izaya/collapseos
1
0
Fork 0
mirror of https://github.com/hsoft/collapseos.git synced 2024-11-01 19:10:56 +11:00
Code Issues Releases Wiki Activity
08328b86a9
collapseos/emul/zasm/glue.asm

131 lines
2.0 KiB
NASM
Raw Normal View History

Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
; Glue code for the emulated environment
zasm: use blkdev IDs as input and output This will soon allow use to seek and tell on input, which is necessary for a second pass which is necessary for forward symbol references. This require making `blkSel` a bit more flexible. Rather than having one global selected blkdev, each app can select its own, in its own memory space.
2019-05-11 08:20:43 +10:00
.equ RAMSTART 0x4000
zasm now bootstraps itself without scas!
2019-05-20 05:03:34 +10:00
.equ USER_CODE 0x4800
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
.equ STDIO_PORT 0x00
zasm: change ioRewind to ioSeek Will need it soon.
2019-05-16 03:41:56 +10:00
.equ STDIN_SEEK 0x01
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
.equ FS_DATA_PORT 0x02
.equ FS_SEEK_PORT 0x03
zasm emul: add dumpSymbolTable debug routine Very helpful...
2019-05-18 03:22:45 +10:00
.equ STDERR_PORT 0x04
Make the shell a userspace app That's my mega-commit you've all been waiting for. The code for the shell share more routines with userspace apps than with kernel units, because, well, its behavior is that of a userspace app, not a device driver. This created a weird situation with libraries and jump tables. Some routine belonging to the `kernel/` directory felt weird there. And then comes `apps/basic`, which will likely share even more code with the shell. I was seeing myself creating huge jump tables to reuse code from the shell. It didn't feel right. Moreover, we'll probably want basic-like apps to optionnally replace the shell. So here I am with this huge change in the project structure. I didn't test all recipes on hardware yet, I will do later. I might have broken some... But now, the structure feels better and the line between what belongs to `kernel` and what belongs to `apps` feels clearer.
2019-11-16 07:37:49 +11:00
.inc "err.h"
.inc "ascii.h"
.inc "blkdev.h"
.inc "fs.h"
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
zasm emul: bring back kernel/user distinction It was a bad idea to remove it. Now that I'm introducing the concept of a per-app glue file, it becomes much easier to build emulated zasm as a userspace app.
2019-05-20 02:57:59 +10:00
jp init ; 3 bytes
; *** JUMP TABLE ***
fs: check for file size bounds in GetC
2019-06-01 01:12:29 +10:00
jp strncmp
jp upcase
jp findchar
jp blkSel
fs: further adjust to previous blkdev refactoring Previous refacoring broke all seek/tell within fs. fs handles now lost the responsibility to keep track of current position. It's blkdev's job.
2019-06-05 10:45:01 +10:00
jp blkSet
fs: check for file size bounds in GetC
2019-06-01 01:12:29 +10:00
jp fsFindFN
jp fsOpen
Rename blockdev's API routines to GetB/PutB The goal is to avoid mixing those routines with "character devices" (acia, vpd, kbd) which aren't block devices and have routines that have different expectations. This is a first step to fixing #64.
2019-10-31 07:59:35 +11:00
jp fsGetB
jp _blkGetB
jp _blkPutB
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
jp _blkSeek
jp _blkTell
zasm: print progress indicator while assembling
2019-06-20 01:42:39 +10:00
jp printstr
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
zasm: rename #inc to .inc scas, it's not needed any more.
2019-10-07 05:32:23 +11:00
.inc "core.asm"
Extract str.asm from core.asm and make core included by userspace Most of register fiddling routines (which is now the only thing contained in care.asm) are used by almost all userspace apps, often in inner loops. That makes the penalty of using jump tables for those a bit too high. Moreover, it burdens jump tables needlessly. Because this unit is very small (now that string routines are out), it makes sense to always include it in binaries.
2019-11-15 01:55:31 +11:00
.inc "str.asm"
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
.equ BLOCKDEV_RAMSTART RAMSTART
.equ BLOCKDEV_COUNT 3
zasm: rename #inc to .inc scas, it's not needed any more.
2019-10-07 05:32:23 +11:00
.inc "blockdev.asm"
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
; List of devices
Rename blockdev's API routines to GetB/PutB The goal is to avoid mixing those routines with "character devices" (acia, vpd, kbd) which aren't block devices and have routines that have different expectations. This is a first step to fixing #64.
2019-10-31 07:59:35 +11:00
.dw emulGetB, unsetZ
.dw unsetZ, emulPutB
.dw fsdevGetB, fsdevPutB
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
zasm: print progress indicator while assembling
2019-06-20 01:42:39 +10:00
.equ STDIO_RAMSTART BLOCKDEV_RAMEND
stdio: remove a layer of indirection in GetC/PutC We use zasm's ability to use labels in .equ directive. We didn't do it before because for a while, we were in between scas and zasm (scas was used in automated tests) so we needed to use the lowest common denominator: zasm doesn't have macros and scas can't use labels in .equ directives. This forced us to add this layer of indirection. But now that we are completely cut from scas' dependency, we can use this nice zasm's ability.
2019-11-05 01:55:12 +11:00
.equ STDIO_GETC noop
.equ STDIO_PUTC stderrPutC
zasm: rename #inc to .inc scas, it's not needed any more.
2019-10-07 05:32:23 +11:00
.inc "stdio.asm"
zasm: print progress indicator while assembling
2019-06-20 01:42:39 +10:00
.equ FS_RAMSTART STDIO_RAMEND
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
.equ FS_HANDLE_COUNT 0
zasm: rename #inc to .inc scas, it's not needed any more.
2019-10-07 05:32:23 +11:00
.inc "fs.asm"
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
init:
di
zasm emul: get rid of the kernel/user distinction Made things complicated for nothing.
2019-05-19 23:14:40 +10:00
ld hl, 0xffff
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
ld sp, hl
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
ld a, 2 ; select fsdev
blockdev: make selection structure opaque I'm about to change that structure and I don't want fs to be messed up by this. I'm proceeding step by step...
2019-06-04 23:56:36 +10:00
ld de, BLOCKDEV_SEL
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
call blkSel
call fsOn
zasm: allow .org to be specified from command line Also important for upcoming mega commit...
2019-11-16 01:57:53 +11:00
; There's a special understanding between zasm.c and this unit: The
; addresses 0xff00 and 0xff01 contain the two ascii chars to send to
; zasm as the 3rd argument.
ld a, (0xff00)
ld (.zasmArgs+4), a
ld a, (0xff01)
ld (.zasmArgs+5), a
Make userspace parse args the same way the shell does
2019-06-03 04:05:20 +10:00
ld hl, .zasmArgs
zasm emul: bring back kernel/user distinction It was a bad idea to remove it. Now that I'm introducing the concept of a per-app glue file, it becomes much easier to build emulated zasm as a userspace app.
2019-05-20 02:57:59 +10:00
call USER_CODE
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
; signal the emulator we're done
halt
Make userspace parse args the same way the shell does
2019-06-03 04:05:20 +10:00
.zasmArgs:
zasm: allow .org to be specified from command line Also important for upcoming mega commit...
2019-11-16 01:57:53 +11:00
.db "0 1 XX", 0
Make userspace parse args the same way the shell does
2019-06-03 04:05:20 +10:00
zasm emul: add dumpSymbolTable debug routine Very helpful...
2019-05-18 03:22:45 +10:00
; *** I/O ***
Rename blockdev's API routines to GetB/PutB The goal is to avoid mixing those routines with "character devices" (acia, vpd, kbd) which aren't block devices and have routines that have different expectations. This is a first step to fixing #64.
2019-10-31 07:59:35 +11:00
emulGetB:
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
; the STDIN_SEEK port works by poking it twice. First poke is for high
; byte, second poke is for low one.
ld a, h
out (STDIN_SEEK), a
ld a, l
out (STDIN_SEEK), a
zasm: Read from and write to streams instead of memory
2019-05-10 05:36:03 +10:00
in a, (STDIO_PORT)
or a ; cp 0
jr z, .eof
cp a ; ensure z
ret
.eof:
zasm: clean up jump table requirements There was a little bit of cruft.
2019-11-12 12:21:13 +11:00
jp unsetZ
zasm: Read from and write to streams instead of memory
2019-05-10 05:36:03 +10:00
Rename blockdev's API routines to GetB/PutB The goal is to avoid mixing those routines with "character devices" (acia, vpd, kbd) which aren't block devices and have routines that have different expectations. This is a first step to fixing #64.
2019-10-31 07:59:35 +11:00
emulPutB:
zasm: Read from and write to streams instead of memory
2019-05-10 05:36:03 +10:00
out (STDIO_PORT), a
zasm: properly raise error on bas ioPutC calls
2019-06-16 06:48:30 +10:00
cp a ; ensure Z
zasm: Read from and write to streams instead of memory
2019-05-10 05:36:03 +10:00
ret
zasm: print progress indicator while assembling
2019-06-20 01:42:39 +10:00
stderrPutC:
out (STDERR_PORT), a
cp a ; ensure Z
ret
Rename blockdev's API routines to GetB/PutB The goal is to avoid mixing those routines with "character devices" (acia, vpd, kbd) which aren't block devices and have routines that have different expectations. This is a first step to fixing #64.
2019-10-31 07:59:35 +11:00
fsdevGetB:
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
ld a, e
out (FS_SEEK_PORT), a
zasm: change ioRewind to ioSeek Will need it soon.
2019-05-16 03:41:56 +10:00
ld a, h
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
out (FS_SEEK_PORT), a
zasm: change ioRewind to ioSeek Will need it soon.
2019-05-16 03:41:56 +10:00
ld a, l
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
out (FS_SEEK_PORT), a
in a, (FS_SEEK_PORT)
or a
ret nz
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
in a, (FS_DATA_PORT)
cp a ; ensure Z
ret
Rename blockdev's API routines to GetB/PutB The goal is to avoid mixing those routines with "character devices" (acia, vpd, kbd) which aren't block devices and have routines that have different expectations. This is a first step to fixing #64.
2019-10-31 07:59:35 +11:00
fsdevPutB:
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
push af
Make blockdev pointers 32 bits This allows us to break through the 64K limit for includes CFS in zasm, a limit we were dangerously close to breaking. In fact, this commit makes us go over that limit. Right in time!
2019-05-29 05:56:39 +10:00
ld a, e
out (FS_SEEK_PORT), a
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
ld a, h
out (FS_SEEK_PORT), a
zasm: iiiiiiiincluuuuuuudes!!1!
2019-05-17 11:15:00 +10:00
ld a, l
out (FS_SEEK_PORT), a
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
in a, (FS_SEEK_PORT)
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
or a
jr nz, .error
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
pop af
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
out (FS_DATA_PORT), a
zasm: properly raise error on bas ioPutC calls
2019-06-16 06:48:30 +10:00
cp a ; ensure Z
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
ret
blockdev: make implementors "random access" This huge refactoring remove the Seek and Tell routine from blockdev implementation requirements and change GetC and PutC's API so that they take an address to read and write (through HL/DE) at each call. The "PTR" approach in blockdev implementation was very redundant from device to device and it made more sense to generalize. It's possible that future device aren't "random access", but we'll be able to add more device types later. Another important change in this commit is that the "blockdev handle" is now opaque. Previously, consumers of the API would happily call routines directly from one of the 4 offsets. We can't do that any more. This makes the API more solid for future improvements. This change forced me to change a lot of things in fs, but overall, things are now simpler. No more `FS_PTR`: the "device handle" now holds the active pointer. Lots, lots of changes, but it also feels a lot cleaner and solid.
2019-06-05 01:53:02 +10:00
.error:
pop af
jp unsetZ ; returns
zasm emul: hook fsdev in glue code
2019-05-17 03:23:23 +10:00
Reference in New Issue Copy Permalink