Re: amd64(gcc4): aptitude dumps core
On Thu, Jan 13, 2005 at 10:36:33AM -0600, Pete Harlan wrote:
>
> If the pure64 version worked for us we would have used it, but it
> didn't. (For details see my post in this list several months ago.)
> Switching to the gcc-3.4 archive finally worked for us.
I'm assuming you're talking about mysql?
I've just build mysql-dfsg 4.0.23-1, ran make test, and got:
All 211 tests were successful.
Then I did the same with mysql-dfsg-4.1 4.1.8a-4 and got:
create [ fail ]
Errors are (from
/usr/src/mysql-dfsg-4.1-4.1.8a/mysql-test/var/log/mysqltest-time)
:
/usr/src/mysql-dfsg-4.1-4.1.8a/client/.libs/mysqltest: At line
413: query 'use test' failed: 1044: Access denied for user
''@'localhost' to database 'test'
(the last lines may be the most important ones)
Aborting: create failed. To continue, re-run with '--force'.
Note that i386 fails in the very same way and this is probably a
bug in the test suite.
As far as I know there wasn't even anything fixed that should
explain why it works now. There were only patched to make it
build on amd64 using nptl instead of linuxthreads. It failed to
build without the patch.
What could be different between pure64 and gcc-3.4 is the speed
by which a package might be available that required a patch to
build. With pure64 we just send the patch the BTS and wait for
the maintainer to apply it. The gcc-3.4 archive on the other
hand applies all kinds of patches, even totaly broken patches,
but it will be in the archive sooner.
Kurt
Reply to: