2019-05-27 08:55:21 +02:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
dma-buf/fence: Fix lock inversion within dma-fence-array
Ages ago Rob Clark noted,
"Currently with fence-array, we have a potential deadlock situation. If
we fence_add_callback() on an array-fence, the array-fence's lock is
acquired first, and in it's ->enable_signaling() callback, it will install
cbs on it's array-member fences, so the array-member's lock is acquired
second.
But in the signal path, the array-member's lock is acquired first, and
the array-fence's lock acquired second."
Rob proposed either extensive changes to dma-fence to unnest the
fence-array signaling, or to defer the signaling onto a workqueue. This
is a more refined version of the later, that should keep the latency
of the fence signaling to a minimum by using an irq-work, which is
executed asap.
Reported-by: Rob Clark <robdclark@gmail.com>
Suggested-by: Rob Clark <robdclark@gmail.com>
References: 1476635975-21981-1-git-send-email-robdclark@gmail.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20171114162719.30958-1-chris@chris-wilson.co.uk
2017-11-14 16:27:19 +00:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
dma-buf/fence: Fix lock inversion within dma-fence-array
Ages ago Rob Clark noted,
"Currently with fence-array, we have a potential deadlock situation. If
we fence_add_callback() on an array-fence, the array-fence's lock is
acquired first, and in it's ->enable_signaling() callback, it will install
cbs on it's array-member fences, so the array-member's lock is acquired
second.
But in the signal path, the array-member's lock is acquired first, and
the array-fence's lock acquired second."
Rob proposed either extensive changes to dma-fence to unnest the
fence-array signaling, or to defer the signaling onto a workqueue. This
is a more refined version of the later, that should keep the latency
of the fence signaling to a minimum by using an irq-work, which is
executed asap.
Reported-by: Rob Clark <robdclark@gmail.com>
Suggested-by: Rob Clark <robdclark@gmail.com>
References: 1476635975-21981-1-git-send-email-robdclark@gmail.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20171114162719.30958-1-chris@chris-wilson.co.uk
2017-11-14 16:27:19 +00:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
dma-buf/fence: Fix lock inversion within dma-fence-array
Ages ago Rob Clark noted,
"Currently with fence-array, we have a potential deadlock situation. If
we fence_add_callback() on an array-fence, the array-fence's lock is
acquired first, and in it's ->enable_signaling() callback, it will install
cbs on it's array-member fences, so the array-member's lock is acquired
second.
But in the signal path, the array-member's lock is acquired first, and
the array-fence's lock acquired second."
Rob proposed either extensive changes to dma-fence to unnest the
fence-array signaling, or to defer the signaling onto a workqueue. This
is a more refined version of the later, that should keep the latency
of the fence signaling to a minimum by using an irq-work, which is
executed asap.
Reported-by: Rob Clark <robdclark@gmail.com>
Suggested-by: Rob Clark <robdclark@gmail.com>
References: 1476635975-21981-1-git-send-email-robdclark@gmail.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20171114162719.30958-1-chris@chris-wilson.co.uk
2017-11-14 16:27:19 +00:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2024-05-25 18:00:31 +02:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:04 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
2016-06-01 15:10:04 +02:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2024-11-08 09:29:48 +01:00
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2024-11-08 09:29:48 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2021-11-29 16:27:27 +01:00
|
|
|
|
2024-11-08 09:29:48 +01:00
|
|
|
|
2021-11-29 16:27:27 +01:00
|
|
|
|
2024-11-08 09:29:48 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2021-11-29 16:27:27 +01:00
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2021-09-03 11:47:57 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2021-09-03 11:47:57 -07:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2024-08-26 10:01:43 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-02-21 17:06:51 -08:00
|
|
|
|
2024-08-26 10:01:43 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:04 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2024-08-26 10:01:43 -07:00
|
|
|
|
|
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2024-08-26 10:01:43 -07:00
|
|
|
|
|
|
|
|
|
2025-11-10 19:58:10 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2026-02-24 10:39:22 -08:00
|
|
|
|
|
|
|
|
|
2022-03-25 16:38:54 +01:00
|
|
|
|
|
|
|
|
|
2024-05-25 18:00:31 +02:00
|
|
|
|
|
|
|
|
|
2025-10-09 16:18:53 +02:00
|
|
|
|
|
|
|
|
|
dma-buf/fence: Fix lock inversion within dma-fence-array
Ages ago Rob Clark noted,
"Currently with fence-array, we have a potential deadlock situation. If
we fence_add_callback() on an array-fence, the array-fence's lock is
acquired first, and in it's ->enable_signaling() callback, it will install
cbs on it's array-member fences, so the array-member's lock is acquired
second.
But in the signal path, the array-member's lock is acquired first, and
the array-fence's lock acquired second."
Rob proposed either extensive changes to dma-fence to unnest the
fence-array signaling, or to defer the signaling onto a workqueue. This
is a more refined version of the later, that should keep the latency
of the fence signaling to a minimum by using an irq-work, which is
executed asap.
Reported-by: Rob Clark <robdclark@gmail.com>
Suggested-by: Rob Clark <robdclark@gmail.com>
References: 1476635975-21981-1-git-send-email-robdclark@gmail.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20171114162719.30958-1-chris@chris-wilson.co.uk
2017-11-14 16:27:19 +00:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
2026-02-24 10:39:22 -08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2025-11-10 19:58:10 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2019-08-11 22:09:02 +01:00
|
|
|
|
|
|
|
|
|
2022-01-19 11:40:21 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2024-08-26 10:01:43 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2025-11-10 19:58:10 +01:00
|
|
|
|
2024-08-26 10:01:43 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2025-11-10 19:58:10 +01:00
|
|
|
|
2022-01-19 11:40:21 +01:00
|
|
|
|
2016-06-01 15:10:03 +02:00
|
|
|
|
|
|
|
|
|
2016-10-25 13:00:45 +01:00
|
|
|
|
2017-03-17 17:34:49 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2022-01-24 14:03:24 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2022-03-25 16:38:54 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2022-01-24 14:03:24 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|