public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Daniel Vetter <daniel@ffwll.ch>
To: Zhao Yakui <yakui.zhao@intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 0/5] drm/i915: Add the support of dual BSD rings on BDW GT3
Date: Wed, 9 Apr 2014 16:45:58 +0200	[thread overview]
Message-ID: <20140409144558.GE9262@phenom.ffwll.local> (raw)
In-Reply-To: <1397008796-3595-1-git-send-email-yakui.zhao@intel.com>

On Wed, Apr 09, 2014 at 09:59:51AM +0800, Zhao Yakui wrote:
> 
> This is the patch set that tries to add the support of dual BSD rings on BDW
> GT3. Based on hardware spec, the BDW GT3 has two independent BSD rings, which
> can be used to process the video commands. To be simpler, it is transparent 
> to user-space driver/middleware. In such case the kernel driver will decide
> which ring is to dispatch the BSD video command.
> 
> As every BSD ring is powerful, it is enough to dispatch the BSD video command
> based on the drm fd. In such case the different BSD ring is used for video playing
> back and encoding. At the same time the coarse dispatch mechanism can help to avoid
> the object synchronization between the BSD rings.

Ok, I've quickly read through it all and commented on a few things. Imo
the last patch should be massively simplified, at least for the first
round. Other things look small.

What's still missing are testcases, and I have two things in mind here:
- Exercise the 2nd ring dispatch and sync a bit. Since the 2nd bsd ring is
  hidden within the kernel I think the right approach would be to open a
  few drm fds (10 or so) and then randomly use them with a dummy reloc. We
  have two testcases which can be used as blueprints that need
  adjustement:

  - gem_ring_sync_loop: Probably easiest to copy it to a new file as
    gem_multi_bsd_sync_loop. This test exercises semaphores.
  - gem_dummy_reloc_loop, subtest mixed: Almost the same as the above, but
    the sync is done _inside_ the loop and hence this exercises gpu/cpu
    sync. We need both tests adjusted, for for this we need a new
    multi-bsd test.

- New testcase to fully test main execbuffer flags. This is simply
  something that's we don't yet have. The next guy to touch execbuf code
  needs to add it, and it looks like that's you ;-) I've done a JIRA task
  for the resource streamer work, but I think the resource streamer wont
  be merged anytime soon. So I'll reassign to you. Jira task is VIZ-3129.

Thanks, Daniel

> 
> 
> Zhao Yakui (5):
>   drm/i915: Split the BDW device definition to prepare for dual BSD
>     rings on BDW GT3
>   drm/i915: Initialize the second BSD ring on BDW GT3 machine
>   drm/i915: Handle the irq interrupt for the second BSD ring
>   drm/i915: Add the VCS2 switch in Intel_ring_setup_status_page for Gen7 to
>     remove the switch check warning
>   drm/i915: Use the coarse mechanism based on drm fd to dispatch the BSD command
>     on BDW GT3
> 
>  drivers/gpu/drm/i915/i915_dma.c            |   14 ++++++
>  drivers/gpu/drm/i915/i915_drv.c            |   24 ++++++++-
>  drivers/gpu/drm/i915/i915_drv.h            |    5 ++
>  drivers/gpu/drm/i915/i915_gem.c            |    9 +++-
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c |   73 +++++++++++++++++++++++++++-
>  drivers/gpu/drm/i915/i915_gpu_error.c      |    1 +
>  drivers/gpu/drm/i915/i915_irq.c            |    5 +-
>  drivers/gpu/drm/i915/i915_reg.h            |    1 +
>  drivers/gpu/drm/i915/intel_ringbuffer.c    |   57 ++++++++++++++++++++++
>  drivers/gpu/drm/i915/intel_ringbuffer.h    |    6 ++-
>  include/drm/i915_pciids.h                  |   10 ++--
>  11 files changed, 197 insertions(+), 8 deletions(-)
> 
> -- 
> 1.7.10.1
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch

  parent reply	other threads:[~2014-04-09 14:46 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-04-09  1:59 [PATCH 0/5] drm/i915: Add the support of dual BSD rings on BDW GT3 Zhao Yakui
2014-04-09  1:59 ` [PATCH 1/5] drm/i915: Split the BDW device definition to prepare for " Zhao Yakui
2014-04-09 14:27   ` Daniel Vetter
2014-04-10  0:44     ` Zhao Yakui
2014-04-09  1:59 ` [PATCH 2/5] drm/i915:Initialize the second BSD ring on BDW GT3 machine Zhao Yakui
2014-04-09  1:59 ` [PATCH 3/5] drm/i915:Handle the irq interrupt for the second BSD ring Zhao Yakui
2014-04-09  1:59 ` [PATCH 4/5] drm/i915:Add the VCS2 switch in Intel_ring_setup_status_page for Gen7 to remove the switch check warning Zhao Yakui
2014-04-09 14:29   ` Daniel Vetter
2014-04-10  0:45     ` Zhao Yakui
2014-04-09  1:59 ` [PATCH 5/5] drm/i915:Use the coarse mechanism based on drm fd to dispatch the BSD command on BDW GT3 Zhao Yakui
2014-04-09 14:34   ` Daniel Vetter
2014-04-10  2:24     ` Zhao Yakui
2014-04-10  6:48       ` Daniel Vetter
2014-04-10  8:04         ` Zhao Yakui
2014-04-10  9:03           ` Daniel Vetter
2014-04-11  0:53             ` Zhao Yakui
2014-04-09 14:45 ` Daniel Vetter [this message]
2014-04-10  3:28   ` [PATCH 0/5] drm/i915: Add the support of dual BSD rings " Zhao Yakui
2014-04-10  6:58     ` Daniel Vetter
2014-04-10  8:28       ` Zhao Yakui
2014-04-10  9:04         ` Daniel Vetter
2014-04-11  0:56           ` Zhao Yakui
2014-04-11  8:57             ` Daniel Vetter
2014-04-14  1:05               ` Zhao Yakui

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=20140409144558.GE9262@phenom.ffwll.local \
    --to=daniel@ffwll.ch \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=yakui.zhao@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