drm/nouveau/instmem/gk20a: fix crash during error path
authorAlexandre Courbot <acourbot@nvidia.com>
Tue, 10 Mar 2015 06:46:08 +0000 (15:46 +0900)
committerBen Skeggs <bskeggs@redhat.com>
Tue, 14 Apr 2015 07:00:49 +0000 (17:00 +1000)
If a memory allocation fails when using the DMA allocator,
gk20a_instobj_dtor_dma() will be called on the failed instmem object.
At this time, node->handle might not be NULL despite the call to
dma_alloc_attrs() having failed. node->cpuaddr is the right member to
check for such a failure, so use it instead.

Reported-by: Vince Hsu <vinceh@nvidia.com>
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Reviewed-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c

index fcba72eb74a3198c07f00ea714fa2e040c143ded..dd0994d9ebfc3f377141de6662b57130d83e0b2b 100644 (file)
@@ -148,7 +148,7 @@ gk20a_instobj_dtor_dma(struct gk20a_instobj_priv *_node)
        struct gk20a_instmem_priv *priv = (void *)nvkm_instmem(node);
        struct device *dev = nv_device_base(nv_device(priv));
 
-       if (unlikely(!node->handle))
+       if (unlikely(!node->cpuaddr))
                return;
 
        dma_free_attrs(dev, _node->mem->size << PAGE_SHIFT, node->cpuaddr,