|
From: | Clemens Koller |
Subject: | Re: [ft-devel] Bug on PowerPC: Illegal Intruction in FT_Get_Name_Index (worksforme) |
Date: | Wed, 13 Apr 2005 19:33:04 +0200 |
User-agent: | Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.3) Gecko/20040910 |
Hi, Werner! I have tried to go for the latest CVS, but I guess I have some troubles to get through the autoconf scripts properly. I did a aclocal sh ./autogen.sh which runs fine. but ./configure --prefix=/usr complains about a missing AC_PROG_LIBTOOL: checking for gzsetparams in -lz... yes checking zlib.h usability... yes checking zlib.h presence... yes checking for zlib.h... yes ./configure: line 5051: AC_PROG_LIBTOOL: command not found configure: creating ./config.status config.status: creating unix-cc.mk config.status: creating unix-def.mk config.status: creating freetype-config config.status: creating freetype2.pc config.status: creating ftconfig.h whereis libtools tells me it's in: /usr/bin/libtool So, nothing special... Well... How do you generate your configure scripts from scratch? Step by step? I am working with the latest versions: pkgconfig-0.17.1 libtool-1.5.14 automake-1.9.5 autoconf-2.59 Is freetype-2.1.10 now officially released? Greets, Clemens Koller _______________________________ R&D Imaging Devices Anagramm GmbH Rupert-Mayer-Str. 45/1 81379 Muenchen Germany http://www.anagramm.de Phone: +49-89-741518-50 Fax: +49-89-741518-19 Werner LEMBERG wrote:
> Wed Mar 10 2004 Mike A. Harris <address@hidden> 2.1.7-3 - > Added -fno-strict-aliasing to CFLAGS and CXXFLAGS to try to fix > SEGV and SIGILL crashes in mkfontscale which have been traced > into freetype and seem to be caused by aliasing issues in > freetype macros (#118021) Okay, here is my latest buildlog: [...]Ah, now I remember! Sorry. Meanwhile, the configure script always adds -fno-strict-aliasing for gcc (regardless of the CFLAGS value).So, my conclusion is (read the bugzilla report) that there still might be a bug either in freetype or in gcc.It has been already fixed in the current CVS since 2004-11-16. With other words, the current CVS should work out of the box. Can you verify this, please?Why does it fail if I don't use -fno-strict-aliasing?Because the FreeType code does some tricks which fail if strict aliasing is active. Fixing this is *not* trivial (and probably not worth the trouble). Werner
[Prev in Thread] | Current Thread | [Next in Thread] |