]>
git.lizzy.rs Git - rust.git/log
Brian Anderson [Sun, 2 Oct 2011 22:30:25 +0000 (15:30 -0700)]
Add Haitao Li to AUTHORS.txt
Haitao Li [Sun, 2 Oct 2011 12:52:07 +0000 (20:52 +0800)]
rustc: Fix link error on Ubuntu Oneiric. Issue #903
Brian Anderson [Sun, 2 Oct 2011 03:12:08 +0000 (20:12 -0700)]
Lots of build system cleanup and documentation
This splits mk/stageN.mk into host.mk and target.mk and makes
the build rules somewhat simpler - there's no more building from stageN
into stageN+1; instead we always build from stageN(host) to
stageN(target) then promote from stageN(target) to stageN+1(host).
Add a big honkin explaination right at the top of Makefile.in
Brian Anderson [Sat, 1 Oct 2011 02:00:19 +0000 (19:00 -0700)]
Have 'make install' do something close to useful
We'll copy the files to the right location. Still need to rpath.
Issue #529
Patrick Walton [Sat, 1 Oct 2011 01:35:13 +0000 (18:35 -0700)]
rustc: Call LLVM on the C stack
Patrick Walton [Sat, 1 Oct 2011 01:49:46 +0000 (18:49 -0700)]
rt: Omit leading underscore on symbol names on Linux
Patrick Walton [Sat, 1 Oct 2011 01:20:28 +0000 (18:20 -0700)]
rustc: Stub a --stack-growth option; it's behind a flag for now because it requires patches to LLVM.
Brian Anderson [Sat, 1 Oct 2011 01:03:54 +0000 (18:03 -0700)]
Register new snapshots
Brian Anderson [Fri, 30 Sep 2011 23:55:18 +0000 (16:55 -0700)]
Change the directory for target libs
This pushes them down from stageN/lib/rustc/$(target) to
stageN/lib/rustc/$(target)/lib in order to make room for a target bin dir
Brian Anderson [Fri, 30 Sep 2011 23:27:18 +0000 (16:27 -0700)]
Refix the snapshot rules
Brian Anderson [Fri, 30 Sep 2011 23:15:02 +0000 (16:15 -0700)]
Refix the fuzzer make rule
Brian Anderson [Fri, 30 Sep 2011 23:11:47 +0000 (16:11 -0700)]
Make config of static stdlib more consistent with rest of makefile
I haven't tested if it actually works. Static stdlib is probably becoming
bitrotted.
Brian Anderson [Fri, 30 Sep 2011 19:24:28 +0000 (12:24 -0700)]
Abstract the build directories further in the makefiles
This defines variables for host bin and lib directories as well as all target
bin and lib directories then uses them everywhere.
Brian Anderson [Fri, 30 Sep 2011 19:08:51 +0000 (12:08 -0700)]
Fix some make variables in tests.mk
Patrick Walton [Fri, 30 Sep 2011 21:26:06 +0000 (14:26 -0700)]
rustc: Remove some debug code
Patrick Walton [Fri, 30 Sep 2011 21:25:10 +0000 (14:25 -0700)]
rustc: Track link_args attributes on C stack cdecl natives as well
Patrick Walton [Fri, 30 Sep 2011 19:07:50 +0000 (12:07 -0700)]
trans: Use inttoptr as appropriate when casting return values of C stack functions
Brian Anderson [Fri, 30 Sep 2011 07:15:28 +0000 (00:15 -0700)]
Add Erik Rose to AUTHORS.txt
Brian Anderson [Fri, 30 Sep 2011 05:29:44 +0000 (22:29 -0700)]
Register snapshots and finish filesystem reorg
Brian Anderson [Fri, 30 Sep 2011 05:58:17 +0000 (22:58 -0700)]
Remove legacy library search paths
Brian Anderson [Fri, 30 Sep 2011 05:01:16 +0000 (22:01 -0700)]
Update snapshot scripts for the new regime
Brian Anderson [Fri, 30 Sep 2011 04:34:46 +0000 (21:34 -0700)]
Tweak makefiles, change the way the fuzzer is built
Since librustc and the fuzzer depend on rustllvm, build them
like rustc, linking to libraries in lib instead of lib/rustc/$(target)
Brian Anderson [Fri, 30 Sep 2011 03:27:40 +0000 (20:27 -0700)]
Remove redundant make clean rules
Brian Anderson [Fri, 30 Sep 2011 03:27:28 +0000 (20:27 -0700)]
Fix make all build rule
Brian Anderson [Fri, 30 Sep 2011 01:30:00 +0000 (18:30 -0700)]
Teach compiletest to construct the PATH variable correctly on win32
Brian Anderson [Fri, 30 Sep 2011 01:11:53 +0000 (18:11 -0700)]
Fix CFG_RUN_TARG on windows
Brian Anderson [Thu, 29 Sep 2011 19:21:58 +0000 (12:21 -0700)]
Fix a variety of things in the makefiles
Make tests and perf work again. Correctly get the host triple in rustc. Add
some FIXMEs about transitional code that needs to be removed.
Brian Anderson [Thu, 29 Sep 2011 19:06:37 +0000 (12:06 -0700)]
Remove the redundant definition of STAGE in makefiles
Brian Anderson [Thu, 29 Sep 2011 18:48:44 +0000 (11:48 -0700)]
Update make clean rules
Brian Anderson [Thu, 29 Sep 2011 18:31:28 +0000 (11:31 -0700)]
Make rustc's host triple agree with the build system's
Brian Anderson [Thu, 29 Sep 2011 00:06:57 +0000 (17:06 -0700)]
Reorganize structure of build directory
Each stage is organized more according to Unix standards and to
accommodate multiple target architectures.
stageN/
bin - rustc lives here
lib - libraries that rustc needs
lib/rustc/$(target_triple/ - target libraries
Patrick Walton [Fri, 30 Sep 2011 01:53:00 +0000 (18:53 -0700)]
rustc: Remove some debug code
Patrick Walton [Fri, 30 Sep 2011 01:52:23 +0000 (18:52 -0700)]
rustc: Make the generic C stack function signature include a usable return value
Patrick Walton [Fri, 30 Sep 2011 01:36:27 +0000 (18:36 -0700)]
stdlib: Un-export str::buf; add a comment
Patrick Walton [Fri, 30 Sep 2011 01:18:37 +0000 (18:18 -0700)]
vim: Add `pure` to the list of keywords
Patrick Walton [Fri, 30 Sep 2011 00:58:51 +0000 (17:58 -0700)]
stdlib: Export str::buf
Patrick Walton [Fri, 30 Sep 2011 00:33:41 +0000 (17:33 -0700)]
rustc: Translate calls on the C stack
Patrick Walton [Fri, 30 Sep 2011 00:33:14 +0000 (17:33 -0700)]
rustc: Fix call signature for call_c_stack
Patrick Walton [Fri, 30 Sep 2011 00:32:28 +0000 (17:32 -0700)]
rt: Fix assembler for C calls to work more like a regular function call
Patrick Walton [Fri, 30 Sep 2011 00:31:11 +0000 (17:31 -0700)]
stdlib: Add vec::eachi to the standard library
Patrick Walton [Thu, 29 Sep 2011 01:15:28 +0000 (18:15 -0700)]
rustc: Add upcall stubs for the new C stack upcalls
Patrick Walton [Wed, 28 Sep 2011 22:15:54 +0000 (15:15 -0700)]
rustc: Generate types in trans for the C-stack native ABI
Brian Anderson [Thu, 29 Sep 2011 23:47:09 +0000 (16:47 -0700)]
doc: Don't merge function and concept index. There is no findex
This is causing errors on my windows install
Marijn Haverbeke [Thu, 29 Sep 2011 10:02:37 +0000 (12:02 +0200)]
Don't build up results for ignored expressions.
Issue #667
Marijn Haverbeke [Thu, 29 Sep 2011 09:50:03 +0000 (11:50 +0200)]
Fail with an informative error when trying to bind a method
Issue #435
Marijn Haverbeke [Thu, 29 Sep 2011 09:18:40 +0000 (11:18 +0200)]
Move last remaining expression types out of trans_expr
Issue #667
Marijn Haverbeke [Thu, 29 Sep 2011 08:46:49 +0000 (10:46 +0200)]
Move expr_cast and expr_copy to DPS style
Issue #667
Marijn Haverbeke [Wed, 28 Sep 2011 13:57:38 +0000 (15:57 +0200)]
Move closure construction over to DPS style
Issue #667
Marijn Haverbeke [Wed, 28 Sep 2011 13:15:29 +0000 (15:15 +0200)]
Reduce move ops to a construct-in-place when possible
Brian Anderson [Wed, 28 Sep 2011 22:26:11 +0000 (15:26 -0700)]
Eliminate glue.o. Closes #990
Brian Anderson [Wed, 28 Sep 2011 21:39:57 +0000 (14:39 -0700)]
Test cases for #985
Other restrictions on pinned kinds happened to fix this
Closes #985
Brian Anderson [Wed, 28 Sep 2011 21:37:28 +0000 (14:37 -0700)]
Don't allow vectors of pinned kinds
Vectors of pinned kinds can't be safe because most interesting uses of vector
perform copies
Brian Anderson [Wed, 28 Sep 2011 21:24:56 +0000 (14:24 -0700)]
Make error patterns in some kind tests more specific
Patrick Walton [Wed, 28 Sep 2011 19:58:03 +0000 (12:58 -0700)]
rustc: Add a new "C stack cdecl" native ABI
Brian Anderson [Wed, 28 Sep 2011 19:24:15 +0000 (12:24 -0700)]
xfail-fast run-pass/alt-range.rs
Brian Anderson [Wed, 28 Sep 2011 19:07:33 +0000 (12:07 -0700)]
Revert "Revert "Implement pattern ranges for all numeric types.""
This reverts commit
a034f87146e60e1db2327c6f6807c47406a1bb0b .
Conflicts:
src/comp/middle/check_alt.rs
src/comp/middle/trans_alt.rs
src/comp/syntax/ast.rs
src/comp/syntax/ast_util.rs
src/comp/syntax/fold.rs
src/comp/syntax/print/pprust.rs
Conflicts:
src/comp/middle/trans_alt.rs
Patrick Walton [Wed, 28 Sep 2011 19:40:34 +0000 (12:40 -0700)]
vim: `const` is a keyword
Patrick Walton [Wed, 28 Sep 2011 19:26:36 +0000 (12:26 -0700)]
rt: Add a call stub that switches to the C stack, untested as of yet
Patrick Walton [Wed, 28 Sep 2011 18:31:44 +0000 (11:31 -0700)]
rt: Add an upcall to allocate space on the C stack. This will be used for native calls on the C stack.
Patrick Walton [Wed, 28 Sep 2011 02:19:19 +0000 (19:19 -0700)]
rt: Expose the C stack inside the Rust scheduler so that C calls can switch to it
Patrick Walton [Wed, 28 Sep 2011 01:42:06 +0000 (18:42 -0700)]
rustc: Switch to cdecl for all calls. This is needed to make stack growth efficient, as I need to use eax and ecx in the function prologue and can't afford to stomp on incoming arguments.
Marijn Haverbeke [Wed, 28 Sep 2011 10:59:44 +0000 (12:59 +0200)]
Remove a few kludges that worked around issue #913
Marijn Haverbeke [Wed, 28 Sep 2011 10:56:54 +0000 (12:56 +0200)]
Register new snapshot
It had been a while. I mostly wanted a fix for #913 in stage0.
Marijn Haverbeke [Wed, 28 Sep 2011 10:23:08 +0000 (12:23 +0200)]
Don't descend into functions from relax_precond_block
Closes #913
Marijn Haverbeke [Wed, 28 Sep 2011 09:12:22 +0000 (11:12 +0200)]
Get rid of might_not_init kludge in init_local.
Initializing something is now safe wrt to cleanups (so the cleanup for the
local is only registered after the initialization.)
Marijn Haverbeke [Wed, 28 Sep 2011 08:41:19 +0000 (10:41 +0200)]
Wait with registering cleanups for locals until after the init code ran
Fixes a double-cleanup that was breaking unwind-tup.rs under valgrind.
Marijn Haverbeke [Tue, 27 Sep 2011 18:21:44 +0000 (20:21 +0200)]
Use DPS for assignment and local initialization
Issue #667
Brian Anderson [Wed, 28 Sep 2011 06:04:34 +0000 (23:04 -0700)]
Add Wade Mealing to AUTHORS.txt
Wade Mealing [Tue, 27 Sep 2011 18:15:24 +0000 (04:15 +1000)]
Patch to error instead of crashing when parsing unmatched double quotes
Patch to error and fail instead of using all available memory
then crashing to detect the error condition of an unmatched
double quote before the end of a file.
I couldn't get it to show nice error messages, so this may not be
the ideal fix.
A test case for this situation has also been added.
Brian Anderson [Wed, 28 Sep 2011 05:40:15 +0000 (22:40 -0700)]
Make it again possible to initialize resource locals via assignment
Some special cases allow both 'let a <- my_resource(x)' and
'let a = my_resource(x)' to work as expected despite ostensibly being
copies and moves.
Brian Anderson [Wed, 28 Sep 2011 01:45:35 +0000 (18:45 -0700)]
Prevent copies of resources into various things
Patrick Walton [Wed, 28 Sep 2011 00:14:18 +0000 (17:14 -0700)]
rt: Store the task in TLS
Brian Anderson [Tue, 27 Sep 2011 22:28:25 +0000 (15:28 -0700)]
Add more comments about kind checking on local move initializers
Brian Anderson [Tue, 27 Sep 2011 22:22:06 +0000 (15:22 -0700)]
Add xfailed tests that resources can't be copied into various things
Brian Anderson [Tue, 27 Sep 2011 21:50:55 +0000 (14:50 -0700)]
Don't ever raise unique kinds of pinned kinds to shared (again)
So *resource, ~resource, [resource] are all pinned. This is counter to the
design of the kind system, but this way is a much clearer path to type safety.
Once we've established a good baseline with lots of tests, then we can try to
make raising pinned kinds work.
Brian Anderson [Tue, 27 Sep 2011 22:56:32 +0000 (15:56 -0700)]
Revert "Use DPS for assignment and local initialization"
This reverts commit
b1b202d302d1e84886360f9c8f146b9562db5f23 .
Brian Anderson [Tue, 27 Sep 2011 21:28:20 +0000 (14:28 -0700)]
Add an xfailed test for bogus deep copying of things containing resources
Brian Anderson [Tue, 27 Sep 2011 21:24:18 +0000 (14:24 -0700)]
Add an xfailed test for bogus vector addition of typarams
Patrick Walton [Tue, 27 Sep 2011 21:01:06 +0000 (14:01 -0700)]
doc: Remove `case` from the list of keywords
Patrick Walton [Tue, 27 Sep 2011 20:57:16 +0000 (13:57 -0700)]
doc: Remove `tup` from the list of keywords
Marijn Haverbeke [Tue, 27 Sep 2011 20:49:33 +0000 (22:49 +0200)]
Fix lurking bug in matching of nil
Which was somehow made visible by the DPS changes.
Marijn Haverbeke [Tue, 27 Sep 2011 18:21:44 +0000 (20:21 +0200)]
Use DPS for assignment and local initialization
Issue #667
Marijn Haverbeke [Tue, 27 Sep 2011 18:20:51 +0000 (20:20 +0200)]
Safeguard against self-assignment for unique types
Jesse Ruderman [Tue, 27 Sep 2011 18:05:16 +0000 (11:05 -0700)]
Test for #924
Marijn Haverbeke [Tue, 27 Sep 2011 11:19:55 +0000 (13:19 +0200)]
Move expr_unary to DPS style
Issue #667
Marijn Haverbeke [Tue, 27 Sep 2011 10:02:01 +0000 (12:02 +0200)]
Make trans_rec and trans_tup understand overwrite dests
Issue #667
Marijn Haverbeke [Tue, 27 Sep 2011 08:50:18 +0000 (10:50 +0200)]
Move expr_binary into trans_expr_dps
Issue #667
Marijn Haverbeke [Tue, 27 Sep 2011 06:42:27 +0000 (08:42 +0200)]
Move expr_lit and expr_vec into trans_expr_dps
Issue #667
Marijn Haverbeke [Tue, 27 Sep 2011 06:03:06 +0000 (08:03 +0200)]
Move more non-value-returning things to trans_expr_dps.
Issue #667
Brian Anderson [Tue, 27 Sep 2011 06:45:51 +0000 (23:45 -0700)]
Add two xfailed tests for invalid copies into vectors
Brian Anderson [Tue, 27 Sep 2011 06:34:54 +0000 (23:34 -0700)]
xfail-win32 -> xfail-fast
Brian Anderson [Tue, 27 Sep 2011 06:34:12 +0000 (23:34 -0700)]
Try to appease windows by xfail-win32'ing a random test
Brian Anderson [Tue, 27 Sep 2011 06:24:53 +0000 (23:24 -0700)]
Test for issue #980
Brian Anderson [Tue, 27 Sep 2011 04:57:42 +0000 (21:57 -0700)]
Enforce copy restrictions on let initializers
Brian Anderson [Tue, 27 Sep 2011 04:17:11 +0000 (21:17 -0700)]
Make some nocopy tests a bit more futureproof
Turn the let assignment initializers into move initializers since
that assignment is going to become illegal
Jesse Ruderman [Tue, 27 Sep 2011 03:44:57 +0000 (20:44 -0700)]
Distinguish #924
Jesse Ruderman [Tue, 27 Sep 2011 03:34:07 +0000 (20:34 -0700)]
Add a cycle-collection fuzzer
Brian Anderson [Tue, 27 Sep 2011 03:03:14 +0000 (20:03 -0700)]
Prevent copying of uncopyable things via the copy op
Brian Anderson [Tue, 27 Sep 2011 02:59:49 +0000 (19:59 -0700)]
Prevent copying of uncopyable things via compound assignment ops
Brian Anderson [Tue, 27 Sep 2011 02:56:38 +0000 (19:56 -0700)]
Don't allow vectors of pinned kinds to be copied