From: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: intel-gfx@lists.freedesktop.org,
Dmitry Ermilov <dmitry.ermilov@intel.com>
Subject: Re: [PATCH 4/5] drm/i915: Disable semaphore busywaits on saturated systems
Date: Fri, 3 May 2019 17:32:01 +0300 [thread overview]
Message-ID: <20190503143201.GP24299@intel.com> (raw)
In-Reply-To: <155689272156.3139.14895194831669934715@skylake-alporthouse-com>
On Fri, May 03, 2019 at 03:12:01PM +0100, Chris Wilson wrote:
> Quoting Ville Syrjälä (2019-05-03 15:04:57)
> > On Mon, Apr 29, 2019 at 07:00:19PM +0100, Chris Wilson wrote:
> > > Asking the GPU to busywait on a memory address, perhaps not unexpectedly
> > > in hindsight for a shared system, leads to bus contention that affects
> > > CPU programs trying to concurrently access memory. This can manifest as
> > > a drop in transcode throughput on highly over-saturated workloads.
> >
> > We can't use the singalling semaphore variant?
>
> That requires us to broadcast a signal to each engine (for which I can
> hear the cries of cross-powerwell wakes), and currently does not work
> with execlists + context-id==0. Or at least it failed in my testing.
Ah. If only we had MI_MWAIT.
--
Ville Syrjälä
Intel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2019-05-03 14:32 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-04-29 18:00 [PATCH 1/5] drm/i915: Wait for the struct_mutex on idling Chris Wilson
2019-04-29 18:00 ` [PATCH 2/5] drm/i915: Only reschedule the submission tasklet if preemption is possible Chris Wilson
2019-04-30 9:35 ` [PATCH v2] " Chris Wilson
2019-04-29 18:00 ` [PATCH 3/5] drm/i915: Delay semaphore submission until the start of the signaler Chris Wilson
2019-04-29 18:00 ` [PATCH 4/5] drm/i915: Disable semaphore busywaits on saturated systems Chris Wilson
2019-04-30 8:55 ` Tvrtko Ursulin
2019-04-30 9:04 ` Chris Wilson
2019-04-30 9:07 ` Chris Wilson
2019-05-03 13:25 ` Tvrtko Ursulin
2019-05-03 14:04 ` Ville Syrjälä
2019-05-03 14:12 ` Chris Wilson
2019-05-03 14:32 ` Ville Syrjälä [this message]
2019-04-29 18:00 ` [PATCH 5/5] drm/i915/execlists: Don't apply priority boost for resets Chris Wilson
2019-04-29 18:08 ` ✗ Fi.CI.SPARSE: warning for series starting with [1/5] drm/i915: Wait for the struct_mutex on idling Patchwork
2019-04-30 7:39 ` ✓ Fi.CI.BAT: success " Patchwork
2019-04-30 11:12 ` ✗ Fi.CI.IGT: failure " Patchwork
2019-04-30 13:12 ` ✗ Fi.CI.SPARSE: warning for series starting with [1/5] drm/i915: Wait for the struct_mutex on idling (rev2) Patchwork
2019-04-30 13:27 ` ✗ Fi.CI.BAT: failure " 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=20190503143201.GP24299@intel.com \
--to=ville.syrjala@linux.intel.com \
--cc=chris@chris-wilson.co.uk \
--cc=dmitry.ermilov@intel.com \
--cc=intel-gfx@lists.freedesktop.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox