On 02/24/15 15:56, Jakub Jelinek wrote:
On Tue, Feb 24, 2015 at 10:21:23PM +0100, Jan Hubicka wrote:
On Tue, Feb 24, 2015 at 08:48:19PM +0100, Jan Hubicka wrote:
Thanks, the i386 parts of the patch are OK, but I think you want to add the
reverse
transformation, too. I.e. if someone compiles with -fPIC but links without.
I've only done it this way because that is what
ix86_option_override_internal was doing, but supposedly only because the
command line option is only about the non-PIC variants.
So I agree that the other direction makes sense too and will adjust it.
My plan to fix the testcase was to put it into ix86_function_specific_restore
which would save need for a new hook. But I am fine either way (just can't
approve the newhook)
The way the streaming in now works is that we don't have a gcc_options
structure anywhere, so if it was done in the *_restore hook, you'd need
to *_save it first and then restore.
I see, we call cl_target_option_restore but not the hook. Oh well, then.
No, we call cl_target_option_stream_in and the new hook is the target
specific overrider of that function.
BTW, the -fpic -> -fno-pic change passed bootstrap/regtest on x86_64-linux
and i686-linux too, here is the updated patch:
2015-02-24 Jakub Jelinek <ja...@redhat.com>
PR lto/64374
* target.def (target_option_stream_in): New target hook.
* tree-streamer-in.c (streamer_read_tree_bitfields): Invoke
targetm.target_option.post_stream_in if non-NULL.
* doc/tm.texi.in: Add @hook TARGET_OPTION_POST_STREAM_IN.
* doc/tm.texi: Updated.
* config/i386/i386.c (ix86_function_specific_post_stream_in): New
function.
(TARGET_OPTION_POST_STREAM_IN): Redefine.
OK.
jeff
