Hi! On Tue, 12 May 2015 19:05:17 +0200, Bernd Schmidt wrote: > > @@ -4266,7 +4266,7 @@ process_command (unsigned int decoded_op > > } > > > > gcc_assert (!IS_ABSOLUTE_PATH (tooldir_base_prefix)); > > - tooldir_prefix2 = concat (tooldir_base_prefix, spec_host_machine, > > + tooldir_prefix2 = concat (tooldir_base_prefix, spec_machine, > > dir_separator_str, NULL); > > > > /* Look for tools relative to the location from which the driver is > > This one, definitely. Indeed; now committed in r228429. (Jim, did this missing patch cause the thousands of FAILs you've seen? The problem was that nvptx offloading found the wrong assembler, and we all know what happens when you try to process PTX assembly with an x86/PowerPC assembler...) commit f2fd5997e0856128c0609cdc6aa4cf9867f94c41 Author: tschwinge Date: Fri Oct 2 21:24:50 2015 +0000 Help the offload gcc driver find the right assembler gcc/ * gcc.c (process_command): Use spec_machine rather than spec_host_machine to build tooldir_prefix2. git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@228429 138bc75d-0d04-0410-961f-82ee72b054a4 --- gcc/ChangeLog | 5 +++++ gcc/gcc.c | 2 +- 2 files changed, 6 insertions(+), 1 deletion(-) diff --git gcc/ChangeLog gcc/ChangeLog index d1235bd..7fe0196 100644 --- gcc/ChangeLog +++ gcc/ChangeLog @@ -1,3 +1,8 @@ +2015-10-02 Bernd Schmidt + + * gcc.c (process_command): Use spec_machine rather than + spec_host_machine to build tooldir_prefix2. + 2015-10-02 Thomas Schwinge Bernd Schmidt diff --git gcc/gcc.c gcc/gcc.c index db90796..7f5a36e 100644 --- gcc/gcc.c +++ gcc/gcc.c @@ -4472,7 +4472,7 @@ process_command (unsigned int decoded_options_count, } gcc_assert (!IS_ABSOLUTE_PATH (tooldir_base_prefix)); - tooldir_prefix2 = concat (tooldir_base_prefix, spec_host_machine, + tooldir_prefix2 = concat (tooldir_base_prefix, spec_machine, dir_separator_str, NULL); /* Look for tools relative to the location from which the driver is Grüße, Thomas