akkartik
/
mu
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
mu/106stream.subx

78 lines
4.2 KiB
Plaintext
Raw Permalink Normal View History

4949
2019-02-03 06:05:11 +00:00
# streams: data structure for operating on arrays in a stateful manner
#
# A stream looks like this:
5924
2020-01-27 08:36:44 +00:00
# write: int # index at which writes go
# read: int # index that we've read until
6182 - start of support for safe handles So far it's unclear how to do this in a series of small commits. Still nibbling around the edges. In this commit we standardize some terminology: The length of an array or stream is denominated in the high-level elements. The _size_ is denominated in bytes. The thing we encode into the type is always the size, not the length. There's still an open question of what to do about the Mu `length` operator. I'd like to modify it to provide the length. Currently it provides the size. If I can't fix that I'll rename it.
2020-04-03 19:35:53 +00:00
# data: (array byte) # prefixed by size as usual
4949
2019-02-03 06:05:11 +00:00
#
4952
2019-02-06 07:29:15 +00:00
# some primitives for operating on streams:
6182 - start of support for safe handles So far it's unclear how to do this in a series of small commits. Still nibbling around the edges. In this commit we standardize some terminology: The length of an array or stream is denominated in the high-level elements. The _size_ is denominated in bytes. The thing we encode into the type is always the size, not the length. There's still an open question of what to do about the Mu `length` operator. I'd like to modify it to provide the length. Currently it provides the size. If I can't fix that I'll rename it.
2020-04-03 19:35:53 +00:00
# - clear-stream (clears everything but the data size)
4952
2019-02-06 07:29:15 +00:00
# - rewind-stream (resets read pointer)
7842 - new directory organization Baremetal is now the default build target and therefore has its sources at the top-level. Baremetal programs build using the phase-2 Mu toolchain that requires a Linux kernel. This phase-2 codebase which used to be at the top-level is now under the linux/ directory. Finally, the phase-2 toolchain, while self-hosting, has a way to bootstrap from a C implementation, which is now stored in linux/bootstrap. The bootstrap C implementation uses some literate programming tools that are now in linux/bootstrap/tools. So the whole thing has gotten inverted. Each directory should build one artifact and include the main sources (along with standard library). Tools used for building it are relegated to sub-directories, even though those tools are often useful in their own right, and have had lots of interesting programs written using them. A couple of things have gotten dropped in this process: - I had old ways to run on just a Linux kernel, or with a Soso kernel. No more. - I had some old tooling for running a single test at the cursor. I haven't used that lately. Maybe I'll bring it back one day. The reorg isn't done yet. Still to do: - redo documentation everywhere. All the README files, all other markdown, particularly vocabulary.md. - clean up how-to-run comments at the start of programs everywhere - rethink what to do with the html/ directory. Do we even want to keep supporting it? In spite of these shortcomings, all the scripts at the top-level, linux/ and linux/bootstrap are working. The names of the scripts also feel reasonable. This is a good milestone to take stock at.
2021-03-04 06:09:50 +00:00
#
# We need to do this in machine code because streams need to be opaque types,
# and we don't yet support opaque types in Mu.
4949
2019-02-03 06:05:11 +00:00
== code
# instruction effective address register displacement immediate
# . op subop mod rm32 base index scale r32
# . 1-3 bytes 3 bits 2 bits 3 bits 3 bits 3 bits 2 bits 2 bits 0/1/2/4 bytes 0/1/2/4 bytes
5924
2020-01-27 08:36:44 +00:00
clear-stream: # f: (addr stream byte)
5698 Thanks Andrew Owen for reporting this typo.
2019-10-16 02:35:19 +00:00
# . prologue
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
55/push-ebp
89/copy 3/mod/direct 5/rm32/ebp . . . 4/r32/esp . . # copy esp to ebp
4949
2019-02-03 06:05:11 +00:00
# . save registers
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
50/push-eax
51/push-ecx
# eax = f
8b/copy 1/mod/*+disp8 5/rm32/ebp . . 0/r32/eax 8/disp8 . # copy *(ebp+8) to eax
6182 - start of support for safe handles So far it's unclear how to do this in a series of small commits. Still nibbling around the edges. In this commit we standardize some terminology: The length of an array or stream is denominated in the high-level elements. The _size_ is denominated in bytes. The thing we encode into the type is always the size, not the length. There's still an open question of what to do about the Mu `length` operator. I'd like to modify it to provide the length. Currently it provides the size. If I can't fix that I'll rename it.
2020-04-03 19:35:53 +00:00
# var count/ecx: int = f->size
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
8b/copy 1/mod/*+disp8 0/rm32/eax . . . 1/r32/ecx 8/disp8 . # copy *(eax+8) to ecx
6182 - start of support for safe handles So far it's unclear how to do this in a series of small commits. Still nibbling around the edges. In this commit we standardize some terminology: The length of an array or stream is denominated in the high-level elements. The _size_ is denominated in bytes. The thing we encode into the type is always the size, not the length. There's still an open question of what to do about the Mu `length` operator. I'd like to modify it to provide the length. Currently it provides the size. If I can't fix that I'll rename it.
2020-04-03 19:35:53 +00:00
# var max/ecx: (addr byte) = &f->data[f->size]
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
8d/copy-address 1/mod/*+disp8 4/rm32/sib 0/base/eax 1/index/ecx . 1/r32/ecx 0xc/disp8 . # copy eax+ecx+12 to ecx
4949
2019-02-03 06:05:11 +00:00
# f->write = 0
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
c7 0/subop/copy 0/mod/direct 0/rm32/eax . . . . . 0/imm32 # copy to *eax
4949
2019-02-03 06:05:11 +00:00
# f->read = 0
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
c7 0/subop/copy 1/mod/*+disp8 0/rm32/eax . . . . 4/disp8 0/imm32 # copy to *(eax+4)
6072 I just did an experiment, and test_apps is spending over half its time (95 seconds that could be 40) redundantly clearing every stream before every test. And some of those streams are _large_; the translators for SubX and Mu use streams sizes pessimistically for the largest possible program they can handle (1MB per segment). A single test (the one in assort.subx) spends 24 seconds initializing 2 1MB-segments before processing a dozen lines of text in the blink of an eye. I'm not going to speed this up. But good to know.
2020-03-01 06:18:26 +00:00
# - clear all stream data
# - this isn't strictly necessary, and it can slow things down *a lot*, but better safe than sorry.
5924
2020-01-27 08:36:44 +00:00
# var curr/eax: (addr byte) = f->data
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
81 0/subop/add 3/mod/direct 0/rm32/eax . . . . . 0xc/imm32 # add to eax
4949
2019-02-03 06:05:11 +00:00
$clear-stream:loop:
5804 Try to make the comments consistent with the type system we'll eventually have.
2019-12-08 21:56:46 +00:00
# if (curr >= max) break
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
39/compare 3/mod/direct 0/rm32/eax . . . 1/r32/ecx . . # compare eax with ecx
5897 - rename comparison instructions Signed and unsigned don't quite capture the essence of what the different combinations of x86 flags are doing for SubX. The crucial distinction is that one set of comparison operators is for integers and the second is for addresses.
2020-01-17 02:31:12 +00:00
73/jump-if-addr>= $clear-stream:end/disp8
5804 Try to make the comments consistent with the type system we'll eventually have.
2019-12-08 21:56:46 +00:00
# *curr = 0
5666 - start of sugar for structured control flow
2019-09-19 07:33:21 +00:00
c6 0/subop/copy-byte 0/mod/direct 0/rm32/eax . . . . . 0/imm8 # copy byte to *eax
5804 Try to make the comments consistent with the type system we'll eventually have.
2019-12-08 21:56:46 +00:00
# ++curr
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
40/increment-eax
4949
2019-02-03 06:05:11 +00:00
eb/jump $clear-stream:loop/disp8
$clear-stream:end:
# . restore registers
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
59/pop-to-ecx
58/pop-to-eax
5698 Thanks Andrew Owen for reporting this typo.
2019-10-16 02:35:19 +00:00
# . epilogue
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
89/copy 3/mod/direct 4/rm32/esp . . . 5/r32/ebp . . # copy ebp to esp
5d/pop-to-ebp
4949
2019-02-03 06:05:11 +00:00
c3/return
4952
2019-02-06 07:29:15 +00:00
5924
2020-01-27 08:36:44 +00:00
rewind-stream: # f: (addr stream byte)
5698 Thanks Andrew Owen for reporting this typo.
2019-10-16 02:35:19 +00:00
# . prologue
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
55/push-ebp
89/copy 3/mod/direct 5/rm32/ebp . . . 4/r32/esp . . # copy esp to ebp
4952
2019-02-06 07:29:15 +00:00
# . save registers
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
50/push-eax
# eax = f
8b/copy 1/mod/*+disp8 5/rm32/ebp . . 0/r32/eax 8/disp8 . # copy *(ebp+8) to eax
4952
2019-02-06 07:29:15 +00:00
# f->read = 0
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
c7 0/subop/copy 1/mod/*+disp8 0/rm32/eax . . . . 4/disp8 0/imm32 # copy to *(eax+4)
4952
2019-02-06 07:29:15 +00:00
$rewind-stream:end:
# . restore registers
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
58/pop-to-eax
5698 Thanks Andrew Owen for reporting this typo.
2019-10-16 02:35:19 +00:00
# . epilogue
5592 - switch register names to lowercase
2019-08-26 18:55:26 +00:00
89/copy 3/mod/direct 4/rm32/esp . . . 5/r32/ebp . . # copy ebp to esp
5d/pop-to-ebp
4952
2019-02-06 07:29:15 +00:00
c3/return
4961
2019-02-15 00:24:20 +00:00
# . . vim:nowrap:textwidth=0