[Intel-gfx] [PATCH 1/2] drm/i915: unpin backing storage in dmabuf_unmap
Maarten Lankhorst
maarten.lankhorst at canonical.com
Wed Aug 7 12:40:11 CEST 2013
Op 07-08-13 12:09, Daniel Vetter schreef:
> This fixes a WARN in i915_gem_free_object when the
> obj->pages_pin_count isn't 0.
>
> v2: Add locking to unmap, noticed by Chris Wilson. Note that even
> though we call unmap with our own dev->struct_mutex held that won't
> result in an immediate deadlock since we never go through the dma_buf
> interfaces for our own, reimported buffers. But it's still easy to
> blow up and anger lockdep, but that's already the case with our ->map
> implementation. Fixing this for real will involve per dma-buf ww mutex
> locking by the callers. And lots of fun. So go with the duct-tape
> approach for now.
>
> Cc: Chris Wilson <chris at chris-wilson.co.uk>
> Reported-by: Maarten Lankhorst <maarten.lankhorst at canonical.com>
> Cc: Maarten Lankhorst <maarten.lankhorst at canonical.com>
> Tested-by: Armin K. <krejzi at email.com> (v1)
> Signed-off-by: Daniel Vetter <daniel.vetter at ffwll.ch>
>
Acked, this was my original patch to solve the issue.
I want to note that locking struct_mutex here will break lockdep, but it's a problem in drm, not this patch.
~Maarten
More information about the Intel-gfx
mailing list