Commit Graph

7742 Commits

Author SHA1 Message Date
Kartik Agaram f626421bc4 7692
It's bad enough that metadata comments are restricted to integer literals;
let's at least make them work on _all_ integer literals.
2021-02-07 10:52:04 -08:00
Kartik Agaram 5268b0e1df 7691 2021-02-07 09:53:14 -08:00
Kartik Agaram 74f1512ff1 7690
Convert comments about magic constants into metadata.
2021-02-07 00:20:29 -08:00
Kartik Agaram 6c4c25555c 7689 - permit metadata on Mu literal integers
Metadata is always ignored. It's purely for documentation purposes. But
as long as Mu has no named constants it's starting to feel increasingly
essential.

I'm still not going to bother to add metadata to other parts of the language.
Let's see if we need them. Even though it's a little warty that the rules
vary throughout the stack:

  - bare SubX: metadata everywhere
  - SubX with syntax sugar: no metadata in calls or addressing-mode sigil-expressions
  - Mu: metadata only for literal integers
2021-02-06 22:24:24 -08:00
Kartik Agaram 9ebe6b3876 7688 2021-02-06 22:05:24 -08:00
Kartik Agaram f0ef1ca231 7687 2021-02-05 20:35:23 -08:00
Kartik Agaram c6b8df9d72 7686 2021-02-04 17:25:00 -08:00
Kartik Agaram 55ace79e44 7685 2021-02-03 09:58:52 -08:00
Kartik Agaram 022ecb94f0 7684 - baremetal will have no mouse
I spent a week on trying to support it, and I am now past the five
stages of grief.

-- Important things I read
https://web.archive.org/web/20040604041507/http://panda.cs.ndsu.nodak.edu/~achapwes/PICmicro/keyboard/atkeyboard.html
https://web.archive.org/web/20040604043149/http://panda.cs.ndsu.nodak.edu/~achapwes/PICmicro/mouse/mouse.html

https://wiki.osdev.org/index.php?title=Mouse_Input&oldid=23086#Waiting_to_Send_Bytes_to_Port_0x60_and_0x64
  says command 0xa8 is optional

SaniK: https://forum.osdev.org/viewtopic.php?t=10247
  recommends command 0xa8 before setting Compaq Status byte

Setting Compaq status byte before 0xa8: https://forum.osdev.org/viewtopic.php?f=1&t=19873
  This thread also suggests that keeping reads from keyboard vs mouse straight
  is non-trivial.

-- Where I got stuck
Following SaniK's recipe doesn't seem to work. It seems like not sending
the 0xa8 command gets us a little closer. I saw the mouse handler
trigger, but it seems to not actually happen in response to mouse
events (vector 0x74 in the interrupt descriptor table).

-- Other options that may be worth considering
USB mouse
Serial mouse

Implementing a PS/2 handler in SubX
  would require somehow referring to SubX labels in this file

It seems clear that a mouse driver is complex enough to need a
higher-level language than just hex bytes. But it's _not_ clear how to
_explain_ a mouse driver. There's just a lot of random rules, historical
anecdotes, just-so stories and sheer black magic here. I'm going to try
to do without it all. Mu will be a keyboard-only computer for the
foreseeable future.
2021-02-03 09:49:40 -08:00
Kartik Agaram 7212178bc6 7683 2021-02-01 22:56:50 -08:00
Kartik Agaram cd83a22a59 7682 2021-02-01 13:26:48 -08:00
Kartik Agaram 0167d5c501 7681 2021-02-01 10:29:31 -08:00
Kartik Agaram 4106cf7538 7680 2021-01-31 21:08:28 -08:00
Kartik Agaram 2a9583414b 7679 - baremetal: delete mouse handler
What I have so far is crap. Roll baremetal/boot.hex back to commit 7673.
2021-01-31 21:03:59 -08:00
Kartik Agaram d66bf07b44 7678 2021-01-30 23:41:11 -08:00
Kartik Agaram fe38e59505 7677 2021-01-30 23:39:05 -08:00
Kartik Agaram 78d3df417a 7676 2021-01-30 23:38:39 -08:00
Kartik Agaram 4984d92205 7675 - drop Travis CI 2021-01-28 09:21:44 -08:00
Kartik Agaram 9e315ddf40 7674 - beginning of mouse driver
No handler yet, just initialization.

Bochs boots up; Qemu gets into a reboot loop.

Unlike the keyboard where I did the minimum necessary for Qemu, here I
am blindly copying something alleged to work on "real hardware." Makes
no difference to the failure modes I'm seeing.

Even in this tiny endeavor I see the abyss open up. Poke bytes at some
sequence of ports, read back bytes from some sequence ports; it's like
sending out prayers.
2021-01-28 00:39:50 -08:00
Kartik Agaram fa0e67c172 7673 2021-01-28 00:25:54 -08:00
Kartik Agaram 49badfb697 7672 2021-01-27 21:52:44 -08:00
Kartik Agaram ca5ac5154e 7671
Make some room for a mouse handler.
2021-01-27 21:16:36 -08:00
Kartik Agaram 2c2ba73f65 7670 2021-01-27 21:09:45 -08:00
Kartik Agaram 9ebc8ad439 7569 2021-01-27 20:53:50 -08:00
Kartik Agaram c98476aa09 7568 2021-01-27 20:24:41 -08:00
Kartik Agaram bcb656a7e8 7567 - baremetal: shift-key support 2021-01-27 19:54:05 -08:00
Kartik Agaram 4c9587b5f4 7566 2021-01-27 16:51:23 -08:00
Kartik Agaram 212d72a2df 7565
Somehow everything worked with this bug.
2021-01-24 22:44:43 -08:00
Kartik Agaram 0373ace5e0 7564 2021-01-24 22:25:10 -08:00
Kartik Agaram 5d5b9b2e9b 7563
Yup, a single read suffices. Might not work on real hardware, but YAGNI.
2021-01-24 22:19:43 -08:00
Kartik Agaram d5ed008900 7562 - bochs working again
Holy crap, perhaps the limitations of int 13h were all a mirage. I just
needed to initialize the stack.
2021-01-24 22:00:53 -08:00
Kartik Agaram 4559206243 7561
Another attempt at reorganizing how I read disks. Everything continues
to work in Qemu, but Bochs still doesn't love me.
2021-01-24 21:24:35 -08:00
Kartik Agaram ec75fe7f28 7560 2021-01-24 20:54:53 -08:00
Kartik Agaram 3844651e49 7559 - reorganize sectors built in raw hex
This was tedious for three reasons beyond the usual one of having to
track and update offsets several time while I debug:
  - The Bochs troubles of the previous commit kept polluting my brain
    even though they were irrelevant.
  - I had to keep some changes locally to allow myself to use Bochs,
    which polluted my working directory.
  - I had to travel the long way to the realization that I'm not
    actually initializing the stack anywhere. BIOS was starting my stack
    off at 0x10000, which was promptly clobbered by my second read from
    disk.

The good news: while I'm here I grow the interrupt descriptor table. So
I don't have to go through this exercise when I get back to supporting
the mouse.
2021-01-24 20:16:27 -08:00
Kartik Agaram 1c77be68d7 7558
Bochs is still broken, but before we can fix it we need to make some
room in the boot sector. So we'll spend a few commits reorganizing
things.
2021-01-24 19:43:23 -08:00
Kartik Agaram 0cf7cecedc 7557
Oh, stupid mistake in segmented address calculation. Now Qemu's working
again everywhere. Bochs is again broken everywhere. But I think we're
getting closer. I think Bochs's BIOS implementation for reading sectors
has two interacting constraints:

  - Can't write to more than 0x10000 bytes past segment register.
  - Can't write across segment alignment boundaries.

Qemu only cares about the first.
2021-01-24 19:43:23 -08:00
Kartik Agaram 1c490388d1 7556 2021-01-24 19:43:23 -08:00
Kartik Agaram 09f2f3075a 7555 - snapshot
While baremetal has been working with Qemu, it's been broken with Bochs
since commit 7547, where we started reading more than 63 sectors (1
track) from disk.

Good to know that Bochs simulates native hardware with so much
verisimilitude!

Unfortunately things aren't fixed yet. The current state:

             - Qemu -                 - Bochs -
  ex2.hex    never switches modes     works
  ex2.subx   never switches modes     works
  ex2.mu     never switches modes     fails unit tests

It sucks that Bochs doesn't have strictly superior verisimilitude
compared to Qemu :(
2021-01-24 19:43:23 -08:00
Kartik Agaram dba18f7f6f 7554 - bug in an error handler 2021-01-24 19:43:23 -08:00
Kartik Agaram 0e4e8355f0 7553 2021-01-24 19:43:23 -08:00
Kartik Agaram 8c0b93a9c3 7552 - I better understand a couple of things 2021-01-23 23:14:34 -08:00
Kartik Agaram ac00ea787d 7551 2021-01-23 23:11:24 -08:00
Kartik Agaram 34599e2c63 7550 2021-01-23 11:16:38 -08:00
Kartik Agaram 4254408a38 7549 2021-01-23 11:14:07 -08:00
Kartik Agaram 63be7b7d0d 7548 - baremetal: better cursor management 2021-01-23 08:45:51 -08:00
Kartik Agaram 0f73127ef1 7547 - baremetal: rpn calculator
Still in progress. Known bugs:
* Cursor management is broken. Every line currently seems to leave
  behind a shadow cursor.
* No shift-key support yet, which means no addition or multiplication.
  (This app doesn't have division yet.)
2021-01-22 23:29:04 -08:00
Kartik Agaram a451b2e14f 7546 - baremetal: disk size check 2021-01-22 23:09:30 -08:00
Kartik Agaram 328d76e4ef 7545 2021-01-22 22:16:59 -08:00
Kartik Agaram 12c9e8fff6 7544 2021-01-22 22:15:17 -08:00
Kartik Agaram a688f8f8a1 7543 2021-01-22 22:03:28 -08:00