@ red-ray I've got an alpha multia UDB box. Do you already have the siv data for that?

weedeewee wrote on 2021-04-01, 19:58:

@ red-ray I've got an alpha multia UDB box. Do you already have the siv data for that?

Thank you for contacting me, as I recall the DEC Multia has a 233 MHz Alpha 21066 and I have never tried SIV on it so save files would be great. I think SIV32A.exe should run OK, but as the 21066 does not have BWX I guess time will tell.

I suspect once I get the save files I will be able to improve what SIV reports. Which OS does it have? SIV should be OK on NT4, but I usually test SIV32A on my DEC MiataGL Alpha 21164A (EV56) Windows 2000 Build 2072 RC1 system.

red-ray wrote on 2021-04-01, 20:10:

weedeewee wrote on 2021-04-01, 19:58:

@ red-ray I've got an alpha multia UDB box. Do you already have the siv data for that?

Thank you for contacting me and as I recall the DEC Multia has a 233 MHz Alpha 21066 and I have never tried SIV on it so save files would be great. I think SIV32A.exe should run OK, but as the 21066 does not have BWX I guess time will tell.

I suspect once I get the save files I will be able to improve what SIV reports. Which OS does it have? SIV should be OK on NT4, but I usually test SIV32A on my DEC MiataGL Windows 2000 Build 2072 RC1 system.

at the moment it doesn't really have any decent os install, since both 2"5 scsi hard drives I have, have started showing bad sectors, so I'll have to do a clean new install on either a sd2scsi device or a 4G Quantum UW scsi drive...
the 4G currently has a copy of the last failing hdd and is a debian upgraded to the latest officially available.
Now that I think of it, that would include finding out how to switch to the ARC firmware, write an NT alpha cd and install it or find the alpha 2k version... Should be fun 😀

weedeewee wrote on 2021-04-01, 20:20:

Now that I think of it, that would include finding out how to switch to the ARC firmware, write an NT alpha cd and install it or find the alpha 2k version... Should be fun 😀

GIMF and I just found REMOVED which is what I run on my DS20 (DP264G_tsunmp).

It was Oct-2015 when I last switched the firmware and can't remember how I did it. My Alphas have BIOS 5.71-R1 03/01/2000 19:48 + BIOS 5.70 07/20/1999 14:58 which I think are the latest available, but don't know about your system, looking at page 40 of the attached it seems easy. Does Re: Alpha hardware for Windows 2000 list it? https://en.wikipedia.org/wiki/DEC_Multia says it will run W2K and http://www.brouhaha.com/~eric/computers/udb.html is worth a look.

When I installed RC1/RC2 before starting the install I advanced the date by 50 years, did the install and then reverted it as by doing this it stops Windows timing out after a while.

red-ray wrote on 2021-04-01, 20:53:

GIMF and I just found REMOVED which is what I run on my DS20 (DP264G_tsunmp).

It was Oct-2015 when I last switched the firmware and can't remember how I did it. My Alphas have BIOS 5.71-R1 03/01/2000 19:48 + BIOS 5.70 07/20/1999 14:58 which I think are the latest available, but don't know about your system. Does Re: Alpha hardware for Windows 2000 list it? https://en.wikipedia.org/wiki/DEC_Multia says it will run W2K and http://www.brouhaha.com/~eric/computers/udb.html is worth a look.

When I installed RC1/RC2 before starting the install I advanced the date by 50 years, did the install and then reverted it as my doing this it stops Windows timing out after a while.

winworldpc is the site I was thinking of downloading it from ! So many old goodies to be found there.
I know I switched from ARC to srm 20 years ago, so the firmware definitely has both, as I think redhat used arc in the installation guidelines, while debian used srm
Though I never read anything about advancing the date 50 years.
I do recall while upgrading debian that at one point I had to update milo due to some code change and for another point in the upgrade, I had to skip the kernel because the framebuffer had a code error which caused it not to function.
Anyway
Thanks for the links ! Saves me to google for it or find the bookmarks I once made, maybe on a previous install, oh well.
When the machine is installed I'll send you the files.

weedeewee wrote on 2021-04-01, 21:19:

Though I never read anything about advancing the date 50 years.

I can't remember where I read it and failed to find anything with Google 🙁

BTW how much memory does it have? I don't have a feel for how much is needed, but my PWS has 1.5GB and typically uses 180MB + 520MB in the system file cache.

red-ray wrote on 2021-04-01, 21:36:

weedeewee wrote on 2021-04-01, 21:19:

Though I never read anything about advancing the date 50 years.

I can't remember where I read it and failed to find anything with Google 🙁

BTW how much memory does it have? I don't have a feel for how much is needed, but my PWS has 1.5GB and typically uses 180MB + 520MB in the system file cache.

I upgraded it last year with 2x 64MB simms so together with the 2 12MB simms it should be 152 MB if memory serves me right which is just wonderful compared to the 40 MB it had before 😁
Should be enough to run NT 4 or a barebones 2k... I hope

Hi

I've since managed to fix the 430VX system, it turned out that it wasn't the VRMs but rather there was a jumper that raised the Vcore voltage by 2.5% and apparently the older P133 could handle it fine, wheras the P233MMX couldn't without becoming unstable! After closing that jumper now it's all fixed.

I've also patched the BIOS of my board to fix the BIOS' 32gb bug (as in, the BIOS would literally crash when detecting hard drives larger than 32gb), I don't have any IDE HDDs on hand to test but we'll see if it works.

Also for some reason I was unable to get the monitor information, SIV had returned a "Error 2 - the specified file is missing" (since my system's locale is set to French) but otherwise I didn't have any bug:

Aside from the BIOS, CPU and the DVD drive (put in a not-so-period-correct-but-beige Pioneer DVD burner in place of the old TSST DVD-ROM drive I had that stuck out like a sore thumb as it had a dark faceplate) nothing much changed, it's still the same machine

Cheers,
Alexis

Vynix wrote on 2021-04-18, 13:48:

Also for some reason I was unable to get the monitor information, SIV had returned a "Error 2 - the specified file is missing" (since my system's locale is set to French) but otherwise I didn't have any bug

Thank you for the report and I am happy to hear you resolved the stability issue.

I suspect the Error 2 is down to HKLM\System\CurrentControlSet\Enum\Display not existing on Windows 9x systems, please can you check if it's present? If it is please post the registry export otherwise I will adjust SIV not to offer Menu->System->Monitors for Windows 9x systems.

Update: I searched my Windows Me system registry for EDID, found it and managed to make Menu->System->Monitors work. How does the attached SIV32L 5.58 EDID-00 test SIV do on your system please?

I'll be sure to do that tomorrow, the 430VX system isn't at my place (at the moment it's stored at my grandparents' house).

I can't rember if this registry key was present on the 430VX... But I'll check again just to be sure.

Vynix wrote on 2021-04-18, 17:28:

I can't remember if this registry key was present on the 430VX... But I'll check again just to be sure.

OK, I guess you did not notice I updated the post, it looks like on Windows 9x HKLM\Enum\Monitor needs to be used so I updated SIV32L 5.58 EDID-00 to use this

Ah right I saw your updated post, I'll try it tomorrow on the 430VX.

Looks like it's working now

Vynix wrote on 2021-04-19, 12:31:

Looks like it's working now

Thank you and yes. I am not sure what to do about the duplicates which I suspect are associated with different GPU instances

Oh the duplicates are due to the different monitors that I used with this system, it's nothing to worry about. I never changed the GPU in this system, it always was a S3 Trio64V2/DX, although I did put it in different slots, so that's possibly why multiple monitors show up.

Long time no see

I crossflashed a bios onto my Crosshair VI Hero t0 run a 5600x, noticed that SIV64X is showing that the chipset stuff is listed as 500 series instead of X370 like it was when a Zen 2 chip was installed even with the crossflashed bios.

The voltage stuff isn't showing due to the crossflashed bios so I have no concern about that.

mastergamma12 wrote on 2021-07-30, 23:49:

I crossflashed a bios onto my Crosshair VI Hero t0 run a 5600x, noticed that SIV64X is showing that the chipset stuff is listed as 500 series instead of X370 like it was when a Zen 2 chip was installed even with the crossflashed bios.

The voltage stuff isn't showing due to the crossflashed bios so I have no concern about that.

Thank you for the report, I thought that Vermeer + X370 was not supported so the SIV code did not allow for X370 with a K19 CPU. Please check SIV 5.61 Beta-01 or later which should report X370, does it?

SIV recognised the ITE IT8665, but no raw data is reported which must be down to the incorrect BIOS.

Update: I just noticed the ITE IT8665 is not enabled so I adjusted the attached SIVX64.sys to enable it. If you replace the Beta-01 driver with it the ITE IT8665 readings may get reported, do they?

I tried it, X370's now showing instead of 500 Series, your ITE fix shows stuff but nothing's remotely accurate.

mastergamma12 wrote on 2021-07-31, 10:32:

your ITE fix shows stuff but nothing's remotely accurate.

Thank you for checking out the changes and I half suspected the ITE IT8665 would need more than just being enabled so I will remove the enable code.

I noticed that your CPU-0 is almost 100% busy processing interrupts which seems wrong, do you know what causes this?

red-ray wrote on 2021-07-31, 11:17:

mastergamma12 wrote on 2021-07-31, 10:32:

your ITE fix shows stuff but nothing's remotely accurate.

Thank you for checking out the changes and I half suspected the ITE IT8665 would need more than just being enabled so I will remove the enable code.

I noticed that your CPU-0 is almost 100% busy processing interrupts which seems wrong, do you know what causes this?

I believe it's related to the enabled ITE code as it never happened before that.

[EDIT] Yep, swapping to the old siv64x.sys file in SIV 5.61 Beta-01 results in normal cpu usage.