From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
To: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>,
intel-gfx@lists.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH] drm/i915: Kill context before taking ctx->mutex
Date: Fri, 3 Jul 2020 12:35:55 +0200 [thread overview]
Message-ID: <df43f488-ca25-1d8b-82dc-7dc2fb968a9f@linux.intel.com> (raw)
In-Reply-To: <e44599b5-b28c-9142-c6df-a9ee62cca05f@linux.intel.com>
Op 02-07-2020 om 16:51 schreef Tvrtko Ursulin:
> On 02/07/2020 14:26, Maarten Lankhorst wrote:
>> Op 30-06-2020 om 16:16 schreef Tvrtko Ursulin:
>>> On 24/06/2020 12:05, Maarten Lankhorst wrote:
>>>> Killing context before taking ctx->mutex fixes a hang in
>>>> gem_ctx_persistence.close-replace-race, where lut_close
>>>> takes obj->resv.lock which is already held by execbuf,
>>>> causing a stalling indefinitely.
>>> If this is the consequence of inverting the locking order I think you need to move the fix earlier in the series, to precede the patch which creates the inversion. Otherwise AFAICT the re-order of kill_context vs lut_close seems fine.
>> Yeah, it was just a bugfix I found when looking at the code, if you review it I can push it now so I don't have to resend. :)
> You are saying it's a bug in drm-tip today?
>
> From the commit:
>
> [ 1904.342847] 2 locks held by gem_ctx_persist/11520:
> [ 1904.342849] #0: ffff8882188e4968 (&ctx->mutex){+.+.}-{3:3}, at: context_close+0xe6/0x850 [i915]
> [ 1904.342941] #1: ffff88821c58a5a8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: lut_close+0x2c2/0xba0 [i915]
> [ 1904.343033] 3 locks held by gem_ctx_persist/11521:
> [ 1904.343035] #0: ffffc900008ff938 (reservation_ww_class_acquire){+.+.}-{0:0}, at: i915_gem_do_execbuffer+0x103d/0x54c0 [i915]
> [ 1904.343157] #1: ffff88821c58a5a8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: eb_validate_vmas+0x602/0x2010 [i915]
> [ 1904.343267] #2: ffff88820afd9200 (&vm->mutex/1){+.+.}-{3:3}, at: i915_vma_pin_ww+0x335/0x2300 [i915]
>
> I don't see two inverted locks in two threads - what is happening causing "stalling" - deadlock? Livelock?
>
> Regards,
>
> Tvrtko
This patch can probably be removed now that lut_lock is split out as a spinlock.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2020-07-03 10:35 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-23 14:28 [Intel-gfx] [PATCH 01/26] Revert "drm/i915/gem: Async GPU relocations only" Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 02/26] drm/i915: Revert relocation chaining commits Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 03/26] Revert "drm/i915/gem: Drop relocation slowpath" Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 04/26] drm/i915: Add an implementation for i915_gem_ww_ctx locking, v2 Maarten Lankhorst
2020-06-24 7:10 ` Thomas Hellström (Intel)
2020-06-24 7:43 ` Chris Wilson
2020-06-24 7:49 ` Thomas Hellström (Intel)
2020-06-24 8:27 ` Chris Wilson
2020-06-29 12:07 ` Tvrtko Ursulin
2020-06-29 12:32 ` Tvrtko Ursulin
2020-06-29 13:44 ` Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 05/26] drm/i915: Remove locking from i915_gem_object_prepare_read/write Maarten Lankhorst
2020-06-26 13:32 ` Thomas Hellström (Intel)
2020-06-29 12:56 ` Tvrtko Ursulin
2020-06-23 14:28 ` [Intel-gfx] [PATCH 06/26] drm/i915: Parse command buffer earlier in eb_relocate(slow) Maarten Lankhorst
2020-06-26 14:41 ` Thomas Hellström (Intel)
2020-06-29 10:40 ` Maarten Lankhorst
2020-06-29 11:15 ` Thomas Hellström (Intel)
2020-06-29 11:18 ` Maarten Lankhorst
2020-06-29 14:42 ` Tvrtko Ursulin
2020-06-23 14:28 ` [Intel-gfx] [PATCH 07/26] Revert "drm/i915/gem: Split eb_vma into its own allocation" Maarten Lankhorst
2020-06-29 15:08 ` Tvrtko Ursulin
2020-06-30 11:52 ` Maarten Lankhorst
2020-06-30 12:31 ` Tvrtko Ursulin
2020-06-30 14:07 ` Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 08/26] drm/i915/gem: Make eb_add_lut interruptible wait on object lock Maarten Lankhorst
2020-06-26 13:52 ` Thomas Hellström (Intel)
2020-06-29 15:14 ` Tvrtko Ursulin
2020-06-30 11:56 ` Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 09/26] drm/i915: Use per object locking in execbuf, v12 Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 10/26] drm/i915: Use ww locking in intel_renderstate Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 11/26] drm/i915: Add ww context handling to context_barrier_task Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 12/26] drm/i915: Nuke arguments to eb_pin_engine Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 13/26] drm/i915: Pin engine before pinning all objects, v4 Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 14/26] drm/i915: Rework intel_context pinning to do everything outside of pin_mutex Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 15/26] drm/i915: Make sure execbuffer always passes ww state to i915_vma_pin Maarten Lankhorst
2020-06-25 14:32 ` Thomas Hellström (Intel)
2020-06-23 14:28 ` [Intel-gfx] [PATCH 16/26] drm/i915: Convert i915_gem_object/client_blt.c to use ww locking as well, v2 Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 17/26] drm/i915: Kill last user of intel_context_create_request outside of selftests Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 18/26] drm/i915: Convert i915_perf to ww locking as well Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 19/26] drm/i915: Dirty hack to fix selftests locking inversion Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 20/26] drm/i915/selftests: Fix locking inversion in lrc selftest Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 21/26] drm/i915: Use ww pinning for intel_context_create_request() Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 22/26] drm/i915: Move i915_vma_lock in the selftests to avoid lock inversion, v2 Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 23/26] drm/i915: Add ww locking to vm_fault_gtt Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 24/26] drm/i915: Add ww locking to pin_to_display_plane Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 25/26] drm/i915: Ensure we hold the pin mutex Maarten Lankhorst
2020-06-23 14:28 ` [Intel-gfx] [PATCH 26/26] drm/i915: Kill context before taking ctx->mutex Maarten Lankhorst
2020-06-24 11:05 ` [Intel-gfx] [PATCH] " Maarten Lankhorst
2020-06-30 14:16 ` Tvrtko Ursulin
2020-07-02 13:26 ` Maarten Lankhorst
2020-07-02 14:51 ` Tvrtko Ursulin
2020-07-03 10:35 ` Maarten Lankhorst [this message]
2020-06-23 15:23 ` [Intel-gfx] [PATCH 01/26] Revert "drm/i915/gem: Async GPU relocations only" Chris Wilson
2020-06-24 11:19 ` Chris Wilson
2020-06-23 15:39 ` [Intel-gfx] ✗ Fi.CI.BUILD: failure for series starting with [01/26] " Patchwork
2020-06-24 11:58 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for series starting with [01/26] Revert "drm/i915/gem: Async GPU relocations only" (rev2) Patchwork
2020-06-24 11:59 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2020-06-24 12:48 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork
2020-07-01 13:10 ` [Intel-gfx] ✗ Fi.CI.IGT: " Patchwork
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=df43f488-ca25-1d8b-82dc-7dc2fb968a9f@linux.intel.com \
--to=maarten.lankhorst@linux.intel.com \
--cc=intel-gfx@lists.freedesktop.org \
--cc=tvrtko.ursulin@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