Long standing ppc64 regression. Any workarounds?

6 msg	Fwd: MetaHTML and the GCC web site
2 msg	C++: operator new and disabled exceptions
3 msg	testsuite for libgomp and so on...
2 msg	GCC 4.2.2 RC2 (finally)
1 msg	context-sensitive keywords, revisited
1 msg	LLVM 2.1 Release
4 msg	Profile information - CFG
4 msg	Question on GGC
4 msg	[libobjc] Shouldn't 'version' take a long?
1 msg	Re: [Bug target/33479] SyncTest Intermittent fa...
3 msg	Require help with the backend in gcc
3 msg	Re: Poor pow() / floating point performance of ...
4 msg	$RANLIB not passed to libiberty
5 msg	missing optimization - don't compute return val...
4 msg	support tnat instruction on IA-64. error occurs...

Long standing ppc64 regression. Any workarounds?
\ Diego Novillo (25 Sep 2007)
. \ David Edelsohn (25 Sep 2007)
. \ Benjamin Kosnik (25 Sep 2007)
. . \ Diego Novillo (25 Sep 2007)

1 msg	Question about how to have an target specific t...
1 msg	getting operands of call arguments and return v...
6 msg	Q about assignment expansion
1 msg	gcc-4.1-20070924 is now available

Subject:	Long standing ppc64 regression. Any workarounds?
Group:	Gcc
From:	Diego Novillo
Date:	25 Sep 2007

I have not been able to get a clean libstdc++ build on ppc64 for more
than a month (since 2007-08-23). The failure is always the same.
While building libstdc++/system-error.cc, I get:

/home/dnovillo/perf/sbox/gcc/local.ppc64/src/libstdc++-v3/src/system_error.cc:67 :
error: std::system_category causes a section type conflict

Ben, that change seems to be yours. Any idea on how can I fix it? Or
at least a workaround?

Thanks.