Mumak wrote on 2021-07-04, 19:22:

RichB93 wrote on 2021-07-04, 17:20:

Working well for me! Just a post to say thank you, and also to submit my log for posterity.

Thank you! What monitor do you have, I retrieved only its ID as EXN0000.

I spotted that too; I'm using an Extron RGB HDMI scaler which reports EDID information based on the output resolution configured.

I see, thanks. Will add that into my database.

Mumak wrote on 2021-07-04, 19:26:

PC@LIVE wrote on 2021-07-04, 19:09:

Yes, thanks it worked well, for the CPU I think it depends on the fact that it is a Cyrix with a different brand, so it is not w […]
Show full quote

Yes, thanks it worked well, for the CPU I think it depends on the fact that it is a Cyrix with a different brand, so it is not wrong to identify it.
Regarding the cache it rightly detects 0KB, because the present one of 256KB is fake.
Today I ran the program on another PC, an S7 P200MMX i430VX chipset, I am attaching the created files, unfortunately I have to zip them because one cannot be attached with its extension.
I will see in the next days to start other old-PCs, and run the program to create the files to attach.

Thanks again! Looks good, is the audio adapter ALS-100 Plus ?

Yes, in fact it says so on the chip.
I put a picture to show you the card, I don't know if it will interest you
I will see in the coming days to start other PCs to create other files to attach.

Just tried 6.1.4 on my VIA Epia ML6000EA board for fun and noticed a couple of minor issues:

* On program startup it thinks I'm running under Windows DOS mode, although I'm booting directly into MSDOS 7.x (copied from Windows 98 SE).

* The CPU clock speed is reporting ~350-360MHz, but it should be more like 667MHz (VIA Eden Nehemiah).

mscdex wrote on 2021-07-05, 05:28:

* On program startup it thinks I'm running under Windows DOS mode, although I'm booting directly into MSDOS 7.x (copied from Windows 98 SE).

I thought MSDOS 7.x was the same as using Windows DOS mode, even when installed on its own.

That's right, MSDOS 7.x is no longer the "good old" real mode DOS as it used to be.
Regarding the wrong CPU clock on VIA Eden, I'd need to see the HWiNFO Report and Debug Files (run HWINFO -r -d) to check what might be wrong.

Mumak wrote on 2021-07-05, 07:29:

That's right, MSDOS 7.x is no longer the "good old" real mode DOS as it used to be.

I wasn't sure how the detection was being done. It was just a bit confusing because Windows is definitely not running or installed even.

Mumak wrote on 2021-07-05, 07:29:

Regarding the wrong CPU clock on VIA Eden, I'd need to see the HWiNFO Report and Debug Files (run HWINFO -r -d) to check what might be wrong.

Attached. I also noticed that some of the CPU features seem to be duplicated in the log? For example the first CMOV and MMX entries both show as enabled but the duplicate copies show as not enabled (x)?

mscdex wrote on 2021-07-05, 14:08:

I wasn't sure how the detection was being done. It was just a bit confusing because Windows is definitely not running or install […]
Show full quote

Mumak wrote on 2021-07-05, 07:29:

That's right, MSDOS 7.x is no longer the "good old" real mode DOS as it used to be.

I wasn't sure how the detection was being done. It was just a bit confusing because Windows is definitely not running or installed even.

Mumak wrote on 2021-07-05, 07:29:

Regarding the wrong CPU clock on VIA Eden, I'd need to see the HWiNFO Report and Debug Files (run HWINFO -r -d) to check what might be wrong.

Attached. I also noticed that some of the CPU features seem to be duplicated in the log? For example the first CMOV and MMX entries both show as enabled but the duplicate copies show as not enabled (x)?

Thanks, will check that. Could you please also attach the HWiNFO Debug File? This is produced when running with the -d switch.

Mumak wrote on 2021-07-05, 14:25:

Could you please also attach the HWiNFO Debug File? This is produced when running with the -d switch.

Sorry about that, updated the post.

Thanks. The problem with CPU clock (and some others) is due to MSDOS 7, which doesn't allow certain privileged operations because it runs in V86 mode.

Today I ran the program on two other PCs, one S7 i430VX with AMD K6-2-360 CPU (it's a 400 set as 6X60), the other is a 486-DX2-66 VLB Cyrix CPU (green), both MB have a problem with the L2 cache, only by disabling it the PCs work.
I put some pictures, in the K6-2-360 I had installed the version for Windows, the program works with w95a installed as OS, the only thing is that I no longer see the references with other PCs, and therefore I cannot compare the values ​​obtained.
I'll put the files created with both PCs as soon as I can.
Returning to the K6-2-360, I noticed that the dos version identifies the MB as PCCHIPS M530, in addition to the AMPTRON models, but in fact the MB PCChips is not M530 but is M549, probably the BIOS used will be the same, also because it is unable to identify at startup the CPU as amd k6-xxx, it recognizes it as -MMX 66MHz, instead your program recognizes it exactly, even if the CPU used by me is a CXT (I don't know if I can identify it from the multi 6X).

Mumak wrote on 2021-07-05, 15:03:

Thanks. The problem with CPU clock (and some others) is due to MSDOS 7, which doesn't allow certain privileged operations because it runs in V86 mode.

With all due respect, I am quite certain that this is not due to running MS-DOS 7 (or 7.1), but more likely to emm386.exe being loaded and switching the CPU into v86 mode .
Case in point, under DOS 7.1, I don't get the error message if EMM386.EXE is not loaded, but I do get it if EMM386.EXE is loaded.
Additionally, if MS-DOS 7.x had broken real-mode compatibility there would be a lot of documented incompatible software issues about it . There are none, that I know of, at the very least .

Without EMM386.EXE under DOS 7.1 :

With EMM386.EXE under DOS 7.1 :

PC@LIVE wrote on 2021-07-05, 19:28:

Today I ran the program on two other PCs, one S7 i430VX with AMD K6-2-360 CPU (it's a 400 set as 6X60), the other is a 486-DX2-6 […]
Show full quote

Today I ran the program on two other PCs, one S7 i430VX with AMD K6-2-360 CPU (it's a 400 set as 6X60), the other is a 486-DX2-66 VLB Cyrix CPU (green), both MB have a problem with the L2 cache, only by disabling it the PCs work.
I put some pictures, in the K6-2-360 I had installed the version for Windows, the program works with w95a installed as OS, the only thing is that I no longer see the references with other PCs, and therefore I cannot compare the values ​​obtained.
I'll put the files created with both PCs as soon as I can.
Returning to the K6-2-360, I noticed that the dos version identifies the MB as PCCHIPS M530, in addition to the AMPTRON models, but in fact the MB PCChips is not M530 but is M549, probably the BIOS used will be the same, also because it is unable to identify at startup the CPU as amd k6-xxx, it recognizes it as -MMX 66MHz, instead your program recognizes it exactly, even if the CPU used by me is a CXT (I don't know if I can identify it from the multi 6X).

Thanks, I will adjust the mainboard model to show M549 too, most likely they all use the same ID.
As for the K6-2 CXT, this should be shown on the second screen with CPU details as stepping.

darry wrote on 2021-07-05, 19:57:

With all due respect, I am quite certain that this is not due to running MS-DOS 7 (or 7.1), but more likely to emm386.exe being […]
Show full quote

Mumak wrote on 2021-07-05, 15:03:

Thanks. The problem with CPU clock (and some others) is due to MSDOS 7, which doesn't allow certain privileged operations because it runs in V86 mode.

With all due respect, I am quite certain that this is not due to running MS-DOS 7 (or 7.1), but more likely to emm386.exe being loaded and switching the CPU into v86 mode .
Case in point, under DOS 7.1, I don't get the error message if EMM386.EXE is not loaded, but I do get it if EMM386.EXE is loaded.
Additionally, if MS-DOS 7.x had broken real-mode compatibility there would be a lot of documented incompatible software issues about it . There are none, that I know of, at the very least .

Without EMM386.EXE under DOS 7.1 :
without_emm386.png

With EMM386.EXE under DOS 7.1 :
with_emm386.png

I guess you're right, it's most likely indeed the EMM386 switching to V86 and hence prohibiting some low-level operations.
Thanks for pointing this out!

darry wrote on 2021-07-05, 19:57:

With all due respect, I am quite certain that this is not due to running MS-DOS 7 (or 7.1), but more likely to emm386.exe being loaded and switching the CPU into v86 mode .
Case in point, under DOS 7.1, I don't get the error message if EMM386.EXE is not loaded, but I do get it if EMM386.EXE is loaded.
Additionally, if MS-DOS 7.x had broken real-mode compatibility there would be a lot of documented incompatible software issues about it . There are none, that I know of, at the very least .

You're right. I just checked with EMM386 disabled and not only is the CPU speed being correctly reported now, but no more warning about running under Windows DOS mode (and of course no more EMM386 warning either).

Mumak wrote on 2021-07-05, 07:29:

That's right, MSDOS 7.x is no longer the "good old" real mode DOS as it used to be.

actually it is, you can install 7.x without windows perfectly fine.

Just ran this on my PS/2 Model 70 and looking good. Can I make one suggestion though - give a little message bottom right when it's working out what to display (like you do when finding machine brand or at the bottom of the screen when looking for dongles in parallel port). This little 386 takes ages to draw some screens 😀 - see here https://photos.app.goo.gl/4Usw6WWfWvowk3G59

SteveC wrote on 2021-07-06, 14:31:

Just ran this on my PS/2 Model 70 and looking good. Can I make one suggestion though - give a little message bottom right when it's working out what to display (like you do when finding machine brand or at the bottom of the screen when looking for dongles in parallel port). This little 386 takes ages to draw some screens 😀 - see here https://photos.app.goo.gl/4Usw6WWfWvowk3G59

Thanks for the feedback, indeed it might need some indication on such machines. Could you please provide the HWiNFO Report + Debug Files for my reference? Don't have many with MCA 😀

Mumak wrote on 2021-07-06, 14:40:

SteveC wrote on 2021-07-06, 14:31:

Just ran this on my PS/2 Model 70 and looking good. Can I make one suggestion though - give a little message bottom right when it's working out what to display (like you do when finding machine brand or at the bottom of the screen when looking for dongles in parallel port). This little 386 takes ages to draw some screens 😀 - see here https://photos.app.goo.gl/4Usw6WWfWvowk3G59

Thanks for the feedback, indeed it might need some indication on such machines. Could you please provide the HWiNFO Report + Debug Files for my reference? Don't have many with MCA 😀

I tried but the Video Info page crashes the machine. The CRT does this - https://photos.app.goo.gl/9sYTPLDfk8t3mHyp6, and an LCD panel just says no signal. Attached are the LOG and DBG files.

Thanks. Looks like it doesn't like some of the VGA checks, but it won't be easy to figure out exactly what is causing this.
It should be displaying detection progress at the bottom, are you perhaps able to spot what's shown as the last one?