Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
Quoting David Gibson (2016-10-31 21:26:39) > On Tue, Oct 25, 2016 at 06:57:33PM -0500, Michael Roth wrote: > > Quoting David Gibson (2016-10-24 20:41:29) > > > On Mon, Oct 17, 2016 at 04:24:31PM -0500, Michael Roth wrote: > > > > Quoting Peter Maydell (2016-10-17 13:45:21) > > > > > On 17 October 2016 at 19:13, Michael Roth > > > > > wrote: > > > > > > We could do both though: use some ad-hoc way to tag for a particular > > > > > > sub-maintainer tree/stable branch, as well as an explicit "not for > > > > > > master" in the cover letter ensure it doesn't go into master. It's > > > > > > a bit > > > > > > more redundant, but flexible in that people can use whatever tagging > > > > > > format they want for a particular tree. > > > > > > > > > > Yes, that would be my preference. Gmail's filtering is not > > > > > very good, and it doesn't seem to be able to support > > > > > multiple or complex matches on the subject line, but > > > > > it can deal with "doesn't include foo in body". > > > > > People who actively want to look for stuff not to go > > > > > into master can filter it however they like. > > > > > > > > Sounds good to me. For my part I think "for-2.7.1" etc. would be > > > > prefereable. No need to resend this patchset though. > > > > > > > > I suppose MAINTAINERS would be the best place to document something > > > > like this? > > > > > > So.. regardless of the outcome in general for future stable merges.. > > > > > > Has this batch been merged for 2.7 stable? Or do I need to resend it > > > in the new style? > > > > No need to resend. I should have the initial staging tree for 2.7 posted > > by Monday and will have this included. > > I haven't spotted the 2.7 stable branch so far. Maybe I don't have > the right remote? Sorry, I was a bit behind getting it posted. I've put up a staging tree here: https://github.com/mdroth/qemu/commits/stable-2.7-staging I'm tentatively planning on posting the initial tree November 7th, setting the freeze for November 11th, and the release for the 16th. I saw your series regarding 2.6<->2.7 CPU migration and had also been hoping to get it sorted for 2.7.1, so let me know if we should consider tweaking the dates. > > -- > David Gibson| I'll have my music baroque, and my code > david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ > | _way_ _around_! > http://www.ozlabs.org/~dgibson
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
On Tue, Oct 25, 2016 at 06:57:33PM -0500, Michael Roth wrote: > Quoting David Gibson (2016-10-24 20:41:29) > > On Mon, Oct 17, 2016 at 04:24:31PM -0500, Michael Roth wrote: > > > Quoting Peter Maydell (2016-10-17 13:45:21) > > > > On 17 October 2016 at 19:13, Michael Roth > > > > wrote: > > > > > We could do both though: use some ad-hoc way to tag for a particular > > > > > sub-maintainer tree/stable branch, as well as an explicit "not for > > > > > master" in the cover letter ensure it doesn't go into master. It's a > > > > > bit > > > > > more redundant, but flexible in that people can use whatever tagging > > > > > format they want for a particular tree. > > > > > > > > Yes, that would be my preference. Gmail's filtering is not > > > > very good, and it doesn't seem to be able to support > > > > multiple or complex matches on the subject line, but > > > > it can deal with "doesn't include foo in body". > > > > People who actively want to look for stuff not to go > > > > into master can filter it however they like. > > > > > > Sounds good to me. For my part I think "for-2.7.1" etc. would be > > > prefereable. No need to resend this patchset though. > > > > > > I suppose MAINTAINERS would be the best place to document something > > > like this? > > > > So.. regardless of the outcome in general for future stable merges.. > > > > Has this batch been merged for 2.7 stable? Or do I need to resend it > > in the new style? > > No need to resend. I should have the initial staging tree for 2.7 posted > by Monday and will have this included. I haven't spotted the 2.7 stable branch so far. Maybe I don't have the right remote? -- David Gibson| I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson signature.asc Description: PGP signature
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
Quoting David Gibson (2016-10-24 20:41:29) > On Mon, Oct 17, 2016 at 04:24:31PM -0500, Michael Roth wrote: > > Quoting Peter Maydell (2016-10-17 13:45:21) > > > On 17 October 2016 at 19:13, Michael Roth > > > wrote: > > > > We could do both though: use some ad-hoc way to tag for a particular > > > > sub-maintainer tree/stable branch, as well as an explicit "not for > > > > master" in the cover letter ensure it doesn't go into master. It's a bit > > > > more redundant, but flexible in that people can use whatever tagging > > > > format they want for a particular tree. > > > > > > Yes, that would be my preference. Gmail's filtering is not > > > very good, and it doesn't seem to be able to support > > > multiple or complex matches on the subject line, but > > > it can deal with "doesn't include foo in body". > > > People who actively want to look for stuff not to go > > > into master can filter it however they like. > > > > Sounds good to me. For my part I think "for-2.7.1" etc. would be > > prefereable. No need to resend this patchset though. > > > > I suppose MAINTAINERS would be the best place to document something > > like this? > > So.. regardless of the outcome in general for future stable merges.. > > Has this batch been merged for 2.7 stable? Or do I need to resend it > in the new style? No need to resend. I should have the initial staging tree for 2.7 posted by Monday and will have this included. > > -- > David Gibson| I'll have my music baroque, and my code > david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ > | _way_ _around_! > http://www.ozlabs.org/~dgibson
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
On Mon, Oct 17, 2016 at 04:24:31PM -0500, Michael Roth wrote: > Quoting Peter Maydell (2016-10-17 13:45:21) > > On 17 October 2016 at 19:13, Michael Roth wrote: > > > We could do both though: use some ad-hoc way to tag for a particular > > > sub-maintainer tree/stable branch, as well as an explicit "not for > > > master" in the cover letter ensure it doesn't go into master. It's a bit > > > more redundant, but flexible in that people can use whatever tagging > > > format they want for a particular tree. > > > > Yes, that would be my preference. Gmail's filtering is not > > very good, and it doesn't seem to be able to support > > multiple or complex matches on the subject line, but > > it can deal with "doesn't include foo in body". > > People who actively want to look for stuff not to go > > into master can filter it however they like. > > Sounds good to me. For my part I think "for-2.7.1" etc. would be > prefereable. No need to resend this patchset though. > > I suppose MAINTAINERS would be the best place to document something > like this? So.. regardless of the outcome in general for future stable merges.. Has this batch been merged for 2.7 stable? Or do I need to resend it in the new style? -- David Gibson| I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson signature.asc Description: PGP signature
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
On 17 October 2016 at 22:24, Michael Roth wrote: > Quoting Peter Maydell (2016-10-17 13:45:21) >> On 17 October 2016 at 19:13, Michael Roth wrote: >> > We could do both though: use some ad-hoc way to tag for a particular >> > sub-maintainer tree/stable branch, as well as an explicit "not for >> > master" in the cover letter ensure it doesn't go into master. It's a bit >> > more redundant, but flexible in that people can use whatever tagging >> > format they want for a particular tree. >> >> Yes, that would be my preference. Gmail's filtering is not >> very good, and it doesn't seem to be able to support >> multiple or complex matches on the subject line, but >> it can deal with "doesn't include foo in body". >> People who actively want to look for stuff not to go >> into master can filter it however they like. > > Sounds good to me. For my part I think "for-2.7.1" etc. would be > prefereable. No need to resend this patchset though. > > I suppose MAINTAINERS would be the best place to document something > like this? We have http://wiki.qemu.org/Contribute/SubmitAPullRequest and I've added a note to it. thanks -- PMM
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
Quoting Peter Maydell (2016-10-17 13:45:21) > On 17 October 2016 at 19:13, Michael Roth wrote: > > We could do both though: use some ad-hoc way to tag for a particular > > sub-maintainer tree/stable branch, as well as an explicit "not for > > master" in the cover letter ensure it doesn't go into master. It's a bit > > more redundant, but flexible in that people can use whatever tagging > > format they want for a particular tree. > > Yes, that would be my preference. Gmail's filtering is not > very good, and it doesn't seem to be able to support > multiple or complex matches on the subject line, but > it can deal with "doesn't include foo in body". > People who actively want to look for stuff not to go > into master can filter it however they like. Sounds good to me. For my part I think "for-2.7.1" etc. would be prefereable. No need to resend this patchset though. I suppose MAINTAINERS would be the best place to document something like this? > > thanks > -- PMM >
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
On 17 October 2016 at 19:13, Michael Roth wrote: > We could do both though: use some ad-hoc way to tag for a particular > sub-maintainer tree/stable branch, as well as an explicit "not for > master" in the cover letter ensure it doesn't go into master. It's a bit > more redundant, but flexible in that people can use whatever tagging > format they want for a particular tree. Yes, that would be my preference. Gmail's filtering is not very good, and it doesn't seem to be able to support multiple or complex matches on the subject line, but it can deal with "doesn't include foo in body". People who actively want to look for stuff not to go into master can filter it however they like. thanks -- PMM
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
Quoting Peter Maydell (2016-10-17 12:33:18) > On 17 October 2016 at 17:51, Michael Roth wrote: > > Maybe just a tag like [PULL for-stable], or [PULL for-2.7]? > > > > The latter seems to mirror how we handle things for patches coming for > > master during freeze. Others who've submitted patches they've > > backported themselves for stable seem to naturally lean toward that > > approach as well. > > > > That said, this might get confusing immediately after a release, where > > there are a lot of patches floating around with such tags, cc:'d for > > stable, that aren't actually meant to be directly pulled into stable. > > So I think I would lean toward "for-stable", or, even better, > > "for-2.7.1", etc. > > > > I don't do automated pulls so it's not a huge deal either way for me, > > but "for-x" in general should hopefully be enough for Peter to filter > > them out for master based on what whether "x" references the next > > major release or not. > > I don't really want to have to update my email filters every > time we do a release, though, and so "for-X.Y" doesn't work because > when we are in the runup to release pull requests targeting > master tend to be marked that way. What about just for-stable, for-stable-2.7, for-ppc-2.8, etc.? Basically just adopt the for-* prefix for these sorts of pulls, but reserve the for-x.y prefix for master, so that anything that doesn't match for-\d\.\d can get filtered out based on that single rule? > > Maybe just having not-for-master pull requests say "not for master" > in the cover letter somewhere ? I tend to treat PULLs cc'd for stable as just having individual patches marked for stable, so it's a bit easier to miss if it's not something obvious like a subject line tag. It also kind of leaves it as an exercise for the reader what branch other than master is actually the intended target for stuff like sub-maintainer pulls (where there might actually be a bit more automation). We could do both though: use some ad-hoc way to tag for a particular sub-maintainer tree/stable branch, as well as an explicit "not for master" in the cover letter ensure it doesn't go into master. It's a bit more redundant, but flexible in that people can use whatever tagging format they want for a particular tree. > > thanks > -- PMM >
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
On 17 October 2016 at 17:51, Michael Roth wrote: > Maybe just a tag like [PULL for-stable], or [PULL for-2.7]? > > The latter seems to mirror how we handle things for patches coming for > master during freeze. Others who've submitted patches they've > backported themselves for stable seem to naturally lean toward that > approach as well. > > That said, this might get confusing immediately after a release, where > there are a lot of patches floating around with such tags, cc:'d for > stable, that aren't actually meant to be directly pulled into stable. > So I think I would lean toward "for-stable", or, even better, > "for-2.7.1", etc. > > I don't do automated pulls so it's not a huge deal either way for me, > but "for-x" in general should hopefully be enough for Peter to filter > them out for master based on what whether "x" references the next > major release or not. I don't really want to have to update my email filters every time we do a release, though, and so "for-X.Y" doesn't work because when we are in the runup to release pull requests targeting master tend to be marked that way. Maybe just having not-for-master pull requests say "not for master" in the cover letter somewhere ? thanks -- PMM
Re: [Qemu-devel] [Qemu-stable] [Qemu-ppc] [PULL 0/4] ppc patches for qemu-2.7 stable branch
Quoting Thomas Huth (2016-10-17 02:44:59) > On 14.10.2016 19:38, Peter Maydell wrote: > > On 14 October 2016 at 09:27, Greg Kurz wrote: > >> On Fri, 14 Oct 2016 09:28:35 +1100 > >> David Gibson wrote: > >> > >>> On Thu, Oct 13, 2016 at 12:57:19PM +0100, Peter Maydell wrote: > On 13 October 2016 at 12:54, Peter Maydell > wrote: > More generally, we need to come up with something for distinguishing > PULL requests not for master, because my current workflow basically > says "anything that says 'for you to fetch changes up to' will get > merged into master... > >>> > >>> Um.. yes.. this was intended for merge to the 2.7 branch, not master. > >>> Any ideas how I should express that? > >>> > >> > >> I'm not aware of any formal process, other than sending a mail to > >> qemu-stable and Cc: Michael Roth. This is often done by simply > >> replying to selected messages in the pull requests for the master > >> branch. > >> > >> Then Michael does all the cherry picking stuff and usually sends a > >> patch round-up two weeks before the stable release, for people to > >> review. > > > > Yes, I think I was partly thrown because in general patches > > don't go into the stable branches via pull requests. > > That said, my current filter/workflow is clearly broken > > so I'm open to any suggestions for easy-for-me-to-filter-for > > ways to flag up that a pull request isn't aimed at master. > > Maybe simply filter out the requests that include qemu-stable in "To:" ? Maybe just a tag like [PULL for-stable], or [PULL for-2.7]? The latter seems to mirror how we handle things for patches coming for master during freeze. Others who've submitted patches they've backported themselves for stable seem to naturally lean toward that approach as well. That said, this might get confusing immediately after a release, where there are a lot of patches floating around with such tags, cc:'d for stable, that aren't actually meant to be directly pulled into stable. So I think I would lean toward "for-stable", or, even better, "for-2.7.1", etc. I don't do automated pulls so it's not a huge deal either way for me, but "for-x" in general should hopefully be enough for Peter to filter them out for master based on what whether "x" references the next major release or not. > > Thomas > >
