The current freebios (1.0) tree won't build correctly when using the FDC37N769 for me. I needed to change
53: cmp %al, $0x4
to
53: cmp $0x4, %al /* 2003-05-12-jordan: order was incorrect for assembler.
in src/superio/SMC/fdc37c669/setup_serial.inc
Jordan
PS: Ron, Thanks for the info on 2.0.
Message: 1 Date: Sun, 11 May 2003 15:58:31 -0600 (MDT) From: ron minnich rminnich@lanl.gov To: linuxbios@clustermatic.org Subject: freebios2
Eric may be too busy to explain freebios2 so I'll pitch in.
We are trying to freeze the current freebios code base (current goal is 5/15 since fixes keep rolling in) and give people something that works. First frozen version will be 1.0.1.
Then we want to move to a new 'cruft-free' code base. This seemed the right time to branch.
I would hope that folks who own mainboards, and who are committers, would start looking at the new code base and seeing what it takes to get moved to it. I will be personally starting work on an Arima K8 mainboard this week.
ron
ADLO appears to choke on older harddrives.
We have a box of bunch of < 512 Meg harddrives that we use for systems where we want a little more space than a 8 or 32 Meg CF to develop on. I wanted to use one to boot a different Linux install than the one I'm booting now.
I tried about 6 or 7 different drives from different Mfgs and they all had a similar issue. ADLO would find the drive get a PCHS value and what sometimes looked like a resonable LCHS values. But the drive size would be 0 Meg. Then the int13 handler would barf about things being out of range. Usually 2 out of the 3 CHS values reported by int13 would be zero.
Perhpas a LBA problem? Is the current HD stuff only good for say like drives that can do LBA32? Anyone else ran into this?
This is mostly just a FYI for the archives as I'll just go cobble up another newer drive and move on.
heh. I just assumed nobody would use those since my assumption was that if someone does port of LinuxBIOS then they would go with new hardware since it is where the monnies are :-)
anyway, have a look at loaders.s specifically:
;----------------------------------------------------- ; V) tell BOCHS' BIOS we want to have LBA translation. ; 0x00 - NONE ; 0x01 - LBA <<<< ; 0x02 - LARGE ; 0x03 - R-CHS ; In future there will be 'fd failover'option in bochs.
mov al, #0x39 ;; cmos_reg out 0x70, al mov al, #0x01 ;; val (LBA) out 0x71, al ;-----------------------------------------------------
try puttting different values (other than LBA) into 'AL' and see if it helps.
On Mon, 12 May 2003, Richard Smith wrote:
ADLO appears to choke on older harddrives.
We have a box of bunch of < 512 Meg harddrives that we use for systems where we want a little more space than a 8 or 32 Meg CF to develop on. I wanted to use one to boot a different Linux install than the one I'm booting now.
I tried about 6 or 7 different drives from different Mfgs and they all had a similar issue. ADLO would find the drive get a PCHS value and what sometimes looked like a resonable LCHS values. But the drive size would be 0 Meg. Then the int13 handler would barf about things being out of range. Usually 2 out of the 3 CHS values reported by int13 would be zero.
Perhpas a LBA problem? Is the current HD stuff only good for say like drives that can do LBA32? Anyone else ran into this?
This is mostly just a FYI for the archives as I'll just go cobble up another newer drive and move on.
Adam Sulmicki wrote:
heh. I just assumed nobody would use those since my assumption was that if someone does port of LinuxBIOS then they would go with new hardware since it is where the monnies are :-)
Here at Bitworks about once per project you end up doing something stupid like shorting 12V to the 5V rail, or the drive gets knocked off the bench, or my personal favorite where a well place short in a 120V control circuit that had the line and neutral reversed make the whole setup float at 120V. Looking for an easy ground the unsuspecting engineer goes to plug his scope ground (which was earthed properly) up to the ground on the HD. *Poof*
When stuff happens it's painfull to know that you just waxed a 30+ gig drive. A < $1.00 ebay drive however is much eaiser to accept and sometimes even fun say if you have a big metal band saw around.
I'll add HD method selection info to my FAQ after I've played with it.
On Mon, 12 May 2003, jarcher wrote:
The current freebios (1.0) tree won't build correctly when using the FDC37N769 for me. I needed to change
53: cmp %al, $0x4
to
53: cmp $0x4, %al /* 2003-05-12-jordan: order was incorrect for assembler.
committed, please try it now.
ron
-
Adam Sulmicki -
jarcher -
Richard Smith -
ron minnich