1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
|
The powerpc assembler can generate .PPC.EMB.apuinfo sections in some
circumstances. These originate, I think, from the old e500 ABI; in
any case, they are only defined for 32-bit ABIs, and there is a
significant amount of code in bfd/elf32-ppc.c to handle them
appropriately, none of which is in elf64-ppc.c. However, the
assembler can generate them when building 64-bit code for processors
such as e5500 (which uses PPC_OPCODE_E500MC). In such circumstances,
they get generated as PROGBITS rather than NOTE sections, which
reportedly can cause problems in a Linux kernel build.
Since these sections are only defined in 32-bit ABIs, I think it's
most appropriate to avoid generating them for 64-bit output, which
this patch does.
Upstream-Status: Backport
-Khem 2013/03/25
From d77411f56a85225e3f7b674e7172c712f6cd61cb Mon Sep 17 00:00:00 2001
From: Joseph Myers <joseph@codesourcery.com>
Date: Fri, 25 Jan 2013 02:02:20 +0000
Subject: [PATCH] * config/tc-ppc.c (md_assemble): Do not generate
APUinfo sections for 64-bit output.
---
gas/ChangeLog | 5 +++++
gas/config/tc-ppc.c | 3 ++-
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/gas/config/tc-ppc.c b/gas/config/tc-ppc.c
index fd15bea..0929e52 100644
--- a/gas/config/tc-ppc.c
+++ b/gas/config/tc-ppc.c
@@ -3249,7 +3249,8 @@ md_assemble (char *str)
#ifdef OBJ_ELF
/* Do we need/want an APUinfo section? */
- if ((ppc_cpu & (PPC_OPCODE_E500 | PPC_OPCODE_E500MC | PPC_OPCODE_VLE)) != 0)
+ if ((ppc_cpu & (PPC_OPCODE_E500 | PPC_OPCODE_E500MC | PPC_OPCODE_VLE)) != 0
+ && !ppc_obj64)
{
/* These are all version "1". */
if (opcode->flags & PPC_OPCODE_SPE)
--
1.7.9.5
|