All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
To: tomba@kernel.org, airlied@linux.ie, daniel@ffwll.ch
Cc: dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
	linux-omap@vger.kernel.org, merlijn@wizzup.org, tony@atomide.com,
	Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
Subject: [PATCH 0/3] drm: omapdrm: Fix excessive GEM buffers DMM/CMA usage
Date: Wed, 19 Jan 2022 12:23:08 +0200	[thread overview]
Message-ID: <1642587791-13222-1-git-send-email-ivo.g.dimitrov.75@gmail.com> (raw)

This patch series fixes excessive DMM or CMA usage of GEM buffers leading to
various runtime allocation failures. The series enables daily usage of devices
without exausting limited resources like CMA or DMM space if GPU rendering is
needed.

The first patch doesn't bring any functional changes, it just moves some
TILER/DMM related code to a separate function, to simplify the review of the
next two patches.

The second patch allows off-CPU rendering to non-scanout buffers. Without that
patch, it is basically impossible to use the driver allocated GEM buffers on
OMAP3 for anything else but a basic CPU rendered examples as if we want GPU
rendering, we must allocate buffers as scanout buffers, which are CMA allocated.
CMA soon gets fragmented and we start seeing allocation failures. Such failres
in Xorg cannot be handeled gracefully, so the system is basically unusable.

Third patch fixes similar issue on OMAP4/5, where DMM/TILER spaces get
fragmented with time, leading to allocation failures.

Series were tested on Motolola Droid4 and Nokia N900, with OMAP DDX and
PVR EXA from https://github.com/maemo-leste/xf86-video-omap

Ivaylo Dimitrov (3):
  drm: omapdrm: simplify omap_gem_pin
  drm: omapdrm: Support exporting of non-contiguous GEM BOs
  drm: omapdrm: Do no allocate non-scanout GEMs through DMM/TILER

 drivers/gpu/drm/omapdrm/omap_gem.c        | 198 +++++++++++++++++-------------
 drivers/gpu/drm/omapdrm/omap_gem.h        |   3 +-
 drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c |   5 +-
 3 files changed, 116 insertions(+), 90 deletions(-)

-- 
1.9.1


WARNING: multiple messages have this Message-ID (diff)
From: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
To: tomba@kernel.org, airlied@linux.ie, daniel@ffwll.ch
Cc: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>,
	tony@atomide.com, merlijn@wizzup.org,
	linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org,
	linux-omap@vger.kernel.org
Subject: [PATCH 0/3] drm: omapdrm: Fix excessive GEM buffers DMM/CMA usage
Date: Wed, 19 Jan 2022 12:23:08 +0200	[thread overview]
Message-ID: <1642587791-13222-1-git-send-email-ivo.g.dimitrov.75@gmail.com> (raw)

This patch series fixes excessive DMM or CMA usage of GEM buffers leading to
various runtime allocation failures. The series enables daily usage of devices
without exausting limited resources like CMA or DMM space if GPU rendering is
needed.

The first patch doesn't bring any functional changes, it just moves some
TILER/DMM related code to a separate function, to simplify the review of the
next two patches.

The second patch allows off-CPU rendering to non-scanout buffers. Without that
patch, it is basically impossible to use the driver allocated GEM buffers on
OMAP3 for anything else but a basic CPU rendered examples as if we want GPU
rendering, we must allocate buffers as scanout buffers, which are CMA allocated.
CMA soon gets fragmented and we start seeing allocation failures. Such failres
in Xorg cannot be handeled gracefully, so the system is basically unusable.

Third patch fixes similar issue on OMAP4/5, where DMM/TILER spaces get
fragmented with time, leading to allocation failures.

Series were tested on Motolola Droid4 and Nokia N900, with OMAP DDX and
PVR EXA from https://github.com/maemo-leste/xf86-video-omap

Ivaylo Dimitrov (3):
  drm: omapdrm: simplify omap_gem_pin
  drm: omapdrm: Support exporting of non-contiguous GEM BOs
  drm: omapdrm: Do no allocate non-scanout GEMs through DMM/TILER

 drivers/gpu/drm/omapdrm/omap_gem.c        | 198 +++++++++++++++++-------------
 drivers/gpu/drm/omapdrm/omap_gem.h        |   3 +-
 drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c |   5 +-
 3 files changed, 116 insertions(+), 90 deletions(-)

-- 
1.9.1


             reply	other threads:[~2022-01-19 10:23 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-01-19 10:23 Ivaylo Dimitrov [this message]
2022-01-19 10:23 ` [PATCH 0/3] drm: omapdrm: Fix excessive GEM buffers DMM/CMA usage Ivaylo Dimitrov
2022-01-19 10:23 ` [PATCH 1/3] drm: omapdrm: simplify omap_gem_pin Ivaylo Dimitrov
2022-01-19 10:23   ` Ivaylo Dimitrov
2022-01-19 10:23 ` [PATCH 2/3] drm: omapdrm: Support exporting of non-contiguous GEM BOs Ivaylo Dimitrov
2022-01-19 10:23   ` Ivaylo Dimitrov
2022-01-19 10:23 ` [PATCH 3/3] drm: omapdrm: Do no allocate non-scanout GEMs through DMM/TILER Ivaylo Dimitrov
2022-01-19 10:23   ` Ivaylo Dimitrov
2022-02-17 12:46   ` Tomi Valkeinen
2022-02-17 12:46     ` Tomi Valkeinen
2022-02-17 15:29     ` Ivaylo Dimitrov
2022-02-17 15:29       ` Ivaylo Dimitrov
2022-08-12  4:35       ` Yongqin Liu
2022-08-12  4:35         ` Yongqin Liu
2022-08-13  6:58         ` Ivaylo Dimitrov
2022-08-13  6:58           ` Ivaylo Dimitrov
2022-08-14 14:27           ` Yongqin Liu
2022-08-14 14:27             ` Yongqin Liu
2022-08-15  6:23             ` Ivaylo Dimitrov
2022-08-15  6:23               ` Ivaylo Dimitrov
2022-08-17  4:52               ` Yongqin Liu
2022-08-17  4:52                 ` Yongqin Liu
2022-08-18 10:23                 ` Ivaylo Dimitrov
2022-08-18 10:23                   ` Ivaylo Dimitrov
2022-08-29  2:51                   ` Yongqin Liu
2022-08-29  2:51                     ` Yongqin Liu
2022-08-29 13:24                     ` Ivaylo Dimitrov
2022-08-29 13:24                       ` Ivaylo Dimitrov
2022-08-29 14:36                       ` Andrew Davis
2022-08-29 14:36                         ` Andrew Davis
2022-08-30 15:08                         ` Yongqin Liu
2022-08-30 15:08                           ` Yongqin Liu
2022-08-30 18:08                           ` Ivaylo Dimitrov
2022-08-30 18:08                             ` Ivaylo Dimitrov
2022-08-30 18:23                             ` Tomi Valkeinen
2022-08-30 18:23                               ` Tomi Valkeinen
2022-09-02 14:13                               ` Ivaylo Dimitrov
2022-09-02 14:13                                 ` Ivaylo Dimitrov
2022-10-30 22:08   ` H. Nikolaus Schaller
2022-10-30 22:08     ` H. Nikolaus Schaller
2022-10-31  7:05     ` Ivaylo Dimitrov
2022-10-31  7:05       ` Ivaylo Dimitrov
2022-10-31  7:44       ` H. Nikolaus Schaller
2022-10-31  7:44         ` H. Nikolaus Schaller
2022-10-31  7:57         ` H. Nikolaus Schaller
2022-10-31  7:57           ` H. Nikolaus Schaller
2022-10-31  9:58           ` Ivaylo Dimitrov
2022-10-31  9:58             ` Ivaylo Dimitrov
2022-10-31 10:07             ` H. Nikolaus Schaller
2022-10-31 10:07               ` H. Nikolaus Schaller
2022-10-31 10:13       ` Tony Lindgren
2022-10-31 10:13         ` Tony Lindgren
2022-10-31 10:30         ` Ivaylo Dimitrov
2022-10-31 10:30           ` Ivaylo Dimitrov
2022-10-31 11:14         ` Ivaylo Dimitrov
2022-10-31 11:14           ` Ivaylo Dimitrov
2022-02-14  7:08 ` [PATCH 0/3] drm: omapdrm: Fix excessive GEM buffers DMM/CMA usage Ivaylo Dimitrov
2022-02-14  7:08   ` Ivaylo Dimitrov
2022-02-16  8:10 ` Thomas Zimmermann
2022-02-16  8:10   ` Thomas Zimmermann
2022-02-17 13:03 ` Tomi Valkeinen
2022-02-17 13:03   ` Tomi Valkeinen
2022-02-17 16:21   ` Ivaylo Dimitrov
2022-02-17 16:21     ` Ivaylo Dimitrov
2022-03-08  8:51     ` Tomi Valkeinen
2022-03-08  8:51       ` Tomi Valkeinen
2022-03-28  9:46 ` Tomi Valkeinen
2022-03-28  9:46   ` Tomi Valkeinen
2022-03-28 15:30   ` Ivaylo Dimitrov
2022-03-28 15:30     ` Ivaylo Dimitrov

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=1642587791-13222-1-git-send-email-ivo.g.dimitrov.75@gmail.com \
    --to=ivo.g.dimitrov.75@gmail.com \
    --cc=airlied@linux.ie \
    --cc=daniel@ffwll.ch \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-omap@vger.kernel.org \
    --cc=merlijn@wizzup.org \
    --cc=tomba@kernel.org \
    --cc=tony@atomide.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 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.