Fwd: [ANNOUNCEMENT] GDB 7.4 release process created!

[Ralf Corsepius]

On 12/13/2011 03:31 PM, Ralf Corsepius wrote:
> On 12/13/2011 03:13 PM, Joel Sherrill wrote:
>> Hi,
>>
>> I am sure Ralf saw this but wanted to pass it along
>> anyway.
>
> Yes, saw it.
>
>> Do the RPMs need to spin for the 7.3.91 tarball?
> It's at least my plan ...
>
> ... but packaging-wise gdb is the most complex package we have ...

Temporary results:
* gdb-7.3.91 installs another file, which is not useful for us and 
likely is not useful in general.

* Building gdb-7.3.91 with system-readline enabled seems to require 
readline > 5. It fails to build on systems with readline5 only 
(Currently CentOS5 and Cygwin). Likely, I will have to resort to enable 
building against the bundled readline on affected systems
(== regression).


Independent of gdb-7.3.91:
* gdb's python support and python's versioning is becoming an increasing 
problem. It currently breaks re-using CentOS packages on "fairly new, 
but discontinued distros" and switching to "supporting one single distro 
for older RTEMS" (like we currently do for rtems-4.9).

If all else fail, I will likely can not avoid having to resort to 
drastic measures, such as
* to disable python in gdb, at least on CentOS (This would remove the 
harmful dependencies)
* to stop trying to reutilize older distros and shutdown repositories 
hard - This would mean to abandon support of older RTEMS toolchains on 
new distros and would mean to abandon providing "legacy packages" on new 
distros.

IMO, both steps are hardly acceptable and should be avoided - However, I 
currently see no solution.

>> Do we have patches that can/should be pushed?
> I doubt it, nevertheless, to be checked.
Check the patch in CVS. It's current and mostly identical to what we had 
in gdb-7.2 ... same remarks as before apply.

> Sorry for being brief ... "end-of-year" lack of time.
Sorry, these "end of year" "release rushes" do not match with my time 
constraints. No time to take care about this around this time of year.

Ralf