Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit b5eb7138 authored by Neeti Desai's avatar Neeti Desai Committed by Matt Wagantall
Browse files

arm64: dma-mapping: Split large pages when allocating in atomic context 



In atomic context, gen_pool_alloc allocates a single page large
enough to accomodate the requested size. However __iommu_create_mapping
always maps pages assuming they are of size 4K. Thus only the first
4K of the buffer is mapped and a translation fault is generated
during an unmap.

Fix this by splitting the larger pages into 4K pages.

Change-Id: Ifcbe29477ad210204028486bd011470fe8b50852
Signed-off-by: default avatarNeeti Desai <neetid@codeaurora.org>
Signed-off-by: default avatarMitchel Humpherys <mitchelh@codeaurora.org>
parent 267ca264

arch/arm64/mm/dma-mapping.c

+42 −13
Original line number Diff line number Diff line
@@ -60,10 +60,12 @@ static int __init early_coherent_pool(char *p) 
}

early_param("coherent_pool", early_coherent_pool);



static void *__alloc_from_pool(size_t size, struct page **ret_page)

static void *__alloc_from_pool(size_t size, struct page **ret_pages)

{

	unsigned long val;

	void *ptr = NULL;

	int count = size >> PAGE_SHIFT;

	int i;



	if (!atomic_pool) {

		WARN(1, "coherent pool not initialised!\n");
@@ -73,8 +75,10 @@ static void *__alloc_from_pool(size_t size, struct page **ret_page) 
	val = gen_pool_alloc(atomic_pool, size);

	if (val) {

		phys_addr_t phys = gen_pool_virt_to_phys(atomic_pool, val);



		*ret_page = phys_to_page(phys);

		for (i = 0; i < count ; i++) {

			ret_pages[i] = phys_to_page(phys);

			phys += 1 << PAGE_SHIFT;

		}

		ptr = (void *)val;

	}
@@ -203,11 +207,23 @@ static void *__dma_alloc_noncoherent(struct device *dev, size_t size, 
	size = PAGE_ALIGN(size);



	if (!(flags & __GFP_WAIT)) {

		struct page *page = NULL;

		void *addr = __alloc_from_pool(size, &page);

		struct page **page = NULL;

		int count = size >> PAGE_SHIFT;

		int array_size = count * sizeof(struct page *);

		void *addr;



		if (array_size <= PAGE_SIZE)

			page = kzalloc(array_size, flags);

		else

			page = vzalloc(array_size);



		if (!page)

			return NULL;



		addr = __alloc_from_pool(size, page);



		if (addr)

			*dma_handle = phys_to_dma(dev, page_to_phys(page));

			*dma_handle = phys_to_dma(dev, page_to_phys(*page));



		return addr;
@@ -847,24 +863,37 @@ static struct page **__iommu_get_pages(void *cpu_addr, struct dma_attrs *attrs) 
	return NULL;

}



static void *__iommu_alloc_atomic(struct device *dev, size_t size,

				  dma_addr_t *handle)

void *__iommu_alloc_atomic(struct device *dev, size_t size,

				  dma_addr_t *handle, gfp_t gfp)

{

	struct page *page;

	struct page **pages;

	int count = size >> PAGE_SHIFT;

	int array_size = count * sizeof(struct page *);

	void *addr;



	addr = __alloc_from_pool(size, &page);

	if (!addr)

	if (array_size <= PAGE_SIZE)

		pages = kzalloc(array_size, gfp);

	else

		pages = vzalloc(array_size);



	if (!pages)

		return NULL;



	*handle = __iommu_create_mapping(dev, &page, size);

	addr = __alloc_from_pool(size, pages);

	if (!addr)

		goto err_free;



	*handle = __iommu_create_mapping(dev, pages, size);

	if (*handle == DMA_ERROR_CODE)

		goto err_mapping;



	kvfree(pages);

	return addr;



err_mapping:

	__free_from_pool(addr, size);

err_free:

	kvfree(pages);

	return NULL;

}
@@ -886,7 +915,7 @@ static void *arm_iommu_alloc_attrs(struct device *dev, size_t size, 
	size = PAGE_ALIGN(size);



	if (!(gfp & __GFP_WAIT))

		return __iommu_alloc_atomic(dev, size, handle);

		return __iommu_alloc_atomic(dev, size, handle, gfp);



	/*

	 * Following is a work-around (a.k.a. hack) to prevent pages