Re: [PATCH] pooling NUMA scheduler with initial load balancing

From: Martin J. Bligh (
Date: Wed Oct 09 2002 - 12:33:19 EST

> This is strange. It works for me really reliably. I added a check
> for non-online CPUs in calc_pool_load and changed the pool_lock to
> be a spinlock. The patches are attached again.

I'm testing on 2.5.41-mm1 ... your patches still apply clean. That
has a whole bunch of nice NUMA mods, you might want to try that
instead? All the changes in there will end up in mainline real soon
anyway ;-)

One minor warning:

arch/i386/kernel/smpboot.c: In function `smp_cpus_done':
arch/i386/kernel/smpboot.c:1199: warning: implicit declaration of function `bld_pools'

And the same panic:

Starting migration thread for cpu 3
Bringing up 4
igrU: 4
ead :or 0060:[<c011422d>] Not tainted
EFLAGS: 00010046
EIP is at calc_pool_load+0x115/0x12c
eax: 00000000 ebx: 00000001 ecx: c028f948 edx: 00000000
esi: c034f380 edi: 00000020 ebp: c3a37efc esp: c3a37ed4
ds: 0068 es: 0068 ss: 0068
Process swapper (pid: 0, threadinfo=c3a36000 task=f01bf060)
Stack: c028f948 c028efa0 f01bf060 00002928 00002944 00002944 ffffffff ffffffff
       c028efa0 ffffffff c3a37f78 c01142d1 c028f948 00000004 00000001 00000001
       c3a36000 c028efa0 f01bf060 00000010 00000008 00000000 00000001 00002a13
Call Trace:
 [<c01142d1>] load_balance+0x8d/0x5c8
 [<c0118a9d>] call_console_drivers+0xdd/0xe4
 [<c0118cd2>] release_console_sem+0x42/0xa4
 [<c0114c21>] schedule+0xd5/0x3ac
 [<c0105300>] default_idle+0x0/0x34
 [<c01053bf>] cpu_idle+0x43/0x48
 [<c0118cd2>] release_console_sem+0x42/0xa4
 [<c0118c2d>] printk+0x125/0x140

Code: f7 3c 9e 89 84 99 80 00 00 00 8b 45 f4 5b 5e 5f 89 ec 5d c3

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

This archive was generated by hypermail 2b29 : Tue Oct 15 2002 - 22:00:32 EST