[00/11] Refactor to expand subslice mask (rev 2)
mbox series

Message ID 20190823160307.180813-1-stuart.summers@intel.com
Headers show
Series
  • Refactor to expand subslice mask (rev 2)
Related show

Message

Summers, Stuart Aug. 23, 2019, 4:02 p.m. UTC
Better match what is presented to userspace through the
I915_QUERY_TOPOLOGY_INFO ioctl. The index into this array is
then calculated:
  slice * subslice stride + subslice index / 8

Note this is the second iteration of an original patch to implement
the same. There are a couple of minor code changes based on changes
since the first series was posted. Additionally, the original patch
has been split into several smaller patches with more isolated
changes based on review feedback in that first series.

Link to the original series:
https://patchwork.freedesktop.org/series/59742/

v2: Fix 32-bit build
v3: Fix typo in haswell sseu info routine and fix SSEU workaround
    print
v4: Merge patch to HSW in previous revision with patch to
    set subslice_mask for each platform and address feedback
    from Chris
v5: No changes in code. Resending due to unrelated failures - confirmed
    no failures with trybot.
v6: Minor change in intel_workarounds.c to use the new helper function
v7: Split set sublice function into multiple more manageable patches
    And add a couple of GEM_BUG_ON checks requested by Chris
v8: Address new comments from Chris including adding/moving
    a couple of ss/eu_stride checks to those previously reviewed
    patches earlier in this series
v9: Use memcpy to set subslices (Chris)

Stuart Summers (11):
  drm/i915: Use variable for debugfs device status
  drm/i915: Add function to set SSEU info per platform
  drm/i915: Add subslice stride runtime parameter
  drm/i915: Add EU stride runtime parameter
  drm/i915: Use local variables for subslice_mask for device info
  drm/i915: Add function to set subslices
  drm/i915: Use subslice stride to set subslices for a given slice
  drm/i915: Add function to determine if a slice has a subslice
  drm/i915: Refactor instdone loops on new subslice functions
  drm/i915: Add new function to copy subslices for a slice
  drm/i915: Expand subslice mask

 drivers/gpu/drm/i915/gt/intel_engine_cs.c    |   3 +-
 drivers/gpu/drm/i915/gt/intel_engine_types.h |  30 +++--
 drivers/gpu/drm/i915/gt/intel_hangcheck.c    |   3 +-
 drivers/gpu/drm/i915/gt/intel_sseu.c         |  37 +++++-
 drivers/gpu/drm/i915/gt/intel_sseu.h         |  32 ++++-
 drivers/gpu/drm/i915/gt/intel_workarounds.c  |   5 +-
 drivers/gpu/drm/i915/i915_debugfs.c          |  49 +++++---
 drivers/gpu/drm/i915/i915_gpu_error.c        |   5 +-
 drivers/gpu/drm/i915/i915_query.c            |  10 +-
 drivers/gpu/drm/i915/intel_device_info.c     | 122 +++++++++----------
 10 files changed, 182 insertions(+), 114 deletions(-)

Comments

Chris Wilson Aug. 23, 2019, 6:15 p.m. UTC | #1
Quoting Patchwork (2019-08-23 19:11:52)
> == Series Details ==
> 
> Series: Refactor to expand subslice mask (rev 2)
> URL   : https://patchwork.freedesktop.org/series/65709/
> State : success
> 
> == Summary ==
> 
> CI Bug Log - changes from CI_DRM_6777 -> Patchwork_14171
> ====================================================
> 
> Summary
> -------
> 
>   **SUCCESS**
> 
>   No regressions found.
> 
>   External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_14171/

The delta from the previous run is small enough for me to not worry
about. Pushed,
-Chris