Re: [PATCH] fix bugs in run_in_dir ()

[James Youngman]

On Fri, May 7, 2010 at 9:21 AM, Kamil Dudka <address@hidden> wrote:
> On Friday 07 of May 2010 10:08:08 James Youngman wrote:
>> Thanks for spotting this.   I can see the bug simply by inspecting
>> your fix.   But, is there any kind of test case which actually
>> demonstrates the problem?   If we add a relevant test we can protect
>> ourselves against reintroducing this kind of bug, too.
>
> As I said, I hit the bug in 4.4.2.  The code that the patch addresses works
> in 4.5.9 without the patch, but not in 4.4.2.  The failing test-case was
> execdir-multiple.exp.

I see.  Thanks.

I have applied and pushed your patch as the two attached git patches
(if you could send your changes in "git format-patch" format in
future, ideally with a ChangeLog entry too, that would streamline
things for me).

Thank you for this contribution to GNU findutils.

> It simply broke the traversal as soon as the first
> execdir had been executed.  FWIW the full backport for upstream bugs #19593
> and #27563 is attached to the corresponding Fedora bug:
>
> https://bugzilla.redhat.com/493143

Thanks for the pointer.   The semantics that Gen Zhang was relying on
(processing all entries in a directory at once) were never guaranteed
anyway.

If you'd like to supply a backported "format-patch" patch against the
4.4.x tree, I can apply that too.

Thanks,
James.

0001-Correctly-initialise-variables-in-run_in_dir.patch
Description: Text Data

0002-Add-Kamil-Dudka-to-the-AUTHORS-file.patch
Description: Text Data