summaryrefslogtreecommitdiff
path: root/poky/meta/recipes-extended/stress-ng/stress-ng-0.13.12/0001-stress-cpu-disable-float128-math-on-powerpc64-to-avo.patch
blob: bb35b3030a775f4df1868230c8a87c04d044eb6b (plain)
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
From ea9ee4dd64ee88e03a959b2c694aa8feb53c7e78 Mon Sep 17 00:00:00 2001
From: He Zhe <zhe.he@windriver.com>
Date: Wed, 28 Sep 2022 16:47:24 +0800
Subject: [PATCH] stress-cpu: disable float128 math on powerpc64 to avoid
 SIGILL

float128 requires instructions of xsmaddqp and xsmsubqp which are added to
qemu since v7.0 by the following commit.
https://github.com/qemu/qemu/commit/3bb1aed246d7b59ceee625a82628f7369d492a8f

While kirkstone is still at v6.2 and thus experiences SIGILL as follow
root@qemuppc64:~# stress-ng --cpu 2 --timeout 30s
stress-ng: info: [972] setting to a 30 second run per stressor
stress-ng: info: [972] dispatching hogs: 2 cpu
stress-ng: info: [973] stressor terminated with unexpected signal signal 4 'SIGILL'
<snip>

Upstream-Status: Inappropriate [This is specific to kirkstone since qemu on
master branch has upgraded to v7.1.]

Signed-off-by: He Zhe <zhe.he@windriver.com>
---
 stress-cpu.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/stress-cpu.c b/stress-cpu.c
index 0a08f1d1..2849e715 100644
--- a/stress-cpu.c
+++ b/stress-cpu.c
@@ -41,6 +41,10 @@
 #undef HAVE_FLOAT_DECIMAL128
 #endif
 
+#if defined(STRESS_ARCH_PPC64)
+#undef HAVE_FLOAT128
+#endif
+
 #define GAMMA 		(0.57721566490153286060651209008240243104215933593992L)
 #define OMEGA		(0.56714329040978387299996866221035554975381578718651L)
 #define PSI		(3.35988566624317755317201130291892717968890513373197L)
-- 
2.25.1