This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Patch to fix delete_noop_move REG_LIBCALL handling.
- To: rth at redhat dot com
- Subject: Re: Patch to fix delete_noop_move REG_LIBCALL handling.
- From: John Wehle <john at feith dot com>
- Date: Wed, 10 Oct 2001 00:31:06 -0400 (EDT)
- Cc: gcc-patches at gcc dot gnu dot org
>On Sun, Oct 07, 2001 at 05:16:47PM -0400, John Wehle wrote:
>> ! /* Handle the last insn of a libcall in a similar fashion. */
>> ! else if ((note = find_reg_note (insn, REG_RETVAL, NULL_RTX))
>> ! && XEXP (note, 0) != insn)
>> ! {
>
>This is incorrect -- the REG_RETVAL insn in special in that
>it may be a noop move *explictly* in order to hide some real
>operation that comes before.
Isn't there a REG_EQUAL note in this event? noop_move_p
returns false if a REG_EQUAL note is present.
Are there useful reasons to keep a noop move that has a REG_RETVAL
note even if there's no REG_EQUAL note? Or, are you saying that
this will never happen so there's no need to handle it?
-- John
-------------------------------------------------------------------------
| Feith Systems | Voice: 1-215-646-8000 | Email: john@feith.com |
| John Wehle | Fax: 1-215-540-5495 | |
-------------------------------------------------------------------------