Thread (24 messages) 24 messages, 8 authors, 2018-11-07

Re: [PATCH] ppc64: fix compilation of when AltiVec is enabled

From: Christian Ehrhardt <hidden>
Date: 2018-08-30 11:58:27

On Thu, Aug 30, 2018 at 12:52 PM Takeshi T Yoshimura [off-list ref]
wrote:
quoted
宛先: adrien.mazarguil@6wind.com, dev [off-list ref], Gowrishankar
Muthukrishnan [off-list ref], Chao Zhu
[off-list ref]
送信元: Christian Ehrhardt
送信者: "dev"
日付: 2018/08/30 07:00PM
Cc: Luca Boccassi <redacted>, Thomas Monjalon
[off-list ref], Christian Ehrhardt
[off-list ref]
件名: [dpdk-dev] [PATCH] ppc64: fix compilation of when AltiVec is
enabled

The definition of almost any newer standard like --stc=c11 will drop
__APPLCE_ALTIVEC__ which otherwise would be defined.
If that is the case then altivec.h will redefine bool to a type
conflicting with those defined by stdbool.h.

This breaks compilation of 18.08 on ppc64 like:
 mlx5_nl_flow.c:407:17: error: incompatible types when assigning to
type
 ‘__vector __bool int’ {aka ‘__vector(4) __bool int’} from type
‘int’
 in_port_id_set = false;

Other alternatives were pursued on [1] but they always ended up being
more
complex than what would be appropriate for the issue we face.

[1]:
INVALID URI REMOVED
chives_dev_2018-2DAugust_109926.html&d=DwIDaQ&c=jf_iaSHvJObTbx-siA1ZO
g&r=EZR6Jx10q0q3dTopeH3WIQ&m=bbU1KVc1ZvNW9Rz7B0MLHfS0f0oZv35d2mpRpHO0
ByY&s=RvMIFfk-cAAGTrYM76-iSSqIYV_X2EptYZzYIweIHRk&e=

Change-Id: I1ed56da954e4951b9d120ca2ac0c0c218b4a0140
Signed-off-by: Christian Ehrhardt <redacted>
---
.../common/include/arch/ppc_64/rte_memcpy.h           | 11
+++++++++++
1 file changed, 11 insertions(+)
diff --git a/lib/librte_eal/common/include/arch/ppc_64/rte_memcpy.h
b/lib/librte_eal/common/include/arch/ppc_64/rte_memcpy.h
index 75f74897b..0b3b89b56 100644
--- a/lib/librte_eal/common/include/arch/ppc_64/rte_memcpy.h
+++ b/lib/librte_eal/common/include/arch/ppc_64/rte_memcpy.h
@@ -37,6 +37,17 @@
#include <string.h>
/*To include altivec.h, GCC version must  >= 4.8 */
#include <altivec.h>
+/*
+ * Compilation workaround for PPC64 targets when AltiVec is fully
+ * enabled e.g. with std=c11. Otherwise there would be a type
conflict
+ * of "bool" between stdbool and altivec.
+ */
+#if defined(__PPC64__) && !defined(__APPLE_ALTIVEC__)
+ #undef bool
+ /* redefine as in stdbool.h */
+ #define bool _Bool
+#endif
+

#ifdef __cplusplus
extern "C" {
--
2.17.1

Hi,
I could reproduce the issue you reported in 18.08 with my ppc64le box with
RedHat 7.5 and GCC4.8.
The patch resolved the issue in my environment. Thanks!
I added your test (tanks) and Adrien's extensive review/discussion as tags
and also addressed a few checkpatch findings.
V2 is up on the list now ...

I am a bit newbie in dpdk-dev, but I will try contacting Chao and other
IBM guys... Sorry for our slow reply.
Thanks for your participation Takeshi,
we at least now have had a few replies after Thomas used the superpowers of
"CPT. CAPSLOCK" \o/.

I also have a call later today to make sure this is brought up inside IBM
to make sure someone is maintaining it for real.

Regards,
Takeshi
-- 
Christian Ehrhardt
Software Engineer, Ubuntu Server
Canonical Ltd
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help