Update gcc-version rsb for 4.11

Chris Johns chrisj at rtems.org
Sat Jan 16 00:34:41 UTC 2016

On 16/01/2016 4:57 AM, Jan Sommer wrote:
> Am Friday 15 January 2016, 11:24:00 schrieb Marcos Díaz:
>> Hi,
>> You can try if the latest snapshot of gcc works:
>> Check this patch, in this case I made a modification for the arm bset, if
>> you use another architecture you have to make the same change in your bset.
> Thank you, that seems to be a better way (I just overwrote the gcc-tarball of rsb with my own).
> However, I meant to suggest that we such a patch upstream in general. The current situation is that you can build a toolchain with Ada-support using the rsb, it will compile, but the programs it creates will fail at least for arm, but I think the error is architecture agnostic. So you need to find out why it fails, apply a patch like yours locally and rebuild which makes the process quite painful.
> I don't know if the policy says, it has to be a stable gcc-release, but the gcc-4.9 branch only accepts regression and docu fixes.
> @Joel+Chris: What do you think?

Is there a ticket in Trac for 4.11 tracking this issue? We need tickets
for anything destined for 4.11.

I would prefer nothing changes for 4.11.0. This change can then be
reviewed and tested on the 4.11-branch and if ok we can look at 4.11.1.
This is not reflection on the importance of the change, it is a
reflection on me having too many moving parts as I attempt to get 4.11.0
released. My hope is getting 4.11.0 out will smooth the path for future

> I am aware that Ada is the ugly little stepchild no one really wants to play with, but would still appreciate an easy way to build a working toolchain. ;-)

I do not see Ada as this. We will support what we can and what is
practical. If members of the community are willing to support Ada, to
test it and to contribute patches then we are willing to accept them. If
are willing to look after Ada, monitor it and test it we would welcome
your efforts.


More information about the devel mailing list