| Message ID | 20240628200729.754669-1-zack.rusin@broadcom.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | Fix various buffer mapping/import issues | expand |
On Fri, Jun 28, 2024 at 11:07 PM Zack Rusin <zack.rusin@broadcom.com> wrote: > > This small series fixes all known prime/dumb_buffer/buffer dirty > tracking issues. Fixing of dumb-buffers turned out to be a lot more > complex than I wanted it to be. There's not much that can be done > there because the driver has to support old userspace (our Xorg driver > expects those to not be gem buffers and special cases a bunch of > functionality) and new userspace (which expects the handles to be > gem buffers, at least to issue GEM_CLOSE). > > The third patch deals with it by making the objects returned from > dumb-buffers both (raw buffers and surfaces referenced by the same > handle), which always works and doesn't require any changes in userspace. > > This fixes the known KDE (KWin's) buffer rendering issues. > > v2: Fix compute_crc in the second patch, as spotted by Martin > > Zack Rusin (4): > drm/vmwgfx: Fix a deadlock in dma buf fence polling > drm/vmwgfx: Make sure the screen surface is ref counted > drm/vmwgfx: Fix handling of dumb buffers > drm/vmwgfx: Add basic support for external buffers > > drivers/gpu/drm/vmwgfx/vmw_surface_cache.h | 10 +- > drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 127 +++--- > drivers/gpu/drm/vmwgfx/vmwgfx_bo.h | 15 +- > drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 40 +- > drivers/gpu/drm/vmwgfx/vmwgfx_fence.c | 26 +- > drivers/gpu/drm/vmwgfx/vmwgfx_gem.c | 62 ++- > drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 453 +++++++-------------- > drivers/gpu/drm/vmwgfx/vmwgfx_kms.h | 17 +- > drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c | 14 +- > drivers/gpu/drm/vmwgfx/vmwgfx_prime.c | 32 +- > drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 27 +- > drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c | 33 +- > drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c | 145 +++---- > drivers/gpu/drm/vmwgfx/vmwgfx_surface.c | 277 ++++++++++++- > drivers/gpu/drm/vmwgfx/vmwgfx_vkms.c | 40 +- > 15 files changed, 788 insertions(+), 530 deletions(-) > > -- > 2.40.1 > To the series: Reviewed-by: Martin Krastev <martin.krastev@broadcom.com> Regards, Martin
On 6/28/24 13:07, Zack Rusin wrote: > This small series fixes all known prime/dumb_buffer/buffer dirty > tracking issues. Fixing of dumb-buffers turned out to be a lot more > complex than I wanted it to be. There's not much that can be done > there because the driver has to support old userspace (our Xorg driver > expects those to not be gem buffers and special cases a bunch of > functionality) and new userspace (which expects the handles to be > gem buffers, at least to issue GEM_CLOSE). > > The third patch deals with it by making the objects returned from > dumb-buffers both (raw buffers and surfaces referenced by the same > handle), which always works and doesn't require any changes in userspace. > > This fixes the known KDE (KWin's) buffer rendering issues. > > v2: Fix compute_crc in the second patch, as spotted by Martin > > Zack Rusin (4): > drm/vmwgfx: Fix a deadlock in dma buf fence polling > drm/vmwgfx: Make sure the screen surface is ref counted > drm/vmwgfx: Fix handling of dumb buffers > drm/vmwgfx: Add basic support for external buffers > > drivers/gpu/drm/vmwgfx/vmw_surface_cache.h | 10 +- > drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 127 +++--- > drivers/gpu/drm/vmwgfx/vmwgfx_bo.h | 15 +- > drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 40 +- > drivers/gpu/drm/vmwgfx/vmwgfx_fence.c | 26 +- > drivers/gpu/drm/vmwgfx/vmwgfx_gem.c | 62 ++- > drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 453 +++++++-------------- > drivers/gpu/drm/vmwgfx/vmwgfx_kms.h | 17 +- > drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c | 14 +- > drivers/gpu/drm/vmwgfx/vmwgfx_prime.c | 32 +- > drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 27 +- > drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c | 33 +- > drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c | 145 +++---- > drivers/gpu/drm/vmwgfx/vmwgfx_surface.c | 277 ++++++++++++- > drivers/gpu/drm/vmwgfx/vmwgfx_vkms.c | 40 +- > 15 files changed, 788 insertions(+), 530 deletions(-) > For the series - LGTM! Reviewed-by: Maaz Mombasawala <maaz.mombasawala@broadcom.com> Thanks, Maaz Mombasawala <maaz.mombasawala@broadcom.com>