Re: [rfa/rfc] Build libcommon.a for gdb and gdbserver

[Yao Qi <yao at codesourcery dot com>]

On 03/01/2011 12:41 PM, Joel Brobecker wrote:
>> Yao> Personally, I still prefer a separated configure/makefile in common/,
>> Yao> because,
>> Yao> 1.  if my patch works, configure/make is not a problem,
>> Yao> 2.  if we look forward, there should be quite a few *.c and *.h files in
>> Yao> common in the future.  Write rules in both gdb/Makefile.in and
>> Yao> gdbserver/Makefile.in doesn't scale.
>>
>> I think the most important thing is that if you want to keep the
>> common/configure stuff, then please fix the existing problems that have
>> been reported.  Maybe it is just the GNU make-ism at this point, I
>> haven't kept track.
> 
> On my end of things, I actually do not like the multiple layers of
> miniature configure scripts. These things just keep doing the same
> checks over and over for the most part.  It's particularly visible
> on Windows, were configure takes ages because spawning a script is
> utterly inefficient.  (and as such, Tom's proposal to remove the
> configure script in gdb/testsuite seems like a good idea to me too)
> 
> That being said, as long as it works, it's not of uber importance
> to me but I am not certain that argument number 2 above from Yao
> really is that much work. So, if we fix things fast, I do not mind
> continuing with the present approach. (does anyone know what the
> remaining issues are, though?)
> 

AFAIK, there are three problems,

1. Build failure on mingw32, reported by Pierre. See
<http://sourceware.org/ml/gdb-patches/2011-02/msg00285.html>
I'll give a patch to fix this.

2. make clean doesn't run in gdbserver/common, reported by Michael.
http://sourceware.org/ml/gdb/2011-02/msg00126.html
Patch was sent to the wrong place.  I'll send it again to gdb-patches@.

3.  GNU make feature,
http://sourceware.org/ml/gdb-patches/2011-02/msg00489.html
Since we decide not to require GNU make, I'll remove its GNU make feature.

Am I missing any problems else?

-- 
Yao (éå)