httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Martin Kraemer <Martin.Krae...@mch.sni.de>
Subject [APR] writev Problem (HAVE_WRITEV / NO_WRITEV / ap_writev)
Date Mon, 10 Apr 2000 09:33:27 GMT
I see a problem in the current apache_2.0 code regarding the portability
to machines that lack the writev system call.

The current situation is like this:
* configure tests whether writev is available and sets HAVE_WRITEV
  accordingly
* file_io/unix/readwrite.c checks HAVE_WRITEV and defines the function
  ap_writev() based on writev() if HAVE_WRITEV is set
* main/iol_file.c calls ap_writev() from file_ap_writev()

IMHO the second step should *ALWAYS* define ap_writev(), possibly as an
emulation wrapper based on write(), *OR*
all functions using ap_writev() should be able to emulate it (based on
ap_write()) if HAVE_WRITEV is not set. Of course, the latter is a stupid
idea.

What was the design idea behind it?

    Martin
--
<Martin.Kraemer@MchP.Siemens.De>             |    Fujitsu Siemens
Fon: +49-89-636-46021, FAX: +49-89-636-41143 | 81730  Munich,  Germany


Mime
View raw message