about summary refs log tree commit diff stats
Commit message (Collapse)AuthorAgeFilesLines
* 2585 - label sandboxes with a numberKartik K. Agaram2016-01-2212-50/+80
| | | | | | | | | It also seems useful that the number maps to the name of the file the sandbox is saved in. However this mapping is currently a happy accident and not actually tested. I'm starting to switch gears and help make the editor useable with many many sandboxes. This is just the first step of several.
* 2584Kartik K. Agaram2016-01-211-1/+1
|
* 2583 - start maintaining refcountsKartik K. Agaram2016-01-212-3/+68
| | | | | | | | | | | | Also start auto-abandoning addresses when their refcount returns to 0. I'm mixing this auto-abandon support with my earlier/hackier support for automatically abandoning default-space created by 'local-scope'. We need to flesh out the story for automatically reclaiming memory using C++-style destructors. But that's a value-add. Memory corruption is far more important to avoid than memory *leaks*.
* 2582Kartik K. Agaram2016-01-201-2/+2
|
* 2581 - make space for the refcount in address:sharedKartik K. Agaram2016-01-207-53/+97
| | | | | | | | | | | We don't yet actually maintain the refcount. That's next. Hardest part of this was debugging the assume-console scenarios in layer 85. That took some detailed manual diffing of traces (because the output of diff was no good). New tracing added in this commit add 8% to .traces LoC. Commented out trace() calls (used during debugging) make that 45%.
* 2580 - check product type of 'maybe-convert'Kartik K. Agaram2016-01-204-28/+70
| | | | I had to undo some over-zealous changes in 2576.
* 2579Kartik K. Agaram2016-01-202-0/+9
| | | | Separate core mu tests from those loaded from the commandline.
* 2578Kartik K. Agaram2016-01-202-4/+4
|
* 2577Kartik K. Agaram2016-01-201-49/+50
|
* 2576 - distinguish allocated addresses from othersKartik K. Agaram2016-01-1956-3300/+3364
| | | | | | | | | | | | | | | | This is the one major refinement on the C programming model I'm planning to introduce in mu. Instead of Rust's menagerie of pointer types and static checking, I want to introduce just one new type, and use it to perform ref-counting at runtime. So far all we're doing is updating new's interface. The actual ref-counting implementation is next. One implication: I might sometimes need duplicate implementations for a recipe with allocated vs vanilla addresses of the same type. So far it seems I can get away with just always passing in allocated addresses; the situations when you want to pass an unallocated address to a recipe should be few and far between.
* 2575 - better messages on trace count failuresKartik K. Agaram2016-01-193-3/+13
|
* 2574Kartik K. Agaram2016-01-192-15/+12
|
* 2573 - check product type of 'new'Kartik K. Agaram2016-01-194-10/+38
| | | | | | | | | | | I realize that my current doesn't allow nesting a:b:c linear type syntax inside a dilated property. So you can't currently say: (recipe address:number) Need to fix that at some point. Non-trivial since linear syntax is oblivious to dilated syntax. I should probably make the dilated syntax more fundamental and introduce it at an earlier layer.
* 2572Kartik K. Agaram2016-01-191-1/+1
|
* 2571Kartik K. Agaram2016-01-196-4/+4
| | | | | Reorganize layers in preparation for a better way to manage heap allocations without ever risking use-after-free errors.
* 2570 - recipe variables are now fully type-checkedKartik K. Agaram2016-01-181-0/+31
|
* 2569Kartik K. Agaram2016-01-184-6/+17
|
* 2568Kartik K. Agaram2016-01-181-0/+17
|
* 2567Kartik K. Agaram2016-01-181-2/+7
|
* 2566 - typecheck 'call' on recipe variablesKartik K. Agaram2016-01-181-1/+25
|
* 2565 - typecheck 'call' on literal recipesKartik K. Agaram2016-01-181-0/+48
|
* 2564Kartik K. Agaram2016-01-182-8/+6
|
* 2563 - bring back 'call'Kartik K. Agaram2016-01-182-1/+59
| | | | Still need to type-check it, though.
* 2562Kartik K. Agaram2016-01-1713-49/+47
| | | | | | | | | | | | We want to use the type 'recipe' for recipe *variables*, because it seems nicer to say `recipe number -> number` rather than recipe-ordinal, etc. To support this we'll allow recipe names to be mentioned without any type. This might make a couple of places in this commit more brittle. I'm dropping error messages, causing them to not happen in some situations. Maybe I should just bite the bullet and require an explicit :recipe-literal. We'll see.
* 2561Kartik K. Agaram2016-01-176-32/+33
| | | | | Reorganize layers in preparation for a better, more type-safe implementation of first-class and higher-order recipes.
* 2560Kartik K. Agaram2016-01-131-5/+3
|
* 2559 - stop using 'next-ingredient' explicitlyKartik K. Agaram2016-01-125-30/+19
| | | | | I still need it in some situations because I have no way to set a non-zero default for an optional ingredient. Open question..
* 2558Kartik K. Agaram2016-01-122-2/+8
|
* 2557 - type-check most ingredients ahead of timeKartik K. Agaram2016-01-125-48/+86
|
* 2556Kartik K. Agaram2016-01-121-1/+2
|
* 2555Kartik K. Agaram2016-01-114-11/+11
|
* 2554 - give up on static pointer analysisKartik K. Agaram2016-01-111-126/+0
| | | | | Eventually you'll need to create a doubly-linked list, and `isolated` won't save you then.
* 2553 - keep failed specializations from generating spurious errorsKartik K. Agaram2015-12-285-17/+3
| | | | Thanks Caleb Couch.
* 2552Kartik K. Agaram2015-12-281-4/+4
|
* 2551 - add a space after variables in stashKartik K. Agaram2015-12-281-1/+8
|
* 2550Kartik K. Agaram2015-12-281-0/+1
|
* 2549Kartik K. Agaram2015-12-281-1/+3
|
* 2548 - teach 'print' to print integersKartik K. Agaram2015-12-2816-98/+164
| | | | | | | | | | Still can't print non-integer numbers, so this is a bit hacky. The big consequence is that you can't print literal characters anymore because of our rules about how we pick which variant to statically dispatch to. You have to save to a character variable first. Maybe I can add an annotation to literals..
* 2547Kartik K. Agaram2015-12-242-1/+7
|
* 2546 - another phase-ordering constraintKartik K. Agaram2015-12-192-2/+20
| | | | Thanks Caleb Couch.
* 2545Kartik K. Agaram2015-12-1534-260/+1182
| | | | update html
* support immutability checks in sandbox/ appKartik K. Agaram2015-12-158-18/+47
|
* three bugs fixedKartik K. Agaram2015-12-158-31/+253
| | | | | | | | | | | | | | | | | - notes bug in edit/ triggers in immutable but not master branch bug triggered by changes to layer 059: we're finding an unspecialized call to 'length' in 'append_6' hard to debug because trace isn't complete just bring out the big hammer: use a new log file length_2 from recipes.mu is not being deleted (bug #1) so reload doesn't switch length to length_2 when variant_already_exists (bug #2) so we end up saving in Recipe for a primitive ordinal so no valid specialization is found for 'length' (bug #3) why doesn't it trigger in a non-interactive scenario? argh, wasn't checking for an empty line at end. ok, confidence restored.
* bugfix after commit 2612Kartik K. Agaram2015-12-151-0/+21
|
* start working on checking address lifetimeKartik K. Agaram2015-12-151-0/+126
| | | | First some baseline tests that should never trigger warnings.
* start using the new checkKartik K. Agaram2015-12-152-8/+6
| | | | | | | | | | I wasn't seeing errors because I wasn't using /contained-in in products yet. But it seems to work fine even after. One reason this isn't an invasive change is that it's opt-in. Most of the time it isn't triggered. You have to add the /contained-in check to trigger a check. But that's just like regular const checks in other languages: if you don't specify immutability you get no checks.
* improve error messageKartik K. Agaram2015-12-151-8/+16
|
* infect immutability across recipesKartik K. Agaram2015-12-151-2/+47
| | | | | | | | | | | | | | | | | If a product is contained-in some ingredient, then the caller will check the product for mutations as well. For example, consider this linked-list operation: b:address:list:number <- next a:address:list:number If 'a' is immutable in the surrounding recipe, you probably want 'b' to be immutable as well. You can achieve this by giving 'next' the following header (ignoring shape-shifting): recipe next a:address:list:number -> b:address:list:number/contained-in:a This is the theory, anyway. Rather to my surprise, this doesn't trigger any issues with existing code. That's probably too good to be true.
* next baby step: infect immutability across copiesKartik K. Agaram2015-12-151-15/+54
|
* fix a crash on unknown typeKartik K. Agaram2015-12-151-0/+11
|