Hi, Jasmine, On 12/3/21 22:49, Jasmine Newsome wrote:
To avoid confusion with deferencing possible null pointer bo->ttm, replace pointer bo->ttm with local pointer ttm in i915_ttm_move as ttm has checks for null before getting passed to __i915_ttm_move
It's OK to use the local variable ttm here because it has previously been assigned from bo->ttm.
However, if a static analyzer gets confused by that and the fact that a NULL pointer might be passed to __i915_ttm_move, (which will not dereference it), then that's a static analyzer false positive.
So the commit message needs to be rewritten saying something like the code is getting a bit more readable if the local variable is used since it was previously assigned from bo->ttm, that's ok.
Also please add your Signed-off-by: To the patch. Thanks, Thomas
--- drivers/gpu/drm/i915/gem/i915_gem_ttm_move.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_ttm_move.c b/drivers/gpu/drm/i915/gem/i915_gem_ttm_move.c index 80df9f592407..56b6573b5c93 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_ttm_move.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_ttm_move.c @@ -763,7 +763,7 @@ int i915_ttm_move(struct ttm_buffer_object *bo, bool evict, return PTR_ERR(dep); }- migration_fence = __i915_ttm_move(bo, clear, dst_mem, bo->ttm,+ migration_fence = __i915_ttm_move(bo, clear, dst_mem, ttm, dst_rsgt, true, dep); dma_fence_put(dep); }
