[Bug manual/17459] New: Manual doesn't mention that timeout argument of select can be modified

[Bug manual/17459] New: Manual doesn't mention that timeout argument of select can be modified

[GavinSmith0123 at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17459

            Bug ID: 17459
           Summary: Manual doesn't mention that timeout argument of select
                    can be modified
           Product: glibc
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: manual
          Assignee: unassigned at sourceware dot org
          Reporter: GavinSmith0123 at gmail dot com
                CC: mtk.manpages at gmail dot com, roland at gnu dot org

In the "Waiting for I/O" node of the manual (which is in the manual/llio.texi
file), in the description of the "select" function, it is mentioned that you
can timeout waiting for input or output by passing a pointer to a struct
timeval object to "select". However, it doesn't say that select can overwrite
this object. This led to a problem when I tried to reuse the struct timeval
object in a loop with a select call.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug manual/17459] Manual doesn't mention that timeout argument of select can be modified

[mtk.manpages at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17459

--- Comment #1 from Michael Kerrisk <mtk.manpages at gmail dot com> ---
(In reply to Gavin Smith from comment #0)
> In the "Waiting for I/O" node of the manual (which is in the
> manual/llio.texi file), in the description of the "select" function, it is
> mentioned that you can timeout waiting for input or output by passing a
> pointer to a struct timeval object to "select". However, it doesn't say that
> select can overwrite this object. This led to a problem when I tried to
> reuse the struct timeval object in a loop with a select call.

Confirmed. POSIX permits this behavior. The select(2) man page documents this
point:

       On Linux, select() modifies timeout to reflect  the  amount  of
       time  not  slept;  most  other  implementations do not do this.
       (POSIX.1-2001 permits either behavior.)  This  causes  problems
       both  when  Linux  code  which reads timeout is ported to other
       operating systems, and when code is ported to Linux that reuses
       a  struct  timeval  for  multiple  select()s  in a loop without
       reinitializing it.  Consider  timeout  to  be  undefined  after
       select() returns.

A fix to the glibc manual is probably in order.

Cheers,

Michael

-- 
You are receiving this mail because:
You are on the CC list for the bug.