The actual lockup occurs within ide_config_drive_speed
(drivers/block/ide.c) at "OUT_BYTE(drive->ctl | 2, IDE_CONTROL_REG);" (on or
about line 2306). It occurs regardless of drive mode (UDMA66/4 UDMA66/3
UDMA33). ide_config_drive_speed is called by hpt366_tune_chipset which is
called by config_chipset_for_dma (both in drivers/block/hpt366.c after
ide.2.2.12.patch has been applied).
The motherboard itself has 4 ide ports, 2 are standard UDMA33 built onto
the motherboard, 2 are UDMA66 controlled through the HPT366 (and a seperate
bios). The problem existed with bios v1.01 as well as 1.10 (an upgrade
included with the BE6 bios update MU). It occurs regardless of IRQ or other
pci/isa devices (occurs when the system is striped). The drives in use are
maxtor 92720U8 (4 of them, but since the bios update to 1.10 the hpt366
driver only detects the first chain - 2 drives.) Each of the UDMA33 ports
has a single UDMA33 drive on it which initialize fine.
The system lock causes a complete loss of response (4 second power-button
required to power down).
The lock occurs immediately after initializing the UDMA33 drives:
hda: Maxtor 88400D8, 8011MB w/256kB Cache, CHS=16278/16/63, UDMA
hdc: Maxtor 91360D8, 12970MB w/256kB Cache, CHS=26353/16/63, UDMA
A bit of clunky debugging revealed the function/line which caused the lock.
Specs:
Abit BE6 motherboard (single slot-1, 2 udma33 ports, 2 udma66 ports via
HPT366)
p-2 266
128 megs sdram
The two UDMA33 drives listed above (8G,13G). Four Maxtor 92720U8 7200rpm
UDMA66 drives.
Redhat 6.0 (though I suspect this doesn't matter since the kernel never
fully initializes.
kernel 2.2.12 w/ ide.2.2.12.patch
(ftp.kernel.org/pub/linux/kernel/people/hedrick/ide.2.2.12.patch.gz -
9/12/1999)
Various cables and drive configurations were attempted.
Any help/references/pointers would be appreciated! :)
Andrew Mann
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/