Re: Commits to trunk and branch: maintainer's role?

[Andrew Cagney <ac131313@cygnus.com>]

> Just committing a change to cris-tdep.c on the 5.1 branch, I noticed
> that a couple of changes on the trunk by other people after the 5.1
> branch was cut never made it to the branch.  Being new in the target
> maintainer role, I'm wondering if those changes should have been made to
> the branch also (if applicable for the branch of course), or if that is
> the maintainer's role.

For changes like the below, it would be up to you.  These changes 
weren't to fix a demonstrated bug but rather to address a potential 
coding problem.  In the case of the first, I noticed a -Werror problem, 
the second was part of multi-arching TARGET_PRINT_INSN.

As they say, if it ain't broke, don't fix it :-)

Andrew


> More specifically, these are the changes I'm thinking about.  At least
> the first one seems branch material to me, the second I'm not sure of
> since it deals with multi-arching.
> 
> revision 1.3
> date: 2001/09/19 21:59:41;  author: cagney;  state: Exp;  lines: +2 -2
> * cris-tdep.c (cris_get_signed_offset): Change return type to an
> explicitly signed char.

This
> revision 1.2
> date: 2001/09/05 23:44:43;  author: ezannoni;  state: Exp;  lines: +1 -1
> 2001-09-05  Elena Zannoni  <ezannoni@redhat.com>
> 
> [snip]
>         * cris-tdep.c (cris_delayed_get_disassembler): Use
>         TARGET_PRINT_INSN, instead of tm_print_insn.
> 
> 
> I'm grateful that other people fix these things, I'm just wondering
> about the proper procedure for committing when there's an active branch.
> 
>