Angel Tsankov wrote: > As far as I get it, the GCC's fixincludes script should not be run > during the second and third passes of GCC. Currently, the LFS book > suggests that the execution of finxincludes be suppressed by applying > a sed substition similar to this one: > > sed -i 's...@\./fixinc\...@-c true@' gcc/Makefile.in > > On the other hand, the gcc/Makefile.in file says that the execution of > fxincludes can be controlled via the STMP_FIXINC variable. Isn't > this a better way to suppress fixincludes? > Well, it might be a more complete way to suppress the execution of fixincludes via the STMP_FIXINC variable, but then GCC won't find some headers. Changing 'STMP_FIXINC = @STMP_FIXINC@' in gcc/Makefile.in to 'STMP_FIXINC =' leads to the following error during compilation of GCC: 'no include path in which to search for limits.h'
--Angel -- http://linuxfromscratch.org/mailman/listinfo/lfs-chat FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
