Re: mm snapshot broken-out-2007-04-25-02-49.tar.gz uploaded

[Jeremy Fitzhardinge <jeremy@goop.org>]

Michal Piotrowski wrote:
> Timing information is broken
>   

This looks like Andi's sched_clock stuff hitting: it returns an
inherently per-cpu result, so using it for printk timestamps isn't very
useful.

> [4294667.296007]      soft-irqs-on + irq-safe-A/21:  ok  |  ok  |  ok  |
> [4294667.296007]        sirq-safe-A => hirqs-on/12:  ok  |  ok  |irq event stamp: 472
> [4294667.296008] hardirqs last  enabled at (472): [<c020fe3a>] irqsafe2A_rlock_12+0xb6/0xc7
> [4294667.296008] hardirqs last disabled at (471): [<c010a9f1>] sched_clock+0x78/0x120
> [4294667.296008] softirqs last  enabled at (468): [<c020fe25>] irqsafe2A_rlock_12+0xa1/0xc7
> [4294667.296008] softirqs last disabled at (464): [<c020fd90>] irqsafe2A_rlock_12+0xc/0xc7
> [4294667.296008] FAILED| [<c01052e0>] dump_trace+0x63/0x1eb
> [4294667.296008]  [<c0105482>] show_trace_log_lvl+0x1a/0x30
> [4294667.296008]  [<c01060c4>] show_trace+0x12/0x14
> [4294667.296008]  [<c0106161>] dump_stack+0x16/0x18
> [4294667.296008]  [<c020da4f>] dotest+0x8d/0x3f4
> [4294667.296008]  [<c02181dd>] locking_selftest+0x915/0x1a58
> [4294667.296008]  [<c04f88ed>] start_kernel+0x277/0x349
> [4294667.296008]  =======================
>   

This is a bit mysterious, in that it seems to be triggered by my
softlockup watchdog patch, which doesn't seem to have anything to do
with the locking tests at all.  Filed under "Ingo's mystery".

> [4294667.511000] BUG: at arch/i386/kernel/sched-clock.c:175 init_sched_clock()
> [4294667.511000]  [<c01052e0>] dump_trace+0x63/0x1eb
> [4294667.511000]  [<c0105482>] show_trace_log_lvl+0x1a/0x30
> [4294667.511000]  [<c01060c4>] show_trace+0x12/0x14
> [4294667.511000]  [<c0106161>] dump_stack+0x16/0x18
> [4294667.511000]  [<c04fd968>] init_sched_clock+0x6d/0xcc
> [4294667.512000]  [<c04f8505>] kernel_init+0x14e/0x2bf
> [4294667.512000]  [<c0104ec3>] kernel_thread_helper+0x7/0x10
> [4294667.512000]  =======================

This also seems to be the sched_clock change; the WARN_ON is asserting
that it got initialized early enough, but it isn't. 

    J