Re: [RFC][PATCH 1/3] radix priority search tree - objrmap complexity fix

[Andrea Arcangeli]

Let's forget the "should we allow people to use rw_swap_page_sync to
swapout/swapin anonymous pages" discussion, there's a major issue that
my latest patch still doesn't work:

Writing data to swap (5354 pages): .<1>Unable to handle kernel NULL pointer dereference at virtual address 00000004
 printing eip:
c01d9b34
*pde = 00000000
Oops: 0000 [#1]
CPU:    0
EIP:    0060:[<c01d9b34>]    Not tainted
EFLAGS: 00010082   (2.6.4-41.8-default)
EIP is at radix_tree_delete+0x14/0x160
eax: 00000004   ebx: c10361c0   ecx: 00000016   edx: 000023ee
esi: 000023ee   edi: 00000000   ebp: 000000d0   esp: cdee5e1c
ds: 007b   es: 007b   ss: 0068
Process bash (pid: 1, threadinfo=cdee4000 task=cdf9d7b0)
Stack: 00000000 f7b0d200 00000004 00000016 c041d440 c03ffe2e c0108d48 c041d440
       00000000 000003fd 000026b6 c041d440 c03ffe2e 00000320 0000007b ffff007b
       ffffff00 c021a39e 00000060 c10361c0 c0341d20 00000056 00000056 00000056
Call Trace:
 [<c0108d48>] common_interrupt+0x18/0x20
 [<c021a39e>] serial_in+0x1e/0x40
 [<c014fc3c>] swap_free+0x1c/0x30
 [<c0151597>] remove_exclusive_swap_page+0x97/0x155
 [<c013bc1f>] __remove_from_page_cache+0x3f/0xa0
 [<c013bc9b>] remove_from_page_cache+0x1b/0x27
 [<c014eb59>] rw_swap_page_sync+0xa9/0x1d0
 [<c013588d>] do_magic_suspend_2+0x27d/0x7d0
 [<c0275c2d>] do_magic+0x4d/0x130
 [<c0135310>] software_suspend+0xd0/0xe0
 [<c01fad86>] acpi_system_write_sleep+0xb5/0xd2
 [<c01facd1>] acpi_system_write_sleep+0x0/0xd2
 [<c0153e4e>] vfs_write+0xae/0xf0
 [<c0153f2c>] sys_write+0x2c/0x50
 [<c0107dc9>] sysenter_past_esp+0x52/0x79

Code: 8b 28 8d 7c 24 10 3b 14 ad 00 99 41 c0 0f 87 18 01 00 00 8d
 <0>Kernel panic: Attempted to kill init!


Pavel told me a SMP kernel cannot suspend, that's probably why I
couldn't reproduce, I'll recompile UP and hopefully I will be able to
reproduce, so I can debug it, and I can try latest Andrew's patch too
(the one allowing anonymous memory swapin/swapouts too).
-
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/