git-rev-list --merge-order hangs

git-rev-list --merge-order hangs

[Radoslaw Szkodzinski]

I've got a Linux git tree with some merges and some other commits
interwined.
I want to print all my changes in merge order, so I do:

git-rev-list --pretty --merge-order v2.6.12-rc6-astorm1 ^v2.6.12-rc6

It prints the first 3 commits (linear), but then hangs, eating all
available CPU.
Without --merge-order it works fine, but obviously lists the commits by
parent.

AstralStorm

Re: git-rev-list --merge-order hangs

[Jon Seymour]

On 6/8/05, Radoslaw Szkodzinski <astralstorm@gorzow.mm.pl> wrote:
> I've got a Linux git tree with some merges and some other commits
> interwined.
> I want to print all my changes in merge order, so I do:
> 
> git-rev-list --pretty --merge-order v2.6.12-rc6-astorm1 ^v2.6.12-rc6
> 
> It prints the first 3 commits (linear), but then hangs, eating all
> available CPU.
> Without --merge-order it works fine, but obviously lists the commits by
> parent.

Can you git-rev-list >/dev/null without arguments, then do a
git-rev-list --pretty --merge-order again and see it works any better?

If it still doesn't work, can you send me the output of:

        git-rev-list --parents v2.6.12-rc6-astorm1 ^v2.6.12-rc6 

I may then need to request a tarball of the related git objects so
that I can reproduce the problem on my end, but I'll send you the list
I need to see.

Regards,

jon.

Re: git-rev-list --merge-order hangs

[Jon Seymour]

Actually, one case that I may not have considered properly is a commit
near the head that has no parents.

      git-rev-list --parents v2.6.12-rc6-astorm1 ^v2.6.12-rc6

should tell you if this case exists. Ideally such a case should not
occur, but that isn't a logical certainity, so I should handle it
better than I currently do.

Please let me know if this may explain your case. 

Regards,

jon.

Re: git-rev-list --merge-order hangs

[Jon Seymour]

Ok, I reproduced exponential behaviour when a commit with no parents
is merged near the head of the kernel and a ^ argument is specified.

Fortunately a work-around is a one line change - a patch will be
forthcoming shortly.
The workaround will still require a full-graph scan but that is linear
rather than
exponential so is tolerable.

A better solution which will not require a full-graph scan will be
delivered in a future
patch.

Thanks for the report.

jon.

Re: git-rev-list --merge-order hangs

[Radoslaw Szkodzinski]

Jon Seymour wrote:

>Actually, one case that I may not have considered properly is a commit
>near the head that has no parents.
>
>      git-rev-list --parents v2.6.12-rc6-astorm1 ^v2.6.12-rc6
>
>should tell you if this case exists. Ideally such a case should not
>occur, but that isn't a logical certainity, so I should handle it
>better than I currently do.
>
>Please let me know if this may explain your case. 
>  
>
This just threw tons of SHA1 hashes at me without complaining.
I've also tried git-rev-list HEAD >/dev/null without any chage whatsoever.

The tree is not secret, You can try (very slow) rsync at:
rsync://astralstorm.servegame.com/linux-2.6-astorm

Just update latest linux-2.6 with it to save time.
(mine is up to commit 1d6757fbff5bc86e94e59ab0d7bdd7e71351d839)

AstralStorm