When I was running tests for this [0] revert, I saw different performance consequences than expected (not reversal of the original [1] regression). After rebasing, I am seeing yet another set of unexpected performance differences.
This has reminded me of Stabilizer [2] utility, I learned about it by watching this [3] video. What do you think, is it possible that also for VPP, some regressions are caused just by compiler changing the memory layout (without any "real" change in the program logic)? If yes, should we try to apply Stabilizer to VPP? (Re [0], I am leaning towards giving up on it.) Vratko. [0] https://gerrit.fd.io/r/c/vpp/+/35164 [1] https://gerrit.fd.io/r/c/vpp/+/34568/2#message-63e1cdafce46e7d389ffb45d79df4fb622681888 [2] https://github.com/ccurtsinger/stabilizer [3] https://www.youtube.com/watch?v=r-TLSBdHe1A
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#20889): https://lists.fd.io/g/vpp-dev/message/20889 Mute This Topic: https://lists.fd.io/mt/89321880/21656 Group Owner: vpp-dev+ow...@lists.fd.io Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
