numa: fix numa test error with non-continuous nodes
PATCH v2 1/2 numa: fix numa test error with non-continuous nodes
Numa test failed on such machine which has non-continuous numa nodes, it gets wrong data because of the below syntax rule is not applicable to that special situation.
` numastat -p $pid |awk '/^Total/ {print $'$((node+2))'}'
In this patch, we fix that via a bit complex way of awk to get the allocated memory in specified node.
# numactl -H
available: 2 nodes (0,8)
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 32 33 34
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60 61 62 63
node 0 size: 257741 MB
node 0 free: 253158 MB
node 8 cpus: 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
96 97 98 99 100 101 102 103 104 105 106 107 108
109 110 111 112 113 114 115 116 117 118 119 120
121 122 123 124 125 126 127
node 8 size: 261752 MB
node 8 free: 240933 MB
node distances:
node 0 8
0: 10 40
8: 40 10 # numastat -p $pid
Per-node process memory usage (in MBs) for PID 34168 (support_numa)
Node 0 Node 8 Total
--------------- --------------- ---------------
Huge 0.00 0.00 0.00
Heap 0.00 0.12 0.12
Stack 0.00 0.06 0.06
Private 1.62 1.50 3.12
---------------- --------------- --------------- ---------------
Total 1.62 1.69 3.31--- testcases/kernel/numa/numa01.sh | 26 ++++++++++++++++++++++---- 1 file changed, 22 insertions(+), 4 deletions(-)
testcases/kernel/numa/numa01.sh +22 -4
@@ -52,9 +52,18 @@ TST_NEEDS_CMDS="awk bc numactl numastat" extract_numastat_p() { local pid=$1 - local node=$(($2 + 2)) - - echo $(numastat -p $pid |awk '/^Total/ {print $'$node'}') + local node=$2 + + echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \ + } \ + /^Total/ {print $col}' \ + ) } check_for_support_numa()
@@ -363,7 +372,16 @@ test9() pid=$! TST_RETRY_FUNC "check_for_support_numa $pid" 0 - Mem_huge=$(echo $(numastat -p $pid |awk '/^Huge/ {print $'$((node+2))'}')) + Mem_huge=$(echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \ + } \ + /^Huge/ {print $col}') \ + ) Mem_huge=$((${Mem_huge%.*} * 1024)) if [ "$Mem_huge" -lt "$HPAGE_SIZE" ]; then
Signature · 1 line
2.20.1
PATCH v2 2/2 libltpnuma: remove restrictions on numa node-id
For some ppc64le systems, it has non-continuous numa nodes in hardware configuration. So we're hitting the below warnings while running set_mempolicy tests on that. To fix this issue, let's just remove restrictions on numa node-id in get_mempolicy().
Error Log
---------
tst_test.c:1096: INFO: Timeout per run is 0h 50m 00s tst_numa.c:190: INFO: Found 2 NUMA memory nodes set_mempolicy01.c:63: PASS: set_mempolicy(MPOL_BIND) node 0 tst_numa.c:26: INFO: Node 0 allocated 16 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:82: PASS: child: Node 0 allocated 16 set_mempolicy01.c:63: PASS: set_mempolicy(MPOL_BIND) node 8 tst_numa.c:92: WARN: get_mempolicy(...) returned invalid node 8 tst_numa.c:92: WARN: get_mempolicy(...) returned invalid node 8 tst_numa.c:92: WARN: get_mempolicy(...) returned invalid node 8
...
tst_numa.c:26: INFO: Node 0 allocated 0 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:86: FAIL: child: Node 8 allocated 0, expected 16
lscpu
-----
Architecture: ppc64le
...
CPU(s): 128 Core(s) per socket: 16 Socket(s): 2 NUMA node(s): 2 Model name: POWER9, altivec supported ... NUMA node0 CPU(s): 0-63 NUMA node8 CPU(s): 64-127
--- libs/libltpnuma/tst_numa.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
libs/libltpnuma/tst_numa.c +3 -2
@@ -88,8 +88,9 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" ); -if (node <0 || (unsigned int )node >= nodes->cnt) { -tst_res (TWARN ,"get_mempolicy(...) returned invalid node %i\n" , node); +if (node <0 ) { +tst_res (TWARN , +"get_mempolicy(...) returned invalid node %i\n" , node);continue ; }
Signature · 1 line
2.20.1
PATCH v2 2/2 libltpnuma: remove restrictions on numa node-id
Quoted 46 lines · Li Wang
For some ppc64le systems, it has non-continuous numa nodes in hardware configuration. So we're hitting the below warnings while running
set_mempolicytests on that. To fix this issue, let's just remove restrictions on numa node-id inget_mempolicy().Error Log
---------
tst_test.c:1096: INFO: Timeout per run is 0h 50m 00s tst_numa.c:190: INFO: Found 2 NUMA memory nodes set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 0 tst_numa.c:26: INFO: Node 0 allocated 16 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:82: PASS: child: Node 0 allocated 16 set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8
...
tst_numa.c:26: INFO: Node 0 allocated 0 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:86: FAIL: child: Node 8 allocated 0, expected 16lscpu
-----
Architecture: ppc64le
...CPU(s): 128 Core(s) per socket: 16 Socket(s): 2 NUMA node(s): 2 Model name: POWER9, altivec supported ... NUMA node0 CPU(s): 0-63 NUMA node8 CPU(s): 64-127Signed-off-by Li Wang <[email protected]>Cc Cyril Hrubis <[email protected]>--- libs/libltpnuma/tst_numa.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)libs/libltpnuma/tst_numa.c +0
-0
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 91
Quoted 6 lines
libs/libltpnuma/tst_numa.c +1
-2@@ -88,8 +88,9 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" ); -if (node <0 || (unsigned int )node >= nodes->cnt) { -tst_res (TWARN ,"get_mempolicy(...) returned invalid node %i\n" , node); +if (node <0 ) {
It is better to have check for powerpc and remove restriction as it will still hold good for other arch.
-- Bala
Quoted 11 lines
libs/libltpnuma/tst_numa.c +2
-0+tst_res (TWARN , +"get_mempolicy(...) returned invalid node %i\n" , node);continue ; }Signature · 5 lines
2.20.1 -- Mailing list info: https://lists.linux.it/listinfo/ltp
PATCH v2 2/2 libltpnuma: remove restrictions on numa node-id
Quoted 50 lines · Balamuruhan S
Quoted 49 lines · Li Wang
For some ppc64le systems, it has non-continuous numa nodes in hardware configuration. So we're hitting the below warnings while running
set_mempolicytests on that. To fix this issue, let's just remove restrictions on numa node-id inget_mempolicy().Error Log
---------
tst_test.c:1096: INFO: Timeout per run is 0h 50m 00s tst_numa.c:190: INFO: Found 2 NUMA memory nodes set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 0 tst_numa.c:26: INFO: Node 0 allocated 16 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:82: PASS: child: Node 0 allocated 16 set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8
...
tst_numa.c:26: INFO: Node 0 allocated 0 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:86: FAIL: child: Node 8 allocated 0, expected 16lscpu
-----
Architecture: ppc64le
...CPU(s): 128 Core(s) per socket: 16 Socket(s): 2 NUMA node(s): 2 Model name: POWER9, altivec supported ... NUMA node0 CPU(s): 0-63 NUMA node8 CPU(s): 64-127Signed-off-by Li Wang <[email protected]>Cc Cyril Hrubis <[email protected]>--- libs/libltpnuma/tst_numa.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)libs/libltpnuma/tst_numa.c +0
-0@@ -88,8 +88,9 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" );
Quoted 8 lines
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 92
Quoted 2 lines
libs/libltpnuma/tst_numa.c +0
-2-if (node <0 || (unsigned int )node >= nodes->cnt) { - tst_res(TWARN, "get_mempolicy(...) returnedinvalid node %i\n", node);
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 91
Quoted 1 line
libs/libltpnuma/tst_numa.c +1
-0+if (node <0 ) {It is better to have check for powerpc and remove restriction as it will still hold good for other arch.
If the non-continuous numa nodes only appear in ppc machine, that should be necessary to check arch before removing the restriction.
then the patch probably like:
libs/libltpnuma/tst_numa.c +4 -0
@@ -88,7 +88,11 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" ); +#if defined (__powerpc64__) || (__powerpc__) +if (node <0 ) { +#elif if (node <0 || (unsigned int )node >= nodes->cnt) { +#endif tst_res(TWARN, "get_mempolicy(...) returned invalid
node %i\n", node);
continue;
}Signature · 5 lines
Regards, Li Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linux.it/pipermail/ltp/attachments/20190509/e93d1f8f/attachment-0001.html>
PATCH v2 2/2 libltpnuma: remove restrictions on numa node-id
Quoted 70 lines · Li Wang
Quoted 50 lines · Balamuruhan S
Quoted 49 lines · Li Wang
For some ppc64le systems, it has non-continuous numa nodes in hardware configuration. So we're hitting the below warnings while running
set_mempolicytests on that. To fix this issue, let's just remove restrictions on numa node-id inget_mempolicy().Error Log
---------
tst_test.c:1096: INFO: Timeout per run is 0h 50m 00s tst_numa.c:190: INFO: Found 2 NUMA memory nodes set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 0 tst_numa.c:26: INFO: Node 0 allocated 16 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:82: PASS: child: Node 0 allocated 16 set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8
...
tst_numa.c:26: INFO: Node 0 allocated 0 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:86: FAIL: child: Node 8 allocated 0, expected 16lscpu
-----
Architecture: ppc64le
...CPU(s): 128 Core(s) per socket: 16 Socket(s): 2 NUMA node(s): 2 Model name: POWER9, altivec supported ... NUMA node0 CPU(s): 0-63 NUMA node8 CPU(s): 64-127Signed-off-by Li Wang <[email protected]>Cc Cyril Hrubis <[email protected]>--- libs/libltpnuma/tst_numa.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)libs/libltpnuma/tst_numa.c +0
-0@@ -88,8 +88,9 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" );Quoted 8 lines
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 92
Quoted 2 lines
libs/libltpnuma/tst_numa.c +0
-2-if (node <0 || (unsigned int )node >= nodes->cnt) { - tst_res(TWARN, "get_mempolicy(...) returnedinvalid node %i\n", node);
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 92
Quoted 1 line
libs/libltpnuma/tst_numa.c +1
-0+if (node <0 ) {It is better to have check for powerpc and remove restriction as it will still hold good for other arch.
If the non-continuous numa nodes only appear in ppc machine, that should be necessary to check arch before removing the restriction.
then the patch probably like:
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 95
Quoted 8 lines
libs/libltpnuma/tst_numa.c +4
-0@@ -88,7 +88,11 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" ); +#if defined (__powerpc64__) || (__powerpc__) +if (node <0 ) { +#elif if (node <0 || (unsigned int )node >= nodes->cnt) { +#endif
Looks good to me. Thanks!
-- Bala
Quoted 9 lines
libs/libltpnuma/tst_numa.c +0
-0tst_res(TWARN, "get_mempolicy(...) returned invalid node %i\n", node); continue;}Signature · 2 lines
Regards, Li Wang
PATCH v2 2/2 libltpnuma: remove restrictions on numa node-id
In reply to Balamuruhan S----- Original Message -----
Quoted 50 lines · Li Wang
Quoted 49 lines
For some ppc64le systems, it has non-continuous numa nodes in hardware configuration. So we're hitting the below warnings while running
set_mempolicytests on that. To fix this issue, let's just remove restrictions on numa node-id inget_mempolicy().Error Log
---------
tst_test.c:1096: INFO: Timeout per run is 0h 50m 00s tst_numa.c:190: INFO: Found 2 NUMA memory nodes set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 0 tst_numa.c:26: INFO: Node 0 allocated 16 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:82: PASS: child: Node 0 allocated 16 set_mempolicy01.c:63: PASS:set_mempolicy(MPOL_BIND)node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8 tst_numa.c:92: WARN:get_mempolicy(...)returned invalid node 8
...
tst_numa.c:26: INFO: Node 0 allocated 0 pages tst_numa.c:26: INFO: Node 8 allocated 0 pages set_mempolicy01.c:86: FAIL: child: Node 8 allocated 0, expected 16lscpu
-----
Architecture: ppc64le
...CPU(s): 128 Core(s) per socket: 16 Socket(s): 2 NUMA node(s): 2 Model name: POWER9, altivec supported ... NUMA node0 CPU(s): 0-63 NUMA node8 CPU(s): 64-127Signed-off-by Li Wang <[email protected]>Cc Cyril Hrubis <[email protected]>--- libs/libltpnuma/tst_numa.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)libs/libltpnuma/tst_numa.c +0
-0@@ -88,8 +88,9 @@ void tst_nodemap_count_pages(struct tst_nodemap *nodes,if (ret <0 )tst_brk (TBROK |TERRNO ,"get_mempolicy() failed" );
Quoted 6 lines
libs/libltpnuma/tst_numa.c void tst_nodemap_count_pages(struct tst_nodemap *nodes, · line 91
Quoted 3 lines
libs/libltpnuma/tst_numa.c +1
-2-if (node <0 || (unsigned int )node >= nodes->cnt) { -tst_res (TWARN ,"get_mempolicy(...) returned invalid node %i\n" , node); +if (node <0 ) {It is better to have check for powerpc and remove restriction as it will still hold good for other arch.
Unless memory gets offlined or test harness restricts LTP to run only on some nodes (via numactl). Here's example of former:
# uname -m x86_64
# numactl -H available: 8 nodes (0-7) node 0 cpus: 0 4 8 12 node 0 size: 4061 MB node 0 free: 3437 MB node 1 cpus: 16 20 24 28 node 1 size: 0 MB node 1 free: 0 MB node 2 cpus: 1 5 9 13 node 2 size: 4096 MB node 2 free: 3871 MB node 3 cpus: 17 21 25 29 node 3 size: 0 MB node 3 free: 0 MB node 4 cpus: 2 6 10 14 node 4 size: 4096 MB node 4 free: 3568 MB node 5 cpus: 18 22 26 30 node 5 size: 0 MB node 5 free: 0 MB node 6 cpus: 19 23 27 31 node 6 size: 0 MB node 6 free: 0 MB node 7 cpus: 3 7 11 15 node 7 size: 4095 MB node 7 free: 3802 MB
tst_test.c:1096: INFO: Timeout per run is 0h 05m 00s tst_numa.c:190: INFO: Found 4 NUMA memory nodes
...
tst_numa.c:92: WARN: get_mempolicy(...) returned invalid node 4 tst_numa.c:92: WARN: get_mempolicy(...) returned invalid node 7
Quoted 19 lines
-- Bala
Quoted 11 lines
libs/libltpnuma/tst_numa.c +2
-1+tst_res (TWARN , +"get_mempolicy(...) returned invalid node %i\n" , node);continue ; } --2.20.1
--
Mailing list info: https://lists.linux.it/listinfo/ltp--
Mailing list info: https://lists.linux.it/listinfo/ltp
PATCH v2 2/2 libltpnuma: remove restrictions on numa node-id
In reply to Balamuruhan SHi!
Quoted 2 lines
It is better to have check for powerpc and remove restriction as it will still hold good for other arch.
I would like to avoid adding arch specific ifdefs to the code, so for the release I've commited the original fix and we can add better checks later on. As I wrote checking that the numa id is <= than maximal used id would be much better and would work in all cases.
Signature · 2 lines
Cyril Hrubis [email protected]
PATCH v2 1/2 numa: fix numa test error with non-continuous nodes
In reply to Li WangQuoted 59 lines · Li Wang
Numa test failed on such machine which has non-continuous numa nodes, it gets wrong data because of the below syntax rule is not applicable to that special situation.
` numastat -p $pid |awk '/^Total/ {print $'$((node+2))'}'
In this patch, we fix that via a bit complex way of awk to get the allocated memory in specified node.# numactl -H available: 2 nodes (0,8) 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 node 0 size: 257741 MB node 0 free: 253158 MB node 8 cpus: 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 node 8 size: 261752 MB node 8 free: 240933 MB node distances: node 0 8 0: 10 40 8: 40 10# numastat -p $pid Per-node process memory usage (in MBs) for PID 34168 (support_numa) Node 0 Node 8 Total --------------- --------------- --------------- Huge 0.00 0.00 0.00 Heap 0.00 0.12 0.12 Stack 0.00 0.06 0.06 Private 1.62 1.50 3.12 ---------------- --------------- --------------- --------------- Total 1.62 1.69 3.31Signed-off-by Li Wang <[email protected]>Cc Chunyu Hu <[email protected]>Cc Cyril Hrubis <[email protected]>--- testcases/kernel/numa/numa01.sh | 26 ++++++++++++++++++++++---- 1 file changed, 22 insertions(+), 4 deletions(-)testcases/kernel/numa/numa01.sh +3
-3@@ -52,9 +52,18 @@ TST_NEEDS_CMDS="awk bc numactl numastat" extract_numastat_p() { local pid=$1 - local node=$(($2 + 2)) - - echo $(numastat -p $pid |awk '/^Total/ {print $'$node'}') + local node=$2 + + echo $(numastat -p $pid | \
testcases/kernel/numa/numa01.sh TST_NEEDS_CMDS="awk bc numactl numastat" · line 67
Quoted 10 lines
testcases/kernel/numa/numa01.sh +9
-0+ awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \ + } \ + /^Total/ {print $col}' \ + ) }
If we can use extract_numastat_p0 with node_index then the existing code works, for example if can index while iterating $node_list and use this with extract_numastat_p0 then it should work.
This is just my thoughts, please feel free to correct me if I miss or it is something wrong. Thanks!
-- Bala
Quoted 26 lines
testcases/kernel/numa/numa01.sh +10
-1check_for_support_numa()@@ -363,7 +372,16 @@ test9() pid=$! TST_RETRY_FUNC "check_for_support_numa $pid" 0 - Mem_huge=$(echo $(numastat -p $pid |awk '/^Huge/ {print $'$((node+2))'}')) + Mem_huge=$(echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \ + } \ + /^Huge/ {print $col}') \ + ) Mem_huge=$((${Mem_huge%.*} * 1024)) if [ "$Mem_huge" -lt "$HPAGE_SIZE" ]; thenSignature · 5 lines
2.20.1 -- Mailing list info: https://lists.linux.it/listinfo/ltp
PATCH v2 1/2 numa: fix numa test error with non-continuous nodes
Quoted 67 lines · Balamuruhan S
Quoted 38 lines · Li Wang
Numa test failed on such machine which has non-continuous numa nodes, it gets wrong data because of the below syntax rule is not applicable to that special situation.
` numastat -p $pid |awk '/^Total/ {print $'$((node+2))'}'
In this patch, we fix that via a bit complex way of awk to get the allocated memory in specified node.# numactl -H available: 2 nodes (0,8) 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 node 0 size: 257741 MB node 0 free: 253158 MB node 8 cpus: 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 node 8 size: 261752 MB node 8 free: 240933 MB node distances: node 0 8 0: 10 40 8: 40 10# numastat -p $pid Per-node process memory usage (in MBs) for PID 34168 (support_numa) Node 0 Node 8 Total --------------- --------------- --------------- Huge 0.00 0.00 0.00 Heap 0.00 0.12 0.12 Stack 0.00 0.06 0.06 Private 1.62 1.50 3.12 ---------------- --------------- --------------- --------------- Total 1.62 1.69 3.31Review
Quoted 8 lines
Signed-off-by Li Wang <[email protected]>Cc Chunyu Hu <[email protected]>Cc Cyril Hrubis <[email protected]>--- testcases/kernel/numa/numa01.sh | 26 ++++++++++++++++++++++---- 1 file changed, 22 insertions(+), 4 deletions(-)(quoted diff) +0
-0
b/testcases/kernel/numa/numa01.shQuoted 19 lines
index
33393ac8d..7f3dee849100755testcases/kernel/numa/numa01.sh +9
-3@@ -52,9 +52,18 @@ TST_NEEDS_CMDS="awk bc numactl numastat" extract_numastat_p() { local pid=$1 - local node=$(($2 + 2)) - - echo $(numastat -p $pid |awk '/^Total/ {print $'$node'}') + local node=$2 + + echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \
Quoted 9 lines
testcases/kernel/numa/numa01.sh TST_NEEDS_CMDS="awk bc numactl numastat" · line 67
Quoted 4 lines
testcases/kernel/numa/numa01.sh +3
-0+ } \ + /^Total/ {print $col}' \ + ) }If we can use
extract_numastat_p0withnode_indexthen the existing code works, for example if can index while iterating $node_list and use this withextract_numastat_p0then it should work.
Hi Bala,
Sorry, I don't fully understand what's you mean here :(. Could you explain something more for this method?
Quoted 34 lines
This is just my thoughts, please feel free to correct me if I miss or it is something wrong. Thanks!
-- Bala
testcases/kernel/numa/numa01.sh test9() · line 366
Quoted 7 lines
testcases/kernel/numa/numa01.sh +0
-1check_for_support_numa()@@ -363,7 +372,16 @@ test9() pid=$! TST_RETRY_FUNC "check_for_support_numa $pid" 0 - Mem_huge=$(echo $(numastat -p $pid |awk '/^Huge/ {print$'$((node+2))'}'))
Quoted 19 lines
testcases/kernel/numa/numa01.sh +10
-1+ Mem_huge=$(echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \ + } \ + /^Huge/ {print $col}') \ + ) Mem_huge=$((${Mem_huge%.*} * 1024)) if [ "$Mem_huge" -lt "$HPAGE_SIZE" ]; then --2.20.1
--
Mailing list info: https://lists.linux.it/listinfo/ltp
Signature · 5 lines
Regards, Li Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linux.it/pipermail/ltp/attachments/20190509/81d3c6c9/attachment.html>
PATCH v2 1/2 numa: fix numa test error with non-continuous nodes
Quoted 80 lines · Li Wang
Quoted 67 lines · Balamuruhan S
Quoted 38 lines · Li Wang
Numa test failed on such machine which has non-continuous numa nodes, it gets wrong data because of the below syntax rule is not applicable to that special situation.
` numastat -p $pid |awk '/^Total/ {print $'$((node+2))'}'
In this patch, we fix that via a bit complex way of awk to get the allocated memory in specified node.# numactl -H available: 2 nodes (0,8) 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 node 0 size: 257741 MB node 0 free: 253158 MB node 8 cpus: 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 node 8 size: 261752 MB node 8 free: 240933 MB node distances: node 0 8 0: 10 40 8: 40 10# numastat -p $pid Per-node process memory usage (in MBs) for PID 34168 (support_numa) Node 0 Node 8 Total --------------- --------------- --------------- Huge 0.00 0.00 0.00 Heap 0.00 0.12 0.12 Stack 0.00 0.06 0.06 Private 1.62 1.50 3.12 ---------------- --------------- --------------- --------------- Total 1.62 1.69 3.31Review
Quoted 8 lines
Signed-off-by Li Wang <[email protected]>Cc Chunyu Hu <[email protected]>Cc Cyril Hrubis <[email protected]>--- testcases/kernel/numa/numa01.sh | 26 ++++++++++++++++++++++---- 1 file changed, 22 insertions(+), 4 deletions(-)(quoted diff) +0
-0
b/testcases/kernel/numa/numa01.shQuoted 19 lines
index
33393ac8d..7f3dee849100755testcases/kernel/numa/numa01.sh +9
-3@@ -52,9 +52,18 @@ TST_NEEDS_CMDS="awk bc numactl numastat" extract_numastat_p() { local pid=$1 - local node=$(($2 + 2)) - - echo $(numastat -p $pid |awk '/^Total/ {print $'$node'}') + local node=$2 + + echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \Quoted 9 lines
testcases/kernel/numa/numa01.sh TST_NEEDS_CMDS="awk bc numactl numastat" · line 67
Quoted 4 lines
testcases/kernel/numa/numa01.sh +3
-0+ } \ + /^Total/ {print $col}' \ + ) }If we can use
extract_numastat_p0withnode_indexthen the existing code works, for example if can index while iterating $node_list and use this withextract_numastat_p0then it should work.
Quoted 4 lines
Hi Bala,
Sorry, I don't fully understand what's you mean here :(. Could you explain something more for this method?
instead of changing extract_numastat_p0() can do something like this ?
testcases/kernel/numa/numa01.sh +3 -1
@@ -94,6 +94,7 @@ setup() test1() { Mem_curr=0 + node_index=0 for node in $nodes_list; do numactl --cpunodebind=$node --membind=$node support_numa alloc_1MB &
@@ -101,7 +102,8 @@ test1() TST_RETRY_FUNC "check_for_support_numa $pid" 0 - Mem_curr=$(echo "$(extract_numastat_p $pid $node) * $MB" |bc) + Mem_curr=$(echo "$(extract_numastat_p $pid $node_index) * $MB" |bc) + let node_index++ if [ $(echo "$Mem_curr < $MB" | bc) -eq 1 ]; then tst_res TFAIL \ "NUMA memory allocated in node$node is less than expected"
-- Bala
Quoted 41 lines
Quoted 37 lines
This is just my thoughts, please feel free to correct me if I miss or it is something wrong. Thanks!
-- Bala
testcases/kernel/numa/numa01.sh test9() · line 366
Quoted 7 lines
testcases/kernel/numa/numa01.sh +0
-1check_for_support_numa()@@ -363,7 +372,16 @@ test9() pid=$! TST_RETRY_FUNC "check_for_support_numa $pid" 0 - Mem_huge=$(echo $(numastat -p $pid |awk '/^Huge/ {print$'$((node+2))'}'))
Quoted 19 lines
testcases/kernel/numa/numa01.sh +10
-1+ Mem_huge=$(echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \ + col=i+1; \ + next \ + } \ + /^Huge/ {print $col}') \ + ) Mem_huge=$((${Mem_huge%.*} * 1024)) if [ "$Mem_huge" -lt "$HPAGE_SIZE" ]; then --2.20.1
--
Mailing list info: https://lists.linux.it/listinfo/ltpSignature · 2 lines
Regards, Li Wang
PATCH v2 1/2 numa: fix numa test error with non-continuous nodes
Hi Bala,
Quoted 54 lines · Balamuruhan S
[...snip...]
Quoted 16 lines
Quoted 16 lines
Quoted 16 lines
testcases/kernel/numa/numa01.sh +7
-3@@ -52,9 +52,18 @@ TST_NEEDS_CMDS="awk bc numactl numastat" extract_numastat_p() { local pid=$1 - local node=$(($2 + 2)) - - echo $(numastat -p $pid |awk '/^Total/ {print $'$node'}') + local node=$2 + + echo $(numastat -p $pid | \ + awk -v node=$node '/Node/ { \ + gsub("Node", ""); \ + for (i=0; i<NF; i++) \ + if ($i == node) \Quoted 8 lines
Quoted 8 lines
testcases/kernel/numa/numa01.sh TST_NEEDS_CMDS="awk bc numactl numastat" · line 67
Quoted 6 lines
testcases/kernel/numa/numa01.sh +5
-0+ col=i+1; \ + next \ + } \ + /^Total/ {print $col}' \ + ) }If we can use
extract_numastat_p0withnode_indexthen the existingcode
Quoted 7 lines
Quoted 3 lines
works, for example if can index while iterating $node_list and use this with
extract_numastat_p0then it should work.Hi Bala,
Sorry, I don't fully understand what's you mean here :(. Could you
explain
Quoted 1 line
something more for this method?
instead of changing
extract_numastat_p0()can do something like this ?testcases/kernel/numa/numa01.sh +1
-0@@ -94,6 +94,7 @@ setup() test1() { Mem_curr=0 + node_index=0 for node in $nodes_list; do numactl --cpunodebind=$node --membind=$node support_numa
alloc_1MB&
testcases/kernel/numa/numa01.sh test1() · line 109
Quoted 9 lines
testcases/kernel/numa/numa01.sh +2
-1@@ -101,7 +102,8 @@ test1() TST_RETRY_FUNC "check_for_support_numa $pid" 0 - Mem_curr=$(echo "$(extract_numastat_p $pid $node) * $MB" |bc) + Mem_curr=$(echo "$(extract_numastat_p $pid $node_index) * $MB" |bc) + let node_index++
I guess it can be work, but the disadvantage of that is we have to involve a new variable(node_index) in each of the tests (from test1 to test10). Hence I don't think it is much better than my patch. For which way to go, I'd leave this to Cyril to make a choice. Or, maybe he has different thoughts on this:).
Anyway, thanks for your suggestion.
Signature · 5 lines
Regards, Li Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linux.it/pipermail/ltp/attachments/20190509/55f8a078/attachment.html>
PATCH v2 1/2 numa: fix numa test error with non-continuous nodes
Hi!
Quoted 18 lines
Quoted 18 lines
testcases/kernel/numa/numa01.sh +1
-0@@ -94,6 +94,7 @@ setup() test1() { Mem_curr=0 + node_index=0 for node in $nodes_list; do numactl --cpunodebind=$node --membind=$node support_numa
alloc_1MB&testcases/kernel/numa/numa01.sh +0
-0@@ -101,7 +102,8 @@ test1() TST_RETRY_FUNC "check_for_support_numa $pid" 0
Quoted 12 lines
testcases/kernel/numa/numa01.sh test1() · line 109
Quoted 6 lines
testcases/kernel/numa/numa01.sh +2
-1- Mem_curr=$(echo "$(extract_numastat_p $pid $node) * $MB" |bc) + Mem_curr=$(echo "$(extract_numastat_p $pid $node_index) * $MB" |bc) + let node_index++I guess it can be work, but the disadvantage of that is we have to involve a new
variable(node_index)in each of the tests (from test1 to test10). Hence I don't think it is much better than my patch. For which way to go, I'd leave this to Cyril to make a choice. Or, maybe he has different thoughts on this:).
I actually do not care that much about the numa01.sh tests, because these are broken in more ways than this and were never correct to begin with.
I've started to rewrite these into proper tests, the set_mempolicy() was first part of that effort, the mbind() tests are continuation of that and the end goal is to get rid of these broken tests eventually.
Signature · 2 lines
Cyril Hrubis [email protected]
Hello,
Quoted 32 lines · Cyril Hrubis
Hi!
Quoted 18 lines
Quoted 18 lines
testcases/kernel/numa/numa01.sh +1
-0@@ -94,6 +94,7 @@ setup() test1() { Mem_curr=0 + node_index=0 for node in $nodes_list; do numactl --cpunodebind=$node --membind=$node support_numa
alloc_1MB&testcases/kernel/numa/numa01.sh +0
-0@@ -101,7 +102,8 @@ test1() TST_RETRY_FUNC "check_for_support_numa $pid" 0Quoted 12 lines
testcases/kernel/numa/numa01.sh test1() · line 109
Quoted 6 lines
testcases/kernel/numa/numa01.sh +2
-1- Mem_curr=$(echo "$(extract_numastat_p $pid $node) * $MB" |bc) + Mem_curr=$(echo "$(extract_numastat_p $pid $node_index) * $MB" |bc) + let node_index++I guess it can be work, but the disadvantage of that is we have to involve a new
variable(node_index)in each of the tests (from test1 to test10). Hence I don't think it is much better than my patch. For which way to go, I'd leave this to Cyril to make a choice. Or, maybe he has different thoughts on this:).
Quoted 7 lines
I actually do not care that much about the
numa01.shtests, because these are broken in more ways than this and were never correct to begin with.I've started to rewrite these into proper tests, the
set_mempolicy()was first part of that effort, thembind()tests are continuation of that and the end goal is to get rid of these broken tests eventually.
We have quite a few mbind and set_mempolicy tests now. So should we delete this test?
Quoted 4 lines
Signature · 3 lines
-- Cyril Hrubis [email protected]
Signature · 5 lines
Thank you, Richard. -- Mailing list info: https://lists.linux.it/listinfo/ltp