Just a heads up, it looks like this landed sometime last week for platforms that support PGO.
This has an unintended consequence of making it look like perf data for integration branches went awol, but in fact you need to switch from looking at "opt" data to "pgo" data. Unfortunately since we didn't used to run PGO on integration branches you'll also need to include opt data in your view for continuity. For example I've resorted to using an opt/pgo mashup to track memory regressions [1]. This seems to work okay for memory, but I can imagine it won't be a fair comparison for talos tests. -e [1] https://treeherder.mozilla.org/perf.html#/graphs?series=mozilla-central,1888424,1,4&series=mozilla-inbound,1890241,1,4&series=autoland,1885459,1,4&series=mozilla-central,1684866,1,4&series=mozilla-inbound,1684808,1,4&series=autoland,1684872,1,4 On Mon, Jan 21, 2019 at 2:38 AM James Graham <ja...@hoppipolla.co.uk> wrote: > On 21/01/2019 10:18, Jan de Mooij wrote: > > On Fri, Jan 18, 2019 at 10:36 PM Joel Maher <joel.ma...@gmail.com> > wrote: > > > >> Are there any concerns with this latest proposal? > >> > > > > This proposal sounds great to me. Thank you! > > +1. This seems like the right first step to me. > _______________________________________________ > dev-platform mailing list > dev-platform@lists.mozilla.org > https://lists.mozilla.org/listinfo/dev-platform > _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
