From: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
To: igt-dev@lists.freedesktop.org
Cc: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
"Matthew Brost" <matthew.brost@intel.com>,
"Maarten Lankhorst" <maarten.lankhorst@linux.intel.com>,
"Michal Mrozek" <michal.mrozek@intel.com>,
"John Falkowski" <john.falkowski@intel.com>,
"Rodrigo Vivi" <rodrigo.vivi@intel.com>,
"Lahtinen Joonas" <joonas.lahtinen@intel.com>
Subject: [PATCH i-g-t 0/4] xe: watch queue event support and VM restart recovery
Date: Fri, 12 Jun 2026 13:06:15 +0200 [thread overview]
Message-ID: <20260612110619.103198-1-thomas.hellstrom@linux.intel.com> (raw)
This series adds IGT support for two new Xe kernel interfaces: the watch
queue event notification mechanism and the VM restart IOCTL.
The watch queue allows a userspace process to subscribe to device-scoped
events delivered through a pipe — notably VM errors caused by memory
pressure. The VM restart IOCTL lets userspace synchronously trigger a
re-run of the preempt-rebind worker to recover a VM after such an error.
Together they enable a fault-tolerance loop: a listener thread receives
VM error events carrying an error code and VM id; when the error is
-ENOMEM or -ENOSPC the test can immediately attempt to restart the
affected VM rather than letting it stay faulted.
The xe_exec_compute_mode test is extended to demonstrate this: it
subscribes to file events at fixture time, logs all notifications, and
automatically calls VM restart on memory-pressure errors.
Thomas Hellström (4):
lib/xe: add xe_vm_restart ioctl helper
lib/xe: add xe_watch listener for watch queue events
tests/intel/xe_exec_compute_mode: Add a listener for file events
tests/intel/xe_exec_compute_mode: Restart VM on ENOMEM/ENOSPC errors
include/drm-uapi/xe_drm.h | 84 +++++++++++
include/drm-uapi/xe_drm_events.h | 71 +++++++++
lib/meson.build | 1 +
lib/xe/xe_ioctl.c | 44 ++++++
lib/xe/xe_ioctl.h | 2 +
lib/xe/xe_watch.c | 221 +++++++++++++++++++++++++++++
lib/xe/xe_watch.h | 80 +++++++++++
tests/intel/xe_exec_compute_mode.c | 92 +++++++++++-
8 files changed, 591 insertions(+), 4 deletions(-)
create mode 100644 include/drm-uapi/xe_drm_events.h
create mode 100644 lib/xe/xe_watch.c
create mode 100644 lib/xe/xe_watch.h
--
2.54.0
next reply other threads:[~2026-06-12 11:07 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-12 11:06 Thomas Hellström [this message]
2026-06-12 11:06 ` [PATCH i-g-t 1/4] lib/xe: add xe_vm_restart ioctl helper Thomas Hellström
2026-06-12 14:38 ` Kamil Konieczny
2026-06-12 11:06 ` [PATCH i-g-t 2/4] lib/xe: add xe_watch listener for watch queue events Thomas Hellström
2026-06-12 11:06 ` [PATCH i-g-t 3/4] tests/intel/xe_exec_compute_mode: Add a listener for file events Thomas Hellström
2026-06-12 11:06 ` [PATCH i-g-t 4/4] tests/intel/xe_exec_compute_mode: Restart VM on ENOMEM/ENOSPC errors Thomas Hellström
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=20260612110619.103198-1-thomas.hellstrom@linux.intel.com \
--to=thomas.hellstrom@linux.intel.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=john.falkowski@intel.com \
--cc=joonas.lahtinen@intel.com \
--cc=maarten.lankhorst@linux.intel.com \
--cc=matthew.brost@intel.com \
--cc=michal.mrozek@intel.com \
--cc=rodrigo.vivi@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 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.