* [LTP] [PATCH v3] mbind01, get_mempolicy01: dont use hardcoded node 0
@ 2012-05-25 7:05 Jan Stancek
2012-05-25 7:14 ` Wanlong Gao
0 siblings, 1 reply; 2+ messages in thread
From: Jan Stancek @ 2012-05-25 7:05 UTC (permalink / raw)
To: ltp-list; +Cc: Jeffrey Burke
[-- Attachment #1: Type: text/plain, Size: 884 bytes --]
These tests are failing on following setup:
# numactl -H
available: 2 nodes (0-1)
node 0 cpus: 0 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
node 0 size: 0 MB
node 0 free: 0 MB
node 1 cpus:
node 1 size: 12288 MB
node 1 free: 9998 MB
node distances:
node 0 1
0: 10 40
1: 40 10
Avoid hardcoded node 0 by picking first available node obtained using
get_mempolicy(..., MPOL_F_MEMS_ALLOWED);
Using whole nodemask doesn't work with MPOL_PREFERRED, because set_mempolicy
will use only first node and subsequent check for bitmask equality fails.
Signed-off-by: Jan Stancek <jstancek@redhat.com>
---
.../syscalls/get_mempolicy/get_mempolicy01.c | 22 ++++++++++++++++++-
testcases/kernel/syscalls/mbind/mbind01.c | 22 ++++++++++++++++++-
2 files changed, 40 insertions(+), 4 deletions(-)
[-- Attachment #2: 0001-mbind01-get_mempolicy01-dont-use-hardcoded-node-0.patch --]
[-- Type: text/x-patch, Size: 2828 bytes --]
diff --git a/testcases/kernel/syscalls/get_mempolicy/get_mempolicy01.c b/testcases/kernel/syscalls/get_mempolicy/get_mempolicy01.c
index 0565d6f..d2de92b 100644
--- a/testcases/kernel/syscalls/get_mempolicy/get_mempolicy01.c
+++ b/testcases/kernel/syscalls/get_mempolicy/get_mempolicy01.c
@@ -231,15 +231,33 @@ static int do_test(struct test_case *tc)
#endif
char *p = NULL;
unsigned long len = MEM_LENGTH;
+ struct bitmask *allowed = NULL;
+ int i, test_node = -1;
+
+ allowed = numa_allocate_nodemask();
+ /*
+ * avoid numa_get_mems_allowed(), because of bug in getpol()
+ * utility function in older versions:
+ * http://www.spinics.net/lists/linux-numa/msg00849.html
+ */
+ TEST(syscall(__NR_get_mempolicy, NULL, allowed->maskp,
+ allowed->size, 0, MPOL_F_MEMS_ALLOWED));
+ if (TEST_RETURN < 0)
+ tst_resm(TFAIL|TERRNO, "get_mempolicy: MPOL_F_MEMS_ALLOWED");
+ for (i = 0; i < allowed->size; i++)
+ if (numa_bitmask_isbitset(allowed, i)) {
+ test_node = i;
+ break;
+ }
#if !defined(LIBNUMA_API_VERSION) || LIBNUMA_API_VERSION < 2
nodemask = malloc(sizeof(nodemask_t));
nodemask_zero(nodemask);
- nodemask_set(nodemask, 0);
+ nodemask_set(nodemask, test_node);
getnodemask = malloc(sizeof(nodemask_t));
nodemask_zero(getnodemask);
#else
- numa_bitmask_setbit(nodemask, 0);
+ numa_bitmask_setbit(nodemask, test_node);
#endif
switch (tc->ttype) {
case DEFAULT:
diff --git a/testcases/kernel/syscalls/mbind/mbind01.c b/testcases/kernel/syscalls/mbind/mbind01.c
index 4b25db0..cdc1721 100644
--- a/testcases/kernel/syscalls/mbind/mbind01.c
+++ b/testcases/kernel/syscalls/mbind/mbind01.c
@@ -217,15 +217,33 @@ static int do_test(struct test_case *tc)
unsigned long maxnode = NUMA_NUM_NODES;
unsigned long len = MEM_LENGTH;
unsigned long *invalid_nodemask;
+ struct bitmask *allowed = NULL;
+ int i, test_node = -1;
+
+ allowed = numa_allocate_nodemask();
+ /*
+ * avoid numa_get_mems_allowed(), because of bug in getpol()
+ * utility function in older versions:
+ * http://www.spinics.net/lists/linux-numa/msg00849.html
+ */
+ TEST(syscall(__NR_get_mempolicy, NULL, allowed->maskp,
+ allowed->size, 0, MPOL_F_MEMS_ALLOWED));
+ if (TEST_RETURN < 0)
+ tst_resm(TFAIL|TERRNO, "get_mempolicy: MPOL_F_MEMS_ALLOWED");
+ for (i = 0; i < allowed->size; i++)
+ if (numa_bitmask_isbitset(allowed, i)) {
+ test_node = i;
+ break;
+ }
#if !defined(LIBNUMA_API_VERSION) || LIBNUMA_API_VERSION < 2
nodemask = malloc(sizeof(nodemask_t));
nodemask_zero(nodemask);
- nodemask_set(nodemask, 0);
+ nodemask_set(nodemask, test_node);
getnodemask = malloc(sizeof(nodemask_t));
nodemask_zero(getnodemask);
#else
- numa_bitmask_setbit(nodemask, 0);
+ numa_bitmask_setbit(nodemask, test_node);
#endif
p = mmap(NULL, len, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS,
0, 0);
[-- Attachment #3: Type: text/plain, Size: 395 bytes --]
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
[-- Attachment #4: Type: text/plain, Size: 155 bytes --]
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [LTP] [PATCH v3] mbind01, get_mempolicy01: dont use hardcoded node 0
2012-05-25 7:05 [LTP] [PATCH v3] mbind01, get_mempolicy01: dont use hardcoded node 0 Jan Stancek
@ 2012-05-25 7:14 ` Wanlong Gao
0 siblings, 0 replies; 2+ messages in thread
From: Wanlong Gao @ 2012-05-25 7:14 UTC (permalink / raw)
To: Jan Stancek; +Cc: ltp-list, Jeffrey Burke
On 05/25/2012 03:05 PM, Jan Stancek wrote:
>
> These tests are failing on following setup:
> # numactl -H
> available: 2 nodes (0-1)
> node 0 cpus: 0 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
> node 0 size: 0 MB
> node 0 free: 0 MB
> node 1 cpus:
> node 1 size: 12288 MB
> node 1 free: 9998 MB
> node distances:
> node 0 1
> 0: 10 40
> 1: 40 10
>
> Avoid hardcoded node 0 by picking first available node obtained using
> get_mempolicy(..., MPOL_F_MEMS_ALLOWED);
>
> Using whole nodemask doesn't work with MPOL_PREFERRED, because set_mempolicy
> will use only first node and subsequent check for bitmask equality fails.
>
Thank you Jan,
Acked-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
> Signed-off-by: Jan Stancek <jstancek@redhat.com>
> ---
> .../syscalls/get_mempolicy/get_mempolicy01.c | 22 ++++++++++++++++++-
> testcases/kernel/syscalls/mbind/mbind01.c | 22 ++++++++++++++++++-
> 2 files changed, 40 insertions(+), 4 deletions(-)
>
>
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>
>
>
> _______________________________________________
> Ltp-list mailing list
> Ltp-list@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ltp-list
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2012-05-25 7:15 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-05-25 7:05 [LTP] [PATCH v3] mbind01, get_mempolicy01: dont use hardcoded node 0 Jan Stancek
2012-05-25 7:14 ` Wanlong Gao
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox