--=-qyDqNhhe90WQS2HGpF6j
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Tue, 2005-12-13 at 01:24 +0100, Johnny Billquist wrote:
>=20
> The specific problem you hit upon is that when the tools build comes to=20
> groff, it will fail late in the build, when creating some document,=20
> because grn segfaults.

Funnily enough I've seen a similar problem with a friend of mine
cross-building for sparc on a Fedora Core 4 Linux box.

> If I were to guess that is because some code generated by gcc is wrong=20
> when using -O2. Nobody ever seems to try to find the problem, but it has=20
> been verified that using -O0 in those few cases where it happens will wor=
k.
> I found it while building one file in kermit as well.
> It's rare, but it do happen occasionally.

This particular problem seems to get caught by the version of glibc on
Fedora which has some malloc() checking. It detects a "double free or
corruption" and aborts. Disabling/ignoring the checks allows the build
to continue in this case.

IIRC this was building on the netbsd-2 branch.

Matt

--=-qyDqNhhe90WQS2HGpF6j
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQBDnh4aKP58eR+X2TMRAtTmAKCVG4xfkw/7gQty0yptliNq9sWcdgCgj4Yn
BpgAJXj8UebtvpjivMvcCas=
=Iybi
-----END PGP SIGNATURE-----

--=-qyDqNhhe90WQS2HGpF6j--