Re: Linus GIT 3.17.0+ - BUG: unable to handle kernel paging request - IP: [<ffffffff81109e8d>] scan_block+0x59/0x100

[Catalin Marinas]

Hi Miles,

On 15 October 2014 04:43, Miles Lane <miles.lane@xxxxxxxxx> wrote:
> [   68.164917] BUG: unable to handle kernel paging request at ffff88021dc52000
> [   68.164957] IP: [<ffffffff81109e8d>] scan_block+0x59/0x100
> [   68.164987] PGD 2902067 PUD 2905067 PMD 21fdf2067 PTE 800000021dc52060

My reading of the x86 _PAGE_BIT_* macros show that this is not present.

> [   68.165020] Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC

And this could be caused by CONFIG_DEBUG_PAGEALLOC, though kmemleak
should not try to access such memory.

> [   68.165945] Call Trace:
> [   68.165960]  [<ffffffff81528559>] ? _raw_spin_lock_irqsave+0x48/0x52
> [   68.165988]  [<ffffffff8110a016>] ? scan_gray_list+0xe2/0x15f
> [   68.166015]  [<ffffffff81109feb>] scan_gray_list+0xb7/0x15f
> [   68.166040]  [<ffffffff8110a32f>] kmemleak_scan+0x29c/0x466
> [   68.166066]  [<ffffffff8110a84d>] ? kmemleak_write+0x354/0x354
> [   68.166092]  [<ffffffff8110a8cf>] kmemleak_scan_thread+0x82/0xa4

I have a couple of questions:

1. Are there any messages prior to the dump you posted, like kmemleak
getting disabled?
2. Can you still reproduce the problem with !CONFIG_DEBUG_PAGEALLOC?

Thanks,

Catalin
--
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/