| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
|
|
| |
Let's start adding ':end' labels in all functions, just because it helps
us visualize where function calls end in traces, thanks to the '--map'
commandline argument.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some automated commenting cleanup. Still needs more careful manual scanning.
sed -i 's/^# 1-3/# . 1-3/' *.subx */*.subx
sed -i 's/^# op/# . op/' *.subx */*.subx
sed -i 's/# vim/# . . vim/' *.subx */*.subx
sed -i 's/^ # push args/ # . . push args/' *.subx */*.subx
sed -i 's/^ # discard args/ # . . discard args/' *.subx */*.subx
sed -i 's/^ # call/ # . . call/' *.subx */*.subx
sed -i 's/^ # prolog/ # . prolog/' *.subx */*.subx
sed -i 's/^ # epilog/ # . epilog/' *.subx */*.subx
sed -i 's/^ # save registers/ # . save registers/' *.subx */*.subx
sed -i 's/^ # restore registers/ # . restore registers/' *.subx */*.subx
sed -i 's/ operand / register /' *.subx */*.subx
|
|
|
|
| |
Reindent all SubX code to make some room for the new comment style.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
Long-standing and long-copied typo has been messing with our exit status
on test failures.
|
| |
|
| |
|
|
|
|
| |
Extract a helper that we'll need for 'read'.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
Extract a helper for appending strings to raw buffers.
I'd been resisting this idea, but it actually turns out to be a pretty
clean abstraction in the end.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
All tests now once again run the same natively and on VM.
|
|
|
|
|
|
|
|
|
|
| |
Turns out the tests for 'trace' have been broken in native mode since the
original commit (4674). Dangers of running my tests on Darwin, where I
can't run them natively.
The test failures didn't get flagged on CI because I'd forgotten to update
the exit code of the factorial app in commit 4664. At least that's fixed
in this commit.
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
Maps have definitely helped with debugging. Even having just the top of
the call stack is very helpful.
We're soon gonna need setup/teardown for tests. I'm not sure how compiling
run-tests will work then.
|
|
|
|
|
|
| |
On second thoughts, let's not use Mu's "null is real hardware" convention
for traces. There's no real difference between a real and fake trace stream,
so we'll just always explicitly pass in *Trace-stream in production code.
|
| |
|
|
subx: append to trace
|