PROBLEM: 2.4.19 & 2.4.20 hang without oops...

PROBLEM: 2.4.19 & 2.4.20 hang without oops...

[Charlton Harrison]

Hello Everyone,


I'd like to report a (probable) kernel bug in v2.4.19 and v2.4.20.
This problem does NOT appear in 2.4.18 and I've had to go back to
that kernel for now.

The problem is as follows:
While attempting to copy about 50GB of data onto an NFS-mounted partition,
the `cp -a` process will go for a while, then my machine will hang/freeze up.

I can reproduce it very easily and quickly on kernel 2.4.19 and 2.4.20,
and most of the time happens before even copying 10GB worth of data.
I am unable to reproduce the problem on kernel 2.4.18.

Here are the specifics:

HARDWARE:

Dual (SMP) P3-500,  supermicro MB,  512MB ECC buffered SDRAM.
3c905b ethernet card.
EIDE hard drive operating on a 16MHz (UDMA *disabled*) bus.


SOFTWARE:

Compiled the kernel with Redhat GCC v2.96.
(I also compiled v2.4.18 the same way and it works)


I am very concerned about this bug and that it doesn't seem to be fixed
in the latest Linux 2.4 kernels.

For additional information,  please feel free to e-mail me.  Or, if you
happen to know something about this bug already,  please e-mail me.

Thanks!


Charlton Harrison
charlton@dynet.com

Re: PROBLEM: 2.4.19 & 2.4.20 hang without oops...

[Matthew Costello]

I've just run in to a similar problem with 2.4.20.
The system hangs using 2.4.20, and everything is
fine with 2.4.18-10bigmem (rh73).

Points of similarity:

Dual (SMP) P3-450, supermicro (BX) motherboard
kernel 2.4.20
red hat (7.3 w/ updates)

Differences:

no IDE
main disks on AHA2940U2W
Firewire (aua-3121) or scsi AHA-1542

I've found two separate hangs.  The most interesting is that I have
two (different) firewire disks.  I can access either of the disks all
day, but as soon as I access both at the same time the system hangs.
I am unable to test this on 2.4.18 as one of the firewire disks
(WD 200GB) is not recognized correctly under 2.4.18, which is
why I am running 2.4.20 in the first place.

The second hang seems related to ejecting an audio cd-rom from
a CD-RW attached to its private AHA-1542 scsi controller.

Both hangs are "hard" without any sort of output to the console.
I also lose the console echo.  I'd guess that I'm running into a
SMP deadlock.

--- Matthew Costello <mattc@san.rr.com>

Re: PROBLEM: 2.4.19 & 2.4.20 hang without oops...

[Alan Cox]

On Tue, 2003-01-07 at 06:52, Charlton Harrison wrote:
> While attempting to copy about 50GB of data onto an NFS-mounted partition,
> the `cp -a` process will go for a while, then my machine will hang/freeze up.
> 
> I can reproduce it very easily and quickly on kernel 2.4.19 and 2.4.20,
> and most of the time happens before even copying 10GB worth of data.
> I am unable to reproduce the problem on kernel 2.4.18.
> 
> Here are the specifics:
> 
> HARDWARE:
> 
> Dual (SMP) P3-500,  supermicro MB,  512MB ECC buffered SDRAM.

What chipset is this system - is it 450NX by any chance ?

Re: PROBLEM: 2.4.19 & 2.4.20 hang without oops...

[Matthew Costello]

My motherboard is a Supermicro P6DBE with the 440BX chipset,
2 x P3 (Katmai) at 450MHz, 512MB RAM.
The output of lspci:

00:00.0 Host bridge: Intel Corp. 440BX/ZX/DX - 82443BX/ZX/DX Host bridge 
(rev 03)
00:01.0 PCI bridge: Intel Corp. 440BX/ZX/DX - 82443BX/ZX/DX AGP bridge 
(rev 03)
00:07.0 ISA bridge: Intel Corp. 82371AB/EB/MB PIIX4 ISA (rev 02)
00:07.1 IDE interface: Intel Corp. 82371AB/EB/MB PIIX4 IDE (rev 01)
00:07.2 USB Controller: Intel Corp. 82371AB/EB/MB PIIX4 USB (rev 01)
00:07.3 Bridge: Intel Corp. 82371AB/EB/MB PIIX4 ACPI (rev 02)
00:0f.0 Ethernet controller: Intel Corp. 82557/8/9 [Ethernet Pro 100] 
(rev 08)
00:10.0 SCSI storage controller: Adaptec AHA-2940U2/U2W
00:12.0 PCI bridge: Hint Corp HB1-SE33 PCI-PCI Bridge (rev 13)
00:14.0 Ethernet controller: Intel Corp. 82557/8/9 [Ethernet Pro 100] 
(rev 08)
01:08.0 USB Controller: NEC Corporation USB (rev 41)
01:08.1 USB Controller: NEC Corporation USB (rev 41)
01:08.2 USB Controller: NEC Corporation USB 2.0 (rev 02)
01:0b.0 FireWire (IEEE 1394): Texas Instruments TSB12LV26 IEEE-1394 
Controller (Link)
02:00.0 VGA compatible controller: nVidia Corporation RIVA TNT2 Model 64 
(rev 15)

Also two ISA cards:
Adaptec AHA1542C
Digiboard PC/8i

I'd be happy to help track the hang down.  If not I'll have to backport
the firewire changes in 2.4.19/20 to 2.4.18 since the WD firewire drive
is not recognized correctly in 2.4.20.
--- Matthew Costello <mattc@san.rr.com>