I modified patch accordingly to Alexander comments. Is it OK for trunk?
2015-02-25 15:38 GMT+03:00 Alexander Monakov <amona...@ispras.ru>: > > > On Wed, 25 Feb 2015, Yuri Rumyantsev wrote: > >> Here is updated patch accordingly to Alexander comments. >> >> BTW another function using HID interface is do_reorder_for_imul and it >> is called from ix86_sched_reorder. > > do_reorder_for_imul uses dependency list iteration macros, which use HDID, not > HID, and HDID is populated during selective scheduling. How exactly does > do_reorder_for_imul access HID? > > Alexander
patch.3
Description: Binary data
