| Message ID | 1483627291-7532-1-git-send-email-matthew.auld@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, Jan 05, 2017 at 02:41:31PM +0000, Matthew Auld wrote: > If we find a suitable victim node on our first pass, then ret > will be uninitialized which could lead to some funny business later. > > Fixes: 9332f3b1b99a ("drm/i915: Combine loops within i915_gem_evict_something") > Cc: Chris Wilson <chris@chris-wilson.co.uk> > Signed-off-by: Matthew Auld <matthew.auld@intel.com> > --- > drivers/gpu/drm/i915/i915_gem_evict.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c > index 50129ec1caab..19716548c455 100644 > --- a/drivers/gpu/drm/i915/i915_gem_evict.c > +++ b/drivers/gpu/drm/i915/i915_gem_evict.c > @@ -109,7 +109,7 @@ i915_gem_evict_something(struct i915_address_space *vm, > }, **phase; > struct i915_vma *vma, *next; > struct drm_mm_node *node; > - int ret; > + int ret = 0; Please don't randomly initialise locals. It is meant to be initialised just prior to the unbind loop. It appears that my patches are out-of-order. :| -Chris
On 5 January 2017 at 15:13, Chris Wilson <chris@chris-wilson.co.uk> wrote: > On Thu, Jan 05, 2017 at 02:41:31PM +0000, Matthew Auld wrote: >> If we find a suitable victim node on our first pass, then ret >> will be uninitialized which could lead to some funny business later. >> >> Fixes: 9332f3b1b99a ("drm/i915: Combine loops within i915_gem_evict_something") >> Cc: Chris Wilson <chris@chris-wilson.co.uk> >> Signed-off-by: Matthew Auld <matthew.auld@intel.com> >> --- >> drivers/gpu/drm/i915/i915_gem_evict.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c >> index 50129ec1caab..19716548c455 100644 >> --- a/drivers/gpu/drm/i915/i915_gem_evict.c >> +++ b/drivers/gpu/drm/i915/i915_gem_evict.c >> @@ -109,7 +109,7 @@ i915_gem_evict_something(struct i915_address_space *vm, >> }, **phase; >> struct i915_vma *vma, *next; >> struct drm_mm_node *node; >> - int ret; >> + int ret = 0; > > Please don't randomly initialise locals. It is meant to be initialised > just prior to the unbind loop. It appears that my patches are > out-of-order. :| Oh, then I misread the intention of the code, sorry.
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index 50129ec1caab..19716548c455 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -109,7 +109,7 @@ i915_gem_evict_something(struct i915_address_space *vm, }, **phase; struct i915_vma *vma, *next; struct drm_mm_node *node; - int ret; + int ret = 0; lockdep_assert_held(&vm->i915->drm.struct_mutex); trace_i915_gem_evict(vm, min_size, alignment, flags);
If we find a suitable victim node on our first pass, then ret will be uninitialized which could lead to some funny business later. Fixes: 9332f3b1b99a ("drm/i915: Combine loops within i915_gem_evict_something") Cc: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Matthew Auld <matthew.auld@intel.com> --- drivers/gpu/drm/i915/i915_gem_evict.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)