> -----Original Message----- > From: Stefan Hajnoczi <stefa...@gmail.com> > Sent: Tuesday, November 15, 2022 10:52 AM > To: Taylor Simpson <tsimp...@quicinc.com> > Cc: qemu-devel@nongnu.org; richard.hender...@linaro.org; > phi...@linaro.org; peter.mayd...@linaro.org; Brian Cain > <bc...@quicinc.com>; Matheus Bernardino (QUIC) > <quic_mathb...@quicinc.com>; stefa...@redhat.com > Subject: Re: [PULL 00/11] Hexagon bug fixes and performance improvement > > On Tue, 15 Nov 2022 at 11:16, Taylor Simpson <tsimp...@quicinc.com> > wrote: > > > > OK. I wasn't sure if performance improvements would be considered new > features or not. > > No problem! If there is a performance regression in the upcoming release, > then fixes will be accepted. For example, if QEMU 7.1 was fast but the > upcoming QEMU 7.2 release is going to be slow then a performance fix will > be accepted to avoid a regression in 7.2. > > On the other hand, if it's a fix for something that was already slow in the > last > release (7.1), then it's less likely to be accepted during freeze.
The performance improvements fall into this bucket. > > These are general guidelines and maintainers have a say in what gets > merged. In this case I looked at the pull request and I wasn't sure if you had > decided based on these guidelines or not. It helps when it's clear from the > commit message (or from the commit description in more involved cases) > that the commit fixes a bug or has some other justification. I'm the maintainer for the directories touched by these patches (target/hexagon and tests/tcg/hexagon), but I'll defer you as a more senior maintainer to decide not to merge if it is too risky at this stage. Taylor
