This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: Pre-Release binutils 2.17; [PATCH] Request for inserting bug-fix prior to releasing 2.17.
- From: Björn Haase <bjoern dot m dot haase at web dot de>
- To: Nick Clifton <nickc at redhat dot com>
- Cc: binutils at sourceware dot org
- Date: Tue, 16 May 2006 08:50:05 +0200
- Subject: Re: Pre-Release binutils 2.17; [PATCH] Request for inserting bug-fix prior to releasing 2.17.
- References: <200605151305.49186.bjoern.m.haase@web.de> <44698F0A.9030802@redhat.com>
Nick Clifton wrote on Dienstag, 16. Mai 2006 10:36 :
> Hi Björn,
>
> > 2006-05-10 Bjoern Haase <bjoern.m.haase@web.de>
> >
> > * gas/config/tc-avr.h
> > TC_VALIDATE_FIX: handle case of immediate constant values.
>
> I have applied this patch to the mainline and 2.17 branch.
>
> A few small points:
>
> * Your ChangeLog entry is not correctly formatted. The filename
> should be relative to the location of the ChangeLog. So, since this
> ChangeLog is in the gas/ directory, the filename should be
> config/tc-avr.h not gas/config/tc-avr.h. Also the context for the
> change (in this case the TC_VALIDATE_FIX macro) should, if possible,
> be include on the same line and enclosed in parentheses. Finally
> the description of the change should be formatted as a proper
> English sentence, complete with a capital initial letter. Ideally
> the description should allow the reader to deduce the contents of
> the patch without actually seeing the changed code.
>
> Thus I reformatted your ChangeLog entry as:
>
> * config/tc-avr.h (TC_VALIDATE_FIX): Allow fixups for immediate
> constant values.
Understood. I will be reworking the change log entries also for the large
pending (avr6) patch. I'll wait for the remarks of the avr maintainers and
provide a possibly corrected patch and new change log entries.
> * Finally, you mentioned that without this fix the assembler could
> encounter a seg-fault. It would be very helpful if you could create
> a small gas testsuite case for this, so that we can be sure that
> this problem does not reappear in the future.
I'll have a look at how the testsuite operates and plan to add the test to the
avr6 patch so that there is less work for those having CVS write permissions.
Thank's for the fast reply.
Bjoern.