This is the mail archive of the
newlib@sourceware.org
mailing list for the newlib project.
Re: [PATCH] Move _wrapup_reent() to new file
- From: Joel Sherrill <joel dot sherrill at oarcorp dot com>
- To: newlib at sourceware dot org
- Date: Fri, 26 Apr 2013 09:52:46 -0500
- Subject: Re: [PATCH] Move _wrapup_reent() to new file
- References: <1366881156-9673-1-git-send-email-sebastian dot huber at embedded-brains dot de> <5179424D dot 8060107 at oarcorp dot com> <20130425152744 dot GG2875 at calimero dot vinschen dot de> <51794C88 dot 3070401 at oarcorp dot com> <20130425155828 dot GH2875 at calimero dot vinschen dot de>
On 4/25/2013 10:58 AM, Corinna Vinschen wrote:
On Apr 25 10:32, Joel Sherrill wrote:
On 4/25/2013 10:27 AM, Corinna Vinschen wrote:
On Apr 25 09:48, Joel Sherrill wrote:
I am OK with this. I was looking to see if the new file
should have some comment block at the top but
there is no notice in reent.c and the method isn't
public. So that doesn't leave much.
Corinna?
Given that _wrapup_reent seems to be unused, I wonder if we shouldn't
rather make an experiment: Just #if 0 it. If somebody complains
we can easily revert that. If not...
How long would you like to leave it disabled before killing it?
That's my only concern. Code gets disabled and never removed. I
can set a calendar alarm to remind me about this. But after how long?
If it's #if 0'd it doesn't hurt at all, even if we keep it for a long
time. If you add a comment, archaeologists will be able to find the
reason and still remove it any time. Apart from that, next newlib
release is in December. That's when we can remove it.
How about a patch like this which adds a warning comment to speak up
and an #if 0/#endif
Index: reent/reent.c
===================================================================
RCS file: /cvs/src/src/newlib/libc/reent/reent.c,v
retrieving revision 1.11
diff -u -r1.11 reent.c
--- reent/reent.c 23 Nov 2009 17:02:20 -0000 1.11
+++ reent/reent.c 26 Apr 2013 14:52:30 -0000
@@ -122,9 +122,15 @@
}
/*
+ * _wraup_reent is not prototyped in a .h. There is no known use.
+ * Disabling in anticipation of removal. If you need this routine,
+ * please speak up. --joel April 2013
+ */
+#if 0
+/*
* Do atexit() processing and cleanup
*
- * NOTE: This is to be executed at task exit. It does not tear anything
+ * NOTE: This is to be executed at task exit. It does not tear
anythingkkk
* down which is used on a global basis.
*/
@@ -148,4 +154,5 @@
if (ptr->__cleanup)
(*ptr->__cleanup) (ptr);
}
+#endif
Corinna
--
Joel Sherrill, Ph.D. Director of Research & Development
joel.sherrill@OARcorp.com On-Line Applications Research
Ask me about RTEMS: a free RTOS Huntsville AL 35805
Support Available (256) 722-9985