* Re: [RFC] get_sha1() shorthands for blob/tree objects
From: Andreas Ericsson @ 2006-04-19 8:15 UTC (permalink / raw)
To: Linus Torvalds; +Cc: Junio C Hamano, git
In-Reply-To: <Pine.LNX.4.64.0604181745410.3701@g5.osdl.org>
Linus Torvalds wrote:
>
> On Tue, 18 Apr 2006, Linus Torvalds wrote:
>
>>And I thought we already disallowed ':' in branch names because cogito
>>uses them for the strange <rev>:<rev> syntax..
>
>
> Btw, pathnames with ':' in them aren't a problem. It's only revision
> names that can't have ':' in them and still be used together with this
> syntax.
>
> If you have a pathname with ':', it's fine to do
>
> git cat-file v1.2.4:pathname:with:colon
>
> because the magic revision parsing only cares about the _first_ colon, and
> will split that into "v1.2.4" and "pathname:with:colon" without ever even
> looking at the other ones.
>
Except that you'll have to explicitly state HEAD:pathname:with:colon, or
does it try finding a file with the argument verbatim first?
--
Andreas Ericsson andreas.ericsson@op5.se
OP5 AB www.op5.se
Tel: +46 8-230225 Fax: +46 8-230231
^ permalink raw reply
* Re: Clone with local alternates?
From: David Woodhouse @ 2006-04-19 8:39 UTC (permalink / raw)
To: Shawn Pearce; +Cc: git
In-Reply-To: <20060418235658.GB8915@spearce.org>
On Tue, 2006-04-18 at 19:56 -0400, Shawn Pearce wrote:
> git clone --reference=/foo git://remote/foo
That's nice. What would be even _nicer_ is if I could enforce its usage
on the server. I keep a clean copy of Linus' kernel tree on the machine
with the git dæmon, so I'd like to make it refuse to send any objects
which are in that tree and which people should already have fetched from
elsewhere anyway.
--
dwmw2
^ permalink raw reply
* Re: Clone with local alternates?
From: Eric W. Biederman @ 2006-04-19 9:36 UTC (permalink / raw)
To: David Woodhouse; +Cc: git
In-Reply-To: <1145404132.16166.97.camel@shinybook.infradead.org>
David Woodhouse <dwmw2@infradead.org> writes:
> Often I want to clone a remote repository but would like to use an
> existing local source tree as 'alternates'.
>
> One way of doing this is to clone the local tree with 'git-clone -l -s',
> find the latest common commit shared with the remote tree to be fetched,
> revert to that with 'git-reset --head $last' and then pulling from the
> remote.
Simpler than the git-reset you can simply do
git-fetch remote branch:branch and git will find common ancestor
for you and create a new branch that mirrors the old one.
I mention this because it can be interesting to have branches
from several remote repositories all in one local repository.
Eric
^ permalink raw reply
* Re: Clone with local alternates?
From: Eric W. Biederman @ 2006-04-19 9:42 UTC (permalink / raw)
To: David Woodhouse; +Cc: Shawn Pearce, git
In-Reply-To: <1145435957.13200.39.camel@pmac.infradead.org>
David Woodhouse <dwmw2@infradead.org> writes:
> On Tue, 2006-04-18 at 19:56 -0400, Shawn Pearce wrote:
>> git clone --reference=/foo git://remote/foo
>
> That's nice. What would be even _nicer_ is if I could enforce its usage
> on the server. I keep a clean copy of Linus' kernel tree on the machine
> with the git dæmon, so I'd like to make it refuse to send any objects
> which are in that tree and which people should already have fetched from
> elsewhere anyway.
Given how large the linux kernel archive is I can see the desire.
At git startup there is an exchange of common commits. to
find the best place to build the pack from so what you want
is possible. I haven't a clue how you would describe it
though in the face of multiple branches, coming off the
trunk at different times.
Eric
^ permalink raw reply
* Re: cg-clone produces "___" file and no working tree
From: Petr Baudis @ 2006-04-19 9:49 UTC (permalink / raw)
To: Zack Brown; +Cc: git
In-Reply-To: <20060419053640.GA16334@tumblerings.org>
Dear diary, on Wed, Apr 19, 2006 at 07:36:40AM CEST, I got a letter
where Zack Brown <zbrown@tumblerings.org> said that...
> When I do something like
> cg-clone rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/git.git
>
> The first few lines of output are:
>
> defaulting to local storage area
> warning: templates not found /home/zbrown/share/git-core/templates/
> /home/zbrown/git/cogito/cg-clone: line 137: .git/info/cg-fetch-earlydie: No such file or directory
> /home/zbrown/git/cogito/cg-clone: line 148: .git/info/cg-fetch-initial: No such file or directory
>
> The rest of the process seems to go without incident. However, when I look
> at the repository I see:
>
> $ ls -A
> .git ___
> $
Could you please list the contents of the .git subdirectory? It seems
that git-init-db did not create the .git/info subdirectory.
I suspect that something went wrong with your installation and in fact
you are using much older git version. Check git --version and if `which
git-init-db` corresponds to `which git`.
--
Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
Right now I am having amnesia and deja-vu at the same time. I think
I have forgotten this before.
^ permalink raw reply
* Re: Next problem: empty ident <joern@limerick.(none)> not allowed
From: Petr Baudis @ 2006-04-19 9:51 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Jörn Engel, git
In-Reply-To: <7vr73uei1w.fsf@assigned-by-dhcp.cox.net>
Dear diary, on Tue, Apr 18, 2006 at 10:37:47PM CEST, I got a letter
where Junio C Hamano <junkio@cox.net> said that...
> Jörn Engel <joern@wohnheim.fh-wedel.de> writes:
>
> > And now I have some questions:
> > 1. Why didn't the environment variables work?
> > 2. Why is there a check for commit information when I pull from some
> > tree?
>
> Because "pull" means "fetch and merge the local modifications if
> any". When you merge (and you _did_ merge), you create a new
> commit of your own, and the commit records who committed it.
>
> You need GIT_COMMITTER_EMAIL.
You mean GIT_COMMITTER_NAME. :-)
--
Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
Right now I am having amnesia and deja-vu at the same time. I think
I have forgotten this before.
^ permalink raw reply
* GIT Error issue
From: Shyamal Sadanshio @ 2006-04-19 11:16 UTC (permalink / raw)
To: git
Hi,
I am newbie to GIT and facing some problem with the GIT usage. I am
trying to clone a repository at www.linux-mips.org with the
following command, I get error message as below:
**************************************************************************
[root@sshyamal git-tutorial]# git clone
git://ftp.linux-mips.org/pub/scm/linux-malta.git linux-malta.git
git: warning: invalid extra options ignored
GNU Interactive Tools 4.3.20 (i686-pc-linux-gnu), 12:39:46 Apr 18 2006
GIT is free software; you can redistribute it and/or modify it under the
terms of the GNU General Public License as published by the Free
Software
Foundation; either version 2, or (at your option) any later version.
Copyright (C) 1993-1999 Free Software Foundation, Inc.
Written by Tudor Hulubei and Andrei Pitis, Bucharest, Romania
git: fatal error: `chdir' failed: permission denied.
***************************************************************************
I am exercising this command in root mode.
I have browsed through FAQs/archived mails but haven't come across
appropriate solution to this.
Can anyone please let me know the problem cause for this issue?
Thanks and Regards,
Shyamal
^ permalink raw reply
* Re: GIT Error issue
From: Sam Ravnborg @ 2006-04-19 11:20 UTC (permalink / raw)
To: Shyamal Sadanshio; +Cc: git
In-Reply-To: <3857255c0604190416j62abeae8va164896c5100f6ee@mail.gmail.com>
On Wed, Apr 19, 2006 at 04:46:57PM +0530, Shyamal Sadanshio wrote:
> Hi,
>
> I am newbie to GIT and facing some problem with the GIT usage. I am
> trying to clone a repository at www.linux-mips.org with the
> following command, I get error message as below:
>
> **************************************************************************
> [root@sshyamal git-tutorial]# git clone
> git://ftp.linux-mips.org/pub/scm/linux-malta.git linux-malta.git
> git: warning: invalid extra options ignored
>
> GNU Interactive Tools 4.3.20 (i686-pc-linux-gnu), 12:39:46 Apr 18 2006
You are running the wrong git.
The git you are running are the "GNU Interactive Tools" and not git from
the git-suite.
Fix - uninstall "GNU Interactive Tools"
Sam
^ permalink raw reply
* Re: GIT Error issue
From: Krzysiek Pawlik @ 2006-04-19 11:20 UTC (permalink / raw)
To: Shyamal Sadanshio; +Cc: git
In-Reply-To: <3857255c0604190416j62abeae8va164896c5100f6ee@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 525 bytes --]
Shyamal Sadanshio wrote:
> **************************************************************************
> [root@sshyamal git-tutorial]# git clone
> git://ftp.linux-mips.org/pub/scm/linux-malta.git linux-malta.git
> git: warning: invalid extra options ignored
>
> GNU Interactive Tools 4.3.20 (i686-pc-linux-gnu), 12:39:46 Apr 18 2006
^ ^ ^
Use git - the content manager, not interactive tools. Use full path to
git executable.
--
Krzysiek Pawlik (Nelchael)
RLU #322999 GPG Key ID: 0xBC555551
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 191 bytes --]
^ permalink raw reply
* Re: GIT Error issue
From: Erik Mouw @ 2006-04-19 11:21 UTC (permalink / raw)
To: Shyamal Sadanshio; +Cc: git
In-Reply-To: <3857255c0604190416j62abeae8va164896c5100f6ee@mail.gmail.com>
On Wed, Apr 19, 2006 at 04:46:57PM +0530, Shyamal Sadanshio wrote:
> I am newbie to GIT and facing some problem with the GIT usage. I am
> trying to clone a repository at www.linux-mips.org with the
> following command, I get error message as below:
>
> **************************************************************************
> [root@sshyamal git-tutorial]# git clone
> git://ftp.linux-mips.org/pub/scm/linux-malta.git linux-malta.git
> git: warning: invalid extra options ignored
>
> GNU Interactive Tools 4.3.20 (i686-pc-linux-gnu), 12:39:46 Apr 18 2006
> GIT is free software; you can redistribute it and/or modify it under the
> terms of the GNU General Public License as published by the Free
> Software
> Foundation; either version 2, or (at your option) any later version.
> Copyright (C) 1993-1999 Free Software Foundation, Inc.
> Written by Tudor Hulubei and Andrei Pitis, Bucharest, Romania
>
> git: fatal error: `chdir' failed: permission denied.
You're using the wrong GIT. Remove the GNU Interactive Tools and use
git from kernel.org.
> ***************************************************************************
>
> I am exercising this command in root mode.
There is absolutely zero reason to run git as root.
Erik
--
+-- Erik Mouw -- www.harddisk-recovery.com -- +31 70 370 12 90 --
| Lab address: Delftechpark 26, 2628 XH, Delft, The Netherlands
^ permalink raw reply
* Default refspec for branches
From: Josh Boyer @ 2006-04-19 12:58 UTC (permalink / raw)
To: git
Is there a way to change the default refspec that git pull uses on a
per branch basis? I know that you can create .git/remotes/<foo> and
do 'git pull <foo>' to pull from whatever is listed in there, but that
isn't quite what I'm looking for.
What I'd like to be able to do is create a branch and have 'git pull'
simply pull from the remote tree. I tried listing multiple refspecs
in .git/remotes/origin, but git didn't seem to like that.
For example, I clone Linus' tree. Then I create a branch called mtd.
When I'm working in that branch, I want 'git pull' to pull from
git://git.infradead.org/mtd-2.6.git.
Any ideas?
josh
^ permalink raw reply
* git-daemon memory usage, disconnection.
From: David Woodhouse @ 2006-04-19 13:22 UTC (permalink / raw)
To: git
I'm running git-daemon from xinetd and it seems a little greedy...
Cpu(s): 2.7% us, 6.4% sy, 0.0% ni, 1.7% id, 87.7% wa, 1.4% hi, 0.0% si
Mem: 253680k total, 250076k used, 3604k free, 568k buffers
Swap: 500960k total, 500864k used, 96k free, 24696k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
31232 nobody 18 0 155m 29m 7224 D 1.3 11.9 0:25.56 git-rev-list
30743 nobody 18 0 179m 29m 9480 D 0.7 11.9 0:42.60 git-rev-list
31277 nobody 18 0 147m 28m 7476 D 2.6 11.4 0:20.90 git-rev-list
30314 nobody 18 0 233m 26m 7696 D 0.0 10.6 1:20.24 git-rev-list
30612 nobody 18 0 204m 23m 7432 D 1.3 9.4 0:59.19 git-rev-list
30574 nobody 18 0 190m 20m 7608 D 0.3 8.3 0:50.77 git-rev-list
30208 nobody 18 0 140m 14m 7632 D 0.3 5.9 0:15.23 git-pack-object
Now, this wouldn't be _so_ bad if there were only two of them running.
The clients for the other four have actually given up and disconnected
long ago, but git-daemon doesn't seem to have reacted to that.
--
dwmw2
^ permalink raw reply
* Re: cg-clone produces "___" file and no working tree
From: Zack Brown @ 2006-04-19 14:16 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git
In-Reply-To: <7vd5feawel.fsf@assigned-by-dhcp.cox.net>
On Tue, Apr 18, 2006 at 11:53:38PM -0700, Junio C Hamano wrote:
> Zack Brown <zbrown@tumblerings.org> writes:
>
> > What is going on? I'm completely unable to clone a repository.
>
> I have no idea how cg-* is broken, so I'll let Pasky answer
> that, but I suspect your git installation is broken.
>
> > If I try
> > "git clone rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/git.git",
> > I get this error: "git: 'clone' is not a git-command", and it prints a usage
> > page, but "clone" is listed on that usage page.
>
> That sounds intersting. Although rsync is deprecated for a long
> time and git:// is the preferred transport, I do not get "is not
> a git-command" error. Are you installing things correctly?
I think so. I was able to reproduce this behavior using the tarballs that are
used to start out a new user with git and cogito; as well as with the latest
version in the repository.
But I also thought the rsync deprecation was not really true. Wasn't that
discussed here recently? I thought the conclusion was that folks *wanted*
to deprecate it, but at the moment it was still the best way to accomplish
certain things.
>
> For example, as the first paragraph of INSTALL says, if you
> override prefix= from the make command line, you need to do so
> consistently when you build and when you install.
>
> What do these command say?
>
> $ git --exec-path
> $ ls -l "`git --exec-path`/git-clone"
22:07:05 [zbrown] ~$ git --exec-path
/home/zbrown/bin
07:10:34 [zbrown] ~$ ls -l "`git --exec-path`/git-clone"
ls: /home/zbrown/bin/git-clone: No such file or directory
Does that mean it's looking in /home/zbrown/bin for the git binaries? That's
weird. I have /home/zbrown/git/git and /home/zbrown/git/cogito in my $PATH
specifically to hold those executables. There's no git stuff in
/home/zbrown/bin.
Be well,
Zack
>
--
Zack Brown
^ permalink raw reply
* Re: cg-clone produces "___" file and no working tree
From: Zack Brown @ 2006-04-19 14:21 UTC (permalink / raw)
To: Petr Baudis; +Cc: git
In-Reply-To: <20060419094916.GD27689@pasky.or.cz>
Hi Petr,
On Wed, Apr 19, 2006 at 11:49:16AM +0200, Petr Baudis wrote:
> Dear diary, on Wed, Apr 19, 2006 at 07:36:40AM CEST, I got a letter
> where Zack Brown <zbrown@tumblerings.org> said that...
> > When I do something like
> > cg-clone rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/git.git
> >
> > The first few lines of output are:
> >
> > defaulting to local storage area
> > warning: templates not found /home/zbrown/share/git-core/templates/
> > /home/zbrown/git/cogito/cg-clone: line 137: .git/info/cg-fetch-earlydie: No such file or directory
> > /home/zbrown/git/cogito/cg-clone: line 148: .git/info/cg-fetch-initial: No such file or directory
> >
> > The rest of the process seems to go without incident. However, when I look
> > at the repository I see:
> >
> > $ ls -A
> > .git ___
> > $
>
> Could you please list the contents of the .git subdirectory? It seems
> that git-init-db did not create the .git/info subdirectory.
07:19:57 [zbrown] ~/git/trees/tmp/git/.git$ ls -F
total 28
4 HEAD 4 branches/ 4 config 4 index 4 info/ 4 objects/ 4 refs/
>
> I suspect that something went wrong with your installation and in fact
> you are using much older git version. Check git --version and if `which
> git-init-db` corresponds to `which git`.
07:11:13 [zbrown] ~$ git --version
git version 1.3.0.g2473-dirty
07:18:38 [zbrown] ~$ which git-init-db
/home/zbrown/git/git//git-init-db
07:18:52 [zbrown] ~$ which git
/home/zbrown/git/git//git
07:18:54 [zbrown] ~$
I assume the "-dirty" in the version number means it thinks I've made some
changes, but I really haven't. Maybe the repo got corrupted somehow.
Be well,
Zack
>
> --
> Petr "Pasky" Baudis
> Stuff: http://pasky.or.cz/
> Right now I am having amnesia and deja-vu at the same time. I think
> I have forgotten this before.
--
Zack Brown
^ permalink raw reply
* Re: [RFC] get_sha1() shorthands for blob/tree objects
From: Linus Torvalds @ 2006-04-19 14:44 UTC (permalink / raw)
To: Andreas Ericsson; +Cc: Junio C Hamano, git
In-Reply-To: <4445F1B0.4060105@op5.se>
On Wed, 19 Apr 2006, Andreas Ericsson wrote:
>
> Except that you'll have to explicitly state HEAD:pathname:with:colon, or does
> it try finding a file with the argument verbatim first?
If you have a pathname:with:colon and you _just_ want to use it as a
pathname, you'd do one of either:
- use "--" to separate the pathnames from the revision info. That always
works.
- even without the "--", if the first part of the pathname:with:colon (ie
the "pathname" part) cannot be looked up as a SHA1 tag, then the
pathname:with:colon is left alone and seen as a normal file.
So "--" is always the dis-ambiguator. But it almost certainly will never
be needed.
Linus
^ permalink raw reply
* Re: cg-clone produces "___" file and no working tree
From: Petr Baudis @ 2006-04-19 14:48 UTC (permalink / raw)
To: Zack Brown; +Cc: git
In-Reply-To: <20060419142131.GD4104@tumblerings.org>
Hi,
Dear diary, on Wed, Apr 19, 2006 at 04:21:31PM CEST, I got a letter
where Zack Brown <zbrown@tumblerings.org> said that...
> On Wed, Apr 19, 2006 at 11:49:16AM +0200, Petr Baudis wrote:
> > Dear diary, on Wed, Apr 19, 2006 at 07:36:40AM CEST, I got a letter
> > where Zack Brown <zbrown@tumblerings.org> said that...
> > > When I do something like
> > > cg-clone rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/git.git
> > >
> > > The first few lines of output are:
> > >
> > > defaulting to local storage area
> > > warning: templates not found /home/zbrown/share/git-core/templates/
> > > /home/zbrown/git/cogito/cg-clone: line 137: .git/info/cg-fetch-earlydie: No such file or directory
> > > /home/zbrown/git/cogito/cg-clone: line 148: .git/info/cg-fetch-initial: No such file or directory
> > >
> > > The rest of the process seems to go without incident. However, when I look
> > > at the repository I see:
> > >
> > > $ ls -A
> > > .git ___
> > > $
> >
> > Could you please list the contents of the .git subdirectory? It seems
> > that git-init-db did not create the .git/info subdirectory.
>
> 07:19:57 [zbrown] ~/git/trees/tmp/git/.git$ ls -F
> total 28
> 4 HEAD 4 branches/ 4 config 4 index 4 info/ 4 objects/ 4 refs/
hmm, could you please do this just after running git-init-db in an
empty directory? I just realized cg-fetch will mkdir -p the .git/info/
directory.
If the .git/info/ directory would be there after git-init-db, I
couldn't explain the
/home/zbrown/git/cogito/cg-clone: line 137: .git/info/cg-fetch-earlydie: No such file or directory
error. If the .git/info/ directory is not there after git-init-db,
either it is somehow broken in git-1.3.0, or it belongs to a much older
git version.
> 07:18:38 [zbrown] ~$ which git-init-db
> /home/zbrown/git/git//git-init-db
> 07:18:52 [zbrown] ~$ which git
> /home/zbrown/git/git//git
It might be a good idea to compare the ctimes.
--
Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
Right now I am having amnesia and deja-vu at the same time. I think
I have forgotten this before.
^ permalink raw reply
* Re: git-daemon memory usage, disconnection.
From: Linus Torvalds @ 2006-04-19 14:59 UTC (permalink / raw)
To: David Woodhouse; +Cc: git
In-Reply-To: <1145452967.13200.92.camel@pmac.infradead.org>
On Wed, 19 Apr 2006, David Woodhouse wrote:
>
> I'm running git-daemon from xinetd and it seems a little greedy...
>
> Cpu(s): 2.7% us, 6.4% sy, 0.0% ni, 1.7% id, 87.7% wa, 1.4% hi, 0.0% si
> Mem: 253680k total, 250076k used, 3604k free, 568k buffers
> Swap: 500960k total, 500864k used, 96k free, 24696k cached
>
> PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
> 31232 nobody 18 0 155m 29m 7224 D 1.3 11.9 0:25.56 git-rev-list
> 30743 nobody 18 0 179m 29m 9480 D 0.7 11.9 0:42.60 git-rev-list
> 31277 nobody 18 0 147m 28m 7476 D 2.6 11.4 0:20.90 git-rev-list
> 30314 nobody 18 0 233m 26m 7696 D 0.0 10.6 1:20.24 git-rev-list
> 30612 nobody 18 0 204m 23m 7432 D 1.3 9.4 0:59.19 git-rev-list
> 30574 nobody 18 0 190m 20m 7608 D 0.3 8.3 0:50.77 git-rev-list
> 30208 nobody 18 0 140m 14m 7632 D 0.3 5.9 0:15.23 git-pack-object
Well, you've probably got two issues:
- it looks like you aren't packing your archives (which explains why the
disk accesses are horrid, which in turn explains the "D" part).
For a git server, you _really_ want all trees to be mostly packed, or
you want absolutely tons of memory (and 256kB is definitely not "tons"
as far as git is concerned).
- git-rev-list won't notice that there is nobody listening until it gets
a EPIPE, and it won't get an EPIPE until it actually outputs something,
and it won't output anything until it is largely done traversing the
tree..
> Now, this wouldn't be _so_ bad if there were only two of them running.
> The clients for the other four have actually given up and disconnected
> long ago, but git-daemon doesn't seem to have reacted to that.
Well, the way things work under UNIX, you normally don't notice that the
other end isn't interested until you try to write, and you get a "nobody
is listening". And sadly, the packing stuff does most (not all) of the
heavy lifting before it can even start to write things out.
That said, I should probably take a look at git-rev-list --objects memory
usage once again. It's neve rbeen exactly "lean" (and it can't really be:
it does end up needing the total object list in memory for a full clone,
and with something like the kernel, that's about 250 _thousand_ objects).
We should probably also make send-pack.c use the nice revision library,
because right now it's doing that pipe to git-rev-list for no good reason.
Linus
^ permalink raw reply
* Re: git-daemon memory usage, disconnection.
From: David Woodhouse @ 2006-04-19 15:27 UTC (permalink / raw)
To: Linus Torvalds; +Cc: git
In-Reply-To: <Pine.LNX.4.64.0604190749160.3701@g5.osdl.org>
On Wed, 2006-04-19 at 07:59 -0700, Linus Torvalds wrote:
> Well, you've probably got two issues:
>
> - it looks like you aren't packing your archives (which explains why the
> disk accesses are horrid, which in turn explains the "D" part).
Hm, good point. They're fairly new trees -- I had foolishly assumed that
they would at least start off packed. That isn't the case though --
perhaps it should be? Did the original clone receive a pack on the wire
and then _split_ it?
If the tools would automatically pack when the number of unpacked
objects reaches a threshold, that would be useful.
Since this repo is only available through git:// and git+ssh:// URLs, I
can safely use git-repack's '-a -d' options, right?
I'll do 'git-repack -l' nightly and 'git-repack -a -d -l' weekly -- does
that seem sane?
> For a git server, you _really_ want all trees to be mostly packed, or
> you want absolutely tons of memory (and 256kB is definitely not "tons"
> as far as git is concerned).
>
> Well, the way things work under UNIX, you normally don't notice that the
> other end isn't interested until you try to write, and you get a "nobody
> is listening". And sadly, the packing stuff does most (not all) of the
> heavy lifting before it can even start to write things out.
Well, it does that with SIGALRM happening periodically, theoretically
for the purpose of providing progress output. Perhaps we could do a
getpeername() or something else to check on the output fd each time?
--
dwmw2
^ permalink raw reply
* Re: git-daemon memory usage, disconnection.
From: Linus Torvalds @ 2006-04-19 15:49 UTC (permalink / raw)
To: David Woodhouse; +Cc: git
In-Reply-To: <1145460469.11909.25.camel@pmac.infradead.org>
On Wed, 19 Apr 2006, David Woodhouse wrote:
> On Wed, 2006-04-19 at 07:59 -0700, Linus Torvalds wrote:
> > Well, you've probably got two issues:
> >
> > - it looks like you aren't packing your archives (which explains why the
> > disk accesses are horrid, which in turn explains the "D" part).
>
> Hm, good point. They're fairly new trees -- I had foolishly assumed that
> they would at least start off packed. That isn't the case though --
> perhaps it should be? Did the original clone receive a pack on the wire
> and then _split_ it?
For old versions of git, yes.
> If the tools would automatically pack when the number of unpacked
> objects reaches a threshold, that would be useful.
Well, packing is still best done in the background: you don't generally
want the tools to just stop for a minute to repack while you're doing
something. You'd normally want to do a cron run at 4AM or something, see
if there is lots to pack, and repack that.
The one exception is probably a large conversion process (from CVS, SVN,
whatever). The conversion process itself probably takes ages, and it will
be even slower if it were to keep the potentially huge result unpacked all
the time.
But for normal ops, you really don't want to repack synchronously.
> Since this repo is only available through git:// and git+ssh:// URLs, I
> can safely use git-repack's '-a -d' options, right?
Yes.
> I'll do 'git-repack -l' nightly and 'git-repack -a -d -l' weekly -- does
> that seem sane?
Absolutely. The one exception might be trees that really don't change very
much (which is quite common), so you might make it conditional on seeing
if there are _any_ objects at all in .git/objects/00/, for example. Not
that repack will be very expensive, but still..
> Well, it does that with SIGALRM happening periodically, theoretically
> for the purpose of providing progress output. Perhaps we could do a
> getpeername() or something else to check on the output fd each time?
Yes, that's possibly a good idea. Of course, for git-rev-list, it's just a
pipe, and it's hard to do that check at least portably. On Linux, doing a
"poll()" on a pipe for writing, with newer kernels you'll get a POLLERR if
the other side has hung up, but that's by no means portable.
(On some other systems, doing a zero-sized write() _might_ do it, but at
least Linux will happily say "ok, wrote 0 bytes" even if the other end
isn't listening).
And git-rev-list isn't doing the SIGALARM anyway.
In other words, to do this, we'd have to change send-pack to use the
revision library. Which, as mentioned, is worth-while anyway, but it's not
totally trivial.
Linus
^ permalink raw reply
* Allow "git repack" users to specify repacking window/depth
From: Linus Torvalds @ 2006-04-19 17:05 UTC (permalink / raw)
To: Junio C Hamano, Git Mailing List
.. but don't even bother documenting it. I don't think any normal person
is supposed to ever really care, but it simplifies testing when you want
to use the "git repack" wrapper rather than forcing you to use the core
programs (which already do support the window/depth arguments, of course).
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
diff --git a/git-repack.sh b/git-repack.sh
index a5d349f..7a72c2c 100755
--- a/git-repack.sh
+++ b/git-repack.sh
@@ -7,7 +7,7 @@ USAGE='[-a] [-d] [-f] [-l] [-n] [-q]'
. git-sh-setup
no_update_info= all_into_one= remove_redundant=
-local= quiet= no_reuse_delta=
+local= quiet= no_reuse_delta= extra=
while case "$#" in 0) break ;; esac
do
case "$1" in
@@ -17,6 +17,8 @@ do
-q) quiet=-q ;;
-f) no_reuse_delta=--no-reuse-delta ;;
-l) local=--local ;;
+ --window=*) extra="$extra $1" ;;
+ --depth=*) extra="$extra $1" ;;
*) usage ;;
esac
shift
@@ -40,7 +42,7 @@ case ",$all_into_one," in
find . -type f \( -name '*.pack' -o -name '*.idx' \) -print`
;;
esac
-pack_objects="$pack_objects $local $quiet $no_reuse_delta"
+pack_objects="$pack_objects $local $quiet $no_reuse_delta $extra"
name=$(git-rev-list --objects --all $rev_list 2>&1 |
git-pack-objects --non-empty $pack_objects .tmp-pack) ||
exit 1
^ permalink raw reply related
* git log: don't do merge diffs by default
From: Linus Torvalds @ 2006-04-19 17:20 UTC (permalink / raw)
To: Junio C Hamano, Git Mailing List
I personally prefer "ignore_merges" to be on by default, because quite
often the merge diff is distracting and not interesting. That's true both
with "-p" and with "--stat" output.
If you want output from merges, you can trivially use the "-m", "-c" or
"--cc" flags to tell that you're interested in merges, which also tells
the diff generator what kind of diff to do (for --stat, any of the three
will do, of course, but they differ for plain patches or for
--patch-with-stat).
This trivial patch just removes the two lines that tells "git log" not to
ignore merges. It will still show the commit log message, of course, due
to the "always_show_header" part.
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
diff --git a/git.c b/git.c
index 0be14bb..40b7e42 100644
--- a/git.c
+++ b/git.c
@@ -331,8 +331,6 @@ static int cmd_log(int argc, const char
init_revisions(&rev);
rev.always_show_header = 1;
rev.diffopt.recursive = 1;
- rev.combine_merges = 1;
- rev.ignore_merges = 0;
return cmd_log_wc(argc, argv, envp, &rev);
}
^ permalink raw reply related
* Re: Default refspec for branches
From: Junio C Hamano @ 2006-04-19 17:45 UTC (permalink / raw)
To: Josh Boyer; +Cc: git
In-Reply-To: <625fc13d0604190558tf0e8b69n5f5a830a3131f1d4@mail.gmail.com>
"Josh Boyer" <jwboyer@gmail.com> writes:
> Is there a way to change the default refspec that git pull uses on a
> per branch basis?
No.
But it is a _very_ reasonable thing people would want to be able
to. Please look at the last entry in:
Subject: Recent unresolved issues
Date: Fri, 14 Apr 2006 02:31:36 -0700
Message-ID: <7v64lcqz9j.fsf@assigned-by-dhcp.cox.net>
^ permalink raw reply
* Re: Default refspec for branches
From: Josh Boyer @ 2006-04-19 17:52 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git
In-Reply-To: <7v3bg9bgsz.fsf@assigned-by-dhcp.cox.net>
On 4/19/06, Junio C Hamano <junkio@cox.net> wrote:
> "Josh Boyer" <jwboyer@gmail.com> writes:
>
> > Is there a way to change the default refspec that git pull uses on a
> > per branch basis?
>
> No.
>
> But it is a _very_ reasonable thing people would want to be able
> to. Please look at the last entry in:
>
> Subject: Recent unresolved issues
> Date: Fri, 14 Apr 2006 02:31:36 -0700
> Message-ID: <7v64lcqz9j.fsf@assigned-by-dhcp.cox.net>
Hm, ok.
I'd offer to work on that, but I have a bit of a learning curve in
front of me before I could make something useful. If others beat me
to it, I certainly wouldn't mind ;)
josh
^ permalink raw reply
* [PATCH] sha1_name.c: prepare to make get_tree_entry() reusable from others.
From: Junio C Hamano @ 2006-04-19 20:01 UTC (permalink / raw)
To: git
Make the mode information extracted from the tree available.
Signed-off-by: Junio C Hamano <junkio@cox.net>
---
* This comes on top of the extended SHA1 expression sha1:path
patch Linus did.
sha1_name.c | 19 ++++++++++---------
1 files changed, 10 insertions(+), 9 deletions(-)
diff --git a/sha1_name.c b/sha1_name.c
index 0cd1139..35e8dfb 100644
--- a/sha1_name.c
+++ b/sha1_name.c
@@ -450,18 +450,17 @@ static int get_sha1_1(const char *name,
return get_short_sha1(name, len, sha1, 0);
}
-static int get_tree_entry(const unsigned char *, const char *, unsigned char *);
+static int get_tree_entry(const unsigned char *, const char *, unsigned char *, unsigned *);
-static int find_tree_entry(struct tree_desc *t, const char *name, unsigned char *result)
+static int find_tree_entry(struct tree_desc *t, const char *name, unsigned char *result, unsigned *mode)
{
int namelen = strlen(name);
while (t->size) {
const char *entry;
const unsigned char *sha1;
int entrylen, cmp;
- unsigned mode;
- sha1 = tree_entry_extract(t, &entry, &mode);
+ sha1 = tree_entry_extract(t, &entry, mode);
update_tree_entry(t);
entrylen = strlen(entry);
if (entrylen > namelen)
@@ -477,18 +476,18 @@ static int find_tree_entry(struct tree_d
}
if (name[entrylen] != '/')
continue;
- if (!S_ISDIR(mode))
+ if (!S_ISDIR(*mode))
break;
if (++entrylen == namelen) {
memcpy(result, sha1, 20);
return 0;
}
- return get_tree_entry(sha1, name + entrylen, result);
+ return get_tree_entry(sha1, name + entrylen, result, mode);
}
return -1;
}
-static int get_tree_entry(const unsigned char *tree_sha1, const char *name, unsigned char *sha1)
+static int get_tree_entry(const unsigned char *tree_sha1, const char *name, unsigned char *sha1, unsigned *mode)
{
int retval;
void *tree;
@@ -498,7 +497,7 @@ static int get_tree_entry(const unsigned
if (!tree)
return -1;
t.buf = tree;
- retval = find_tree_entry(&t, name, sha1);
+ retval = find_tree_entry(&t, name, sha1, mode);
free(tree);
return retval;
}
@@ -510,6 +509,7 @@ static int get_tree_entry(const unsigned
int get_sha1(const char *name, unsigned char *sha1)
{
int ret;
+ unsigned unused;
prepare_alt_odb();
ret = get_sha1_1(name, strlen(name), sha1);
@@ -518,7 +518,8 @@ int get_sha1(const char *name, unsigned
if (cp) {
unsigned char tree_sha1[20];
if (!get_sha1_1(name, cp-name, tree_sha1))
- return get_tree_entry(tree_sha1, cp+1, sha1);
+ return get_tree_entry(tree_sha1, cp+1, sha1,
+ &unused);
}
}
return ret;
--
1.3.0.g2c4a
^ permalink raw reply related
* [PATCH] sha1_name.c: no need to include diff.h; tree-walk.h will do.
From: Junio C Hamano @ 2006-04-19 20:01 UTC (permalink / raw)
To: git
Signed-off-by: Junio C Hamano <junkio@cox.net>
---
* Just a clean-up, comes on top of the one that returns mode
from get_tree_entry().
sha1_name.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/sha1_name.c b/sha1_name.c
index 35e8dfb..7ad20b5 100644
--- a/sha1_name.c
+++ b/sha1_name.c
@@ -3,7 +3,7 @@ #include "tag.h"
#include "commit.h"
#include "tree.h"
#include "blob.h"
-#include "diff.h"
+#include "tree-walk.h"
static int find_short_object_filename(int len, const char *name, unsigned char *sha1)
{
--
1.3.0.g2c4a
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox