This is the mail archive of the
mailing list for the elfutils project.
Re: [commit] [patch] readelf.c: Use real output width [Re: [patch] Implement reglocs for s390/s390x]
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: elfutils-devel at lists dot fedorahosted dot org
- Date: Sat, 13 Oct 2012 21:14:17 +0200
- Subject: Re: [commit] [patch] readelf.c: Use real output width [Re: [patch] Implement reglocs for s390/s390x]
On Fri, 12 Oct 2012 20:56:32 +0200, Jan Kratochvil wrote:
> On Fri, 12 Oct 2012 20:11:01 +0200, Roland McGrath wrote:
> > Well, I think there should still be a comment somewhere saying the intent
> > is to avoid having the wrapping depend on the values.
> + Line breaks should not depend on the particular values although that
> + may happen in some cases of the core items. */
> - But we do not want the line breaks to depend on
> - the particular values. */
pro: It never changed line breaks
con: It could print unusual values very wide, up to column 150.
pro: It will always fit in 75 columns.
con: Unusual values may change line breaks.
The problem was that S390_HIGH_GPRS are those "unusual values" which got
printed very wide.
So if your the most preferred value is to never change line breaks this
check-in should be reverted and we should check-in the previous one:
+ unsigned int wrap = (item->group && strcmp (item->group, "register") == 0
+ ? REGISTER_WRAP_COLUMN : ITEM_WRAP_COLUMN);
But you said for that one:
On Wed, 10 Oct 2012 20:47:13 +0200, Roland McGrath wrote:
> The special case on group name is bletcherous.