Igt-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Kamil Konieczny <kamil.konieczny@linux.intel.com>
To: igt-dev@lists.freedesktop.org
Cc: Kamil Konieczny <kamil.konieczny@linux.intel.com>,
	Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com>
Subject: [PATCH i-g-t v6 0/8] introduce Xe multigpu and other multi-GPU helpers
Date: Thu,  1 Feb 2024 16:50:39 +0100	[thread overview]
Message-ID: <20240201155047.48608-1-kamil.konieczny@linux.intel.com> (raw)

Introduce some multi-gpu function helpers and macros. Change in drmtest
base opening function allows it to use __drm_open_driver_another(N, ...)
out of order, for example on board with three discrete GPUs:
__drm_open_driver_another(2, DRIVER_INTEL);
__drm_open_driver_another(0, DRIVER_INTEL);

Second solution for Xe is based on filtered views. Both allows to
quickly write tests. There is still drawback of not printing in
children logs <g:gpu-number> for first opened device. I renamed
lib from lib/i915/igt_multigpu.* to lib/intel_multigu.* as I do not
know of any non Intel developer using igt for such tests.

v2: corrected two patches which introduced multigpu lib (Dominik)
  rebased patch intoducing Xe multigpu macro, corrected description
v3: corrected include in first multigpu patch (Dominik)
  fixed typo (Dominik), refactoring code with Xe and gem macro (Kamil)
v4: corrected typo in macro (Dominik)
v5: removed cached names and relaxing checks for filtered devices
  in 3/8 patch, added Janusz to cc in 1/8...5/8 (Kamil)
v6: extended description of patches 1 and 2, fixed bug in 3/8 (Janusz)
  improved usage of _is_already_opened() (Kamil)

Cc: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com>

Dominik Karol Piątkowski (4):
  lib/igt_device_scan: Introduce filtering out non-PCI devices
  lib/drmtest: Introduced drm_open_driver_another
  lib/intel_multigpu: Introduce library for multi-GPU scenarios
  lib/intel_multigpu: Introduced gem_multigpu_count_class and
    igt_multi_fork_foreach_gpu

Kamil Konieczny (4):
  lib/drmtest: allow opening cards out of order
  lib/intel_multigpu: Add xe_multi_fork_foreach_gpu
  tests/intel/xe_exec_basic: add multigpu subtests
  tests/intel/gem_mmap: add basic multi-GPU subtest

 lib/drmtest.c               |  85 ++++++++++++++++++-----
 lib/drmtest.h               |   1 +
 lib/igt_device_scan.c       |  21 ++++++
 lib/igt_device_scan.h       |   2 +
 lib/intel_multigpu.c        | 135 ++++++++++++++++++++++++++++++++++++
 lib/intel_multigpu.h        |  36 ++++++++++
 lib/meson.build             |   1 +
 tests/intel/gem_mmap.c      |  77 ++++++++++++--------
 tests/intel/xe_exec_basic.c |  36 ++++++++++
 9 files changed, 350 insertions(+), 44 deletions(-)
 create mode 100644 lib/intel_multigpu.c
 create mode 100644 lib/intel_multigpu.h

-- 
2.42.0


             reply	other threads:[~2024-02-01 15:50 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-01 15:50 Kamil Konieczny [this message]
2024-02-01 15:50 ` [PATCH i-g-t v6 1/8] lib/igt_device_scan: Introduce filtering out non-PCI devices Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 2/8] lib/drmtest: Introduced drm_open_driver_another Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 3/8] lib/drmtest: allow opening cards out of order Kamil Konieczny
2024-02-02 11:38   ` Janusz Krzysztofik
2024-02-02 15:32     ` Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 4/8] lib/intel_multigpu: Introduce library for multi-GPU scenarios Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 5/8] lib/intel_multigpu: Introduced gem_multigpu_count_class and igt_multi_fork_foreach_gpu Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 6/8] lib/intel_multigpu: Add xe_multi_fork_foreach_gpu Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 7/8] tests/intel/xe_exec_basic: add multigpu subtests Kamil Konieczny
2024-02-01 15:50 ` [PATCH i-g-t v6 8/8] tests/intel/gem_mmap: add basic multi-GPU subtest Kamil Konieczny
2024-02-01 17:03 ` ✓ CI.xeBAT: success for introduce Xe multigpu and other multi-GPU helpers (rev6) Patchwork
2024-02-01 17:21 ` ✓ Fi.CI.BAT: " Patchwork
2024-02-01 19:14 ` ✓ 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=20240201155047.48608-1-kamil.konieczny@linux.intel.com \
    --to=kamil.konieczny@linux.intel.com \
    --cc=igt-dev@lists.freedesktop.org \
    --cc=janusz.krzysztofik@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