[PATCH] D60000: [llvm-exegesis] Post-processing for chained instrs in latency mode (PR41275)

[Roman Lebedev via Phabricator via llvm-commits]

lebedev.ri created this revision.
lebedev.ri added reviewers: courbet, gchatelet.
lebedev.ri added a project: LLVM.
Herald added subscribers: jdoerfert, tschuett, mgorny.

Ok, so this turned out to be easier than i expected.
Also, i initially thought that other modes might need this post-processing,
but i'm not sure which opcodes are affected there, if any.

The results look much better.
On BdVer2 this exposes at least one stable sched cluster
that has inconsistent values from from the measurements,
and a dozen or so somewhat-unstable clusters that also are inconsistent.
F8563140: latency-clusters-stable.html <https://reviews.llvm.org/F8563140> F8563135: latency-clusters-unstable.html <https://reviews.llvm.org/F8563135>

Resolves(?) PR41275 <https://bugs.llvm.org/show_bug.cgi?id=41275>


Repository:
  rL LLVM

https://reviews.llvm.org/D60000

Files:
  test/tools/llvm-exegesis/X86/analysis-latency-instruction-chaining.test
  tools/llvm-exegesis/lib/Analysis.cpp
  tools/llvm-exegesis/lib/Analysis.h
  tools/llvm-exegesis/lib/CMakeLists.txt
  tools/llvm-exegesis/lib/PostProcessing.cpp
  tools/llvm-exegesis/lib/PostProcessing.h
  tools/llvm-exegesis/llvm-exegesis.cpp

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D60000.192864.patch
Type: text/x-patch
Size: 12714 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190329/8b988d7c/attachment.bin>