pkgsrc-Bugs archive

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

Re: pkg/53319: text relocations fixes for multimedia/mplayer others



The following reply was made to PR pkg/53319; it has been noted by GNATS.

From: Onno van der Linden <o.vd.linden%quicknet.nl@localhost>
To: Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>
Cc: gnats-bugs%netbsd.org@localhost
Subject: Re: pkg/53319: text relocations fixes for multimedia/mplayer others
Date: Mon, 13 Aug 2018 23:19:51 +0200

 On Tue, Aug 14, 2018 at 05:50:52AM +0900, Izumi Tsutsui wrote:
 > > So, the question is why I don't have any problem running these
 > > programs with asm *enabled* on -current ? Could be paxctl
 > > and FORTIFY_SOURCE related, but why am I not seeing the
 > > error messages you were seeing with the programs that were build
 > > using the i386 assemebler code ? Were the 8.0 binaries build
 > > with stricter options ?
 > 
 > How did you enable x86 asm in ffmpeg3 from pkgsrc-2018Q2?
 
 Compiled and installed ffmpeg3 on Mar 19 2018 using the
 -current pkgsrc from that timestamp with a -current system
 from that timestamp  when the noasm option wasn't
 around in the Makefile.
 
 > 
 > mplayer doesn't start if it's enabled as the following:
 
 It does for me, with -current.
 
 However, I noticed your mplayer is compiled with gcc 5.5.0.
 Mine is compiled with gcc 6.4.0. No idea if that's the
 reason for you getting the "Cannot write-enable text segment"
 error messages.
 
 > Anyway, there are strong complaints about text relocations:
 >  http://mail-index.netbsd.org/pkgsrc-bugs/2018/04/28/msg063868.html
 >  http://mail-index.netbsd.org/pkgsrc-bugs/2018/04/28/msg063878.html
 >  http://mail-index.netbsd.org/pkgsrc-bugs/2018/04/28/msg063881.html
 > 
 > That was the reason why I filed this PR to disable them completely.
 
 If disabling i386 assembler code fixes the
 "Cannot write-enable text segment: Permission denied" errors, fine.
 But, I have yet to see any such message and the text reloction
 messages don't bother me that much. I'd love to see a more subtle
 way to fix this.
 
 Question still remains why am I only getting the text relocation
 warning messages and no error message whatsoever ? Is it due
 to gcc 6.4 vs gcc 5.5 and its handling of FORTIFY_SOURCE ?
 
 Regards,
 
 Onno
 


Home | Main Index | Thread Index | Old Index