From: Rodrigo Vivi <rodrigo.vivi@intel.com>
To: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: <intel-xe@lists.freedesktop.org>,
Lucas De Marchi <lucas.demarchi@intel.com>
Subject: Re: [PATCH v2 0/6] Refactor PF debugfs
Date: Mon, 29 Sep 2025 19:21:03 -0400 [thread overview]
Message-ID: <aNsUX8w6JLoDQPa_@intel.com> (raw)
In-Reply-To: <20250928140029.198847-1-michal.wajdeczko@intel.com>
On Sun, Sep 28, 2025 at 04:00:22PM +0200, Michal Wajdeczko wrote:
> In new debugfs layout, all SR-IOV files are grouped under single
> "sriov" directory, where each SR-IOV function is represented by
> its own separate directory, with individual tile/GT structure.
>
> This allows to place GGTT/VRAM attributes correctly under a tile,
> without polluting 'native' entry with many function directories.
>
> /sys/kernel/debug/dri/0000:4d:00.0/
> ├── ...
> ├── gt0 -> tile0/gt0
> ├── sriov
> │ ├── pf
> │ │ ├── tile0
> │ │ │ ├── ggtt_available
> │ │ │ ├── ggtt_provisioned
> │ │ │ ├── ggtt_spare
> │ │ │ ├── gt0
> │ │ │ │ ├── adverse_events
> │ │ │ │ ├── contexts_provisioned
> │ │ │ │ ├── contexts_spare
> │ │ │ │ ├── doorbells_provisioned
> │ │ │ │ ├── doorbells_spare
> │ │ │ │ ├── exec_quantum_ms
> │ │ │ │ ├── ggtt_available -> ../ggtt_available
> │ │ │ │ ├── ggtt_provisioned -> ../ggtt_provisioned
> │ │ │ │ ├── ggtt_spare -> ../ggtt_spare
> │ │ │ │ ├── lmem_provisioned -> ../vram_provisioned
> │ │ │ │ ├── lmem_spare -> ../vram_spare
> │ │ │ │ ├── preempt_timeout_us
> │ │ │ │ ├── reset_engine
> │ │ │ │ ├── runtime_registers
> │ │ │ │ ├── sample_period_ms
> │ │ │ │ ├── sched_if_idle
> │ │ │ │ ├── sched_priority
> │ │ │ │ ├── threshold_cat_error_count
> │ │ │ │ ├── threshold_doorbell_time_us
> │ │ │ │ ├── threshold_engine_reset_count
> │ │ │ │ ├── threshold_guc_time_us
> │ │ │ │ ├── threshold_irq_time_us
> │ │ │ │ └── threshold_page_fault_count
> │ │ │ ├── vram_provisioned
> │ │ │ └── vram_spare
> │ │ ├── versions
> │ │ └── vfs
> │ ├── vf1
> │ │ └── tile0
> │ │ ├── ggtt_quota
> │ │ ├── gt0
> │ │ │ ├── config_blob
> │ │ │ ├── contexts_quota
> │ │ │ ├── control
> │ │ │ ├── doorbells_quota
> │ │ │ ├── exec_quantum_ms
> │ │ │ ├── ggtt_quota -> ../ggtt_quota
> │ │ │ ├── guc_state
> │ │ │ ├── lmem_quota -> ../vram_quota
> │ │ │ ├── preempt_timeout_us
> │ │ │ ├── sched_priority
> │ │ │ ├── threshold_cat_error_count
> │ │ │ ├── threshold_doorbell_time_us
> │ │ │ ├── threshold_engine_reset_count
> │ │ │ ├── threshold_guc_time_us
> │ │ │ ├── threshold_irq_time_us
> │ │ │ └── threshold_page_fault_count
> │ │ └── vram_quota
> │ ├── vf2
> │ :
> ├── tile0
> │ ├── ggtt
> │ ├── gt0
> │ │ ├── ...
> │ │ ├── pf -> ../../sriov/pf/tile0/gt0
> │ │ ├── vf1 -> ../../sriov/vf1/tile0/gt0
> │ │ ├── vf2 -> ../../sriov/vf2/tile0/gt0
> : : └── ...
Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
>
> v1: https://patchwork.freedesktop.org/series/155064/#rev1
> v2: use VRAM instead of LMEM (Lucas, Rodrigo)
>
> Cc: Lucas De Marchi <lucas.demarchi@intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
>
> Michal Wajdeczko (6):
> drm/xe/pf: Promote PF debugfs function to its own file
> drm/xe/pf: Create separate debugfs tree for SR-IOV files
> drm/xe/pf: Populate SR-IOV debugfs tree with tiles
> drm/xe/pf: Move SR-IOV GT debugfs files to new tree
> drm/xe/debugfs: Promote xe_tile_debugfs_simple_show
> drm/xe/pf: Make GGTT/LMEM debugfs files per-tile
>
> drivers/gpu/drm/xe/Makefile | 4 +-
> drivers/gpu/drm/xe/xe_debugfs.c | 2 +-
> drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c | 389 ++++++++++--------
> drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.h | 1 +
> drivers/gpu/drm/xe/xe_sriov_pf.c | 42 --
> drivers/gpu/drm/xe/xe_sriov_pf.h | 5 -
> drivers/gpu/drm/xe/xe_sriov_pf_debugfs.c | 108 +++++
> drivers/gpu/drm/xe/xe_sriov_pf_debugfs.h | 18 +
> drivers/gpu/drm/xe/xe_tile_debugfs.c | 14 +-
> drivers/gpu/drm/xe/xe_tile_debugfs.h | 3 +
> drivers/gpu/drm/xe/xe_tile_sriov_pf_debugfs.c | 247 +++++++++++
> drivers/gpu/drm/xe/xe_tile_sriov_pf_debugfs.h | 15 +
> 12 files changed, 621 insertions(+), 227 deletions(-)
> create mode 100644 drivers/gpu/drm/xe/xe_sriov_pf_debugfs.c
> create mode 100644 drivers/gpu/drm/xe/xe_sriov_pf_debugfs.h
> create mode 100644 drivers/gpu/drm/xe/xe_tile_sriov_pf_debugfs.c
> create mode 100644 drivers/gpu/drm/xe/xe_tile_sriov_pf_debugfs.h
>
> --
> 2.47.1
>
prev parent reply other threads:[~2025-09-29 23:21 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-09-28 14:00 [PATCH v2 0/6] Refactor PF debugfs Michal Wajdeczko
2025-09-28 14:00 ` [PATCH v2 1/6] drm/xe/pf: Promote PF debugfs function to its own file Michal Wajdeczko
2025-09-28 14:00 ` [PATCH v2 2/6] drm/xe/pf: Create separate debugfs tree for SR-IOV files Michal Wajdeczko
2025-09-29 21:19 ` Lucas De Marchi
2025-09-28 14:00 ` [PATCH v2 3/6] drm/xe/pf: Populate SR-IOV debugfs tree with tiles Michal Wajdeczko
2025-09-28 14:00 ` [PATCH v2 4/6] drm/xe/pf: Move SR-IOV GT debugfs files to new tree Michal Wajdeczko
2025-09-28 14:00 ` [PATCH v2 5/6] drm/xe/debugfs: Promote xe_tile_debugfs_simple_show Michal Wajdeczko
2025-09-28 14:00 ` [PATCH v2 6/6] drm/xe/pf: Make GGTT/LMEM debugfs files per-tile Michal Wajdeczko
2025-09-29 21:21 ` Lucas De Marchi
2025-09-28 14:07 ` ✗ CI.checkpatch: warning for Refactor PF debugfs Patchwork
2025-09-28 14:08 ` ✓ CI.KUnit: success " Patchwork
2025-09-28 14:45 ` ✓ Xe.CI.BAT: " Patchwork
2025-09-28 15:59 ` ✗ Xe.CI.Full: failure " Patchwork
2025-09-29 21:56 ` Michal Wajdeczko
2025-09-29 23:21 ` Rodrigo Vivi [this message]
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=aNsUX8w6JLoDQPa_@intel.com \
--to=rodrigo.vivi@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=lucas.demarchi@intel.com \
--cc=michal.wajdeczko@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.