Re: cvs commit: apr/test Makefile.in test_apr.h testall.c testprocmutex.c

[Jeff Trawick <tr...@attglobal.net>]

jorton@apache.org wrote:
> jorton      2003/07/02 05:12:30
> 
>   Modified:    test     Makefile.in test_apr.h testall.c testprocmutex.c
>   Log:
>   - dropped support for specifying a lock filename, hopefully that is
>   not critical

At one point, the test could not be run naturally on a machine with home 
dir over NFS.

Maybe if you pass NULL for lockname it would work since APR would create 
it in /tmp (ISTR)?

It looks like the previous code normally passed NULL, which may be a 
change after the lockname option was added which negated the need for 
the lockname to be specified.

Also

+    const char *lockname = "tpm.lock";
+    const char *shmname = "tpm.shm";

Did you intend "tmp.lock" and "tmp.shm"?

Re: cvs commit: apr/test Makefile.in test_apr.h testall.c testprocmutex.c

[Joe Orton <jo...@manyfish.co.uk>]

On Wed, Jul 02, 2003 at 08:59:19AM -0400, Jeff Trawick wrote:
> jorton@apache.org wrote:
> >jorton      2003/07/02 05:12:30
> >
> >  Modified:    test     Makefile.in test_apr.h testall.c testprocmutex.c
> >  Log:
> >  - dropped support for specifying a lock filename, hopefully that is
> >  not critical
> 
> At one point, the test could not be run naturally on a machine with home 
> dir over NFS.
> 
> Maybe if you pass NULL for lockname it would work since APR would create 
> it in /tmp (ISTR)?
>
> It looks like the previous code normally passed NULL, which may be a 
> change after the lockname option was added which negated the need for 
> the lockname to be specified.

Ah, good point. Yes, from looking at the code NULL does indeed look like
the better default choice; the header said a filename should always be
used so I just did that.

> Also
> 
> +    const char *lockname = "tpm.lock";
> +    const char *shmname = "tpm.shm";
> 
> Did you intend "tmp.lock" and "tmp.shm"?

No, "tpm" was my little acronym for "test process mutex" :)

Thanks for the review.