Re: [patch 1/2] x86, fpu: split FPU state from task struct - v2

[Ingo Molnar]

* Ingo Molnar <mingo@xxxxxxx> wrote:

> > 2) allocate the right size for the actual cpu rather than 512 bytes 
> > always. Patches enabling xsave/xrstor support (coming shortly) will 
> > take advantage of this.
> 
> thanks, i've applied your two patches to x86.git#testing.

hm, qa immediately got problems with it:

[   20.054821] BUG: sleeping function called from invalid context at mm/slab.c:3055
[   20.057253] in_atomic():0, irqs_disabled():1
[   20.057253] no locks held by awk/1294.
[   20.057253] irq event stamp: 5745
[   20.057253] hardirqs last  enabled at (5745): [<ffffffff8079aaf0>] trace_hardirqs_on_thunk+0x35/0x3a
[   20.057253] hardirqs last disabled at (5744): [<ffffffff8079ab2a>] trace_hardirqs_off_thunk+0x35/0x37
[   20.057253] softirqs last  enabled at (0): [<ffffffff80237800>] copy_process+0x310/0x16b0
[   20.057253] softirqs last disabled at (0): [<0000000000000000>] 0x0
[   20.057253] Pid: 1294, comm: awk Not tainted 2.6.25-rc3 #2
[   20.057253] 
[   20.057253] Call Trace:
[   20.057253]  [<ffffffff8025bfb0>] ? print_irqtrace_events+0x110/0x120
[   20.057253]  [<ffffffff80230de5>] __might_sleep+0xe5/0x110
[   20.057253]  [<ffffffff802a9c53>] kmem_cache_alloc+0xe3/0x100
[   20.057253]  [<ffffffff80213f84>] init_fpu+0x94/0xf0
[   20.057253]  [<ffffffff8020e6d8>] math_state_restore+0x58/0x70
[   20.057253]  [<ffffffff8079bb39>] error_exit+0x0/0xb8
[   20.057253] 
[   35.722463] device: 'vcs1': device_add

that's a non-atomic alloc in an atomic (irqs off) section ...

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/