Os Test Version Out for Test Report

[Slasher]

Hi,
at http://www.geocities.com/akabakwu/os.html you'll find my os I've been working on. It works off a floppy disk so no FEAR of it damaging your hard disk.
You can write the floppy image with WinImage,rawwrite or partcopy. Please try it out and tell me if it boots on you system and if when you press the 'a' and 'z' keys the console changes. 'a' moves to the next console while 'z' moves to the previous console. Also the floppy driver just reads and write a test sector and writes some bytes read in hex on the first console in MAGENTA(its a test task).

Console 1 shows your memory layout and memory allocated to tasks
Concole 2 shows your VESA Info
In console 4, when you hold down keys on the keyboard and press 'c' it shows a map that corressponds to the SCANCODE of all the keys you held down when pressing 'c'.
this is the basis for my multiple keypress interface for games and co

I've tried this OS
in Boch 2.0 with no problems
on a P200 32mb 1GB HD ps2 keyboard and mouse NVidia TNT2 32 video card system with no problems
on an Athalon XP 2.6GHZ 256mb 80GB HD ps2 keyboard and mouse NVidia TNT2 32 video card system with no problems

[Peter_Vigren]

I like the page... I know this haven't anything to do with the testing... but I just wanted you to know... The only thing that is bothering me is that you use Times New Roman... especially when your heading is not... Else... I like it.

[Unwritten Axiom]

What are your OS's system requirements?
On my 386 (4MB RAM), it starts and hangs before it even prints any text on the screen.
This isn't likely your intended target system, but it doesn't run it, just so you know.

[Slasher]

Well,
a 386 with 4mb, color vga(min), keyboard ps2 mouse(optitional no serial support yet),floppy drive is all that this version tests and works with so it should work.
It could be a bad download or bad image write(happens a lot ) that caused the problem cause if you get past the blue options screen then you are in the os, but before then its the boot sector that's running and nothing fancy is in there.
Try deleting kboot.bin (you can rename it) to see if a red error message is displayed,this is in the boot code.
if you get that message, put kboot.bin back and delete (rename) setup.bin and see if you get a green error message, this is in the 2nd stage loader.

i used VESA function 0x4xxx to get the VESA info and memory map function 0xe820 (or something similar) for the memory map info. I didn't perform a check to see if the hardware supports them so this could be the problem but then again these functions are call WELL into the 2nd stage loader and by then a lot of text has been displayed

[bkilgore]

I tried testing with bochs 2.0.2 with the following results:

- The boot itself works fine. It asks me for what size screen I want (I picked 640x480) how many colors (i chose 256) and then about the mouse (I said ps/2).

- At this point, it looked like some red and green text flashed on the screen, and then bochs bugged out with the following error:

Bochs is exiting with the following message:
[PIT ] pit: io write to port 0043, len=2

I tried changing the screen size but that didn't help, nor did selecting a different mouse type.

I tried grabbing a screen shot between when the text displayed and bochs died. The best I could get was this, and it flashed too quickly for me to be able to say if there was more than this or not...

Kernel Start: 9000
Kernel End: 10000
Kernel Size: 7000 bytes
A20 Line Set
DMA Controller Initialized
Total Memory Size [31 MB]
Page Dir: 65536
Page Table: 69632
page dir: 10000
table index: 0
page index: 200
table ptr: 11000

Hope this helps...

- Brandon

[Ozguxxx]

Hey I got same errors with bkilgore with bochs test but when I set mouse type as 3rd option (its usb I think) in blue font screen, it just hung, you can check last state below from bochsout.txt:
00059148000i[CPU ] | EAX=00170000 EBX=0000001e ECX=0017000a EDX=00000000
00059148000i[CPU ] | ESP=0000077a EBP=00000000 ESI=000007b0 EDI=000007ae
00059148000i[CPU ] | IOPL=0 NV UP DI PL ZR NA PE NC
00059148000i[CPU ] | SEG selector base limit G D
00059148000i[CPU ] | SEG sltr(index|ti|rpl) base limit G D
00059148000i[CPU ] | DS:0040( 0000| 0| 0) 00000400 0000ffff 0 0
00059148000i[CPU ] | ES:b800( 0000| 0| 0) 000b8000 0000ffff 0 0
00059148000i[CPU ] | FS:0000( 0000| 0| 0) 00000000 0000ffff 0 0
00059148000i[CPU ] | GS:0000( 0000| 0| 0) 00000000 0000ffff 0 0
00059148000i[CPU ] | SS:0000( 0000| 0| 0) 00000000 0000ffff 0 0
00059148000i[CPU ] | CS:f000( 0000| 0| 0) 000f0000 0000ffff 0 0
00059148000i[CPU ] | EIP=0000e869 (0000e869)
00059148000i[CPU ] | CR0=0x60000010 CR1=0x00000000 CR2=0x00000000
00059148000i[CPU ] | CR3=0x00000000 CR4=0x00000000

Hope this helps you, good luck.

[Slasher]

Can you please compare your bochs setup text file with mine here.

[PIT ] pit: io write to port 0043, len=2

is not an error,I'm reprogramming the PIT to 100Hz. choose to continue,the os will run.
In my bochs setup i have,

romimage: file=../BIOS-bochs-latest, address=0xf0000
vgaromimage: ../VGABIOS-lgpl-latest
panic: action=ignore
error: action=report
info: action=report
debug: action=ignore

The setup screen where you choose your screen size,number of colors etc does not affect the os yet,its just a place holder for gathering the info when i want to add the gui.
Thanks

[Ozguxxx]

Now some screen giving time and other stuff comes up but I cant do anything with a and z, no console changes... Also when I move the mouse screen is filled with mouse int all over the place...

[Slasher]

hi, the input is case sensitive. Is the capslock off? turn it off. I've replaced the download with another one thats NOT case sensitive.
The time output and 'mouse int ' message is a good thing ;D means it is working. the other stuff is the output from the floppy driver, it reads a sector from the disk and display the first 16bytes. Thanks a lot

[Slasher]

hi,
has anyone tried it on REAL pc hardware? If you have, could you please state the componets and if it worked. Thanks

[Pype.Clicker]

Can you please compare your bochs setup text file with mine here.

[PIT ] pit: io write to port 0043, len=2

is not an error,I'm reprogramming the PIT to 100Hz. choose to continue,the os will run.
In my bochs setup i have,

romimage: file=../BIOS-bochs-latest, address=0xf0000
vgaromimage: ../VGABIOS-lgpl-latest
panic: action=ignore

Imho, ignoring the panics from BOCHS is a bad idea ...
and the "write len=2" *is* an error for bochs. The PIT ports are 8bits-wide in the specs, and you do 16-bits out (probably some out dx,ax instead of out dx,al somewhere), which is not standard-compliant and could therefore lead to unpredictable results on some real hardware platform...

[Slasher]

thanks pype, changed the out dx,ax to out dx,al in my PIT programming code. No more panics. ;D
Anyone tried on real hardware yet? Works fine on mine (specs stated in previous post)