Re: [GIT PULL v2] Early boot SLAB for 2.6.31

From: Pekka Enberg
Date: Thu Jun 11 2009 - 07:56:23 EST


On Thu, 2009-06-11 at 13:48 +0200, Ingo Molnar wrote:
> * Ingo Molnar <mingo@xxxxxxx> wrote:
>
> > * Ingo Molnar <mingo@xxxxxxx> wrote:
> >
> > > Hm, with this pulled on a testbox i'm still getting:
> > >
> > > [ 0.000000] Experimental hierarchical RCU init done.
> > > [ 0.000000] NR_IRQS:4352 nr_irqs:256
> > > [ 0.000000] ------------[ cut here ]------------
> > > [ 0.000000] WARNING: at mm/bootmem.c:537 alloc_arch_preferred_bootmem+0x40/0x7e()
> >
> > Another testbox crashed on bootup. I'm collecting a serial log
> > from it - config attached meanwhile.
>
> We have a hard crash in the WP-protect code:
>
> [ 0.000000] Checking if this processor honours the WP bit even in supervisor mode...BUG: Int 14: CR2 ffcff000
> [ 0.000000] EDI 00000188 ESI 00000ac7 EBP c17eaf9c ESP c17eaf8c
> [ 0.000000] EBX 000014e0 EDX 0000000e ECX 01856067 EAX 00000001
> [ 0.000000] err 00000003 EIP c10135b1 CS 00000060 flg 00010002
> [ 0.000000] Stack: c17eafa8 c17fd410 c16747bc c17eafc4 c17fd7e5 000011fd f8616000 c18237cc
> [ 0.000000] 00099800 c17bb000 c17eafec c17f1668 000001c5 c17f1322 c166e039 c1822bf0
> [ 0.000000] c166e033 c153a014 c18237cc 00020800 c17eaff8 c17f106a 00020800 01ba5003
> [ 0.000000] Pid: 0, comm: swapper Not tainted 2.6.30-tip-02161-g7a74539-dirty #52203
> [ 0.000000] Call Trace:
> [ 0.000000] [<c15357c2>] ? printk+0x14/0x16
> [ 0.000000] [<c10135b1>] ? do_test_wp_bit+0x19/0x23
> [ 0.000000] [<c17fd410>] ? test_wp_bit+0x26/0x64
> [ 0.000000] [<c17fd7e5>] ? mem_init+0x1ba/0x1d8
> [ 0.000000] [<c17f1668>] ? start_kernel+0x164/0x2f7
> [ 0.000000] [<c17f1322>] ? unknown_bootoption+0x0/0x19c
> [ 0.000000] [<c17f106a>] ? __init_begin+0x6a/0x6f
>
> (full bootlog attached)
>
> Just a quick analysis from the place we crash (without looking into
> any details): the WP test is the first time we really make use of
> the MMU during bootup - crashes there are often a sign of messed up
> pagetables, which is easy if the bootmem allocator is changed.

No, mem_init() happens _before_ slab is set up so it's probably just
this.

And yes, the patch set clearly needs more work.

Pekka

diff --git a/init/main.c b/init/main.c
index 859af21..6d38f96 100644
--- a/init/main.c
+++ b/init/main.c
@@ -588,6 +588,8 @@ asmlinkage void __init start_kernel(void)
*/
pidhash_init();
vfs_caches_init_early();
+ sort_main_extable();
+ trap_init();
/*
* Set up kernel memory allocators
*/
@@ -610,8 +612,6 @@ asmlinkage void __init start_kernel(void)
"enabled *very* early, fixing it\n");
local_irq_disable();
}
- sort_main_extable();
- trap_init();
rcu_init();
/* init some links before init_ISA_irqs() */
early_irq_init();


--
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/