Error (?) in man page for ppoll(2)

Error (?) in man page for ppoll(2)

[Alan Stern]

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

Re: Error (?) in man page for ppoll(2)

[Michael Kerrisk (man-pages)]

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/

Re: Error (?) in man page for ppoll(2)

[Alan Stern]

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

Re: Error (?) in man page for ppoll(2)

[Michael Kerrisk (man-pages)]

>> +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/