Repo Man11 wrote on 2021-12-17, 02:08:

tech3475 wrote on 2021-12-16, 09:53:

Well I have egg on my face right now. […]
Show full quote

Well I have egg on my face right now.

First of all thanks to everyone for their feedback and sorry for the late response as I had things come up and I wanted to test something.

Just before posting I decided to check the system using my PCI diag card to see if it would suggest anything.

Whilst I was looking up the code in the manual while it hung.....it booted from the SSD.

Turns out, it does work but it's incredibly slow to boot the drive, in comparison to anything else including the floppy as a proxy.

For some reason it seems to just be slow when on "FF 63" (which according to the manual is the INT 19 boot attempt).

That said, it's not ideal, so I'll probably be looking at the different options, but at least I know the thing can boot on it's own.

Interesting. For the sake of experimenting, I used Super Fdisk to partion a 500 gigabyte SATA drive to 120 gigs in a much newer computer and tried that in my Asus TXP4 (which has one of Jan's excellent patched BIOS) just using a SATA to IDE adapter, but it would hang when it it tried to detect the drive. Using the 120 gig SSD that I normally do worked fine in the IDE port with the adapter.

Have you tried just manually setting the drive in the bios to the max it can detect?
I've got an old 440BX board that goes to max <64G and just have it set at the max, manually, in the bios. Can't recall how I sized up the partitions but I know I can boot from it and use the whole drive.

edit: just to clarify, if I leave the board at autodetect or try to do an autodetect in the bios, it hangs. That's why I've set it manually to the max allowed in bios.

weedeewee wrote on 2021-12-17, 08:01:

Have you tried just manually setting the drive in the bios to the max it can detect? I've got an old 440BX board that goes to m […]
Show full quote

Repo Man11 wrote on 2021-12-17, 02:08:

tech3475 wrote on 2021-12-16, 09:53:

Well I have egg on my face right now. […]
Show full quote

Well I have egg on my face right now.

First of all thanks to everyone for their feedback and sorry for the late response as I had things come up and I wanted to test something.

Just before posting I decided to check the system using my PCI diag card to see if it would suggest anything.

Whilst I was looking up the code in the manual while it hung.....it booted from the SSD.

Turns out, it does work but it's incredibly slow to boot the drive, in comparison to anything else including the floppy as a proxy.

For some reason it seems to just be slow when on "FF 63" (which according to the manual is the INT 19 boot attempt).

That said, it's not ideal, so I'll probably be looking at the different options, but at least I know the thing can boot on it's own.

Interesting. For the sake of experimenting, I used Super Fdisk to partion a 500 gigabyte SATA drive to 120 gigs in a much newer computer and tried that in my Asus TXP4 (which has one of Jan's excellent patched BIOS) just using a SATA to IDE adapter, but it would hang when it it tried to detect the drive. Using the 120 gig SSD that I normally do worked fine in the IDE port with the adapter.

Have you tried just manually setting the drive in the bios to the max it can detect?
I've got an old 440BX board that goes to max <64G and just have it set at the max, manually, in the bios. Can't recall how I sized up the partitions but I know I can boot from it and use the whole drive.

edit: just to clarify, if I leave the board at autodetect or try to do an autodetect in the bios, it hangs. That's why I've set it manually to the max allowed in bios.

Unfortunately I don't seem to have that option, just the 'auto detect' menu. I tried manually setting the different modes (LBA, Normal and Large) but no difference.

care to post a screenshot of your bios hard disk settings ?

Repo Man11 wrote on 2021-12-17, 02:08:

Interesting. For the sake of experimenting, I used Super Fdisk to partion a 500 gigabyte SATA drive to 120 gigs in a much newer computer and tried that in my Asus TXP4 (which has one of Jan's excellent patched BIOS) just using a SATA to IDE adapter, but it would hang when it it tried to detect the drive. Using the 120 gig SSD that I normally do worked fine in the IDE port with the adapter.

Hi Repo Man11,

Great, you put my assumptions to the test! 😀
I still believe this 127.5 GiB workaround on a 500GB drive is valid on 2000-2003 Award BIOSes. However, you ran into the 128GiB limit of my patched BIOS!

Back in 2002, when I learned that another BIOS patcher found a fix for the infamous Award 32GB limit bug, I started to use that fix in my patched K6+ BIOSes as well. His name is Petr Soucek and he uses a simple and elegant 15 bytes patch to fix this bug. You can find a description on this website: http://www.ryston.cz/petr/bios/

This same patch is used in all patched BIOSes by the Wimsbios team as well, and team member Rainbow still has these BIOSes available at http://wims.rainbow-software.org/

However, all these patched BIOSes hang during POST when a drive with 267,386,880 or more sectors (127.5GiB) is attached. This new limit is a side-effect of the code used in the patch and coincides nicely with the 28-bit LBA limit. So this was of no consequence at the time, until bigger drives (with 48-bit LBA support) like 160 or 200GB came on the market.

Recently I have been looking into the code that Award itself used to fix the 32GB limit bug in late 1999/2000 BIOSes, and I found a way to use this logic to make a new bugfix for 1997/98/99 Socket 7 BIOSes. This new 32GB bugfix now uses 32-bit registers i.s.o. 16-bit to do the calculations, thereby avoiding a divide overflow at 127.5GiB. The new limit for this fix is 640GiB, just like in Award’s new code, although the BIOS will still only 'see' the first 127.5GiB due to the 28-bit LBA limitation.

This new 32GB bugfix works fine now on a few patched BIOSes, and I can put it in your patched 0112l-1 Asus TXP4 BIOS as well.
It took me years to patch all those socket 7 BIOSes and I will not redo them all, but I like to adapt the TXP4 BIOS so you can finish this workaround experiment. 😉
Just give me a few days to get the patched BIOS ready.

Jan.

(bangs head on desk)

It was the DVD drive.

I forgot that I had also changed the DVD drive because my old one kept closing itself. The replacement DVD drive worked fine in my Amiga 1200 but in this it seems to just hang the system (my boot priority is A>CDROM>C).

What's weird is that when booting from CDs using it I had no problems, it's only when going from the DVD to SSD/HDDs that it has any problems.

So I've just changed the boot priority (A>C) for now and it's working just fine.

Sorry for the false alarm.

Thank you.

Any chance it could have something to do with the master/slave setup? What is the DVD set to? Primary slave jumpered as cable select, primary slave jumpered as slave, or secondary master jumpered as cable select or master? There have been some instances I've run into where using cable select has caused a hang or other weird issue.

weedeewee wrote on 2021-12-17, 08:01:

Have you tried just manually setting the drive in the bios to the max it can detect? I've got an old 440BX board that goes to m […]
Show full quote

Repo Man11 wrote on 2021-12-17, 02:08:

tech3475 wrote on 2021-12-16, 09:53:

Well I have egg on my face right now. […]
Show full quote

Well I have egg on my face right now.

First of all thanks to everyone for their feedback and sorry for the late response as I had things come up and I wanted to test something.

Just before posting I decided to check the system using my PCI diag card to see if it would suggest anything.

Whilst I was looking up the code in the manual while it hung.....it booted from the SSD.

Turns out, it does work but it's incredibly slow to boot the drive, in comparison to anything else including the floppy as a proxy.

For some reason it seems to just be slow when on "FF 63" (which according to the manual is the INT 19 boot attempt).

That said, it's not ideal, so I'll probably be looking at the different options, but at least I know the thing can boot on it's own.

Interesting. For the sake of experimenting, I used Super Fdisk to partion a 500 gigabyte SATA drive to 120 gigs in a much newer computer and tried that in my Asus TXP4 (which has one of Jan's excellent patched BIOS) just using a SATA to IDE adapter, but it would hang when it it tried to detect the drive. Using the 120 gig SSD that I normally do worked fine in the IDE port with the adapter.

Have you tried just manually setting the drive in the bios to the max it can detect?
I've got an old 440BX board that goes to max <64G and just have it set at the max, manually, in the bios. Can't recall how I sized up the partitions but I know I can boot from it and use the whole drive.

edit: just to clarify, if I leave the board at autodetect or try to do an autodetect in the bios, it hangs. That's why I've set it manually to the max allowed in bios.

I will give this a try when I get the chance.

Chkcpu wrote on 2021-12-17, 15:28:

Hi Repo Man11, […]
Show full quote

Repo Man11 wrote on 2021-12-17, 02:08:

Interesting. For the sake of experimenting, I used Super Fdisk to partion a 500 gigabyte SATA drive to 120 gigs in a much newer computer and tried that in my Asus TXP4 (which has one of Jan's excellent patched BIOS) just using a SATA to IDE adapter, but it would hang when it it tried to detect the drive. Using the 120 gig SSD that I normally do worked fine in the IDE port with the adapter.

Hi Repo Man11,

Great, you put my assumptions to the test! 😀
I still believe this 127.5 GiB workaround on a 500GB drive is valid on 2000-2003 Award BIOSes. However, you ran into the 128GiB limit of my patched BIOS!

Back in 2002, when I learned that another BIOS patcher found a fix for the infamous Award 32GB limit bug, I started to use that fix in my patched K6+ BIOSes as well. His name is Petr Soucek and he uses a simple and elegant 15 bytes patch to fix this bug. You can find a description on this website: http://www.ryston.cz/petr/bios/

This same patch is used in all patched BIOSes by the Wimsbios team as well, and team member Rainbow still has these BIOSes available at http://wims.rainbow-software.org/

However, all these patched BIOSes hang during POST when a drive with 267,386,880 or more sectors (127.5GiB) is attached. This new limit is a side-effect of the code used in the patch and coincides nicely with the 28-bit LBA limit. So this was of no consequence at the time, until bigger drives (with 48-bit LBA support) like 160 or 200GB came on the market.

Recently I have been looking into the code that Award itself used to fix the 32GB limit bug in late 1999/2000 BIOSes, and I found a way to use this logic to make a new bugfix for 1997/98/99 Socket 7 BIOSes. This new 32GB bugfix now uses 32-bit registers i.s.o. 16-bit to do the calculations, thereby avoiding a divide overflow at 127.5GiB. The new limit for this fix is 640GiB, just like in Award’s new code, although the BIOS will still only 'see' the first 127.5GiB due to the 28-bit LBA limitation.

This new 32GB bugfix works fine now on a few patched BIOSes, and I can put it in your patched 0112l-1 Asus TXP4 BIOS as well.
It took me years to patch all those socket 7 BIOSes and I will not redo them all, but I like to adapt the TXP4 BIOS so you can finish this workaround experiment. 😉
Just give me a few days to get the patched BIOS ready.

Jan.

That would be cool, but don't go to all of that trouble on my account - that computer works very well as is, I was just trying this out for the sake of knowing what works and what does not.

Brickpad wrote on 2021-12-18, 02:08:

Any chance it could have something to do with the master/slave setup? What is the DVD set to? Primary slave jumpered as cable select, primary slave jumpered as slave, or secondary master jumpered as cable select or master? There have been some instances I've run into where using cable select has caused a hang or other weird issue.

I thought I had it set to secondary slave, but I’ll have to have a look again when I get the chance. Secondary master being an mSD to IDE adapter (which on a side note, has proven to be extremely useful as it acts as a HDD, definitely recommend since DOS can see it so I use it to install W98, etc.).

I’ll take a look and maybe try experimenting when I get a chance.

Thanks for the new and improved BIOS Jan! I used the same 500 gigabyte Seagate hybrid SATA drive as before, but with the new BIOS it was detected properly and it booted right into the Win98 install I'd already put on the 120 gigabyte partition while it was attached to the SiL3114. The drive speed does suffer a bit when compared to using the PCI SATA card, but mission accomplished.

@ Repo Man11,

Thanks for testing the improved 32GB bugfix on your TXP4! Looks fine indeed. 😀
As I have now several tested patched K6+ BIOSes with the improved 32GB bugfix, I will put them on my k6plus webpage at the next update.

Jan

Repo Man11 wrote on 2021-12-20, 01:05:

Thanks for the new and improved BIOS Jan! I used the same 500 gigabyte Seagate hybrid SATA drive as before, but with the new BIOS it was detected properly and it booted right into the Win98 install I'd already put on the 120 gigabyte partition while it was attached to the SiL3114. The drive speed does suffer a bit when compared to using the PCI SATA card, but mission accomplished.

That seems abnormally slow to me, under Windows 98 SE
I get this with an SSD on my P3B-F's integrated IDE port (DMA enabled manually in device manager)

and this on a SIL3114 in the same P3B-F but with a mechanical 1TB hard drive .

Good catch - I forget that you need to check that with Win98, and DMA was not enabled. The results are now more in line with an ATA33 controller.

Chkcpu wrote on 2021-12-20, 11:32:

@ Repo Man11, […]
Show full quote

@ Repo Man11,

Thanks for testing the improved 32GB bugfix on your TXP4! Looks fine indeed. 😀
As I have now several tested patched K6+ BIOSes with the improved 32GB bugfix, I will put them on my k6plus webpage at the next update.

Jan

Hey Jan, could you update the s1564D/S bios with the latest fix?

Re: Tyan S1564D to tyan s1564D dual processor conversion

Yes, I can update the 32GB bugfix in the patched Tyan S1564 BIOS as well.
I expect to have time for that after the holidays. 😉

@Sphere478, I will send you a PM when its ready.

Merry Christmas everybody,
Jan

Chkcpu wrote on 2021-12-24, 12:27:

Yes, I can update the 32GB bugfix in the patched Tyan S1564 BIOS as well. I expect to have time for that after the holidays. ;) […]
Show full quote

Yes, I can update the 32GB bugfix in the patched Tyan S1564 BIOS as well.
I expect to have time for that after the holidays. 😉

@Sphere478, I will send you a PM when its ready.

Merry Christmas everybody,
Jan

woohoo!!