CuPid wrote on 2020-08-11, 15:00:

I've just released a new version of cpu-z vintage edition (1.02). The new features are : - AMD K5 model 0 (SSA/5). - Improved su […]
Show full quote

I've just released a new version of cpu-z vintage edition (1.02).
The new features are :
- AMD K5 model 0 (SSA/5).
- Improved support of Cyrix Cx486DRx2/DRu2.
- Improved support of Ti486SXL/SXL2.
- Intel 430FX chipset.
- Opti 82C822 chipset.
- VIA VT82C505/VT82C486 chipset.

Official version : https://download.cpuid.com/cpu-z/cpu-z_1.02-win9x.zip
Beta version : http://download.cpuid.com/betas/cpuz_w95.zip

I'm currently working on adding the support of ISA/VLB video cards, if you have such a video card and are willing to help, you can use one the two versions listed above, save a Video BIOS dump from the tools button, and post it here. Thank you !

Hi,

I could try this with a Cirrus Logic VLB card that I've got lying somewhere, though I won't immediately be able to pull the BIOS dump (it's currently inside a 80486DX4-100 system that I need to get from my uncle's house).

Give me maybe 2-3 weeks and I'll try to get this sorted out.

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck.
CPUID 1.02.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. The beginning of the video memory is scrambled.

Windows 98 SE.
BIOS labels CPU as "VIA C3 - 1 Giga Pro".
CHKCPU identifies the CPU as
CPU-ID Vendor String: CentaurHauls
CPU-ID Name String: VIA Samuel 2
CPU-ID Signature: 0673
CPU is 6x86 class, but does not support CMOV instruction.

Disruptor wrote on 2020-09-19, 10:08:

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. CPUID 1.02.0.w9x hangs at 5 % on a VI […]
Show full quote

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck.
CPUID 1.02.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. The beginning of the video memory is scrambled.

Windows 98 SE.
BIOS labels CPU as "VIA C3 - 1 Giga Pro".
CHKCPU identifies the CPU as
CPU-ID Vendor String: CentaurHauls
CPU-ID Name String: VIA Samuel 2
CPU-ID Signature: 0673
CPU is 6x86 class, but does not support CMOV instruction.

Thanks for the report.
Can you please copy that file cpuz.ini in the same folder as cpuz_w9x.exe, and try to run the application again ?

CuPid wrote on 2020-09-20, 16:29:

Disruptor wrote on 2020-09-19, 10:08:

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. CPUID 1.02.0.w9x hangs at 5 % on a VI […]
Show full quote

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck.
CPUID 1.02.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. The beginning of the video memory is scrambled.

Windows 98 SE.
BIOS labels CPU as "VIA C3 - 1 Giga Pro".
CHKCPU identifies the CPU as
CPU-ID Vendor String: CentaurHauls
CPU-ID Name String: VIA Samuel 2
CPU-ID Signature: 0673
CPU is 6x86 class, but does not support CMOV instruction.

Thanks for the report.
Can you please copy that file cpuz.ini in the same folder as cpuz_w9x.exe, and try to run the application again ?

Same result. Video memory is not scrambled anymore.

Also same result with this cpuz.ini
Video memory is also not scrambled.

1[CPU-Z]
2VERSION=1.0.2.0
3TextFontName=
4TextFontSize=14
5TextFontColor=000080
6LabelFontName=
7LabelFontSize=14
8ACPI=0
9PCI=0
10MaxPCIBus=256
11DMI=0
12Sensor=0
13SMBus=0
14Display=0
15UseDisplayAPI=0
16BusClock=0
17Chipset=0
18SPD=0
19XOC=0
20CheckUpdates=0
21ValidUsername=
22ValidEmail=
23ValidPublish=1

Disruptor wrote on 2020-09-20, 18:04:

Same result. Video memory is not scrambled anymore. […]
Show full quote

CuPid wrote on 2020-09-20, 16:29:

Disruptor wrote on 2020-09-19, 10:08:

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. CPUID 1.02.0.w9x hangs at 5 % on a VI […]
Show full quote

CPUID 1.01.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck.
CPUID 1.02.0.w9x hangs at 5 % on a VIA C3 - 1 Giga Pro (VIA Samuel 2 CPU). Mouse is stuck. The beginning of the video memory is scrambled.

Windows 98 SE.
BIOS labels CPU as "VIA C3 - 1 Giga Pro".
CHKCPU identifies the CPU as
CPU-ID Vendor String: CentaurHauls
CPU-ID Name String: VIA Samuel 2
CPU-ID Signature: 0673
CPU is 6x86 class, but does not support CMOV instruction.

Thanks for the report.
Can you please copy that file cpuz.ini in the same folder as cpuz_w9x.exe, and try to run the application again ?

Same result. Video memory is not scrambled anymore.

Also same result with this cpuz.ini
Video memory is also not scrambled.

Copy code to clipboard

1[CPU-Z]
2VERSION=1.0.2.0
3TextFontName=
4TextFontSize=14
5TextFontColor=000080
6LabelFontName=
7LabelFontSize=14
8ACPI=0
9PCI=0
10MaxPCIBus=256
11DMI=0
12Sensor=0
13SMBus=0
14Display=0
15UseDisplayAPI=0
16BusClock=0
17Chipset=0
18SPD=0
19XOC=0
20CheckUpdates=0
21ValidUsername=
22ValidEmail=
23ValidPublish=1

OK thank you, I think I know what's wrong, can you please download and try that version ?

https://download.cpuid.com/betas/cpuz_w95.zip

CuPid wrote on 2020-09-21, 17:33:

OK thank you, I think I know what's wrong, can you please download and try that version ?

Sorry, you were wrong. Same result.

Disruptor wrote on 2020-09-21, 18:13:

CuPid wrote on 2020-09-21, 17:33:

OK thank you, I think I know what's wrong, can you please download and try that version ?

Sorry, you were wrong. Same result.

OK thank you.
Here is a new test version, can you please use the ini included in the zip ?
Thanks.

https://download.cpuid.com/betas/cpuz_w9x_test_0.zip

I'm sorry. There is no other behaviour.

Perhaps we can get closer to the problem when you add some commands to print status codes on port 80h

Hum this one is a struggle.

CuPid wrote on 2020-09-24, 18:14:

Hum this one is a struggle.

I joined the struggle. It turns out that there is no agreement between different environment on how to handle the x86 instruction bytes 0F 3F 07 0B. If you just enter these bytes into Google, you will quickly find out that this sequence is commonly disassembled as vpcext 7, 0Bh. This is not a processor instruction, and will fault as undefined opcode on most x86 processors. The Hypervisor in Microsoft VirtualPC uses 0F 3F as entry point, and this instruction is used to detect the presence of this Hypervisor.

On the other hand, there is the "Alternate Instruction Set" of the VIA processors. As Wikipedia says

Wikipedia wrote:

The Alternate Instruction Set (AIS) is a second 32-bit instruction set architecture found in some x86 CPUs made by VIA Technologies. On these VIA C3 processors, the second hidden processor mode is accessed by executing the x86 instruction ALTINST (0F 3F). If AIS mode has been enabled, the processor will perform a JMP EAX[1] and begin executing AIS instructions at the address of the EAX register."

It seems the PC-Chips M787CLR mainboard boots into the operating system with the AIS enabled, which is known to be a very bad idea for system security, as the AIS (if enabled) can be used to fully bypass any kind of memory protection. CPU-Z calls the VirtualPC Hypervisor with EAX equal to 1, which is not at all a good address to jump to. I overwrote the offending instruction by 0F 0B 90 90, an opcode that is guaranteed by specification to raise an undefined opcode exception, followed by two NOPs. Now CPU-Z works on this machine (but it would fail to detect VirtualPC).

Hello!

I noticed this little project of yours, so I decided I'd check out Vintage Edition v1.02 on my Pentium III-S 1.4Ghz rig to see how it stacks up against one of the last versions of CPU-Z supported by Win98SE that I know of (v1.57). I will say, I love the improvements to the readability of the text (I increased the size of the text in the .inf file for v1.57 for the text to be more readable) as well as some minor layout improvements. I also see that some more information is being displayed in the VE than in the older version. Keep up the great work!

To further help, I did notice one regression between v1.57 and VE v1.02 that I would like to report. It has to do with the VE not reporting my mainboard's "LCPIO", whatever that is. 😀

Here's the missing information from the VE report that's in the report from v1.57:

1LPCIO Vendor		AsusTek
2LPCIO Model		AS99127F
3LPCIO Vendor ID		0x5CA3
4LPCIO Chip ID		0x59
5LPCIO Revision ID	0x53
6Config Mode I/O address	0x2E
7Config Mode LDN		0xB
8Config Mode registers	
9		00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 
10	00	FF FF FF FF FF FF FF 0B FF FF FF FF FF FF FF FF 
11	10	FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
12	20	59 53 FF FE 80 00 00 FF 00 0C 7C F0 1F 00 00 FF 
13	30	00 FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
14	40	FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
15	50	FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
16	60	00 00 FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
17	70	00 FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 

I have included both reports as well as the above screenshot in the zip file attached for your assistance. If there's anything else you need from me, feel free to ask.

CPUZ fails to report tRFC, should it?

the_ultra_code wrote on 2020-10-29, 21:42:

Hello! […]
Show full quote

Hello!

I noticed this little project of yours, so I decided I'd check out Vintage Edition v1.02 on my Pentium III-S 1.4Ghz rig to see how it stacks up against one of the last versions of CPU-Z supported by Win98SE that I know of (v1.57). I will say, I love the improvements to the readability of the text (I increased the size of the text in the .inf file for v1.57 for the text to be more readable) as well as some minor layout improvements. I also see that some more information is being displayed in the VE than in the older version. Keep up the great work!

To further help, I did notice one regression between v1.57 and VE v1.02 that I would like to report. It has to do with the VE not reporting my mainboard's "LCPIO", whatever that is. 😀

Here's the missing information from the VE report that's in the report from v1.57:

Copy code to clipboard

1LPCIO Vendor		AsusTek
2LPCIO Model		AS99127F
3LPCIO Vendor ID		0x5CA3
4LPCIO Chip ID		0x59
5LPCIO Revision ID	0x53
6Config Mode I/O address	0x2E
7Config Mode LDN		0xB
8Config Mode registers	
9		00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 
10	00	FF FF FF FF FF FF FF 0B FF FF FF FF FF FF FF FF 
11	10	FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
12	20	59 53 FF FE 80 00 00 FF 00 0C 7C F0 1F 00 00 FF 
13	30	00 FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
14	40	FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
15	50	FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
16	60	00 00 FF FF FF FF FF FF FF FF FF FF FF FF FF FF 
17	70	00 FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF 

I have included both reports as well as the above screenshot in the zip file attached for your assistance. If there's anything else you need from me, feel free to ask.

Thanks, I'll have a look. Amazingly the LPCIO as a device disappeared, but the hardware monitoring outputs of this LPCIO appeared. Most likely a bug somewhere.

red-ray wrote on 2020-11-06, 22:36:

CPUZ fails to report tRFC, should it?

no, but it's a missing. Thanks !

A Windows 3.11 version (with Win32s installed) of CPU-Z would be fine. I think my TI486slc/IIT387 machine would be too slow for Win 9x.

Hello, i tested CPU-Z Vintage edition 1.02.0 on my current PC for Win98 games (installed system is Windows ME) and almost all is detected, except graphics card. On this tab is only Revision field filled and that's all. And second thing: on SPD tab, first memory module don't have manufacturing week and year filled, but in txt report is reported correctly - interesting. Report attached
GPU is Nvidia GF2 MX400 with 64 MB
Driver version 45.23

https://valid.x86.fr/tmse16

It's actually an Embedded Pentium 266, not a mobile one. But I guess there's no way for CPU-Z to tell the difference as both apparently use the same exact die.

STB Nitro DVD 4MB AGP.

Here's a few. AST Bravo MS Motherboard (HX chipset) with an Intel P133 (@133MHz), a Cyrix 6x86L P200+ (@133MHz - I can't do 75MHz bus on this board!) and a Cyrix 6x86 P150+ (@120MHz, and a million bent pins, and this got HOT!).

The board has 16Mb (2x8MB FPM SIMMS), no L2 cache (no COAST module) and an ATI Mach64 graphics chipset.

Cheers,
Steve