[PATCH] Fix ftime

[PATCH] Fix ftime

[Jakub Jelinek]

Hi!

I couldn't find anything which would mandate this, but at least it is very
weird to have millitm = 1000 returned (in 0.0999% of cases) when it is supposed
to mean millisecond part of the time.

2002-03-04  Jakub Jelinek  <jakub@redhat.com>

	* sysdeps/unix/bsd/ftime.c (ftime): Don't return 1000 in millitm.

--- libc/sysdeps/unix/bsd/ftime.c.jj	Thu Aug 23 18:50:29 2001
+++ libc/sysdeps/unix/bsd/ftime.c	Mon Mar  4 16:47:20 2002
@@ -1,4 +1,4 @@
-/* Copyright (C) 1994, 1997 Free Software Foundation, Inc.
+/* Copyright (C) 1994, 1997, 2002 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -33,5 +33,10 @@ ftime (timebuf)
   timebuf->millitm = (tv.tv_usec + 999) / 1000;
   timebuf->timezone = tz.tz_minuteswest;
   timebuf->dstflag = tz.tz_dsttime;
+  if (timebuf->millitm == 1000)
+    {
+      timebuf->time++;
+      timebuf->millitm = 0;
+    }
   return 0;
 }

	Jakub

Re: [PATCH] Fix ftime

[Andreas Schwab]

Jakub Jelinek <jakub@redhat.com> writes:

|> Hi!
|> 
|> I couldn't find anything which would mandate this, but at least it is very
|> weird to have millitm = 1000 returned (in 0.0999% of cases) when it is supposed
|> to mean millisecond part of the time.
|> 
|> 2002-03-04  Jakub Jelinek  <jakub@redhat.com>
|> 
|> 	* sysdeps/unix/bsd/ftime.c (ftime): Don't return 1000 in millitm.
|> 
|> --- libc/sysdeps/unix/bsd/ftime.c.jj	Thu Aug 23 18:50:29 2001
|> +++ libc/sysdeps/unix/bsd/ftime.c	Mon Mar  4 16:47:20 2002
|> @@ -1,4 +1,4 @@
|> -/* Copyright (C) 1994, 1997 Free Software Foundation, Inc.
|> +/* Copyright (C) 1994, 1997, 2002 Free Software Foundation, Inc.
|>     This file is part of the GNU C Library.
|>  
|>     The GNU C Library is free software; you can redistribute it and/or
|> @@ -33,5 +33,10 @@ ftime (timebuf)
|>    timebuf->millitm = (tv.tv_usec + 999) / 1000;

Perhaps we should actually truncate here instead of rounding up.

|>    timebuf->timezone = tz.tz_minuteswest;
|>    timebuf->dstflag = tz.tz_dsttime;
|> +  if (timebuf->millitm == 1000)
|> +    {
|> +      timebuf->time++;
|> +      timebuf->millitm = 0;
|> +    }
|>    return 0;
|>  }

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE GmbH, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [PATCH] Fix ftime

[Ulrich Drepper]

[-- Attachment #1: Type: text/plain, Size: 564 bytes --]

On Mon, 2002-03-04 at 07:57, Andreas Schwab wrote:

> Perhaps we should actually truncate here instead of rounding up.

I wonder way anything but rounding to nearest is used.  I've checked in
a patch to do along with Jakub's change.  Truncating is also possible
but the rules for stable arithmetic really suggest round to nearest.

-- 
---------------.                          ,-.   1325 Chesapeake Terrace
Ulrich Drepper  \    ,-------------------'   \  Sunnyvale, CA 94089 USA
Red Hat          `--' drepper at redhat.com   `------------------------

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 232 bytes --]