Fwd: [btrfs/i_size] xfstests generic/299 TFAIL
by Fengguang Wu
----- Forwarded message from Fengguang Wu <fengguang.wu(a)intel.com> -----
Date: Thu, 30 Jan 2014 11:54:40 +0800
From: Fengguang Wu <fengguang.wu(a)intel.com>
To: Steven Whitehouse <swhiteho(a)redhat.com>
Cc: Al Viro <viro(a)zeniv.linux.org.uk>, linux-fsdevel(a)vger.kernel.org, "linux-btrfs(a)vger.kernel.org" <linux-btrfs(a)vger.kernel.org>, LKML
<linux-kernel(a)vger.kernel.org>
Subject: [btrfs/i_size] xfstests generic/299 TFAIL
User-Agent: Heirloom mailx 12.5 6/20/10
Hi Steven,
We noticed xfstests generic/299 TFAIL on btrfs since
commit 9fe55eea7e4b444bafc42fa0000cc2d1d2847275
Author: Steven Whitehouse <swhiteho(a)redhat.com>
AuthorDate: Fri Jan 24 14:42:22 2014 +0000
Commit: Al Viro <viro(a)zeniv.linux.org.uk>
CommitDate: Sun Jan 26 08:26:42 2014 -0500
Fix race when checking i_size on direct i/o read
More changes that might help debugging:
2796e4cec525a2b 9fe55eea7e4b444bafc42fa00
--------------- -------------------------
0 +Inf% 1 ~ 0% xfstests.generic.299.fail
6601 ~11% +55547.3% 3673721 ~18% slabinfo.btrfs_extent_map.active_objs
49 ~ 6% +6181.0% 3115 ~19% slabinfo.btrfs_extent_buffer.num_slabs
85 ~18% +776.4% 750 ~14% slabinfo.buffer_head.num_slabs
30584 ~ 0% +1105.5% 368688 ~ 0% time.maximum_resident_set_size
85 ~18% +776.4% 750 ~14% slabinfo.buffer_head.active_slabs
3367 ~18% +769.2% 29268 ~14% slabinfo.buffer_head.num_objs
3304 ~19% +783.1% 29180 ~14% slabinfo.buffer_head.active_objs
49 ~ 6% +6181.0% 3115 ~19% slabinfo.btrfs_extent_buffer.active_slabs
1249 ~ 6% +6134.8% 77897 ~19% slabinfo.btrfs_extent_buffer.num_objs
1102 ~ 3% +6957.3% 77771 ~19% slabinfo.btrfs_extent_buffer.active_objs
255 ~11% +55224.5% 141298 ~18% slabinfo.btrfs_extent_map.num_slabs
255 ~11% +55224.5% 141298 ~18% slabinfo.btrfs_extent_map.active_slabs
6645 ~10% +55181.5% 3673784 ~18% slabinfo.btrfs_extent_map.num_objs
2850 ~ 7% +434.8% 15242 ~ 9% slabinfo.ext4_extent_status.num_objs
2841 ~ 8% +429.5% 15047 ~10% slabinfo.ext4_extent_status.active_objs
44659 ~ 2% +1329.9% 638573 ~17% meminfo.SReclaimable
61541 ~ 2% +964.6% 655186 ~17% meminfo.Slab
27 ~ 8% +447.8% 149 ~ 9% slabinfo.ext4_extent_status.num_slabs
9188 ~ 3% +666.4% 70420 ~ 9% interrupts.TLB
2642 ~ 5% +425.0% 13874 ~14% slabinfo.ext3_xattr.active_objs
2662 ~ 5% +424.9% 13973 ~14% slabinfo.ext3_xattr.num_objs
57 ~ 5% +428.2% 303 ~14% slabinfo.ext3_xattr.num_slabs
57 ~ 5% +428.2% 303 ~14% slabinfo.ext3_xattr.active_slabs
27 ~ 8% +447.8% 149 ~ 9% slabinfo.ext4_extent_status.active_slabs
0 ~ 0% +Inf% 138193 ~ 0% proc-vmstat.unevictable_pgs_culled
379 ~13% +45684.1% 173705 ~ 0% proc-vmstat.pgdeactivate
8107 ~16% +3196.9% 267299 ~ 0% proc-vmstat.pgactivate
11160 ~ 2% +1329.0% 159479 ~17% proc-vmstat.nr_slab_reclaimable
6577 ~ 3% +387.4% 32059 ~24% proc-vmstat.nr_tlb_remote_flush
6684 ~ 3% +380.8% 32142 ~24% proc-vmstat.nr_tlb_remote_flush_received
15707 ~31% +282.3% 60043 ~17% meminfo.Dirty
6380554 ~ 0% +259.8% 22954274 ~ 7% proc-vmstat.pgfault
22901 ~ 3% +290.9% 89514 ~18% proc-vmstat.nr_active_file
4067 ~29% +268.0% 14966 ~17% proc-vmstat.nr_dirty
91655 ~ 3% +291.3% 358640 ~18% meminfo.Active(file)
3088362 ~ 0% +211.5% 9618749 ~ 6% proc-vmstat.pgalloc_dma32
3090040 ~ 0% +211.3% 9619232 ~ 6% proc-vmstat.pgfree
3046221 ~ 0% +211.2% 9479249 ~ 6% proc-vmstat.numa_local
3046221 ~ 0% +211.2% 9479249 ~ 6% proc-vmstat.numa_hit
23371 ~ 3% +218.6% 74472 ~29% softirqs.TIMER
51894 ~ 2% +202.5% 156994 ~23% interrupts.LOC
207400 ~ 2% +142.2% 502386 ~10% meminfo.Active
101124 ~ 1% +151.8% 254632 ~17% proc-vmstat.nr_tlb_local_flush_all
30294 ~ 8% -50.7% 14930 ~17% slabinfo.btrfs_extent_state.active_objs
725 ~ 7% -49.5% 366 ~15% slabinfo.btrfs_extent_state.num_slabs
725 ~ 7% -49.5% 366 ~15% slabinfo.btrfs_extent_state.active_slabs
30490 ~ 7% -49.5% 15409 ~15% slabinfo.btrfs_extent_state.num_objs
63861 ~11% +90.7% 121757 ~ 9% softirqs.RCU
849659 ~ 1% +105.7% 1747978 ~15% proc-vmstat.nr_tlb_local_flush_one
1034500 ~ 0% +94.1% 2007885 ~ 3% proc-vmstat.pgpgin
232831 ~14% +90.8% 444281 ~13% interrupts.RES
169 ~ 3% +91.2% 323 ~15% uptime.boot
7332 ~ 8% +104.1% 14968 ~36% softirqs.SCHED
59342 ~17% +60.4% 95197 ~23% interrupts.43:PCI-MSI-edge.virtio1-requests
555 ~ 8% +70.4% 946 ~13% slabinfo.blkdev_requests.num_objs
526 ~ 7% +65.0% 867 ~18% slabinfo.kmalloc-2048.active_objs
525 ~ 8% +66.0% 872 ~15% slabinfo.blkdev_requests.active_objs
648109 ~ 1% -36.8% 409436 ~ 9% proc-vmstat.nr_free_pages
2594146 ~ 1% -36.9% 1635776 ~ 9% meminfo.MemFree
603 ~ 8% +60.5% 968 ~16% slabinfo.kmalloc-2048.num_objs
2587973 ~ 1% -36.7% 1637486 ~ 9% vmstat.memory.free
433 ~ 4% +71.8% 745 ~25% uptime.idle
104603 ~ 0% +49.4% 156274 ~ 9% proc-vmstat.nr_unevictable
418413 ~ 0% +49.3% 624828 ~ 9% meminfo.Unevictable
81418 ~ 0% -25.4% 60757 ~ 2% proc-vmstat.nr_dirty_background_threshold
162839 ~ 0% -25.4% 121516 ~ 2% proc-vmstat.nr_dirty_threshold
956619 ~12% +30.5% 1248532 ~11% proc-vmstat.nr_written
968744 ~12% +29.9% 1258046 ~11% proc-vmstat.nr_dirtied
12837 ~ 7% -23.1% 9877 ~17% interrupts.IWI
305754 ~ 3% +27.7% 390352 ~ 4% proc-vmstat.nr_file_pages
2490 ~11% +24.1% 3089 ~ 6% slabinfo.kmalloc-96.num_objs
1221055 ~ 3% +19.2% 1455334 ~ 5% meminfo.Cached
1223056 ~ 3% +19.0% 1455025 ~ 5% vmstat.memory.cache
172852 ~ 6% -20.0% 138300 ~12% proc-vmstat.nr_inactive_file
689411 ~ 5% -19.7% 553897 ~12% meminfo.Inactive(file)
2471 ~11% +18.8% 2935 ~ 6% slabinfo.kmalloc-96.active_objs
711198 ~ 5% -18.6% 579097 ~12% meminfo.Inactive
42.28 ~21% +367.9% 197.85 ~10% time.system_time
5.06 ~ 6% +616.9% 36.29 ~ 9% time.user_time
5711222 ~ 0% +279.1% 21648853 ~ 8% time.minor_page_faults
32 ~16% +148.4% 80 ~19% time.percent_of_cpu_this_job_got
85616 ~27% +110.2% 179944 ~18% time.involuntary_context_switches
2067193 ~ 0% +94.1% 4013246 ~ 3% time.file_system_inputs
144 ~ 4% +106.7% 298 ~16% time.elapsed_time
144296 ~ 4% -53.4% 67248 ~17% vmstat.io.bo
41865918 ~ 2% -11.7% 36960769 ~ 7% time.file_system_outputs
Thanks,
Fengguang
----- End forwarded message -----
7 years, 2 months
Re: [LKP] x86 idle: pigz.throughput regression
by Brown, Len
> Is this on a NHM-EX or WSM-EX machine?
>
> The regressed test case is lkp-nex04/micro/pigz/100%, where lkp-nex04 is
> a NHM-EX machine.
>
> v3.13-rc4 40e2d7f9b5dae048789c64672
> --------------- -------------------------
> 414 ~ 0% -6.1% 389 ~ 0% pigz.throughput
What do these numbers mean?
What is the typical variance in this test when
run multiple times?
> > How much idle time/idle transitions are there in this workload?
>
> CPU idle time is 12%. What do you mean by idle transitions and how to
> measure it?
Transitions to idle are counted in
/sys/devices/system/cpu/cpu*/cpuidle/*/usage
so you can snapshot those counts,
run the workload, snapshot again,
and the difference is the number of transitions
by each processor to each state during the test.
Also interesting would be the output on this machine of
dmesg | grep idle
grep . /sys/devices/system/cpu/cpu*/cpuidle/*/*
and the output from
# turbostat -v {pigz invocation here} > pigz.out 2>&1
Can you tell me exactly how to run this test
so that I can reproduce on my WSM-EX?
thanks,
-Len
>
> Detailed numbers in
> /lkp/result/lkp-nex04/micro/pigz/100%/x86_64-
> lkp/40e2d7f9b5dae048789c64672bf3027fbb663ffa/matrix.json:
>
> "vmstat.cpu.us": [
> 86,
> 87,
> 86,
> 86,
> 86
> ],
> "vmstat.cpu.sy": [
> 1,
> 1,
> 1,
> 1,
> 1
> ],
> "vmstat.cpu.id": [
> 12,
> 11,
> 12,
> 12,
> 12
> ],
>
> Thanks,
> Fengguang
7 years, 3 months
