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

Commit a9162f23 authored by Carsten Otte's avatar Carsten Otte Committed by Martin Schwidefsky
Browse files

[S390] fix possible deadlock in gmap_map_segment 



Fix possible deadlock reported by lockdep:
qemu-system-s39/2963 is trying to acquire lock:
(&mm->mmap_sem){++++++}, at: gmap_alloc_table+0x9c/0x120
but task is already holding lock:
(&mm->mmap_sem){++++++}, at: gmap_map_segment+0xa6/0x27c

Actually gmap_alloc_table is the only called in gmap_map_segment with
mmap_sem held, thus it's safe to simply remove the inner lock.

Signed-off-by: default avatarCarsten Otte <cotte@de.ibm.com>
Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
parent 69ba9743

arch/s390/mm/pgtable.c

+3 −2
Original line number Diff line number Diff line
@@ -256,6 +256,9 @@ void gmap_disable(struct gmap *gmap) 
}

EXPORT_SYMBOL_GPL(gmap_disable);



/*

 * gmap_alloc_table is assumed to be called with mmap_sem held

 */

static int gmap_alloc_table(struct gmap *gmap,

			       unsigned long *table, unsigned long init)

{
@@ -267,14 +270,12 @@ static int gmap_alloc_table(struct gmap *gmap, 
		return -ENOMEM;

	new = (unsigned long *) page_to_phys(page);

	crst_table_init(new, init);

	down_read(&gmap->mm->mmap_sem);

	if (*table & _REGION_ENTRY_INV) {

		list_add(&page->lru, &gmap->crst_list);

		*table = (unsigned long) new | _REGION_ENTRY_LENGTH |

			(*table & _REGION_ENTRY_TYPE_MASK);

	} else

		__free_pages(page, ALLOC_ORDER);

	up_read(&gmap->mm->mmap_sem);

	return 0;

}