Re: Stats for XDP actions

On 6/20/19 2:42 PM, Toke Høiland-Jørgensen wrote: >>> I don't recall seeing any follow-up on this. Did you have a chance to >>> formulate your ideas? :) >>> >> >> Not yet. Almost done with the nexthop changes. Once that is out of the >> way I can come back to this. > > Ping? :) > Definitely

Re: [PATCH 1/1] scsi: virtio_scsi: remove unused 'affinity_hint_set'

Dongli, > The 'affinity_hint_set' is not used any longer since > commit 0d9f0a52c8b9 ("virtio_scsi: use virtio IRQ affinity"). Applied to 5.3/scsi-queue. Thanks! -- Martin K. Petersen Oracle Linux Engineering ___ Virtualization mailing list

Re: Stats for XDP actions

David Ahern writes: > On 4/18/19 8:24 AM, Toke Høiland-Jørgensen wrote: >>> >>> Understood. Hopefully in March I will get some time to come back to this >>> and propose an idea on what I would like to see - namely, the admin has >>> a config option at load time to enable driver counters

[PATCH 3/6] drm/qxl: use embedded gem object

Drop drm_gem_object from qxl_bo, use the ttm_buffer_object.base instead. Signed-off-by: Gerd Hoffmann --- drivers/gpu/drm/qxl/qxl_drv.h | 6 +++--- drivers/gpu/drm/qxl/qxl_object.h | 4 ++-- drivers/gpu/drm/qxl/qxl_cmd.c | 4 ++-- drivers/gpu/drm/qxl/qxl_debugfs.c | 2 +-

[PATCH 5/6] drm/ttm: use gem vma_node

Drop vma_node from ttm_buffer_object, use the gem struct (base.vma_node) instead. Signed-off-by: Gerd Hoffmann --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.h | 2 +- drivers/gpu/drm/qxl/qxl_object.h | 2 +- drivers/gpu/drm/radeon/radeon_object.h | 2 +-

[PATCH 3/3] drm: drop DEFINE_DRM_GEM_SHMEM_FOPS

DEFINE_DRM_GEM_SHMEM_FOPS is identical to DEFINE_DRM_GEM_FOPS now, drop it. Signed-off-by: Gerd Hoffmann --- include/drm/drm_gem_shmem_helper.h | 26 - drivers/gpu/drm/cirrus/cirrus.c | 2 +- drivers/gpu/drm/panfrost/panfrost_drv.c | 2 +-

[PATCH v4 09/12] drm/virtio: rework virtio_gpu_object_create fencing

Use gem reservation helpers and direct reservation_object_* calls instead of ttm. v3: Due to using the gem reservation object it is initialized and ready for use before calling ttm_bo_init, so we can also drop the tricky fence logic which checks whenever the command is in flight still. We can

[PATCH v4 08/12] drm/virtio: rework virtio_gpu_execbuffer_ioctl fencing

Use gem reservation helpers and direct reservation_object_* calls instead of ttm. v3: Also attach the array of gem objects to the virtio command buffer, so we can drop the object references in the completion callback. Needed because ttm fence helpers grab a reference for us, but gem helpers

[PATCH v4 10/12] drm/virtio: drop virtio_gpu_object_list_validate/virtio_gpu_unref_list

No users left. Signed-off-by: Gerd Hoffmann Acked-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_drv.h | 3 -- drivers/gpu/drm/virtio/virtgpu_ioctl.c | 39 -- 2 files changed, 42 deletions(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h

[PATCH v4 01/12] drm/virtio: pass gem reservation object to ttm init

With this gem and ttm will use the same reservation object, so mixing and matching ttm / gem reservation helpers should work fine. Signed-off-by: Gerd Hoffmann Reviewed-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_object.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff

[PATCH v4 02/12] drm/virtio: switch virtio_gpu_wait_ioctl() to gem helper.

Use drm_gem_reservation_object_wait() in virtio_gpu_wait_ioctl(). This also makes the ioctl run lockless. v2: use reservation_object_test_signaled_rcu for VIRTGPU_WAIT_NOWAIT. Signed-off-by: Gerd Hoffmann Reviewed-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_ioctl.c | 24

[PATCH v4 05/12] drm/virtio: drop no_wait argument from virtio_gpu_object_reserve

All callers pass no_wait = false. Signed-off-by: Gerd Hoffmann Reviewed-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_drv.h | 5 ++--- drivers/gpu/drm/virtio/virtgpu_gem.c | 4 ++-- drivers/gpu/drm/virtio/virtgpu_ioctl.c | 4 ++-- 3 files changed, 6 insertions(+), 7 deletions(-)

[PATCH v4 11/12] drm/virtio: switch from ttm to gem shmem helpers

virtio-gpu basically needs a sg_table for the bo, to tell the host where the backing pages for the object are. So the gem shmem helpers are a perfect fit. Some drm_gem_object_funcs need thin wrappers to update the host state, but otherwise the helpers handle everything just fine. Once the

[PATCH v4 03/12] drm/virtio: simplify cursor updates

No need to do the reservation dance, we can just wait on the fence directly. Signed-off-by: Gerd Hoffmann Reviewed-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_plane.c | 13 +++-- 1 file changed, 3 insertions(+), 10 deletions(-) diff --git

[PATCH v4 06/12] drm/virtio: remove ttm calls from in virtio_gpu_object_{reserve, unreserve}

Call reservation_object_* directly instead of using ttm_bo_{reserve,unreserve}. v4: check for EINTR only. v3: check for EINTR too. Signed-off-by: Gerd Hoffmann Reviewed-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_drv.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff

[PATCH v4 04/12] drm/virtio: remove virtio_gpu_object_wait

No users left. Signed-off-by: Gerd Hoffmann Reviewed-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_drv.h| 1 - drivers/gpu/drm/virtio/virtgpu_object.c | 13 - 2 files changed, 14 deletions(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h

[PATCH v4 12/12] drm/virtio: remove virtio_gpu_alloc_object

Thin wrapper around virtio_gpu_object_create(), but calling that directly works equally well. Signed-off-by: Gerd Hoffmann Acked-by: Daniel Vetter --- drivers/gpu/drm/virtio/virtgpu_drv.h | 4 drivers/gpu/drm/virtio/virtgpu_gem.c | 23 ---

[PATCH v4 07/12] drm/virtio: add virtio_gpu_object_array & helpers

Some helper functions to manage an array of gem objects. v4: make them virtio-private instead of generic helpers. Signed-off-by: Gerd Hoffmann --- drivers/gpu/drm/virtio/virtgpu_drv.h | 10 ++ drivers/gpu/drm/virtio/virtgpu_gem.c | 50 2 files changed, 60

Re: [PATCH v3 08/12] drm/virtio: rework virtio_gpu_execbuffer_ioctl fencing

Hi, > Also, I strongly recommend you do a very basic igt to exercise this, i.e. > allocate some buffers, submit them in a dummby op, then close the entire > drmfd. The old version should at least have tripped over kasan, maybe even > oopses somewhere. Hmm, I suspect I have to extend igt for