Igt-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Adam Miszczak <adam.miszczak@linux.intel.com>
To: igt-dev@lists.freedesktop.org
Cc: marcin.bernatowicz@linux.intel.com, kamil.konieczny@linux.intel.com
Subject: [PATCH i-g-t 0/3] tools/vmtb: Parallel multi-VM migration and tests reconfiguration
Date: Thu, 15 Oct 2026 09:10:34 +0200	[thread overview]
Message-ID: <20261015071037.10898-1-adam.miszczak@linux.intel.com> (raw)

This series extends VMTB VF migration validation coverage with parallel
multi-VM migration and reconfigures test variants to auto provisioning.

Patch 1:
Extend BaseTestBusyMigration to migrate multiple VMs concurrently.
Source/destination pairs go from the edges to the center
(VM0->VM(N-1), VM1->VM(N-2), ...).
Save and load use two-phase async operations: all migrations start
first, then each one is polled for completion.
The IGT workload runs on every source VM.
The single-VM (same source and destination VF) flow is kept.
For odd VM counts, the center VM is not migrated.

Patch 2:
Skip double migration tests when the debugfs 'resfix_stoppers' hook
isn't available (KMD no-debug build), instead of failing them.

Patch 3:
Reconfigure the tests setup:
- switch basic and migration test variants to auto resources and
  scheduling provisioning
- drop the 'full' migration variant; busy migration tests
  1xVF and 2xVF config scope by default
- add TestBusyMigrationParallel with 4xVF and 6xVF variants, which
  uses the parallel migration support from patch 1
- remove the manual EQ/PT override in test_2vm_pause_resume, which is
  no longer needed since the driver provides finite defaults
- limit TestVfDriverLoadRemove to 2 VMs to avoid exhausting host
  memory on systems with limited host RAM (<=16GB)

Signed-off-by: Adam Miszczak <adam.miszczak@linux.intel.com>
Cc: Marcin Bernatowicz <marcin.bernatowicz@linux.intel.com>
Cc: Kamil Konieczny <kamil.konieczny@linux.intel.com>

Adam Miszczak (3):
  tools/vmtb: Support parallel multi-VM busy migration
  tools/vmtb: Skip double migration tests on no-debug config
  tools/vmtb: Reconfigure tests setup

 tools/vmtb/vmm_flows/test_basic.py     |  15 +-
 tools/vmtb/vmm_flows/test_migration.py | 235 ++++++++++++++++---------
 2 files changed, 157 insertions(+), 93 deletions(-)

-- 
2.39.1


             reply	other threads:[~2026-10-15  7:52 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-15  7:10 Adam Miszczak [this message]
2026-10-15  7:10 ` [PATCH i-g-t 1/3] tools/vmtb: Support parallel multi-VM busy migration Adam Miszczak
2026-10-16 17:01   ` Bernatowicz, Marcin
2026-10-15  7:10 ` [PATCH i-g-t 2/3] tools/vmtb: Skip double migration tests on no-debug config Adam Miszczak
2026-10-16 17:02   ` Bernatowicz, Marcin
2026-10-15  7:10 ` [PATCH i-g-t 3/3] tools/vmtb: Reconfigure tests setup Adam Miszczak
2026-10-16 17:02   ` Bernatowicz, Marcin
2026-10-15  9:05 ` ✓ i915.CI.BAT: success for tools/vmtb: Parallel multi-VM migration and tests reconfiguration Patchwork
2026-10-15 11:37 ` ✓ i915.CI.Full: " Patchwork
2026-10-15 15:21 ` ✓ Xe.CI.BAT: " Patchwork
2026-10-15 20:09 ` ✗ Xe.CI.FULL: failure " Patchwork
2026-10-15 21:35 ` 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=20261015071037.10898-1-adam.miszczak@linux.intel.com \
    --to=adam.miszczak@linux.intel.com \
    --cc=igt-dev@lists.freedesktop.org \
    --cc=kamil.konieczny@linux.intel.com \
    --cc=marcin.bernatowicz@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