Re: Compile error of last version - Lenz Grimmer - com.mysql.lists.packagers

10 messages in com.mysql.lists.packagersRe: Compile error of last version

From	Sent On	Attachments
Mauro Sacchetto	08 Dec 2005 13:02
Lenz Grimmer	09 Dec 2005 09:12
Mauro Sacchetto	09 Dec 2005 09:26
Mauro Sacchetto	09 Dec 2005 16:41
Mauro Sacchetto	10 Dec 2005 04:45
Tom Lane	10 Dec 2005 08:29
Lenz Grimmer	12 Dec 2005 05:40
Lenz Grimmer	12 Dec 2005 05:48
Tom Lane	12 Dec 2005 14:07
Lenz Grimmer	12 Dec 2005 23:56

Subject:	Re: Compile error of last version
From:	Lenz Grimmer (le...@mysql.com)
Date:	12/12/2005 05:40:41 AM
List:	com.mysql.lists.packagers

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1

Hi,

On Sat, 10 Dec 2005, Tom Lane wrote:

Mauro Sacchetto <maur...@alice.it> writes:

I communicate it to bugzilla. The kernel team replied:

Additionally, last I checked MySQL abused atomic.h in a way it was never meant to be used (for using the inline functions in userspace).

Could you please give me the related Bugzilla ID for reference?

I compiled from myself the headers for the most recent kernel. With 2.6.13 mysql compilation worls fine, with 2.6.14 no. Please, consider if the problem is not of mysql itself...

It's a MySQL problem, related to the detection of atomic operations in "configure".

The Red Hat distribution of mysql ran into this problem quite some time ago: the "atomic add/sub" assembly code doesn't work on any of RH's 64-bit platforms. Fortunately there is a configure-driven option to not use that code. We fix it by hacking configure.in (see attached patch, which presumes that you run autoconf afterwards) but I suppose you could also do it by modifying configure's output files. The patch was generated against 4.1.x, as you can see, but it still applies to 5.0.x.

Thanks! Could you please alert us about such issues via our bug database at http://bugs.mysql.com/ in the future? That should make sure you don't have to maintain such fixes by yourself forever and we don't have to trawl through all vendors' SRPMs for useful patches :)

The idea of userland using <asm/atomic.h> has our kernel hackers all bent out of shape, so don't do it. Suppressing this has another benefit too, which is we avoid the 4-vs-8-byte problem in BDB that formerly required mysql-threads.patch.

Agreed, this is ugly. It seems like this particular issue was reported in our Bugs DB already:

http://bugs.mysql.com/bug.php?id=13621 http://bugs.mysql.com/bug.php?id=15244

I will nudge our devs about this one again. Thanks for the reminder!

iD8DBQFDnX3xSVDhKrJykfIRAl1nAJ4r9EzuCaCmwfZx08ckWmapGLcRhACfR/ls 1fJG28/3vobCepY98yo7iHY= =La0C -----END PGP SIGNATURE-----