No, just realmode.

Here a new build to improve dumping of MSRs on IBM BL3. Also adds recognition of SLC2/3 and BL2/3 model based on actual ratio.
@ruthan: Please run this build with -d parameter and attach the HWINFO.DBG file it produced to analyze the issue.
[build pulled, newer available]

Here is log.

ruthan wrote on 2020-05-17, 20:16:

Here is log.

Thanks. Does this build detect the CPU properly now?
[build pulled, newer available]

No its still broken - CPU type is fine but, TDP is still reported with value 130W, i dunno if you are getting it from some HW registers, or you are just using some clever list with these values its wrong. Intel has on ARK 65W.

Other other issue, is that after is mane of MB reported.. it take at least 45s to print out rest of info, its quite a lot, it would be nice to show some progressbar or something like that, because otherwise some less patient user could easily lose patience and reset the system prematurely.

ruthan wrote on 2020-05-17, 22:16:

No its still broken - CPU type is fine but, TDP is still reported with value 130W, i dunno if you are getting it from some HW registers, or you are just using some clever list with these values its wrong. Intel has on ARK 65W.

Other other issue, is that after is mane of MB reported.. it take at least 45s to print out rest of info, its quite a lot, it would be nice to show some progressbar or something like that, because otherwise some less patient user could easily lose patience and reset the system prematurely.

Can you please attach a new report file, so I can see the exact data?
Regarding the very long wait, that's not expected. Can you please take a screenshot where exactly it takes so long?

First menu option - Mainboard.., after MB name is printed.. is that long wait.

ruthan wrote on 2020-05-17, 23:13:

First menu option - Mainboard.., after MB name is printed.. is that long wait.

I guess this is down to some strange behavior of the system PCI BIOS - it reports the system has 128 buses (which is not correct) and scanning them all takes so much time.

I heard that these MBs have more than 1 PCI bus to bypass 133 MB/s .

ruthan wrote on 2020-05-18, 10:05:

I heard that these MBs have more than 1 PCI bus to bypass 133 MB/s .

1,2,3 would be OK, but 128 is way more than really present 😀
If you can provide the new DBG file, I will check the problem with reporting TDP.

I will, but it not my priority, so i could be in hour, or it could take a week..

Here's the latest HWiNFO on the IBM 486BL3. It shows the 3x multiplier.

feipoa wrote on 2020-05-19, 09:56:

Here's the latest HWiNFO on the IBM 486BL3. It shows the 3x multiplier.

Nice, thank you!

Here is new long, TDP is still reported 130W.

What about Sata drives detection crash?

Version 6.1.1 released:
- Added Debug Mode via -d command-line switch.
- Improved detection of IBM Blue Lightning CPUs.
- Improved detection of some 386/486 chipsets.
- Added reporting of supported/current AGP mode.
- Improved detection of some SiS graphics adapters.
- Improved L1/L2 cache detection.

I cannot edit the initial post anymore because of too many edits, so it's attached here.

Tested on my i7 and:
* video detection improved, instead of VGA it shows NVidia unknown (1189) - it's GTX670
* video memory size detected 1GB instead of 2GB
* video mem.clk detected 8MHz
* superIO now detected OK as IT8728F
* missing ATAPI device (DVD) on SATA secondary master (HWI 6.0.5 show it but only sometimes)
* crash happen when do sequence: go to peripheral info, serial port enumeration (IRQs probed) go back, go to peripheral info again

RayeR wrote on 2020-05-21, 02:28:

Tested on my i7 and: * video detection improved, instead of VGA it shows NVidia unknown (1189) - it's GTX670 * video memory size […]
Show full quote

Tested on my i7 and:
* video detection improved, instead of VGA it shows NVidia unknown (1189) - it's GTX670
* video memory size detected 1GB instead of 2GB
* video mem.clk detected 8MHz
* superIO now detected OK as IT8728F
* missing ATAPI device (DVD) on SATA secondary master (HWI 6.0.5 show it but only sometimes)
* crash happen when do sequence: go to peripheral info, serial port enumeration (IRQs probed) go back, go to peripheral info again

That GPU is too new for the DOS version and it's not possible to support it under DOS anyway. On Windows one needs to rely on NVIDIA APIs, which is not available under DOS.

Ok I understand but how from where the mem size and clock came? Maybe rather not display that wrong values for unsupported HW. The clock could be read like nvclock do directly from MMIO but of course it's very specific for various chips and not documented for newer so it's beyond the scope of gereric hwinfo...

RayeR wrote on 2020-05-21, 08:43:

Ok I understand but how from where the mem size and clock came? Maybe rather not display that wrong values for unsupported HW. The clock could be read like nvclock do directly from MMIO but of course it's very specific for various chips and not documented for newer so it's beyond the scope of gereric hwinfo...

Those came from using a method for previous generations, which can't be applied here. I will check how to remove that for later ones.

Here's the log and report files for the IBM 486BL3 using v6.1.1.