Re: git 1.5.3.5 error over NFS

This is the almost same message I cced to lk and nfs, but with
Git-interasting parts added.

Bill Lear, Mon, Nov 12, 2007 16:39:15 +0100:

On Saturday, November 10, 2007 at 00:21:06 (+0100) Alex Riesen writes:

quoted

Bill Lear, Fri, Nov 09, 2007 16:31:39 +0100:

quoted

I've brought this up before, but I don't recall a resolution to it.

We have an NFS-mounted filesystem, and git pull is choking on it.

% uname -a
Linux uhlr.zopyra.com 2.6.9-42.0.2.ELsmp #1 SMP Wed Aug 23 13:38:27 BST 2006 x86_64 x86_64 x86_64 GNU/Linux

It is a really old kernel... Maybe you could try with some of the
recent ones?

quoted

quoted

% git --version
git version 1.5.3.5

% git pull
remote: Generating pack...
remote: Done counting 998 objects.
remote: Result has 836 objects.
remote: Deltifying 836 objects.
remote:  100% (836/836) done
Indexing 836 objects...
remote: Total 836 (delta 526), reused 688 (delta 380)
  100% (836/836) done
Resolving 526 deltas...
fatal: cannot pread pack file: No such file or directory

Could you please strace it? With strace -ff?

quoted

fatal: index-pack died with error code 128
fatal: Fetch failure: git://source/repo

I looked through the archives of this list and did not see a final
resolution, other than a suspected bug in the OS NFS code.

Strace, just to be on the safe side

Ok, I've done the strace -ff.  It has generated 176 strace.out.<pid>
files.  I have placed a tarball of these files on my home server:

  http://www.zopyra.com/~rael/git/git-trace.tar.bz2

The file strace.out.25526 has, at the very end:

Yes, this is the file.

I extend the part you quoted. The file is opened here:

open(".git/objects/pack_awOTNW", O_RDWR|O_CREAT|O_EXCL, 0600) = 3
write(3, "PACK\0\0\0\2\0\0\3r", 12)     = 12
... lots of data. Counted writes: 488538...
write(3, "v\351\247V\325\362\327/\240\265\211\211\322,\261\210\301"..., 4096) = 4096
write(3, "\202\202g\232Bf\211Bf\261Bb\36X\22\4\364\365\25\22\223"..., 1114) = 1114
write(3, "M", 1)                        = 1

write(2, "Resolving 551 deltas.\n", 22)                               = 22
pread64(3, "", 242, 541) = 0

That's the problem.

"Git-interesting parts"

write(2, "fatal: ", 7)                  = 7
write(2, "cannot pread pack file: Success", 31) = 31

This is strange. The current git should not produce anything like
this (and does not, here). The code in question is:

	do {
		ssize_t n = pread(pack_fd, data + rdy, len - rdy, from + rdy);
		if (n <= 0)
			die("cannot pread pack file: %s", strerror(errno));
		rdy += n;
	} while (rdy < len);

Either your strace or that kernel seem to have clobbered errno, too.

fstat64(0, 0xffff907c)                  = 0
open("/etc/mtab", O_RDONLY)             = 4
fstat64(0x4, 0xffff6810)                = 0

Interesing. What in git index-pack can read mtab and what for?
Is it vanilla Git or have you instrumented it somehow?

[ Process PID=25347 runs in 32 bit mode. ]
umovestr: Input/output error
umovestr: Input/output error

Is it your strace?
Do you have anything unusual in syslog?

BTW, you could try defining NO_PREAD in config.mak, and see if it
works the problem around for you.