[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#632863: eglibc: add support for arch powerpcspe



On Fri, Jul 08, 2011 at 05:55:50PM +0200, David Kuehling wrote:
> >>>>> "Sebastian" == Sebastian Andrzej Siewior <bigeasy@linutronix.de> writes:
> 
> > severity 632863 wishlist thanks
> 
> > David Kuehling wrote:
> >> Severity: minor
> > I've been this is always wishlist because we are not in main archive.
> 
> > Please use
> 
> >  User: debian-powerpcspe@breakpoint.cc Usertags: powerpcspe
> 
> > So all ppcspe related bugs are together somehow.
> 
> Just sent a mail to control@bugs.debian.org that should fix it.
> 
> >> The attached patch makes eglibc 2.13 work on architecture powerpcspe
> >> (an unofficial port hosted on www.debian-ports.org, [1]).
> 
> > Thanks for that. If you have sometime then you could add two
> > files. One of them says which tests-cases are known to fail so if we
> > get another failure it aborts. The second file contains the symbol
> > list for libc. Both were available for ppc the last time I was looking
> > into it.
> 
> Attached below is an updated version of the patch.  

Thanks for the patch, it's now in the SVN (with minor changes, see
below), will be in the next upload.

>  - patches in debian/patches/powerpcspe are renamed to branch-* since
>  they were just committed to eglibc svn (branch 2.13, 2.14 and trunk).

Given these patches have been committed in the 2.13 branch, I have just
updated the svn-updates.diff patch instead.

>  - added file
>  testsuite-checking/expected-results-powerpc-linux-gnuspe-libc, tested
>  to pass here

Added.

>  - added debian/libc6.symbols.powerpcspe, which is currently just copied
>  from debian/libc6.symbols.powerpc

Added.

>  running 'dpkg-gensymbols -Pdebian/tmp-libc -plibc6 -v2.13' (output
>  attached below), I get a huge load of deltas involving 'symver' lines
>  like
> 
>     #MISSING: 2.13# (symver|optional)GCC_3.0 2.3.6
>     (symver|optional)GLIBC_2.1.2 2.1.2
> 
>   which looks non-critical, but I'm not sure why these aren't suppressed
>   by symbols.wildcards.  Filtering that noise, it shows something that
>   looks like permutations in symbol or library order, or just an
>   artefact of the "noise" lines that makes diff think that stuff lines
>   moved around, although joust noise lines were added in between.
> 
>   No symbols are missing, though, and no symbols have been added.  I
>   wonder whether it's the same on normal 'powerpc' builds.

This is something that has appeared with recent versions of dpkg-dev on
all architectures using symbol files, I haven't yet tracked why. Anyway
the generated symbol files are correct, so there is no need to worry
here.

Aurelien

-- 
Aurelien Jarno	                        GPG: 1024D/F1BCDB73
aurelien@aurel32.net                 http://www.aurel32.net



Reply to: