cacao [Thu, 1 Apr 2004 16:35:39 +0000 (16:35 +0000)]
some words for intro and overview by andi
edwin [Wed, 31 Mar 2004 22:44:07 +0000 (22:44 +0000)]
added field and method resolution
edwin [Wed, 31 Mar 2004 20:13:14 +0000 (20:13 +0000)]
added exceptions to field/method resolution
twisti [Wed, 31 Mar 2004 19:16:23 +0000 (19:16 +0000)]
- removoed unused functions
- removed remaining proto_ comments
- disass prints addresses instead of method offsets, it's easier to find
something (jumps need to be fixed, i think)
twisti [Wed, 31 Mar 2004 13:34:45 +0000 (13:34 +0000)]
500k is not enough, we go with 1M.
twisti [Tue, 30 Mar 2004 23:01:45 +0000 (23:01 +0000)]
- usage of new_exception
twisti [Tue, 30 Mar 2004 23:00:14 +0000 (23:00 +0000)]
Huh, heap size was too small.
twisti [Tue, 30 Mar 2004 22:56:04 +0000 (22:56 +0000)]
- usage of new_exception
- some cosmetic changes
twisti [Tue, 30 Mar 2004 22:49:03 +0000 (22:49 +0000)]
Replaced native_new_and_init with new_exception.
twisti [Tue, 30 Mar 2004 22:44:28 +0000 (22:44 +0000)]
- removed unused exception class_ declarations
- added exception string_'s for code generation
- new_exception* functions to instantiate runtime exceptions
twisti [Tue, 30 Mar 2004 22:42:57 +0000 (22:42 +0000)]
- removed references to proto_ exceptions
- use new_exception instead of native_new_and_init
twisti [Tue, 30 Mar 2004 22:34:40 +0000 (22:34 +0000)]
- removed unused exception class_ declarations
- removed unused exception proto_'s, except NullPointer
- removed create_system_exception_classes(),
create_system_exception_proto_classes()
twisti [Tue, 30 Mar 2004 21:59:27 +0000 (21:59 +0000)]
- removed unused exception class_ declarations
- added exception string_'s for code generation
- new_exception* functions to instantiate runtime exceptions
twisti [Tue, 30 Mar 2004 21:49:28 +0000 (21:49 +0000)]
- heap_init works again and we can set initial and maximun heap size
- implemented Java_java_lang_Runtime_maxMemory, also in boehm itself
(GC_get_max_heap_size)
twisti [Tue, 30 Mar 2004 21:41:36 +0000 (21:41 +0000)]
- removed warnings with newest autoconf version (except 1)
- removed default THREADS=no, for native thread support
twisti [Tue, 30 Mar 2004 21:39:17 +0000 (21:39 +0000)]
Removed warnings for newest autconf versions. We AC_PREREQ version 2.59,
because boehmm wants 2.53 and classpath 2.59.
twisti [Tue, 30 Mar 2004 21:36:03 +0000 (21:36 +0000)]
Remove stamp-nat in cleanfiles.
twisti [Tue, 30 Mar 2004 21:26:36 +0000 (21:26 +0000)]
Not needed anymore or it should not reside in cvs.
stefan [Mon, 29 Mar 2004 11:22:34 +0000 (11:22 +0000)]
Native threads almost working
stefan [Mon, 29 Mar 2004 11:03:56 +0000 (11:03 +0000)]
Native threads almost working
stefan [Mon, 29 Mar 2004 11:03:06 +0000 (11:03 +0000)]
*** empty log message ***
stefan [Mon, 29 Mar 2004 07:13:42 +0000 (07:13 +0000)]
Removed cast_lock
stefan [Mon, 29 Mar 2004 07:12:38 +0000 (07:12 +0000)]
*** empty log message ***
stefan [Mon, 29 Mar 2004 07:02:38 +0000 (07:02 +0000)]
crash fix
twisti [Sun, 28 Mar 2004 23:32:53 +0000 (23:32 +0000)]
This will come, but not yet!
twisti [Sun, 28 Mar 2004 23:08:07 +0000 (23:08 +0000)]
Implemented some functions: gc_get_free_bytes, gc_get_heap_size,
gc_finalize_all
twisti [Sun, 28 Mar 2004 23:06:43 +0000 (23:06 +0000)]
Ooops, small type. But gcc does not complain about #elsif.
twisti [Sun, 28 Mar 2004 22:37:16 +0000 (22:37 +0000)]
Implemented some functions: totalMemory, freeMemory, availableProcessors
runFinalization
twisti [Fri, 26 Mar 2004 00:34:51 +0000 (00:34 +0000)]
Removed gcc3 compiler warnings and some other warnings.
twisti [Thu, 25 Mar 2004 23:47:49 +0000 (23:47 +0000)]
Pass code pointer to disassinstr instead of one instruction, so we can
print the instruction address.
twisti [Thu, 25 Mar 2004 23:46:22 +0000 (23:46 +0000)]
Replaced proto_ exceptions with dynamic ones.
twisti [Thu, 25 Mar 2004 23:45:51 +0000 (23:45 +0000)]
Print real addresses instead of method byte offset. So we can find
exception pc's better.
twisti [Thu, 25 Mar 2004 18:37:52 +0000 (18:37 +0000)]
Replaced #ifdef __ALPHA__ with #ifndef __I386__. I thought this was already in cvs.
twisti [Thu, 25 Mar 2004 18:18:52 +0000 (18:18 +0000)]
These should not reside in cvs.
twisti [Thu, 25 Mar 2004 18:09:45 +0000 (18:09 +0000)]
Removed unused asm_switchstackandcall
jowenn [Thu, 25 Mar 2004 17:31:13 +0000 (17:31 +0000)]
inlining starts working again for the first time after typecheck modifications. -inv doesn't work yet, and the rest isn't tested excaustive yet,but a hello world works. I had to introduce a class member on instruction level, first I had thought basic block level would be enough, but inlined functions do not start at a basic block boundary. I'm not sure if that is intended or a bug
twisti [Thu, 25 Mar 2004 15:19:16 +0000 (15:19 +0000)]
- added gnu header, emacs tailer
- removed log_text
edwin [Wed, 24 Mar 2004 22:48:01 +0000 (22:48 +0000)]
added functions for resolving fields and methods
edwin [Wed, 24 Mar 2004 22:44:29 +0000 (22:44 +0000)]
more debug output
twisti [Wed, 24 Mar 2004 00:16:07 +0000 (00:16 +0000)]
Removed unused functions and .globl's.
twisti [Mon, 22 Mar 2004 20:51:52 +0000 (20:51 +0000)]
Generate all autmake and autoconf files.
jowenn [Fri, 19 Mar 2004 17:52:42 +0000 (17:52 +0000)]
patching asm_builtin_new away and do direct calls to builtin_new the next time
jowenn [Thu, 18 Mar 2004 14:29:03 +0000 (14:29 +0000)]
additional fields for line numbers added to data segment, codegen generates padding bytes, debug output in monitorenter/exit commented out
twisti [Wed, 17 Mar 2004 15:58:01 +0000 (15:58 +0000)]
- -verbosecall fixes
- save return register before monitorexit
twisti [Wed, 17 Mar 2004 11:38:26 +0000 (11:38 +0000)]
Finished calljavafunction2 copy loop stuff.
jowenn [Mon, 15 Mar 2004 14:52:43 +0000 (14:52 +0000)]
compile and run again on alpha. No new features from i386 ported yet. It only runs with threads disabled on alpha right now
jowenn [Mon, 15 Mar 2004 07:37:49 +0000 (07:37 +0000)]
exception back traces for i386
twisti [Mon, 15 Mar 2004 00:37:59 +0000 (00:37 +0000)]
- fixed -verbosecall
- replaced proto_ exceptions with dynamic ones
twisti [Mon, 15 Mar 2004 00:35:39 +0000 (00:35 +0000)]
- added gnu header and emacs tailer
- added asm_check_clinit
- make asm_calljavafunction2 work with current stuff, but not ready yet!
twisti [Sun, 14 Mar 2004 23:41:11 +0000 (23:41 +0000)]
Added some functions from native used in loader.
twisti [Sun, 14 Mar 2004 23:39:31 +0000 (23:39 +0000)]
- show message in ExceptionInInitializerError
- make code more readable
twisti [Sun, 14 Mar 2004 21:02:49 +0000 (21:02 +0000)]
POWERPC defines for time outputs.
twisti [Sun, 14 Mar 2004 21:01:12 +0000 (21:01 +0000)]
POWERPC defines in builtin_trace_args and builtin_displaymethodstop.
jowenn [Sun, 14 Mar 2004 14:22:24 +0000 (14:22 +0000)]
testcase including fillInStackTrace and rethrow examples
jowenn [Sat, 13 Mar 2004 12:51:30 +0000 (12:51 +0000)]
the line number <-> mpc table is now stored into the data segment (i386)
stefan [Fri, 12 Mar 2004 10:47:33 +0000 (10:47 +0000)]
configure script from the gc6.2 package - thread support working now
stefan [Thu, 11 Mar 2004 23:02:26 +0000 (23:02 +0000)]
GC_LINUX_THREADS only for linux
stefan [Thu, 11 Mar 2004 21:52:35 +0000 (21:52 +0000)]
GC_LINUX_THREADS
jowenn [Thu, 11 Mar 2004 17:30:03 +0000 (17:30 +0000)]
cacao is now again able to run kjc on i386
twisti [Sun, 7 Mar 2004 23:52:44 +0000 (23:52 +0000)]
- replaced proto_ exceptions with dynamic ones
twisti [Sun, 7 Mar 2004 22:16:31 +0000 (22:16 +0000)]
native_new_and_init_int was not defined
twisti [Sun, 7 Mar 2004 22:12:28 +0000 (22:12 +0000)]
- replace proto_ exceptions with dynamic ones
twisti [Sun, 7 Mar 2004 22:02:29 +0000 (22:02 +0000)]
- replaced proto_ exception with dynamic ones
- removed softnull define
- XXX extest does not work with extest!
twisti [Sun, 7 Mar 2004 17:14:46 +0000 (17:14 +0000)]
Check dynamic inline exceptions (generated after actual code).
twisti [Sat, 6 Mar 2004 21:33:21 +0000 (21:33 +0000)]
Ok, once again: variable declarations at block beginnings! Not everyone has
a gcc-3.x, not even a gcc.
twisti [Sat, 6 Mar 2004 20:50:31 +0000 (20:50 +0000)]
Variables must be declared at block beginnings, damn!
twisti [Sat, 6 Mar 2004 20:20:38 +0000 (20:20 +0000)]
- built with autoconf-2.59 and automake-1.8.2
- gnuclasspath is configured with AC_CONFIG_SUBDIRS
jowenn [Sat, 6 Mar 2004 17:32:16 +0000 (17:32 +0000)]
missing extern and ;
jowenn [Sat, 6 Mar 2004 17:27:56 +0000 (17:27 +0000)]
retrieve line number information
jowenn [Sat, 6 Mar 2004 14:04:15 +0000 (14:04 +0000)]
Make Method.getExceptionTypes work
twisti [Sat, 6 Mar 2004 13:57:41 +0000 (13:57 +0000)]
- removed inc, dec macros -- they only write partially the flag register
and this is slower on new processors
- replaced proto_ exception with dynamic ones, stuff as StringBuffer(-1)
should work now
twisti [Sat, 6 Mar 2004 00:56:43 +0000 (00:56 +0000)]
Added reg to branchref structure for ArrayIndexOutOfBounds exceptions
twisti [Sat, 6 Mar 2004 00:45:23 +0000 (00:45 +0000)]
Pass register number of index register to codegen_addxboundrefs
twisti [Fri, 5 Mar 2004 23:49:31 +0000 (23:49 +0000)]
Display exceptions classname correctly.
twisti [Fri, 5 Mar 2004 23:49:12 +0000 (23:49 +0000)]
- utf_display_classname which converts / into . for classnames
- made code more readable
twisti [Fri, 5 Mar 2004 23:20:53 +0000 (23:20 +0000)]
- new_exception functions for dynamic exceptions
- some strings defined for code generation
jowenn [Fri, 5 Mar 2004 21:27:21 +0000 (21:27 +0000)]
better handling of for name
twisti [Fri, 5 Mar 2004 17:58:27 +0000 (17:58 +0000)]
Testcase for SecurityManager.getClassContext() and Class.forName()
jowenn [Fri, 5 Mar 2004 17:04:35 +0000 (17:04 +0000)]
use the old code on alpha for now
jowenn [Tue, 2 Mar 2004 21:18:23 +0000 (21:18 +0000)]
native functions, getClassContext, currentClassLoader implementation (with stack unwinding across native functions, as far as I know that is needed), test cases, usage of the gnu systemclassloader, storage of classloader in classinfo structure, ...
At the moment the stack unwinding is only implemented for i386, alpha will follow during the next days, so currently only i386 will compile and work. I'll do a cleanup of debug code and #if 0 blocks, once I have the alpha platform code in cvs too
twisti [Thu, 26 Feb 2004 00:20:02 +0000 (00:20 +0000)]
- don't use asm_builtin trace functions anymore
- use check_clinit for PUT/GETSTATIC
twisti [Thu, 26 Feb 2004 00:18:36 +0000 (00:18 +0000)]
Define functions prototypes
twisti [Thu, 26 Feb 2004 00:17:51 +0000 (00:17 +0000)]
Changed names of asm_calljava* functions
twisti [Thu, 26 Feb 2004 00:16:50 +0000 (00:16 +0000)]
- removed builtin trace functions
- implemented check_clinit
twisti [Thu, 26 Feb 2004 00:09:51 +0000 (00:09 +0000)]
Removed GC_DEBUG for x86_64
twisti [Thu, 26 Feb 2004 00:00:38 +0000 (00:00 +0000)]
- restructured
- exception handling
edwin [Tue, 24 Feb 2004 13:28:08 +0000 (13:28 +0000)]
fix bug when jit_compile is entered recursively
edwin [Tue, 24 Feb 2004 13:26:24 +0000 (13:26 +0000)]
avoid 'label at end of compound statement' warning
jowenn [Mon, 16 Feb 2004 04:00:59 +0000 (04:00 +0000)]
fix for Object[] oa = (Object[])java.lang.reflect.Array.newInstance(Runnable.class, 1);
jowenn [Mon, 16 Feb 2004 03:59:26 +0000 (03:59 +0000)]
testcase for twisti's Array.NewInstance problem
edwin [Sun, 8 Feb 2004 20:24:57 +0000 (20:24 +0000)]
changed savedstack implementation
edwin [Sun, 8 Feb 2004 19:54:52 +0000 (19:54 +0000)]
fixed exception typechecking
edwin [Sun, 8 Feb 2004 18:13:05 +0000 (18:13 +0000)]
restore input stack types changed by <init> call
edwin [Sun, 8 Feb 2004 15:17:10 +0000 (15:17 +0000)]
don't override private methods
edwin [Thu, 5 Feb 2004 23:13:19 +0000 (23:13 +0000)]
silently remove ACC_SUPER on interfaces
twisti [Thu, 5 Feb 2004 21:44:29 +0000 (21:44 +0000)]
Added java.security.ProtectionDomain, java.lang.VMString to header
generation.
twisti [Thu, 5 Feb 2004 21:23:19 +0000 (21:23 +0000)]
We have to add the string to cacao's constant pool when String.intern() is
called.
twisti [Thu, 5 Feb 2004 21:20:33 +0000 (21:20 +0000)]
Pass string offset to literalstring_u2, fixed to store them properly. Still
one thing left with copymode == false.
carolyn [Wed, 4 Feb 2004 11:42:41 +0000 (11:42 +0000)]
RT update for loader changes
twisti [Mon, 2 Feb 2004 23:09:49 +0000 (23:09 +0000)]
Safety first.