Re: c++/3082

[Loren James Rittle <rittle at latour dot rsch dot comm dot mot dot com>]

Hi Nathan,

In article <3B20D206.C5FCA6F3@codesourcery.com>,
Nathan Sidwell <nathan@codesourcery.com> writes:

> http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3082&database=gcc
> I agreee, this should be high priority.

>> Patch posted to libstdc++-v3, gcc-patches and installed on mainline
> are you refering to http://gcc.gnu.org/ml/gcc-patches/2001-06/msg00325.html

Sorry about not clarifying which patch set exactly.  Yes, that patch
along with:

http://gcc.gnu.org/ml/gcc-patches/2001-06/msg00330.html

http://gcc.gnu.org/ml/gcc-patches/2001-06/msg00400.html (if you want
alpha/ELF working as best as possible with the other patches, but it
is not strictly required)

Additional commentary:

http://gcc.gnu.org/ml/gcc-patches/2001-06/msg00327.html

> or something else? Have you booted that patch on a 3.0 prerelease?
> Please provide cluons :-)

From that e-mail:

i386-unknown-freebsd4.2 (many earlier versions on mainline)
i386-unknown-freebsd4.2 (this version on 3.0 branch)
alpha-unknown-freebsd4.2 (this version on 3.0 branch)
i386-*-linux* (very close to this version on 3.0 branch)
hppa-*-hpux10* (very close to this version on 3.0 branch)
hppa-*-hpux11* (very close to this version on 3.0 branch)
sparc-sun-solaris2.7 (this version on 3.0 branch however
 I don't usually build this target and every C++ test fails with:
 ld.so.1: [...]: fatal: relocation error: R_SPARC_32: file [...]/libstdc++.so.3:
 symbol __gxx_personality_v0:  offset 0xff31cbe2 is non-aligned ---
 This could be because I have an old/bad binutils; I used
 --with-gnu-as and --with-gnu-ld or whatever.  I have checked at the
 source level that things are being setup properly for threads on this
 port and I think the failure is independent of this patch but I have
 not verified that yet.)