From: Jakub Jelinek <jakub@redhat.com>
To: Florian Weimer <fweimer@redhat.com>
Cc: "Andrew Pinski" <pinskia@gmail.com>,
"Segher Boessenkool" <segher@kernel.crashing.org>,
"Marek Polacek" <polacek@redhat.com>,
"Martin Liška" <mliska@suse.cz>,
"Joseph Myers" <joseph@codesourcery.com>,
"Maxim Kuvyrkov" <maxim.kuvyrkov@linaro.org>,
"GCC Patches" <gcc-patches@gcc.gnu.org>,
"Jason Merrill" <jason@redhat.com>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Eric Raymond" <esr@thyrsus.com>
Subject: Re: [Contrib PATCH] Add scripts to convert GCC repo from SVN to Git
Date: Mon, 20 May 2019 14:25:00 -0000 [thread overview]
Message-ID: <20190520142455.GT19695@tucnak> (raw)
In-Reply-To: <871s0tjkks.fsf@oldenburg2.str.redhat.com>
On Mon, May 20, 2019 at 03:56:35PM +0200, Florian Weimer wrote:
> * Andrew Pinski:
>
> > On Sun, May 19, 2019 at 12:54 PM Segher Boessenkool
> > <segher@kernel.crashing.org> wrote:
>
> >> Git can bisect automatically just fine, there is no upside to doing things
> >> manually. In git there are various handy ways of referring to commits; you
> >> can say master@{3 days ago} for example, or zut@{31} to get the 31st
> >> commit back on branch "zut", etc. See "man gitrevisions".
> >
> > Well one thing is if you have prebuilt cc1/cc1plus. So it is not
> > really doing a manual bisect per-say but rather it is doing a manual
> > bisect using prebuilt binaries and knowing which one comes before
> > which one.
>
> If GCC policy is to reject merge commits, a command similar to
> âgit log --pretty=oneline | wc -lâ gives something that is very
> much like a Subversion revision number, in the sense that it matches
> the commit ordering and that the assigned numbers remain stable
> over time.
That is way too slow for our purposes, note we have at least 150000 trunk
commits at least, git log --pretty=oneline | wc -l takes on my box more than
1.4 seconds.
So far the best suggestion I was given for this was:
> so, if we have tags like r163 for 163000'th commit then git describe --all --match 'r[0-9]*' whatever | sed 's/^r\([0-9]\)*-\([0-9]\)*-.*$/\1\2/'
> would give us the r163147 (except it doesn't handle the r163000 commit or < 100 commits after it, to be fixed)
matz> yeah. With the other direction then being as discussed above (limiting the git log revision to just the couple thousands in range). One thing to consider: having a zillion tags locally can make other operations slow (as the mapping from SHA1 to tag is slow), so if really using rXXX tags it should probably be a larger range than just 1000 revisions.
Thus perhaps tag in a post-commit hook every 5000th commit and handle the
rest in some git alias command or script that. We need a quick way to map
between these revisions and hashes bidirectionally.
Unfortunately, the above is just a per-branch number rather than branch
number common to trunk and official release branches.
Jakub
next prev parent reply other threads:[~2019-05-20 14:25 UTC|newest]
Thread overview: 103+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-05-14 16:11 Maxim Kuvyrkov
2019-05-14 21:20 ` Segher Boessenkool
2019-05-15 8:34 ` Maxim Kuvyrkov
2019-05-15 18:47 ` Segher Boessenkool
2019-05-16 9:44 ` Maxim Kuvyrkov
2019-05-15 11:19 ` Richard Biener
2019-05-15 12:08 ` Maxim Kuvyrkov
2019-05-15 18:42 ` Eric Gallager
2019-05-16 0:33 ` Paul Koning
2019-05-16 9:53 ` Maxim Kuvyrkov
2019-05-16 16:22 ` Jeff Law
2019-05-16 16:40 ` Maxim Kuvyrkov
2019-05-16 18:36 ` Ramana Radhakrishnan
2019-05-16 19:07 ` Jeff Law
2019-05-16 22:04 ` Jonathan Wakely
2019-05-17 11:33 ` Martin Liška
2019-05-16 23:54 ` Joseph Myers
2019-05-17 8:19 ` Richard Sandiford
2019-05-17 19:51 ` Segher Boessenkool
2019-05-17 20:59 ` Steve Ellcey
2019-05-17 21:23 ` Jason Merrill
2019-05-20 22:42 ` Joseph Myers
2019-05-21 14:24 ` Richard Earnshaw (lists)
2019-05-21 14:45 ` Jeff Law
2019-05-21 15:02 ` Richard Earnshaw (lists)
2019-05-21 16:44 ` Segher Boessenkool
2019-05-23 22:33 ` Joseph Myers
2019-05-24 8:58 ` Segher Boessenkool
2019-05-24 12:02 ` Florian Weimer
2019-05-29 1:50 ` Joseph Myers
2019-05-29 13:04 ` Segher Boessenkool
2019-05-31 0:16 ` Joseph Myers
2019-06-02 23:13 ` Segher Boessenkool
2019-06-03 22:33 ` Joseph Myers
2019-06-03 22:49 ` Segher Boessenkool
2019-06-05 18:04 ` Jason Merrill
2019-06-06 10:14 ` Richard Earnshaw (lists)
2019-06-06 23:41 ` Joseph Myers
2019-06-06 23:50 ` Ian Lance Taylor
2019-06-07 9:32 ` Richard Earnshaw (lists)
2019-06-06 23:36 ` Joseph Myers
2019-07-22 9:05 ` Maxim Kuvyrkov
2019-05-16 23:06 ` Joseph Myers
2019-05-17 12:22 ` Martin Liška
2019-05-17 12:39 ` Jakub Jelinek
2019-05-19 7:35 ` Martin Liška
2019-05-19 8:11 ` Segher Boessenkool
2019-05-19 19:21 ` Marek Polacek
2019-05-19 19:46 ` Andreas Schwab
2019-05-19 19:54 ` Segher Boessenkool
2019-05-19 20:01 ` Andrew Pinski
2019-05-19 20:06 ` Marek Polacek
2019-05-20 7:29 ` Martin Liška
2019-05-20 13:56 ` Florian Weimer
2019-05-20 14:18 ` Segher Boessenkool
2019-05-20 14:25 ` Jakub Jelinek [this message]
2019-05-20 14:26 ` Andreas Schwab
2019-05-20 14:29 ` Jakub Jelinek
2019-05-20 14:36 ` Andreas Schwab
2019-05-20 15:04 ` Segher Boessenkool
2019-05-17 14:59 ` Maxim Kuvyrkov
2019-05-19 7:09 ` Martin Liška
2019-05-17 14:56 ` Maxim Kuvyrkov
2019-05-17 13:07 ` Jason Merrill
2019-05-17 15:08 ` Maxim Kuvyrkov
2019-05-20 22:48 ` Joseph Myers
2019-05-28 10:44 ` Maxim Kuvyrkov
2019-07-16 10:21 ` Maxim Kuvyrkov
2019-07-16 12:40 ` Jason Merrill
2019-07-16 14:27 ` Maxim Kuvyrkov
2019-07-20 11:24 ` Maxim Kuvyrkov
2019-07-22 9:35 ` Maxim Kuvyrkov
2019-08-01 20:43 ` Jason Merrill
2019-08-02 8:41 ` Maxim Kuvyrkov
2019-08-02 8:57 ` Richard Biener
2019-08-02 10:27 ` Martin Liška
2019-08-02 10:54 ` Maxim Kuvyrkov
2019-08-02 11:01 ` Martin Liška
2019-08-02 11:06 ` Richard Biener
2019-08-02 11:35 ` Martin Liška
2019-08-02 22:31 ` Jason Merrill
2019-08-05 13:20 ` Martin Liška
2019-08-05 15:20 ` Monotonically increasing counter (was Re: [Contrib PATCH] Add scripts to convert GCC repo from SVN to Git) Jason Merrill
2019-08-05 15:34 ` Jakub Jelinek
2019-08-05 15:45 ` Richard Earnshaw (lists)
2019-08-05 18:22 ` Jason Merrill
2019-08-14 18:49 ` Jason Merrill
2019-09-19 19:29 ` Jason Merrill
2019-09-21 18:18 ` Segher Boessenkool
2019-09-21 20:31 ` Nicholas Krause
2019-09-21 21:32 ` Jason Merrill
2019-09-22 0:20 ` Segher Boessenkool
2019-08-02 14:35 ` [Contrib PATCH] Add scripts to convert GCC repo from SVN to Git Segher Boessenkool
2019-08-02 14:55 ` Maxim Kuvyrkov
2019-08-05 16:43 ` Mike Stump
2019-08-05 8:24 ` Maxim Kuvyrkov
2019-08-06 11:16 ` Maxim Kuvyrkov
2019-08-23 8:27 ` Maxim Kuvyrkov
2019-08-23 22:08 ` Joseph Myers
2019-09-13 7:20 ` Maxim Kuvyrkov
2019-08-02 8:35 ` Maxim Kuvyrkov
2019-08-02 14:14 ` Maxim Kuvyrkov
2019-08-02 15:47 ` Segher Boessenkool
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190520142455.GT19695@tucnak \
--to=jakub@redhat.com \
--cc=esr@thyrsus.com \
--cc=fweimer@redhat.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=jason@redhat.com \
--cc=joseph@codesourcery.com \
--cc=maxim.kuvyrkov@linaro.org \
--cc=mliska@suse.cz \
--cc=pbonzini@redhat.com \
--cc=pinskia@gmail.com \
--cc=polacek@redhat.com \
--cc=segher@kernel.crashing.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).