2014-10-10 13:30 GMT+02:00 cg :

How can I configure to build x86_64?

When I build GHC (with msys2), it always builds i386 and I haven't spotted the option in ./configure to choose a x86_64 release.

This is implicitly determined by the toolchain you use. So, probably you have the i686 msys2 installed, while you would need the x86_64 version. Given, that your operating system (and thus your hardware) is also x86_64.

Indeed, but this looks like completely unrelated to the issue originally reported. Kind of libffi misdetection of target platform? i.e. why it compiles win32 related file on macosx? Just trying to categorize not to decrease importance of this issue! Karel On 10/10/14 10:47 PM, Carter Schonwald wrote:

likewise, 32bit OS X seems to be broken on HEAD too

http://lpaste.net/112412 is the relevant bit

make[5]: Nothing to be done for `all'. depbase=`echo src/x86/win32.lo | sed 's|[^/]*$|.deps/&|;s|\.lo$||'`;\ /bin/sh ./libtool --mode=compile gcc-4.9 -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -I. -I../include -Iinclude -I../src -U__i686 -m32 -fno-stack-protector -w -MT src/x86/win32.lo -MMD -MP -MF $depbase.Tpo -c -o src/x86/win32.lo ../src/x86/win32.S&&\ mv -f $depbase.Tpo $depbase.Plo libtool: compile: gcc-4.9 -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -I. -I../include -Iinclude -I../src -U__i686 -m32 -fno-stack-protector -w -MT src/x86/win32.lo -MMD -MP -MF src/x86/.deps/win32.Tpo -c ../src/x86/win32.S -fno-common -DPIC -o src/x86/.libs/win32.o ../src/x86/win32.S:1283:section difference relocatable subtraction expression, ".LFE5" minus ".LFB5" using a symbol at the end of section will not produce an assembly time constant ../src/x86/win32.S:1283:use a symbol with a constant value created with an assignment instead of the expression, L_const_sym = .LFE5 - .LFB5 ../src/x86/win32.S:1275:section difference relocatable subtraction expression, ".LEFDE5" minus ".LASFDE5" using a symbol at the end of section will not produce an assembly time constant ../src/x86/win32.S:1275:use a symbol with a constant value created with an assignment instead of the expression, L_const_sym = .LEFDE5 - .LASFDE5 ../src/x86/win32.S:unknown:missing indirect symbols for section (__IMPORT,__jump_table) make[5]: *** [src/x86/win32.lo] Error 1

On Fri, Oct 10, 2014 at 9:40 AM, Páli Gábor János mailto:pali.gabor@gmail.com> wrote:

2014-10-10 13:30 GMT+02:00 cg mailto:chengang31@gmail.com>: > How can I configure to build x86_64? > > When I build GHC (with msys2), it always builds i386 and I haven't spotted > the option in ./configure to choose a x86_64 release.

This is implicitly determined by the toolchain you use. So, probably you have the i686 msys2 installed, while you would need the x86_64 version. Given, that your operating system (and thus your hardware) is also x86_64. _______________________________________________ ghc-devs mailing list ghc-devs@haskell.org mailto:ghc-devs@haskell.org http://www.haskell.org/mailman/listinfo/ghc-devs

_______________________________________________ ghc-devs mailing list ghc-devs@haskell.org http://www.haskell.org/mailman/listinfo/ghc-devs

I believe I've just pushed a fix for this, let me know if you still have problems. Cheers, Simon On 10/10/2014 13:47, Carter Schonwald wrote:

likewise, 32bit OS X seems to be broken on HEAD too

http://lpaste.net/112412 is the relevant bit

make[5]: Nothing to be done for `all'. depbase=`echo src/x86/win32.lo | sed 's|[^/]*$|.deps/&|;s|\.lo$||'`;\ /bin/sh ./libtool --mode=compile gcc-4.9 -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -I. -I../include -Iinclude -I../src -U__i686 -m32 -fno-stack-protector -w -MT src/x86/win32.lo -MMD -MP -MF $depbase.Tpo -c -o src/x86/win32.lo ../src/x86/win32.S &&\ mv -f $depbase.Tpo $depbase.Plo libtool: compile: gcc-4.9 -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -I. -I../include -Iinclude -I../src -U__i686 -m32 -fno-stack-protector -w -MT src/x86/win32.lo -MMD -MP -MF src/x86/.deps/win32.Tpo -c ../src/x86/win32.S -fno-common -DPIC -o src/x86/.libs/win32.o ../src/x86/win32.S:1283:section difference relocatable subtraction expression, ".LFE5" minus ".LFB5" using a symbol at the end of section will not produce an assembly time constant ../src/x86/win32.S:1283:use a symbol with a constant value created with an assignment instead of the expression, L_const_sym = .LFE5 - .LFB5 ../src/x86/win32.S:1275:section difference relocatable subtraction expression, ".LEFDE5" minus ".LASFDE5" using a symbol at the end of section will not produce an assembly time constant ../src/x86/win32.S:1275:use a symbol with a constant value created with an assignment instead of the expression, L_const_sym = .LEFDE5 - .LASFDE5 ../src/x86/win32.S:unknown:missing indirect symbols for section (__IMPORT,__jump_table) make[5]: *** [src/x86/win32.lo] Error 1

On Fri, Oct 10, 2014 at 9:40 AM, Páli Gábor János mailto:pali.gabor@gmail.com> wrote:

2014-10-10 13:30 GMT+02:00 cg mailto:chengang31@gmail.com>: > How can I configure to build x86_64? > > When I build GHC (with msys2), it always builds i386 and I haven't spotted > the option in ./configure to choose a x86_64 release.

This is implicitly determined by the toolchain you use. So, probably you have the i686 msys2 installed, while you would need the x86_64 version. Given, that your operating system (and thus your hardware) is also x86_64. _______________________________________________ ghc-devs mailing list ghc-devs@haskell.org mailto:ghc-devs@haskell.org http://www.haskell.org/mailman/listinfo/ghc-devs

_______________________________________________ ghc-devs mailing list ghc-devs@haskell.org http://www.haskell.org/mailman/listinfo/ghc-devs

Sorry about this folks, I'm going to guess at the fix and validate it. Cheers, Simon On 10/10/2014 10:51, Karel Gardas wrote:

On 10/10/14 09:19 AM, Páli Gábor János wrote:

...

Hello there,

Looks one of the recent commits broke the x86 builds on multiple platforms [1][2][3]. The common error message basically is as follows:

rts/Linker.c: In function 'mkOc': rts/Linker.c:2372:6: error: 'ObjectCode' has no member named 'symbol_extras'

It looks like this patch is the culprit:

5300099ed rts/Linker.c (Simon Marlow 2014-10-01 13:15:05 +0100 2372) oc->symbol_extras = NULL;

...

Note that the x86_64 counterparts of the affected builders completed their builds fine -- except for Solaris, but that has been broken already for a long time (relatively).

Yeah, I'll need to fix that. So far I invested my energy in fixing Solaris/i386 testcase issues...

Karel