Age | Commit message (Collapse) | Author |
|
Make sure we don't have any space in there.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
In the case where an expression is a single variable like
expression="$EuThreadsCount" we should resolve that last variable to a
codegen expression.
Take the opportunity to uniformize the hash table keys (require "$"
prefix everywhere).
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Rather than having all GTs types, just rely on the chipset name to
find the set of registers that are relevant.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
There is an effort to drop the "Gen" prefix from much of our codebase.
This just applies this to the metrics.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
List all supported generations so we can fail sooner.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Yet another file.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Adding a register, similar to what was done for RenderBasic.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Simply moving the counter around.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
The availability of those counters depends on the topology.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Simple order change.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Programming an additional FLEX register.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Working with an offset means we can have the PERFCNT snapshots stored
at an offset in the existing accumulator rather than having a specific
accumulator just for them.
This isn't really useful for IGT because it only looks at the OA
buffer where those registers aren't stored, but it is useful for perf
queries in Mesa.
This bumps the version of the i915_perf library to reflect the API
change (see perf.h).
v2: Also bump so version
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Petri Latvala <petri.latvala@intel.com>
|
|
New XML files for DG1.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
New XML files for RKL.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Those metrics are mostly about exposing the amount of FPU usage with a
compute filter in the FLEX EU per registers.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Programming for GT1s & GT2s appear to be different.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
No functional changes.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
This script was previous run in python2, but because IGT is python3
only there are differences in what it generates. In particular the
attribute order differs in the tostring() of the ElementTree
function...
Reorder them alphabetically manually.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 53f8f541caae ("lib: Add i915_perf library")
Acked-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
It got removed in Python 3.9 and ever since Python 3.3 it's:
from xml.etree.ElementTree import *
Link: https://docs.python.org/3.9/whatsnew/3.9.html#removed
Issue: https://gitlab.freedesktop.org/drm/igt-gpu-tools/-/issues/86#note_520308
Cc: Petri Latvala <petri.latvala@intel.com>
Cc: Lionel Landwerlin <lionel.g.landwerlin@linux.intel.com>
Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>
|
|
Some HW woes are forcing us to store EU activity counters in a
different way.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
|
|
Initially all the generated code was per generation. Eventually we
grouped it into a single file to reuse as much as possible equation
code (this reduce binary size by a factor). So many equations are just
the same from generation to generation.
But this generated file is 200k lines long...
This change puts all the equations into a single file, so that we
reuse as much code as possible, and then breaks down the metric sets &
register configurations into per generation files.
v2: Split registers away from the metric set descriptions
v3: Fix automake
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
|
|
Gputop had support for generating mathml figures of the post
processing equations. It's not useful here.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
|
|
This library contains the database of performance metrics for each
supported generation as well as its associated programming.
It also allow for the user to get human readable data out of HW
generated data.
v2: Switch a few python scripts to python3
v3: Fix some python2-ism
s/lkf/ehl/
Autotools support
v4: set soversion on i915_perf.so
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Petri Latvala <petri.latvala@intel.com>
|