Re: [PATCH 00/15] Add new debug infrastructure for configfs

["Summers, Stuart" <stuart.summers@intel.com>]

Looks like there's a bug in this rev, sorry about that. I'll respin
with a fix, please ignore this one.

Thanks,
Stuart

On Mon, 2026-10-05 at 19:06 +0000, Stuart Summers wrote:
> Add a new configfs debug group. The intent of this structure is
> to allow us to separate ABI facing configfs entries from those
> which are purely for debug purposes. And it allows us more
> flexibility
> in how we define those parameters used for debug.
> 
> Add a new infrastructure to this debug configfs group that lets us
> easily define the parameters in a quick list. This is primarily
> useful for simple, single-type parameters such as enable/disable
> features or simple values passed. For more complex parameters,
> we will still need to define these separately.
> 
> Pull the GuC target related changes from [1] to fit within
> this new structure and add a new definition for guc_log_level
> on top of the existing module parameter (to ensure we aren't
> impacting existing users of the module parameter).
> 
> Note that the debug parameters here are all to be used "at your
> own risk". Without having in depth knowledge of how these impact
> the software and hardware, there could be unforeseen consequences
> of setting them. As such, they are all wrapped in a
> CONFIG_DRM_XE_DEBUG configfs option.
> 
> In terms of the patches here, I'm sorting the existing parameters
> by name/type. I know we have a few other module parameters that
> could migrate here, but I didn't want to overload this series
> too much, so the focus for now is on the existing configfs entries
> and demonstrating the new structures with the GuC log level and
> target parameters.
> 
> I used GitHub Copilot with Claude pretty extensively through the
> process here and attributed as such. Happy to answer any questions
> around this. Took a bit of time getting back to this series around
> other work, and in that time I was playing around with a few
> different
> models, hence some of the patches are showing multiple of them. I
> tried to attribute each as I was implementing the changes.
> 
> I also decided to drop John Harrison from the NPK patch. It has
> been modified quite a bit from the original, but more importantly
> John is no longer with Intel and that email address isn't available
> any more. If it makes a difference here, John and I had both
> separately
> implemented this same change at different occasions for debug. The
> one I used to start that initial series was cherry-picked from his
> latest variant.
> 
> v2:
>  - In this second revision I did confirm that the guc_log_level
>    module parameter is taking precedence over the configfs parameter
>    and ensured the other parameters seem to be autogenerating and
>    working as expected.
>  - I tried to address all the review feedback from the first
>    revision, [2].
>  - I also did another pass on the sorting since there were a few
>    discrepancies I noticed in the first revision. I kept Gustavo's
>    R-B on that one, but would like an ack before merging at least
>    to confirm the patch is sane.
>  - And finally I moved the getter functions into the X-macro
>    generators so we can autogenerate more of the similar functions
>    between the different parameters in that debug param list.
> v3:
>  - Address a couple of comments from Sashiko around GuC log level
>    input checking and proper guard implementation.
> v4:
>  - More review feedback from Sashiko addressed...
> v5:
>  - Move the goto to a return (more Sashiko feedback) in the GuC
>    log level setter before moving to the X-macro solution.
> v6:
>  - Make the autogenerated X-macro function names more specific to
>    avoid naming collisions (Sashiko again).
> v7:
>  - Fix the couple of pre-existing bugs called out by Sashiko in
>    the prior rev...
>  - Make CONFIGFS_FS a required config for xe to avoid issues with
>    stale values in the fallback getters.
>  - Renamed disable_vram_page_offline to enable_vram_page_offline
>    for a more consistent naming scheme (this was a new configfs
>    entry added since the prior rev).
>  - Converted survivability_mode to a u8 bitmap to allow for
>    extendability in the future. Only bit 0 is defined, so the
> behavior
>    should be the same.
>  - Added an enable_media module parameter at the end of the series.
>    gt_types_allowed is debug-only, so this gives production builds a
>    supported way to leave the media GT alone. The modparam takes
>    precedence over configfs.
>  - Adjust the sorting to be alphabetical for the documentation
>    specifically (Matt)
> v8:
>  - Revert the CONFIGFS_FS require change from the previous rev (Matt)
>  - Rebase on the latest which adds a new ulls configfs that is going
>    into the production list.
>  - Use Assisted-by: LLM as documented on the latest kernel.org site
> for
>    AI attribution.
> 
> Stuart Summers (15):
>   drm/xe: Guard configfs attribute reads in getters
>   drm/xe/configfs: Fix out-of-bounds read in parse_wa_bb_lines()
>   drm/xe/configfs: Copy wa_bb out under the configfs lock
>   drm/xe: Invert vram_page_offline configfs attribute
>   drm/xe: Make survivability_mode configfs attribute a bitmap
>   drm/xe: Sort xe_config_device fields
>   drm/xe: Split out configfs data structures
>   drm/xe: Add a new debug focused configfs group
>   drm/xe: Move debug configfs entries to xe_configfs_debug.c
>   drm/xe/guc: Add configfs support for guc_log_level
>   drm/xe/guc: Add support for NPK as a GuC log target
>   drm/xe: Add infrastructure for debug configfs parameters
>   drm/xe: Migrate existing debug configfs entries to params
>     infrastructure
>   drm/xe: Taint kernel when debug configfs parameters are set
>   drm/xe: Add enable_media module parameter
> 
>  drivers/gpu/drm/xe/Makefile                   |    3 +
>  drivers/gpu/drm/xe/abi/guc_log_abi.h          |    8 +
>  drivers/gpu/drm/xe/xe_configfs.c              | 1110 ++-------------
> --
>  drivers/gpu/drm/xe/xe_configfs.h              |  105 +-
>  drivers/gpu/drm/xe/xe_configfs_debug.c        |  899 +++++++++++++
>  drivers/gpu/drm/xe/xe_configfs_debug.h        |   51 +
>  drivers/gpu/drm/xe/xe_configfs_debug_params.c |  158 +++
>  drivers/gpu/drm/xe/xe_configfs_debug_params.h |  194 +++
>  drivers/gpu/drm/xe/xe_configfs_types.h        |   61 +
>  drivers/gpu/drm/xe/xe_defaults.h              |    6 +
>  drivers/gpu/drm/xe/xe_drm_ras_types.h         |    4 +-
>  drivers/gpu/drm/xe/xe_guc.c                   |   14 +-
>  drivers/gpu/drm/xe/xe_guc_ads.c               |    1 +
>  drivers/gpu/drm/xe/xe_guc_log.c               |    3 +-
>  drivers/gpu/drm/xe/xe_hw_engine.c             |    1 +
>  drivers/gpu/drm/xe/xe_lrc.c                   |   39 +-
>  drivers/gpu/drm/xe/xe_module.c                |    6 +
>  drivers/gpu/drm/xe/xe_module.h                |    1 +
>  drivers/gpu/drm/xe/xe_pci.c                   |   20 +-
>  drivers/gpu/drm/xe/xe_psmi.c                  |    3 +-
>  drivers/gpu/drm/xe/xe_ras.c                   |    9 +-
>  drivers/gpu/drm/xe/xe_rtp.c                   |    3 +-
>  drivers/gpu/drm/xe/xe_survivability_mode.c    |    7 +-
>  drivers/gpu/drm/xe/xe_ttm_vram_mgr.c          |    2 +-
>  24 files changed, 1646 insertions(+), 1062 deletions(-)
>  create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.c
>  create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.h
>  create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug_params.c
>  create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug_params.h
>  create mode 100644 drivers/gpu/drm/xe/xe_configfs_types.h
>