Hi Am 16.06.23 um 19:00 schrieb Jeffrey Hugo: > On 6/15/2023 3:31 AM, Thomas Zimmermann wrote: >> Set drm_gem_prime_handle_to_fd() and drm_gem_prime_fd_to_handle() >> for all DRM drivers. Even drivers that do not support PRIME import >> or export of dma-bufs can now import their own buffer objects. This >> is required by some userspace, such as wlroots-based compositors, to >> share buffers among processes. >> >> The only driver that does not use the drm_gem_prime_*() helpers is >> vmwgfx. Once it has been converted, the callbacks in struct drm_driver >> can be removed. >> >> Simon Ser implemented the feature for drivers based on GEM VRAM helpers >> in [1]. This patchset generalizes the code for all drivers that do not >> otherwise support PRIME. Tested by running sway with gma500 hardware. >> >> [1] >> https://lore.kernel.org/dri-devel/20230302143502.500661-1-contact@emersion.fr/ >> >> Thomas Zimmermann (3): >>    drm: Enable PRIME import/export for all drivers >>    drm: Clear fd/handle callbacks in struct drm_driver >>    drm/prime: Unexport helpers for fd/handle conversion >> >>   Documentation/gpu/todo.rst                    | 14 +++++ >>   drivers/accel/ivpu/ivpu_drv.c                 |  2 - >>   drivers/accel/qaic/qaic_drv.c                 |  1 - >>   drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c       |  2 - >>   drivers/gpu/drm/armada/armada_drv.c           |  2 - >>   drivers/gpu/drm/drm_ioctl.c                   |  4 +- >>   drivers/gpu/drm/drm_prime.c                   | 62 ++++++++++--------- >>   drivers/gpu/drm/etnaviv/etnaviv_drv.c         |  2 - >>   drivers/gpu/drm/exynos/exynos_drm_drv.c       |  2 - >>   drivers/gpu/drm/i915/i915_driver.c            |  2 - >>   drivers/gpu/drm/lima/lima_drv.c               |  2 - >>   drivers/gpu/drm/mediatek/mtk_drm_drv.c        |  2 - >>   drivers/gpu/drm/msm/msm_drv.c                 |  2 - >>   drivers/gpu/drm/nouveau/nouveau_drm.c         |  2 - >>   drivers/gpu/drm/omapdrm/omap_drv.c            |  2 - >>   drivers/gpu/drm/panfrost/panfrost_drv.c       |  2 - >>   drivers/gpu/drm/pl111/pl111_drv.c             |  2 - >>   drivers/gpu/drm/qxl/qxl_drv.c                 |  2 - >>   drivers/gpu/drm/radeon/radeon_drv.c           |  2 - >>   drivers/gpu/drm/renesas/rcar-du/rcar_du_drv.c |  2 - >>   drivers/gpu/drm/rockchip/rockchip_drm_drv.c   |  2 - >>   drivers/gpu/drm/tegra/drm.c                   |  2 - >>   drivers/gpu/drm/v3d/v3d_drv.c                 |  2 - >>   drivers/gpu/drm/virtio/virtgpu_drv.c          |  2 - >>   drivers/gpu/drm/xen/xen_drm_front.c           |  2 - >>   include/drm/drm_drv.h                         |  6 ++ >>   include/drm/drm_gem_dma_helper.h              | 12 ++-- >>   include/drm/drm_gem_shmem_helper.h            |  6 +- >>   include/drm/drm_gem_vram_helper.h             |  4 +- >>   include/drm/drm_prime.h                       |  7 --- >>   30 files changed, 62 insertions(+), 96 deletions(-) >> > > Surely the maintainers of the 22 drivers touched by this series should > be in the to list or something?  Speaking for myself, threads that > directly address me trigger my "priority" filters. My apologies. I didn't want to snub you or anyone else. This change should not have an effect on *any* of these drivers. It's just clearing a pointer assignment that is now the default. And adding the maintainers of all affected drivers would add all maintainers of DRM code, hence I decided to not do that. Please note that the one maintainer that actually is affected, Zack, is in CC. > > Series looks sane to me, but I'd like to see what you do for a v2 to > address the existing comments. You'll be cc'ed. > > For the QAIC part - > > Acked-by: Jeffrey Hugo Thanks a lot. Best regards Thomas > > -Jeff -- Thomas Zimmermann Graphics Driver Developer SUSE Software Solutions Germany GmbH Frankenstrasse 146, 90461 Nuernberg, Germany GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman HRB 36809 (AG Nuernberg)