index
:
mesa/shader-db
dualt
loops
master
sso
turnip-support
unchanged
Collection of shaders for offline analysis. (mirrored from https://gitlab.freedesktop.org/mesa/shader-db)
root
summary
refs
log
tree
commit
diff
log msg
author
committer
range
path:
root
/
report.py
Age
Commit message (
Expand
)
Author
Files
Lines
2022-05-24
report.py: Fix scheduler mode descriptions
Kenneth Graunke
1
-2
/
+1
2022-03-07
report.py: Add support for nouveau-formatted shader-db output.
Emma Anholt
1
-12
/
+28
2022-03-07
report.py: Add support for shader-db without a loops count.
Emma Anholt
1
-1
/
+1
2021-06-23
report: Consider "waves" along with "threads" as more_is_better.
Emma Anholt
1
-1
/
+1
2021-01-19
report.py: Don't require spills/fills
Matt Turner
1
-0
/
+3
2020-12-02
report.py: fix crash on get_sched_mode
Erico Nunes
1
-0
/
+3
2020-07-07
report: Fix thread count printing
Alyssa Rosenzweig
1
-4
/
+4
2020-04-23
report.py: Handle scheduler mode changes
Matt Turner
1
-5
/
+27
2020-04-23
report.py: Print spill/fill changes beside instruction/cycle changes
Matt Turner
1
-4
/
+15
2020-04-23
report.py: Remove semicolons
Matt Turner
1
-5
/
+5
2020-04-23
report.py: Remove tabs
Matt Turner
1
-5
/
+5
2019-10-30
report: Don't use len(x) > 0
Ian Romanick
1
-2
/
+1
2019-10-24
report: add argument to ignore loops
Rob Clark
1
-1
/
+2
2019-10-24
report: add arg to filter on shader stage
Rob Clark
1
-3
/
+7
2019-05-29
report: Fix heuristic for default set of measurements
Caio Marcelo de Oliveira Filho
1
-2
/
+2
2019-05-28
report: fix '-m inst' case
Alejandro PiƱeiro
1
-0
/
+8
2019-05-16
report: Stop hard-coding the default set of measurements.
Eric Anholt
1
-1
/
+11
2019-05-09
report: Print out total CPU time
Jason Ekstrand
1
-1
/
+22
2019-05-09
report: Stop having argparse call get_results
Jason Ekstrand
1
-19
/
+22
2019-04-04
report: For a thread count, flip the polarity for hurt vs helped.
Eric Anholt
1
-1
/
+3
2019-03-08
report: Generalize the results collection.
Eric Anholt
1
-14
/
+27
2019-03-01
report.py: Gather and log some statistics about the helped / hurt data
Ian Romanick
1
-1
/
+100
2018-03-06
report.py: Print small percentages as <.01%
Ian Romanick
1
-1
/
+9
2018-03-06
report.py: Add option to only display measurements that have changes
Ian Romanick
1
-14
/
+24
2018-03-06
report.py: Add option to only display the final summary
Ian Romanick
1
-28
/
+32
2018-03-06
report.py: Don't use 'len(x) > 0' to check list emptiness
Ian Romanick
1
-4
/
+4
2016-04-28
report.py: Prevent division by zero.
Matt Turner
1
-4
/
+2
2016-04-25
report.py: Add spill/fill reports
Kenneth Graunke
1
-2
/
+6
2016-01-05
report.py: Handle TCS and TES shaders.
Kenneth Graunke
1
-1
/
+1
2016-01-05
report.py: Explicitly require python3
Kenneth Graunke
1
-1
/
+1
2016-01-05
Fix Python2ism that crept into report.py.
Kenneth Graunke
1
-1
/
+1
2015-12-01
report.py: rework and update for cycle info
Connor Abbott
1
-75
/
+89
2015-10-01
report.py: Report loop changes separately.
Matt Turner
1
-11
/
+32
2015-01-20
report.py: Print counts of helped/hurt programs.
Kenneth Graunke
1
-4
/
+8
2015-01-13
report.py: Use the format printed by Matt's new C runner.
Kenneth Graunke
1
-1
/
+1
2014-11-12
Sort the list of helped programs
Jason Ekstrand
1
-5
/
+7
2014-07-16
report.py: call get_result method on file arguments
Dylan Baker
1
-2
/
+2
2014-07-13
report.py: Fix PEP8 issues
Dylan Baker
1
-11
/
+17
2014-07-13
report.py: use argparse instead of getopt
Dylan Baker
1
-38
/
+13
2014-07-13
python: 2 blanks between toplevel functions and classes
Dylan Baker
1
-0
/
+8
2014-07-13
python: cleanup imports
Dylan Baker
1
-2
/
+1
2014-07-13
python: Replace tabs with spaces
Dylan Baker
1
-1
/
+1
2013-10-14
Summarize the count of shaders lost/gained.
Eric Anholt
1
-1
/
+2
2013-03-25
report.py: Also port to Python 3.
Kenneth Graunke
1
-19
/
+12
2013-03-25
Remove debug code that slipped into the report.
Eric Anholt
1
-1
/
+0
2013-03-25
Make the report script handle extra whitespace in the run.py output.
Eric Anholt
1
-1
/
+2
2012-06-06
Convert the scripts to python and make them way better.
Eric Anholt
1
-0
/
+144