From: Boris Brezillon <boris.brezillon@collabora.com>
To: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
Cc: Danilo Krummrich <dakr@redhat.com>,
airlied@gmail.com, daniel@ffwll.ch, matthew.brost@intel.com,
sarah.walker@imgtec.com, donald.robson@imgtec.com,
christian.koenig@amd.com, faith.ekstrand@collabora.com,
dri-devel@lists.freedesktop.org, nouveau@lists.freedesktop.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Date: Wed, 13 Sep 2023 09:03:11 +0200 [thread overview]
Message-ID: <20230913090311.5eeb026a@collabora.com> (raw)
In-Reply-To: <a9ef04d2-2525-65c0-2eda-45ca9a95a3a0@linux.intel.com>
On Tue, 12 Sep 2023 18:20:32 +0200
Thomas Hellström <thomas.hellstrom@linux.intel.com> wrote:
> > +/**
> > + * get_next_vm_bo_from_list() - get the next vm_bo element
> > + * @__gpuvm: The GPU VM
> > + * @__list_name: The name of the list we're iterating on
> > + * @__local_list: A pointer to the local list used to store already iterated items
> > + * @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo()
> > + *
> > + * This helper is here to provide lockless list iteration. Lockless as in, the
> > + * iterator releases the lock immediately after picking the first element from
> > + * the list, so list insertion deletion can happen concurrently.
>
> Are the list spinlocks needed for that async state update from within
> the dma-fence critical section we've discussed previously?
Any driver calling _[un]link() from its drm_gpu_scheduler::run_job()
hook will be in this situation (Panthor at the moment, PowerVR soon). I
get that Xe and Nouveau don't need that because they update the VM
state early (in the ioctl path), but I keep thinking this will hurt us
if we don't think it through from the beginning, because once you've
set this logic to depend only on resv locks, it will be pretty hard to
get back to a solution which lets synchronous VM_BINDs take precedence
on asynchronous request, and, with vkQueueBindSparse() passing external
deps (plus the fact the VM_BIND queue might be pretty deep), it can
take a long time to get your synchronous VM_BIND executed...
Now, maybe the solution is something different, with early VM state
update for everyone (creation of to-be-[un]mapped drm_gpuva entries,
some of them being shadowed by already existing drm_gpuva that's
encoding the currently mapped region), and VM state patching when a
synchronous VM_BIND kicks in (we need to patch the previously queued
requests too, so they always have enough resources for the map/unmap
operations to succeed).
next prev parent reply other threads:[~2023-09-13 7:03 UTC|newest]
Thread overview: 77+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-09 15:31 [PATCH drm-misc-next v3 0/7] [RFC] DRM GPUVA Manager GPU-VM features Danilo Krummrich
2023-09-09 15:31 ` [PATCH drm-misc-next v3 1/7] drm/gpuvm: rename struct drm_gpuva_manager to struct drm_gpuvm Danilo Krummrich
2023-09-09 18:23 ` kernel test robot
2023-09-09 15:31 ` [PATCH drm-misc-next v3 2/7] drm/gpuvm: allow building as module Danilo Krummrich
2023-09-11 13:09 ` Christian König
2023-09-09 15:31 ` [PATCH drm-misc-next v3 3/7] drm/nouveau: uvmm: rename 'umgr' to 'base' Danilo Krummrich
2023-09-09 15:31 ` [PATCH drm-misc-next v3 4/7] drm/gpuvm: common dma-resv per struct drm_gpuvm Danilo Krummrich
2023-09-11 12:00 ` Boris Brezillon
2023-09-11 16:16 ` Danilo Krummrich
2023-09-09 15:31 ` [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination Danilo Krummrich
2023-09-11 17:19 ` Thomas Hellström
2023-09-11 17:49 ` Danilo Krummrich
2023-09-11 18:37 ` Thomas Hellström
2023-09-12 7:42 ` Thomas Hellström
2023-09-12 10:06 ` Danilo Krummrich
2023-09-12 10:33 ` Thomas Hellström
2023-09-12 11:05 ` Danilo Krummrich
2023-09-09 15:31 ` [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation Danilo Krummrich
2023-09-09 20:16 ` kernel test robot
2023-09-11 10:35 ` Boris Brezillon
2023-09-11 16:23 ` Danilo Krummrich
2023-09-11 12:54 ` Boris Brezillon
2023-09-11 14:45 ` Boris Brezillon
2023-09-11 16:30 ` Danilo Krummrich
2023-09-12 16:20 ` Thomas Hellström
2023-09-12 16:50 ` Danilo Krummrich
2023-09-12 19:23 ` Thomas Hellström
2023-09-12 23:36 ` Danilo Krummrich
2023-09-13 9:14 ` Thomas Hellström
2023-09-13 12:16 ` Danilo Krummrich
2023-09-13 14:26 ` Christian König
2023-09-13 15:13 ` Thomas Hellström
2023-09-13 15:26 ` Christian König
2023-09-13 15:15 ` Danilo Krummrich
2023-09-13 15:33 ` Christian König
2023-09-13 15:46 ` Danilo Krummrich
2023-09-19 12:07 ` Christian König
2023-09-19 12:21 ` Thomas Hellström
2023-09-19 15:16 ` Danilo Krummrich
2023-09-19 15:23 ` Thomas Hellström
2023-09-20 5:37 ` Christian König
2023-09-20 7:44 ` Thomas Hellström
2023-09-20 8:29 ` Thomas Hellström
2023-09-20 10:51 ` Christian König
2023-09-20 12:06 ` Thomas Hellström
2023-09-20 13:06 ` Christian König
2023-09-20 13:38 ` Thomas Hellström
2023-09-20 13:48 ` Christian König
2023-09-20 14:02 ` Thomas Hellström
2023-09-20 14:11 ` Christian König
2023-09-14 10:57 ` [Nouveau] " Danilo Krummrich
2023-09-14 11:32 ` Thomas Hellström
2023-09-14 15:27 ` Danilo Krummrich
2023-09-14 17:13 ` Thomas Hellström
2023-09-14 17:15 ` Danilo Krummrich
2023-09-18 11:21 ` Danilo Krummrich
2023-09-13 7:03 ` Boris Brezillon [this message]
2023-09-13 7:05 ` Dave Airlie
2023-09-13 7:19 ` Boris Brezillon
2023-09-13 10:39 ` Thomas Hellström
2023-09-13 11:33 ` Boris Brezillon
2023-09-13 12:01 ` Danilo Krummrich
2023-09-13 13:22 ` Thomas Hellström
2023-09-13 14:01 ` Boris Brezillon
2023-09-13 14:29 ` Thomas Hellström
2023-09-13 15:17 ` Boris Brezillon
2023-09-14 8:20 ` Boris Brezillon
2023-09-14 10:45 ` Thomas Hellström
2023-09-14 11:54 ` Boris Brezillon
2023-09-14 13:33 ` Thomas Hellström
2023-09-14 15:37 ` Boris Brezillon
2023-09-14 13:48 ` Thomas Hellström
2023-09-14 16:36 ` Danilo Krummrich
2023-09-14 17:21 ` Thomas Hellström
2023-09-14 17:25 ` Danilo Krummrich
2023-09-14 19:14 ` Thomas Hellström
2023-09-09 15:31 ` [PATCH drm-misc-next v3 7/7] drm/nouveau: GPUVM dma-resv/extobj handling, " Danilo Krummrich
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=20230913090311.5eeb026a@collabora.com \
--to=boris.brezillon@collabora.com \
--cc=airlied@gmail.com \
--cc=christian.koenig@amd.com \
--cc=dakr@redhat.com \
--cc=daniel@ffwll.ch \
--cc=donald.robson@imgtec.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=faith.ekstrand@collabora.com \
--cc=linux-kernel@vger.kernel.org \
--cc=matthew.brost@intel.com \
--cc=nouveau@lists.freedesktop.org \
--cc=sarah.walker@imgtec.com \
--cc=thomas.hellstrom@linux.intel.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).