Here are two extracts from the man page for ppoll(2) (from the
man-pages 4.16 package; the 5.01 version is the same):
Specifying a negative value in timeout means an infinite timeout.
Other than the difference in the precision of the timeout argument, the
following ppoll() call:
ready = ppoll(&fds, nfds, tmo_p, &sigmask);
is equivalent to atomically executing the following calls:
sigset_t origmask;
int timeout;
timeout = (tmo_p == NULL) ? -1 :
(tmo_p->tv_sec * 1000 + tmo_p->tv_nsec / 1000000);
pthread_sigmask(SIG_SETMASK, &sigmask, &origmask);
ready = poll(&fds, nfds, timeout);
pthread_sigmask(SIG_SETMASK, &origmask, NULL);
But if tmo_p->tv_sec is negative, the ppoll() call is not equivalent to
the corresponding poll() call. The kernel rejects negative values of
tv_sec with an EINVAL error; it does not interpret the value as meaning
an infinite timeout.
(Yes, the kernel interprets tmo_p == NULL as an infinite timeout, but
the man page is still wrong for the case tmo_p->tv_sec < 0.)
Suggested fix: Following the end of the second extract above, add:
except that negative time values in tmo_p are not interpreted
as an infinite timeout.
Also, in the ERRORS section, change the text for EINVAL to:
EINVAL The nfds value exceeds the RLIMIT_NOFILE value or
*tmo_p contains an invalid (negative) time value.
Alan Stern
Hello Alan,
On 7/21/19 5:32 PM, Alan Stern wrote:
> Here are two extracts from the man page for ppoll(2) (from the
> man-pages 4.16 package; the 5.01 version is the same):
>
> Specifying a negative value in timeout means an infinite timeout.
>
>
> Other than the difference in the precision of the timeout argument, the
> following ppoll() call:
>
> ready = ppoll(&fds, nfds, tmo_p, &sigmask);
>
> is equivalent to atomically executing the following calls:
>
> sigset_t origmask;
> int timeout;
>
> timeout = (tmo_p == NULL) ? -1 :
> (tmo_p->tv_sec * 1000 + tmo_p->tv_nsec / 1000000);
> pthread_sigmask(SIG_SETMASK, &sigmask, &origmask);
> ready = poll(&fds, nfds, timeout);
> pthread_sigmask(SIG_SETMASK, &origmask, NULL);
>
> But if tmo_p->tv_sec is negative, the ppoll() call is not equivalent to
> the corresponding poll() call. The kernel rejects negative values of
> tv_sec with an EINVAL error; it does not interpret the value as meaning
> an infinite timeout.
>
> (Yes, the kernel interprets tmo_p == NULL as an infinite timeout, but
> the man page is still wrong for the case tmo_p->tv_sec < 0.)
>
> Suggested fix: Following the end of the second extract above, add:
>
> except that negative time values in tmo_p are not interpreted
> as an infinite timeout.
>
> Also, in the ERRORS section, change the text for EINVAL to:
>
> EINVAL The nfds value exceeds the RLIMIT_NOFILE value or
> *tmo_p contains an invalid (negative) time value.
Thanks for the report, and the text change suggestions.
I've applied the patch below.
Thanks,
Michael
diff --git a/man2/poll.2 b/man2/poll.2
index 0b023e0a5..3eacb88b7 100644
--- a/man2/poll.2
+++ b/man2/poll.2
@@ -271,7 +271,7 @@ ready = ppoll(&fds, nfds, tmo_p, &sigmask);
.EE
.in
.PP
-is equivalent to
+is nearly equivalent to
.I atomically
executing the following calls:
.PP
@@ -288,6 +288,17 @@ pthread_sigmask(SIG_SETMASK, &origmask, NULL);
.EE
.in
.PP
+The above code segment is described as
+.I nearly
+equivalent because whereas a negative
+.I timeout
+value for
+.BR poll ()
+is interpreted as an infinite timeout, a negative value expressed in
+.IR *tmo_p
+results in an error from
+.BR ppoll ().
+.PP
See the description of
.BR pselect (2)
for an explanation of why
@@ -354,6 +365,12 @@ value exceeds the
.B RLIMIT_NOFILE
value.
.TP
+.B EINVAL
+.RB ( ppoll ())
+The timeout value expressed in
+.IR *ip
+is invalid (nragtive).
+.TP
.B ENOMEM
There was no space to allocate file descriptor tables.
.SH VERSIONS
--
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/
On Mon, 29 Jul 2019, Michael Kerrisk (man-pages) wrote:
> Hello Alan,
>
> On 7/21/19 5:32 PM, Alan Stern wrote:
> > Here are two extracts from the man page for ppoll(2) (from the
> > man-pages 4.16 package; the 5.01 version is the same):
> >
> > Specifying a negative value in timeout means an infinite timeout.
> >
> >
> > Other than the difference in the precision of the timeout argument, the
> > following ppoll() call:
> >
> > ready = ppoll(&fds, nfds, tmo_p, &sigmask);
> >
> > is equivalent to atomically executing the following calls:
> >
> > sigset_t origmask;
> > int timeout;
> >
> > timeout = (tmo_p == NULL) ? -1 :
> > (tmo_p->tv_sec * 1000 + tmo_p->tv_nsec / 1000000);
> > pthread_sigmask(SIG_SETMASK, &sigmask, &origmask);
> > ready = poll(&fds, nfds, timeout);
> > pthread_sigmask(SIG_SETMASK, &origmask, NULL);
> >
> > But if tmo_p->tv_sec is negative, the ppoll() call is not equivalent to
> > the corresponding poll() call. The kernel rejects negative values of
> > tv_sec with an EINVAL error; it does not interpret the value as meaning
> > an infinite timeout.
> >
> > (Yes, the kernel interprets tmo_p == NULL as an infinite timeout, but
> > the man page is still wrong for the case tmo_p->tv_sec < 0.)
> >
> > Suggested fix: Following the end of the second extract above, add:
> >
> > except that negative time values in tmo_p are not interpreted
> > as an infinite timeout.
> >
> > Also, in the ERRORS section, change the text for EINVAL to:
> >
> > EINVAL The nfds value exceeds the RLIMIT_NOFILE value or
> > *tmo_p contains an invalid (negative) time value.
>
> Thanks for the report, and the text change suggestions.
> I've applied the patch below.
>
> Thanks,
>
> Michael
>
> diff --git a/man2/poll.2 b/man2/poll.2
> index 0b023e0a5..3eacb88b7 100644
> --- a/man2/poll.2
> +++ b/man2/poll.2
> @@ -271,7 +271,7 @@ ready = ppoll(&fds, nfds, tmo_p, &sigmask);
> .EE
> .in
> .PP
> -is equivalent to
> +is nearly equivalent to
> .I atomically
> executing the following calls:
> .PP
> @@ -288,6 +288,17 @@ pthread_sigmask(SIG_SETMASK, &origmask, NULL);
> .EE
> .in
> .PP
> +The above code segment is described as
> +.I nearly
> +equivalent because whereas a negative
> +.I timeout
> +value for
> +.BR poll ()
> +is interpreted as an infinite timeout, a negative value expressed in
> +.IR *tmo_p
> +results in an error from
> +.BR ppoll ().
> +.PP
> See the description of
> .BR pselect (2)
> for an explanation of why
> @@ -354,6 +365,12 @@ value exceeds the
> .B RLIMIT_NOFILE
> value.
> .TP
> +.B EINVAL
> +.RB ( ppoll ())
> +The timeout value expressed in
> +.IR *ip
> +is invalid (nragtive).
----------------^^^
Typo. Otherwise, I approve of the patch.
Thank you,
Alan
> +.TP
> .B ENOMEM
> There was no space to allocate file descriptor tables.
> .SH VERSIONS
>> +The timeout value expressed in
>> +.IR *ip
>> +is invalid (nragtive).
> ----------------^^^
>
> Typo. Otherwise, I approve of the patch.
Thanks! Fixed.
Cheers,
Michael
--
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/