From patchwork Sat Oct 13 09:33:21 2012
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: [0/6] Thread pointer built-in functions / [SH] PR 54760
Date: Fri, 12 Oct 2012 23:33:21 -0000
From: Chung-Lin Tang <cltang@codesourcery.com>
X-Patchwork-Id: 191289
Message-Id: <50793561.1000203@codesourcery.com>
To: Oleg Endo <oleg.endo@t-online.de>
Cc: gcc-patches <gcc-patches@gcc.gnu.org>, Richard Henderson <rth@redhat.com>, 
 Richard Sandiford <rdsandiford@googlemail.com>,
 Ramana Radhakrishnan	<ramana.radhakrishnan@arm.com>,
 Sterling Augustine <saugustine@google.com>,
 Andreas Krebbel <krebbel@linux.vnet.ibm.com>

On 2012/10/12 06:55 AM, Oleg Endo wrote:
> This broke the recently added thread pointer built-ins on SH, but I was
> prepared for that, so no problem here.  The attached patch is a straight
> forward fix.
> 
> However, with the patch applied I get an ICE on one of the SH thread
> pointer tests:  gcc/testsuite/gcc.target/sh/pr54760-3.c, function
> test04:
> 
> internal compiler error: in expand_insn, at optabs.c:8208
>     __builtin_set_thread_pointer (xx[i]);

Looks like I was supposed to use create_input_operand() there instead.
I've committed the attached patch as obvious. This should be fixed now.

Thanks,
Chung-Lin

        * builtins.c (expand_builtin_set_thread_pointer): Use
        create_input_operand() instead of create_fixed_operand().


Index: builtins.c
===================================================================
--- builtins.c	(revision 192421)
+++ builtins.c	(revision 192422)
@@ -5776,7 +5776,7 @@
       struct expand_operand op;
       rtx val = expand_expr (CALL_EXPR_ARG (exp, 0), NULL_RTX,
 			     Pmode, EXPAND_NORMAL);      
-      create_fixed_operand (&op, val);
+      create_input_operand (&op, val, Pmode);
       expand_insn (icode, 1, &op);
       return;
     }