I've build it a month ago with the CVS version of GCC against libc 2.3.2
(LFS), offcourse the main problem was the boottrapping which i did on a
system with gcc 2.95 and libc 2.3.1 (Slackware) with that executable i
could actually link the stuff what 3.3-cvs produced although i didn't got
ObjectIO to work.
Greetings,
Stefan de Konink
On 19 Oct 2003, Peter Simons wrote:
> Hi,
>
> has anyone successfully built the current GHC sources using GCC 3.3.1?
> I'm asking because I'm having lot's of problems doing it. The version
> I use here is the "hardened-gcc" package from Gentoo Linux [1]
>
> gcc (GCC) 3.3.1 20030927 (Gentoo Linux 3.3.1-r5, propolice)
>
> which adds transparent support for PaX/etdyn ASLR and Propolice/SSP
> stack protection to every binary.
>
> When compiling with -O, chokes on the stack-securing start-up code:
>
> | ghc -cpp -O -funbox-strict-fields -package util -package text -fffi \
> | -c Autoconf.lhs
> | Warning: retaining unknown function `__i686.get_pc_thunk.bx' in \
> | output from C compiler
> |
> | Prologue junk?: .globl __stginit_Autoconf
> | .type __stginit_Autoconf, @function
> | __stginit_Autoconf:
> | movl %ebx, 8(%esp)
> | call __i686.get_pc_thunk.bx
> | addl $_GLOBAL_OFFSET_TABLE_, %ebx
>
> Without -O, everything works fine (and the code seems to work as
> well). Now, I "fixed" the problem by disabling the special stack
> protection features by adding
>
> -optc-yet_exec -optc-yno_propolice
>
> to the command line. With these options, GHC compiles everything as
> usual, no problems.
>
> So I tried to compile the latest GHC itself:
>
> | GCC="gcc -yet_exec -yno_propolice ${CFLAGS}"
> | GHC="ghc-6.0 -optc-yet_exec -optc-yno_propolice"
> | autoconf &&
> | autoheader &&
> | ./configure --prefix=/usr/local/ghc-current \
> | --with-gcc="${GCC}" \
> | --with-ghc="${GHC}" \
> | --enable-hopengl \
> | --enable-threaded-rts \
> | --enable-src-tree-happy \
> | --enable-src-tree-haddock \
> | --enable-src-tree-alex
>
> The build completes the first stage successfully, but apparently does
> NOT produce a valid executable: The generated ghc-inplace binary
> doesn't do anything. It doesn't fail nor produce any error message,
> but it doesn't produce any output either. I can see no sign it was
> alive neither on the console nor in the system's log file.
>
> Since stack protection code _is_ disabled, I figured this could be a
> problem with gcc itself.
>
> Any ideas, anyone?
>
> Peter
>
>
>
> [1] http://www.gentoo.org/proj/en/hardened/etdyn-ssp.xml
> _______________________________________________
> Cvs-ghc mailing list
> [EMAIL PROTECTED]
> http://www.haskell.org/mailman/listinfo/cvs-ghc
>
_______________________________________________
Cvs-ghc mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/cvs-ghc
