Re: [PowerPC] 2.6.30-git14 boot failure with SLAB
From: Sachin Sant
Date: Sat Jun 20 2009 - 03:26:53 EST
Benjamin Herrenschmidt wrote:
That is strange. If I revert that commit, I get breakages on machines
here. It would be interesting to understand what the problem is here,
as we -do- use that kmem cache for allocating page tables, so we do
need it initialized that early. (IE, we can't allow vmalloc for example
to be called before the page table caches are initialized).
This will need more debugging and understanding as to why it hangs.
Hi Ben,
Looks like the control enters pgtable_cache_init but rever returns. The
machine just hangs. I triggered a system reset via HMC to see what's
happening on the cpu. Here is the xmon o/p after a system reset.
The code that was executed was __mutex_lock_slowpath..
cpu 0x0: Vector: 100 (System Reset) at [c000000000b138e0]
pc: c00000000060a4b8: .__mutex_lock_slowpath+0x9c/0x1f4
lr: c00000000060abc8: .mutex_lock+0x50/0x70
sp: c000000000b13b60
msr: 8000000000081032
current = 0xc000000000a3ab70
paca = 0xc000000000be2400
pid = 0, comm = swapper
enter ? for help
[c000000000b13c30] c00000000060abc8 .mutex_lock+0x50/0x70
[c000000000b13cb0] c00000000008c7f0 .get_online_cpus+0x4c/0x84
[c000000000b13d40] c00000000014a120 .kmem_cache_create+0xcc/0x5f4
[c000000000b13e50] c000000000033f38 .pgtable_cache_init+0x28/0x78
[c000000000b13ee0] c0000000008809a4 .start_kernel+0x1f8/0x568
[c000000000b13f90] c0000000000083d8 .start_here_common+0x1c/0x44
0:mon>
0:mon> di $.__mutex_lock_slowpath
c00000000060a41c fba1ffe8 std r29,-24(r1)
c00000000060a420 7c0802a6 mflr r0
.... SNIP .....
c00000000060a46c 7fe4fb78 mr r4,r31
c00000000060a470 419e0014 beq cr7,c00000000060a484 # .__mutex_lock_slowpath+0x68/0x1f4
c00000000060a474 4ba6859d bl c000000000072a10 # .mutex_spin_on_owner+0x0/0xbc
c00000000060a478 60000000 nop
c00000000060a47c 2fa30000 cmpdi cr7,r3,0
c00000000060a480 419e0078 beq cr7,c00000000060a4f8 # .__mutex_lock_slowpath+0xdc/0x1f4
c00000000060a484 93010070 stw r24,112(r1)
c00000000060a488 93210074 stw r25,116(r1)
c00000000060a48c 81210070 lwz r9,112(r1)
c00000000060a490 80010074 lwz r0,116(r1)
c00000000060a494 7d2907b4 extsw r9,r9
c00000000060a498 7c0007b4 extsw r0,r0
0:mon>
c00000000060a49c 7c2004ac lwsync
c00000000060a4a0 7d60e828 lwarx r11,0,r29
c00000000060a4a4 7c0b4800 cmpw r11,r9
c00000000060a4a8 40c20010 bne- c00000000060a4b8 # .__mutex_lock_slowpath+0x9c/0x1f4
c00000000060a4ac 7c00e92d stwcx. r0,0,r29
c00000000060a4b0 40c2fff0 bne- c00000000060a4a0 # .__mutex_lock_slowpath+0x84/0x1f4
c00000000060a4b4 4c00012c isync
c00000000060a4b8 2f8b0001 cmpwi cr7,r11,1
^^^^^ PC points to this instruction ^^^^^^^^
c00000000060a4bc 2f3f0000 cmpdi cr6,r31,0
c00000000060a4c0 409e0010 bne cr7,c00000000060a4d0 # .__mutex_lock_slowpath+0xb4/0x1f4
c00000000060a4c4 78200464 rldicr r0,r1,0,49
c00000000060a4c8 f81d0030 std r0,48(r29)
c00000000060a4cc 48000118 b c00000000060a5e4 # .__mutex_lock_slowpath+0x1c8/0x1f4
c00000000060a4d0 409a001c bne cr6,c00000000060a4ec # .__mutex_lock_slowpath+0xd0/0x1f4
c00000000060a4d4 e81b0000 ld r0,0(r27)
c00000000060a4d8 7809f7e3 rldicl. r9,r0,62,63
0:mon> r
R00 = 0000000000000000 R16 = 0000000002bc4b68
R01 = c000000000b13b60 R17 = 0000000000000000
R02 = c000000000b0bca0 R18 = c0000000008c4b68
R03 = c000000000d07fd0 R19 = 0000000001b1fc90
R04 = 0000000000000000 R20 = 00000000000000b8
R05 = 000000000000005e R21 = c0000000007ec008
R06 = 0000000000040000 R22 = 00000000007c28bb
R07 = c000000000a95288 R23 = c0000000007cbdd5
R08 = 0000000000000000 R24 = 0000000000000001
R09 = 0000000000000001 R25 = 0000000000000000
R10 = 0000000000000000 R26 = c000000000d08000
R11 = 00000000ffffffff R27 = c000000000b10080
R12 = 0000000024000082 R28 = c000000000a3ab70
R13 = c000000000be2400 R29 = c000000000d07fd0
R14 = c0000000008c4c30 R30 = c000000000a75be8
R15 = c000000000a95288 R31 = 0000000000000000
pc = c00000000060a4b8 .__mutex_lock_slowpath+0x9c/0x1f4
lr = c00000000060abc8 .mutex_lock+0x50/0x70
msr = 8000000000081032 cr = 84000022
ctr = 0000000000136f8c xer = 0000000000000001 trap = 100
0:mon>
Let me know if i can provide more information.
Thanks
-Sachin
--
---------------------------------
Sachin Sant
IBM Linux Technology Center
India Systems and Technology Labs
Bangalore, India
---------------------------------
--
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/