From: "Christian König" <christian.koenig@amd.com>
To: Daniel Stone <daniel@fooishbar.org>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>,
intel-gfx <intel-gfx@lists.freedesktop.org>,
dri-devel <dri-devel@lists.freedesktop.org>,
Simon Ser <contact@emersion.fr>,
Sumit Semwal <sumit.semwal@linaro.org>
Subject: Re: [Intel-gfx] [PATCH 5/7] dma-buf: Add an API for exporting sync files (v11)
Date: Wed, 26 May 2021 14:46:17 +0200 [thread overview]
Message-ID: <770eb698-1dde-9e46-da83-44911d96abec@amd.com> (raw)
In-Reply-To: <CAPj87rN_7vmmWcL1vqhvM=etaTjwts_L0vYbsgvYF0_qZw83gw@mail.gmail.com>
Am 26.05.21 um 13:31 schrieb Daniel Stone:
> Hi Christian,
>
> On Wed, 26 May 2021 at 12:02, Christian König <christian.koenig@amd.com> wrote:
>> Am 25.05.21 um 23:17 schrieb Jason Ekstrand:
>>> This new IOCTL solves this problem by allowing us to get a snapshot of
>>> the implicit synchronization state of a given dma-buf in the form of a
>>> sync file. It's effectively the same as a poll() or I915_GEM_WAIT only,
>>> instead of CPU waiting directly, it encapsulates the wait operation, at
>>> the current moment in time, in a sync_file so we can check/wait on it
>>> later. As long as the Vulkan driver does the sync_file export from the
>>> dma-buf before we re-introduce it for rendering, it will only contain
>>> fences from the compositor or display. This allows to accurately turn
>>> it into a VkFence or VkSemaphore without any over- synchronization.
>> Regarding that, why do we actually use a syncfile and not a drm_syncobj
>> here?
>>
>> The later should be much closer to a Vulkan timeline semaphore.
> How would we insert a syncobj+val into a resv though? Like, if we pass
> an unmaterialised syncobj+val here to insert into the resv, then an
> implicit-only media user (or KMS) goes to sync against the resv, what
> happens?
Well this is for exporting, not importing. So we don't need to worry
about that.
It's just my thinking because the drm_syncobj is the backing object on
VkSemaphore implementations these days, isn't it?
Christian.
>
> Cheers,
> Daniel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
WARNING: multiple messages have this Message-ID (diff)
From: "Christian König" <christian.koenig@amd.com>
To: Daniel Stone <daniel@fooishbar.org>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>,
intel-gfx <intel-gfx@lists.freedesktop.org>,
dri-devel <dri-devel@lists.freedesktop.org>,
Jason Ekstrand <jason@jlekstrand.net>
Subject: Re: [Intel-gfx] [PATCH 5/7] dma-buf: Add an API for exporting sync files (v11)
Date: Wed, 26 May 2021 14:46:17 +0200 [thread overview]
Message-ID: <770eb698-1dde-9e46-da83-44911d96abec@amd.com> (raw)
In-Reply-To: <CAPj87rN_7vmmWcL1vqhvM=etaTjwts_L0vYbsgvYF0_qZw83gw@mail.gmail.com>
Am 26.05.21 um 13:31 schrieb Daniel Stone:
> Hi Christian,
>
> On Wed, 26 May 2021 at 12:02, Christian König <christian.koenig@amd.com> wrote:
>> Am 25.05.21 um 23:17 schrieb Jason Ekstrand:
>>> This new IOCTL solves this problem by allowing us to get a snapshot of
>>> the implicit synchronization state of a given dma-buf in the form of a
>>> sync file. It's effectively the same as a poll() or I915_GEM_WAIT only,
>>> instead of CPU waiting directly, it encapsulates the wait operation, at
>>> the current moment in time, in a sync_file so we can check/wait on it
>>> later. As long as the Vulkan driver does the sync_file export from the
>>> dma-buf before we re-introduce it for rendering, it will only contain
>>> fences from the compositor or display. This allows to accurately turn
>>> it into a VkFence or VkSemaphore without any over- synchronization.
>> Regarding that, why do we actually use a syncfile and not a drm_syncobj
>> here?
>>
>> The later should be much closer to a Vulkan timeline semaphore.
> How would we insert a syncobj+val into a resv though? Like, if we pass
> an unmaterialised syncobj+val here to insert into the resv, then an
> implicit-only media user (or KMS) goes to sync against the resv, what
> happens?
Well this is for exporting, not importing. So we don't need to worry
about that.
It's just my thinking because the drm_syncobj is the backing object on
VkSemaphore implementations these days, isn't it?
Christian.
>
> Cheers,
> Daniel
next prev parent reply other threads:[~2021-05-26 12:46 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-25 21:17 [Intel-gfx] [PATCH 0/7] dma-buf: Add an API for exporting sync files (v11) Jason Ekstrand
2021-05-25 21:17 ` Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] [PATCH 1/7] dma-buf: Add dma_fence_array_for_each (v2) Jason Ekstrand
2021-05-25 21:17 ` Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] [PATCH 2/7] dma-buf: Rename dma_resv helpers from _rcu to _unlocked (v2) Jason Ekstrand
2021-05-25 21:17 ` Jason Ekstrand
2021-05-26 10:57 ` [Intel-gfx] " Christian König
2021-05-26 10:57 ` Christian König
2021-05-27 10:39 ` [Intel-gfx] " Daniel Vetter
2021-05-27 10:39 ` Daniel Vetter
2021-05-27 11:58 ` [Intel-gfx] " Christian König
2021-05-27 11:58 ` Christian König
2021-05-27 13:21 ` [Intel-gfx] " Daniel Vetter
2021-05-27 13:21 ` Daniel Vetter
2021-05-27 13:25 ` [Intel-gfx] " Daniel Vetter
2021-05-27 13:25 ` Daniel Vetter
2021-05-27 13:41 ` [Intel-gfx] " Christian König
2021-05-27 13:41 ` Christian König
2021-06-01 14:34 ` [Intel-gfx] " Daniel Vetter
2021-06-01 14:34 ` Daniel Vetter
2021-06-01 17:27 ` [Intel-gfx] " Christian König
2021-06-01 17:27 ` Christian König
2021-06-01 17:29 ` [Intel-gfx] " Christian König
2021-06-01 17:29 ` Christian König
2021-05-25 21:17 ` [PATCH 3/7] dma-buf: Add dma_resv_get_singleton_unlocked (v5) Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] " Jason Ekstrand
2021-05-25 21:17 ` [PATCH 4/7] dma-buf: Document DMA_BUF_IOCTL_SYNC Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] " Jason Ekstrand
2021-05-27 10:38 ` Daniel Vetter
2021-05-27 10:38 ` Daniel Vetter
2021-05-27 11:12 ` [Intel-gfx] " Sumit Semwal
2021-05-27 11:12 ` Sumit Semwal
2021-06-10 20:57 ` [Intel-gfx] " Jason Ekstrand
2021-06-10 20:57 ` Jason Ekstrand
2021-05-25 21:17 ` [PATCH 5/7] dma-buf: Add an API for exporting sync files (v11) Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] " Jason Ekstrand
2021-05-26 11:02 ` Christian König
2021-05-26 11:02 ` Christian König
2021-05-26 11:31 ` [Intel-gfx] " Daniel Stone
2021-05-26 11:31 ` Daniel Stone
2021-05-26 12:46 ` Christian König [this message]
2021-05-26 12:46 ` Christian König
2021-05-26 13:12 ` Daniel Stone
2021-05-26 13:12 ` Daniel Stone
2021-05-26 13:23 ` Christian König
2021-05-26 13:23 ` Christian König
2021-05-27 10:33 ` Daniel Vetter
2021-05-27 10:33 ` Daniel Vetter
2021-05-27 10:48 ` Simon Ser
2021-05-27 10:48 ` Simon Ser
2021-05-27 12:01 ` Christian König
2021-05-27 12:01 ` Christian König
[not found] ` <CAOFGe95Zdn8P3=sOT0HkE9_+ac70g36LxpmLOyR2bKTTeS-xvQ@mail.gmail.com>
[not found] ` <fef50d81-399a-af09-1d13-de4db1b3fab8@amd.com>
2021-05-27 15:39 ` Jason Ekstrand
2021-05-27 15:39 ` Jason Ekstrand
2021-05-25 21:17 ` [PATCH 6/7] RFC: dma-buf: Add an extra fence to dma_resv_get_singleton_unlocked Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] " Jason Ekstrand
2021-05-25 21:17 ` [PATCH 7/7] RFC: dma-buf: Add an API for importing sync files (v7) Jason Ekstrand
2021-05-25 21:17 ` [Intel-gfx] " Jason Ekstrand
2021-05-26 17:09 ` Daniel Vetter
2021-05-26 17:09 ` Daniel Vetter
2021-05-25 21:44 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for dma-buf: Add an API for exporting sync files (v11) Patchwork
2021-05-25 21:46 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2021-05-25 22:14 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2021-05-26 4:20 ` [Intel-gfx] ✗ Fi.CI.IGT: failure " Patchwork
2021-06-10 20:10 ` [Intel-gfx] [PATCH 0/7] " Chia-I Wu
2021-06-10 20:10 ` Chia-I Wu
2021-06-10 20:26 ` [Intel-gfx] " Jason Ekstrand
2021-06-10 20:26 ` Jason Ekstrand
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=770eb698-1dde-9e46-da83-44911d96abec@amd.com \
--to=christian.koenig@amd.com \
--cc=contact@emersion.fr \
--cc=daniel.vetter@ffwll.ch \
--cc=daniel@fooishbar.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=intel-gfx@lists.freedesktop.org \
--cc=sumit.semwal@linaro.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.