| Message ID | 9a0b09e6b5851f0d4428b72dd6b8b4c0d0ef4206.1572293305.git.robin.murphy@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm/panfrost: Don't dereference bogus MMU pointers | expand |
On Mon, Oct 28, 2019 at 3:08 PM Robin Murphy <robin.murphy@arm.com> wrote: > > It seems that killing an application while faults are occurring > (particularly with a GPU in FPGA at a whopping 40MHz) can lead to > handling a lingering page fault after all the address space contexts > have already been freed. In this situation, the LRU list is empty so > addr_to_drm_mm_node() ends up dereferencing the list head as if it were > a struct panfrost_mmu entry; this leaves "mmu->as" actually pointing at > the pfdev->alloc_mask bitmap, which is also empty, and given that the > fault has a high likelihood of being in AS0, hilarity ensues. > > Sadly, the cleanest solution seems to involve another goto. Oh well, at > least it's robust... > > Signed-off-by: Robin Murphy <robin.murphy@arm.com> > --- > drivers/gpu/drm/panfrost/panfrost_mmu.c | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) Applied to drm-misc-fixes Rob
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c b/drivers/gpu/drm/panfrost/panfrost_mmu.c index e61984e26e0a..508765f80cfe 100644 --- a/drivers/gpu/drm/panfrost/panfrost_mmu.c +++ b/drivers/gpu/drm/panfrost/panfrost_mmu.c @@ -413,11 +413,11 @@ addr_to_drm_mm_node(struct panfrost_device *pfdev, int as, u64 addr) spin_lock(&pfdev->as_lock); list_for_each_entry(mmu, &pfdev->as_lru_list, list) { if (as == mmu->as) - break; + goto found_mmu; } - if (as != mmu->as) - goto out; + goto out; +found_mmu: priv = container_of(mmu, struct panfrost_file_priv, mmu); spin_lock(&priv->mm_lock);
It seems that killing an application while faults are occurring (particularly with a GPU in FPGA at a whopping 40MHz) can lead to handling a lingering page fault after all the address space contexts have already been freed. In this situation, the LRU list is empty so addr_to_drm_mm_node() ends up dereferencing the list head as if it were a struct panfrost_mmu entry; this leaves "mmu->as" actually pointing at the pfdev->alloc_mask bitmap, which is also empty, and given that the fault has a high likelihood of being in AS0, hilarity ensues. Sadly, the cleanest solution seems to involve another goto. Oh well, at least it's robust... Signed-off-by: Robin Murphy <robin.murphy@arm.com> --- drivers/gpu/drm/panfrost/panfrost_mmu.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)