akkartik/mu
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
5,375 Commits 2 Branches 0 Tags 72 MiB
022075e59e
Commit Graph

78 Commits

Author SHA1 Message Date
Kartik Agaram
022075e59e . 2019-07-11 23:45:03 -07:00
Kartik Agaram
a0c877f8cc one failure remaining in test-compute-offsets 
'curr-segment-name' is now a string, and it's stored in a register
rather than a global.

Paradoxically, this leaks *less* than before. Before, every call to
`get-or-insert-slice` leaked memory. Now we leak one string for every
new segment. Which is trivial.
 2019-07-11 23:41:37 -07:00
Kartik Agaram
7fed7232c4 . 2019-07-11 22:27:01 -07:00
Kartik Agaram
2c45de094b . 
Pseudocode is a little more truthful now about what variables are on the
stack.
 2019-07-11 21:52:56 -07:00
Kartik Agaram
98994d5bcc the problem: curr-segment-name is stale 
It's a slice into the 'line' stream. But we want to preserve the current
segment name across lines.

Let's leak some memory.
 2019-07-11 21:47:00 -07:00
Kartik Agaram
bbfa2acaca . 
Make the trace a little more consistent.
 2019-07-11 21:42:54 -07:00
Kartik Agaram
554bd09968 label offset computation still has a bug 
I changed the test a little to make it obvious.
Basically there's no way we can compute the segment offset correctly
without knowing the segment name in the previous assertion.
 2019-07-11 21:38:25 -07:00
Kartik Agaram
be995e2193 revert compute-offsets to segment-relative offsets 
The pseudocode was a mess here :/ I was saving the segment-offset but
tracing the file-offset.

Segments need file offsets (to tweak their starting address).
Labels need segment offsets (to add to segment starting address).
 2019-07-11 21:31:00 -07:00
Kartik Agaram
f57a458e7b . 2019-07-11 21:24:09 -07:00
nc
3a3e7a90d7 move discard instruction to correct spot 2019-07-11 22:31:56 -04:00
nc
2805308785 updated test so 'x' is relative to file-offset not segment offset 2019-07-11 22:21:32 -04:00
nc
c2725b65b3 made test 2 pass 2019-07-11 22:13:03 -04:00
Kartik Agaram
538f24c296 . 2019-07-10 11:57:08 -07:00
Kartik Agaram
c5f7d9dd57 . 
I think we're calling the wrong variant here.
 2019-07-10 11:38:24 -07:00
Kartik Agaram
304bce5895 start distinguishing table lookups from inserts 2019-07-10 11:32:46 -07:00
Kartik Agaram
84aa2fad2d . 
Fix infinite loop in the 2 remaining failing tests; now it's a segfault.
 2019-07-10 09:39:38 -07:00
Kartik Agaram
7e9cdf3c7a . 2019-07-10 09:35:24 -07:00
Kartik Agaram
48aabc860a mostly done with emit-output 
Some nooks and crannies will need light final debugging with xxd, but
emit-hex-output covers most of the logic.
 2019-07-09 23:41:32 -07:00
Kartik Agaram
20a527702b done with emit-segments 
Only failures now are the first two tests in survey.subx.
 2019-07-09 22:11:44 -07:00
Kartik Agaram
7bc9240c3c . 2019-07-09 17:50:22 -07:00
Kartik Agaram
4a739023d9 . 2019-07-09 17:41:52 -07:00
Kartik Agaram
c7564b7688 . 2019-07-09 16:55:57 -07:00
Kartik Agaram
b14457a4cb . 2019-07-09 16:55:57 -07:00
Kartik Agaram
5821a0f612 clean up after a few calls 2019-07-09 16:55:55 -07:00
Kartik Agaram
f6ea25613f preserve truthiness of non-booleans 
Everywhere we check if something is true, we check it by comparing
against 0, not 1.
 2019-07-09 16:55:37 -07:00
Kartik Agaram
7a752f8d3d this looks like a bug 
But of course the test is still infinite-looping so we can't be sure.
I'm still just reading the code.
 2019-07-09 16:55:13 -07:00
Kartik Agaram
ca32bfbed8 . 2019-07-09 16:54:48 -07:00
Kartik Agaram
d6c6aebef4 . 2019-07-09 16:37:01 -07:00
Kartik Agaram
ba93234603 move 'segment-start' to a global variable as well 
It's ugly to have function locals in global variables, but we'll figure
out later how to deal with it.
 2019-07-09 16:32:06 -07:00
Kartik Agaram
eb4b45db40 . 2019-07-09 16:27:48 -07:00
Kartik Agaram
dc7066c7bf there's a variant of compute-width for slices 
Looks like the original compute-width is now dead. But still seems
useful to have around.
 2019-07-09 16:25:31 -07:00
Kartik Agaram
acc83b644b there's a variant of get-or-insert for slices 
The compute-offsets test now goes into an infinite loop :( But I figure
all these changes are useful anyway, we should just debug the error
separately.
 2019-07-09 16:23:14 -07:00
Kartik Agaram
b8b92251e9 keep labels definitions on a separate line 
SubX in SubX doesn't support mixing labels with other stuff :)
 2019-07-09 16:02:00 -07:00
Kartik Agaram
7761194ec1 switch to global Heap 2019-07-09 16:00:55 -07:00
Kartik Agaram
476049a09d . 
Be more consistent about names of ends of a slice.
(In the opposite direction compared to last night's 925fc490d2ce8b8d411de87bd0af5b3a8a704213.)
 2019-07-08 17:02:09 -07:00
Kartik Agaram
b5da792904 . 
minor style tweaks
 2019-07-08 16:43:35 -07:00
Kartik Agaram
01b45300ed colocate compute-offsets with its globals 2019-07-08 16:38:32 -07:00
Kartik Agaram
925fc490d2 . 
Be more consistent about names of ends of a slice.
 2019-07-08 01:05:23 -07:00
Kartik Agaram
597ce7a943 . 2019-07-08 00:17:42 -07:00
Kartik Agaram
195d62f167 build num-bytes 2019-07-07 23:56:23 -07:00
Kartik Agaram
c38b36e260 move is-label? to subx-common 2019-07-07 22:49:04 -07:00
nc
c0a904382b made first compute-offset test pass 2019-07-07 21:19:50 -04:00
nc
c17643b7a8 remove segfaults in survey.subx 2019-07-07 19:41:26 -04:00
Kartik Agaram
fb8474af88 new failing test: emit-segments 
Now the only piece I plan to not write tests for is emit-headers.
 2019-07-07 13:19:51 -07:00
Kartik Agaram
633c289f30 . 2019-07-07 13:13:38 -07:00
Kartik Agaram
764b5726df move phase 3 out of helpers 2019-07-07 13:13:38 -07:00
nc
b95060262e cleanup in compute-offset and fix bug in compute-width 2019-07-07 15:39:35 -04:00
nc
9ac5834194 more progress in compute-offset 2019-07-04 22:59:15 -04:00
nc
f80a025e52 implement segment section in compute-offsets 2019-07-04 12:21:41 -04:00
Kartik Agaram
c015791e80 . 
Fix a couple of syntax errors.

survey.subx still failing tests.
 2019-07-04 00:02:53 -07:00