apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stas Bekman <s...@stason.org>
Subject Re: cvs commit: apr/file_io/win32 filedup.c
Date Tue, 23 Sep 2003 00:05:07 GMT
William A. Rowe, Jr. wrote:
> At 12:25 PM 9/22/2003, Brad Nicholes wrote:
>> NetWare uses the Unix implementation of apr_file_dup() and
>>apr_file_dup2().  There is not a NetWare specific implementation of
>>these functions.  After looking at the apr_file_t structures for Unix
>>and NetWare, and comparing them to the initialization that is done in
>>the functions, it appears that the only missing initialization is 
>>flags     (if not dup2 call)
>>Is there a reason why these fields would not be set to the original
>>values after a call to apr_file_dup() or apr_file_dup2()?
> Offhand, is_pipe should be dupped, I believe.
> flags and timeout are very tricky.  Are they dup()ed by all OS's?  
> Some side effects, like close-on-exec sort of stuff should NOT be dupped,
> we want the new handle to close-on-exec even if the original did not (think
> of dupping stderr - we don't want to close fd 2 but we would certainly intend
> to have fd 712 close unless explicitly changed.)

may I suggest that apr needs to provide a coverage test for the dup 
functionality? sorry, but I'm not volunteering to do that. Just something 
that's needed before releasing 1.0.

Stas Bekman            JAm_pH ------> Just Another mod_perl Hacker
http://stason.org/     mod_perl Guide ---> http://perl.apache.org
mailto:stas@stason.org http://use.perl.org http://apacheweek.com
http://modperlbook.org http://apache.org   http://ticketmaster.com

View raw message