threads / discuss / 1174

Linus kernel tree corrupt?

Subject: Linus kernel tree corrupt?

## tl;dr

7 messages between Jul 8, 2005 and Jul 10, 2005.

replies: 6people: 5as markdown or json

Jon Smirl· Jul 8, 2005, 15:57 UTC · lore

What happened in this session...

[jonsmirl@jonsmirl linus]$ cg update 11:52:32 URL:http://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.g it/refs/heads/master [41/41] -> "refs/heads/origin" [1] progress: 34 objects, 102329 bytes error: File 9cd17004869977026b473030ebb5b6fb71bfcebf (http://www.kernel.org/pub/ scm/linux/kernel/git/torvalds/linux-2.6.git/objects/9c/d17004869977026b473030ebb 5b6fb71bfcebf) corrupt

Cannot obtain needed tree 9cd17004869977026b473030ebb5b6fb71bfcebf while processing commit 0000000000000000000000000000000000000000. cg-pull: objects pull failed [jonsmirl@jonsmirl linus]$ cg update 11:53:26 URL:http://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.g it/refs/heads/master [41/41] -> "refs/heads/origin" [1] Up to date.

Applying changes... Branch already fully merged. [jonsmirl@jonsmirl linus]$ cg pull 11:53:33 URL:http://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.g it/refs/heads/master [41/41] -> "refs/heads/origin" [1] Up to date. [jonsmirl@jonsmirl linus]$ cg diff error: cannot read sha1_file for 9cd17004869977026b473030ebb5b6fb71bfcebf fatal: unable to read tree object 7fd73e9f39bf6003cc3188a10426b62d8c47ab40 [jonsmirl@jonsmirl linus]$

-- 
Jon Smirl
jonsmirl@gmail.com

Tony Luck· Jul 8, 2005, 17:06 UTC · re: Jon Smirl · lore

Re: Linus kernel tree corrupt?

On 7/8/05, Jon Smirl <jonsmirl@gmail.com> wrote:

> What happened in this session...

Linus has "packed" his GIT tree ... and now http-pull doesn't work. rsync still does (provided you have a new enough cogito).

-Tony

Russell King· Jul 9, 2005, 10:43 UTC · re: Tony Luck · lore

Re: Linus kernel tree corrupt?

On Fri, Jul 08, 2005 at 10:06:09AM -0700, Tony Luck wrote:

Show 6 quoted lines

> On 7/8/05, Jon Smirl <jonsmirl@gmail.com> wrote:
> > What happened in this session...
> 
> Linus has "packed" his GIT tree ... and now http-pull doesn't work. 
> rsync still does (provided
> you have a new enough cogito).

So does that mean we should ignore cogito's whinging about rsync being deprecated?

-- 
Russell King

Petr Baudis· Jul 9, 2005, 11:55 UTC · re: Russell King · lore

Re: Linus kernel tree corrupt?

Dear diary, on Sat, Jul 09, 2005 at 12:43:03PM CEST, I got a letter where Russell King <rmk@arm.linux.org.uk> told me that...

Show 10 quoted lines

> On Fri, Jul 08, 2005 at 10:06:09AM -0700, Tony Luck wrote:
> > On 7/8/05, Jon Smirl <jonsmirl@gmail.com> wrote:
> > > What happened in this session...
> > 
> > Linus has "packed" his GIT tree ... and now http-pull doesn't work. 
> > rsync still does (provided
> > you have a new enough cogito).
> 
> So does that mean we should ignore cogito's whinging about rsync being
> deprecated?

Yes, please do. I deprecated rsync a day before Linus "broke" http-pull. It's un-deprecated again for now in the latest Cogito.

-- 
				Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
<Espy> be careful, some twit might quote you out of context..

H. Peter Anvin· Jul 9, 2005, 18:03 UTC · re: Petr Baudis · lore

Re: Linus kernel tree corrupt?

Petr Baudis wrote:

> 
> Yes, please do. I deprecated rsync a day before Linus "broke" http-pull.
> It's un-deprecated again for now in the latest Cogito.
>

Presumably for packed repos you want to drop the --ignore-existing --whole-file options I assume?

Also, pulling with cogito-0.12:

cg-clone -s rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git/

I still get:

receiving file list ... done master

sent 147 bytes received 863 bytes 2020.00 bytes/sec total size is 41 speedup is 0.04

receiving file list ... done progress: 1929 objects, 5796451 bytes, 100% done pack/pack-e3117bbaf6a59cb53c3f6f0d9b17b9433f0e4135.idx pack/pack-e3117bbaf6a59cb53c3f6f0d9b17b9433f0e4135.pack

sent 42599 bytes received 73373124 bytes 451789.06 bytes/sec total size is 73178114 speedup is 1.00

receiving file list ... done v2.6.11 v2.6.11-tree v2.6.12 v2.6.12-rc2 v2.6.12-rc3 v2.6.12-rc4 v2.6.12-rc5 v2.6.12-rc6 v2.6.13-rc1 v2.6.13-rc2

sent 339 bytes received 1802 bytes 4282.00 bytes/sec total size is 410 speedup is 0.19 Missing object of tag v2.6.11... different source (obsolete tag?) Missing object of tag v2.6.11-tree... different source (obsolete tag?) Missing object of tag v2.6.12... different source (obsolete tag?) Missing object of tag v2.6.12-rc2... different source (obsolete tag?) Missing object of tag v2.6.12-rc3... different source (obsolete tag?) Missing object of tag v2.6.12-rc4... different source (obsolete tag?) Missing object of tag v2.6.12-rc5... different source (obsolete tag?) Missing object of tag v2.6.12-rc6... different source (obsolete tag?) Missing object of tag v2.6.13-rc1... different source (obsolete tag?) Missing object of tag v2.6.13-rc2... different source (obsolete tag?) New branch: 0109fd37046de64e8459f8c4f4706df9ac7cc82c Cloned (origin rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git available as branch "origin") Cloned to ./ (origin rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git available as branch "origin")

Is the "missing objects" thing spurious?

	-hpa

Petr Baudis· Jul 10, 2005, 15:23 UTC · re: H. Peter Anvin · lore

Re: Linus kernel tree corrupt?

Dear diary, on Sat, Jul 09, 2005 at 08:03:32PM CEST, I got a letter where "H. Peter Anvin" <hpa@zytor.com> told me that...

Show 8 quoted lines

> Petr Baudis wrote:
> >
> >Yes, please do. I deprecated rsync a day before Linus "broke" http-pull.
> >It's un-deprecated again for now in the latest Cogito.
> >
> 
> Presumably for packed repos you want to drop the --ignore-existing 
> --whole-file options I assume?

It _is_ unsafe for individual objects, and your packfile will be corrupt if you break it in the middle and not have --whole-file turned on, I assume. It would be ideal if we could make rsync allow resuming download of the file if interrupted, but not under the final name but in that hidden file it uses.

Show 21 quoted lines

> sent 339 bytes  received 1802 bytes  4282.00 bytes/sec
> total size is 410  speedup is 0.19
> Missing object of tag v2.6.11... different source (obsolete tag?)
> Missing object of tag v2.6.11-tree... different source (obsolete tag?)
> Missing object of tag v2.6.12... different source (obsolete tag?)
> Missing object of tag v2.6.12-rc2... different source (obsolete tag?)
> Missing object of tag v2.6.12-rc3... different source (obsolete tag?)
> Missing object of tag v2.6.12-rc4... different source (obsolete tag?)
> Missing object of tag v2.6.12-rc5... different source (obsolete tag?)
> Missing object of tag v2.6.12-rc6... different source (obsolete tag?)
> Missing object of tag v2.6.13-rc1... different source (obsolete tag?)
> Missing object of tag v2.6.13-rc2... different source (obsolete tag?)
> New branch: 0109fd37046de64e8459f8c4f4706df9ac7cc82c
> Cloned (origin 
> rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git 
> available as branch "origin")
> Cloned to ./ (origin 
> rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git 
> available as branch "origin")
> 
> Is the "missing objects" thing spurious?

Yes, already fixed. Will release 0.12.1 soon.

-- 
				Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
<Espy> be careful, some twit might quote you out of context..

H. Peter Anvin· Jul 10, 2005, 20:35 UTC · re: Petr Baudis · lore

Re: Linus kernel tree corrupt?

Petr Baudis wrote:

Show 7 quoted lines

> 
> It _is_ unsafe for individual objects, and your packfile will be corrupt
> if you break it in the middle and not have --whole-file turned on, I
> assume. It would be ideal if we could make rsync allow resuming download
> of the file if interrupted, but not under the final name but in that
> hidden file it uses.
>

I think if you just don't give it --partial you're fine (rsync always creates a second copy, --partial decides if a partial file should be thrown away or not.)

	-hpa