RESOLVED -- Promise ATA/Ultra66 Hang on Boot... -- RESOLVED

Message

gholam · Post by **gholam** » Thu May 29, 2003 7:42 am

I don't have an exact paste (yet) because my boot sequence doesn't progress far enough to write dmesg to disk. I'm going to have to copy it by hand (meh). However, I shall do my best to describe this problem.

I have a PCI Promise Ultra66 ATA card that uses the pdc202xx driver. During boot, the kernel recognises my onboard IDE no problem, ide0 and ide1, reads the drives, hda hdb hdc, then goes on to say "Okay, I'm using the pdc driver you compiled into me" and recognises ide2 and ide3, and the drives therein: hde, hdf, hdg. So far, so good.

Now the problem... the kernel goes back over the drives found (hda,hdb,hdc,hde,hdf,hdg), then spits out the ide info (ide0, ide1...) but completely hard locks before saying "ide2, ide3". I am baffled!

The LiveCD loads the pdc driver and boots with no problems at all! I've tried different kernel versions with different options, but everything locks at this point!!! Help! I've been pulling my hair for two weeks now trying to figure out what's happening.

gholam · Post by **gholam** » Thu May 29, 2003 8:20 am

Okay I went through the pains of manually copying the dmesg and here it is as follows.

Code: Select all

SIS5513: IDE controller on PCI bus 00 dev 15
SIS5513: chipset revision 208
SIS5513: not 100% native mode: will probe irqs later
SiS5513
    ide0: BM-DMA at 0xd800-0xd807, BIOS settings: hda:DMA, hdb:DMA
    ide1: BM-DMA at 0xd808-0xd80f, BIOS settings: hdc:DMA, hdd:pio
PDC20262: IDE controller on PCI bus 00 dev 48
PCI: Found IRQ 10 for device 00:09.0
PCI: Sharing IRQ 10 with 00:05.0
PDC20262: chipset revision 1
PDC20262: not 100% native mode: will probe irqs later
PDC20262: (U)DMA Burst Bit ENABLED Primary PCI Mode Secondary PCI Mode
    ide2: BM-DMA at 0x9400-0x9407, BIOS settings: hde:pio, hdf:pio
    ide3: BM-DMA at 0x9408-0x940f, BIOS settings: hdg:pio, hdh:pio
hda: QUANTUM FIREBALL AS40.0, ATA DISK drive
hdb: WDC WD400BB-32AUA1, ATA DISK drive
hdc: MATSHITADVD-ROM SR-8586, ATAPI CD/DVD-ROM drive
hde: QUANTUM FIREBALL LM20.5, ATA DISK drive
hdf: Maxtor 4W080H6, ATA DISK drive
hdg: ZIPCD1024INT-A, ATAPI CD/DVD-ROM drive
ide0 at 0x1f0-0x1f7, 0x3f6 on irq 14
ide1 at 0x170-0x177, 0x376 on irq 15
spurious 8259A interrupt: IRQ7.

... and then it hangs.

This is with the 2.4.20-gentoo-r5 kernel SMP enabled (or disabled.. they both lock up -- also, the pre-emp doesn't do anything). I've also tried 2.4.21-ac1-rc4, 2.4.21-ac2-rc2, 2.4.20 vanilla, 2.4.21-rc2, 2.4.20-xfs, and a few others. They all lock.

[add]
Btw, I'm running this:
AthlonXP 1900+
A7S333 ASUS mobo
256mb 333mhz DDR
3com NIC
GF2 GTS/Pro

ckrenson · Post by **ckrenson** » Thu Jun 12, 2003 4:29 am

not to spoil your hopes or anything, i just wanted to say that i'm having the exact same problem with my promise ultra 66 card. while i have been using r4 recently to try and boot, i doubt r5 will work either based on your message. if i find anything that works i'll let you know.

S_aIN_t · Post by **S_aIN_t** » Thu Jun 12, 2003 1:43 pm

i had the same problem.. the only way i found to deal with it, was to replace the udma66 card with udma133 card.

works like a charm now.

ckrenson · Post by **ckrenson** » Thu Jun 12, 2003 3:11 pm

from my reply above, i tried to boot my computer up with the redhat 9 isos, which properly identified my hard drive and all partitions on it, allowing for modifications etc, so what i'm thinking is maybe somehow make a mixed boot cd with the vmlinuz from the redhat iso or something and then make sure to build ultra 66 compatibility into the kernel when you get to the makemenuconfig stage.

gholam · Post by **gholam** » Tue Jun 17, 2003 5:14 pm

Okay, I had a false sense of security that was induced by the card actually working with the LiveCD and windozeXP (as well as redhat8 last year). I'd ignored a friend's suggestion of moving the card's PCI slot two weeks ago and I finally tried it the other night when he asked me again if I'd done it.
Boy did I feel silly.
I swapped the slot with my NIC and it worked like a charm. I've heard other flaky PCI problems with the A7S333 board, and I've actually had one before that I resolved with the help of a little googling. -- I should post that as another thread...
Anyway, thanks for the help you guys. It just goes to show that 10^x minds are better than 10^0.

defknol · Post by **defknol** » Tue Nov 04, 2003 11:13 pm

I had similar issues and found that I had to compile in the pdc202xx_new driver into the kernel and not as a module.

Rroet · Post by **Rroet** » Sun Nov 23, 2003 10:36 pm

I still have this problem.

The weird part is, it's only with gentoo-sources 2.4.20-rX

the 2.4.22-rX range works like a charm, but I can't use it because the kernel developers put in a version of I2C which is too new for my PVR card to work with

defknol · Post by **defknol** » Mon Nov 24, 2003 2:43 am

I'm no expert at kernel hacking, but I might just copy/replace the .c file for that module into the 2.4.20 sources from the 2.4.22 sources tree - then compile and see if that works. - compiling into the kernel.

It's a hack, but who knows what might happen.

The one I'm using is: /usr/src/linux/drivers/ide/pci/pdc202xx_new.c

I'm no C user, so I'm not sure if this would work - though I'd be curious to see if that works for you.

Rroet · Post by **Rroet** » Mon Nov 24, 2003 10:54 am

found it myself.

it's a ACPI / APM problem again in the kernel

It's fixed in the 2.4.22 vanilla branch