Re: INFO: no limit on AIO_MAX
Andrew, George:
Sorry about that, the test results of concern are:
T.aio_cancel 4 Failed
T.aio_error 1 Failed
T.aio_return 3 Failed
T.aio_suspend 4 Failed
T.aio_suspend 5 Failed
The INFO message is being printed out just before the actual failure
message (e.g. "aio_cancel for completed read request failed ..."). It
is not a big deal, I just wanted to make you aware that this message is
showing up when it probably should not. Thanks.
Amos W.
On Thu, Jun 27, 2002 at 06:40:38AM +0100, Andrew Josey wrote:
> Amos
> Its just informational, and the test result is probably UNTESTED
> or NOTINUSE (you do not say) so this helps in determining why
> the test result is as it is.
> regards
> Andrew
>
> On Jun 26, 3:03pm in "INFO: no limit on AI", Amos Waterland wrote:
> > Most of the LSB tests for aio, when run against glibc 2.2.5, are
> > printing out a message that says:
> >
> > INFO: No limit on AIO_MAX
> >
> > which is produced by code that looks like this:
> >
> > if ((aio_maxl = sysconf(_SC_AIO_MAX)) == -1) {
> > if (errno == 0) {
> > in_rpt("INFO: No limit on AIO_MAX");
> > aio_maxl = LONG_MAX;
> > } else {
> > vsrt_sysfail("sysconf(_SC_AIO_MAX)");
> > DBUG_VOID_RETURN
> > }
> > }
> >
> > I asked about something related to this on libc-alpha, and it was
> > explained to me that glibc's AIO_MAX is indefinite (in principle greater
> > than _POSIX_AIO_MAX but indeterminate and so is not defined in
> > <limits.h>), so sysconf() returns -1 with errno == 0, as per the
> > standard.
> >
> > Is the message that is being printed out intended as a warning? If so,
> > I do not think it should exist, as I believe that glibc is behaving
> > correctly. Thanks.
> >
> > Amos W.
--
To UNSUBSCRIBE, email to lsb-test-request@lists.linuxbase.org
with subject of "unsubscribe". Trouble? Email listmaster@lists.linuxbase.org
Reply to: