SysInfo 4.3 ready, finally

[Geijer]

Changelog:

4.3:

• Bootable disk image (adf) distribution
• Identifies APOLLO-Core 68080 CPU/FPU
• Better hanfling of 68060 CPU/FPU identification
• Bugfix: 68040 non FPU guru fixed
• Experimental adf release http://sysinfo.d0.se
• Documentation in AmigaGuide format and experimental html

4.2

• Intermediate release with better utilization of pipelining in 68040 and newer CPUs in "Dhrystone" calculation implementation.

SysInfo_4.3.lha or test the new adf version SysInfo_4.3.adf

Looking forward for feedback.

[FOL]

Originally Posted by Geijer

Changelog:

4.3:

• Bootable disk image (adf) distribution
• Identifies APOLLO-Core 68080 CPU/FPU
• Better hanfling of 68060 CPU/FPU identification
• Bugfix: 68040 non FPU guru fixed
• Experimental adf release http://sysinfo.d0.se
• Documentation in AmigaGuide format and experimental html

4.2

• Intermediate release with better utilization of pipelining in 68040 and newer CPUs in "Dhrystone" calculation implementation.

SysInfo_4.3.lha or test the new adf version SysInfo_4.3.adf

Looking forward for feedback.

lha link broken. Not found.

[Geijer]

Ouch, character case error, fixed now. Thanks

[LanceT]

Both links are for the LHA-file can manually change it to .adf.

Under WinUAE icon looks distorted (problem solved second image).
Kind of funny comments though?

[Geijer]

Originally Posted by LanceT

Both links are for the LHA-file can manually change it to .adf.

Under WinUAE icon looks distorted (problem solved second image).

Fixed the link, will try to reproduce the mouse pointer error.

Thanks

[Toni Wilen]

It is normal if OCS Agnus + ECS Denise + KS 3.1. OCS Agnus + ECS Denise is not supported by KS, it assumes ECS Agnus if ECS Denise is found. (Commonly, especially in emulation sprite corruption is not always visible because by default chip ram is cleared after powerup)

Last line having garbage is another clue.

[AMIGASYSTEM]

Because I on an RTG system get lower results from the user LanceT despite I have a very fast and performing system

[AMIGASYSTEM]

@LanceT

What PC are you using which CPU?

[Mark sealey]

Aweomme...I`ll give this a try on my A600 Furia.

[LanceT]

Originally Posted by AMIGASYSTEM

@LanceT

What PC are you using which CPU?

Still using a system I built 8 or so years ago.
Motherboard:
CPU Type :HexaCore AMD Phenom II X6 Black Edition 1100T, 3647 MHz (18.5 x 197) Base = 3.8GHz Boost = 4.2GHZ
Motherboard Name : Asus M4A89GTD Pro/USB3 (2 PCI, 1 PCI-E x1, 1 PCI-E x4, 2 PCI-E x16, 4 DDR3 DIMM, Audio, Video, Gigabit LAN, IEEE-1394)


*** And I kind of cheated. I booted straight into the SYSINFO file. Not to Workbench ***

[AMIGASYSTEM]

Originally Posted by LanceT

*** And I kind of cheated. I booted straight into the SYSINFO file. Not to Workbench ***

The difference is not made by direct boot from Floppy , the difference is made by the PC in use , i used an old Pentium Dual-Core (2005) E6300 2.80 GHz 4GB RAM with Win7 32Bit

[PeterK]

I would suggest to remove the decimal point and the 2 places behind it and expand the range for the benchmark results by 3 additional digits, because emulators like WinUAE are far too fast, their results are causing output overflows and then the highest digits are truncated on the left side. You can see that in the 2. screenshot from LanceT in post #4.

This range expansion should be done for output values of 1000 or higher.

Btw, it seems that the field for the Dhrystones result is not properly cleared before another speed test writes out a new value.

[honx]

on aca500plus with aca1221 sysinfo still displays 68ec020, which is wrong... don't mix up aca1221 with aca1221(ec).
i have aca1221 (real 020, no "ec"), 28mhz and 62mb fast ram on it. 020 "ec" cpu could not address more than 16mb.

[Geijer]

Originally Posted by PeterK

I would suggest to remove the decimal point and the 2 places behind it and expand the range for the benchmark results by 3 additional digits, because emulators like WinUAE are far too fast, their results are causing output overflows and then the highest digits are truncated on the left side. You can see that in the 2. screenshot from LanceT in post #4.

In progress on thins one, but decided to do small increments and release more often instead of trying to fix all at once and never release

Originally Posted by honx

on aca500plus with aca1221 sysinfo still displays 68ec020, which is wrong... don't mix up aca1221 with aca1221(ec).
i have aca1221 (real 020, no "ec"), 28mhz and 62mb fast ram on it. 020 "ec" cpu could not address more than 16mb.

Hmm, shouldn't be possible, need to look into this one. Will add it as a bug. Thanks.
The code isn't 100% but shouldn't miss a non EC020 when fastram exists outside of 24bit adresses.

[Geijer]

Originally Posted by PeterK

Btw, it seems that the field for the Dhrystones result is not properly cleared before another speed test writes out a new value.

Thanks, added to the list of bugs...

[solarmon]

I get the following yellow Guru, when I use the ADF version:

68040 Installation Alert.
The 68040.library requires an 68040, do not use it on a 68060.
No exceptions, no excuses, no discussion. It won't work this way.

But this is on an A1200 rev 2B with only an A1208 memory expansion card in 4mb mode.

I don't understand why this is shown in my setup?

[Geijer]

Originally Posted by solarmon

I get the following yellow Guru, when I use the ADF version:

But this is on an A1200 rev 2B with only an A1208 memory expansion card in 4mb mode.

I don't understand why this is shown in my setup?

Hmm, it's not a guru, it's a Recoverable DisplayAlert (easiest form of pop-up dialogue on Amiga). It informs you that SysInfo is trying to load 68040.library on a system without 68040. Thought I had fixed it. Nothing will break, the 68040.library just won't load on your system.

It is included to support as many different combinations of CPU / FPU / MMU as possible.

[solarmon]

Thanks for the correction - I was just too lazy to lookup the correct term for it.

OK, I'll just ignore it then.

Thank you for continuing to update this great tool!

[nogginthenog]

An idea: I appreciate Commodore never release a 68060 Amiga (I think!) but it would be nice to include 68060@50Mhz in the CPU comparison list.

[Locutus]

All the other speed bars are stock configurations.

With the 060 it becomes the question, comparison against which 060 board? There is a big difference between a 68060 on a 3640 and on a Cyberstorm MKIII.

(Okay, this benchmark might be small enough it fits in 060 instruction cache and thus differences in memory speed might not matter that much but you get the idea)