Rewrite NASM programs to FASM #116

IgorA · 2025-03-06T21:25:06+01:00

IgorA commented

2025-03-06 21:25:06 +01:00

Переписаны программы из компилятора NASM под компилятор FASM

IgorA added 8 commits 2025-03-06 21:25:09 +01:00

mxlgv force-pushed nasm-apps-revrite-to-fasm from 002e1677ee to fabe56be85

2025-03-06 22:13:25 +01:00

Compare

mxlgv changed title from ~~nasm-apps-revrite-to-fasm~~ to Nasm apps rewrite to fasm

2025-03-06 22:14:22 +01:00

mxlgv force-pushed nasm-apps-revrite-to-fasm from fabe56be85 to 4bfe30fc86

2025-03-06 22:21:49 +01:00

Compare

mxlgv force-pushed nasm-apps-revrite-to-fasm from 4bfe30fc86 to a5aaa4f6a5

2025-03-06 22:30:12 +01:00

Compare

mxlgv requested review from mxlgv 2025-03-06 22:30:40 +01:00

Sweetbread changed title from ~~Nasm apps rewrite to fasm~~ to Rewrite NASM programs to FASM

2025-03-06 22:39:02 +01:00

mxlgv force-pushed nasm-apps-revrite-to-fasm from a5aaa4f6a5 to 72543c1f5d

2025-03-06 23:26:57 +01:00

Compare

mxlgv force-pushed nasm-apps-revrite-to-fasm from 72543c1f5d to e55099db13

2025-03-07 00:43:53 +01:00

Compare

mxlgv approved these changes 2025-03-07 01:10:21 +01:00

mxlgv merged commit 98cccc438c into main

2025-03-07 01:10:49 +01:00

mxlgv deleted branch nasm-apps-revrite-to-fasm

2025-03-07 01:10:53 +01:00

dunkaist commented

2025-03-07 01:28:25 +01:00

@IgorA, I have checked each pair of a nasm and a fasm binaries, they all are different. Even by size, but also by data. I have no idea how to verify the new programs don't bring new bugs. My advice for similar future works, if any, is to rewrite byte-to-byte first and only then add fixes etc.

1947 aclock.fasm
1695 aclock.nasm
6932 c4.fasm
6693 c4.nasm
341  fillscr.fasm
347  fillscr.nasm
1383 genfiles.fasm
1384 genfiles.nasm
950  lod.fasm
938  lod.nasm
527  mstate.fasm
532  mstate.nasm
1035 timer.fasm
1077 timer.nasm
3655 tinfo.fasm
3632 tinfo.nasm

@IgorA, I have checked each pair of a nasm and a fasm binaries, they all are different. Even by size, but also by data. I have no idea how to verify the new programs don't bring new bugs. My advice for similar future works, if any, is to rewrite byte-to-byte first and only then add fixes etc. ``` 1947 aclock.fasm 1695 aclock.nasm 6932 c4.fasm 6693 c4.nasm 341 fillscr.fasm 347 fillscr.nasm 1383 genfiles.fasm 1384 genfiles.nasm 950 lod.fasm 938 lod.nasm 527 mstate.fasm 532 mstate.nasm 1035 timer.fasm 1077 timer.nasm 3655 tinfo.fasm 3632 tinfo.nasm ```

Leency commented

2025-03-10 15:34:14 +01:00

Test by hand is easier than write byte-to-byte, so the request is not valid for me. Also this blocks some tweaks and improvements.

dunkaist commented

2025-03-15 04:02:14 +01:00

Test by hand is easier than write byte-to-byte, so the request is not valid for me. Also this blocks some tweaks and improvements.

That wasn't a request, that was a regret. Testing by hand is definitely easier because it doesn't provide any guarantee.

What tweaks and improvements did you mean?

> Test by hand is easier than write byte-to-byte, so the request is not valid for me. Also this blocks some tweaks and improvements. That wasn't a request, that was a regret. Testing by hand is definitely easier because it doesn't provide any guarantee. What tweaks and improvements did you mean?

IgorA commented

2025-03-15 13:56:46 +01:00

This is because some minor changes have been added. For example, some labels have been aligned to 4 bytes. Or the mcall macro has been added, which may write values to registers in the wrong order, etc.

Sign in to join this conversation.

4 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: KolibriOS/kolibrios#116