From mboxrd@z Thu Jan  1 00:00:00 1970
From: matthew green <mrg@cygnus.com>
To: Ben Elliston <bje@redhat.com>
Cc: sid@sources.redhat.com
Subject: re: RFA: gloss component extensions 
Date: Sun, 07 Jan 2001 18:38:00 -0000
Message-id: <28706.978921505@cygnus.com>
References: <Pine.LNX.4.31.0101081253570.27669-100000@moshpit.cygnus.com>
X-SW-Source: 2001-q1/msg00020.html

   
   Before you commit, I have some questions:

actually, i already have as frank approved it.
   
      2000-12-30  matthew green  <mrg@redhat.com>
   
      	* gloss.cxx (do_sys_time): New function.
      	(do_sys_times): Likewise.
      	(do_sys_gettimeofday): Likewise.
      	(syscall_trap): Call do_sys_time().  Call new do_sys_times() and
      	do_sys_gettimeofday() for times() and .
   
   `And ..' .. what?

` .. and gettimeofday, respectively.' is what the ChangeLog entry now says.
   
      + #else
      +   set_int_result(-1);
      + #endif
   
   In the case of missing gettimeofday() and times() functions, the newlib stub
   functions should set ENOSYS.  Do they use the -1 to detect this?

ah, i should probably also add a call:

	set_error_result(newlib::eNoSys);

then?
   
   Other than that, the only thing missing is your test cases to prove that
   these work.

actually, the entire gloss32 component has no testsuite at all.  :-(
(parts of the hardware specific ones do, but that is it.)