***PLEASE SEE MY LATEST POST with attachment v.30

I'm trying to verify that IRQ 12 is working by performing a cpu reset whenever the mouse generates the required int. However, nothing I have tried will trigger irq12. Enclosed is a zip file of my code for examination. My keyboard irq works just fine.. thanks

well here it is 2 weeks later and I still ain't figured it out... guess nobody else has a clue either.. lol

Could it be simply that your USB mouse doesn't generate IRQ12?

Can you point out where the mouse code is?

Two things that could be the problem are: (1) The mouse and IRQ12 haven't been enabled on the keyboard [804x] controller (2) there are other (higher priority) IRQs firing first that have not been cleared on the IRQ [8259] controller, preventing IRQ12 being serviced.

to activate the mouse you need some specific initialization code beyond the Keyboard Controller. I looked through your code but couldn't find that code anywhere. its fairly complex as you need to wait for the Keyboard Controller to respond to about 3 commands (in a specific order). the mouse won't do anything until you send it the signal to activate.

on a lighter note; are you using a PS/2 keyboard with a USB mouse? it might have USB support disabled in the BIOS.

- Stefan

Ok, after taking several months off, cause trying to get the mouse working had me ready to jump off a cliff, i've come back to it. I've included the mouse code to activate in the syscom.asm file in the new attachment, but now, it automatically triggers IRQ12 upon activation and gets stuck. The keyboard will no longer work. I just don't understand what's happening.. can someone please help me to understand where I'm going wrong? I can't continue to build on my OS without the mouse. To go where I'm going with it, the mouse is essential to its development. Thanks.

When you say "stuck", does that mean it can't release the IRQ and continually goes back to IRQ12, or that it keeps looping within the IRQ12 handler and never gets to the IRET?

i believe it's not getting to the IRET. it's like it triggers irq12, displays message and freezes..

Can you try to turn the "belief" into a proof? Print a character to the screen at strategic points in the handler to see where things start to go bad. "A" at entry, "B" at some decision point, "C" at another decision point, etc.

ok. it just prints character after character.. it wont issue the IRET. I'm certain.

it won't even jump to another location outside the IRET

when enabling the mouse, this code here triggers the IRQ12.


..and the IRET doesn't get fired to send it back to the kernel loop where you can type on the keyboard. If I omit this part of the mouse activation, of course, IRQ12 never happens and the system(OS) works as intended.

So you need to determine why. What condition is making it loop? Why doesn't the condition change?

that I don't have a clue.. even if I omit all the code in the IRQ - so that it is just a blank call, it still gets stuck..

Become friends with the character printing. Without a debugger, printing is a good way to see what is happening. Create a function to print all the registers and call it at various point in the handler. There is so much you can learn about the code by printing the current state, or printing some specific values, or seeing which bits are set/cleared, etc.

ok.. thanks. I will give it a shot.

the last file you published (I'm the 1 downloader apparently) has a severe error in file 'syscom.asm'. line 27 embeds a character string in the code. I just started looking at it.

Stefan

ok.. i deleted lines 27 and 28. I wasn't sure about those either. I'm still trying to understand what I'm coding. Anyways, something strange happened, or at least to me it's strange. If I add my keyboard routine to IRQ12, both the mouse and keyboard work as expected. However, if I omit the keyboard routine from either IRQ1 or IRQ12, it gets hung, nothing happens.