Re: [patch] VP-2.6.9-rc4-mm1-T8
From: K.R. Foley
Date: Wed Oct 13 2004 - 09:02:47 EST
Ingo Molnar wrote:
* K.R. Foley <kr@xxxxxxxxxx> wrote:
Booted part way this time but then soft locked. [...]
the soft lock it seems was due to a mutex assert in the TCP code. Now i
see that you have the ipv6 module loaded. Would it be possible to
disable ipv6 briefly, to see whether there are other issues?
Yes. I don't need and will disable when building T9.
Mouse was still working but not KB. Any idea why the KB is not working
with these? I am rebuilding with all of the pwr managment off because
I am still getting acpi messages in the boot :-/ [...]
hm, the keyboard should be working - i have not seen any specific
keyboard problems with PREEMPT_REALTIME enabled (except in the very
early stages of this feature). So it must be something else. Is this a
PS2 or an USB keyboard?
PS2 keyboard. For clarification, I have intermittent problems with the
keyboard on this machine with anything to do with preempt. Sometimes it
works on boot sometimes it does not. Actually since it is so
unpredictable, I can't even say for sure that it has to do with preempt,
but it hasn't failed yet on rc4-mm1. I have tried to find a pattern and
have yet to do so. Possibly pertinent info:
Iwill MB dual 2.6G Xeon 512 ram
00:00.0 Host bridge: Intel Corp. E7505 Memory Controller Hub (rev 03)
00:00.1 Class ff00: Intel Corp. E7000 Series RAS Controller (rev 03)
00:01.0 PCI bridge: Intel Corp. E7000 Series Processor to AGP Controller
(rev 03)
00:02.0 PCI bridge: Intel Corp. E7000 Series Hub Interface B PCI-to-PCI
Bridge (rev 03)
00:1d.0 USB Controller: Intel Corp. 82801DB (ICH4) USB UHCI #1 (rev 02)
00:1d.1 USB Controller: Intel Corp. 82801DB (ICH4) USB UHCI #2 (rev 02)
00:1d.2 USB Controller: Intel Corp. 82801DB (ICH4) USB UHCI #3 (rev 02)
00:1d.7 USB Controller: Intel Corp. 82801DB (ICH4) USB2 EHCI Controller
(rev 02)
00:1e.0 PCI bridge: Intel Corp. 82801BA/CA/DB/EB/ER Hub interface to PCI
Bridge (rev 82)
00:1f.0 ISA bridge: Intel Corp. 82801DB (ICH4) LPC Bridge (rev 02)
00:1f.1 IDE interface: Intel Corp. 82801DB (ICH4) Ultra ATA 100 Storage
Controller (rev 02)
00:1f.3 SMBus: Intel Corp. 82801DB/DBM (ICH4) SMBus Controller (rev 02)
01:00.0 VGA compatible controller: nVidia Corporation NV17 [GeForce4 MX
440] (rev a3)
02:1c.0 PIC: Intel Corp. 82870P2 P64H2 I/OxAPIC (rev 03)
02:1d.0 PCI bridge: Intel Corp. 82870P2 P64H2 Hub PCI Bridge (rev 03)
02:1e.0 PIC: Intel Corp. 82870P2 P64H2 I/OxAPIC (rev 03)
02:1f.0 PCI bridge: Intel Corp. 82870P2 P64H2 Hub PCI Bridge (rev 03)
04:04.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5702X
Gigabit Ethernet (rev 02)
05:01.0 Ethernet controller: 3Com Corporation 3c905C-TX/TX-M [Tornado]
(rev 78)
Oct 12 15:33:39 swdev14 kernel: NET: Registered protocol family 10
Oct 12 15:33:39 swdev14 kernel: IPv6 over IPv4 tunneling driver
Oct 12 15:33:40 swdev14 kernel: ------------[ cut here ]------------
Oct 12 15:33:40 swdev14 kernel: kernel BUG at kernel/mutex.c:185!
Oct 12 15:33:40 swdev14 kernel: EIP is at _rw_mutex_write_unlock+0x25/0x2f
Oct 12 15:33:40 swdev14 kernel: Call Trace:
Oct 12 15:33:40 swdev14 kernel: [<c0253738>] tcp_listen_start+0x175/0x1d1
Oct 12 15:33:40 swdev14 kernel: [<c029a33a>] _spin_unlock_bh+0x1a/0x34
Oct 12 15:33:40 swdev14 kernel: [<c0275a8c>] inet_listen+0x65/0x7a
Oct 12 15:33:40 swdev14 kernel: [<c02287ea>] sys_listen+0x5c/0x74
Oct 12 15:33:40 swdev14 kernel: [<c02294a4>] sys_socketcall+0xb1/0x239
Oct 12 15:33:40 swdev14 kernel: [<c015ae81>] sys_close+0x75/0x91
this assert says that a write_unlock() was done before the mutex was
initialized - which is a no-no. Note that the stock kernel does not do
this checking and there's a chance that it has a true bug here which it
ignores silently. The more likely scenario is that the kernel mutex code
somewhere changed an assumption which broke the code. I'll try to
reproduce this.
Ingo
-
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/