Help me follow...

On 14 July 2012 16:57, Perry Smith <pedzsan@gmail.com> wrote:

I hit "bug 5665". I see threads like this:

http://archives.postgresql.org/pgsql-bugs/2011-01/msg00241.php

Where / how can I find the diff that went in to fix this? Mostly I'm curious. I can work around the problem but wanted to create a patch file for myself to make deploying to other servers easier.

You can search through commits here:

http://git.postgresql.org/gitweb/?p=postgresql.git;a=summary

Sometimes a reference to the bug number may appear within the commit
message, but this won't always happen. If you can't find the commit
based on that, try the original reporter's name. We ordinarily credit
bug reporters in commit messages that fix the bug.

--
Peter Geoghegan http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training and Services

Peter Geoghegan <peter@2ndquadrant.com> writes:

On 14 July 2012 16:57, Perry Smith <pedzsan@gmail.com> wrote:

I hit "bug 5665". I see threads like this:

http://archives.postgresql.org/pgsql-bugs/2011-01/msg00241.php

Where / how can I find the diff that went in to fix this? Mostly I'm curious. I can work around the problem but wanted to create a patch file for myself to make deploying to other servers easier.

You can search through commits here:
http://git.postgresql.org/gitweb/?p=postgresql.git;a=summary
Sometimes a reference to the bug number may appear within the commit
message, but this won't always happen. If you can't find the commit
based on that, try the original reporter's name. We ordinarily credit
bug reporters in commit messages that fix the bug.

I'm not sure we ever did fix it; I see no references to that bug in the
commit history, nor to the similar bug #6059, nor does it look like we
changed anything in the relevant make rules.

I can no longer reproduce a failure when building out-of-tree from a
tarball on my Fedora 16 box. That's still make 3.82, but Fedora is
carrying half a dozen patches to the upstream tarball:
http://pkgs.fedoraproject.org/gitweb/?p=make.git;a=tree;h=refs/heads/f16;hb=f16
It's not immediately obvious which of those might be related, but
I think this was a bug in gmake and the fix is in there somewhere.

regards, tom lane

On 14 July 2012 18:56, Tom Lane <tgl@sss.pgh.pa.us> wrote:

I can no longer reproduce a failure when building out-of-tree from a
tarball on my Fedora 16 box. That's still make 3.82, but Fedora is
carrying half a dozen patches to the upstream tarball:
http://pkgs.fedoraproject.org/gitweb/?p=make.git;a=tree;h=refs/heads/f16;hb=f16
It's not immediately obvious which of those might be related, but
I think this was a bug in gmake and the fix is in there somewhere.

Agreed. Since Fedora 14 is now EOL, I'm not inclined to worry about it.

--
Peter Geoghegan http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training and Services

On Jul 14, 2012, at 2:19 PM, Peter Geoghegan wrote:

On 14 July 2012 18:56, Tom Lane <tgl@sss.pgh.pa.us> wrote:

I can no longer reproduce a failure when building out-of-tree from a
tarball on my Fedora 16 box. That's still make 3.82, but Fedora is
carrying half a dozen patches to the upstream tarball:
http://pkgs.fedoraproject.org/gitweb/?p=make.git;a=tree;h=refs/heads/f16;hb=f16
It's not immediately obvious which of those might be related, but
I think this was a bug in gmake and the fix is in there somewhere.

Agreed. Since Fedora 14 is now EOL, I'm not inclined to worry about it.

I'm on AIX 6.1. Using gmake 3.82. PostgreSQL 9.0.4

Its definitely not a show stopper but from the email traffic, it looked like someone was going to make the Makefiles "less cutting edge" :-)

Thank you for your time,
Perry