aboutsummaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
...
* Encode and decode tag types in dwarf properly. Add list module to std. Shift ↵Graydon Hoare2010-10-1510-114/+243
| | | | rustc to use std.util.option. Fix various dependent bugs. Closes #73.
* Finish the fold drivers in fold.rs.Graydon Hoare2010-10-142-12/+110
|
* rustc: Parse field and vector indexing expressionsPatrick Walton2010-10-141-34/+53
|
* Enable resolve pass in rustc. Doesn't do much yet aside from an incomplete ↵Graydon Hoare2010-10-141-0/+2
| | | | identity fold, but no longer crashes.
* Stop using bound refs (don't work; leak) and work around auto-deref ↵Graydon Hoare2010-10-141-23/+33
| | | | vec-append bug.
* Fix mem += mem bug in vec-append.Graydon Hoare2010-10-141-2/+2
|
* Fix typos in Type.friendly_stringify.Graydon Hoare2010-10-141-3/+3
|
* rustc: Start work on lvalsPatrick Walton2010-10-143-59/+137
|
* Typecheck tags in "alt" patternsPatrick Walton2010-10-143-0/+26
|
* Fix crasher in rustc.Graydon Hoare2010-10-144-5/+14
|
* Move the friendly-names table to semant, reuse it in the name mangler.Graydon Hoare2010-10-133-16/+19
|
* Fetch typarams from the outermost item frame, when inside an iter-block. One ↵Graydon Hoare2010-10-134-30/+75
| | | | less crash in rustc.
* Use "friendly" types throughout the typecheckerPatrick Walton2010-10-131-38/+40
|
* Don't run tidy inside etc/. Should put out the burning tinderbox.Patrick Walton2010-10-131-1/+1
|
* Use the user-defined type aliases when reporting type errorsPatrick Walton2010-10-131-21/+71
|
* rustc: Lex identifiers that have numbers in them tooPatrick Walton2010-10-131-2/+6
|
* rustc: Parse vector typesPatrick Walton2010-10-132-0/+8
|
* First stab at a Vim syntax highlighter for RustPatrick Walton2010-10-131-0/+35
|
* Fix horribly embarassing signedness bug in backend, plus related regressions.Graydon Hoare2010-10-124-6/+22
|
* rustc: Add mutability to tuple literalsPatrick Walton2010-10-123-10/+30
|
* rustc: Parse box typesPatrick Walton2010-10-121-0/+2
|
* rustc: Parse tuple typesPatrick Walton2010-10-121-0/+21
|
* Git index wins again.Graydon Hoare2010-10-129-65/+105
|
* Changes to make rustboot compile on OCaml 3.12Graydon Hoare2010-10-129-46/+46
|
* rustc: Add the tuple type to the ASTPatrick Walton2010-10-111-0/+1
|
* Parse autosPatrick Walton2010-10-111-9/+22
|
* rustc: Parse letsPatrick Walton2010-10-113-6/+33
|
* rustc: Say "expected expression" instead of "expected literal" when we ↵Patrick Walton2010-10-111-8/+14
| | | | expect an expression
* Use new and delete instead of alloca(). Should put out the burning tinderbox.Patrick Walton2010-10-111-2/+3
|
* Try to print backtraces on failurePatrick Walton2010-10-114-17/+46
|
* Bind labels explicitly in fe/cexp.ml. Should fix issue #169.Patrick Walton2010-10-111-2/+2
|
* Remove obsolete comment, issue was fixed with last checkin.Graydon Hoare2010-10-101-3/+0
|
* Teach bind to actually bind typarams, as it claims to.Graydon Hoare2010-10-105-61/+112
|
* Subject copying code to inline heuristic, cut 30kb from rustc.Graydon Hoare2010-10-091-10/+31
|
* Actually do what I meant to in the inline threshold. Don't drink and code. ↵Graydon Hoare2010-10-091-4/+4
| | | | Cut rustc by 50kb.
* Elide mark and sever glue in -minimal mode.Graydon Hoare2010-10-091-2/+9
|
* More type-fold caches, shave another second off compile time.Graydon Hoare2010-10-093-5/+16
|
* Implement a -minimal mode prologue, cut 100kb from rustc.Graydon Hoare2010-10-082-32/+78
|
* Quad counters are revealing. Move drops out-of-line, cut 200kb from rustc.Graydon Hoare2010-10-081-103/+168
|
* Add quad-counting system.Graydon Hoare2010-10-083-9/+81
|
* Fix logging bug.Graydon Hoare2010-10-081-1/+1
|
* Add a tag cache for the cases we can't avoid hitting fold; cut rustc compile ↵Graydon Hoare2010-10-082-8/+19
| | | | time by 50%.
* More fold caches, cut rustc compile time by 60% again.Graydon Hoare2010-10-082-7/+27
|
* Cache referent-type calculations in rustboot; cut rustc compile time by 60%.Graydon Hoare2010-10-081-1/+7
|
* Add -lpath mechanism for logging only a subset of a pass (by module-path prefix)Graydon Hoare2010-10-0815-103/+142
|
* rustc: Make functions output a type, not a slotPatrick Walton2010-10-083-9/+8
|
* Add a little code to resolve, to check that the fold interface works.Graydon Hoare2010-10-072-5/+27
|
* Make minimal-mode slightly less aggressive.Graydon Hoare2010-10-071-7/+3
|
* Revise folds to be much-less-polymorphic after some discussion; flesh out ↵Graydon Hoare2010-10-071-172/+309
| | | | stmt and expr fold drivers.
* Simplify type-mismatch messages.Graydon Hoare2010-10-071-5/+158
|