pkgsrc-Bugs archive

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

Re: pkg/47543: graphics/cairo on DragonFly i386 Define CAIRO_NO_MUTEX to 1

The following reply was made to PR pkg/47543; it has been noted by GNATS.

From: David Shao <>
Subject: Re: pkg/47543: graphics/cairo on DragonFly i386 Define CAIRO_NO_MUTEX
 to 1
Date: Sat, 23 Feb 2013 10:57:46 -0800

 Note the following is strictly my personal opinion and some if not
 most may be disputed by the DragonFly development team.
 The fix from
 commit 4a0e0cd7377c45dc21d5c71bcab931b5e5727e74
 Date:   Mon Feb 18 01:15:59 2013 +0100
     csu: Fix .eh_frame_hdr errors seen on i386
 corrects cairo's building on DragonFly i386 (I don't think there was a
 problem on x86_64) using the gcc 4.7.2 compiler.
 There therefore seem to me to be two viable options for now on
 DragonFly with regards to pkgsrc, but one will over time become much
 less viable:
 1) Stick with using gcc 4.4.  This has the advantage that modular xorg
 still starts using startx.  The problem I have seen appears to be in
 compiling the combination of libdrm and modular-xorg-server with gcc
 4.7.2 on DragonFly, which builds, but just does not start instead
 leaving the dreaded black screen of nothingness.  One way to stick
 with gcc 4.4, but note that it is getting more painful as time
 progresses, is to in
 insert the lines
 I don't know that OBJFORMAT_BUILTIN_CCVER_DEFAULT is necessary, only
 if one reads the patch that set the default compiler to gcc 4.7, this
 variable determines the object format which seems an important thing.
 Sticking with gcc 4.4 has the disadvantage that devel/xulrunner and I
 suppose www/firefox have to be manually built with gcc 4.7:
 # env CCVER=gcc47 bmake replace clean
 2) Accept using gcc 4.7.  Just remember to manually rebuild libdrm and
 modular-org-server (I don't know if both are needed, I just gave up
 after determining rebuilding those two were enough).

Home | Main Index | Thread Index | Old Index