_arch_destroy_vm(struct kvm *kvm) +{ + kvmppc_free_vcpus(kvm); + kvm_free_physmem(kvm); + kfree(kvm); +} + +int kvm_dev_ioctl_check_extension(long ext) +{ + int r; + + switch (ext) { + case KVM_CAP_USER_MEMORY: + r = 1; + break; + default: + r = 0; + break; + } + return r; + +} + +long kvm_arch_dev_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + return -EINVAL; +} + +int kvm_arch_set_memory_region(struct kvm *kvm, + struct kvm_userspace_memory_region *mem, + struct kvm_memory_slot old, + int user_alloc) +{ + return 0; +} + +struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id) +{ + struct kvm_vcpu *vcpu; + int err; + + vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL); + if (!vcpu) { + err = -ENOMEM; + goto out; + } + + err = kvm_vcpu_init(vcpu, kvm, id); + if (err) + goto free_vcpu; + + return vcpu; + +free_vcpu: + kmem_cache_free(kvm_vcpu_cache, vcpu); +out: + return ERR_PTR(err); +} + +void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu) +{ + kvm_vcpu_uninit(vcpu); + kmem_cache_free(kvm_vcpu_cache, vcpu); +} + +void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) +{ + kvm_arch_vcpu_free(vcpu); +} + +int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu) +{ + unsigned int priority = exception_priority[BOOKE_INTERRUPT_DECREMENTER]; + + return test_bit(priority, &vcpu->arch.pending_exceptions); +} + +static void kvmppc_decrementer_func(unsigned long data) +{ + struct kvm_vcpu *vcpu = (struct kvm_vcpu *)data; + + kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_DECREMENTER); +} + +int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) +{ + setup_timer(&vcpu->arch.dec_timer, kvmppc_decrementer_func, + (unsigned long)vcpu); + + return 0; +} + +void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) +{ +} + +void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) +{ +} + +void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) +{ +} + +void decache_vcpus_on_cpu(int cpu) +{ +} + +int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu, + struct kvm_debug_guest *dbg) +{ + return -ENOTSUPP; +} + +static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu, + struct kvm_run *run) +{ + u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr]; + *gpr = run->dcr.data; +} + +static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu, + struct kvm_run *run) +{ + u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr]; + + if (run->mmio.len > sizeof(*gpr)) { + printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len); + return; + } + + if (vcpu->arch.mmio_is_bigendian) { + switch (run->mmio.len) { + case 4: *gpr = *(u32 *)run->mmio.data; break; + case 2: *gpr = *(u16 *)run->mmio.data; break; + case 1: *gpr = *(u8 *)run->mmio.data; break; + } + } else { + /* Convert BE data from userland back to LE. */ + switch (run->mmio.len) { + case 4: *gpr = ld_le32((u32 *)run->mmio.data); break; + case 2: *gpr = ld_le16((u16 *)run->mmio.data); break; + case 1: *gpr = *(u8 *)run->mmio.data; break; + } + } +} + +int kvmppc_handle_load(struct kvm_run *run, struct kvm_vcpu *vcpu, + unsigned int rt, unsigned int bytes, int is_bigendian) +{ + if (bytes > sizeof(run->mmio.data)) { + printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__, + run->mmio.len); + } + + run->mmio.phys_addr = vcpu->arch.paddr_accessed; + run->mmio.len = bytes; + run->mmio.is_write = 0; + + vcpu->arch.io_gpr = rt; + vcpu->arch.mmio_is_bigendian = is_bigendian; + vcpu->mmio_needed = 1; + vcpu->mmio_is_write = 0; + + return EMULATE_DO_MMIO; +} + +int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu, + u32 val, unsigned int bytes, int is_bigendian) +{ + void *data = run->mmio.data; + + if (bytes > sizeof(run->mmio.data)) { + printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__, + run->mmio.len); + } + + run->mmio.phys_addr = vcpu->arch.paddr_accessed; + run->mmio.len = bytes; + run->mmio.is_write = 1; + vcpu->mmio_needed = 1; + vcpu->mmio_is_write = 1; + + /* Store the value at the lowest bytes in 'data'. */ + if (is_bigendian) { + switch (bytes) { + case 4: *(u32 *)data = val; break; + case 2: *(u16 *)data = val; break; + case 1: *(u8 *)data = val; break; + } + } else { + /* Store LE value into 'data'. */ + switch (bytes) { + case 4: st_le32(data, val); break; + case 2: st_le16(data, val); break; + case 1: *(u8 *)data = val; break; + } + } + + return EMULATE_DO_MMIO; +} + +int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) +{ + int r; + sigset_t sigsaved; + + if (vcpu->sigset_active) + sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved); + + if (vcpu->mmio_needed) { + if (!vcpu->mmio_is_write) + kvmppc_complete_mmio_load(vcpu, run); + vcpu->mmio_needed = 0; + } else if (vcpu->arch.dcr_needed) { + if (!vcpu->arch.dcr_is_write) + kvmppc_complete_dcr_load(vcpu, run); + vcpu->arch.dcr_needed = 0; + } + + kvmppc_check_and_deliver_interrupts(vcpu); + + local_irq_disable(); + kvm_guest_enter(); + r = __kvmppc_vcpu_run(run, vcpu); + kvm_guest_exit(); + local_irq_enable(); + + if (vcpu->sigset_active) + sigprocmask(SIG_SETMASK, &sigsaved, NULL); + + return r; +} + +int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq) +{ + kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_EXTERNAL); + return 0; +} + +int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu, + struct kvm_mp_state *mp_state) +{ + return -EINVAL; +} + +int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu, + struct kvm_mp_state *mp_state) +{ + return -EINVAL; +} + +long kvm_arch_vcpu_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + struct kvm_vcpu *vcpu = filp->private_data; + void __user *argp = (void __user *)arg; + long r; + + switch (ioctl) { + case KVM_INTERRUPT: { + struct kvm_interrupt irq; + r = -EFAULT; + if (copy_from_user(&irq, argp, sizeof(irq))) + goto out; + r = kvm_vcpu_ioctl_interrupt(vcpu, &irq); + break; + } + default: + r = -EINVAL; + } + +out: + return r; +} + +int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm, struct kvm_dirty_log *log) +{ + return -ENOTSUPP; +} + +long kvm_arch_vm_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + long r; + + switch (ioctl) { + default: + r = -EINVAL; + } + + return r; +} + +int kvm_arch_init(void *opaque) +{ + return 0; +} + +void kvm_arch_exit(void) +{ +} diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index f6a68e178fc52952f31e4e95e7123af44252a919..8f5f02160ffcef2d57c458790537456ee3c06626 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -62,6 +62,10 @@ config GENERIC_LOCKBREAK default y depends on SMP && PREEMPT +config PGSTE + bool + default y if KVM + mainmenu "Linux Kernel Configuration" config S390 @@ -69,6 +73,7 @@ config S390 select HAVE_OPROFILE select HAVE_KPROBES select HAVE_KRETPROBES + select HAVE_KVM if 64BIT source "init/Kconfig" @@ -515,6 +520,13 @@ config ZFCPDUMP Select this option if you want to build an zfcpdump enabled kernel. Refer to for more details on this. +config S390_GUEST +bool "s390 guest support (EXPERIMENTAL)" + depends on 64BIT && EXPERIMENTAL + select VIRTIO + select VIRTIO_RING + help + Select this option if you want to run the kernel under s390 linux endmenu source "net/Kconfig" @@ -536,3 +548,5 @@ source "security/Kconfig" source "crypto/Kconfig" source "lib/Kconfig" + +source "arch/s390/kvm/Kconfig" diff --git a/arch/s390/Makefile b/arch/s390/Makefile index f708be367b030b672ffe58684e20fbcebc90c77a..792a4e7743cee19ef0b41d4b896d104df48aa73e 100644 --- a/arch/s390/Makefile +++ b/arch/s390/Makefile @@ -87,7 +87,7 @@ LDFLAGS_vmlinux := -e start head-y := arch/s390/kernel/head.o arch/s390/kernel/init_task.o core-y += arch/s390/mm/ arch/s390/kernel/ arch/s390/crypto/ \ - arch/s390/appldata/ arch/s390/hypfs/ + arch/s390/appldata/ arch/s390/hypfs/ arch/s390/kvm/ libs-y += arch/s390/lib/ drivers-y += drivers/s390/ drivers-$(CONFIG_MATHEMU) += arch/s390/math-emu/ diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 540a67f979b64820e4b45319451b1e3897ebb0e0..68ec4083bf73910641d32334494c809318ae7c17 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -144,6 +144,10 @@ static noinline __init void detect_machine_type(void) /* Running on a P/390 ? */ if (cpuinfo->cpu_id.machine == 0x7490) machine_flags |= 4; + + /* Running under KVM ? */ + if (cpuinfo->cpu_id.version == 0xfe) + machine_flags |= 64; } #ifdef CONFIG_64BIT diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 7141147e6b639a2035e4542c167ac7d8a3b50c59..a9d18aafa5f47f610db88996985270c1bda30f35 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -316,7 +316,11 @@ static int __init early_parse_ipldelay(char *p) early_param("ipldelay", early_parse_ipldelay); #ifdef CONFIG_S390_SWITCH_AMODE +#ifdef CONFIG_PGSTE +unsigned int switch_amode = 1; +#else unsigned int switch_amode = 0; +#endif EXPORT_SYMBOL_GPL(switch_amode); static void set_amode_and_uaccess(unsigned long user_amode, @@ -797,9 +801,13 @@ setup_arch(char **cmdline_p) "This machine has an IEEE fpu\n" : "This machine has no IEEE fpu\n"); #else /* CONFIG_64BIT */ - printk((MACHINE_IS_VM) ? - "We are running under VM (64 bit mode)\n" : - "We are running native (64 bit mode)\n"); + if (MACHINE_IS_VM) + printk("We are running under VM (64 bit mode)\n"); + else if (MACHINE_IS_KVM) { + printk("We are running under KVM (64 bit mode)\n"); + add_preferred_console("ttyS", 1, NULL); + } else + printk("We are running native (64 bit mode)\n"); #endif /* CONFIG_64BIT */ /* Save unparsed command line copy for /proc/cmdline */ diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c index c5f05b3fb2c30f1548a062280be08fc12411340d..ca90ee3f930edc0ffc364d8a094826df731b3415 100644 --- a/arch/s390/kernel/vtime.c +++ b/arch/s390/kernel/vtime.c @@ -110,6 +110,7 @@ void account_system_vtime(struct task_struct *tsk) S390_lowcore.steal_clock -= cputime << 12; account_system_time(tsk, 0, cputime); } +EXPORT_SYMBOL_GPL(account_system_vtime); static inline void set_vtimer(__u64 expires) { diff --git a/arch/s390/kvm/Kconfig b/arch/s390/kvm/Kconfig new file mode 100644 index 0000000000000000000000000000000000000000..1761b74d639be88e3fbc870cda7a4942c7b50ac9 --- /dev/null +++ b/arch/s390/kvm/Kconfig @@ -0,0 +1,46 @@ +# +# KVM configuration +# +config HAVE_KVM + bool + +menuconfig VIRTUALIZATION + bool "Virtualization" + default y + ---help--- + Say Y here to get to see options for using your Linux host to run other + operating systems inside virtual machines (guests). + This option alone does not add any kernel code. + + If you say N, all options in this submenu will be skipped and disabled. + +if VIRTUALIZATION + +config KVM + tristate "Kernel-based Virtual Machine (KVM) support" + depends on HAVE_KVM && EXPERIMENTAL + select PREEMPT_NOTIFIERS + select ANON_INODES + select S390_SWITCH_AMODE + select PREEMPT + ---help--- + Support hosting paravirtualized guest machines using the SIE + virtualization capability on the mainframe. This should work + on any 64bit machine. + + This module provides access to the hardware capabilities through + a character device node named /dev/kvm. + + To compile this as a module, choose M here: the module + will be called kvm. + + If unsure, say N. + +config KVM_TRACE + bool + +# OK, it's a little counter-intuitive to do this, but it puts it neatly under +# the virtualization menu. +source drivers/virtio/Kconfig + +endif # VIRTUALIZATION diff --git a/arch/s390/kvm/Makefile b/arch/s390/kvm/Makefile new file mode 100644 index 0000000000000000000000000000000000000000..e5221ec0b8e3eeb5b008aefb891841e795d98221 --- /dev/null +++ b/arch/s390/kvm/Makefile @@ -0,0 +1,14 @@ +# Makefile for kernel virtual machines on s390 +# +# Copyright IBM Corp. 2008 +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License (version 2 only) +# as published by the Free Software Foundation. + +common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o) + +EXTRA_CFLAGS += -Ivirt/kvm -Iarch/s390/kvm + +kvm-objs := $(common-objs) kvm-s390.o sie64a.o intercept.o interrupt.o priv.o sigp.o diag.o +obj-$(CONFIG_KVM) += kvm.o diff --git a/arch/s390/kvm/diag.c b/arch/s390/kvm/diag.c new file mode 100644 index 0000000000000000000000000000000000000000..f639a152869fab72c1c2eff9fd2a2f4df2e65fc8 --- /dev/null +++ b/arch/s390/kvm/diag.c @@ -0,0 +1,67 @@ +/* + * diag.c - handling diagnose instructions + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + * Christian Borntraeger + */ + +#include +#include +#include "kvm-s390.h" + +static int __diag_time_slice_end(struct kvm_vcpu *vcpu) +{ + VCPU_EVENT(vcpu, 5, "%s", "diag time slice end"); + vcpu->stat.diagnose_44++; + vcpu_put(vcpu); + schedule(); + vcpu_load(vcpu); + return 0; +} + +static int __diag_ipl_functions(struct kvm_vcpu *vcpu) +{ + unsigned int reg = vcpu->arch.sie_block->ipa & 0xf; + unsigned long subcode = vcpu->arch.guest_gprs[reg] & 0xffff; + + VCPU_EVENT(vcpu, 5, "diag ipl functions, subcode %lx", subcode); + switch (subcode) { + case 3: + vcpu->run->s390_reset_flags = KVM_S390_RESET_CLEAR; + break; + case 4: + vcpu->run->s390_reset_flags = 0; + break; + default: + return -ENOTSUPP; + } + + atomic_clear_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); + vcpu->run->s390_reset_flags |= KVM_S390_RESET_SUBSYSTEM; + vcpu->run->s390_reset_flags |= KVM_S390_RESET_IPL; + vcpu->run->s390_reset_flags |= KVM_S390_RESET_CPU_INIT; + vcpu->run->exit_reason = KVM_EXIT_S390_RESET; + VCPU_EVENT(vcpu, 3, "requesting userspace resets %lx", + vcpu->run->s390_reset_flags); + return -EREMOTE; +} + +int kvm_s390_handle_diag(struct kvm_vcpu *vcpu) +{ + int code = (vcpu->arch.sie_block->ipb & 0xfff0000) >> 16; + + switch (code) { + case 0x44: + return __diag_time_slice_end(vcpu); + case 0x308: + return __diag_ipl_functions(vcpu); + default: + return -ENOTSUPP; + } +} diff --git a/arch/s390/kvm/gaccess.h b/arch/s390/kvm/gaccess.h new file mode 100644 index 0000000000000000000000000000000000000000..4e0633c413f3841dc1132c39d06dc5645b57f66d --- /dev/null +++ b/arch/s390/kvm/gaccess.h @@ -0,0 +1,274 @@ +/* + * gaccess.h - access guest memory + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + */ + +#ifndef __KVM_S390_GACCESS_H +#define __KVM_S390_GACCESS_H + +#include +#include +#include + +static inline void __user *__guestaddr_to_user(struct kvm_vcpu *vcpu, + u64 guestaddr) +{ + u64 prefix = vcpu->arch.sie_block->prefix; + u64 origin = vcpu->kvm->arch.guest_origin; + u64 memsize = vcpu->kvm->arch.guest_memsize; + + if (guestaddr < 2 * PAGE_SIZE) + guestaddr += prefix; + else if ((guestaddr >= prefix) && (guestaddr < prefix + 2 * PAGE_SIZE)) + guestaddr -= prefix; + + if (guestaddr > memsize) + return (void __user __force *) ERR_PTR(-EFAULT); + + guestaddr += origin; + + return (void __user *) guestaddr; +} + +static inline int get_guest_u64(struct kvm_vcpu *vcpu, u64 guestaddr, + u64 *result) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + BUG_ON(guestaddr & 7); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return get_user(*result, (u64 __user *) uptr); +} + +static inline int get_guest_u32(struct kvm_vcpu *vcpu, u64 guestaddr, + u32 *result) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + BUG_ON(guestaddr & 3); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return get_user(*result, (u32 __user *) uptr); +} + +static inline int get_guest_u16(struct kvm_vcpu *vcpu, u64 guestaddr, + u16 *result) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + BUG_ON(guestaddr & 1); + + if (IS_ERR(uptr)) + return PTR_ERR(uptr); + + return get_user(*result, (u16 __user *) uptr); +} + +static inline int get_guest_u8(struct kvm_vcpu *vcpu, u64 guestaddr, + u8 *result) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return get_user(*result, (u8 __user *) uptr); +} + +static inline int put_guest_u64(struct kvm_vcpu *vcpu, u64 guestaddr, + u64 value) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + BUG_ON(guestaddr & 7); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return put_user(value, (u64 __user *) uptr); +} + +static inline int put_guest_u32(struct kvm_vcpu *vcpu, u64 guestaddr, + u32 value) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + BUG_ON(guestaddr & 3); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return put_user(value, (u32 __user *) uptr); +} + +static inline int put_guest_u16(struct kvm_vcpu *vcpu, u64 guestaddr, + u16 value) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + BUG_ON(guestaddr & 1); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return put_user(value, (u16 __user *) uptr); +} + +static inline int put_guest_u8(struct kvm_vcpu *vcpu, u64 guestaddr, + u8 value) +{ + void __user *uptr = __guestaddr_to_user(vcpu, guestaddr); + + if (IS_ERR((void __force *) uptr)) + return PTR_ERR((void __force *) uptr); + + return put_user(value, (u8 __user *) uptr); +} + + +static inline int __copy_to_guest_slow(struct kvm_vcpu *vcpu, u64 guestdest, + const void *from, unsigned long n) +{ + int rc; + unsigned long i; + const u8 *data = from; + + for (i = 0; i < n; i++) { + rc = put_guest_u8(vcpu, guestdest++, *(data++)); + if (rc < 0) + return rc; + } + return 0; +} + +static inline int copy_to_guest(struct kvm_vcpu *vcpu, u64 guestdest, + const void *from, unsigned long n) +{ + u64 prefix = vcpu->arch.sie_block->prefix; + u64 origin = vcpu->kvm->arch.guest_origin; + u64 memsize = vcpu->kvm->arch.guest_memsize; + + if ((guestdest < 2 * PAGE_SIZE) && (guestdest + n > 2 * PAGE_SIZE)) + goto slowpath; + + if ((guestdest < prefix) && (guestdest + n > prefix)) + goto slowpath; + + if ((guestdest < prefix + 2 * PAGE_SIZE) + && (guestdest + n > prefix + 2 * PAGE_SIZE)) + goto slowpath; + + if (guestdest < 2 * PAGE_SIZE) + guestdest += prefix; + else if ((guestdest >= prefix) && (guestdest < prefix + 2 * PAGE_SIZE)) + guestdest -= prefix; + + if (guestdest + n > memsize) + return -EFAULT; + + if (guestdest + n < guestdest) + return -EFAULT; + + guestdest += origin; + + return copy_to_user((void __user *) guestdest, from, n); +slowpath: + return __copy_to_guest_slow(vcpu, guestdest, from, n); +} + +static inline int __copy_from_guest_slow(struct kvm_vcpu *vcpu, void *to, + u64 guestsrc, unsigned long n) +{ + int rc; + unsigned long i; + u8 *data = to; + + for (i = 0; i < n; i++) { + rc = get_guest_u8(vcpu, guestsrc++, data++); + if (rc < 0) + return rc; + } + return 0; +} + +static inline int copy_from_guest(struct kvm_vcpu *vcpu, void *to, + u64 guestsrc, unsigned long n) +{ + u64 prefix = vcpu->arch.sie_block->prefix; + u64 origin = vcpu->kvm->arch.guest_origin; + u64 memsize = vcpu->kvm->arch.guest_memsize; + + if ((guestsrc < 2 * PAGE_SIZE) && (guestsrc + n > 2 * PAGE_SIZE)) + goto slowpath; + + if ((guestsrc < prefix) && (guestsrc + n > prefix)) + goto slowpath; + + if ((guestsrc < prefix + 2 * PAGE_SIZE) + && (guestsrc + n > prefix + 2 * PAGE_SIZE)) + goto slowpath; + + if (guestsrc < 2 * PAGE_SIZE) + guestsrc += prefix; + else if ((guestsrc >= prefix) && (guestsrc < prefix + 2 * PAGE_SIZE)) + guestsrc -= prefix; + + if (guestsrc + n > memsize) + return -EFAULT; + + if (guestsrc + n < guestsrc) + return -EFAULT; + + guestsrc += origin; + + return copy_from_user(to, (void __user *) guestsrc, n); +slowpath: + return __copy_from_guest_slow(vcpu, to, guestsrc, n); +} + +static inline int copy_to_guest_absolute(struct kvm_vcpu *vcpu, u64 guestdest, + const void *from, unsigned long n) +{ + u64 origin = vcpu->kvm->arch.guest_origin; + u64 memsize = vcpu->kvm->arch.guest_memsize; + + if (guestdest + n > memsize) + return -EFAULT; + + if (guestdest + n < guestdest) + return -EFAULT; + + guestdest += origin; + + return copy_to_user((void __user *) guestdest, from, n); +} + +static inline int copy_from_guest_absolute(struct kvm_vcpu *vcpu, void *to, + u64 guestsrc, unsigned long n) +{ + u64 origin = vcpu->kvm->arch.guest_origin; + u64 memsize = vcpu->kvm->arch.guest_memsize; + + if (guestsrc + n > memsize) + return -EFAULT; + + if (guestsrc + n < guestsrc) + return -EFAULT; + + guestsrc += origin; + + return copy_from_user(to, (void __user *) guestsrc, n); +} +#endif diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c new file mode 100644 index 0000000000000000000000000000000000000000..349581a2610344f23c1af6873dc57d6c23761b47 --- /dev/null +++ b/arch/s390/kvm/intercept.c @@ -0,0 +1,216 @@ +/* + * intercept.c - in-kernel handling for sie intercepts + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + * Christian Borntraeger + */ + +#include +#include +#include + +#include + +#include "kvm-s390.h" +#include "gaccess.h" + +static int handle_lctg(struct kvm_vcpu *vcpu) +{ + int reg1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4; + int reg3 = vcpu->arch.sie_block->ipa & 0x000f; + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16) + + ((vcpu->arch.sie_block->ipb & 0xff00) << 4); + u64 useraddr; + int reg, rc; + + vcpu->stat.instruction_lctg++; + if ((vcpu->arch.sie_block->ipb & 0xff) != 0x2f) + return -ENOTSUPP; + + useraddr = disp2; + if (base2) + useraddr += vcpu->arch.guest_gprs[base2]; + + reg = reg1; + + VCPU_EVENT(vcpu, 5, "lctg r1:%x, r3:%x,b2:%x,d2:%x", reg1, reg3, base2, + disp2); + + do { + rc = get_guest_u64(vcpu, useraddr, + &vcpu->arch.sie_block->gcr[reg]); + if (rc == -EFAULT) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + break; + } + useraddr += 8; + if (reg == reg3) + break; + reg = (reg + 1) % 16; + } while (1); + return 0; +} + +static int handle_lctl(struct kvm_vcpu *vcpu) +{ + int reg1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4; + int reg3 = vcpu->arch.sie_block->ipa & 0x000f; + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u64 useraddr; + u32 val = 0; + int reg, rc; + + vcpu->stat.instruction_lctl++; + + useraddr = disp2; + if (base2) + useraddr += vcpu->arch.guest_gprs[base2]; + + VCPU_EVENT(vcpu, 5, "lctl r1:%x, r3:%x,b2:%x,d2:%x", reg1, reg3, base2, + disp2); + + reg = reg1; + do { + rc = get_guest_u32(vcpu, useraddr, &val); + if (rc == -EFAULT) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + break; + } + vcpu->arch.sie_block->gcr[reg] &= 0xffffffff00000000ul; + vcpu->arch.sie_block->gcr[reg] |= val; + useraddr += 4; + if (reg == reg3) + break; + reg = (reg + 1) % 16; + } while (1); + return 0; +} + +static intercept_handler_t instruction_handlers[256] = { + [0x83] = kvm_s390_handle_diag, + [0xae] = kvm_s390_handle_sigp, + [0xb2] = kvm_s390_handle_priv, + [0xb7] = handle_lctl, + [0xeb] = handle_lctg, +}; + +static int handle_noop(struct kvm_vcpu *vcpu) +{ + switch (vcpu->arch.sie_block->icptcode) { + case 0x10: + vcpu->stat.exit_external_request++; + break; + case 0x14: + vcpu->stat.exit_external_interrupt++; + break; + default: + break; /* nothing */ + } + return 0; +} + +static int handle_stop(struct kvm_vcpu *vcpu) +{ + int rc; + + vcpu->stat.exit_stop_request++; + atomic_clear_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); + spin_lock_bh(&vcpu->arch.local_int.lock); + if (vcpu->arch.local_int.action_bits & ACTION_STORE_ON_STOP) { + vcpu->arch.local_int.action_bits &= ~ACTION_STORE_ON_STOP; + rc = __kvm_s390_vcpu_store_status(vcpu, + KVM_S390_STORE_STATUS_NOADDR); + if (rc >= 0) + rc = -ENOTSUPP; + } + + if (vcpu->arch.local_int.action_bits & ACTION_STOP_ON_STOP) { + vcpu->arch.local_int.action_bits &= ~ACTION_STOP_ON_STOP; + VCPU_EVENT(vcpu, 3, "%s", "cpu stopped"); + rc = -ENOTSUPP; + } else + rc = 0; + spin_unlock_bh(&vcpu->arch.local_int.lock); + return rc; +} + +static int handle_validity(struct kvm_vcpu *vcpu) +{ + int viwhy = vcpu->arch.sie_block->ipb >> 16; + vcpu->stat.exit_validity++; + if (viwhy == 0x37) { + fault_in_pages_writeable((char __user *) + vcpu->kvm->arch.guest_origin + + vcpu->arch.sie_block->prefix, + PAGE_SIZE); + return 0; + } + VCPU_EVENT(vcpu, 2, "unhandled validity intercept code %d", + viwhy); + return -ENOTSUPP; +} + +static int handle_instruction(struct kvm_vcpu *vcpu) +{ + intercept_handler_t handler; + + vcpu->stat.exit_instruction++; + handler = instruction_handlers[vcpu->arch.sie_block->ipa >> 8]; + if (handler) + return handler(vcpu); + return -ENOTSUPP; +} + +static int handle_prog(struct kvm_vcpu *vcpu) +{ + vcpu->stat.exit_program_interruption++; + return kvm_s390_inject_program_int(vcpu, vcpu->arch.sie_block->iprcc); +} + +static int handle_instruction_and_prog(struct kvm_vcpu *vcpu) +{ + int rc, rc2; + + vcpu->stat.exit_instr_and_program++; + rc = handle_instruction(vcpu); + rc2 = handle_prog(vcpu); + + if (rc == -ENOTSUPP) + vcpu->arch.sie_block->icptcode = 0x04; + if (rc) + return rc; + return rc2; +} + +static const intercept_handler_t intercept_funcs[0x48 >> 2] = { + [0x00 >> 2] = handle_noop, + [0x04 >> 2] = handle_instruction, + [0x08 >> 2] = handle_prog, + [0x0C >> 2] = handle_instruction_and_prog, + [0x10 >> 2] = handle_noop, + [0x14 >> 2] = handle_noop, + [0x1C >> 2] = kvm_s390_handle_wait, + [0x20 >> 2] = handle_validity, + [0x28 >> 2] = handle_stop, +}; + +int kvm_handle_sie_intercept(struct kvm_vcpu *vcpu) +{ + intercept_handler_t func; + u8 code = vcpu->arch.sie_block->icptcode; + + if (code & 3 || code > 0x48) + return -ENOTSUPP; + func = intercept_funcs[code >> 2]; + if (func) + return func(vcpu); + return -ENOTSUPP; +} diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c new file mode 100644 index 0000000000000000000000000000000000000000..fcd1ed8015c1c2d3b4f4dbdca4f202ffc2dd281a --- /dev/null +++ b/arch/s390/kvm/interrupt.c @@ -0,0 +1,592 @@ +/* + * interrupt.c - handling kvm guest interrupts + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + */ + +#include +#include +#include +#include "kvm-s390.h" +#include "gaccess.h" + +static int psw_extint_disabled(struct kvm_vcpu *vcpu) +{ + return !(vcpu->arch.sie_block->gpsw.mask & PSW_MASK_EXT); +} + +static int psw_interrupts_disabled(struct kvm_vcpu *vcpu) +{ + if ((vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PER) || + (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_IO) || + (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_EXT)) + return 0; + return 1; +} + +static int __interrupt_is_deliverable(struct kvm_vcpu *vcpu, + struct interrupt_info *inti) +{ + switch (inti->type) { + case KVM_S390_INT_EMERGENCY: + if (psw_extint_disabled(vcpu)) + return 0; + if (vcpu->arch.sie_block->gcr[0] & 0x4000ul) + return 1; + return 0; + case KVM_S390_INT_SERVICE: + if (psw_extint_disabled(vcpu)) + return 0; + if (vcpu->arch.sie_block->gcr[0] & 0x200ul) + return 1; + return 0; + case KVM_S390_INT_VIRTIO: + if (psw_extint_disabled(vcpu)) + return 0; + if (vcpu->arch.sie_block->gcr[0] & 0x200ul) + return 1; + return 0; + case KVM_S390_PROGRAM_INT: + case KVM_S390_SIGP_STOP: + case KVM_S390_SIGP_SET_PREFIX: + case KVM_S390_RESTART: + return 1; + default: + BUG(); + } + return 0; +} + +static void __set_cpu_idle(struct kvm_vcpu *vcpu) +{ + BUG_ON(vcpu->vcpu_id > KVM_MAX_VCPUS - 1); + atomic_set_mask(CPUSTAT_WAIT, &vcpu->arch.sie_block->cpuflags); + set_bit(vcpu->vcpu_id, vcpu->arch.local_int.float_int->idle_mask); +} + +static void __unset_cpu_idle(struct kvm_vcpu *vcpu) +{ + BUG_ON(vcpu->vcpu_id > KVM_MAX_VCPUS - 1); + atomic_clear_mask(CPUSTAT_WAIT, &vcpu->arch.sie_block->cpuflags); + clear_bit(vcpu->vcpu_id, vcpu->arch.local_int.float_int->idle_mask); +} + +static void __reset_intercept_indicators(struct kvm_vcpu *vcpu) +{ + atomic_clear_mask(CPUSTAT_ECALL_PEND | + CPUSTAT_IO_INT | CPUSTAT_EXT_INT | CPUSTAT_STOP_INT, + &vcpu->arch.sie_block->cpuflags); + vcpu->arch.sie_block->lctl = 0x0000; +} + +static void __set_cpuflag(struct kvm_vcpu *vcpu, u32 flag) +{ + atomic_set_mask(flag, &vcpu->arch.sie_block->cpuflags); +} + +static void __set_intercept_indicator(struct kvm_vcpu *vcpu, + struct interrupt_info *inti) +{ + switch (inti->type) { + case KVM_S390_INT_EMERGENCY: + case KVM_S390_INT_SERVICE: + case KVM_S390_INT_VIRTIO: + if (psw_extint_disabled(vcpu)) + __set_cpuflag(vcpu, CPUSTAT_EXT_INT); + else + vcpu->arch.sie_block->lctl |= LCTL_CR0; + break; + case KVM_S390_SIGP_STOP: + __set_cpuflag(vcpu, CPUSTAT_STOP_INT); + break; + default: + BUG(); + } +} + +static void __do_deliver_interrupt(struct kvm_vcpu *vcpu, + struct interrupt_info *inti) +{ + const unsigned short table[] = { 2, 4, 4, 6 }; + int rc, exception = 0; + + switch (inti->type) { + case KVM_S390_INT_EMERGENCY: + VCPU_EVENT(vcpu, 4, "%s", "interrupt: sigp emerg"); + vcpu->stat.deliver_emergency_signal++; + rc = put_guest_u16(vcpu, __LC_EXT_INT_CODE, 0x1201); + if (rc == -EFAULT) + exception = 1; + + rc = copy_to_guest(vcpu, __LC_EXT_OLD_PSW, + &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = copy_from_guest(vcpu, &vcpu->arch.sie_block->gpsw, + __LC_EXT_NEW_PSW, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + break; + + case KVM_S390_INT_SERVICE: + VCPU_EVENT(vcpu, 4, "interrupt: sclp parm:%x", + inti->ext.ext_params); + vcpu->stat.deliver_service_signal++; + rc = put_guest_u16(vcpu, __LC_EXT_INT_CODE, 0x2401); + if (rc == -EFAULT) + exception = 1; + + rc = copy_to_guest(vcpu, __LC_EXT_OLD_PSW, + &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = copy_from_guest(vcpu, &vcpu->arch.sie_block->gpsw, + __LC_EXT_NEW_PSW, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = put_guest_u32(vcpu, __LC_EXT_PARAMS, inti->ext.ext_params); + if (rc == -EFAULT) + exception = 1; + break; + + case KVM_S390_INT_VIRTIO: + VCPU_EVENT(vcpu, 4, "interrupt: virtio parm:%x,parm64:%lx", + inti->ext.ext_params, inti->ext.ext_params2); + vcpu->stat.deliver_virtio_interrupt++; + rc = put_guest_u16(vcpu, __LC_EXT_INT_CODE, 0x2603); + if (rc == -EFAULT) + exception = 1; + + rc = put_guest_u16(vcpu, __LC_CPU_ADDRESS, 0x0d00); + if (rc == -EFAULT) + exception = 1; + + rc = copy_to_guest(vcpu, __LC_EXT_OLD_PSW, + &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = copy_from_guest(vcpu, &vcpu->arch.sie_block->gpsw, + __LC_EXT_NEW_PSW, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = put_guest_u32(vcpu, __LC_EXT_PARAMS, inti->ext.ext_params); + if (rc == -EFAULT) + exception = 1; + + rc = put_guest_u64(vcpu, __LC_PFAULT_INTPARM, + inti->ext.ext_params2); + if (rc == -EFAULT) + exception = 1; + break; + + case KVM_S390_SIGP_STOP: + VCPU_EVENT(vcpu, 4, "%s", "interrupt: cpu stop"); + vcpu->stat.deliver_stop_signal++; + __set_intercept_indicator(vcpu, inti); + break; + + case KVM_S390_SIGP_SET_PREFIX: + VCPU_EVENT(vcpu, 4, "interrupt: set prefix to %x", + inti->prefix.address); + vcpu->stat.deliver_prefix_signal++; + vcpu->arch.sie_block->prefix = inti->prefix.address; + vcpu->arch.sie_block->ihcpu = 0xffff; + break; + + case KVM_S390_RESTART: + VCPU_EVENT(vcpu, 4, "%s", "interrupt: cpu restart"); + vcpu->stat.deliver_restart_signal++; + rc = copy_to_guest(vcpu, offsetof(struct _lowcore, + restart_old_psw), &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = copy_from_guest(vcpu, &vcpu->arch.sie_block->gpsw, + offsetof(struct _lowcore, restart_psw), sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + break; + + case KVM_S390_PROGRAM_INT: + VCPU_EVENT(vcpu, 4, "interrupt: pgm check code:%x, ilc:%x", + inti->pgm.code, + table[vcpu->arch.sie_block->ipa >> 14]); + vcpu->stat.deliver_program_int++; + rc = put_guest_u16(vcpu, __LC_PGM_INT_CODE, inti->pgm.code); + if (rc == -EFAULT) + exception = 1; + + rc = put_guest_u16(vcpu, __LC_PGM_ILC, + table[vcpu->arch.sie_block->ipa >> 14]); + if (rc == -EFAULT) + exception = 1; + + rc = copy_to_guest(vcpu, __LC_PGM_OLD_PSW, + &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + rc = copy_from_guest(vcpu, &vcpu->arch.sie_block->gpsw, + __LC_PGM_NEW_PSW, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + break; + + default: + BUG(); + } + + if (exception) { + VCPU_EVENT(vcpu, 1, "%s", "program exception while delivering" + " interrupt"); + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + if (inti->type == KVM_S390_PROGRAM_INT) { + printk(KERN_WARNING "kvm: recursive program check\n"); + BUG(); + } + } +} + +static int __try_deliver_ckc_interrupt(struct kvm_vcpu *vcpu) +{ + int rc, exception = 0; + + if (psw_extint_disabled(vcpu)) + return 0; + if (!(vcpu->arch.sie_block->gcr[0] & 0x800ul)) + return 0; + rc = put_guest_u16(vcpu, __LC_EXT_INT_CODE, 0x1004); + if (rc == -EFAULT) + exception = 1; + rc = copy_to_guest(vcpu, __LC_EXT_OLD_PSW, + &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + rc = copy_from_guest(vcpu, &vcpu->arch.sie_block->gpsw, + __LC_EXT_NEW_PSW, sizeof(psw_t)); + if (rc == -EFAULT) + exception = 1; + + if (exception) { + VCPU_EVENT(vcpu, 1, "%s", "program exception while delivering" \ + " ckc interrupt"); + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + return 0; + } + + return 1; +} + +int kvm_cpu_has_interrupt(struct kvm_vcpu *vcpu) +{ + struct local_interrupt *li = &vcpu->arch.local_int; + struct float_interrupt *fi = vcpu->arch.local_int.float_int; + struct interrupt_info *inti; + int rc = 0; + + if (atomic_read(&li->active)) { + spin_lock_bh(&li->lock); + list_for_each_entry(inti, &li->list, list) + if (__interrupt_is_deliverable(vcpu, inti)) { + rc = 1; + break; + } + spin_unlock_bh(&li->lock); + } + + if ((!rc) && atomic_read(&fi->active)) { + spin_lock_bh(&fi->lock); + list_for_each_entry(inti, &fi->list, list) + if (__interrupt_is_deliverable(vcpu, inti)) { + rc = 1; + break; + } + spin_unlock_bh(&fi->lock); + } + + if ((!rc) && (vcpu->arch.sie_block->ckc < + get_clock() + vcpu->arch.sie_block->epoch)) { + if ((!psw_extint_disabled(vcpu)) && + (vcpu->arch.sie_block->gcr[0] & 0x800ul)) + rc = 1; + } + + return rc; +} + +int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu) +{ + return 0; +} + +int kvm_s390_handle_wait(struct kvm_vcpu *vcpu) +{ + u64 now, sltime; + DECLARE_WAITQUEUE(wait, current); + + vcpu->stat.exit_wait_state++; + if (kvm_cpu_has_interrupt(vcpu)) + return 0; + + if (psw_interrupts_disabled(vcpu)) { + VCPU_EVENT(vcpu, 3, "%s", "disabled wait"); + __unset_cpu_idle(vcpu); + return -ENOTSUPP; /* disabled wait */ + } + + if (psw_extint_disabled(vcpu) || + (!(vcpu->arch.sie_block->gcr[0] & 0x800ul))) { + VCPU_EVENT(vcpu, 3, "%s", "enabled wait w/o timer"); + goto no_timer; + } + + now = get_clock() + vcpu->arch.sie_block->epoch; + if (vcpu->arch.sie_block->ckc < now) { + __unset_cpu_idle(vcpu); + return 0; + } + + sltime = (vcpu->arch.sie_block->ckc - now) / (0xf4240000ul / HZ) + 1; + + vcpu->arch.ckc_timer.expires = jiffies + sltime; + + add_timer(&vcpu->arch.ckc_timer); + VCPU_EVENT(vcpu, 5, "enabled wait timer:%lx jiffies", sltime); +no_timer: + spin_lock_bh(&vcpu->arch.local_int.float_int->lock); + spin_lock_bh(&vcpu->arch.local_int.lock); + __set_cpu_idle(vcpu); + vcpu->arch.local_int.timer_due = 0; + add_wait_queue(&vcpu->arch.local_int.wq, &wait); + while (list_empty(&vcpu->arch.local_int.list) && + list_empty(&vcpu->arch.local_int.float_int->list) && + (!vcpu->arch.local_int.timer_due) && + !signal_pending(current)) { + set_current_state(TASK_INTERRUPTIBLE); + spin_unlock_bh(&vcpu->arch.local_int.lock); + spin_unlock_bh(&vcpu->arch.local_int.float_int->lock); + vcpu_put(vcpu); + schedule(); + vcpu_load(vcpu); + spin_lock_bh(&vcpu->arch.local_int.float_int->lock); + spin_lock_bh(&vcpu->arch.local_int.lock); + } + __unset_cpu_idle(vcpu); + __set_current_state(TASK_RUNNING); + remove_wait_queue(&vcpu->wq, &wait); + spin_unlock_bh(&vcpu->arch.local_int.lock); + spin_unlock_bh(&vcpu->arch.local_int.float_int->lock); + del_timer(&vcpu->arch.ckc_timer); + return 0; +} + +void kvm_s390_idle_wakeup(unsigned long data) +{ + struct kvm_vcpu *vcpu = (struct kvm_vcpu *)data; + + spin_lock_bh(&vcpu->arch.local_int.lock); + vcpu->arch.local_int.timer_due = 1; + if (waitqueue_active(&vcpu->arch.local_int.wq)) + wake_up_interruptible(&vcpu->arch.local_int.wq); + spin_unlock_bh(&vcpu->arch.local_int.lock); +} + + +void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu) +{ + struct local_interrupt *li = &vcpu->arch.local_int; + struct float_interrupt *fi = vcpu->arch.local_int.float_int; + struct interrupt_info *n, *inti = NULL; + int deliver; + + __reset_intercept_indicators(vcpu); + if (atomic_read(&li->active)) { + do { + deliver = 0; + spin_lock_bh(&li->lock); + list_for_each_entry_safe(inti, n, &li->list, list) { + if (__interrupt_is_deliverable(vcpu, inti)) { + list_del(&inti->list); + deliver = 1; + break; + } + __set_intercept_indicator(vcpu, inti); + } + if (list_empty(&li->list)) + atomic_set(&li->active, 0); + spin_unlock_bh(&li->lock); + if (deliver) { + __do_deliver_interrupt(vcpu, inti); + kfree(inti); + } + } while (deliver); + } + + if ((vcpu->arch.sie_block->ckc < + get_clock() + vcpu->arch.sie_block->epoch)) + __try_deliver_ckc_interrupt(vcpu); + + if (atomic_read(&fi->active)) { + do { + deliver = 0; + spin_lock_bh(&fi->lock); + list_for_each_entry_safe(inti, n, &fi->list, list) { + if (__interrupt_is_deliverable(vcpu, inti)) { + list_del(&inti->list); + deliver = 1; + break; + } + __set_intercept_indicator(vcpu, inti); + } + if (list_empty(&fi->list)) + atomic_set(&fi->active, 0); + spin_unlock_bh(&fi->lock); + if (deliver) { + __do_deliver_interrupt(vcpu, inti); + kfree(inti); + } + } while (deliver); + } +} + +int kvm_s390_inject_program_int(struct kvm_vcpu *vcpu, u16 code) +{ + struct local_interrupt *li = &vcpu->arch.local_int; + struct interrupt_info *inti; + + inti = kzalloc(sizeof(*inti), GFP_KERNEL); + if (!inti) + return -ENOMEM; + + inti->type = KVM_S390_PROGRAM_INT;; + inti->pgm.code = code; + + VCPU_EVENT(vcpu, 3, "inject: program check %d (from kernel)", code); + spin_lock_bh(&li->lock); + list_add(&inti->list, &li->list); + atomic_set(&li->active, 1); + BUG_ON(waitqueue_active(&li->wq)); + spin_unlock_bh(&li->lock); + return 0; +} + +int kvm_s390_inject_vm(struct kvm *kvm, + struct kvm_s390_interrupt *s390int) +{ + struct local_interrupt *li; + struct float_interrupt *fi; + struct interrupt_info *inti; + int sigcpu; + + inti = kzalloc(sizeof(*inti), GFP_KERNEL); + if (!inti) + return -ENOMEM; + + switch (s390int->type) { + case KVM_S390_INT_VIRTIO: + VM_EVENT(kvm, 5, "inject: virtio parm:%x,parm64:%lx", + s390int->parm, s390int->parm64); + inti->type = s390int->type; + inti->ext.ext_params = s390int->parm; + inti->ext.ext_params2 = s390int->parm64; + break; + case KVM_S390_INT_SERVICE: + VM_EVENT(kvm, 5, "inject: sclp parm:%x", s390int->parm); + inti->type = s390int->type; + inti->ext.ext_params = s390int->parm; + break; + case KVM_S390_PROGRAM_INT: + case KVM_S390_SIGP_STOP: + case KVM_S390_INT_EMERGENCY: + default: + kfree(inti); + return -EINVAL; + } + + mutex_lock(&kvm->lock); + fi = &kvm->arch.float_int; + spin_lock_bh(&fi->lock); + list_add_tail(&inti->list, &fi->list); + atomic_set(&fi->active, 1); + sigcpu = find_first_bit(fi->idle_mask, KVM_MAX_VCPUS); + if (sigcpu == KVM_MAX_VCPUS) { + do { + sigcpu = fi->next_rr_cpu++; + if (sigcpu == KVM_MAX_VCPUS) + sigcpu = fi->next_rr_cpu = 0; + } while (fi->local_int[sigcpu] == NULL); + } + li = fi->local_int[sigcpu]; + spin_lock_bh(&li->lock); + atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags); + if (waitqueue_active(&li->wq)) + wake_up_interruptible(&li->wq); + spin_unlock_bh(&li->lock); + spin_unlock_bh(&fi->lock); + mutex_unlock(&kvm->lock); + return 0; +} + +int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu, + struct kvm_s390_interrupt *s390int) +{ + struct local_interrupt *li; + struct interrupt_info *inti; + + inti = kzalloc(sizeof(*inti), GFP_KERNEL); + if (!inti) + return -ENOMEM; + + switch (s390int->type) { + case KVM_S390_PROGRAM_INT: + if (s390int->parm & 0xffff0000) { + kfree(inti); + return -EINVAL; + } + inti->type = s390int->type; + inti->pgm.code = s390int->parm; + VCPU_EVENT(vcpu, 3, "inject: program check %d (from user)", + s390int->parm); + break; + case KVM_S390_SIGP_STOP: + case KVM_S390_RESTART: + case KVM_S390_SIGP_SET_PREFIX: + case KVM_S390_INT_EMERGENCY: + VCPU_EVENT(vcpu, 3, "inject: type %x", s390int->type); + inti->type = s390int->type; + break; + case KVM_S390_INT_VIRTIO: + case KVM_S390_INT_SERVICE: + default: + kfree(inti); + return -EINVAL; + } + + mutex_lock(&vcpu->kvm->lock); + li = &vcpu->arch.local_int; + spin_lock_bh(&li->lock); + if (inti->type == KVM_S390_PROGRAM_INT) + list_add(&inti->list, &li->list); + else + list_add_tail(&inti->list, &li->list); + atomic_set(&li->active, 1); + if (inti->type == KVM_S390_SIGP_STOP) + li->action_bits |= ACTION_STOP_ON_STOP; + atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags); + if (waitqueue_active(&li->wq)) + wake_up_interruptible(&vcpu->arch.local_int.wq); + spin_unlock_bh(&li->lock); + mutex_unlock(&vcpu->kvm->lock); + return 0; +} diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c new file mode 100644 index 0000000000000000000000000000000000000000..98d1e73e01f11f31cdac4d693f54c1bfddedc871 --- /dev/null +++ b/arch/s390/kvm/kvm-s390.c @@ -0,0 +1,685 @@ +/* + * s390host.c -- hosting zSeries kernel virtual machines + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + * Christian Borntraeger + * Heiko Carstens + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "kvm-s390.h" +#include "gaccess.h" + +#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU + +struct kvm_stats_debugfs_item debugfs_entries[] = { + { "userspace_handled", VCPU_STAT(exit_userspace) }, + { "exit_validity", VCPU_STAT(exit_validity) }, + { "exit_stop_request", VCPU_STAT(exit_stop_request) }, + { "exit_external_request", VCPU_STAT(exit_external_request) }, + { "exit_external_interrupt", VCPU_STAT(exit_external_interrupt) }, + { "exit_instruction", VCPU_STAT(exit_instruction) }, + { "exit_program_interruption", VCPU_STAT(exit_program_interruption) }, + { "exit_instr_and_program_int", VCPU_STAT(exit_instr_and_program) }, + { "instruction_lctg", VCPU_STAT(instruction_lctg) }, + { "instruction_lctl", VCPU_STAT(instruction_lctl) }, + { "deliver_emergency_signal", VCPU_STAT(deliver_emergency_signal) }, + { "deliver_service_signal", VCPU_STAT(deliver_service_signal) }, + { "deliver_virtio_interrupt", VCPU_STAT(deliver_virtio_interrupt) }, + { "deliver_stop_signal", VCPU_STAT(deliver_stop_signal) }, + { "deliver_prefix_signal", VCPU_STAT(deliver_prefix_signal) }, + { "deliver_restart_signal", VCPU_STAT(deliver_restart_signal) }, + { "deliver_program_interruption", VCPU_STAT(deliver_program_int) }, + { "exit_wait_state", VCPU_STAT(exit_wait_state) }, + { "instruction_stidp", VCPU_STAT(instruction_stidp) }, + { "instruction_spx", VCPU_STAT(instruction_spx) }, + { "instruction_stpx", VCPU_STAT(instruction_stpx) }, + { "instruction_stap", VCPU_STAT(instruction_stap) }, + { "instruction_storage_key", VCPU_STAT(instruction_storage_key) }, + { "instruction_stsch", VCPU_STAT(instruction_stsch) }, + { "instruction_chsc", VCPU_STAT(instruction_chsc) }, + { "instruction_stsi", VCPU_STAT(instruction_stsi) }, + { "instruction_stfl", VCPU_STAT(instruction_stfl) }, + { "instruction_sigp_sense", VCPU_STAT(instruction_sigp_sense) }, + { "instruction_sigp_emergency", VCPU_STAT(instruction_sigp_emergency) }, + { "instruction_sigp_stop", VCPU_STAT(instruction_sigp_stop) }, + { "instruction_sigp_set_arch", VCPU_STAT(instruction_sigp_arch) }, + { "instruction_sigp_set_prefix", VCPU_STAT(instruction_sigp_prefix) }, + { "instruction_sigp_restart", VCPU_STAT(instruction_sigp_restart) }, + { "diagnose_44", VCPU_STAT(diagnose_44) }, + { NULL } +}; + + +/* Section: not file related */ +void kvm_arch_hardware_enable(void *garbage) +{ + /* every s390 is virtualization enabled ;-) */ +} + +void kvm_arch_hardware_disable(void *garbage) +{ +} + +void decache_vcpus_on_cpu(int cpu) +{ +} + +int kvm_arch_hardware_setup(void) +{ + return 0; +} + +void kvm_arch_hardware_unsetup(void) +{ +} + +void kvm_arch_check_processor_compat(void *rtn) +{ +} + +int kvm_arch_init(void *opaque) +{ + return 0; +} + +void kvm_arch_exit(void) +{ +} + +/* Section: device related */ +long kvm_arch_dev_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + if (ioctl == KVM_S390_ENABLE_SIE) + return s390_enable_sie(); + return -EINVAL; +} + +int kvm_dev_ioctl_check_extension(long ext) +{ + return 0; +} + +/* Section: vm related */ +/* + * Get (and clear) the dirty memory log for a memory slot. + */ +int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm, + struct kvm_dirty_log *log) +{ + return 0; +} + +long kvm_arch_vm_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + struct kvm *kvm = filp->private_data; + void __user *argp = (void __user *)arg; + int r; + + switch (ioctl) { + case KVM_S390_INTERRUPT: { + struct kvm_s390_interrupt s390int; + + r = -EFAULT; + if (copy_from_user(&s390int, argp, sizeof(s390int))) + break; + r = kvm_s390_inject_vm(kvm, &s390int); + break; + } + default: + r = -EINVAL; + } + + return r; +} + +struct kvm *kvm_arch_create_vm(void) +{ + struct kvm *kvm; + int rc; + char debug_name[16]; + + rc = s390_enable_sie(); + if (rc) + goto out_nokvm; + + rc = -ENOMEM; + kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL); + if (!kvm) + goto out_nokvm; + + kvm->arch.sca = (struct sca_block *) get_zeroed_page(GFP_KERNEL); + if (!kvm->arch.sca) + goto out_nosca; + + sprintf(debug_name, "kvm-%u", current->pid); + + kvm->arch.dbf = debug_register(debug_name, 8, 2, 8 * sizeof(long)); + if (!kvm->arch.dbf) + goto out_nodbf; + + spin_lock_init(&kvm->arch.float_int.lock); + INIT_LIST_HEAD(&kvm->arch.float_int.list); + + debug_register_view(kvm->arch.dbf, &debug_sprintf_view); + VM_EVENT(kvm, 3, "%s", "vm created"); + + try_module_get(THIS_MODULE); + + return kvm; +out_nodbf: + free_page((unsigned long)(kvm->arch.sca)); +out_nosca: + kfree(kvm); +out_nokvm: + return ERR_PTR(rc); +} + +void kvm_arch_destroy_vm(struct kvm *kvm) +{ + debug_unregister(kvm->arch.dbf); + free_page((unsigned long)(kvm->arch.sca)); + kfree(kvm); + module_put(THIS_MODULE); +} + +/* Section: vcpu related */ +int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) +{ + return 0; +} + +void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) +{ + /* kvm common code refers to this, but does'nt call it */ + BUG(); +} + +void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) +{ + save_fp_regs(&vcpu->arch.host_fpregs); + save_access_regs(vcpu->arch.host_acrs); + vcpu->arch.guest_fpregs.fpc &= FPC_VALID_MASK; + restore_fp_regs(&vcpu->arch.guest_fpregs); + restore_access_regs(vcpu->arch.guest_acrs); + + if (signal_pending(current)) + atomic_set_mask(CPUSTAT_STOP_INT, + &vcpu->arch.sie_block->cpuflags); +} + +void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) +{ + save_fp_regs(&vcpu->arch.guest_fpregs); + save_access_regs(vcpu->arch.guest_acrs); + restore_fp_regs(&vcpu->arch.host_fpregs); + restore_access_regs(vcpu->arch.host_acrs); +} + +static void kvm_s390_vcpu_initial_reset(struct kvm_vcpu *vcpu) +{ + /* this equals initial cpu reset in pop, but we don't switch to ESA */ + vcpu->arch.sie_block->gpsw.mask = 0UL; + vcpu->arch.sie_block->gpsw.addr = 0UL; + vcpu->arch.sie_block->prefix = 0UL; + vcpu->arch.sie_block->ihcpu = 0xffff; + vcpu->arch.sie_block->cputm = 0UL; + vcpu->arch.sie_block->ckc = 0UL; + vcpu->arch.sie_block->todpr = 0; + memset(vcpu->arch.sie_block->gcr, 0, 16 * sizeof(__u64)); + vcpu->arch.sie_block->gcr[0] = 0xE0UL; + vcpu->arch.sie_block->gcr[14] = 0xC2000000UL; + vcpu->arch.guest_fpregs.fpc = 0; + asm volatile("lfpc %0" : : "Q" (vcpu->arch.guest_fpregs.fpc)); + vcpu->arch.sie_block->gbea = 1; +} + +int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) +{ + atomic_set(&vcpu->arch.sie_block->cpuflags, CPUSTAT_ZARCH); + vcpu->arch.sie_block->gmslm = 0xffffffffffUL; + vcpu->arch.sie_block->gmsor = 0x000000000000; + vcpu->arch.sie_block->ecb = 2; + vcpu->arch.sie_block->eca = 0xC1002001U; + setup_timer(&vcpu->arch.ckc_timer, kvm_s390_idle_wakeup, + (unsigned long) vcpu); + get_cpu_id(&vcpu->arch.cpu_id); + vcpu->arch.cpu_id.version = 0xfe; + return 0; +} + +struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, + unsigned int id) +{ + struct kvm_vcpu *vcpu = kzalloc(sizeof(struct kvm_vcpu), GFP_KERNEL); + int rc = -ENOMEM; + + if (!vcpu) + goto out_nomem; + + vcpu->arch.sie_block = (struct sie_block *) get_zeroed_page(GFP_KERNEL); + + if (!vcpu->arch.sie_block) + goto out_free_cpu; + + vcpu->arch.sie_block->icpua = id; + BUG_ON(!kvm->arch.sca); + BUG_ON(kvm->arch.sca->cpu[id].sda); + kvm->arch.sca->cpu[id].sda = (__u64) vcpu->arch.sie_block; + vcpu->arch.sie_block->scaoh = (__u32)(((__u64)kvm->arch.sca) >> 32); + vcpu->arch.sie_block->scaol = (__u32)(__u64)kvm->arch.sca; + + spin_lock_init(&vcpu->arch.local_int.lock); + INIT_LIST_HEAD(&vcpu->arch.local_int.list); + vcpu->arch.local_int.float_int = &kvm->arch.float_int; + spin_lock_bh(&kvm->arch.float_int.lock); + kvm->arch.float_int.local_int[id] = &vcpu->arch.local_int; + init_waitqueue_head(&vcpu->arch.local_int.wq); + vcpu->arch.local_int.cpuflags = &vcpu->arch.sie_block->cpuflags; + spin_unlock_bh(&kvm->arch.float_int.lock); + + rc = kvm_vcpu_init(vcpu, kvm, id); + if (rc) + goto out_free_cpu; + VM_EVENT(kvm, 3, "create cpu %d at %p, sie block at %p", id, vcpu, + vcpu->arch.sie_block); + + try_module_get(THIS_MODULE); + + return vcpu; +out_free_cpu: + kfree(vcpu); +out_nomem: + return ERR_PTR(rc); +} + +void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) +{ + VCPU_EVENT(vcpu, 3, "%s", "destroy cpu"); + free_page((unsigned long)(vcpu->arch.sie_block)); + kfree(vcpu); + module_put(THIS_MODULE); +} + +int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu) +{ + /* kvm common code refers to this, but never calls it */ + BUG(); + return 0; +} + +static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu) +{ + vcpu_load(vcpu); + kvm_s390_vcpu_initial_reset(vcpu); + vcpu_put(vcpu); + return 0; +} + +int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +{ + vcpu_load(vcpu); + memcpy(&vcpu->arch.guest_gprs, ®s->gprs, sizeof(regs->gprs)); + vcpu_put(vcpu); + return 0; +} + +int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +{ + vcpu_load(vcpu); + memcpy(®s->gprs, &vcpu->arch.guest_gprs, sizeof(regs->gprs)); + vcpu_put(vcpu); + return 0; +} + +int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu, + struct kvm_sregs *sregs) +{ + vcpu_load(vcpu); + memcpy(&vcpu->arch.guest_acrs, &sregs->acrs, sizeof(sregs->acrs)); + memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs)); + vcpu_put(vcpu); + return 0; +} + +int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu, + struct kvm_sregs *sregs) +{ + vcpu_load(vcpu); + memcpy(&sregs->acrs, &vcpu->arch.guest_acrs, sizeof(sregs->acrs)); + memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs)); + vcpu_put(vcpu); + return 0; +} + +int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu) +{ + vcpu_load(vcpu); + memcpy(&vcpu->arch.guest_fpregs.fprs, &fpu->fprs, sizeof(fpu->fprs)); + vcpu->arch.guest_fpregs.fpc = fpu->fpc; + vcpu_put(vcpu); + return 0; +} + +int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu) +{ + vcpu_load(vcpu); + memcpy(&fpu->fprs, &vcpu->arch.guest_fpregs.fprs, sizeof(fpu->fprs)); + fpu->fpc = vcpu->arch.guest_fpregs.fpc; + vcpu_put(vcpu); + return 0; +} + +static int kvm_arch_vcpu_ioctl_set_initial_psw(struct kvm_vcpu *vcpu, psw_t psw) +{ + int rc = 0; + + vcpu_load(vcpu); + if (atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_RUNNING) + rc = -EBUSY; + else + vcpu->arch.sie_block->gpsw = psw; + vcpu_put(vcpu); + return rc; +} + +int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu, + struct kvm_translation *tr) +{ + return -EINVAL; /* not implemented yet */ +} + +int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu, + struct kvm_debug_guest *dbg) +{ + return -EINVAL; /* not implemented yet */ +} + +int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu, + struct kvm_mp_state *mp_state) +{ + return -EINVAL; /* not implemented yet */ +} + +int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu, + struct kvm_mp_state *mp_state) +{ + return -EINVAL; /* not implemented yet */ +} + +static void __vcpu_run(struct kvm_vcpu *vcpu) +{ + memcpy(&vcpu->arch.sie_block->gg14, &vcpu->arch.guest_gprs[14], 16); + + if (need_resched()) + schedule(); + + vcpu->arch.sie_block->icptcode = 0; + local_irq_disable(); + kvm_guest_enter(); + local_irq_enable(); + VCPU_EVENT(vcpu, 6, "entering sie flags %x", + atomic_read(&vcpu->arch.sie_block->cpuflags)); + sie64a(vcpu->arch.sie_block, vcpu->arch.guest_gprs); + VCPU_EVENT(vcpu, 6, "exit sie icptcode %d", + vcpu->arch.sie_block->icptcode); + local_irq_disable(); + kvm_guest_exit(); + local_irq_enable(); + + memcpy(&vcpu->arch.guest_gprs[14], &vcpu->arch.sie_block->gg14, 16); +} + +int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + int rc; + sigset_t sigsaved; + + vcpu_load(vcpu); + + if (vcpu->sigset_active) + sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved); + + atomic_set_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); + + BUG_ON(vcpu->kvm->arch.float_int.local_int[vcpu->vcpu_id] == NULL); + + switch (kvm_run->exit_reason) { + case KVM_EXIT_S390_SIEIC: + vcpu->arch.sie_block->gpsw.mask = kvm_run->s390_sieic.mask; + vcpu->arch.sie_block->gpsw.addr = kvm_run->s390_sieic.addr; + break; + case KVM_EXIT_UNKNOWN: + case KVM_EXIT_S390_RESET: + break; + default: + BUG(); + } + + might_sleep(); + + do { + kvm_s390_deliver_pending_interrupts(vcpu); + __vcpu_run(vcpu); + rc = kvm_handle_sie_intercept(vcpu); + } while (!signal_pending(current) && !rc); + + if (signal_pending(current) && !rc) + rc = -EINTR; + + if (rc == -ENOTSUPP) { + /* intercept cannot be handled in-kernel, prepare kvm-run */ + kvm_run->exit_reason = KVM_EXIT_S390_SIEIC; + kvm_run->s390_sieic.icptcode = vcpu->arch.sie_block->icptcode; + kvm_run->s390_sieic.mask = vcpu->arch.sie_block->gpsw.mask; + kvm_run->s390_sieic.addr = vcpu->arch.sie_block->gpsw.addr; + kvm_run->s390_sieic.ipa = vcpu->arch.sie_block->ipa; + kvm_run->s390_sieic.ipb = vcpu->arch.sie_block->ipb; + rc = 0; + } + + if (rc == -EREMOTE) { + /* intercept was handled, but userspace support is needed + * kvm_run has been prepared by the handler */ + rc = 0; + } + + if (vcpu->sigset_active) + sigprocmask(SIG_SETMASK, &sigsaved, NULL); + + vcpu_put(vcpu); + + vcpu->stat.exit_userspace++; + return rc; +} + +static int __guestcopy(struct kvm_vcpu *vcpu, u64 guestdest, const void *from, + unsigned long n, int prefix) +{ + if (prefix) + return copy_to_guest(vcpu, guestdest, from, n); + else + return copy_to_guest_absolute(vcpu, guestdest, from, n); +} + +/* + * store status at address + * we use have two special cases: + * KVM_S390_STORE_STATUS_NOADDR: -> 0x1200 on 64 bit + * KVM_S390_STORE_STATUS_PREFIXED: -> prefix + */ +int __kvm_s390_vcpu_store_status(struct kvm_vcpu *vcpu, unsigned long addr) +{ + const unsigned char archmode = 1; + int prefix; + + if (addr == KVM_S390_STORE_STATUS_NOADDR) { + if (copy_to_guest_absolute(vcpu, 163ul, &archmode, 1)) + return -EFAULT; + addr = SAVE_AREA_BASE; + prefix = 0; + } else if (addr == KVM_S390_STORE_STATUS_PREFIXED) { + if (copy_to_guest(vcpu, 163ul, &archmode, 1)) + return -EFAULT; + addr = SAVE_AREA_BASE; + prefix = 1; + } else + prefix = 0; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, fp_regs), + vcpu->arch.guest_fpregs.fprs, 128, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, gp_regs), + vcpu->arch.guest_gprs, 128, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, psw), + &vcpu->arch.sie_block->gpsw, 16, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, pref_reg), + &vcpu->arch.sie_block->prefix, 4, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, + addr + offsetof(struct save_area_s390x, fp_ctrl_reg), + &vcpu->arch.guest_fpregs.fpc, 4, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, tod_reg), + &vcpu->arch.sie_block->todpr, 4, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, timer), + &vcpu->arch.sie_block->cputm, 8, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, clk_cmp), + &vcpu->arch.sie_block->ckc, 8, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, addr + offsetof(struct save_area_s390x, acc_regs), + &vcpu->arch.guest_acrs, 64, prefix)) + return -EFAULT; + + if (__guestcopy(vcpu, + addr + offsetof(struct save_area_s390x, ctrl_regs), + &vcpu->arch.sie_block->gcr, 128, prefix)) + return -EFAULT; + return 0; +} + +static int kvm_s390_vcpu_store_status(struct kvm_vcpu *vcpu, unsigned long addr) +{ + int rc; + + vcpu_load(vcpu); + rc = __kvm_s390_vcpu_store_status(vcpu, addr); + vcpu_put(vcpu); + return rc; +} + +long kvm_arch_vcpu_ioctl(struct file *filp, + unsigned int ioctl, unsigned long arg) +{ + struct kvm_vcpu *vcpu = filp->private_data; + void __user *argp = (void __user *)arg; + + switch (ioctl) { + case KVM_S390_INTERRUPT: { + struct kvm_s390_interrupt s390int; + + if (copy_from_user(&s390int, argp, sizeof(s390int))) + return -EFAULT; + return kvm_s390_inject_vcpu(vcpu, &s390int); + } + case KVM_S390_STORE_STATUS: + return kvm_s390_vcpu_store_status(vcpu, arg); + case KVM_S390_SET_INITIAL_PSW: { + psw_t psw; + + if (copy_from_user(&psw, argp, sizeof(psw))) + return -EFAULT; + return kvm_arch_vcpu_ioctl_set_initial_psw(vcpu, psw); + } + case KVM_S390_INITIAL_RESET: + return kvm_arch_vcpu_ioctl_initial_reset(vcpu); + default: + ; + } + return -EINVAL; +} + +/* Section: memory related */ +int kvm_arch_set_memory_region(struct kvm *kvm, + struct kvm_userspace_memory_region *mem, + struct kvm_memory_slot old, + int user_alloc) +{ + /* A few sanity checks. We can have exactly one memory slot which has + to start at guest virtual zero and which has to be located at a + page boundary in userland and which has to end at a page boundary. + The memory in userland is ok to be fragmented into various different + vmas. It is okay to mmap() and munmap() stuff in this slot after + doing this call at any time */ + + if (mem->slot) + return -EINVAL; + + if (mem->guest_phys_addr) + return -EINVAL; + + if (mem->userspace_addr & (PAGE_SIZE - 1)) + return -EINVAL; + + if (mem->memory_size & (PAGE_SIZE - 1)) + return -EINVAL; + + kvm->arch.guest_origin = mem->userspace_addr; + kvm->arch.guest_memsize = mem->memory_size; + + /* FIXME: we do want to interrupt running CPUs and update their memory + configuration now to avoid race conditions. But hey, changing the + memory layout while virtual CPUs are running is usually bad + programming practice. */ + + return 0; +} + +gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn) +{ + return gfn; +} + +static int __init kvm_s390_init(void) +{ + return kvm_init(NULL, sizeof(struct kvm_vcpu), THIS_MODULE); +} + +static void __exit kvm_s390_exit(void) +{ + kvm_exit(); +} + +module_init(kvm_s390_init); +module_exit(kvm_s390_exit); diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h new file mode 100644 index 0000000000000000000000000000000000000000..3893cf12eacf4a4bc595b87ee632f84e6b8c7f89 --- /dev/null +++ b/arch/s390/kvm/kvm-s390.h @@ -0,0 +1,64 @@ +/* + * kvm_s390.h - definition for kvm on s390 + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + * Christian Borntraeger + */ + +#ifndef ARCH_S390_KVM_S390_H +#define ARCH_S390_KVM_S390_H + +#include +#include + +typedef int (*intercept_handler_t)(struct kvm_vcpu *vcpu); + +int kvm_handle_sie_intercept(struct kvm_vcpu *vcpu); + +#define VM_EVENT(d_kvm, d_loglevel, d_string, d_args...)\ +do { \ + debug_sprintf_event(d_kvm->arch.dbf, d_loglevel, d_string "\n", \ + d_args); \ +} while (0) + +#define VCPU_EVENT(d_vcpu, d_loglevel, d_string, d_args...)\ +do { \ + debug_sprintf_event(d_vcpu->kvm->arch.dbf, d_loglevel, \ + "%02d[%016lx-%016lx]: " d_string "\n", d_vcpu->vcpu_id, \ + d_vcpu->arch.sie_block->gpsw.mask, d_vcpu->arch.sie_block->gpsw.addr,\ + d_args); \ +} while (0) + +static inline int __cpu_is_stopped(struct kvm_vcpu *vcpu) +{ + return atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_STOP_INT; +} + +int kvm_s390_handle_wait(struct kvm_vcpu *vcpu); +void kvm_s390_idle_wakeup(unsigned long data); +void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu); +int kvm_s390_inject_vm(struct kvm *kvm, + struct kvm_s390_interrupt *s390int); +int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu, + struct kvm_s390_interrupt *s390int); +int kvm_s390_inject_program_int(struct kvm_vcpu *vcpu, u16 code); + +/* implemented in priv.c */ +int kvm_s390_handle_priv(struct kvm_vcpu *vcpu); + +/* implemented in sigp.c */ +int kvm_s390_handle_sigp(struct kvm_vcpu *vcpu); + +/* implemented in kvm-s390.c */ +int __kvm_s390_vcpu_store_status(struct kvm_vcpu *vcpu, + unsigned long addr); +/* implemented in diag.c */ +int kvm_s390_handle_diag(struct kvm_vcpu *vcpu); + +#endif diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c new file mode 100644 index 0000000000000000000000000000000000000000..1465946325c538a85fe183c0fb1aa83c1ece44d3 --- /dev/null +++ b/arch/s390/kvm/priv.c @@ -0,0 +1,323 @@ +/* + * priv.c - handling privileged instructions + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + * Christian Borntraeger + */ + +#include +#include +#include +#include +#include +#include +#include "gaccess.h" +#include "kvm-s390.h" + +static int handle_set_prefix(struct kvm_vcpu *vcpu) +{ + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u64 operand2; + u32 address = 0; + u8 tmp; + + vcpu->stat.instruction_spx++; + + operand2 = disp2; + if (base2) + operand2 += vcpu->arch.guest_gprs[base2]; + + /* must be word boundary */ + if (operand2 & 3) { + kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + goto out; + } + + /* get the value */ + if (get_guest_u32(vcpu, operand2, &address)) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + goto out; + } + + address = address & 0x7fffe000u; + + /* make sure that the new value is valid memory */ + if (copy_from_guest_absolute(vcpu, &tmp, address, 1) || + (copy_from_guest_absolute(vcpu, &tmp, address + PAGE_SIZE, 1))) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + goto out; + } + + vcpu->arch.sie_block->prefix = address; + vcpu->arch.sie_block->ihcpu = 0xffff; + + VCPU_EVENT(vcpu, 5, "setting prefix to %x", address); +out: + return 0; +} + +static int handle_store_prefix(struct kvm_vcpu *vcpu) +{ + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u64 operand2; + u32 address; + + vcpu->stat.instruction_stpx++; + operand2 = disp2; + if (base2) + operand2 += vcpu->arch.guest_gprs[base2]; + + /* must be word boundary */ + if (operand2 & 3) { + kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + goto out; + } + + address = vcpu->arch.sie_block->prefix; + address = address & 0x7fffe000u; + + /* get the value */ + if (put_guest_u32(vcpu, operand2, address)) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + goto out; + } + + VCPU_EVENT(vcpu, 5, "storing prefix to %x", address); +out: + return 0; +} + +static int handle_store_cpu_address(struct kvm_vcpu *vcpu) +{ + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u64 useraddr; + int rc; + + vcpu->stat.instruction_stap++; + useraddr = disp2; + if (base2) + useraddr += vcpu->arch.guest_gprs[base2]; + + if (useraddr & 1) { + kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + goto out; + } + + rc = put_guest_u16(vcpu, useraddr, vcpu->vcpu_id); + if (rc == -EFAULT) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + goto out; + } + + VCPU_EVENT(vcpu, 5, "storing cpu address to %lx", useraddr); +out: + return 0; +} + +static int handle_skey(struct kvm_vcpu *vcpu) +{ + vcpu->stat.instruction_storage_key++; + vcpu->arch.sie_block->gpsw.addr -= 4; + VCPU_EVENT(vcpu, 4, "%s", "retrying storage key operation"); + return 0; +} + +static int handle_stsch(struct kvm_vcpu *vcpu) +{ + vcpu->stat.instruction_stsch++; + VCPU_EVENT(vcpu, 4, "%s", "store subchannel - CC3"); + /* condition code 3 */ + vcpu->arch.sie_block->gpsw.mask &= ~(3ul << 44); + vcpu->arch.sie_block->gpsw.mask |= (3 & 3ul) << 44; + return 0; +} + +static int handle_chsc(struct kvm_vcpu *vcpu) +{ + vcpu->stat.instruction_chsc++; + VCPU_EVENT(vcpu, 4, "%s", "channel subsystem call - CC3"); + /* condition code 3 */ + vcpu->arch.sie_block->gpsw.mask &= ~(3ul << 44); + vcpu->arch.sie_block->gpsw.mask |= (3 & 3ul) << 44; + return 0; +} + +static unsigned int kvm_stfl(void) +{ + asm volatile( + " .insn s,0xb2b10000,0(0)\n" /* stfl */ + "0:\n" + EX_TABLE(0b, 0b)); + return S390_lowcore.stfl_fac_list; +} + +static int handle_stfl(struct kvm_vcpu *vcpu) +{ + unsigned int facility_list = kvm_stfl(); + int rc; + + vcpu->stat.instruction_stfl++; + facility_list &= ~(1UL<<24); /* no stfle */ + + rc = copy_to_guest(vcpu, offsetof(struct _lowcore, stfl_fac_list), + &facility_list, sizeof(facility_list)); + if (rc == -EFAULT) + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + else + VCPU_EVENT(vcpu, 5, "store facility list value %x", + facility_list); + return 0; +} + +static int handle_stidp(struct kvm_vcpu *vcpu) +{ + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u64 operand2; + int rc; + + vcpu->stat.instruction_stidp++; + operand2 = disp2; + if (base2) + operand2 += vcpu->arch.guest_gprs[base2]; + + if (operand2 & 7) { + kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + goto out; + } + + rc = put_guest_u64(vcpu, operand2, vcpu->arch.stidp_data); + if (rc == -EFAULT) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + goto out; + } + + VCPU_EVENT(vcpu, 5, "%s", "store cpu id"); +out: + return 0; +} + +static void handle_stsi_3_2_2(struct kvm_vcpu *vcpu, struct sysinfo_3_2_2 *mem) +{ + struct float_interrupt *fi = &vcpu->kvm->arch.float_int; + int cpus = 0; + int n; + + spin_lock_bh(&fi->lock); + for (n = 0; n < KVM_MAX_VCPUS; n++) + if (fi->local_int[n]) + cpus++; + spin_unlock_bh(&fi->lock); + + /* deal with other level 3 hypervisors */ + if (stsi(mem, 3, 2, 2) == -ENOSYS) + mem->count = 0; + if (mem->count < 8) + mem->count++; + for (n = mem->count - 1; n > 0 ; n--) + memcpy(&mem->vm[n], &mem->vm[n - 1], sizeof(mem->vm[0])); + + mem->vm[0].cpus_total = cpus; + mem->vm[0].cpus_configured = cpus; + mem->vm[0].cpus_standby = 0; + mem->vm[0].cpus_reserved = 0; + mem->vm[0].caf = 1000; + memcpy(mem->vm[0].name, "KVMguest", 8); + ASCEBC(mem->vm[0].name, 8); + memcpy(mem->vm[0].cpi, "KVM/Linux ", 16); + ASCEBC(mem->vm[0].cpi, 16); +} + +static int handle_stsi(struct kvm_vcpu *vcpu) +{ + int fc = (vcpu->arch.guest_gprs[0] & 0xf0000000) >> 28; + int sel1 = vcpu->arch.guest_gprs[0] & 0xff; + int sel2 = vcpu->arch.guest_gprs[1] & 0xffff; + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u64 operand2; + unsigned long mem; + + vcpu->stat.instruction_stsi++; + VCPU_EVENT(vcpu, 4, "stsi: fc: %x sel1: %x sel2: %x", fc, sel1, sel2); + + operand2 = disp2; + if (base2) + operand2 += vcpu->arch.guest_gprs[base2]; + + if (operand2 & 0xfff && fc > 0) + return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + + switch (fc) { + case 0: + vcpu->arch.guest_gprs[0] = 3 << 28; + vcpu->arch.sie_block->gpsw.mask &= ~(3ul << 44); + return 0; + case 1: /* same handling for 1 and 2 */ + case 2: + mem = get_zeroed_page(GFP_KERNEL); + if (!mem) + goto out_fail; + if (stsi((void *) mem, fc, sel1, sel2) == -ENOSYS) + goto out_mem; + break; + case 3: + if (sel1 != 2 || sel2 != 2) + goto out_fail; + mem = get_zeroed_page(GFP_KERNEL); + if (!mem) + goto out_fail; + handle_stsi_3_2_2(vcpu, (void *) mem); + break; + default: + goto out_fail; + } + + if (copy_to_guest_absolute(vcpu, operand2, (void *) mem, PAGE_SIZE)) { + kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + goto out_mem; + } + free_page(mem); + vcpu->arch.sie_block->gpsw.mask &= ~(3ul << 44); + vcpu->arch.guest_gprs[0] = 0; + return 0; +out_mem: + free_page(mem); +out_fail: + /* condition code 3 */ + vcpu->arch.sie_block->gpsw.mask |= 3ul << 44; + return 0; +} + +static intercept_handler_t priv_handlers[256] = { + [0x02] = handle_stidp, + [0x10] = handle_set_prefix, + [0x11] = handle_store_prefix, + [0x12] = handle_store_cpu_address, + [0x29] = handle_skey, + [0x2a] = handle_skey, + [0x2b] = handle_skey, + [0x34] = handle_stsch, + [0x5f] = handle_chsc, + [0x7d] = handle_stsi, + [0xb1] = handle_stfl, +}; + +int kvm_s390_handle_priv(struct kvm_vcpu *vcpu) +{ + intercept_handler_t handler; + + handler = priv_handlers[vcpu->arch.sie_block->ipa & 0x00ff]; + if (handler) + return handler(vcpu); + return -ENOTSUPP; +} diff --git a/arch/s390/kvm/sie64a.S b/arch/s390/kvm/sie64a.S new file mode 100644 index 0000000000000000000000000000000000000000..934fd6a885f69724a00440a714b4fb9498a64c8f --- /dev/null +++ b/arch/s390/kvm/sie64a.S @@ -0,0 +1,47 @@ +/* + * sie64a.S - low level sie call + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Heiko Carstens + */ + +#include +#include + +SP_R5 = 5 * 8 # offset into stackframe +SP_R6 = 6 * 8 + +/* + * sie64a calling convention: + * %r2 pointer to sie control block + * %r3 guest register save area + */ + .globl sie64a +sie64a: + lgr %r5,%r3 + stmg %r5,%r14,SP_R5(%r15) # save register on entry + lgr %r14,%r2 # pointer to sie control block + lmg %r0,%r13,0(%r3) # load guest gprs 0-13 +sie_inst: + sie 0(%r14) + lg %r14,SP_R5(%r15) + stmg %r0,%r13,0(%r14) # save guest gprs 0-13 + lghi %r2,0 + lmg %r6,%r14,SP_R6(%r15) + br %r14 + +sie_err: + lg %r14,SP_R5(%r15) + stmg %r0,%r13,0(%r14) # save guest gprs 0-13 + lghi %r2,-EFAULT + lmg %r6,%r14,SP_R6(%r15) + br %r14 + + .section __ex_table,"a" + .quad sie_inst,sie_err + .previous diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c new file mode 100644 index 0000000000000000000000000000000000000000..0a236acfb5f6f472226845684697e77732ee1a75 --- /dev/null +++ b/arch/s390/kvm/sigp.c @@ -0,0 +1,288 @@ +/* + * sigp.c - handlinge interprocessor communication + * + * Copyright IBM Corp. 2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Carsten Otte + * Christian Borntraeger + */ + +#include +#include +#include "gaccess.h" +#include "kvm-s390.h" + +/* sigp order codes */ +#define SIGP_SENSE 0x01 +#define SIGP_EXTERNAL_CALL 0x02 +#define SIGP_EMERGENCY 0x03 +#define SIGP_START 0x04 +#define SIGP_STOP 0x05 +#define SIGP_RESTART 0x06 +#define SIGP_STOP_STORE_STATUS 0x09 +#define SIGP_INITIAL_CPU_RESET 0x0b +#define SIGP_CPU_RESET 0x0c +#define SIGP_SET_PREFIX 0x0d +#define SIGP_STORE_STATUS_ADDR 0x0e +#define SIGP_SET_ARCH 0x12 + +/* cpu status bits */ +#define SIGP_STAT_EQUIPMENT_CHECK 0x80000000UL +#define SIGP_STAT_INCORRECT_STATE 0x00000200UL +#define SIGP_STAT_INVALID_PARAMETER 0x00000100UL +#define SIGP_STAT_EXT_CALL_PENDING 0x00000080UL +#define SIGP_STAT_STOPPED 0x00000040UL +#define SIGP_STAT_OPERATOR_INTERV 0x00000020UL +#define SIGP_STAT_CHECK_STOP 0x00000010UL +#define SIGP_STAT_INOPERATIVE 0x00000004UL +#define SIGP_STAT_INVALID_ORDER 0x00000002UL +#define SIGP_STAT_RECEIVER_CHECK 0x00000001UL + + +static int __sigp_sense(struct kvm_vcpu *vcpu, u16 cpu_addr, u64 *reg) +{ + struct float_interrupt *fi = &vcpu->kvm->arch.float_int; + int rc; + + if (cpu_addr >= KVM_MAX_VCPUS) + return 3; /* not operational */ + + spin_lock_bh(&fi->lock); + if (fi->local_int[cpu_addr] == NULL) + rc = 3; /* not operational */ + else if (atomic_read(fi->local_int[cpu_addr]->cpuflags) + & CPUSTAT_RUNNING) { + *reg &= 0xffffffff00000000UL; + rc = 1; /* status stored */ + } else { + *reg &= 0xffffffff00000000UL; + *reg |= SIGP_STAT_STOPPED; + rc = 1; /* status stored */ + } + spin_unlock_bh(&fi->lock); + + VCPU_EVENT(vcpu, 4, "sensed status of cpu %x rc %x", cpu_addr, rc); + return rc; +} + +static int __sigp_emergency(struct kvm_vcpu *vcpu, u16 cpu_addr) +{ + struct float_interrupt *fi = &vcpu->kvm->arch.float_int; + struct local_interrupt *li; + struct interrupt_info *inti; + int rc; + + if (cpu_addr >= KVM_MAX_VCPUS) + return 3; /* not operational */ + + inti = kzalloc(sizeof(*inti), GFP_KERNEL); + if (!inti) + return -ENOMEM; + + inti->type = KVM_S390_INT_EMERGENCY; + + spin_lock_bh(&fi->lock); + li = fi->local_int[cpu_addr]; + if (li == NULL) { + rc = 3; /* not operational */ + kfree(inti); + goto unlock; + } + spin_lock_bh(&li->lock); + list_add_tail(&inti->list, &li->list); + atomic_set(&li->active, 1); + atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags); + if (waitqueue_active(&li->wq)) + wake_up_interruptible(&li->wq); + spin_unlock_bh(&li->lock); + rc = 0; /* order accepted */ +unlock: + spin_unlock_bh(&fi->lock); + VCPU_EVENT(vcpu, 4, "sent sigp emerg to cpu %x", cpu_addr); + return rc; +} + +static int __sigp_stop(struct kvm_vcpu *vcpu, u16 cpu_addr, int store) +{ + struct float_interrupt *fi = &vcpu->kvm->arch.float_int; + struct local_interrupt *li; + struct interrupt_info *inti; + int rc; + + if (cpu_addr >= KVM_MAX_VCPUS) + return 3; /* not operational */ + + inti = kzalloc(sizeof(*inti), GFP_KERNEL); + if (!inti) + return -ENOMEM; + + inti->type = KVM_S390_SIGP_STOP; + + spin_lock_bh(&fi->lock); + li = fi->local_int[cpu_addr]; + if (li == NULL) { + rc = 3; /* not operational */ + kfree(inti); + goto unlock; + } + spin_lock_bh(&li->lock); + list_add_tail(&inti->list, &li->list); + atomic_set(&li->active, 1); + atomic_set_mask(CPUSTAT_STOP_INT, li->cpuflags); + if (store) + li->action_bits |= ACTION_STORE_ON_STOP; + li->action_bits |= ACTION_STOP_ON_STOP; + if (waitqueue_active(&li->wq)) + wake_up_interruptible(&li->wq); + spin_unlock_bh(&li->lock); + rc = 0; /* order accepted */ +unlock: + spin_unlock_bh(&fi->lock); + VCPU_EVENT(vcpu, 4, "sent sigp stop to cpu %x", cpu_addr); + return rc; +} + +static int __sigp_set_arch(struct kvm_vcpu *vcpu, u32 parameter) +{ + int rc; + + switch (parameter & 0xff) { + case 0: + printk(KERN_WARNING "kvm: request to switch to ESA/390 mode" + " not supported"); + rc = 3; /* not operational */ + break; + case 1: + case 2: + rc = 0; /* order accepted */ + break; + default: + rc = -ENOTSUPP; + } + return rc; +} + +static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address, + u64 *reg) +{ + struct float_interrupt *fi = &vcpu->kvm->arch.float_int; + struct local_interrupt *li; + struct interrupt_info *inti; + int rc; + u8 tmp; + + /* make sure that the new value is valid memory */ + address = address & 0x7fffe000u; + if ((copy_from_guest(vcpu, &tmp, + (u64) (address + vcpu->kvm->arch.guest_origin) , 1)) || + (copy_from_guest(vcpu, &tmp, (u64) (address + + vcpu->kvm->arch.guest_origin + PAGE_SIZE), 1))) { + *reg |= SIGP_STAT_INVALID_PARAMETER; + return 1; /* invalid parameter */ + } + + inti = kzalloc(sizeof(*inti), GFP_KERNEL); + if (!inti) + return 2; /* busy */ + + spin_lock_bh(&fi->lock); + li = fi->local_int[cpu_addr]; + + if ((cpu_addr >= KVM_MAX_VCPUS) || (li == NULL)) { + rc = 1; /* incorrect state */ + *reg &= SIGP_STAT_INCORRECT_STATE; + kfree(inti); + goto out_fi; + } + + spin_lock_bh(&li->lock); + /* cpu must be in stopped state */ + if (atomic_read(li->cpuflags) & CPUSTAT_RUNNING) { + rc = 1; /* incorrect state */ + *reg &= SIGP_STAT_INCORRECT_STATE; + kfree(inti); + goto out_li; + } + + inti->type = KVM_S390_SIGP_SET_PREFIX; + inti->prefix.address = address; + + list_add_tail(&inti->list, &li->list); + atomic_set(&li->active, 1); + if (waitqueue_active(&li->wq)) + wake_up_interruptible(&li->wq); + rc = 0; /* order accepted */ + + VCPU_EVENT(vcpu, 4, "set prefix of cpu %02x to %x", cpu_addr, address); +out_li: + spin_unlock_bh(&li->lock); +out_fi: + spin_unlock_bh(&fi->lock); + return rc; +} + +int kvm_s390_handle_sigp(struct kvm_vcpu *vcpu) +{ + int r1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4; + int r3 = vcpu->arch.sie_block->ipa & 0x000f; + int base2 = vcpu->arch.sie_block->ipb >> 28; + int disp2 = ((vcpu->arch.sie_block->ipb & 0x0fff0000) >> 16); + u32 parameter; + u16 cpu_addr = vcpu->arch.guest_gprs[r3]; + u8 order_code; + int rc; + + order_code = disp2; + if (base2) + order_code += vcpu->arch.guest_gprs[base2]; + + if (r1 % 2) + parameter = vcpu->arch.guest_gprs[r1]; + else + parameter = vcpu->arch.guest_gprs[r1 + 1]; + + switch (order_code) { + case SIGP_SENSE: + vcpu->stat.instruction_sigp_sense++; + rc = __sigp_sense(vcpu, cpu_addr, + &vcpu->arch.guest_gprs[r1]); + break; + case SIGP_EMERGENCY: + vcpu->stat.instruction_sigp_emergency++; + rc = __sigp_emergency(vcpu, cpu_addr); + break; + case SIGP_STOP: + vcpu->stat.instruction_sigp_stop++; + rc = __sigp_stop(vcpu, cpu_addr, 0); + break; + case SIGP_STOP_STORE_STATUS: + vcpu->stat.instruction_sigp_stop++; + rc = __sigp_stop(vcpu, cpu_addr, 1); + break; + case SIGP_SET_ARCH: + vcpu->stat.instruction_sigp_arch++; + rc = __sigp_set_arch(vcpu, parameter); + break; + case SIGP_SET_PREFIX: + vcpu->stat.instruction_sigp_prefix++; + rc = __sigp_set_prefix(vcpu, cpu_addr, parameter, + &vcpu->arch.guest_gprs[r1]); + break; + case SIGP_RESTART: + vcpu->stat.instruction_sigp_restart++; + /* user space must know about restart */ + default: + return -ENOTSUPP; + } + + if (rc < 0) + return rc; + + vcpu->arch.sie_block->gpsw.mask &= ~(3ul << 44); + vcpu->arch.sie_block->gpsw.mask |= (rc & 3ul) << 44; + return 0; +} diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index fd072013f88ce11658adab8b7db4818a6479f50b..5c1aea97cd1229b2e231c7dd838ee71663087316 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -30,11 +30,27 @@ #define TABLES_PER_PAGE 4 #define FRAG_MASK 15UL #define SECOND_HALVES 10UL + +void clear_table_pgstes(unsigned long *table) +{ + clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE/4); + memset(table + 256, 0, PAGE_SIZE/4); + clear_table(table + 512, _PAGE_TYPE_EMPTY, PAGE_SIZE/4); + memset(table + 768, 0, PAGE_SIZE/4); +} + #else #define ALLOC_ORDER 2 #define TABLES_PER_PAGE 2 #define FRAG_MASK 3UL #define SECOND_HALVES 2UL + +void clear_table_pgstes(unsigned long *table) +{ + clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE/2); + memset(table + 256, 0, PAGE_SIZE/2); +} + #endif unsigned long *crst_table_alloc(struct mm_struct *mm, int noexec) @@ -153,7 +169,7 @@ unsigned long *page_table_alloc(struct mm_struct *mm) unsigned long *table; unsigned long bits; - bits = mm->context.noexec ? 3UL : 1UL; + bits = (mm->context.noexec || mm->context.pgstes) ? 3UL : 1UL; spin_lock(&mm->page_table_lock); page = NULL; if (!list_empty(&mm->context.pgtable_list)) { @@ -170,7 +186,10 @@ unsigned long *page_table_alloc(struct mm_struct *mm) pgtable_page_ctor(page); page->flags &= ~FRAG_MASK; table = (unsigned long *) page_to_phys(page); - clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE); + if (mm->context.pgstes) + clear_table_pgstes(table); + else + clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE); spin_lock(&mm->page_table_lock); list_add(&page->lru, &mm->context.pgtable_list); } @@ -191,7 +210,7 @@ void page_table_free(struct mm_struct *mm, unsigned long *table) struct page *page; unsigned long bits; - bits = mm->context.noexec ? 3UL : 1UL; + bits = (mm->context.noexec || mm->context.pgstes) ? 3UL : 1UL; bits <<= (__pa(table) & (PAGE_SIZE - 1)) / 256 / sizeof(unsigned long); page = pfn_to_page(__pa(table) >> PAGE_SHIFT); spin_lock(&mm->page_table_lock); @@ -228,3 +247,43 @@ void disable_noexec(struct mm_struct *mm, struct task_struct *tsk) mm->context.noexec = 0; update_mm(mm, tsk); } + +/* + * switch on pgstes for its userspace process (for kvm) + */ +int s390_enable_sie(void) +{ + struct task_struct *tsk = current; + struct mm_struct *mm; + int rc; + + task_lock(tsk); + + rc = 0; + if (tsk->mm->context.pgstes) + goto unlock; + + rc = -EINVAL; + if (!tsk->mm || atomic_read(&tsk->mm->mm_users) > 1 || + tsk->mm != tsk->active_mm || tsk->mm->ioctx_list) + goto unlock; + + tsk->mm->context.pgstes = 1; /* dirty little tricks .. */ + mm = dup_mm(tsk); + tsk->mm->context.pgstes = 0; + + rc = -ENOMEM; + if (!mm) + goto unlock; + mmput(tsk->mm); + tsk->mm = tsk->active_mm = mm; + preempt_disable(); + update_mm(mm, tsk); + cpu_set(smp_processor_id(), mm->cpu_vm_mask); + preempt_enable(); + rc = 0; +unlock: + task_unlock(tsk); + return rc; +} +EXPORT_SYMBOL_GPL(s390_enable_sie); diff --git a/arch/um/Kconfig.x86_64 b/arch/um/Kconfig.x86_64 index 3fbe69e359edcd6ddcc7b18e701c7a04071741fa..5696e7b374b31904b7559d045b7235b4e25bcf3e 100644 --- a/arch/um/Kconfig.x86_64 +++ b/arch/um/Kconfig.x86_64 @@ -1,3 +1,10 @@ + +menu "Host processor type and features" + +source "arch/x86/Kconfig.cpu" + +endmenu + config UML_X86 bool default y diff --git a/arch/um/os-Linux/helper.c b/arch/um/os-Linux/helper.c index f4bd349d441222fb7c400ac63d8e7b8c390784a2..f25c29a12d007676d736ed8b8fcb9c9e47a0eaed 100644 --- a/arch/um/os-Linux/helper.c +++ b/arch/um/os-Linux/helper.c @@ -14,6 +14,7 @@ #include "os.h" #include "um_malloc.h" #include "user.h" +#include struct helper_data { void (*pre_exec)(void*); diff --git a/arch/um/sys-i386/Makefile b/arch/um/sys-i386/Makefile index 964dc1a04c37a7774f56a5d0842d737d477c85c4..598b5c1903af11f10c5c171402b6d283ba9e7565 100644 --- a/arch/um/sys-i386/Makefile +++ b/arch/um/sys-i386/Makefile @@ -6,7 +6,7 @@ obj-y = bug.o bugs.o checksum.o delay.o fault.o ksyms.o ldt.o ptrace.o \ ptrace_user.o setjmp.o signal.o stub.o stub_segv.o syscalls.o sysrq.o \ sys_call_table.o tls.o -subarch-obj-y = lib/bitops_32.o lib/semaphore_32.o lib/string_32.o +subarch-obj-y = lib/semaphore_32.o lib/string_32.o subarch-obj-$(CONFIG_HIGHMEM) += mm/highmem_32.o subarch-obj-$(CONFIG_MODULES) += kernel/module_32.o diff --git a/arch/um/sys-x86_64/Makefile b/arch/um/sys-x86_64/Makefile index 3c22de532088b4c5baef1bee93cb85d9cd668840..c8b4cce9cfe1ce333dd877c8644f5b7d4fab3558 100644 --- a/arch/um/sys-x86_64/Makefile +++ b/arch/um/sys-x86_64/Makefile @@ -10,7 +10,7 @@ obj-y = bug.o bugs.o delay.o fault.o ldt.o mem.o ptrace.o ptrace_user.o \ obj-$(CONFIG_MODULES) += um_module.o -subarch-obj-y = lib/bitops_64.o lib/csum-partial_64.o lib/memcpy_64.o lib/thunk_64.o +subarch-obj-y = lib/csum-partial_64.o lib/memcpy_64.o lib/thunk_64.o subarch-obj-$(CONFIG_MODULES) += kernel/module_64.o ldt-y = ../sys-i386/ldt.o diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 4d350b5cbc71fdaf7f8c8ed0d6a84d4aa96b9de4..e5790fe9e330acf69482fd7e988c47b9075295b5 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -142,6 +142,9 @@ config AUDIT_ARCH config ARCH_SUPPORTS_AOUT def_bool y +config ARCH_SUPPORTS_OPTIMIZED_INLINING + def_bool y + # Use the generic interrupt handling code in kernel/irq/: config GENERIC_HARDIRQS bool @@ -370,6 +373,25 @@ config VMI at the moment), by linking the kernel to a GPL-ed ROM module provided by the hypervisor. +config KVM_CLOCK + bool "KVM paravirtualized clock" + select PARAVIRT + depends on !(X86_VISWS || X86_VOYAGER) + help + Turning on this option will allow you to run a paravirtualized clock + when running over the KVM hypervisor. Instead of relying on a PIT + (or probably other) emulation by the underlying device model, the host + provides the guest with timing infrastructure such as time of day, and + system time + +config KVM_GUEST + bool "KVM Guest support" + select PARAVIRT + depends on !(X86_VISWS || X86_VOYAGER) + help + This option enables various optimizations for running under the KVM + hypervisor. + source "arch/x86/lguest/Kconfig" config PARAVIRT @@ -1049,9 +1071,9 @@ config MTRR See for more information. config X86_PAT - def_bool y + bool prompt "x86 PAT support" - depends on MTRR && NONPROMISC_DEVMEM + depends on MTRR help Use PAT attributes to setup page level cache control. diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index 57072f2716f98881fd67a6addd82cffa27ab7817..7ef18b01f0bca49c4da9ed1ed78b330ab1506da4 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -21,8 +21,8 @@ config M386 Here are the settings recommended for greatest speed: - "386" for the AMD/Cyrix/Intel 386DX/DXL/SL/SLC/SX, Cyrix/TI - 486DLC/DLC2, UMC 486SX-S and NexGen Nx586. Only "386" kernels - will run on a 386 class machine. + 486DLC/DLC2, and UMC 486SX-S. Only "386" kernels will run on a 386 + class machine. - "486" for the AMD/Cyrix/IBM/Intel 486DX/DX2/DX4 or SL/SLC/SLC2/SLC3/SX/SX2 and UMC U5D or U5S. - "586" for generic Pentium CPUs lacking the TSC @@ -278,6 +278,11 @@ config GENERIC_CPU endchoice +config X86_CPU + def_bool y + select GENERIC_FIND_FIRST_BIT + select GENERIC_FIND_NEXT_BIT + config X86_GENERIC bool "Generic x86 support" depends on X86_32 @@ -398,7 +403,7 @@ config X86_TSC # generates cmov. config X86_CMOV def_bool y - depends on (MK7 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || MVIAC3_2 || MVIAC7) + depends on (MK7 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || MVIAC3_2 || MVIAC7 || X86_64) config X86_MINIMUM_CPU_FAMILY int diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug index 239fd9fba0a5fd8737aff15585494ae1d0176611..5b1979a45a1ea07f6ffc8d6b563fc1ddcb64dae3 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -257,3 +257,16 @@ config CPA_DEBUG Do change_page_attr() self-tests every 30 seconds. endmenu + +config OPTIMIZE_INLINING + bool "Allow gcc to uninline functions marked 'inline'" + default y + help + This option determines if the kernel forces gcc to inline the functions + developers have marked 'inline'. Doing so takes away freedom from gcc to + do what it thinks is best, which is desirable for the gcc 3.x series of + compilers. The gcc 4.x series have a rewritten inlining algorithm and + disabling this option will generate a smaller kernel there. Hopefully + this algorithm is so good that allowing gcc4 to make the decision can + become the default in the future, until then this option is there to + test gcc for this. diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S index 6d2df8d61c54089a553c0cdee879dde6f421130c..af86e431acfae2af91957a9932c3bf5fb96bd5fb 100644 --- a/arch/x86/boot/header.S +++ b/arch/x86/boot/header.S @@ -120,7 +120,7 @@ _start: # Part 2 of the header, from the old setup.S .ascii "HdrS" # header signature - .word 0x0208 # header version number (>= 0x0105) + .word 0x0209 # header version number (>= 0x0105) # or else old loadlin-1.5 will fail) .globl realmode_swtch realmode_swtch: .word 0, 0 # default_switch, SETUPSEG @@ -227,6 +227,10 @@ hardware_subarch_data: .quad 0 payload_offset: .long input_data payload_length: .long input_data_end-input_data +setup_data: .quad 0 # 64-bit physical pointer to + # single linked list of + # struct setup_data + # End of setup header ##################################################### .section ".inittext", "ax" diff --git a/arch/x86/configs/i386_defconfig b/arch/x86/configs/i386_defconfig index 3df340b54e57e93ab09dc6430184f38a2e4b49f3..ad7ddaaff588aaa619010fabe29af2ee2559d54d 100644 --- a/arch/x86/configs/i386_defconfig +++ b/arch/x86/configs/i386_defconfig @@ -1421,6 +1421,7 @@ CONFIG_DEBUG_BUGVERBOSE=y # CONFIG_DEBUG_VM is not set # CONFIG_DEBUG_LIST is not set # CONFIG_FRAME_POINTER is not set +CONFIG_OPTIMIZE_INLINING=y # CONFIG_RCU_TORTURE_TEST is not set # CONFIG_LKDTM is not set # CONFIG_FAULT_INJECTION is not set diff --git a/arch/x86/configs/x86_64_defconfig b/arch/x86/configs/x86_64_defconfig index eef98cb00c629f118ae3982c4f654290de6c2102..2d6f5b2809d2f4644e82ccfd323acbe587693bf5 100644 --- a/arch/x86/configs/x86_64_defconfig +++ b/arch/x86/configs/x86_64_defconfig @@ -1346,6 +1346,7 @@ CONFIG_DEBUG_BUGVERBOSE=y # CONFIG_DEBUG_VM is not set # CONFIG_DEBUG_LIST is not set # CONFIG_FRAME_POINTER is not set +CONFIG_OPTIMIZE_INLINING=y # CONFIG_RCU_TORTURE_TEST is not set # CONFIG_LKDTM is not set # CONFIG_FAULT_INJECTION is not set diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c index 05e155d3fb6c6eebb668ac74ac52fa916f53df5b..bbed3a26ce5567abb107f0371651c906b5f368a7 100644 --- a/arch/x86/ia32/ia32_signal.c +++ b/arch/x86/ia32/ia32_signal.c @@ -499,11 +499,6 @@ int ia32_setup_frame(int sig, struct k_sigaction *ka, regs->cs = __USER32_CS; regs->ss = __USER32_DS; - set_fs(USER_DS); - regs->flags &= ~(X86_EFLAGS_TF | X86_EFLAGS_DF); - if (test_thread_flag(TIF_SINGLESTEP)) - ptrace_notify(SIGTRAP); - #if DEBUG_SIG printk(KERN_DEBUG "SIG deliver (%s:%d): sp=%p pc=%lx ra=%u\n", current->comm, current->pid, frame, regs->ip, frame->pretcode); @@ -599,11 +594,6 @@ int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, regs->cs = __USER32_CS; regs->ss = __USER32_DS; - set_fs(USER_DS); - regs->flags &= ~(X86_EFLAGS_TF | X86_EFLAGS_DF); - if (test_thread_flag(TIF_SINGLESTEP)) - ptrace_notify(SIGTRAP); - #if DEBUG_SIG printk(KERN_DEBUG "SIG deliver (%s:%d): sp=%p pc=%lx ra=%u\n", current->comm, current->pid, frame, regs->ip, frame->pretcode); diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index ae7158bce4d62d6b0ec1bfec5c52ead3b3e48049..b5e329da166cfa2a48a74339a47539739ec939ac 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -430,7 +430,7 @@ ia32_sys_call_table: .quad sys_setuid16 .quad sys_getuid16 .quad compat_sys_stime /* stime */ /* 25 */ - .quad sys32_ptrace /* ptrace */ + .quad compat_sys_ptrace /* ptrace */ .quad sys_alarm .quad sys_fstat /* (old)fstat */ .quad sys_pause diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 90e092d0af0c639211932fc44a0f5dd959ec590b..fa19c3819540d0f823bd675f61457bab2fc31b30 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -80,6 +80,8 @@ obj-$(CONFIG_DEBUG_RODATA_TEST) += test_rodata.o obj-$(CONFIG_DEBUG_NX_TEST) += test_nx.o obj-$(CONFIG_VMI) += vmi_32.o vmiclock_32.o +obj-$(CONFIG_KVM_GUEST) += kvm.o +obj-$(CONFIG_KVM_CLOCK) += kvmclock.o obj-$(CONFIG_PARAVIRT) += paravirt.o paravirt_patch_$(BITS).o ifdef CONFIG_INPUT_PCSPKR diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 057ccf1d5ad453891fa78f4c85622386ccfe8728..977ed5cdeaa3b1b6a8d5f78b3446df7be7e21b53 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -697,10 +697,6 @@ static int __init acpi_parse_hpet(struct acpi_table_header *table) #define HPET_RESOURCE_NAME_SIZE 9 hpet_res = alloc_bootmem(sizeof(*hpet_res) + HPET_RESOURCE_NAME_SIZE); - if (!hpet_res) - return 0; - - memset(hpet_res, 0, sizeof(*hpet_res)); hpet_res->name = (void *)&hpet_res[1]; hpet_res->flags = IORESOURCE_MEM; snprintf((char *)hpet_res->name, HPET_RESOURCE_NAME_SIZE, "HPET %u", diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c index 8317401170b86845d3346251c3d44ef150830c05..4b99b1bdeb6cbb090e5868138922b259da7d98fe 100644 --- a/arch/x86/kernel/apic_32.c +++ b/arch/x86/kernel/apic_32.c @@ -451,7 +451,8 @@ void __init setup_boot_APIC_clock(void) } /* Calculate the scaled math multiplication factor */ - lapic_clockevent.mult = div_sc(delta, TICK_NSEC * LAPIC_CAL_LOOPS, 32); + lapic_clockevent.mult = div_sc(delta, TICK_NSEC * LAPIC_CAL_LOOPS, + lapic_clockevent.shift); lapic_clockevent.max_delta_ns = clockevent_delta2ns(0x7FFFFF, &lapic_clockevent); lapic_clockevent.min_delta_ns = diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c index bf83157337e428bb84619478420d540e01fce466..5910020c3f24f20f4a1be57650d1e2d9c9a3a1fa 100644 --- a/arch/x86/kernel/apic_64.c +++ b/arch/x86/kernel/apic_64.c @@ -360,7 +360,8 @@ static void __init calibrate_APIC_clock(void) result / 1000 / 1000, result / 1000 % 1000); /* Calculate the scaled math multiplication factor */ - lapic_clockevent.mult = div_sc(result, NSEC_PER_SEC, 32); + lapic_clockevent.mult = div_sc(result, NSEC_PER_SEC, + lapic_clockevent.shift); lapic_clockevent.max_delta_ns = clockevent_delta2ns(0x7FFFFF, &lapic_clockevent); lapic_clockevent.min_delta_ns = diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index f0030a0999c74f01785c4acd7a499778de58a4ff..e4ea362e84800a6c28bf8056ec340aff5fea4bad 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -904,6 +904,7 @@ static void apm_cpu_idle(void) original_pm_idle(); else default_idle(); + local_irq_disable(); jiffies_since_last_check = jiffies - last_jiffies; if (jiffies_since_last_check > idle_period) goto recalc; @@ -911,6 +912,8 @@ static void apm_cpu_idle(void) if (apm_idle_done) apm_do_busy(); + + local_irq_enable(); } /** diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index ee7c45235e547a769f091842a8fe3cf959ce5fcc..a0c6f8190887f4220b4b76e531e57ad471fc1acb 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -11,7 +11,6 @@ obj-$(CONFIG_X86_32) += cyrix.o obj-$(CONFIG_X86_32) += centaur.o obj-$(CONFIG_X86_32) += transmeta.o obj-$(CONFIG_X86_32) += intel.o -obj-$(CONFIG_X86_32) += nexgen.o obj-$(CONFIG_X86_32) += umc.o obj-$(CONFIG_X86_MCE) += mcheck/ diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 0173065dc3b7bd81995f0d8ed5d84c7876624452..245866828294c6be3fcf369d0ea76a0a3eed59cb 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -343,10 +343,4 @@ static struct cpu_dev amd_cpu_dev __cpuinitdata = { .c_size_cache = amd_size_cache, }; -int __init amd_init_cpu(void) -{ - cpu_devs[X86_VENDOR_AMD] = &amd_cpu_dev; - return 0; -} - cpu_vendor_dev_register(X86_VENDOR_AMD, &amd_cpu_dev); diff --git a/arch/x86/kernel/cpu/mcheck/mce_64.c b/arch/x86/kernel/cpu/mcheck/mce_64.c index 9a699ed035982ddeac48c4800478d5c8d4a0a5c4..e07e8c068ae00c464cdbf828cfbc8c67aaf76202 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_64.c @@ -49,7 +49,7 @@ static int banks; static unsigned long bank[NR_BANKS] = { [0 ... NR_BANKS-1] = ~0UL }; static unsigned long notify_user; static int rip_msr; -static int mce_bootlog = 1; +static int mce_bootlog = -1; static atomic_t mce_events; static char trigger[128]; @@ -471,13 +471,15 @@ static void mce_init(void *dummy) static void __cpuinit mce_cpu_quirks(struct cpuinfo_x86 *c) { /* This should be disabled by the BIOS, but isn't always */ - if (c->x86_vendor == X86_VENDOR_AMD && c->x86 == 15) { - /* disable GART TBL walk error reporting, which trips off - incorrectly with the IOMMU & 3ware & Cerberus. */ - clear_bit(10, &bank[4]); - /* Lots of broken BIOS around that don't clear them - by default and leave crap in there. Don't log. */ - mce_bootlog = 0; + if (c->x86_vendor == X86_VENDOR_AMD) { + if(c->x86 == 15) + /* disable GART TBL walk error reporting, which trips off + incorrectly with the IOMMU & 3ware & Cerberus. */ + clear_bit(10, &bank[4]); + if(c->x86 <= 17 && mce_bootlog < 0) + /* Lots of broken BIOS around that don't clear them + by default and leave crap in there. Don't log. */ + mce_bootlog = 0; } } diff --git a/arch/x86/kernel/cpu/nexgen.c b/arch/x86/kernel/cpu/nexgen.c deleted file mode 100644 index 5d5e1c134123b7edef8d4e1cf318c253f36401da..0000000000000000000000000000000000000000 --- a/arch/x86/kernel/cpu/nexgen.c +++ /dev/null @@ -1,59 +0,0 @@ -#include -#include -#include -#include - -#include "cpu.h" - -/* - * Detect a NexGen CPU running without BIOS hypercode new enough - * to have CPUID. (Thanks to Herbert Oppmann) - */ - -static int __cpuinit deep_magic_nexgen_probe(void) -{ - int ret; - - __asm__ __volatile__ ( - " movw $0x5555, %%ax\n" - " xorw %%dx,%%dx\n" - " movw $2, %%cx\n" - " divw %%cx\n" - " movl $0, %%eax\n" - " jnz 1f\n" - " movl $1, %%eax\n" - "1:\n" - : "=a" (ret) : : "cx", "dx"); - return ret; -} - -static void __cpuinit init_nexgen(struct cpuinfo_x86 *c) -{ - c->x86_cache_size = 256; /* A few had 1 MB... */ -} - -static void __cpuinit nexgen_identify(struct cpuinfo_x86 *c) -{ - /* Detect NexGen with old hypercode */ - if (deep_magic_nexgen_probe()) - strcpy(c->x86_vendor_id, "NexGenDriven"); -} - -static struct cpu_dev nexgen_cpu_dev __cpuinitdata = { - .c_vendor = "Nexgen", - .c_ident = { "NexGenDriven" }, - .c_models = { - { .vendor = X86_VENDOR_NEXGEN, - .family = 5, - .model_names = { [1] = "Nx586" } - }, - }, - .c_init = init_nexgen, - .c_identify = nexgen_identify, -}; - -int __init nexgen_init_cpu(void) -{ - cpu_devs[X86_VENDOR_NEXGEN] = &nexgen_cpu_dev; - return 0; -} diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c index b943e10ad81438aa528ee1c7dad5f7a528488a38..f9ae93adffe5a58619b2c687e39cfbb43e00d987 100644 --- a/arch/x86/kernel/cpu/perfctr-watchdog.c +++ b/arch/x86/kernel/cpu/perfctr-watchdog.c @@ -614,16 +614,6 @@ static struct wd_ops intel_arch_wd_ops __read_mostly = { .evntsel = MSR_ARCH_PERFMON_EVENTSEL1, }; -static struct wd_ops coreduo_wd_ops = { - .reserve = single_msr_reserve, - .unreserve = single_msr_unreserve, - .setup = setup_intel_arch_watchdog, - .rearm = p6_rearm, - .stop = single_msr_stop_watchdog, - .perfctr = MSR_ARCH_PERFMON_PERFCTR0, - .evntsel = MSR_ARCH_PERFMON_EVENTSEL0, -}; - static void probe_nmi_watchdog(void) { switch (boot_cpu_data.x86_vendor) { @@ -637,8 +627,8 @@ static void probe_nmi_watchdog(void) /* Work around Core Duo (Yonah) errata AE49 where perfctr1 doesn't have a working enable bit. */ if (boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 14) { - wd_ops = &coreduo_wd_ops; - break; + intel_arch_wd_ops.perfctr = MSR_ARCH_PERFMON_PERFCTR0; + intel_arch_wd_ops.evntsel = MSR_ARCH_PERFMON_EVENTSEL0; } if (cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) { wd_ops = &intel_arch_wd_ops; diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index 2251d0ae9570de24259a6c132aec2c4ef2d05f46..2685538179097a1e8e073ebf07cd49e898ef4f6e 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -25,6 +25,7 @@ #include #include #include +#include #include @@ -117,7 +118,7 @@ static void nmi_shootdown_cpus(void) } #endif -void machine_crash_shutdown(struct pt_regs *regs) +void native_machine_crash_shutdown(struct pt_regs *regs) { /* This function is only called after the system * has panicked or is otherwise in a critical state. diff --git a/arch/x86/kernel/e820_64.c b/arch/x86/kernel/e820_64.c index cbd42e51cb082d82b8f287eaa1c244913268d64c..645ee5e32a27184c714355c8e916951ef3605711 100644 --- a/arch/x86/kernel/e820_64.c +++ b/arch/x86/kernel/e820_64.c @@ -84,14 +84,41 @@ void __init reserve_early(unsigned long start, unsigned long end, char *name) strncpy(r->name, name, sizeof(r->name) - 1); } -void __init early_res_to_bootmem(void) +void __init free_early(unsigned long start, unsigned long end) +{ + struct early_res *r; + int i, j; + + for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { + r = &early_res[i]; + if (start == r->start && end == r->end) + break; + } + if (i >= MAX_EARLY_RES || !early_res[i].end) + panic("free_early on not reserved area: %lx-%lx!", start, end); + + for (j = i + 1; j < MAX_EARLY_RES && early_res[j].end; j++) + ; + + memcpy(&early_res[i], &early_res[i + 1], + (j - 1 - i) * sizeof(struct early_res)); + + early_res[j - 1].end = 0; +} + +void __init early_res_to_bootmem(unsigned long start, unsigned long end) { int i; + unsigned long final_start, final_end; for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { struct early_res *r = &early_res[i]; - printk(KERN_INFO "early res: %d [%lx-%lx] %s\n", i, - r->start, r->end - 1, r->name); - reserve_bootmem_generic(r->start, r->end - r->start); + final_start = max(start, r->start); + final_end = min(end, r->end); + if (final_start >= final_end) + continue; + printk(KERN_INFO " early res: %d [%lx-%lx] %s\n", i, + final_start, final_end - 1, r->name); + reserve_bootmem_generic(final_start, final_end - final_start); } } diff --git a/arch/x86/kernel/genapic_64.c b/arch/x86/kernel/genapic_64.c index 9546ef408b92c95177f58ba5cce6287951789735..021624c835832d81f00f1ed6ef5d5437e9303813 100644 --- a/arch/x86/kernel/genapic_64.c +++ b/arch/x86/kernel/genapic_64.c @@ -51,7 +51,7 @@ void __init setup_apic_routing(void) else #endif - if (cpus_weight(cpu_possible_map) <= 8) + if (num_possible_cpus() <= 8) genapic = &apic_flat; else genapic = &apic_physflat; diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index 993c767732564afe285e7ddc23228511108b7b11..e25c57b8aa844937742f246288e9ffa2e2be5229 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -11,6 +11,7 @@ #include #include #include +#include #include #include @@ -22,6 +23,7 @@ #include #include #include +#include static void __init zap_identity_mappings(void) { @@ -49,7 +51,6 @@ static void __init copy_bootdata(char *real_mode_data) } } -#define BIOS_EBDA_SEGMENT 0x40E #define BIOS_LOWMEM_KILOBYTES 0x413 /* @@ -80,8 +81,7 @@ static void __init reserve_ebda_region(void) lowmem <<= 10; /* start of EBDA area */ - ebda_addr = *(unsigned short *)__va(BIOS_EBDA_SEGMENT); - ebda_addr <<= 4; + ebda_addr = get_bios_ebda(); /* Fixup: bios puts an EBDA in the top 64K segment */ /* of conventional memory, but does not adjust lowmem. */ @@ -101,6 +101,24 @@ static void __init reserve_ebda_region(void) reserve_early(lowmem, 0x100000, "BIOS reserved"); } +static void __init reserve_setup_data(void) +{ + struct setup_data *data; + unsigned long pa_data; + char buf[32]; + + if (boot_params.hdr.version < 0x0209) + return; + pa_data = boot_params.hdr.setup_data; + while (pa_data) { + data = early_ioremap(pa_data, sizeof(*data)); + sprintf(buf, "setup data %x", data->type); + reserve_early(pa_data, pa_data+sizeof(*data)+data->len, buf); + pa_data = data->next; + early_iounmap(data, sizeof(*data)); + } +} + void __init x86_64_start_kernel(char * real_mode_data) { int i; @@ -157,6 +175,7 @@ void __init x86_64_start_kernel(char * real_mode_data) #endif reserve_ebda_region(); + reserve_setup_data(); /* * At this point everything still needed from the boot loader diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 36652ea1a2658d10d729647e0f613c01ccc88ee6..9007f9ea64eed80e9537be44d391fe55d6f7f749 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -218,7 +218,7 @@ static void hpet_legacy_clockevent_register(void) hpet_freq = 1000000000000000ULL; do_div(hpet_freq, hpet_period); hpet_clockevent.mult = div_sc((unsigned long) hpet_freq, - NSEC_PER_SEC, 32); + NSEC_PER_SEC, hpet_clockevent.shift); /* Calculate the min / max delta */ hpet_clockevent.max_delta_ns = clockevent_delta2ns(0x7FFFFFFF, &hpet_clockevent); diff --git a/arch/x86/kernel/i8253.c b/arch/x86/kernel/i8253.c index 8540abe86ade2d0b1f3fd26ad0dff18b5733a6e1..c1b5e3ece1f241f9ef8716abd02d98ad6721ffa7 100644 --- a/arch/x86/kernel/i8253.c +++ b/arch/x86/kernel/i8253.c @@ -115,7 +115,8 @@ void __init setup_pit_timer(void) * IO_APIC has been initialized. */ pit_clockevent.cpumask = cpumask_of_cpu(smp_processor_id()); - pit_clockevent.mult = div_sc(CLOCK_TICK_RATE, NSEC_PER_SEC, 32); + pit_clockevent.mult = div_sc(CLOCK_TICK_RATE, NSEC_PER_SEC, + pit_clockevent.shift); pit_clockevent.max_delta_ns = clockevent_delta2ns(0x7FFF, &pit_clockevent); pit_clockevent.min_delta_ns = @@ -224,7 +225,8 @@ static int __init init_pit_clocksource(void) pit_clockevent.mode != CLOCK_EVT_MODE_PERIODIC) return 0; - clocksource_pit.mult = clocksource_hz2mult(CLOCK_TICK_RATE, 20); + clocksource_pit.mult = clocksource_hz2mult(CLOCK_TICK_RATE, + clocksource_pit.shift); return clocksource_register(&clocksource_pit); } arch_initcall(init_pit_clocksource); diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c index 2e2f42074e18e0fa0de175b47325c5b895635c4d..696b8e4e66bbf89bd9b71499f549697b0f68f745 100644 --- a/arch/x86/kernel/io_apic_32.c +++ b/arch/x86/kernel/io_apic_32.c @@ -2068,7 +2068,7 @@ static void __init setup_nmi(void) * cycles as some i82489DX-based boards have glue logic that keeps the * 8259A interrupt line asserted until INTA. --macro */ -static inline void unlock_ExtINT_logic(void) +static inline void __init unlock_ExtINT_logic(void) { int apic, pin, i; struct IO_APIC_route_entry entry0, entry1; diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c index 9ba11d07920f794c71e11de98d51d8430e129cf2..ef1a8dfcc529a36643bdbe34f835eb1819c5f265 100644 --- a/arch/x86/kernel/io_apic_64.c +++ b/arch/x86/kernel/io_apic_64.c @@ -1599,7 +1599,7 @@ static void __init setup_nmi(void) * cycles as some i82489DX-based boards have glue logic that keeps the * 8259A interrupt line asserted until INTA. --macro */ -static inline void unlock_ExtINT_logic(void) +static inline void __init unlock_ExtINT_logic(void) { int apic, pin, i; struct IO_APIC_route_entry entry0, entry1; diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index 6ea67b76a21452f3b7ffacdec52ae8b155f2a21a..00bda7bcda63741178db3ad46b444fe25c8b87a7 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -134,7 +134,7 @@ unsigned int do_IRQ(struct pt_regs *regs) : "=a" (arg1), "=d" (arg2), "=b" (bx) : "0" (irq), "1" (desc), "2" (isp), "D" (desc->handle_irq) - : "memory", "cc" + : "memory", "cc", "ecx" ); } else #endif diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c index 73354302fda76c3ee5a326e63deed6f37acca3a9..c03205991718b30da8adfed31d915f7f325b2aa0 100644 --- a/arch/x86/kernel/kdebugfs.c +++ b/arch/x86/kernel/kdebugfs.c @@ -6,23 +6,171 @@ * * This file is released under the GPLv2. */ - #include +#include #include #include +#include +#include #include #ifdef CONFIG_DEBUG_BOOT_PARAMS +struct setup_data_node { + u64 paddr; + u32 type; + u32 len; +}; + +static ssize_t +setup_data_read(struct file *file, char __user *user_buf, size_t count, + loff_t *ppos) +{ + struct setup_data_node *node = file->private_data; + unsigned long remain; + loff_t pos = *ppos; + struct page *pg; + void *p; + u64 pa; + + if (pos < 0) + return -EINVAL; + if (pos >= node->len) + return 0; + + if (count > node->len - pos) + count = node->len - pos; + pa = node->paddr + sizeof(struct setup_data) + pos; + pg = pfn_to_page((pa + count - 1) >> PAGE_SHIFT); + if (PageHighMem(pg)) { + p = ioremap_cache(pa, count); + if (!p) + return -ENXIO; + } else { + p = __va(pa); + } + + remain = copy_to_user(user_buf, p, count); + + if (PageHighMem(pg)) + iounmap(p); + + if (remain) + return -EFAULT; + + *ppos = pos + count; + + return count; +} + +static int setup_data_open(struct inode *inode, struct file *file) +{ + file->private_data = inode->i_private; + return 0; +} + +static const struct file_operations fops_setup_data = { + .read = setup_data_read, + .open = setup_data_open, +}; + +static int __init +create_setup_data_node(struct dentry *parent, int no, + struct setup_data_node *node) +{ + struct dentry *d, *type, *data; + char buf[16]; + int error; + + sprintf(buf, "%d", no); + d = debugfs_create_dir(buf, parent); + if (!d) { + error = -ENOMEM; + goto err_return; + } + type = debugfs_create_x32("type", S_IRUGO, d, &node->type); + if (!type) { + error = -ENOMEM; + goto err_dir; + } + data = debugfs_create_file("data", S_IRUGO, d, node, &fops_setup_data); + if (!data) { + error = -ENOMEM; + goto err_type; + } + return 0; + +err_type: + debugfs_remove(type); +err_dir: + debugfs_remove(d); +err_return: + return error; +} + +static int __init create_setup_data_nodes(struct dentry *parent) +{ + struct setup_data_node *node; + struct setup_data *data; + int error, no = 0; + struct dentry *d; + struct page *pg; + u64 pa_data; + + d = debugfs_create_dir("setup_data", parent); + if (!d) { + error = -ENOMEM; + goto err_return; + } + + pa_data = boot_params.hdr.setup_data; + + while (pa_data) { + node = kmalloc(sizeof(*node), GFP_KERNEL); + if (!node) { + error = -ENOMEM; + goto err_dir; + } + pg = pfn_to_page((pa_data+sizeof(*data)-1) >> PAGE_SHIFT); + if (PageHighMem(pg)) { + data = ioremap_cache(pa_data, sizeof(*data)); + if (!data) { + error = -ENXIO; + goto err_dir; + } + } else { + data = __va(pa_data); + } + + node->paddr = pa_data; + node->type = data->type; + node->len = data->len; + error = create_setup_data_node(d, no, node); + pa_data = data->next; + + if (PageHighMem(pg)) + iounmap(data); + if (error) + goto err_dir; + no++; + } + return 0; + +err_dir: + debugfs_remove(d); +err_return: + return error; +} + static struct debugfs_blob_wrapper boot_params_blob = { - .data = &boot_params, - .size = sizeof(boot_params), + .data = &boot_params, + .size = sizeof(boot_params), }; static int __init boot_params_kdebugfs_init(void) { - int error; struct dentry *dbp, *version, *data; + int error; dbp = debugfs_create_dir("boot_params", NULL); if (!dbp) { @@ -41,7 +189,13 @@ static int __init boot_params_kdebugfs_init(void) error = -ENOMEM; goto err_version; } + error = create_setup_data_nodes(dbp); + if (error) + goto err_data; return 0; + +err_data: + debugfs_remove(data); err_version: debugfs_remove(version); err_dir: @@ -61,5 +215,4 @@ static int __init arch_kdebugfs_init(void) return error; } - arch_initcall(arch_kdebugfs_init); diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c new file mode 100644 index 0000000000000000000000000000000000000000..8b7a3cf37d2b8192cbe1a1fe68ff93740dd0d59f --- /dev/null +++ b/arch/x86/kernel/kvm.c @@ -0,0 +1,248 @@ +/* + * KVM paravirt_ops implementation + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * + * Copyright (C) 2007, Red Hat, Inc., Ingo Molnar + * Copyright IBM Corporation, 2007 + * Authors: Anthony Liguori + */ + +#include +#include +#include +#include +#include +#include +#include + +#define MMU_QUEUE_SIZE 1024 + +struct kvm_para_state { + u8 mmu_queue[MMU_QUEUE_SIZE]; + int mmu_queue_len; + enum paravirt_lazy_mode mode; +}; + +static DEFINE_PER_CPU(struct kvm_para_state, para_state); + +static struct kvm_para_state *kvm_para_state(void) +{ + return &per_cpu(para_state, raw_smp_processor_id()); +} + +/* + * No need for any "IO delay" on KVM + */ +static void kvm_io_delay(void) +{ +} + +static void kvm_mmu_op(void *buffer, unsigned len) +{ + int r; + unsigned long a1, a2; + + do { + a1 = __pa(buffer); + a2 = 0; /* on i386 __pa() always returns <4G */ + r = kvm_hypercall3(KVM_HC_MMU_OP, len, a1, a2); + buffer += r; + len -= r; + } while (len); +} + +static void mmu_queue_flush(struct kvm_para_state *state) +{ + if (state->mmu_queue_len) { + kvm_mmu_op(state->mmu_queue, state->mmu_queue_len); + state->mmu_queue_len = 0; + } +} + +static void kvm_deferred_mmu_op(void *buffer, int len) +{ + struct kvm_para_state *state = kvm_para_state(); + + if (state->mode != PARAVIRT_LAZY_MMU) { + kvm_mmu_op(buffer, len); + return; + } + if (state->mmu_queue_len + len > sizeof state->mmu_queue) + mmu_queue_flush(state); + memcpy(state->mmu_queue + state->mmu_queue_len, buffer, len); + state->mmu_queue_len += len; +} + +static void kvm_mmu_write(void *dest, u64 val) +{ + __u64 pte_phys; + struct kvm_mmu_op_write_pte wpte; + +#ifdef CONFIG_HIGHPTE + struct page *page; + unsigned long dst = (unsigned long) dest; + + page = kmap_atomic_to_page(dest); + pte_phys = page_to_pfn(page); + pte_phys <<= PAGE_SHIFT; + pte_phys += (dst & ~(PAGE_MASK)); +#else + pte_phys = (unsigned long)__pa(dest); +#endif + wpte.header.op = KVM_MMU_OP_WRITE_PTE; + wpte.pte_val = val; + wpte.pte_phys = pte_phys; + + kvm_deferred_mmu_op(&wpte, sizeof wpte); +} + +/* + * We only need to hook operations that are MMU writes. We hook these so that + * we can use lazy MMU mode to batch these operations. We could probably + * improve the performance of the host code if we used some of the information + * here to simplify processing of batched writes. + */ +static void kvm_set_pte(pte_t *ptep, pte_t pte) +{ + kvm_mmu_write(ptep, pte_val(pte)); +} + +static void kvm_set_pte_at(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, pte_t pte) +{ + kvm_mmu_write(ptep, pte_val(pte)); +} + +static void kvm_set_pmd(pmd_t *pmdp, pmd_t pmd) +{ + kvm_mmu_write(pmdp, pmd_val(pmd)); +} + +#if PAGETABLE_LEVELS >= 3 +#ifdef CONFIG_X86_PAE +static void kvm_set_pte_atomic(pte_t *ptep, pte_t pte) +{ + kvm_mmu_write(ptep, pte_val(pte)); +} + +static void kvm_set_pte_present(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, pte_t pte) +{ + kvm_mmu_write(ptep, pte_val(pte)); +} + +static void kvm_pte_clear(struct mm_struct *mm, + unsigned long addr, pte_t *ptep) +{ + kvm_mmu_write(ptep, 0); +} + +static void kvm_pmd_clear(pmd_t *pmdp) +{ + kvm_mmu_write(pmdp, 0); +} +#endif + +static void kvm_set_pud(pud_t *pudp, pud_t pud) +{ + kvm_mmu_write(pudp, pud_val(pud)); +} + +#if PAGETABLE_LEVELS == 4 +static void kvm_set_pgd(pgd_t *pgdp, pgd_t pgd) +{ + kvm_mmu_write(pgdp, pgd_val(pgd)); +} +#endif +#endif /* PAGETABLE_LEVELS >= 3 */ + +static void kvm_flush_tlb(void) +{ + struct kvm_mmu_op_flush_tlb ftlb = { + .header.op = KVM_MMU_OP_FLUSH_TLB, + }; + + kvm_deferred_mmu_op(&ftlb, sizeof ftlb); +} + +static void kvm_release_pt(u32 pfn) +{ + struct kvm_mmu_op_release_pt rpt = { + .header.op = KVM_MMU_OP_RELEASE_PT, + .pt_phys = (u64)pfn << PAGE_SHIFT, + }; + + kvm_mmu_op(&rpt, sizeof rpt); +} + +static void kvm_enter_lazy_mmu(void) +{ + struct kvm_para_state *state = kvm_para_state(); + + paravirt_enter_lazy_mmu(); + state->mode = paravirt_get_lazy_mode(); +} + +static void kvm_leave_lazy_mmu(void) +{ + struct kvm_para_state *state = kvm_para_state(); + + mmu_queue_flush(state); + paravirt_leave_lazy(paravirt_get_lazy_mode()); + state->mode = paravirt_get_lazy_mode(); +} + +static void paravirt_ops_setup(void) +{ + pv_info.name = "KVM"; + pv_info.paravirt_enabled = 1; + + if (kvm_para_has_feature(KVM_FEATURE_NOP_IO_DELAY)) + pv_cpu_ops.io_delay = kvm_io_delay; + + if (kvm_para_has_feature(KVM_FEATURE_MMU_OP)) { + pv_mmu_ops.set_pte = kvm_set_pte; + pv_mmu_ops.set_pte_at = kvm_set_pte_at; + pv_mmu_ops.set_pmd = kvm_set_pmd; +#if PAGETABLE_LEVELS >= 3 +#ifdef CONFIG_X86_PAE + pv_mmu_ops.set_pte_atomic = kvm_set_pte_atomic; + pv_mmu_ops.set_pte_present = kvm_set_pte_present; + pv_mmu_ops.pte_clear = kvm_pte_clear; + pv_mmu_ops.pmd_clear = kvm_pmd_clear; +#endif + pv_mmu_ops.set_pud = kvm_set_pud; +#if PAGETABLE_LEVELS == 4 + pv_mmu_ops.set_pgd = kvm_set_pgd; +#endif +#endif + pv_mmu_ops.flush_tlb_user = kvm_flush_tlb; + pv_mmu_ops.release_pte = kvm_release_pt; + pv_mmu_ops.release_pmd = kvm_release_pt; + pv_mmu_ops.release_pud = kvm_release_pt; + + pv_mmu_ops.lazy_mode.enter = kvm_enter_lazy_mmu; + pv_mmu_ops.lazy_mode.leave = kvm_leave_lazy_mmu; + } +} + +void __init kvm_guest_init(void) +{ + if (!kvm_para_available()) + return; + + paravirt_ops_setup(); +} diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c new file mode 100644 index 0000000000000000000000000000000000000000..ddee04043aeb36716bc0910f5830e48d852c8b6a --- /dev/null +++ b/arch/x86/kernel/kvmclock.c @@ -0,0 +1,187 @@ +/* KVM paravirtual clock driver. A clocksource implementation + Copyright (C) 2008 Glauber de Oliveira Costa, Red Hat Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA +*/ + +#include +#include +#include +#include +#include +#include +#include + +#define KVM_SCALE 22 + +static int kvmclock = 1; + +static int parse_no_kvmclock(char *arg) +{ + kvmclock = 0; + return 0; +} +early_param("no-kvmclock", parse_no_kvmclock); + +/* The hypervisor will put information about time periodically here */ +static DEFINE_PER_CPU_SHARED_ALIGNED(struct kvm_vcpu_time_info, hv_clock); +#define get_clock(cpu, field) per_cpu(hv_clock, cpu).field + +static inline u64 kvm_get_delta(u64 last_tsc) +{ + int cpu = smp_processor_id(); + u64 delta = native_read_tsc() - last_tsc; + return (delta * get_clock(cpu, tsc_to_system_mul)) >> KVM_SCALE; +} + +static struct kvm_wall_clock wall_clock; +static cycle_t kvm_clock_read(void); +/* + * The wallclock is the time of day when we booted. Since then, some time may + * have elapsed since the hypervisor wrote the data. So we try to account for + * that with system time + */ +unsigned long kvm_get_wallclock(void) +{ + u32 wc_sec, wc_nsec; + u64 delta; + struct timespec ts; + int version, nsec; + int low, high; + + low = (int)__pa(&wall_clock); + high = ((u64)__pa(&wall_clock) >> 32); + + delta = kvm_clock_read(); + + native_write_msr(MSR_KVM_WALL_CLOCK, low, high); + do { + version = wall_clock.wc_version; + rmb(); + wc_sec = wall_clock.wc_sec; + wc_nsec = wall_clock.wc_nsec; + rmb(); + } while ((wall_clock.wc_version != version) || (version & 1)); + + delta = kvm_clock_read() - delta; + delta += wc_nsec; + nsec = do_div(delta, NSEC_PER_SEC); + set_normalized_timespec(&ts, wc_sec + delta, nsec); + /* + * Of all mechanisms of time adjustment I've tested, this one + * was the champion! + */ + return ts.tv_sec + 1; +} + +int kvm_set_wallclock(unsigned long now) +{ + return 0; +} + +/* + * This is our read_clock function. The host puts an tsc timestamp each time + * it updates a new time. Without the tsc adjustment, we can have a situation + * in which a vcpu starts to run earlier (smaller system_time), but probes + * time later (compared to another vcpu), leading to backwards time + */ +static cycle_t kvm_clock_read(void) +{ + u64 last_tsc, now; + int cpu; + + preempt_disable(); + cpu = smp_processor_id(); + + last_tsc = get_clock(cpu, tsc_timestamp); + now = get_clock(cpu, system_time); + + now += kvm_get_delta(last_tsc); + preempt_enable(); + + return now; +} +static struct clocksource kvm_clock = { + .name = "kvm-clock", + .read = kvm_clock_read, + .rating = 400, + .mask = CLOCKSOURCE_MASK(64), + .mult = 1 << KVM_SCALE, + .shift = KVM_SCALE, + .flags = CLOCK_SOURCE_IS_CONTINUOUS, +}; + +static int kvm_register_clock(void) +{ + int cpu = smp_processor_id(); + int low, high; + low = (int)__pa(&per_cpu(hv_clock, cpu)) | 1; + high = ((u64)__pa(&per_cpu(hv_clock, cpu)) >> 32); + + return native_write_msr_safe(MSR_KVM_SYSTEM_TIME, low, high); +} + +static void kvm_setup_secondary_clock(void) +{ + /* + * Now that the first cpu already had this clocksource initialized, + * we shouldn't fail. + */ + WARN_ON(kvm_register_clock()); + /* ok, done with our trickery, call native */ + setup_secondary_APIC_clock(); +} + +/* + * After the clock is registered, the host will keep writing to the + * registered memory location. If the guest happens to shutdown, this memory + * won't be valid. In cases like kexec, in which you install a new kernel, this + * means a random memory location will be kept being written. So before any + * kind of shutdown from our side, we unregister the clock by writting anything + * that does not have the 'enable' bit set in the msr + */ +#ifdef CONFIG_KEXEC +static void kvm_crash_shutdown(struct pt_regs *regs) +{ + native_write_msr_safe(MSR_KVM_SYSTEM_TIME, 0, 0); + native_machine_crash_shutdown(regs); +} +#endif + +static void kvm_shutdown(void) +{ + native_write_msr_safe(MSR_KVM_SYSTEM_TIME, 0, 0); + native_machine_shutdown(); +} + +void __init kvmclock_init(void) +{ + if (!kvm_para_available()) + return; + + if (kvmclock && kvm_para_has_feature(KVM_FEATURE_CLOCKSOURCE)) { + if (kvm_register_clock()) + return; + pv_time_ops.get_wallclock = kvm_get_wallclock; + pv_time_ops.set_wallclock = kvm_set_wallclock; + pv_time_ops.sched_clock = kvm_clock_read; + pv_apic_ops.setup_secondary_clock = kvm_setup_secondary_clock; + machine_ops.shutdown = kvm_shutdown; +#ifdef CONFIG_KEXEC + machine_ops.crash_shutdown = kvm_crash_shutdown; +#endif + clocksource_register(&kvm_clock); + } +} diff --git a/arch/x86/kernel/mfgpt_32.c b/arch/x86/kernel/mfgpt_32.c index b402c0f3f192213f84595197f067c28043545ea1..cfc2648d25ff924e704b9316286f5cca1bae3049 100644 --- a/arch/x86/kernel/mfgpt_32.c +++ b/arch/x86/kernel/mfgpt_32.c @@ -364,7 +364,8 @@ int __init mfgpt_timer_setup(void) geode_mfgpt_write(mfgpt_event_clock, MFGPT_REG_SETUP, val); /* Set up the clock event */ - mfgpt_clockevent.mult = div_sc(MFGPT_HZ, NSEC_PER_SEC, 32); + mfgpt_clockevent.mult = div_sc(MFGPT_HZ, NSEC_PER_SEC, + mfgpt_clockevent.shift); mfgpt_clockevent.min_delta_ns = clockevent_delta2ns(0xF, &mfgpt_clockevent); mfgpt_clockevent.max_delta_ns = clockevent_delta2ns(0xFFFE, diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index 70744e344fa1f7541e249299f1a056d2689cf7c3..3e2c54dc8b29f2e907021bf84d55081b5f65ffa6 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -686,13 +686,11 @@ void __init get_smp_config(void) static int __init smp_scan_config(unsigned long base, unsigned long length, unsigned reserve) { - extern void __bad_mpf_size(void); unsigned int *bp = phys_to_virt(base); struct intel_mp_floating *mpf; Dprintk("Scan SMP from %p for %ld bytes.\n", bp, length); - if (sizeof(*mpf) != 16) - __bad_mpf_size(); + BUILD_BUG_ON(sizeof(*mpf) != 16); while (length > 0) { mpf = (struct intel_mp_floating *)bp; @@ -801,7 +799,6 @@ void __init find_smp_config(void) #ifdef CONFIG_X86_IO_APIC #define MP_ISA_BUS 0 -#define MP_MAX_IOAPIC_PIN 127 extern struct mp_ioapic_routing mp_ioapic_routing[MAX_IO_APICS]; @@ -820,7 +817,7 @@ static int mp_find_ioapic(int gsi) return -1; } -static u8 uniq_ioapic_id(u8 id) +static u8 __init uniq_ioapic_id(u8 id) { #ifdef CONFIG_X86_32 if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && @@ -909,14 +906,7 @@ void __init mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */ intsrc.mpc_dstirq = pin; /* INTIN# */ - Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, %d-%d\n", - intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, - (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, - intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq); - - mp_irqs[mp_irq_entries] = intsrc; - if (++mp_irq_entries == MAX_IRQ_SOURCES) - panic("Max # of irq sources exceeded!\n"); + MP_intsrc_info(&intsrc); } int es7000_plat; @@ -985,23 +975,14 @@ void __init mp_config_acpi_legacy_irqs(void) intsrc.mpc_srcbusirq = i; /* Identity mapped */ intsrc.mpc_dstirq = i; - Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, " - "%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, - (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, - intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, - intsrc.mpc_dstirq); - - mp_irqs[mp_irq_entries] = intsrc; - if (++mp_irq_entries == MAX_IRQ_SOURCES) - panic("Max # of irq sources exceeded!\n"); + MP_intsrc_info(&intsrc); } } int mp_register_gsi(u32 gsi, int triggering, int polarity) { - int ioapic = -1; - int ioapic_pin = 0; - int idx, bit = 0; + int ioapic; + int ioapic_pin; #ifdef CONFIG_X86_32 #define MAX_GSI_NUM 4096 #define IRQ_COMPRESSION_START 64 @@ -1041,15 +1022,13 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) * with redundant pin->gsi mappings (but unique PCI devices); * we only program the IOAPIC on the first. */ - bit = ioapic_pin % 32; - idx = (ioapic_pin < 32) ? 0 : (ioapic_pin / 32); - if (idx > 3) { + if (ioapic_pin > MP_MAX_IOAPIC_PIN) { printk(KERN_ERR "Invalid reference to IOAPIC pin " "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, ioapic_pin); return gsi; } - if ((1 << bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) { + if (test_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed)) { Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", mp_ioapic_routing[ioapic].apic_id, ioapic_pin); #ifdef CONFIG_X86_32 @@ -1059,7 +1038,7 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) #endif } - mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1 << bit); + set_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed); #ifdef CONFIG_X86_32 /* * For GSI >= 64, use IRQ compression diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 2edee22e9c307610f71b47799c792ea8009713be..e28ec497e142f49cc4847f833ff5a5f8f96833f6 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -43,6 +43,7 @@ #include #include #include +#include #ifdef CONFIG_CALGARY_IOMMU_ENABLED_BY_DEFAULT int use_calgary __read_mostly = 1; diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 3004d716539dde2b882acd4e0576676531382965..67e9b4a1e89d5a1ac3a3851beaab9d48bbe22c0b 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -4,6 +4,8 @@ #include #include #include +#include +#include struct kmem_cache *task_xstate_cachep; @@ -42,3 +44,118 @@ void arch_task_cache_init(void) __alignof__(union thread_xstate), SLAB_PANIC, NULL); } + +static void do_nothing(void *unused) +{ +} + +/* + * cpu_idle_wait - Used to ensure that all the CPUs discard old value of + * pm_idle and update to new pm_idle value. Required while changing pm_idle + * handler on SMP systems. + * + * Caller must have changed pm_idle to the new value before the call. Old + * pm_idle value will not be used by any CPU after the return of this function. + */ +void cpu_idle_wait(void) +{ + smp_mb(); + /* kick all the CPUs so that they exit out of pm_idle */ + smp_call_function(do_nothing, NULL, 0, 1); +} +EXPORT_SYMBOL_GPL(cpu_idle_wait); + +/* + * This uses new MONITOR/MWAIT instructions on P4 processors with PNI, + * which can obviate IPI to trigger checking of need_resched. + * We execute MONITOR against need_resched and enter optimized wait state + * through MWAIT. Whenever someone changes need_resched, we would be woken + * up from MWAIT (without an IPI). + * + * New with Core Duo processors, MWAIT can take some hints based on CPU + * capability. + */ +void mwait_idle_with_hints(unsigned long ax, unsigned long cx) +{ + if (!need_resched()) { + __monitor((void *)¤t_thread_info()->flags, 0, 0); + smp_mb(); + if (!need_resched()) + __mwait(ax, cx); + } +} + +/* Default MONITOR/MWAIT with no hints, used for default C1 state */ +static void mwait_idle(void) +{ + if (!need_resched()) { + __monitor((void *)¤t_thread_info()->flags, 0, 0); + smp_mb(); + if (!need_resched()) + __sti_mwait(0, 0); + else + local_irq_enable(); + } else + local_irq_enable(); +} + + +static int __cpuinit mwait_usable(const struct cpuinfo_x86 *c) +{ + if (force_mwait) + return 1; + /* Any C1 states supported? */ + return c->cpuid_level >= 5 && ((cpuid_edx(5) >> 4) & 0xf) > 0; +} + +/* + * On SMP it's slightly faster (but much more power-consuming!) + * to poll the ->work.need_resched flag instead of waiting for the + * cross-CPU IPI to arrive. Use this option with caution. + */ +static void poll_idle(void) +{ + local_irq_enable(); + cpu_relax(); +} + +void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) +{ + static int selected; + + if (selected) + return; +#ifdef CONFIG_X86_SMP + if (pm_idle == poll_idle && smp_num_siblings > 1) { + printk(KERN_WARNING "WARNING: polling idle and HT enabled," + " performance may degrade.\n"); + } +#endif + if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) { + /* + * Skip, if setup has overridden idle. + * One CPU supports mwait => All CPUs supports mwait + */ + if (!pm_idle) { + printk(KERN_INFO "using mwait in idle threads.\n"); + pm_idle = mwait_idle; + } + } + selected = 1; +} + +static int __init idle_setup(char *str) +{ + if (!strcmp(str, "poll")) { + printk("using polling idle threads.\n"); + pm_idle = poll_idle; + } else if (!strcmp(str, "mwait")) + force_mwait = 1; + else + return -1; + + boot_option_idle_override = 1; + return 0; +} +early_param("idle", idle_setup); + diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 77de848bd1fbbca9d498a07f4c0fbc65cf05c43b..f8476dfbb60d9818f5da08d3a5b95a223930964f 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -111,12 +111,10 @@ void default_idle(void) */ smp_mb(); - local_irq_disable(); - if (!need_resched()) { + if (!need_resched()) safe_halt(); /* enables interrupts racelessly */ - local_irq_disable(); - } - local_irq_enable(); + else + local_irq_enable(); current_thread_info()->status |= TS_POLLING; } else { local_irq_enable(); @@ -128,17 +126,6 @@ void default_idle(void) EXPORT_SYMBOL(default_idle); #endif -/* - * On SMP it's slightly faster (but much more power-consuming!) - * to poll the ->work.need_resched flag instead of waiting for the - * cross-CPU IPI to arrive. Use this option with caution. - */ -static void poll_idle(void) -{ - local_irq_enable(); - cpu_relax(); -} - #ifdef CONFIG_HOTPLUG_CPU #include /* We don't actually take CPU down, just spin without interrupts. */ @@ -196,6 +183,7 @@ void cpu_idle(void) if (cpu_is_offline(cpu)) play_dead(); + local_irq_disable(); __get_cpu_var(irq_stat).idle_timestamp = jiffies; idle(); } @@ -206,104 +194,6 @@ void cpu_idle(void) } } -static void do_nothing(void *unused) -{ -} - -/* - * cpu_idle_wait - Used to ensure that all the CPUs discard old value of - * pm_idle and update to new pm_idle value. Required while changing pm_idle - * handler on SMP systems. - * - * Caller must have changed pm_idle to the new value before the call. Old - * pm_idle value will not be used by any CPU after the return of this function. - */ -void cpu_idle_wait(void) -{ - smp_mb(); - /* kick all the CPUs so that they exit out of pm_idle */ - smp_call_function(do_nothing, NULL, 0, 1); -} -EXPORT_SYMBOL_GPL(cpu_idle_wait); - -/* - * This uses new MONITOR/MWAIT instructions on P4 processors with PNI, - * which can obviate IPI to trigger checking of need_resched. - * We execute MONITOR against need_resched and enter optimized wait state - * through MWAIT. Whenever someone changes need_resched, we would be woken - * up from MWAIT (without an IPI). - * - * New with Core Duo processors, MWAIT can take some hints based on CPU - * capability. - */ -void mwait_idle_with_hints(unsigned long ax, unsigned long cx) -{ - if (!need_resched()) { - __monitor((void *)¤t_thread_info()->flags, 0, 0); - smp_mb(); - if (!need_resched()) - __sti_mwait(ax, cx); - else - local_irq_enable(); - } else - local_irq_enable(); -} - -/* Default MONITOR/MWAIT with no hints, used for default C1 state */ -static void mwait_idle(void) -{ - local_irq_enable(); - mwait_idle_with_hints(0, 0); -} - -static int __cpuinit mwait_usable(const struct cpuinfo_x86 *c) -{ - if (force_mwait) - return 1; - /* Any C1 states supported? */ - return c->cpuid_level >= 5 && ((cpuid_edx(5) >> 4) & 0xf) > 0; -} - -void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) -{ - static int selected; - - if (selected) - return; -#ifdef CONFIG_X86_SMP - if (pm_idle == poll_idle && smp_num_siblings > 1) { - printk(KERN_WARNING "WARNING: polling idle and HT enabled," - " performance may degrade.\n"); - } -#endif - if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) { - /* - * Skip, if setup has overridden idle. - * One CPU supports mwait => All CPUs supports mwait - */ - if (!pm_idle) { - printk(KERN_INFO "using mwait in idle threads.\n"); - pm_idle = mwait_idle; - } - } - selected = 1; -} - -static int __init idle_setup(char *str) -{ - if (!strcmp(str, "poll")) { - printk("using polling idle threads.\n"); - pm_idle = poll_idle; - } else if (!strcmp(str, "mwait")) - force_mwait = 1; - else - return -1; - - boot_option_idle_override = 1; - return 0; -} -early_param("idle", idle_setup); - void __show_registers(struct pt_regs *regs, int all) { unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L; diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 131c2ee7ac5618e9fc92f576f4db19676bd235af..e2319f39988b042f364a73174a0d2ac254a5b60d 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -106,26 +106,13 @@ void default_idle(void) * test NEED_RESCHED: */ smp_mb(); - local_irq_disable(); - if (!need_resched()) { + if (!need_resched()) safe_halt(); /* enables interrupts racelessly */ - local_irq_disable(); - } - local_irq_enable(); + else + local_irq_enable(); current_thread_info()->status |= TS_POLLING; } -/* - * On SMP it's slightly faster (but much more power-consuming!) - * to poll the ->need_resched flag instead of waiting for the - * cross-CPU IPI to arrive. Use this option with caution. - */ -static void poll_idle(void) -{ - local_irq_enable(); - cpu_relax(); -} - #ifdef CONFIG_HOTPLUG_CPU DECLARE_PER_CPU(int, cpu_state); @@ -192,110 +179,6 @@ void cpu_idle(void) } } -static void do_nothing(void *unused) -{ -} - -/* - * cpu_idle_wait - Used to ensure that all the CPUs discard old value of - * pm_idle and update to new pm_idle value. Required while changing pm_idle - * handler on SMP systems. - * - * Caller must have changed pm_idle to the new value before the call. Old - * pm_idle value will not be used by any CPU after the return of this function. - */ -void cpu_idle_wait(void) -{ - smp_mb(); - /* kick all the CPUs so that they exit out of pm_idle */ - smp_call_function(do_nothing, NULL, 0, 1); -} -EXPORT_SYMBOL_GPL(cpu_idle_wait); - -/* - * This uses new MONITOR/MWAIT instructions on P4 processors with PNI, - * which can obviate IPI to trigger checking of need_resched. - * We execute MONITOR against need_resched and enter optimized wait state - * through MWAIT. Whenever someone changes need_resched, we would be woken - * up from MWAIT (without an IPI). - * - * New with Core Duo processors, MWAIT can take some hints based on CPU - * capability. - */ -void mwait_idle_with_hints(unsigned long ax, unsigned long cx) -{ - if (!need_resched()) { - __monitor((void *)¤t_thread_info()->flags, 0, 0); - smp_mb(); - if (!need_resched()) - __mwait(ax, cx); - } -} - -/* Default MONITOR/MWAIT with no hints, used for default C1 state */ -static void mwait_idle(void) -{ - if (!need_resched()) { - __monitor((void *)¤t_thread_info()->flags, 0, 0); - smp_mb(); - if (!need_resched()) - __sti_mwait(0, 0); - else - local_irq_enable(); - } else { - local_irq_enable(); - } -} - - -static int __cpuinit mwait_usable(const struct cpuinfo_x86 *c) -{ - if (force_mwait) - return 1; - /* Any C1 states supported? */ - return c->cpuid_level >= 5 && ((cpuid_edx(5) >> 4) & 0xf) > 0; -} - -void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) -{ - static int selected; - - if (selected) - return; -#ifdef CONFIG_X86_SMP - if (pm_idle == poll_idle && smp_num_siblings > 1) { - printk(KERN_WARNING "WARNING: polling idle and HT enabled," - " performance may degrade.\n"); - } -#endif - if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) { - /* - * Skip, if setup has overridden idle. - * One CPU supports mwait => All CPUs supports mwait - */ - if (!pm_idle) { - printk(KERN_INFO "using mwait in idle threads.\n"); - pm_idle = mwait_idle; - } - } - selected = 1; -} - -static int __init idle_setup(char *str) -{ - if (!strcmp(str, "poll")) { - printk("using polling idle threads.\n"); - pm_idle = poll_idle; - } else if (!strcmp(str, "mwait")) - force_mwait = 1; - else - return -1; - - boot_option_idle_override = 1; - return 0; -} -early_param("idle", idle_setup); - /* Prints also some state that isn't saved in the pt_regs */ void __show_regs(struct pt_regs * regs) { diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 559c1b027417f66823d8f214b45c61f31bbb77f1..fb03ef380f0e230ebf5f3835bc7ec57e1c9fed20 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -1207,96 +1207,15 @@ static int genregs32_set(struct task_struct *target, return ret; } -static long ptrace32_siginfo(unsigned request, u32 pid, u32 addr, u32 data) +long compat_arch_ptrace(struct task_struct *child, compat_long_t request, + compat_ulong_t caddr, compat_ulong_t cdata) { - siginfo_t __user *si = compat_alloc_user_space(sizeof(siginfo_t)); - compat_siginfo_t __user *si32 = compat_ptr(data); - siginfo_t ssi; - int ret; - - if (request == PTRACE_SETSIGINFO) { - memset(&ssi, 0, sizeof(siginfo_t)); - ret = copy_siginfo_from_user32(&ssi, si32); - if (ret) - return ret; - if (copy_to_user(si, &ssi, sizeof(siginfo_t))) - return -EFAULT; - } - ret = sys_ptrace(request, pid, addr, (unsigned long)si); - if (ret) - return ret; - if (request == PTRACE_GETSIGINFO) { - if (copy_from_user(&ssi, si, sizeof(siginfo_t))) - return -EFAULT; - ret = copy_siginfo_to_user32(si32, &ssi); - } - return ret; -} - -asmlinkage long sys32_ptrace(long request, u32 pid, u32 addr, u32 data) -{ - struct task_struct *child; - struct pt_regs *childregs; + unsigned long addr = caddr; + unsigned long data = cdata; void __user *datap = compat_ptr(data); int ret; __u32 val; - switch (request) { - case PTRACE_TRACEME: - case PTRACE_ATTACH: - case PTRACE_KILL: - case PTRACE_CONT: - case PTRACE_SINGLESTEP: - case PTRACE_SINGLEBLOCK: - case PTRACE_DETACH: - case PTRACE_SYSCALL: - case PTRACE_OLDSETOPTIONS: - case PTRACE_SETOPTIONS: - case PTRACE_SET_THREAD_AREA: - case PTRACE_GET_THREAD_AREA: -#ifdef X86_BTS - case PTRACE_BTS_CONFIG: - case PTRACE_BTS_STATUS: - case PTRACE_BTS_SIZE: - case PTRACE_BTS_GET: - case PTRACE_BTS_CLEAR: - case PTRACE_BTS_DRAIN: -#endif - return sys_ptrace(request, pid, addr, data); - - default: - return -EINVAL; - - case PTRACE_PEEKTEXT: - case PTRACE_PEEKDATA: - case PTRACE_POKEDATA: - case PTRACE_POKETEXT: - case PTRACE_POKEUSR: - case PTRACE_PEEKUSR: - case PTRACE_GETREGS: - case PTRACE_SETREGS: - case PTRACE_SETFPREGS: - case PTRACE_GETFPREGS: - case PTRACE_SETFPXREGS: - case PTRACE_GETFPXREGS: - case PTRACE_GETEVENTMSG: - break; - - case PTRACE_SETSIGINFO: - case PTRACE_GETSIGINFO: - return ptrace32_siginfo(request, pid, addr, data); - } - - child = ptrace_get_task_struct(pid); - if (IS_ERR(child)) - return PTR_ERR(child); - - ret = ptrace_check_attach(child, request == PTRACE_KILL); - if (ret < 0) - goto out; - - childregs = task_pt_regs(child); - switch (request) { case PTRACE_PEEKUSR: ret = getreg32(child, addr, &val); @@ -1343,12 +1262,14 @@ asmlinkage long sys32_ptrace(long request, u32 pid, u32 addr, u32 data) sizeof(struct user32_fxsr_struct), datap); + case PTRACE_GET_THREAD_AREA: + case PTRACE_SET_THREAD_AREA: + return arch_ptrace(child, request, addr, data); + default: return compat_ptrace_request(child, request, addr, data); } - out: - put_task_struct(child); return ret; } diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 1791a751a772d0d6d1a6256591707ad5fca8c237..a4a838306b2cc69f2856f1b8244b6045c86165c9 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -399,7 +399,7 @@ static void native_machine_emergency_restart(void) } } -static void native_machine_shutdown(void) +void native_machine_shutdown(void) { /* Stop the cpus and apics */ #ifdef CONFIG_SMP @@ -470,7 +470,10 @@ struct machine_ops machine_ops = { .shutdown = native_machine_shutdown, .emergency_restart = native_machine_emergency_restart, .restart = native_machine_restart, - .halt = native_machine_halt + .halt = native_machine_halt, +#ifdef CONFIG_KEXEC + .crash_shutdown = native_machine_crash_shutdown, +#endif }; void machine_power_off(void) @@ -498,3 +501,9 @@ void machine_halt(void) machine_ops.halt(); } +#ifdef CONFIG_KEXEC +void machine_crash_shutdown(struct pt_regs *regs) +{ + machine_ops.crash_shutdown(regs); +} +#endif diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c index 455d3c80960bf93e63d0914c56a170f88482ab70..2283422af7946fef30a86189dddb0a7a12fe3fc1 100644 --- a/arch/x86/kernel/setup_32.c +++ b/arch/x86/kernel/setup_32.c @@ -47,6 +47,7 @@ #include #include #include +#include #include