OS X 10.5 build failure

Judah Jacobson judah.jacobson at gmail.com
Wed Jul 30 13:48:45 EDT 2008


On Tue, Jul 29, 2008 at 10:06 PM, Roman Leshchinskiy <rl at cse.unsw.edu.au> wrote:
> With the latest HEAD, I get this:
>
> ar: creating archive dist-stage2/build/libHSghc-6.9.a
> ld: atom sorting error for _ghczm6zi9_LibFFI_Czuffizucif_closure_tbl and
> _ghczm6zi9_LibFFI_Czuffizutype_closure_tbl in dist-stage2/build/LibFFI.o
> ld: atom sorting error for _ghczm6zi9_LibFFI_Czuffizucif_closure_tbl and
> _ghczm6zi9_LibFFI_Czuffizutype_closure_tbl in dist-stage2/build/LibFFI.o
> ranlib: file: dist-stage2/build/libHSghc-6.9.a(unicode.o) has no symbols
> ld: scattered reloc r_address too large for inferred architecture i386
> make[3]: *** [dist-stage2/build/HSghc-6.9.o] Error 1
>

I managed to track this bug down, only to find that it has already
been fixed by the following patch:

Wed Jul 30 03:52:03 PDT 2008  Simon Marlow <marlowsd at gmail.com>
 * oops, fix a small pessimisation made in previous refactoring

For reference, the original patch that caused this problem was the following:

Tue Jul 22 02:21:13 PDT 2008  Simon Marlow <marlowsd at gmail.com>
  * refactoring/tidyup: (not.is64BitInteger) -> is32BitInteger



-Judah



More information about the Cvs-ghc mailing list