1807 Commits

Author SHA1 Message Date
James Miller
761fc16c60 Fix-up PP Code to reflect new lifetime param syntax 2013-06-22 12:38:40 +12:00
James Miller
048ed1486f Move count-llvm-insn code into task-local storage 2013-06-22 12:38:40 +12:00
James Miller
0b0c756c9c Fix warnings in trans 2013-06-22 12:38:40 +12:00
James Miller
81cf72c264 Finish up Type refactoring 2013-06-22 12:35:35 +12:00
James Miller
57a75374d6 Initial Type Refactoring done 2013-06-22 12:32:11 +12:00
James Miller
befbd3a680 Add the rest of the atomic operations.
This makes the handling of atomic operations more generic, which
does impose a specific naming convention for the intrinsics, but
that seems ok with me, rather than having an individual case for
each name.

It also adds the intrinsics to the the intrinsics file.
2013-06-22 12:26:33 +12:00
James Miller
fd83b92b59 More Type refactorings 2013-06-22 12:26:33 +12:00
James Miller
1968622798 Start refacting LLVM Type handling 2013-06-22 12:24:21 +12:00
James Miller
66d8e8b481 Make log_fn_time a method 2013-06-22 12:24:21 +12:00
James Miller
b4b2cbb299 Change calls for TypeName stuff to methods 2013-06-22 12:24:20 +12:00
James Miller
3dbdb3a364 Methodize TypeNames 2013-06-22 12:24:20 +12:00
bors
f886520d24 auto merge of #7219 : sstewartgallus/rust/cleanup, r=graydon 2013-06-21 16:01:59 -07:00
Corey Richardson
116897fa6c Remove ast::pure_fn and all concept of pure from the compiler 2013-06-21 18:10:56 -04:00
bors
544f6159f7 auto merge of #7259 : dotdash/rust/ir_improvement, r=graydon
The changes in these commits improve the IR codegen by removing unnecessary copies for certain function call arguments and stopping to allocate return values for functions returning nil. They reduce compile times by about 10% in total.
2013-06-21 07:52:55 -07:00
Björn Steinbrink
4fb2c09541 Avoid pointless allocas for "nil" return values
By using "void" instead of "{}" as the LLVM type for nil, we can avoid
the alloca/store/load sequence for the return value, resulting in less
and simpler IR code.

This reduces compile times by about 10%.
2013-06-21 13:36:25 +02:00
bors
b0e3ffd380 auto merge of #7263 : thestinger/rust/iterator, r=graydon 2013-06-21 01:49:50 -07:00
Daniel Micay
06bec77faf replace vec::find with the IteratorUtil method 2013-06-21 03:24:03 -04:00
Daniel Micay
883c966d5c vec: replace position with iter().position_ 2013-06-21 03:23:59 -04:00
Daniel Micay
49c74524e2 vec: rm old_iter implementations, except BaseIter
The removed test for issue #2611 is well covered by the `std::iterator`
module itself.

This adds the `count` method to `IteratorUtil` to replace `EqIter`.
2013-06-21 03:20:22 -04:00
Daniel Micay
cbad1da3db vec: remove eachi
replaced by the `enumerate` method from std::iterator
2013-06-21 03:20:22 -04:00
bors
ba05af7b1c auto merge of #7203 : msullivan/rust/default-methods, r=graydon
This fixes the large number of problems that prevented cross crate
methods from ever working. It also fixes a couple lingering bugs with
polymorphic default methods and cleans up some of the code paths.

Closes #4102. Closes #4103.

r? nikomatsakis
2013-06-20 22:28:52 -07:00
Luqman Aden
31b4b53797 librustc: Don't allow enum struct variants to shadow structs. 2013-06-20 19:14:57 -04:00
Michael Sullivan
1a8969f64b Get cross crate default methods working.
This fixes the large number of problems that prevented cross crate
methods from ever working. It also fixes a couple lingering bugs with
polymorphic default methods and cleans up some of the code paths.

Closes #4102. Closes #4103.
2013-06-20 15:42:16 -07:00
Björn Steinbrink
dc262d9aa7 Avoid unnecessary scratch datums for by-copy function arguments
Currently, by-copy function arguments are always stored into a scratch
datum, which serves two purposes.  First, it is required to be able to
have a temporary cleanup, in case that the call fails before the callee
actually takes ownership of the value. Second, if the argument is to be
passed by reference, the copy is required, so that the function doesn't
get a reference to the original value.

But in case that the datum does not need a drop glue call and it is
passed by value, there's no need to perform the extra copy.
2013-06-20 23:53:26 +02:00
James Miller
229a88217f librust: cleanup warnings (except 1) 2013-06-21 02:43:03 +12:00
James Miller
3bc4d1a120 Remove all #[cfg(stage0)]-protected code
New snapshot means this can all go. Also removes places that have
comments that say they are workarounds for stage0 errors.
2013-06-21 02:43:02 +12:00
Michael Woerister
6eb3c0f30d debuginfo.rs: Removed trailing whitespace. 2013-06-20 11:50:09 +02:00
Michael Woerister
60474f395c debuginfo.rs: Explained why a function is called although one could think it should not be. 2013-06-20 11:50:09 +02:00
Michael Woerister
38743f51ad debuginfo.rs: Removed some static constants not used anymore. 2013-06-20 11:50:09 +02:00
Michael Woerister
290d35312a Reordered debuginfo.rs and commented public fns 2013-06-20 11:50:09 +02:00
Michael Woerister
f1bd3e7ca2 Fixed compiler warnings about unnecessary string allocations in debuginfo.rs 2013-06-20 11:50:09 +02:00
Michael Woerister
5d5311dc74 Added overview of the debuginfo module at the beginning of debuginfo.rs. 2013-06-20 11:50:09 +02:00
Brian Anderson
b7a6919899 rustc: Dispose of LLVM passes in test cases 2013-06-19 15:18:25 -07:00
Ben Blum
3ba8404c88 Oops, hide 'unless once fn' error message hint behind -Z once-fns too. 2013-06-19 16:04:18 -04:00
Ben Blum
643be38cfe Hide the once-move-out privilege for stack fns behind '-Z once-fns' 2013-06-19 15:53:05 -04:00
Ben Blum
1496216db6 Permit moving out of captured upvars in once fns. Close #2549. 2013-06-19 14:25:51 -04:00
Ben Blum
2c7903d599 Consume once fns when calling them (#2549). 2013-06-19 14:25:51 -04:00
Steven Stewart-Gallus
b8efa2ebe0 Cleanup src/librustc/back 2013-06-18 18:50:37 -07:00
Graydon Hoare
d904c72af8 replace #[inline(always)] with #[inline]. r=burningtree. 2013-06-18 14:48:48 -07:00
bors
4bf074cc66 auto merge of #7134 : vadimcn/rust/DIBuilder, r=jdm
This commit fixes rustc's debug info generation and turns debug-info tests back on.

The old generator used to write out LLVM metadata directly, however it seems that debug metadata format is not stable and keeps changing from release to release.  So I wrapped LLVM's official debug info API - the DIBuilder class, and now rustc will use that.

One bit of old functionality that still doesn't work, is debug info for function arguments.  Someone more familiar with the compiler guts will need to look into that.

Also, unfortunately, debug info is still won't work on Windows,- due to a LLVM bug (http://llvm.org/bugs/show_bug.cgi?id=16249).

Resolves issues #5836, #5848, #6814
2013-06-17 14:01:35 -07:00
Brian Anderson
fd09e40f97 Revert "stop using an absolute rpath"
This reverts commit 708395d65d689080cc96dd091adab3fc4c5c3f27.
2013-06-17 12:37:06 -07:00
Graydon Hoare
6914ff9d01 rustc: map node ids through a table that ensures bitset indexes in dataflow are dense 2013-06-17 10:37:42 -07:00
Vadim Chugunov
adff46250e Fixed rebase fallout . 2013-06-17 08:42:05 -07:00
Vadim Chugunov
1e682e29eb Refactoring and tidy warnings cleanup. 2013-06-17 08:41:26 -07:00
Vadim Chugunov
2b45591daf Made unimplemented debuginfo a note rather than a compiler error. 2013-06-17 08:41:26 -07:00
Vadim Chugunov
65dd6218af Fixed remaining issues to pass debug-test/* tests.
Made debugger scripts source line insensitive.
2013-06-17 08:41:25 -07:00
Vadim Chugunov
62e86e044d Fixed compile warnings.
Fixed whitespace "errors".
2013-06-17 08:41:25 -07:00
Vadim Chugunov
00bb15bf72 Removed extraneous string allocations.
Misc refactoring.
2013-06-17 08:41:24 -07:00
Vadim Chugunov
6db3302b35 Debug loc for local var declarations 2013-06-17 08:41:24 -07:00
Vadim Chugunov
1079e41143 Move "return" basic block after all other function blocks. 2013-06-17 08:41:23 -07:00