]>
git.rm.cloudns.org Git - xonotic/gmqcc.git/log
Dale Weiler [Tue, 4 Jun 2013 17:44:45 +0000 (17:44 +0000)]
Some typos.
Dale Weiler [Tue, 4 Jun 2013 17:43:20 +0000 (17:43 +0000)]
Gentoo ebuilds
Dale Weiler [Tue, 4 Jun 2013 02:53:44 +0000 (02:53 +0000)]
Add license header
Dale Weiler [Tue, 4 Jun 2013 02:49:49 +0000 (02:49 +0000)]
Doh
Dale Weiler [Tue, 4 Jun 2013 02:47:07 +0000 (02:47 +0000)]
Major header reworking, this respects the namespaces properly. Makes object dependency more obvious, allows for better make caches, and prevents misuse of library features, i.e use con_* instead of printf.
Wolfgang Bumiller [Sun, 2 Jun 2013 10:35:53 +0000 (12:35 +0200)]
remove the ast tree earlier - saves over 3% peak memory usage
Dale Weiler [Sun, 2 Jun 2013 09:50:39 +0000 (09:50 +0000)]
Fix a divide by zero error
Dale Weiler [Sun, 2 Jun 2013 08:52:30 +0000 (08:52 +0000)]
Fix output
Dale Weiler [Sun, 2 Jun 2013 08:49:16 +0000 (08:49 +0000)]
Fix two bugs
Dale Weiler [Sun, 2 Jun 2013 08:37:22 +0000 (08:37 +0000)]
Some cleanups
Dale Weiler [Sun, 2 Jun 2013 08:28:56 +0000 (08:28 +0000)]
Fix the peak
Dale Weiler [Sun, 2 Jun 2013 08:25:00 +0000 (08:25 +0000)]
Holy whoops
Dale Weiler [Sun, 2 Jun 2013 08:21:06 +0000 (08:21 +0000)]
Major utility rewrite for compiler memory utilization statistics. Cleanups everywhere, no more NOTRACK stuff, all allocates are tracked. Major identifier cleanups as well.
Dale Weiler [Sun, 2 Jun 2013 05:32:37 +0000 (05:32 +0000)]
Some static
Dale Weiler [Sun, 2 Jun 2013 05:09:28 +0000 (05:09 +0000)]
Hashtable statistics too
Dale Weiler [Sun, 2 Jun 2013 04:45:04 +0000 (04:45 +0000)]
Track hashtables too
Dale Weiler [Sun, 2 Jun 2013 04:38:20 +0000 (04:38 +0000)]
Track strdups too in the statistics
Dale Weiler [Sun, 2 Jun 2013 04:32:20 +0000 (04:32 +0000)]
This was an experiment that snuk it's way inside that commit.
Dale Weiler [Sun, 2 Jun 2013 04:29:53 +0000 (04:29 +0000)]
Vector tracing, step two in determining how to lower memory usage.
Dale Weiler [Sun, 2 Jun 2013 00:07:54 +0000 (00:07 +0000)]
Guart statistics by option.
Dale Weiler [Sun, 2 Jun 2013 00:03:27 +0000 (00:03 +0000)]
Add statistics for vector usage to the output. Step one in lowering memory usage. ONE FUCKING MILLION VECTORS for XONOTIC..
Wolfgang Bumiller [Sat, 1 Jun 2013 20:25:58 +0000 (22:25 +0200)]
fix a leak on a parse-error
Dale Weiler [Sat, 1 Jun 2013 20:20:25 +0000 (20:20 +0000)]
Doh
Dale Weiler [Sat, 1 Jun 2013 20:18:53 +0000 (20:18 +0000)]
Fix some possible memleaks.
Wolfgang Bumiller [Sat, 1 Jun 2013 18:12:27 +0000 (20:12 +0200)]
Merge branch 'master' into cooking
Wolfgang Bumiller [Sat, 1 Jun 2013 18:12:14 +0000 (20:12 +0200)]
fix the INCLUDE_DEF flag getting applied regardless of dotranslate...
Wolfgang Bumiller [Sat, 1 Jun 2013 16:16:56 +0000 (18:16 +0200)]
Merge branch 'master' into cooking
Wolfgang Bumiller [Sat, 1 Jun 2013 16:16:50 +0000 (18:16 +0200)]
fix: dotranslate now sets AST_FLAG_INCLUDE_DEF to not get removed by -Ostrip-constant-names
Dale Weiler [Thu, 30 May 2013 19:36:37 +0000 (19:36 +0000)]
Merge branch 'cooking' of github.com:graphitemaster/gmqcc into cooking
Dale Weiler [Thu, 30 May 2013 19:36:01 +0000 (19:36 +0000)]
column printing for warnings and errors now
Wolfgang Bumiller [Wed, 29 May 2013 15:08:03 +0000 (17:08 +0200)]
now a bare 'return;' is not required anymore if return has been assigned anywhere
Wolfgang Bumiller [Wed, 29 May 2013 15:01:26 +0000 (17:01 +0200)]
make sure missing return values still warn
Wolfgang Bumiller [Wed, 29 May 2013 14:58:18 +0000 (16:58 +0200)]
return assignment factorial test
Wolfgang Bumiller [Wed, 29 May 2013 14:56:39 +0000 (16:56 +0200)]
fixing that comment...
Wolfgang Bumiller [Wed, 29 May 2013 14:51:59 +0000 (16:51 +0200)]
assignable return value now lives in ast_function, as globals can get overwritten randomly; removed parser_find_retval; updated parse_return
Dale Weiler [Wed, 29 May 2013 11:32:42 +0000 (11:32 +0000)]
Make static
Dale Weiler [Wed, 29 May 2013 11:30:26 +0000 (11:30 +0000)]
Update test template.
Dale Weiler [Wed, 29 May 2013 11:29:52 +0000 (11:29 +0000)]
Update man page
Dale Weiler [Wed, 29 May 2013 11:13:42 +0000 (11:13 +0000)]
Experimental support for implicit return assignments. This closes #107. To enable return assignment support use -freturn-assignments. This allows you to do the following in QC code. T name() { return = expr_eval_T; return; /* returns expr_eval_T */ }. It allows for concise code and to rid locals in functions. It also saves a tremendous amount of space since only types of certian globals need to be allocated for returns.
Dale Weiler [Wed, 29 May 2013 03:43:49 +0000 (03:43 +0000)]
Merge branch 'union-replacement' into cooking
Conflicts:
ast.c
Dale Weiler [Wed, 29 May 2013 03:39:38 +0000 (03:39 +0000)]
Add back that splint flag
Dale Weiler [Wed, 29 May 2013 03:35:51 +0000 (03:35 +0000)]
Fix that memory leak that I spent hours trying to find months ago and didn't succeed at.
Dale Weiler [Wed, 29 May 2013 03:29:04 +0000 (03:29 +0000)]
Major export cleanup. Anything that was exported but wasn't used outside where it was implemented has been turned into static, and their exports have been removed. This actually makes the compiler compile slightly faster.
Dale Weiler [Wed, 29 May 2013 02:37:45 +0000 (02:37 +0000)]
Prevent negitive numbers from being used as the left operand of shift operator.
Dale Weiler [Wed, 29 May 2013 02:24:12 +0000 (02:24 +0000)]
Result of shift operation on signed integers is bad.
Dale Weiler [Wed, 29 May 2013 02:16:50 +0000 (02:16 +0000)]
Cleanups and fixes
Wolfgang Bumiller [Sat, 11 May 2013 20:34:01 +0000 (22:34 +0200)]
killing the union types
Wolfgang Bumiller [Sat, 11 May 2013 20:05:30 +0000 (22:05 +0200)]
first step to getting rid of those unions we don't use anyway
Wolfgang Bumiller [Tue, 7 May 2013 18:35:20 +0000 (20:35 +0200)]
less confusing
Wolfgang Bumiller [Tue, 7 May 2013 18:13:19 +0000 (20:13 +0200)]
be more strict here
Wolfgang Bumiller [Tue, 7 May 2013 17:59:06 +0000 (19:59 +0200)]
removing this dead code_init call
Wolfgang Bumiller [Tue, 7 May 2013 17:56:41 +0000 (19:56 +0200)]
fixing a few leaks - code_write doesn't delete the code object anymore, code_cleanup has to be called
Wolfgang Bumiller [Tue, 7 May 2013 17:39:10 +0000 (19:39 +0200)]
changing a few ast_delete to ast_unref in places where the to-be-deleted maybe coming from some other place
Wolfgang Bumiller [Sun, 5 May 2013 11:12:14 +0000 (04:12 -0700)]
Merge pull request #111 from matthiaskrgr/cooking_stuff
.gitignore: add binarys
Matthias Krüger [Sun, 5 May 2013 10:53:58 +0000 (12:53 +0200)]
.gitignore: add binarys
Dale Weiler [Thu, 2 May 2013 19:18:59 +0000 (19:18 +0000)]
Escape strings in executor for -printdefs
Dale Weiler [Sun, 28 Apr 2013 04:29:20 +0000 (04:29 +0000)]
Implemented support for having codegen (and lno files) be wrote out to memory, this essentially allows us to embed the compiler in applications now.
Dale Weiler [Sun, 28 Apr 2013 03:50:22 +0000 (03:50 +0000)]
Fix deps (remove duplicates with clever make tricks)
Dale Weiler [Sun, 28 Apr 2013 03:42:04 +0000 (03:42 +0000)]
Merge branch 'cooking' of github.com:graphitemaster/gmqcc into cooking
Dale Weiler [Sun, 28 Apr 2013 03:41:03 +0000 (03:41 +0000)]
Less general name for PAK utility, added install targets for PAK utility, added manpag for PAK utility, this closes #108
Wolfgang Bumiller [Sat, 27 Apr 2013 17:05:06 +0000 (19:05 +0200)]
Merge branch 'master' into cooking
Wolfgang Bumiller [Sat, 27 Apr 2013 17:04:43 +0000 (10:04 -0700)]
Merge pull request #109 from matthiaskrgr/PKGBUILD_master
PKGBUILD: archlinux: fix build by applying a patch (
4c4aa5534c34 ) on 0.2.9 package.
Accepting this with a patch in order for it to keep the 0.2.9 version info.
Matthias Krüger [Sat, 27 Apr 2013 16:45:26 +0000 (18:45 +0200)]
PKGBUILD: archlinux: fix build by applying a patch (
4c4aa5534c34 ) on 0.2.9 package.
Wolfgang Bumiller [Sat, 27 Apr 2013 15:52:54 +0000 (17:52 +0200)]
Merge branch 'master' into cooking
Wolfgang Bumiller [Sat, 27 Apr 2013 15:51:51 +0000 (17:51 +0200)]
dup is marked as warn-unused-result
Dale Weiler [Sat, 27 Apr 2013 15:20:01 +0000 (15:20 +0000)]
Expression has undefined behavior (left operand modifies code->globals->used, used by right operand): (code->globals)[(((vector_t*)((void *)code->globals)) - 1)->used++] = (code_genstring(code, global->constval.vstring)) Code has unspecified behavior. Order of evaluation of function parameters or subexpressions is not defined, so if a value is used and modified in different places not separated by a sequence point constraining evaluation order, then the result of the expression is unspecified.
Wolfgang Bumiller [Sat, 27 Apr 2013 15:00:27 +0000 (17:00 +0200)]
debian Makefile had the same flaw as archlinux/this... fixing
Wolfgang Bumiller [Sat, 27 Apr 2013 14:55:28 +0000 (16:55 +0200)]
Merge branch 'master' into cooking
Wolfgang Bumiller [Sat, 27 Apr 2013 14:54:53 +0000 (16:54 +0200)]
updating archbsd and archlinux release/PKGBUILD files
Wolfgang Bumiller [Sat, 27 Apr 2013 14:33:55 +0000 (16:33 +0200)]
Merge branch 'master' into cooking
Wolfgang Bumiller [Sat, 27 Apr 2013 14:33:47 +0000 (16:33 +0200)]
Update CHANGES file
Wolfgang Bumiller [Sat, 27 Apr 2013 14:30:35 +0000 (16:30 +0200)]
Starting point of 0.3.0
Wolfgang Bumiller [Sat, 27 Apr 2013 14:30:03 +0000 (16:30 +0200)]
this should be tagged 0.2.9
Wolfgang Bumiller [Fri, 26 Apr 2013 15:56:08 +0000 (17:56 +0200)]
fix distro/ Makefiles
Wolfgang Bumiller [Fri, 26 Apr 2013 15:33:56 +0000 (17:33 +0200)]
don't overwrite ldflags/libs env vars
Dale Weiler [Fri, 26 Apr 2013 15:43:13 +0000 (15:43 +0000)]
Fix OSX compiles
Wolfgang Bumiller [Fri, 26 Apr 2013 15:33:56 +0000 (17:33 +0200)]
don't overwrite ldflags/libs env vars
Wolfgang Bumiller [Fri, 26 Apr 2013 08:45:13 +0000 (10:45 +0200)]
ast_value now has an initializer list array, still unused but it's there for later
Wolfgang Bumiller [Fri, 26 Apr 2013 08:31:38 +0000 (10:31 +0200)]
an = before an { is also always required when the declared variable is not a function; adding because of planned array initializers
Dale Weiler [Thu, 25 Apr 2013 17:39:05 +0000 (17:39 +0000)]
Add expressions for builtins test
Wolfgang Bumiller [Thu, 25 Apr 2013 17:18:50 +0000 (19:18 +0200)]
solve it, not hide it
Wolfgang Bumiller [Thu, 25 Apr 2013 17:18:11 +0000 (19:18 +0200)]
Revert "Just fucking initialize everything"
This reverts commit
aed2b1031c6466a2dbcf909faeac1ac1cff0f6e6 .
Wolfgang Bumiller [Thu, 25 Apr 2013 17:13:43 +0000 (19:13 +0200)]
manpage and ini update
Dale Weiler [Thu, 25 Apr 2013 17:11:15 +0000 (17:11 +0000)]
Just fucking initialize everything
Dale Weiler [Thu, 25 Apr 2013 17:08:02 +0000 (17:08 +0000)]
Pushing the -fexpressions-for-builtins stuff and the modff for catching fractional-part builtin numbers.
Wolfgang Bumiller [Thu, 25 Apr 2013 16:47:54 +0000 (18:47 +0200)]
this allows builtin numbers to be constant expressions, not just constants
Wolfgang Bumiller [Thu, 25 Apr 2013 15:39:12 +0000 (17:39 +0200)]
change the value of TOKEN_EOF
Dale Weiler [Thu, 25 Apr 2013 12:22:34 +0000 (12:22 +0000)]
Less extern
Dale Weiler [Thu, 25 Apr 2013 12:08:43 +0000 (12:08 +0000)]
cleaner
Dale Weiler [Thu, 25 Apr 2013 12:08:13 +0000 (12:08 +0000)]
less globals
Dale Weiler [Thu, 25 Apr 2013 09:35:30 +0000 (09:35 +0000)]
No more globals for codegen
Wolfgang Bumiller [Thu, 25 Apr 2013 07:55:58 +0000 (09:55 +0200)]
these can be const now
Dale Weiler [Thu, 25 Apr 2013 03:37:40 +0000 (03:37 +0000)]
That is also not needed anymore
Dale Weiler [Thu, 25 Apr 2013 03:36:57 +0000 (03:36 +0000)]
not needed for arm anymore
Dale Weiler [Thu, 25 Apr 2013 03:34:42 +0000 (03:34 +0000)]
Apply some static where it belongs
Dale Weiler [Thu, 25 Apr 2013 03:17:52 +0000 (03:17 +0000)]
Fix alignment issues (can remove a clang warning, and also makes our code more portable!)
Dale Weiler [Thu, 25 Apr 2013 03:11:50 +0000 (03:11 +0000)]
Use the LIBS properly
Wolfgang Bumiller [Wed, 24 Apr 2013 15:47:42 +0000 (17:47 +0200)]
Grrrr
Wolfgang Bumiller [Wed, 24 Apr 2013 15:42:48 +0000 (17:42 +0200)]
manpage: document and explain __FUNC__ and __NULL__
Wolfgang Bumiller [Wed, 24 Apr 2013 14:07:31 +0000 (16:07 +0200)]
Allow string concatenation in the parser as well, for now only for immediate strings and __FUNC__, maybe later -fpermissive or somethign could allow it on const-delcared string variables