[Bug runtime/4458] New: lockdep annotations

[Bug runtime/4458] New: lockdep annotations

[fche at redhat dot com]

When lockdep logic is on, systemtap modules trigger warnings for the lock objects
that are allocated dynamically.  We need to annotate these lock variables with
code like that in <http://article.gmane.org/gmane.linux.irda.general/1148> for
lockdep to work properly.

-- 
           Summary: lockdep annotations
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: runtime
        AssignedTo: systemtap at sources dot redhat dot com
        ReportedBy: fche at redhat dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep annotations

[joshua dot i dot stone at intel dot com]

------- Additional Comments From joshua dot i dot stone at intel dot com  2007-05-03 16:52 -------
Do you know which locks are the culprit?

There was a similar bug #2989 for the seqlocks used in the time code, which
should be resolved in 2.6.20+ kernels.

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep annotations

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2007-05-03 17:09 -------
Created an attachment (id=1752)
 --> (http://sourceware.org/bugzilla/attachment.cgi?id=1752&action=view)
lockdep warning text


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep annotations

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2007-05-03 17:09 -------
(In reply to comment #1)
> Do you know which locks are the culprit?

In this case, some lock within the pmap logic.  It came from the
examples/iotime.stp, 

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep annotations

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2007-05-03 17:11 -------
Actually, this may be solved automatically if we remove those
locks entirely, a la bug #2224.  (They are redundant since the
translator already emits lock calls for pmap accesses.)


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep annotations

[hunt at redhat dot com]

------- Additional Comments From hunt at redhat dot com  2007-05-03 17:18 -------
(In reply to comment #4)
> Actually, this may be solved automatically if we remove those
> locks entirely, a la bug #2224.  (They are redundant since the
> translator already emits lock calls for pmap accesses.)
> 
I was just thinking the same thing. Thought we removed those. I'll take 2224 and
confirm the stap-generated code always does the appropriate locking now.


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep warnings

[fche at redhat dot com]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|lockdep annotations         |lockdep warnings


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep annotations

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2007-05-29 19:39 -------
With bug #2244 fixed, lockdep on 2.6.21-rc6-mm1 complains about another runtime
issue while running the pfaults.stp test case:

=================================
[ INFO: inconsistent lock state ]
2.6.21-rc6-mm1 #6
---------------------------------
inconsistent {softirq-on-W} -> {in-softirq-W} usage.
swapper/0 [HC0[0]:SC1[1]:HE0:SE0] takes:
 (old_style_spin_init#6){-+..}, at: [<f0aadb0f>] _stp_print_flush+0x49/0xb7
[stap_c2f25c0293a8b5dd273068154fe922d
0_3173]
{softirq-on-W} state was registered at:
  [<c043e286>] mark_lock+0x6d/0x406
  [<c043f0be>] __lock_acquire+0x44b/0xb3e
  [<c043fb70>] lock_acquire+0x56/0x6f
  [<c05eff05>] _spin_lock+0x2b/0x38
  [<f0aadb0f>] _stp_print_flush+0x49/0xb7
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<f0aae6b1>] probe_1533+0x15d/0x1a2 [stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<f0aae10b>] enter_begin_probe+0x9d/0x14f
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<f0ab09c5>] systemtap_module_init+0x1e9/0x517
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<f0ab0cfe>] probe_start+0xb/0x14 [stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<f0ab0d2a>] _stp_handle_start+0x23/0x82
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<f0ab0f14>] _stp_ctl_write_cmd+0x18b/0x718
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
  [<c047dc18>] vfs_write+0xaf/0x163
  [<c047e268>] sys_write+0x3d/0x61
  [<c0404eb4>] syscall_call+0x7/0xb
  [<ffffe402>] 0xffffe402
  [<ffffffff>] 0xffffffff
irq event stamp: 2164189
hardirqs last  enabled at (2164188): [<c05f01e0>] _spin_unlock_irq+0x25/0x2a
hardirqs last disabled at (2164189): [<f0ab14ca>] enter_hrtimer_probe+0x29/0x334
[stap_c2f25c0293a8b5dd273068154f
e922d0_3173]
softirqs last  enabled at (2164174): [<c0427fbe>] __do_softirq+0xdc/0xe4
softirqs last disabled at (2164183): [<c0428003>] do_softirq+0x3d/0x56

other info that might help us debug this:
3 locks held by swapper/0:
 #0:  (& global_pidnames_lock){....}, at: [<f0ab1c41>] probe_1531+0x2c/0xdf6
[stap_c2f25c0293a8b5dd273068154fe922
d0_3173]
 #1:  (& global_faults_lock){....}, at: [<f0ab1c73>] probe_1531+0x5e/0xdf6
[stap_c2f25c0293a8b5dd273068154fe922d0
_3173]
 #2:  (& global_fault_types_lock){....}, at: [<f0ab1ca5>] probe_1531+0x90/0xdf6
[stap_c2f25c0293a8b5dd273068154fe
922d0_3173]

stack backtrace:
 [<c0405ee1>] dump_trace+0x63/0x1eb
 [<c0406083>] show_trace_log_lvl+0x1a/0x30
 [<c0406af7>] show_trace+0x12/0x14
 [<c0406b50>] dump_stack+0x16/0x18
 [<c043dbe1>] print_usage_bug+0x140/0x14a
 [<c043e333>] mark_lock+0x11a/0x406
 [<c043f02f>] __lock_acquire+0x3bc/0xb3e
 [<c043fb70>] lock_acquire+0x56/0x6f
 [<c05eff05>] _spin_lock+0x2b/0x38
 [<f0aadb0f>] _stp_print_flush+0x49/0xb7
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
 [<f0ab2911>] probe_1531+0xcfc/0xdf6 [stap_c2f25c0293a8b5dd273068154fe922d0_3173]
 [<f0ab16ac>] enter_hrtimer_probe+0x20b/0x334
[stap_c2f25c0293a8b5dd273068154fe922d0_3173]
 [<c0436fde>] hrtimer_run_queues+0x23e/0x290
 [<c042adab>] run_timer_softirq+0x24/0x189
 [<c0427f51>] __do_softirq+0x6f/0xe4
 [<c0428003>] do_softirq+0x3d/0x56
 [<c04283a8>] irq_exit+0x4a/0x7e
 [<c0415be9>] smp_apic_timer_interrupt+0x77/0x85
 [<c0405977>] apic_timer_interrupt+0x33/0x38
 [<c0403c7a>] default_idle+0x43/0x5c
 [<c040340b>] cpu_idle+0xae/0xcf
 [<c040140a>] rest_init+0x26/0x28
 [<c07719c0>] start_kernel+0x34d/0x355


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug runtime/4458] lockdep warnings

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2007-05-29 22:08 -------
Patch committed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED


http://sourceware.org/bugzilla/show_bug.cgi?id=4458

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.