]>
git.rm.cloudns.org Git - xonotic/gmqcc.git/log
Dale Weiler [Sat, 22 Dec 2012 07:24:25 +0000 (07:24 +0000)]
Cleanups
Dale Weiler [Fri, 21 Dec 2012 11:13:16 +0000 (06:13 -0500)]
Windows stuff for visual studio now offical works with exception to colorized output to cmd (which actually should work). As well as the testsuite which lacks implementations of task_popen/task_pclose.
Wolfgang Bumiller [Fri, 21 Dec 2012 10:42:23 +0000 (11:42 +0100)]
-Opeephole now also reduces CALL stores
Wolfgang Bumiller [Fri, 21 Dec 2012 10:33:44 +0000 (11:33 +0100)]
Rename ir_function_pass_tailcall->tailrecursion; Take -fadjust-vector-fields into account when storing field type return-values from functions
Wolfgang Bumiller [Fri, 21 Dec 2012 10:02:36 +0000 (11:02 +0100)]
manpage: -Wcpp; added a FILES section which contains gmqcc.ini.example - we should install this to some DATADIR in the install make target
Dale Weiler [Fri, 21 Dec 2012 06:59:34 +0000 (01:59 -0500)]
testsuite compiles on windows now (but doesn't work). Still need to implement bidirectional piping on stdout/stderr/stdin.. Which for some reason is incredibly stupidly complicated on windows. You need to setup SECURITY_ATTRIBUTES for each handle, do crazy handle duplication, and forking via CreateThread. Meanwhile the whole "concept" of threading is entierly broken in fundamental design on windows, which means you constantly need to store the current process handles for backup, and "inherit" from child process. It's just stupid, provide unix pipes via close/open and dup. Like normal people (they're simple and sane).
Dale Weiler [Fri, 21 Dec 2012 06:00:51 +0000 (01:00 -0500)]
Get gmqcc/qcvm compiling on windows again. Plus work in progress support for the testsuite on windows (does not compile yet).
Dale Weiler [Fri, 21 Dec 2012 04:48:01 +0000 (04:48 +0000)]
Implemented preprocessing error directive tests (should fail)
Dale Weiler [Fri, 21 Dec 2012 04:01:47 +0000 (04:01 +0000)]
Document ini.example
Dale Weiler [Fri, 21 Dec 2012 03:14:31 +0000 (03:14 +0000)]
Update ini.example (added UNREACHABLE_CODE, and CPP) as warning options
Dale Weiler [Fri, 21 Dec 2012 03:12:58 +0000 (03:12 +0000)]
added -Wcpp (for turning off cpp warnings defined with #warning like GCC/clang/pathscale .. no more [-Wunused-variable])
Dale Weiler [Fri, 21 Dec 2012 03:08:21 +0000 (03:08 +0000)]
Implemented #warning and #error preprocessor directives, they're functionally equivalent to CPPs (supporting both string constant and non-string constant versions). Warnings however are printed with a [-Wunused-variable] which isn't correct (TODO: allow systematic changes of -W paramaters in relation to warning fields for preprocessor directives.
Dale Weiler [Thu, 20 Dec 2012 23:07:33 +0000 (23:07 +0000)]
Merge branch 'master' of github.com:graphitemaster/gmqcc
Dale Weiler [Thu, 20 Dec 2012 23:07:06 +0000 (23:07 +0000)]
Implemented roboust compile-time endianess check.
Wolfgang Bumiller [Thu, 20 Dec 2012 22:52:54 +0000 (23:52 +0100)]
Remove old tempalloc code; disable -Olocaltemps for now
Wolfgang Bumiller [Thu, 20 Dec 2012 22:48:41 +0000 (23:48 +0100)]
Be more correct with sizes; use a better local-alloc strategy
Wolfgang Bumiller [Thu, 20 Dec 2012 22:27:23 +0000 (23:27 +0100)]
minor: fix ast_function_label
Wolfgang Bumiller [Thu, 20 Dec 2012 21:03:51 +0000 (22:03 +0100)]
Fix util_endianswap; and endianswap the LNO data as well
Wolfgang Bumiller [Thu, 20 Dec 2012 20:05:29 +0000 (21:05 +0100)]
Don't check for is_return but for final at the end of a function so that if 'goto' is the last expression we don't try to append a return
Wolfgang Bumiller [Thu, 20 Dec 2012 19:57:05 +0000 (20:57 +0100)]
*sigh*
Wolfgang Bumiller [Thu, 20 Dec 2012 19:56:49 +0000 (20:56 +0100)]
no need to check the array, the 2nd NOT is a NOT_F
Wolfgang Bumiller [Thu, 20 Dec 2012 19:56:08 +0000 (20:56 +0100)]
manpage: -fcorrect-logic, -ftrue-empty-strings, -ffalse-empty-strings
Wolfgang Bumiller [Thu, 20 Dec 2012 19:55:56 +0000 (20:55 +0100)]
fix -fcorrect-logic to use double-not instead of a single one - wow that was a bad one
Wolfgang Bumiller [Thu, 20 Dec 2012 19:43:58 +0000 (20:43 +0100)]
Testcases for -ftrue-empty-strings and -ffalse-empty-strings
Wolfgang Bumiller [Thu, 20 Dec 2012 19:40:59 +0000 (20:40 +0100)]
Fix the -fcorrect-logic condition
Wolfgang Bumiller [Thu, 20 Dec 2012 19:33:05 +0000 (20:33 +0100)]
Fix -fcorrect-logic
Wolfgang Bumiller [Thu, 20 Dec 2012 19:26:54 +0000 (20:26 +0100)]
fix a leak in the local-allocator
Wolfgang Bumiller [Thu, 20 Dec 2012 19:26:14 +0000 (20:26 +0100)]
Fix a leak: free code_linenums
Wolfgang Bumiller [Thu, 20 Dec 2012 19:23:35 +0000 (20:23 +0100)]
For loops may have an empty condition - don't try processing a NULL
Wolfgang Bumiller [Thu, 20 Dec 2012 19:22:31 +0000 (20:22 +0100)]
while/for/do now all use process_condition, ast_loop got 2 new parameters for more efficient translation
Wolfgang Bumiller [Thu, 20 Dec 2012 18:51:30 +0000 (19:51 +0100)]
process_condition function used by parse_if - need to still add it to loops
Wolfgang Bumiller [Thu, 20 Dec 2012 18:32:57 +0000 (19:32 +0100)]
-ftrue-empty-strings now changes type_not_instr[TYPE_STRING] to INSTR_NOT_F
Wolfgang Bumiller [Thu, 20 Dec 2012 18:19:12 +0000 (19:19 +0100)]
since unary not doesn't fold these, remove them from immediate_is_true
Wolfgang Bumiller [Thu, 20 Dec 2012 18:18:22 +0000 (19:18 +0100)]
introducing an immediate_is_true which follows the new -f flags on strings and vector truthfulness, used in logic-op constant folding; const folding now disabled for functions-constants on NOT because it's stupid
Wolfgang Bumiller [Thu, 20 Dec 2012 18:04:56 +0000 (19:04 +0100)]
unary not now uses NOT_F with -ftrue-empty-strings
Wolfgang Bumiller [Thu, 20 Dec 2012 18:00:23 +0000 (19:00 +0100)]
Making logical 'and' and 'or' use NOT_ in -fcorrect-logic for both operands normally, but only for the first with -fperl-logic
Wolfgang Bumiller [Thu, 20 Dec 2012 17:59:05 +0000 (18:59 +0100)]
Adding -fcorrect-logic, -ftrue-empty-strings, -ffalse-empty-strings flags; not used yet
Wolfgang Bumiller [Thu, 20 Dec 2012 15:56:22 +0000 (16:56 +0100)]
manpage: typo
Wolfgang Bumiller [Thu, 20 Dec 2012 15:55:00 +0000 (16:55 +0100)]
Fix condition on whether or not to print 'error' or 'warning'
Wolfgang Bumiller [Thu, 20 Dec 2012 15:53:09 +0000 (16:53 +0100)]
-Werror-<warning> should actually use opts_setwerror
Wolfgang Bumiller [Thu, 20 Dec 2012 15:49:10 +0000 (16:49 +0100)]
-Werror-<warning>, -Wno-error-<warning>, manpage updated
Wolfgang Bumiller [Thu, 20 Dec 2012 15:41:40 +0000 (16:41 +0100)]
manpage: -Wunreachable-code
Wolfgang Bumiller [Thu, 20 Dec 2012 15:32:36 +0000 (16:32 +0100)]
Reorder the warnings in opts.c to match their oder in opts.def, added WARN_LOCAL_SHADOWS to the list with its default false parameter
Wolfgang Bumiller [Thu, 20 Dec 2012 15:29:32 +0000 (16:29 +0100)]
Turn -fallow-unreachable-code into a -Wunreachable-code
Wolfgang Bumiller [Thu, 20 Dec 2012 15:28:35 +0000 (16:28 +0100)]
Fix object dependency: OBJ -> OBJ_D so they all depend on gmqcc.h again properly
Wolfgang Bumiller [Thu, 20 Dec 2012 15:20:08 +0000 (16:20 +0100)]
Temporarily fix noreturn bug by creating an actual dummy return instruction after the call for now
Wolfgang Bumiller [Thu, 20 Dec 2012 15:04:10 +0000 (16:04 +0100)]
-g sets opts.g and -debug causes blocks to be labeled
Wolfgang Bumiller [Thu, 20 Dec 2012 14:46:31 +0000 (15:46 +0100)]
sscanf_s only for _MSC_VER not WIN32 in exec.c
Wolfgang Bumiller [Thu, 20 Dec 2012 14:45:15 +0000 (15:45 +0100)]
Oh god I hate this so much
Wolfgang Bumiller [Thu, 20 Dec 2012 14:27:15 +0000 (15:27 +0100)]
that should have gone into the second to last commit
Wolfgang Bumiller [Thu, 20 Dec 2012 14:26:45 +0000 (15:26 +0100)]
WINDOWS 8 USERS ARE IDIOTS, YES YOU SHOULD TAKE THIS PERSONALLY
Wolfgang Bumiller [Thu, 20 Dec 2012 14:01:19 +0000 (15:01 +0100)]
ast_return needs the correct context...
Wolfgang Bumiller [Thu, 20 Dec 2012 13:41:16 +0000 (14:41 +0100)]
-fallow-unreachable-code
Wolfgang Bumiller [Thu, 20 Dec 2012 13:12:22 +0000 (14:12 +0100)]
vector initialization tests
Wolfgang Bumiller [Thu, 20 Dec 2012 13:06:25 +0000 (14:06 +0100)]
making the uninitialized-var on vector warning now check all members; ir_value_vector_member now also appends _x,_y,_z since they're used in the used-unintialized message
Wolfgang Bumiller [Thu, 20 Dec 2012 12:44:53 +0000 (13:44 +0100)]
Remove the suffix entirely
Wolfgang Bumiller [Thu, 20 Dec 2012 11:55:21 +0000 (12:55 +0100)]
UL->ULL
Wolfgang Bumiller [Thu, 20 Dec 2012 11:00:39 +0000 (12:00 +0100)]
Added tests for 'noreturn'
Dale Weiler [Thu, 20 Dec 2012 10:21:31 +0000 (10:21 +0000)]
ugh todo: seed per ht, not per hash
Dale Weiler [Thu, 20 Dec 2012 10:20:16 +0000 (10:20 +0000)]
No more hacks
Dale Weiler [Thu, 20 Dec 2012 09:44:47 +0000 (09:44 +0000)]
fixes
Dale Weiler [Thu, 20 Dec 2012 09:01:01 +0000 (09:01 +0000)]
One last fix for mankind
Dale Weiler [Thu, 20 Dec 2012 08:51:29 +0000 (08:51 +0000)]
*sigh*
Dale Weiler [Thu, 20 Dec 2012 08:49:14 +0000 (08:49 +0000)]
Allow uname override
Dale Weiler [Thu, 20 Dec 2012 08:46:39 +0000 (08:46 +0000)]
Fix murmur hash seeding
Dale Weiler [Thu, 20 Dec 2012 08:45:54 +0000 (08:45 +0000)]
More fixes
Dale Weiler [Thu, 20 Dec 2012 08:19:43 +0000 (08:19 +0000)]
Another fix
Dale Weiler [Thu, 20 Dec 2012 08:17:04 +0000 (08:17 +0000)]
FTBFS fixes
Dale Weiler [Thu, 20 Dec 2012 00:22:22 +0000 (00:22 +0000)]
handle cygwin/mingw for makefile
Dale Weiler [Wed, 19 Dec 2012 21:37:37 +0000 (21:37 +0000)]
Typos
Dale Weiler [Wed, 19 Dec 2012 21:35:47 +0000 (21:35 +0000)]
Implemened -fail procedure type for testsuite. Essentially performs compilation on the idea that the compilation should fail, for the test to succeed, otherwise the test actually fails.
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 21:17:22 +0000 (22:17 +0100)]
Applying the actual 'Fix FTFBS stuff' part
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 21:15:38 +0000 (22:15 +0100)]
Revert "Fix FTFBS stuff"
This reverts commit
2cc51b8eb060e6efab5b56cf471594911a79c5c1 .
Dale Weiler [Wed, 19 Dec 2012 21:05:24 +0000 (21:05 +0000)]
Fix FTFBS stuff
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 19:54:54 +0000 (20:54 +0100)]
'noreturn' is - for now - a keyword, and used just like noref/var/const, to mark a function as not-returning
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 19:47:01 +0000 (20:47 +0100)]
ast_call now passes AST_FLAG_NORETURN on to the IR
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 19:45:48 +0000 (20:45 +0100)]
removing ast_expression_common.variadic, adding ast_expression_common.flags, added AST_FLAG_VARIADIC and AST_FLAG_NORETURN
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 19:40:45 +0000 (20:40 +0100)]
NRCALL now sets block->final and block->is_return
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 19:39:44 +0000 (20:39 +0100)]
ir_block_create_{phi,call} now check self->final
Wolfgang (Blub) Bumiller [Wed, 19 Dec 2012 19:38:32 +0000 (20:38 +0100)]
VINSTR_NRCALL, translated like any other call for now; to be used to mark a call which never returns, ie the error builtin
Wolfgang Bumiller [Tue, 18 Dec 2012 21:01:43 +0000 (13:01 -0800)]
Merge pull request #48 from mhofstra/patch-1
Small typo fix to make it compile on x86
mhofstra [Tue, 18 Dec 2012 20:57:16 +0000 (21:57 +0100)]
Small typo fix to make it compile on x86
Rudolf Polzer [Tue, 18 Dec 2012 15:45:18 +0000 (16:45 +0100)]
Fixes for some minor bugs clang-analyzer and cppcheck found
Signed-off-by: Dale Weiler <killfieldengine@gmail.com>
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:58:21 +0000 (16:58 +0100)]
default make target now is 'all'
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:57:38 +0000 (16:57 +0100)]
Tests for perl-logic _types_
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:56:22 +0000 (16:56 +0100)]
-fperl-logic now doesn't allow logic ops with operands of different types, but therefore uses the correct output type
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:52:30 +0000 (16:52 +0100)]
Added intrinsic: __builtin_debug_typestring(expr) to get the type of an expression as string
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:36:06 +0000 (16:36 +0100)]
Remove a TODO message
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:03:48 +0000 (16:03 +0100)]
make install-doc now also installs qcvm.q
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 15:03:29 +0000 (16:03 +0100)]
Adding qcvm.1
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 14:46:26 +0000 (15:46 +0100)]
Improved parameter handling of qcvm a little, added --version and --help
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 14:44:26 +0000 (15:44 +0100)]
fix wrong --version string
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 14:39:20 +0000 (15:39 +0100)]
Better usage message for qcvm, don't exit after any of the information prints - but don't execute if any of them are used
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 14:30:59 +0000 (15:30 +0100)]
BUGS, SEE ALSO and AUTHOR sections in the manpage
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 14:27:12 +0000 (15:27 +0100)]
It's -config=file, not -config file
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 14:26:41 +0000 (15:26 +0100)]
Added -config to usage and manpage, added a CONFIG section in the manpage describing the ini file format, and added a short example to it
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 13:21:01 +0000 (14:21 +0100)]
Add included -fs and -Ws to for each -std in the manpage
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 12:40:59 +0000 (13:40 +0100)]
More jump removal
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 12:13:54 +0000 (13:13 +0100)]
*sigh*
Wolfgang (Blub) Bumiller [Tue, 18 Dec 2012 11:58:15 +0000 (12:58 +0100)]
vector components are floats, not vectors