]> git.lizzy.rs Git - rust.git/log
rust.git
13 years agoAdd Kelly to AUTHORS.
Graydon Hoare [Thu, 5 May 2011 23:40:11 +0000 (16:40 -0700)]
Add Kelly to AUTHORS.

13 years agoAdd quick sort function to the std lib.
Kelly Wilson [Thu, 5 May 2011 20:08:52 +0000 (14:08 -0600)]
Add quick sort function to the std lib.

13 years agoUse symbolic register names so that we get the correct encoding on OS X.
Rafael Ávila de Espíndola [Thu, 5 May 2011 22:49:45 +0000 (18:49 -0400)]
Use symbolic register names so that we get the correct encoding on OS X.

13 years agoChange the setup so that rust_activate_glue returns to rust_exit_task_glue
Rafael Ávila de Espíndola [Thu, 5 May 2011 18:01:16 +0000 (14:01 -0400)]
Change the setup so that rust_activate_glue returns to rust_exit_task_glue
and rust_exit_task_glue calls the rust main.

This is simpler since we only need to setup one frame. It also matches
what ld.so does, so gdb is happy and stops a backtrace at rust_exit_task_glue
instead of continuing past whatever function happened to be before
rust_exit_task_glue is the object file.

This is the rt part and should be merged after the rust0 part.

13 years agoRegister new stage0 snapshots.
Graydon Hoare [Thu, 5 May 2011 22:44:01 +0000 (15:44 -0700)]
Register new stage0 snapshots.

13 years agoChange the setup so that rust_activate_glue returns to rust_exit_task_glue
Rafael Ávila de Espíndola [Thu, 5 May 2011 18:00:50 +0000 (14:00 -0400)]
Change the setup so that rust_activate_glue returns to rust_exit_task_glue
and rust_exit_task_glue calls the rust main.

This is simpler since we only need to setup one frame. It also matches
what ld.so does, so gdb is happy and stops a backtrace at rust_exit_task_glue
instead of continuing past whatever function happened to be before
rust_exit_task_glue is the object file.

This is only the rustc changes and should be merged first.

13 years agorustc: Add Rust intrinsic support to the crate metadata reader and writer
Patrick Walton [Thu, 5 May 2011 21:34:22 +0000 (14:34 -0700)]
rustc: Add Rust intrinsic support to the crate metadata reader and writer

13 years agoAttempt to fix fs.dirname harder.
Graydon Hoare [Thu, 5 May 2011 21:31:10 +0000 (14:31 -0700)]
Attempt to fix fs.dirname harder.

13 years agorustc: Fix bug in equal_abi that was preventing Rust intrinsics from working
Patrick Walton [Thu, 5 May 2011 21:21:21 +0000 (14:21 -0700)]
rustc: Fix bug in equal_abi that was preventing Rust intrinsics from working

13 years agoBecause good hackers at least check to see if the code compiles.
Graydon Hoare [Thu, 5 May 2011 20:53:57 +0000 (13:53 -0700)]
Because good hackers at least check to see if the code compiles.

13 years agoTidy up stage 2 and 3 mk files.
Graydon Hoare [Thu, 5 May 2011 20:43:08 +0000 (13:43 -0700)]
Tidy up stage 2 and 3 mk files.

13 years agoFix assumption that only os_fs.path_sep can separate paths, sigh.
Graydon Hoare [Thu, 5 May 2011 20:42:52 +0000 (13:42 -0700)]
Fix assumption that only os_fs.path_sep can separate paths, sigh.

13 years agoAdd --help, --version and -v flags (not yet supporting GIT_REV env var, waiting on...
Graydon Hoare [Thu, 5 May 2011 19:03:23 +0000 (12:03 -0700)]
Add --help, --version and -v flags (not yet supporting GIT_REV env var, waiting on snapshot).

13 years agorustc: Link with intrinsics.bc; change intrinsics linkage to linkonce_odr
Patrick Walton [Thu, 5 May 2011 19:45:34 +0000 (12:45 -0700)]
rustc: Link with intrinsics.bc; change intrinsics linkage to linkonce_odr

13 years agoPut out burning tinderbox.
Graydon Hoare [Thu, 5 May 2011 18:54:32 +0000 (11:54 -0700)]
Put out burning tinderbox.

13 years agorustc: Add a binding to LLVM's bitcode parser
Patrick Walton [Thu, 5 May 2011 18:34:15 +0000 (11:34 -0700)]
rustc: Add a binding to LLVM's bitcode parser

13 years agoEnforce in typechecker that preds return a bool
Tim Chevalier [Wed, 4 May 2011 22:24:07 +0000 (15:24 -0700)]
Enforce in typechecker that preds return a bool

as well as a test case

13 years agoUpdate docs to reflect preds
Tim Chevalier [Wed, 4 May 2011 21:28:52 +0000 (14:28 -0700)]
Update docs to reflect preds

13 years agoTest cases for pred / check stuff
Tim Chevalier [Wed, 4 May 2011 18:32:06 +0000 (11:32 -0700)]
Test cases for pred / check stuff

13 years agoBring back "pred" syntax for writing predicates for check
Tim Chevalier [Wed, 4 May 2011 18:28:13 +0000 (11:28 -0700)]
Bring back "pred" syntax for writing predicates for check

This commit reinstates the requirement that the predicate in a
"check" must be a manifest call to a special kind of function
declared with the new "pred" keyword instead of "fn". Preds must
have a boolean return type and can only call other preds; they
can't have any effects (as enforced by the typechecker).
The arguments to a predicate in a check expression must be
slot variables or literals.

13 years agoUpdate docs to reflect assert vs. check
Tim Chevalier [Wed, 4 May 2011 18:01:47 +0000 (11:01 -0700)]
Update docs to reflect assert vs. check

Also added the --batch flag to texi2pdf, as it doesn't really ever
seem useful to drop to the TeX prompt during a build.

13 years agoadd dl/ and .pyc files to .gitignore
Tim Chevalier [Wed, 4 May 2011 17:55:32 +0000 (10:55 -0700)]
add dl/ and .pyc files to .gitignore

13 years agoCheck well-formedness of constraints
Tim Chevalier [Mon, 2 May 2011 21:28:35 +0000 (14:28 -0700)]
Check well-formedness of constraints

Check that the operand in a constraint is an explicit name,
and that the operands are all local variables or literals. Still need
to check that the name refers to a pure function.

13 years agoChange checks to asserts in test/bench files
Tim Chevalier [Mon, 2 May 2011 20:50:34 +0000 (13:50 -0700)]
Change checks to asserts in test/bench files

13 years agoCheck that the operand in a check is a call
Tim Chevalier [Mon, 2 May 2011 20:47:41 +0000 (13:47 -0700)]
Check that the operand in a check is a call

In addition, fix bug in fold that was turning asserts into checks.

More typechecking still needs to be done.

13 years agoUn-XFAILed not-a-pred in stage0 (it fails correctly)
Tim Chevalier [Mon, 2 May 2011 19:16:29 +0000 (12:16 -0700)]
Un-XFAILed not-a-pred in stage0 (it fails correctly)

13 years agoRemove 'deprecated mutable...' from our code
Marijn Haverbeke [Thu, 5 May 2011 18:17:44 +0000 (20:17 +0200)]
Remove 'deprecated mutable...' from our code

This should make compilation a bit less noisy.

13 years agoReturn a fresh, unreachable context after ret, break, and cont
Marijn Haverbeke [Thu, 5 May 2011 13:30:21 +0000 (15:30 +0200)]
Return a fresh, unreachable context after ret, break, and cont

This ensures we don't get compile errors on unreachable code (see
test/run-pass/artificial-block.rs for an example of sane code that
wasn't compiling). In the future, we might want to warn about
non-trivial code appearing in an unreachable context, and/or avoid
generating unreachable code altogether (though I'm sure LLVM will weed
it out as well).

13 years agoEradicate fold from capture.rs
Marijn Haverbeke [Thu, 5 May 2011 13:03:37 +0000 (15:03 +0200)]
Eradicate fold from capture.rs

The pass now uses walk.

13 years agoMake ocamlc and optional dependency.
Ralph Giles [Wed, 4 May 2011 23:24:47 +0000 (16:24 -0700)]
Make ocamlc and optional dependency.

Since moving to a downloadable snapshot for stage0, it's been
possible to complete the build without compiling code written
in ocaml. However, the configure script still required it to
be present.

This commit changes detection failure to a non-fatal condition,
allowing the build to complete on systems without an ocaml
compiler.

An info message is also made conditional.

13 years agoMark valgrind's _qzz_res as unused to silence warnings on gcc 4.6.
Ralph Giles [Thu, 5 May 2011 06:19:38 +0000 (23:19 -0700)]
Mark valgrind's _qzz_res as unused to silence warnings on gcc 4.6.

This applies c11675 by Julian Seward from valgrind trunk
(svn://svn.valgrind.org/valgrind/trunk) to rustrt's included
copies of memcheck.h and valgrind.h, effectively backporting
the fix from the unreleased 3.6.2. The commit simply applies
the gcc 'unused' attribute to the relevant declarations.

This change allows compilation of the runtime code under
gcc 4.6 with -Werror, as the makefile currently requests.

13 years agorustc: Add a Link module; move crate writing to it to slim down trans slightly
Patrick Walton [Thu, 5 May 2011 17:48:02 +0000 (10:48 -0700)]
rustc: Add a Link module; move crate writing to it to slim down trans slightly

13 years agorustllvm: Add bindings to the LLVM linker
Patrick Walton [Thu, 5 May 2011 04:27:00 +0000 (21:27 -0700)]
rustllvm: Add bindings to the LLVM linker

13 years agorustllvm: Whitespace police in RustWrapper.cpp
Patrick Walton [Thu, 5 May 2011 03:30:23 +0000 (20:30 -0700)]
rustllvm: Whitespace police in RustWrapper.cpp

13 years agorustc: Detect the system root and allow the user to override if necessary
Patrick Walton [Thu, 5 May 2011 02:27:54 +0000 (19:27 -0700)]
rustc: Detect the system root and allow the user to override if necessary

13 years agobuild: Whitespace police in configure. Puts out burning tinderbox.
Patrick Walton [Thu, 5 May 2011 02:14:01 +0000 (19:14 -0700)]
build: Whitespace police in configure. Puts out burning tinderbox.

13 years agoAdd #env syntax extension for plucking strings out of the compilation environment.
Graydon Hoare [Thu, 5 May 2011 02:05:16 +0000 (19:05 -0700)]
Add #env syntax extension for plucking strings out of the compilation environment.

13 years agoRe-snapshot to pick up fixed crasher bugs in resolve and performance improvements.
Graydon Hoare [Thu, 5 May 2011 02:04:34 +0000 (19:04 -0700)]
Re-snapshot to pick up fixed crasher bugs in resolve and performance improvements.

13 years agobuild: Add missing intrinsics.mk
Patrick Walton [Thu, 5 May 2011 01:35:36 +0000 (18:35 -0700)]
build: Add missing intrinsics.mk

13 years agobuild: Build intrinsics.bc
Patrick Walton [Thu, 5 May 2011 01:28:30 +0000 (18:28 -0700)]
build: Build intrinsics.bc

13 years agoAdd stage-specific test targets (make check-stage1 etc.)
Graydon Hoare [Thu, 5 May 2011 00:32:05 +0000 (17:32 -0700)]
Add stage-specific test targets (make check-stage1 etc.)

13 years agoFix and re-enable destructors
Marijn Haverbeke [Thu, 5 May 2011 00:13:39 +0000 (02:13 +0200)]
Fix and re-enable destructors

There was some confusion on whether the destructors took their
argument by pointer or direct value. They now take it directly, just
like other methods. You no longer get a segfault when a constructor
actually does something with its self value.

13 years agoAdd a structure for passing option flags around the compiler, put it in session,...
Graydon Hoare [Wed, 4 May 2011 23:53:42 +0000 (16:53 -0700)]
Add a structure for passing option flags around the compiler, put it in session, and use it.

13 years agoAdd *.pyc to .gitignore
Brian Anderson [Wed, 4 May 2011 23:33:51 +0000 (19:33 -0400)]
Add *.pyc to .gitignore

13 years agoAdd missing ret statements to check_mod
Brian Anderson [Wed, 4 May 2011 23:29:27 +0000 (19:29 -0400)]
Add missing ret statements to check_mod

Un-XFAIL compile-fail export tests

13 years agoSupport new -g flag, only mangle glue names by type when it's passed.
Graydon Hoare [Wed, 4 May 2011 22:36:42 +0000 (15:36 -0700)]
Support new -g flag, only mangle glue names by type when it's passed.

13 years agoMore export XFAILs.
Graydon Hoare [Wed, 4 May 2011 22:35:41 +0000 (15:35 -0700)]
More export XFAILs.

13 years agoRe-snapshot stage0
Graydon Hoare [Wed, 4 May 2011 21:45:09 +0000 (14:45 -0700)]
Re-snapshot stage0

13 years agoXFAIL some compile-fail tests mysteriously failing on win32 tinderbox.
Graydon Hoare [Wed, 4 May 2011 21:44:57 +0000 (14:44 -0700)]
XFAIL some compile-fail tests mysteriously failing on win32 tinderbox.

13 years agoextend xfail-stage0 to stage1, stage2 on obj-dtor-2.rs
Graydon Hoare [Wed, 4 May 2011 18:23:05 +0000 (11:23 -0700)]
extend xfail-stage0 to stage1, stage2 on obj-dtor-2.rs

13 years agoFix horrible rule-hijacking bug that was causing compile-fail targets to trigger...
Graydon Hoare [Wed, 4 May 2011 18:21:14 +0000 (11:21 -0700)]
Fix horrible rule-hijacking bug that was causing compile-fail targets to trigger honest compile attempts.

13 years agoChange two compile-fail patterns to match rustc output.
Graydon Hoare [Wed, 4 May 2011 18:20:47 +0000 (11:20 -0700)]
Change two compile-fail patterns to match rustc output.

13 years agoTry to fail nicely, again.
Graydon Hoare [Wed, 4 May 2011 15:04:53 +0000 (15:04 +0000)]
Try to fail nicely, again.

13 years agoFail in a slightly more organized fashion for now.
Graydon Hoare [Wed, 4 May 2011 06:19:58 +0000 (06:19 +0000)]
Fail in a slightly more organized fashion for now.

13 years agorustc: Stub Rust intrinsics. We check in the .ll file for now to temporarily dodge...
Patrick Walton [Wed, 4 May 2011 01:49:29 +0000 (18:49 -0700)]
rustc: Stub Rust intrinsics. We check in the .ll file for now to temporarily dodge a dependency on clang.

13 years agoAssume xfail-stage0 implies stage1 and stage2 in tests, for now.
Graydon Hoare [Wed, 4 May 2011 01:13:51 +0000 (18:13 -0700)]
Assume xfail-stage0 implies stage1 and stage2 in tests, for now.

13 years agoFix bug in CFG_TESTLIB that only manifests on stage1, stage2.
Graydon Hoare [Wed, 4 May 2011 01:09:18 +0000 (18:09 -0700)]
Fix bug in CFG_TESTLIB that only manifests on stage1, stage2.

13 years agorustc: Stub support for Rust intrinsics
Patrick Walton [Wed, 4 May 2011 01:03:59 +0000 (18:03 -0700)]
rustc: Stub support for Rust intrinsics

13 years agoRevert "Rename the "llvm" API to "llvm-intrinsic"" due to tinderbox bustage
Patrick Walton [Wed, 4 May 2011 00:49:58 +0000 (17:49 -0700)]
Revert "Rename the "llvm" API to "llvm-intrinsic"" due to tinderbox bustage

This reverts commit 6871c245a67fab222eccc2a21dcb620d11d3b0d0.

13 years agoMore python portability fixes (for windows).
Graydon Hoare [Tue, 3 May 2011 23:53:36 +0000 (16:53 -0700)]
More python portability fixes (for windows).

13 years agoAttempt to use more portable python
Graydon Hoare [Tue, 3 May 2011 23:22:00 +0000 (16:22 -0700)]
Attempt to use more portable python

13 years agoRemove everything to do with rustboot from makefiles.
Graydon Hoare [Tue, 3 May 2011 22:51:21 +0000 (15:51 -0700)]
Remove everything to do with rustboot from makefiles.

13 years agoStart sketching --depend support in rustc.
Graydon Hoare [Tue, 3 May 2011 22:50:56 +0000 (15:50 -0700)]
Start sketching --depend support in rustc.

13 years agoRename the "llvm" API to "llvm-intrinsic"
Patrick Walton [Tue, 3 May 2011 22:55:01 +0000 (15:55 -0700)]
Rename the "llvm" API to "llvm-intrinsic"

13 years agorustc: Refactor metadata.Encode.* to not require a type abbreviation table if abbrevi...
Patrick Walton [Tue, 3 May 2011 21:36:04 +0000 (14:36 -0700)]
rustc: Refactor metadata.Encode.* to not require a type abbreviation table if abbreviation isn't enabled

13 years agorustc: Name type glue properly
Patrick Walton [Tue, 3 May 2011 20:03:32 +0000 (13:03 -0700)]
rustc: Name type glue properly

13 years agoRemove calls to rustboot -rdeps, leave 'touch' in place while working on replacement.
Graydon Hoare [Tue, 3 May 2011 21:18:30 +0000 (14:18 -0700)]
Remove calls to rustboot -rdeps, leave 'touch' in place while working on replacement.

13 years agoUpdate copyrights for completeness.
Graydon Hoare [Tue, 3 May 2011 19:10:03 +0000 (12:10 -0700)]
Update copyrights for completeness.

13 years agoUpdate valgrind.h and memcheck.h to 3.6.0
Graydon Hoare [Tue, 3 May 2011 19:04:00 +0000 (12:04 -0700)]
Update valgrind.h and memcheck.h to 3.6.0

13 years agoAttempt to shift build to stage0-from-snapshots.
Graydon Hoare [Tue, 3 May 2011 18:34:44 +0000 (11:34 -0700)]
Attempt to shift build to stage0-from-snapshots.

13 years agoSmall fixes to snapshot script.
Graydon Hoare [Tue, 3 May 2011 16:25:59 +0000 (09:25 -0700)]
Small fixes to snapshot script.

13 years agoRemove unused variable.
Rafael Ávila de Espíndola [Tue, 3 May 2011 15:06:57 +0000 (11:06 -0400)]
Remove unused variable.

13 years agoSplit start into rustc and rustboot versions. This introduces a bit of
Rafael Ávila de Espíndola [Tue, 3 May 2011 14:19:28 +0000 (10:19 -0400)]
Split start into rustc and rustboot versions. This introduces a bit of
duplication, but we will hopefully drop the rustboot one soon.

This is also a preparation for changing the rustc one to have the activate glue
return to the exit glue which will then call the main function.

This (returning to the function that calls main) matches what happens when
loader stats a program or a new thread. It lets gdb produce good backtraces
and should help with EH too.

13 years agoMore snapshot logic refactoring.
Graydon Hoare [Tue, 3 May 2011 14:23:05 +0000 (07:23 -0700)]
More snapshot logic refactoring.

13 years agoAdd forgotten 'snap.mk'
Graydon Hoare [Tue, 3 May 2011 06:43:13 +0000 (23:43 -0700)]
Add forgotten 'snap.mk'

13 years agoMore hacking on the snapshot system.
Graydon Hoare [Tue, 3 May 2011 06:37:52 +0000 (23:37 -0700)]
More hacking on the snapshot system.

13 years agoExtract ast.is_exported from the resolve module
Brian Anderson [Tue, 3 May 2011 00:29:39 +0000 (20:29 -0400)]
Extract ast.is_exported from the resolve module

13 years agoAdd some tests of tag-export interaction
Brian Anderson [Tue, 3 May 2011 00:29:11 +0000 (20:29 -0400)]
Add some tests of tag-export interaction

13 years agoInitial sketch of snapshot support code, based on code from marijn.
Graydon Hoare [Tue, 3 May 2011 01:56:13 +0000 (18:56 -0700)]
Initial sketch of snapshot support code, based on code from marijn.

13 years agoRevert "Use check instead of assert in export-unexported-dep"
Brian Anderson [Tue, 3 May 2011 00:53:18 +0000 (20:53 -0400)]
Revert "Use check instead of assert in export-unexported-dep"

This reverts commit 480eda0f10fabd6e76a30c91df1d84740d6f81cc. The commit
adding the assert keyword has been restored.

13 years agoAdd stage1, stage2 to Makefile.in default target, so tinderboxes notice breakage.
Graydon Hoare [Tue, 3 May 2011 00:56:15 +0000 (17:56 -0700)]
Add stage1, stage2 to Makefile.in default target, so tinderboxes notice breakage.

13 years agorustc: Time the Rust translation and LLVM passes separately
Patrick Walton [Tue, 3 May 2011 00:45:07 +0000 (17:45 -0700)]
rustc: Time the Rust translation and LLVM passes separately

13 years agorustc: Remove the artificial block in ty.rs. Fixes self-hosting. Add a test case...
Patrick Walton [Tue, 3 May 2011 00:44:22 +0000 (17:44 -0700)]
rustc: Remove the artificial block in ty.rs. Fixes self-hosting. Add a test case for this, XFAIL'd.

13 years agoUn-revert "Use different syntax for checks that matter to typestate", fixing the...
Patrick Walton [Tue, 3 May 2011 00:47:24 +0000 (17:47 -0700)]
Un-revert "Use different syntax for checks that matter to typestate", fixing the problem.

This reverts commit d08b443fffb1181d8d45ae5d061412f202dd4118.

13 years agoUse check instead of assert in export-unexported-dep
Brian Anderson [Tue, 3 May 2011 00:48:52 +0000 (20:48 -0400)]
Use check instead of assert in export-unexported-dep

13 years agoRevert "Use different syntax for checks that matter to typestate"
Graydon Hoare [Mon, 2 May 2011 23:24:09 +0000 (16:24 -0700)]
Revert "Use different syntax for checks that matter to typestate"

This reverts commit aa25f22f197682de3b18fc4c8ba068d1feda220f. It broke stage2, not sure why yet.

13 years agoAdd a regression test that exports can expose unexported items
Brian Anderson [Mon, 2 May 2011 23:42:00 +0000 (19:42 -0400)]
Add a regression test that exports can expose unexported items

While those unexported items can't be used by name outside the module in which
they are defined, they can be used as a sort of ADT.

13 years agorustc: Allocate tydescs on the stack when it's safe to do so. 60% compile speed increase.
Patrick Walton [Mon, 2 May 2011 22:28:22 +0000 (15:28 -0700)]
rustc: Allocate tydescs on the stack when it's safe to do so. 60% compile speed increase.

13 years agoUse different syntax for checks that matter to typestate
Tim Chevalier [Mon, 2 May 2011 18:23:07 +0000 (11:23 -0700)]
Use different syntax for checks that matter to typestate

This giant commit changes the syntax of Rust to use "assert" for
"check" expressions that didn't mean anything to the typestate
system, and continue using "check" for checks that are used as
part of typestate checking.

Most of the changes are just replacing "check" with "assert" in test
cases and rustc.

13 years agorustc: Add a "fat tydesc" LLVM type to trans
Patrick Walton [Mon, 2 May 2011 19:05:21 +0000 (12:05 -0700)]
rustc: Add a "fat tydesc" LLVM type to trans

13 years agorustc: Disable frame pointer omission
Patrick Walton [Mon, 2 May 2011 18:01:28 +0000 (11:01 -0700)]
rustc: Disable frame pointer omission

13 years agoAdd a test that imports can't circumvent exports
Brian Anderson [Sun, 1 May 2011 21:22:20 +0000 (17:22 -0400)]
Add a test that imports can't circumvent exports

13 years agoRemove the search direction from resolve's fold environment
Brian Anderson [Sun, 1 May 2011 20:55:57 +0000 (16:55 -0400)]
Remove the search direction from resolve's fold environment

It's not actually involved in the fold so it can just be passed between the
functions that need it.

13 years agoAdd a regression test for use of unexported fully-qualified paths
Brian Anderson [Sun, 1 May 2011 20:29:52 +0000 (16:29 -0400)]
Add a regression test for use of unexported fully-qualified paths

An unexported foo.baz can't be resolved from inside foo when called as foo.baz
instead of just baz. This behavior may want to change eventually.

13 years agoHide unexported tag variants
Brian Anderson [Sun, 1 May 2011 20:01:27 +0000 (16:01 -0400)]
Hide unexported tag variants

13 years agoImplement simple module export
Brian Anderson [Sun, 24 Apr 2011 23:40:01 +0000 (19:40 -0400)]
Implement simple module export

13 years agoStart splitting up Makefile.in
Graydon Hoare [Sun, 1 May 2011 20:18:52 +0000 (20:18 +0000)]
Start splitting up Makefile.in

13 years agorustc: Preserve dots in input path when using an implicit output path
Brian Anderson [Sun, 1 May 2011 18:42:45 +0000 (14:42 -0400)]
rustc: Preserve dots in input path when using an implicit output path

This avoids outputing, e.g. /test.bc for ../test.rs

13 years agoFix expr-alt-generic-box2.rs and un-XFAIL
Brian Anderson [Sun, 1 May 2011 16:21:08 +0000 (12:21 -0400)]
Fix expr-alt-generic-box2.rs and un-XFAIL

13 years agoUn-XFAIL various tests in stage0
Brian Anderson [Sun, 1 May 2011 16:15:42 +0000 (12:15 -0400)]
Un-XFAIL various tests in stage0