Skip to content

L
linux
Switch branch/tag
  1. 17 Oct, 2023 14 commits
  3. 12 Oct, 2023 23 commits
  5. 10 Oct, 2023 1 commit
  7. 05 Oct, 2023 2 commits
    • Kajol Jain's avatar
      tools/perf: Update call stack check in builtin-lock.c · d7c9ae8d
      Kajol Jain authored 
      The perf test named "kernel lock contention analysis test"
fails in powerpc system with below error:

  [command]# ./perf test 81 -vv
   81: kernel lock contention analysis test                            :
   --- start ---
  test child forked, pid 2140
  Testing perf lock record and perf lock contention
  Testing perf lock contention --use-bpf
  [Skip] No BPF support
  Testing perf lock record and perf lock contention at the same time
  Testing perf lock contention --threads
  Testing perf lock contention --lock-addr
  Testing perf lock contention --type-filter (w/ spinlock)
  Testing perf lock contention --lock-filter (w/ tasklist_lock)
  Testing perf lock contention --callstack-filter (w/ unix_stream)
  [Fail] Recorded result should have a lock from unix_stream:
  test child finished with -1
   ---- end ----
  kernel lock contention analysis test: FAILED!

The test is failing because we get an address entry with 0 in
perf lock samples for powerpc, and code for lock contention
option "--callstack-filter" will not check further entries after
address 0.

Below are some of the samples from test generated perf.data file, which
have 0 address in the 2nd entry of callstack:
 --------
sched-messaging    3409 [001]  7152.904029: lock:contention_begin: 0xc00000c80904ef00 (flags=SPIN)
        c0000000001e926c __traceiter_contention_begin+0x6c ([kernel.kallsyms])
                       0 [unknown] ([unknown])
        c000000000f8a178 native_queued_spin_lock_slowpath+0x1f8 ([kernel.kallsyms])
        c000000000f89f44 _raw_spin_lock_irqsave+0x84 ([kernel.kallsyms])
        c0000000001d9fd0 prepare_to_wait+0x50 ([kernel.kallsyms])
        c000000000c80f50 sock_alloc_send_pskb+0x1b0 ([kernel.kallsyms])
        c000000000e82298 unix_stream_sendmsg+0x2b8 ([kernel.kallsyms])
        c000000000c78980 sock_sendmsg+0x80 ([kernel.kallsyms])

sched-messaging    3408 [005]  7152.904036: lock:contention_begin: 0xc00000c80904ef00 (flags=SPIN)
        c0000000001e926c __traceiter_contention_begin+0x6c ([kernel.kallsyms])
                       0 [unknown] ([unknown])
        c000000000f8a178 native_queued_spin_lock_slowpath+0x1f8 ([kernel.kallsyms])
        c000000000f89f44 _raw_spin_lock_irqsave+0x84 ([kernel.kallsyms])
        c0000000001d9fd0 prepare_to_wait+0x50 ([kernel.kallsyms])
        c000000000c80f50 sock_alloc_send_pskb+0x1b0 ([kernel.kallsyms])
        c000000000e82298 unix_stream_sendmsg+0x2b8 ([kernel.kallsyms])
        c000000000c78980 sock_sendmsg+0x80 ([kernel.kallsyms])
 --------

Based on commit 20002ded ("perf_counter: powerpc: Add callchain support"),
incase of powerpc, the callchain saved by kernel always includes first
three entries as the NIP (next instruction pointer), LR (link register), and
the contents of LR save area in the second stack frame. In certain scenarios
its possible to have invalid kernel instruction addresses in either of LR or the
second stack frame's LR. In that case, kernel will store the address as zer0.
Hence, its possible to have 2nd or 3rd callstack entry as 0.

As per the current code in match_callstack_filter function, we skip the callstack
check incase we get 0 address. And hence the test case is failing in powerpc.

Fix this issue by updating the check in match_callstack_filter function,
to not skip callstack check if the 2nd or 3rd entry have 0 address
for powerpc.

Result in powerpc after patch changes:

  [command]# ./perf test 81 -vv
   81: kernel lock contention analysis test                            :
   --- start ---
  test child forked, pid 4570
  Testing perf lock record and perf lock contention
  Testing perf lock contention --use-bpf
  [Skip] No BPF support
  Testing perf lock record and perf lock contention at the same time
  Testing perf lock contention --threads
  Testing perf lock contention --lock-addr
  Testing perf lock contention --type-filter (w/ spinlock)
  Testing perf lock contention --lock-filter (w/ tasklist_lock)
  [Skip] Could not find 'tasklist_lock'
  Testing perf lock contention --callstack-filter (w/ unix_stream)
  Testing perf lock contention --callstack-filter with task aggregation
  Testing perf lock contention CSV output
  [Skip] No BPF support
  test child finished with 0
   ---- end ----
  kernel lock contention analysis test: Ok

Fixes: ebab2916 ("perf lock contention: Support filters for different aggregation")
Reported-by: default avatarDisha Goel <disgoel@linux.vnet.ibm.com>
Tested-by: default avatarDisha Goel <disgoel@linux.ibm.com>
Signed-off-by: default avatarKajol Jain <kjain@linux.ibm.com>
Cc: maddy@linux.ibm.com
Cc: atrajeev@linux.vnet.ibm.com
Link: https://lore.kernel.org/r/20231003092113.252380-1-kjain@linux.ibm.comSigned-off-by: default avatarNamhyung Kim <namhyung@kernel.org>
      d7c9ae8d
    • Athira Rajeev's avatar
      tools/perf/tests: Fix object code reading to skip address that falls out of text section · 8f5b62a1
      Athira Rajeev authored 
      The testcase "Object code reading" fails in somecases
for "fs_something" sub test as below:

    Reading object code for memory address: 0xc008000007f0142c
    File is: /lib/modules/6.5.0-rc3+/kernel/fs/xfs/xfs.ko
    On file address is: 0x1114cc
    Objdump command is: objdump -z -d --start-address=0x11142c --stop-address=0x1114ac /lib/modules/6.5.0-rc3+/kernel/fs/xfs/xfs.ko
    objdump read too few bytes: 128
    test child finished with -1

This can alo be reproduced when running perf record with
workload that exercises fs_something() code. In the test
setup, this is exercising xfs code since root is xfs.

    # perf record ./a.out
    # perf report -v |grep "xfs.ko"
      0.76% a.out /lib/modules/6.5.0-rc3+/kernel/fs/xfs/xfs.ko  0xc008000007de5efc B [k] xlog_cil_commit
      0.74% a.out  /lib/modules/6.5.0-rc3+/kernel/fs/xfs/xfs.ko  0xc008000007d5ae18 B [k] xfs_btree_key_offset
      0.74% a.out  /lib/modules/6.5.0-rc3+/kernel/fs/xfs/xfs.ko  0xc008000007e11fd4 B [k] 0x0000000000112074

Here addr "0xc008000007e11fd4" is not resolved. since this is a
kernel module, its offset is from the DSO. Xfs module is loaded
at 0xc008000007d00000

   # cat /proc/modules | grep xfs
    xfs 2228224 3 - Live 0xc008000007d00000

And size is 0x220000. So its loaded between  0xc008000007d00000
and 0xc008000007f20000. From objdump, text section is:
    text 0010f7bc  0000000000000000 0000000000000000 000000a0 2**4

Hence perf captured ip maps to 0x112074 which is:
( ip - start of module ) + a0

This offset 0x112074 falls out .text section which is up to 0x10f7bc
In this case for module, the address 0xc008000007e11fd4 is pointing
to stub instructions. This address range represents the module stubs
which is allocated on module load and hence is not part of DSO offset.

To address this issue in "object code reading", skip the sample if
address falls out of text section and is within the module end.
Use the "text_end" member of "struct dso" to do this check.

To address this issue in "perf report", exploring an option of
having stubs range as part of the /proc/kallsyms, so that perf
report can resolve addresses in stubs range

However this patch uses text_end to skip the stub range for
Object code reading testcase.
Reported-by: default avatarDisha Goel <disgoel@linux.ibm.com>
Signed-off-by: default avatarAthira Rajeev <atrajeev@linux.vnet.ibm.com>
Tested-by: Disha Goel<disgoel@linux.ibm.com>
Reviewed-by: default avatarAdrian Hunter <adrian.hunter@intel.com>
Reviewed-by: default avatarKajol Jain <kjain@linux.ibm.com>
Cc: maddy@linux.ibm.com
Cc: disgoel@linux.vnet.ibm.com
Cc: linuxppc-dev@lists.ozlabs.org
Link: https://lore.kernel.org/r/20230928075213.84392-3-atrajeev@linux.vnet.ibm.comSigned-off-by: default avatarNamhyung Kim <namhyung@kernel.org>
      8f5b62a1
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7