On Tue, 2013-01-29 at 12:29 +0100, Samuel Thibault wrote: > Svante Signell, le Tue 29 Jan 2013 10:09:52 +0100, a écrit : > > On Tue, 2013-01-29 at 07:36 +0100, Svante Signell wrote: > > > On Tue, 2013-01-29 at 00:13 +0100, Samuel Thibault wrote: > > > > Svante Signell, le Sun 20 Jan 2013 18:18:16 +0100, a écrit : > > > > > A recent problem popping up is that executables using the -pie flag > > > > > when > > > > > linking segfaults when starting up. Examples are mktable in w3m > > > > > > > > It doesn't seem so simple. I've tried building a small program with > > > > -pie, and it does work. I've tried to use w3m, it does work. elinks does > > > > indeed crash. > > > > No problems with mktable in w3m? > > What do you mean by mktable in w3m?
When building w35-0.5.3-8 with -pie option: ./mktable 100 functable.tab > functable.c /bin/bash: line 1: 6559 Segmentation fault ./mktable 100 functable.tab > functable.c When disabling -pie with export DEB_BUILD_MAINT_OPTIONS = hardening=+all,-pie in debian/rules the build is OK. > > > I know in which module elinks crashes. Disabling it makes things work > > > again. Can this be of interest? > > > > The problems with elinks is when initializing the perl_scripting_module > > in src/scripting/scripting.c defined in src/scripting/perl/perl.c > > Ok, I'd avoid trying to debug that (perl, brrr), and rather debug w3m, > which is much more simple. It loads a lot of stuff when initializing...
