yann@1
|
1 |
Grabbed with
|
yann@1
|
2 |
wget 'http://sources.redhat.com/cgi-bin/cvsweb.cgi/src/opcodes/ppc-opc.c.diff?r1=1.70&r2=1.71&cvsroot=src'
|
yann@1
|
3 |
See
|
yann@1
|
4 |
http://mail.gnu.org/archive/html/bug-binutils/2004-05/msg00071.html
|
yann@1
|
5 |
and
|
yann@1
|
6 |
http://sources.redhat.com/ml/binutils-cvs/2004-05/msg00111.html
|
yann@1
|
7 |
|
yann@1
|
8 |
This fixes problems like
|
yann@1
|
9 |
{standard input}: Assembler messages:
|
yann@1
|
10 |
{standard input}:1: Error: Unrecognized opcode: `dssall'
|
yann@1
|
11 |
which show up in binutils-2.15 when building the Linux kernel,
|
yann@1
|
12 |
or possibly failed compilations when building setjmp/longjmp in glibc.
|
yann@1
|
13 |
|
yann@1
|
14 |
An alternative fix would be to pass -many to the assembler.
|
yann@1
|
15 |
A patch to do that for glibc is at
|
yann@1
|
16 |
wget 'http://sources.redhat.com/cgi-bin/get-raw-msg?listname=libc-alpha&date=2004-05&msgid=40B36E8C.9030609%40us.ibm.com'
|
yann@1
|
17 |
Yet another, better, fix would be for gcc to generate assembly
|
yann@1
|
18 |
that told the assembler which processor type to use.
|
yann@1
|
19 |
Presumably the Linux kernel sources would need a fix, too.
|
yann@1
|
20 |
Probably better to just fix binutils to accept sources that it used to.
|
yann@1
|
21 |
|
yann@1
|
22 |
|
yann@1
|
23 |
===================================================================
|
yann@1
|
24 |
RCS file: /cvs/src/src/opcodes/ppc-opc.c,v
|
yann@1
|
25 |
retrieving revision 1.70
|
yann@1
|
26 |
retrieving revision 1.71
|
yann@1
|
27 |
diff -u -r1.70 -r1.71
|
yann@1
|
28 |
--- src/opcodes/ppc-opc.c 2004/05/05 13:43:36 1.70
|
yann@1
|
29 |
+++ src/opcodes/ppc-opc.c 2004/05/19 05:11:48 1.71
|
yann@1
|
30 |
@@ -1004,8 +1004,13 @@
|
yann@1
|
31 |
|
yann@1
|
32 |
/* If only one bit of the FXM field is set, we can use the new form
|
yann@1
|
33 |
of the instruction, which is faster. Unlike the Power4 branch hint
|
yann@1
|
34 |
- encoding, this is not backward compatible. */
|
yann@1
|
35 |
- else if ((dialect & PPC_OPCODE_POWER4) != 0 && (value & -value) == value)
|
yann@1
|
36 |
+ encoding, this is not backward compatible. Do not generate the
|
yann@1
|
37 |
+ new form unless -mpower4 has been given, or -many and the two
|
yann@1
|
38 |
+ operand form of mfcr was used. */
|
yann@1
|
39 |
+ else if ((value & -value) == value
|
yann@1
|
40 |
+ && ((dialect & PPC_OPCODE_POWER4) != 0
|
yann@1
|
41 |
+ || ((dialect & PPC_OPCODE_ANY) != 0
|
yann@1
|
42 |
+ && (insn & (0x3ff << 1)) == 19 << 1)))
|
yann@1
|
43 |
insn |= 1 << 20;
|
yann@1
|
44 |
|
yann@1
|
45 |
/* Any other value on mfcr is an error. */
|