ACK: [SRU][P:linux][PATCH v2 0/3] drm/xe: Lite restore breaks fdinfo drm-cycles-rcs reporting

John Cabaj john.cabaj at canonical.com
Wed Aug 6 04:07:46 UTC 2025


On 8/5/25 5:15 PM, Thibault Ferrante wrote:
> BugLink: https://bugs.launchpad.net/bugs/2119526
> 
> [ Impact ]
> 
>   * GPU usage reported wrongly.
>   * Upstream bug report: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4532
> 
> [ Fix ]
> 
>   * Upstream fixes:
> 617d824c53 drm/xe: Add WA BB to capture active context utilization
> 9c7632faad drm/xe/lrc: Use a temporary buffer for WA BB
> 741d3ef8b8 drm/xe: Save the gt pointer in lrc and drop the tile
> 
> [ Test Plan ]
> 
>   * Trigger 100% usage of GPU, check for usage.
> 
> [ Where problems could occur ]
> 
>   * Regression in drm/xe driver and GPU management.
> 
> [ Other Info ]
> 
>   * Tested and submitted by intel.
>   * Upstream fix afcad9241177 'drm/xe: Make WA BB part of LRC BO' skipped as deemed not usefull by intel further testing.
> 
> v2:
>   - Switch 617d824c5 to backport (There is only context changes, but seems it can't be solved automatically).
>   - Mention upstream fix.
> 
> Lucas De Marchi (1):
>    drm/xe/lrc: Use a temporary buffer for WA BB
> 
> Umesh Nerlige Ramappa (2):
>    drm/xe: Add WA BB to capture active context utilization
>    drm/xe: Save the gt pointer in lrc and drop the tile
> 
>   drivers/gpu/drm/xe/regs/xe_engine_regs.h |   5 +
>   drivers/gpu/drm/xe/regs/xe_lrc_layout.h  |   2 +
>   drivers/gpu/drm/xe/xe_device_types.h     |   2 +
>   drivers/gpu/drm/xe/xe_exec_queue.c       |   2 +-
>   drivers/gpu/drm/xe/xe_guc_submit.c       |   2 +-
>   drivers/gpu/drm/xe/xe_lrc.c              | 213 +++++++++++++++++++++--
>   drivers/gpu/drm/xe/xe_lrc.h              |   5 +-
>   drivers/gpu/drm/xe/xe_lrc_types.h        |   9 +-
>   drivers/gpu/drm/xe/xe_pci.c              |   2 +
>   drivers/gpu/drm/xe/xe_pci_types.h        |   1 +
>   drivers/gpu/drm/xe/xe_trace_lrc.h        |   8 +-
>   11 files changed, 228 insertions(+), 23 deletions(-)
> 

Acked-by: John Cabaj <john.cabaj at canonical.com>





More information about the kernel-team mailing list