Beta-07 has improved support for the IBM Personal Computer Power Series 850 Model 6070 (RS/6000 7248 (43P)).
lolo799 wrote on 2025-08-13, 18:08:I'll test the new version when I can, in a week or 10 days from now.
I have just released SIV V5.83 Beta-08 which further improves the Windows NT V4.00 SP2 + PowerPC support so please use that.
Sorry, it took a while.
The iBook hdd is having problems coldstarting.
lolo799 wrote on 2025-09-06, 16:44:The iBook hdd is having problems coldstarting.
Thank you and I guess this is down to the age of the system. I had to replace the disk in my Motorola PowerStack as there were a vast number of bad blocks, I hope the replacement keeps working.
Since your last run I changed SIV to read the TB (mftbu+mftb) but the returned values are not what I expected which is why the speed is reported as 133 MHz, what speeds are reported when the system posts?
For PowerPC 604 I need to scale the value by 8 and wonder if I need to scale it by 30 for PowerPC 750cx, but feel I need to see what happens with a few other CPUs to figure out what is needed. Looking at http://www.cap-lore.com/code/TB/ it specifies scaling by 16.
red-ray wrote on 2025-09-06, 17:59:Since your last run I changed SIV to read the TB (mftbu+mftb) but the returned values are not what I expected which is why the speed is reported as 133 MHz, what speeds are reported when the system posts?
The G3 cpu in my iBook is a 500MHz model, with a bus speed of 66MHz.
There's no such thing as a Post screen with system info on a new world mac (there is the Open Firmware shell but it doesn't show up by default) and the NT loader screen only shows the ram available.
You can see what the NT4 setup process looks like using Wack0's fork of the dingusppc emulator, here for example:
https://virtuallyfun.com/2024/07/22/a-quick-v … macintosh-0-05/
lolo799 wrote on 2025-09-07, 09:10:The G3 cpu in my iBook is a 500MHz model, with a bus speed of 66MHz.
I have been looking into what is happening and it seems that the TB register is not derived from the CPU speed, but rather the FSB speed.
After discovering this I have been pondering how to get the FSB to CPU speed ratio, on recent x32/x64 systems (PentiumPro and later) it's quite easy as it's in a MSR, but for PowerPC this is not the case.
I have just uploaded SIV32P V5.83 Beta-09 which should work, please will you try it and post new save files?
Below is what I got in testing, you should get similar, but with a x7.5 rather than a x2 multiplier. The code should also work when the FSB is 100 MHz, etc., but I don't have a system I can test this on.
Here you go:
lolo799 wrote on 2025-09-08, 17:58:Here you go:
Thank you, the ratio calculation worked as I hoped it would. I have just uploaded SIV32P V5.83 Beta-10 which includes a few more improvements, it really needs to be validated with a few more CPU types/speeds though.
Above is using the Tahoma font, below is what happens when it's not installed.
I tried with the Tahoma font, it's not working too well:
lolo799 wrote on 2025-09-11, 19:57:I tried with the Tahoma font, it's not working too well:
In what aspect? Looking in the save file at Menu->Machine->SIV Setup then Arial is reported, I need to see the bottom line of [About] to be sure what font is being displayed.
My system @ 8: English - Font: Tahoma (Proportional Western) Weight 400 Point Size 8 Cell Size 6×13 Line Height 14.5
My system @ 9: English - Font: Tahoma (Proportional Western) Weight 400 Point Size 9 Cell Size 7×14 Line Height 15.75
Old Save File: English - Font: Tahoma (Proportional Western) Weight 400 Point Size 8 Cell Size 8×16 Line Height 18 - Note: Tahoma was not installed
New Save File: English - Font: Arial (Proportional Western) Weight 400 Point Size 8 Cell Size 6×14 Line Height 15.75
With a point size of 8 the Arial cells are about the same size as Tahoma with a point size of 9. You could use Menu->Tools->Font to set Tahoma with a point size of 9
Original
Tahoma 8 point
Tahoma 9 point
System - Arial I suspect
I had made the save local report while using the Arial font.
Tried Tahoma with a 9 font size:
red-ray wrote on 2025-09-12, 08:32:lolo799 wrote on 2025-09-11, 19:57:I tried with the Tahoma font, it's not working too well:
In what aspect?
lolo799 wrote on 2025-09-12, 11:19:Tried Tahoma with a 9 point font size:
Thank you, I am still unclear as to why you feel "it's not working too well".
Using 8 point Ariel I expect many of the SIV screens will to too wide to fit on your screen
red-ray wrote on 2025-09-12, 14:23:Thank you, I am still unclear as to why you feel "it's not working too well".
Using 8 point Ariel I expect many of the SIV screens will to too wide to fit on your screen
The text is cut off in many places, for example on the left side, Chips without the number, below CPU-0 with a missing half, the FSB MHz missing the Hz.
lolo799 wrote on 2025-09-12, 15:01:The text is cut off in many places
Thank you, I am unsure where the issue is as when the panel is rendered on this system there is no truncation.
The CPU-0 + GPU-0 cells are 7 characters wide so should not get truncated.
I felt I should try Windows XP @ 1024 x 786 and again there is no truncation.
I suspect the truncation may be down to the graphics driver.
lolo799 wrote on 2025-09-12, 15:01:The text is cut off in many places, for example on the left side, Chips without the number, below CPU-0 with a missing half, the FSB MHz missing the Hz.
After pondering I adjusted the width of a few columns for pre Windows 2000 systems so the SIV V5.83 release should have less truncation, how does it do on your system please?
Just wanted to mention, I have been testing a rather obscure graphics card based on the equally obscure Avance Logic ALG25128 graphics chip... I could be wrong but I believe this individual card is still the only one that has been pictured online anywhere using this chip. When I run SIV with this card installed, the program will crash (illegal op in SIV32L.exe) with a general protection fault in one of the video card's driver files"AW95G8.drv". When the illegal op window pops up the mouse is frozen (USB or PS/2) but I can still use the keyboard to navigate windows. Once I close the illegal op window the system seems unresponsive, though after a while it will respond to ctrl-alt-del to bring up task manager but I can't interact with anything. Sometimes I can ctrl-alt-del again to reboot, but other times it is just dead and I have to hold the power button.
I know this is probably of pretty limited importance since it is somewhat possible that this is the only ALG25128 that will ever need to be examined by SIV... but for what it's worth, I am willing to help diagnose the issue if you are interested. It is also possible that the video card itself has an issue that is causing the crash, but I have not experienced crashes with this card doing anything else, so I don't think that's it.
This was in Windows 98SE using SIV32L.exe from the latest version 5.84 zip.
Other system specs:
440BX chipset
Pentium III 850Mhz
128MB RAM
No other cards installed.
Ozzuneoj wrote on 2025-11-17, 03:16:When I run SIV with this card installed, the program will crash (illegal op in SIV32L.exe) with a general protection fault in one of the video card's driver files"AW95G8.drv".
Thank you for the report of the issue which I would like to fix, to be clear do both SIV32L and AW95G8.drv crash? Can you post a photo of the screen?
I suspect fixing this will be tricky. From a command window please will you try:
If I could get the issue here it would be much easier to know what is happening, I would run the MSVC 6 remote client (MSVCMON) on the W98 system then from a W10/W11 system use a TCP/IP connection from MSVC 6 to know exactly where the "illegal op in SIV32L.exe" is, but guess you don't have MSVC 6 available. In the SIV crash popup does it list the PC and other register values? Can you post a screen shot/photo?
Yesterday I released SIV V5.85 which can be downloaded via https://workupload.com/file/6BNqKMPdT7E.
It includes the first ever release of SIV32L that has ICH/PCH SMBus support for Windows 9x systems so it can now report DIMM SPD information. The code is the same as SIV has used on Windows NT/XP/7/10/11 for many years, but it's run in user mode rather than kernel mode so there may be timing issues. It works OK on my Pentium III (Katmaï) which has a PIIX4E SMBus, but I don't have many 95/98/Me systems so would be interested to hear how it does on other systems.
By default the SIV32L SMBus support is disabled on Windows 9x, to enable it do SIV32L -SMBUS and to disable it do SIV32L -NOSMBUS. Below is I typically what I got in testing.
I have just released SIV V5.86 Beta-00 which has much improved Windows 9x SMBus support, it's also enabled by default.
Below is what I got with it, notice how all the Peak Times are smaller then they were. Thus far it's worked well on every Windows 9x system I have, it should report DDR4 + DDR5 SPD OK, but I don't have a W9x system I can test this on.
It also reported the W83782D SMBus SIO on my Tyan S2466 Tiger MPX system and should also work for AS99127 SIOs on such as ASUS P3C + P3W + CUSL + A7V , please let me know how it does.
I have now released SIV V5.86 Beta-01 which had much improved Windows 9x SMBus support including ALi + SiS SMBus controllers.
I still need to add NVidia Windows 9x SMBus support and plan to do this once it can be tested.
