about summary refs log tree commit diff stats
path: root/036refcount.cc
Commit message (Collapse)AuthorAgeFilesLines
* 3124Kartik K. Agaram2016-07-221-1/+0
| | | | | Reorganize data structure for lambda cells. Create our first real unit test for the compiler in the process.
* 3059Kartik K. Agaram2016-06-171-0/+2
|
* 3047 - fix another hole in refcount managementKartik K. Agaram2016-06-111-26/+56
| | | | | | This wouldn't have failed silently; I have that fig leaf. If someone had tried to copy an exclusive container containing an exclusive container containing an address Mu would have crashed on them.
* 3045 - generalize core refcounting data structureKartik K. Agaram2016-06-111-24/+67
|
* 3044Kartik K. Agaram2016-06-111-0/+1
|
* 3043Kartik K. Agaram2016-06-101-18/+9
|
* 3042Kartik K. Agaram2016-06-101-2/+2
|
* 3018Kartik K. Agaram2016-05-261-5/+2
|
* 2973 - reclaim refcounts for local scopes againKartik K. Agaram2016-05-181-6/+14
| | | | | | More thorough redo of commit 2767 (Mar 12), which was undone in commit 2810 (Mar 24). It's been a long slog. Next step: write a bunch of mu code in the edit/ app in search of bugs.
* 2972 - abandon recursive containersKartik K. Agaram2016-05-171-8/+14
|
* 2968Kartik K. Agaram2016-05-171-14/+37
| | | | | | | | | | | | | | | | | | | | | | | More reorganization in preparation for implementing recursive abandon(). Refcounts are getting incredibly hairy. I need to juggle containers containing other containers, and containers *pointing* to other containers. For a while I considered getting rid of address_element_info entirely and just going by types for every single update_refcount. But that's definitely more work, and it's unclear that things will be cleaner/shorter/simpler. I haven't measured the speedup, but it seems worth optimizing every pointer copy to make sure we aren't manipulating types at runtime. The key insight now is a) to continue to compute information about nested containers at load time, because that's the common case when updating refcounts, but b) to compute information about *pointed* values at run-time, because that's the uncommon case. As a result, we're going to cheat in the interpreter and use type information at runtime just for abandon(), just because the corresponding task when we get to a compiler will be radically different. It will still be tractable, though.
* 2966Kartik K. Agaram2016-05-171-6/+3
|
* 2965 - update refcounts when copying containersKartik K. Agaram2016-05-151-47/+131
| | | | | | This is hopefully quite thorough. I handle nested containers, as well as exclusive containers that might contain addresses only when the tag has a specific value.
* 2963Kartik K. Agaram2016-05-151-38/+42
| | | | | | To support exclusive containers a recursive answer will be easier to reuse than the current iterative one. First step: figure out the precise boundary and interface of the recursive function.
* 2962Kartik K. Agaram2016-05-151-0/+3
|
* 2961Kartik K. Agaram2016-05-151-2/+2
|
* 2960Kartik K. Agaram2016-05-151-7/+7
| | | | First genuine Travis CI failure fixed! May there be many more.
* 2959Kartik K. Agaram2016-05-141-7/+71
|
* 2958Kartik K. Agaram2016-05-141-1/+30
|
* 2957Kartik K. Agaram2016-05-141-8/+57
|
* 2956Kartik K. Agaram2016-05-141-3/+7
|
* 2955 - back to more refcount housekeepingKartik K. Agaram2016-05-121-0/+108
| | | | | Update refcounts of address elements when copying containers. Still lots to do; see todo list at end of 036refcount.cc.
* 2934 - all layers running againKartik K. Agaram2016-05-071-2/+2
| | | | | | Since I switched to a Mac laptop (commit 2725) I've been lax in running test_all_layers because I have to ssh into a server and whatnot. I should just get CI setup somewhere..
* 2933Kartik K. Agaram2016-05-061-1/+1
| | | | Can't believe I didn't run tests after 2932.
* 2931 - be explicit about making copiesKartik K. Agaram2016-05-061-2/+2
|
* 2923Kartik K. Agaram2016-05-041-3/+0
|
* 2922Kartik K. Agaram2016-05-041-28/+34
| | | | | | | | | A better primitive: manage refcounts for an assignment without actually doing the assignment itself. This way we can add refcounts as a new, independent bit of bookkeeping without littering early returns and duplicate code paths. (OCD: commit number.)
* 2898 - start filling in missing refcountsKartik K. Agaram2016-05-031-6/+91
| | | | | | | This commit covers instructions 'put', 'put-index' and 'maybe-convert'. Next up are the harder ones: 'copy' and 'merge'. In these cases there's a non-scalar being copied, and we need to figure out which locations within it need to update their refcount.
* 2897Kartik K. Agaram2016-05-031-0/+74