about summary refs log tree commit diff stats
Commit message (Collapse)AuthorAgeFilesLines
* 377Kartik K. Agaram2014-11-281-45/+37
|
* 376 - similarly simplify record helpersKartik K. Agaram2014-11-282-8/+25
|
* 375 - reorg testsKartik K. Agaram2014-11-281-239/+189
|
* 374 - simplify array helpersKartik K. Agaram2014-11-282-21/+35
|
* 373Kartik K. Agaram2014-11-281-6/+5
|
* 372 - size check for arraysKartik K. Agaram2014-11-282-2/+10
|
* 371 - bugfix: don't read destination when writing arrayKartik K. Agaram2014-11-282-4/+21
| | | | | Finally we've gotten to one of the top-level bugs uncovered by 'interpolate'.
* 370 - unit tests for 'setm'Kartik K. Agaram2014-11-282-4/+50
|
* 369Kartik K. Agaram2014-11-281-4/+4
|
* 368 - replace 'sz' with our new 'sizeof'Kartik K. Agaram2014-11-282-21/+20
|
* 367 - start of unit tests for 'm'Kartik K. Agaram2014-11-281-0/+15
|
* 366 - reorg run's helpersKartik K. Agaram2014-11-281-142/+155
| | | | | First step to using our new 'deref' and 'absolutize' helpers more coherently.
* 365 - new primitives let us generalize 'addr'Kartik K. Agaram2014-11-282-4/+8
|
* 364Kartik K. Agaram2014-11-281-16/+2
|
* 363Kartik K. Agaram2014-11-282-1/+29
|
* 362Kartik K. Agaram2014-11-282-0/+4
| | | | Gaining confidence..
* 361 - *now* 'sizeof' on array pointersKartik K. Agaram2014-11-282-0/+7
|
* 360 - back up, let's create a new 'deref' helperKartik K. Agaram2014-11-282-0/+21
|
* 359 - 'sizeof' supports 'deref' pointersKartik K. Agaram2014-11-282-1/+8
| | | | Just non-arrays for now.
* 358 - start of 'sizeof' support for arraysKartik K. Agaram2014-11-282-0/+16
|
* 357Kartik K. Agaram2014-11-281-0/+10
|
* 356 - high time 'sizeof' supported operandsKartik K. Agaram2014-11-282-12/+23
|
* 355Kartik K. Agaram2014-11-281-0/+11
|
* 354 - stepping back, test-driving 'addr' helper againKartik K. Agaram2014-11-282-1/+29
|
* 353 - found lots of holes in array supportKartik K. Agaram2014-11-271-34/+34
| | | | | I've uncovered a few while enhancing 'interpolate', but I'll wait to fix them all before I commit the enhancements.
* 352 - 'interpolate' as a poor man's printfKartik K. Agaram2014-11-272-2/+174
|
* 351 - break/loop from nested blocksKartik K. Agaram2014-11-272-10/+58
|
* 350Kartik K. Agaram2014-11-271-9/+9
|
* 349 - jump to labelKartik K. Agaram2014-11-272-3/+49
|
* 348Kartik K. Agaram2014-11-272-7/+7
|
* 347 - allow 'arg' channel to be rewoundKartik K. Agaram2014-11-272-9/+26
|
* 346Kartik K. Agaram2014-11-271-7/+7
|
* 345Kartik K. Agaram2014-11-272-59/+2
|
* 344 - about to give up on rewrite rulesKartik K. Agaram2014-11-273-8/+30
| | | | | | | | | | | | | | | | | I wanted to come up with some way to rewrite 'def-clause foo' to 'after foo/more-clauses', something like: rewrite def-clause [ (fn-name string-address) <- arg (label-name string-address) <- strcat (fn-name string-address deref) ("/more-clauses" literal) (body expr) <- arg reply `(after ,label-name ,body) ] But the quasiquote is still a nested expression that doesn't fit our model well. Still an open question how to do template interpolation in mu.
* 343 - experiment: rewrite rulesKartik K. Agaram2014-11-272-2/+38
| | | | We'll worry later about implementing them in mu.
* 342 - strcatKartik K. Agaram2014-11-272-2/+79
|
* 341 - standardize traces againKartik K. Agaram2014-11-273-1/+51
|
* 340Kartik K. Agaram2014-11-262-2/+3
|
* 339 - string literals for convenienceKartik K. Agaram2014-11-262-7/+37
| | | | | | I'm cheating a little when it comes to strings, making use of Racket primitives. But in principle it seems like a local transform to implement even in machine code.
* 338 - starting to work on string primitivesKartik K. Agaram2014-11-262-1/+13
|
* 337 - rename 'continue' instructions to 'loop'Kartik K. Agaram2014-11-262-28/+28
|
* 336Kartik K. Agaram2014-11-261-1/+0
|
* 335 - tangle at labels scoped to functionsKartik K. Agaram2014-11-262-5/+37
|
* 334Kartik K. Agaram2014-11-261-0/+4
|
* 333Kartik K. Agaram2014-11-261-1/+2
|
* 332 - example function with multiple clausesKartik K. Agaram2014-11-262-0/+32
| | | | | | | | | | | I'm still figuring out the idioms, and it doesn't seem like there's an easy way to avoid redundant allocations and arg reading. At least for generic functions we want to optimize for each call-site, we can create subsidiary labels to jump to. Also, this is the first genuine use for random access to a function's arguments.
* 331Kartik K. Agaram2014-11-261-3/+1
|
* 330Kartik K. Agaram2014-11-261-12/+52
|
* 329Kartik K. Agaram2014-11-261-1/+1
|
* 328Kartik K. Agaram2014-11-261-35/+34
|
generated by cgit-pink 1.4.1-2-gfad0 (git 2.36.2.497.gbbea4dcf42) at 2024-10-08 04:18:16 +0000