On 07/05/ , Matthew Auld wrote:
> On Wed, 5 Jul 2023 at 11:08, Lang Yu <Lang.Yu@xxxxxxx> wrote:
> >
> > bo->kref is increased once(kref_init()) in ttm_bo_release,
> > but decreased twice(ttm_bo_put()) respectively in
> > ttm_bo_delayed_delete and ttm_bo_cleanup_refs,
> > which is unbalanced.
> >
> > Just clean up bo resource in one place for a delayed deleted bo.
> >
> > Fixes: 9bff18d13473 ("drm/ttm: use per BO cleanup workers")
> >
> > [ 67.399887] refcount_t: underflow; use-after-free.
> > [ 67.399901] WARNING: CPU: 0 PID: 3172 at lib/refcount.c:28 refcount_warn_saturate+0xc2/0x110
> > [ 67.400124] RIP: 0010:refcount_warn_saturate+0xc2/0x110
> > [ 67.400173] Call Trace:
> > [ 67.400176] <TASK>
> > [ 67.400181] ttm_mem_evict_first+0x4fe/0x5b0 [ttm]
> > [ 67.400216] ttm_bo_mem_space+0x1e3/0x240 [ttm]
> > [ 67.400239] ttm_bo_validate+0xc7/0x190 [ttm]
> > [ 67.400253] ? ww_mutex_trylock+0x1b1/0x390
> > [ 67.400266] ttm_bo_init_reserved+0x183/0x1c0 [ttm]
> > [ 67.400280] ? __rwlock_init+0x3d/0x70
> > [ 67.400292] amdgpu_bo_create+0x1cd/0x4f0 [amdgpu]
> > [ 67.400607] ? __pfx_amdgpu_bo_user_destroy+0x10/0x10 [amdgpu]
> > [ 67.400980] amdgpu_bo_create_user+0x38/0x70 [amdgpu]
> > [ 67.401291] amdgpu_gem_object_create+0x77/0xb0 [amdgpu]
> > [ 67.401641] ? __pfx_amdgpu_bo_user_destroy+0x10/0x10 [amdgpu]
> > [ 67.401958] amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu+0x228/0xa30 [amdgpu]
> > [ 67.402433] kfd_ioctl_alloc_memory_of_gpu+0x14e/0x390 [amdgpu]
> > [ 67.402824] ? lock_release+0x13f/0x290
> > [ 67.402838] kfd_ioctl+0x1e0/0x640 [amdgpu]
> > [ 67.403205] ? __pfx_kfd_ioctl_alloc_memory_of_gpu+0x10/0x10 [amdgpu]
> > [ 67.403579] ? tomoyo_file_ioctl+0x19/0x20
> > [ 67.403590] __x64_sys_ioctl+0x95/0xd0
> > [ 67.403601] do_syscall_64+0x3b/0x90
> > [ 67.403609] entry_SYSCALL_64_after_hwframe+0x72/0xdc
> >
> > Signed-off-by: Lang Yu <Lang.Yu@xxxxxxx>
> > ---
> > drivers/gpu/drm/ttm/ttm_bo.c | 89 ++++--------------------------------
> > 1 file changed, 10 insertions(+), 79 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
> > index 326a3d13a829..1e073dfb1332 100644
> > --- a/drivers/gpu/drm/ttm/ttm_bo.c
> > +++ b/drivers/gpu/drm/ttm/ttm_bo.c
> > @@ -224,82 +224,6 @@ static void ttm_bo_flush_all_fences(struct ttm_buffer_object *bo)
> > dma_resv_iter_end(&cursor);
> > }
> >
> > -/**
> > - * ttm_bo_cleanup_refs
> > - * If bo idle, remove from lru lists, and unref.
> > - * If not idle, block if possible.
> > - *
> > - * Must be called with lru_lock and reservation held, this function
> > - * will drop the lru lock and optionally the reservation lock before returning.
> > - *
> > - * @bo: The buffer object to clean-up
> > - * @interruptible: Any sleeps should occur interruptibly.
> > - * @no_wait_gpu: Never wait for gpu. Return -EBUSY instead.
> > - * @unlock_resv: Unlock the reservation lock as well.
> > - */
> > -
> > -static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
> > - bool interruptible, bool no_wait_gpu,
> > - bool unlock_resv)
> > -{
> > - struct dma_resv *resv = &bo->base._resv;
> > - int ret;
> > -
> > - if (dma_resv_test_signaled(resv, DMA_RESV_USAGE_BOOKKEEP))
> > - ret = 0;
> > - else
> > - ret = -EBUSY;
> > -
> > - if (ret && !no_wait_gpu) {
> > - long lret;
> > -
> > - if (unlock_resv)
> > - dma_resv_unlock(bo->base.resv);
> > - spin_unlock(&bo->bdev->lru_lock);
> > -
> > - lret = dma_resv_wait_timeout(resv, DMA_RESV_USAGE_BOOKKEEP,
> > - interruptible,
> > - 30 * HZ);
> > -
> > - if (lret < 0)
> > - return lret;
> > - else if (lret == 0)
> > - return -EBUSY;
> > -
> > - spin_lock(&bo->bdev->lru_lock);
> > - if (unlock_resv && !dma_resv_trylock(bo->base.resv)) {
> > - /*
> > - * We raced, and lost, someone else holds the reservation now,
> > - * and is probably busy in ttm_bo_cleanup_memtype_use.
> > - *
> > - * Even if it's not the case, because we finished waiting any
> > - * delayed destruction would succeed, so just return success
> > - * here.
> > - */
> > - spin_unlock(&bo->bdev->lru_lock);
> > - return 0;
> > - }
> > - ret = 0;
> > - }
> > -
> > - if (ret) {
> > - if (unlock_resv)
> > - dma_resv_unlock(bo->base.resv);
> > - spin_unlock(&bo->bdev->lru_lock);
> > - return ret;
> > - }
> > -
> > - spin_unlock(&bo->bdev->lru_lock);
> > - ttm_bo_cleanup_memtype_use(bo);
> > -
> > - if (unlock_resv)
> > - dma_resv_unlock(bo->base.resv);
> > -
> > - ttm_bo_put(bo);
>
> The put() here is indeed broken and leads to some nasty uaf I think,
> but we fixed that a while back in: c00133a9e87e ("drm/ttm: drop extra
> ttm_bo_put in ttm_bo_cleanup_refs"). It looks like you are using an
> old tree? Perhaps the issue you are seeing was also fixed with that?
Thanks. I can see this commit in my tree but it was overrode by other
patches. It fixed this issue.
Regards,
Lang
> > -
> > - return 0;
> > -}
> > -
> > /*
> > * Block for the dma_resv object to become idle, lock the buffer and clean up
> > * the resource and tt object.
> > @@ -622,8 +546,10 @@ int ttm_mem_evict_first(struct ttm_device *bdev,
> > }
> >
> > if (bo->deleted) {
> > - ret = ttm_bo_cleanup_refs(bo, ctx->interruptible,
> > - ctx->no_wait_gpu, locked);
> > + if (locked)
> > + dma_resv_unlock(bo->base.resv);
> > + spin_unlock(&bdev->lru_lock);
> > + ret = ttm_bo_wait_ctx(bo, ctx);
> > ttm_bo_put(bo);
> > return ret;
> > }
> > @@ -1146,7 +1072,12 @@ int ttm_bo_swapout(struct ttm_buffer_object *bo, struct ttm_operation_ctx *ctx,
> > }
> >
> > if (bo->deleted) {
> > - ret = ttm_bo_cleanup_refs(bo, false, false, locked);
> > + struct ttm_operation_ctx ctx = { false, false };
> > +
> > + if (locked)
> > + dma_resv_unlock(bo->base.resv);
> > + spin_unlock(&bo->bdev->lru_lock);
> > + ret = ttm_bo_wait_ctx(bo, &ctx);
> > ttm_bo_put(bo);
> > return ret == -EBUSY ? -ENOSPC : ret;
> > }
> > --
> > 2.25.1
> >
