#1 $Id: ChangeLog 35854 2009-01-21 08:08:42Z chromatic $ 2009.01.20999 chromatic * Released 0.9.0 See NEWS for more. 2008.12.16999 Whiteknight * Released 0.8.2 See NEWS for more. 2008.11.18999 chromatic * Released 0.8.1 See NEWS for more. 2008.10.21999 particle * Released 0.8.0 See NEWS for more. 2008.09.16999 pmichaud * Released 0.7.1 See NEWS for more. 2008.08.19999 rgrjr * Released 0.7.0 See NEWS for more. 2008.07.15999 barney * Released 0.6.4 See NEWS for more. 2008.06.17999 smash * Released 0.6.3 See NEWS for more. 2008.05.20999 chromatic * Released 0.6.2 See NEWS for more. 2008.04.15999 particle * Released 0.6.1 See NEWS for more. 2008.03.18999 bernhard * Released 0.6.0 See NEWS for more. 2008.02.20999 pmichaud * Released 0.5.3 See NEWS for more. 2008.01.15999 rgrjr * Released 0.5.2 See NEWS for more. 2007.12.18999 jonathan * Released 0.5.1 See NEWS for more. 2007.11.20999 chromatic * Released 0.5.0 See NEWS for more. 2007.10.17999 coke * Released 0.4.17 See NEWS for more. 2007.09.18999 particle * Released 0.4.16 See NEWS for more. 2007.08.22999 pmichaud * Released 0.4.15 See NEWS for more. 2007.07.17999 particle * Released 0.4.14 See NEWS for more. 2007.06.19999 allison * Released 0.4.13 See NEWS for more. 2007.05.15999 chromatic * Released 0.4.12 See NEWS for more. 2007.04.17999 mdiep * Released 0.4.11 See NEWS for more. 2007.03.20999 coke * Released 0.4.10 See NEWS for more. 2007.3.10999 coke * removed parakeet 2007.02.22999 pmichaud * Released 0.4.9 See NEWS for more. 2007.01.17999 chromatic * changed 'name' method on namespaces to 'get_name' per pdd21 2007.01.16999 particle * Released 0.4.8 See NEWS for more. 2006.12.12999 particle * added C opcode 2006.11.14999 chip * Released 0.4.7 See NEWS for more. 2006.11.07999 tewk * added :init pragma functionality to subs 2006.10.28999 jonathan * added :vtable pragma functionality to subs 2006.09.18999 chromatic * added Parrot::Embed under ext/Parrot-Embed 2006.08.18999 leo * remove the None PMC - if needed it can trivially be reimplemented in languages from r14189 2006.08.16999 creiss Merged Software Transactional Memory branch into trunk, adds: * several new opcodes whose names start with stm_ * read-only PMC support * several new PMC types (whose names start with STM*) * support for threads (again) 2006.08.08999 chip * Released 0.4.6 See NEWS for more. 2006.08.02999 chip * opcodes: "in PMC" is no longer a synonym for "invar PMC"; there are opcodes now that want the real "in PMC". * namespace and global opcodes now accept name arrays again, as well as keys; the corresponding namespace PMC methods are gone. 2006.07.21999 leo * Parrot_exit, Parrot_on_exit take now an Interp * argument 2006.07.25999 chip * add experimental op find_type_i_p * fix longstanding bug that sometimes didn't encode pasm keys as Key PMCs (specifically "[S0]" and "['stringliteral']"). 2006.07.24999 chip further implementation of new pdd21: * new opcodes {get,set}*global (old {find,store}*global will be supported for a short while) 2006.07.20999 audreyt * --maintainer prerequesites are now bison 2.2+ and flex 2.5.3x (or some such --leo) (at the moment code is checking for 2.0 -- ambs) 2006.07.17999 chip partial implementation of new pdd21: * old: "get_namespace KEY". new: "get_hll_namespace KEY" * get_namespace and find_global opcodes no longer accept arrays of strings. To use arrays, call namespace object methods. 2006.06.19999 leo * Released 0.4.5 See NEWS for more. 2006.06.03999 leo * I and N registers are never cleared to 0 / 0.0 on subroutine entry 2006.05.14999 leo * Released 0.4.4 See NEWS for more. 2006.04.30999 bernhard * Make the Perl PMCs dynamic and move them to src/dynpmc 2006.04.02999 leo * Released 0.4.3 See NEWS for more. 2006.03.06999 leo * Shifted argument type bits left by 2 according to r11754 change of docs/pdds/pdd03_calling_conventions.pod * if you are creating PASM then code needs adjustment 2006.02.22999 leo * Released 0.4.2 See NEWS for more. 2006.01.16999 bernhard * rename src/dynclasses to src/dynpmc * rename src/classes to src/pmc 2006.01.15999 leo * find_global and friends returns a .Null PMC instead of .Undef on lookup failure, iff lookup exceptions are off 2006.01.08999 leo * Released 0.4.1 See NEWS for more. 2005.12.28999 particle * renamed all .imc files to .pir * modified all doc references to .pir as well 2005.12.13999 particle * moved jit/ directory to src/jit/ 2005.12.12999 particle * moved io/ directory to src/io/ * moved classes/ directory to src/classes/ * moved dynclasses/ directory to src/dynclasses/ 2005.12.08999 particle * moved types/ directory to src/types/ * moved pf/ directory to src/packfile/ * moved ops/ directory to src/ops/ 2005.12.08999 particle * moved charset/ directory to src/charset/ * moved encodings/ directory to src/encodings/ 2005.12.07999 particle * moved imcc/ directory to compilers/imcc/ 2005.12.06999 particle * moved util/ directory to tools/util/ * removed imcc/examples/ directory * moved ast/ directory to compilers/ast/ 2005.12.05999 particle * moved imcc/docs/ directory to docs/imcc/ * moved build_tools/ directory to tools/build/ * modified MANIFEST, MANIFEST.generated, and MANIFEST.SKIP 2005.12.04999 leo * Released 0.4.0 * See NEWS for changes 2005.11.06999 leo * Released 0.3.1 * See NEWS for changes 2005.10.01999 leo * Released 0.3.0 * See NEWS for changes 2005.08.05999 leo * Released 0.2.3 * See NEWS for changes 2005.07.03999 leo * Released 0.2.2 * See NEWS for changes 2005.06.04999 leo * Released 0.2.1 * allow short names for object attributes * string encoding and charset can now be set independently * experimental mmap IO layer for slurping files * enable compile opcode support for compilers written in PASM/PIR * improved HLL support with .HLL and n_operators pragmas * simplified hash code * more strength reduce optimizations * distinct debug and trace flag settings * glob support in PGE * new character classification opcodes and interfaces * numerous bug fixes and enhancements 2005-05-08999 leo * Released 0.2.0 * parrot repository is now under subversion * MMD (Multi Method Dispatch) enhanced * new unary and infix opcodes that return new result PMCs * dynamic scalar PMCs inherit now almost all from Parrot core PMCs * more unification of PMCs and ParrotObjects * tailcalls for functions and methods * PGE (Parrot Grammar Engine) reworked * Pugs creates Parrot code and Pugs is a registered compiler now * new languages/lisp * the usual bug fixes and improvements 2001-12-08 21:24 simon * Parrot/PackFile.pm, docs/parrot.pod, docs/parrotbyte.pod, docs/vtables.pod, lib/Parrot/PackFile.pm: Getting the documentation up to scratch. 2001-12-08 21:21 simon * classes/pmc2c.pl: Allow inheriting from default class. 2001-12-07 20:43 gregor * examples/assembly/queens.pasm: Regenerated the file with the latest jakoc. 2001-12-07 20:29 simon * core.ops, t/op/stacks.t: Rename "rotate" to "rotate_up". 2001-12-07 20:15 simon * t/op/stacks.t: Some tests for rotate. 2001-12-07 00:55 gregor * core.ops: Cast the result of the pointer arithmetic, since without the cast there are warnings with some compilers, and without the parens around the pointer arithmetic, precedence breaks the meaning and causes crashes. 2001-12-06 23:37 simon * core.ops: Temporary. I prefer compiler warnings to segfaults, thanks. 2001-12-06 23:34 simon * runops_cores.c: Make "trace" work again. 2001-12-06 21:23 dan * Parrot/Test.pm, lib/Parrot/Test.pm: Undo twiddles to test.pm 2001-12-06 21:22 dan * core.ops, packfile.c, Parrot/Test.pm, include/parrot/stacks.h, lib/Parrot/Test.pm: Fix up source issues that made Dec C unhappy 2001-12-06 19:22 simon * core.ops: Yet another "How the *%$! did that compile?" moment. 2001-12-06 18:28 dan * hints/vms.pl: Fix to build, with a HERE logical required 2001-12-06 17:48 dan * core.ops, interpreter.c, pbc2c.pl, pdump.c, test_main.c, docs/parrot_assembly.pod, include/parrot/interpreter.h: Sensible changes to the interpreter--flags are now part of the interp structure. Courtesy of "Melvin Smith" 2001-12-06 14:02 gregor * docs/intro.pod: s/little_languages/languages/ in Jako section. 2001-12-06 13:15 simon * Makefile.in: Ensure the documentation gets made. 2001-12-06 13:12 simon * MANIFEST, docs/intro.pod: Documentation! Added the "Some Assembly Required" article with minor updates. 2001-12-06 00:39 simon * core.ops: More "eq" support. Courtesy of: Nguon Hao Ching 2001-12-06 00:37 simon * test_main.c: Add the "-d" debug flag. Doesn't actually *do* anything yet. Courtesy of: Melvin 2001-12-06 00:11 tom * string.c, chartypes/unicode.c, chartypes/usascii.c, encodings/singlebyte.c, encodings/utf16.c, encodings/utf8.c, include/parrot/chartype.h, include/parrot/encoding.h, include/parrot/parrot.h: Moved string to number conversion code out of the encoding layer and into the main string layer as neither the encoding layer nor the character type layer knows enough on it's own to do the conversion correctly. 2001-12-05 23:12 tom * Makefile.in: Rebuild the classes directory if any of the object files for the classes are missing rather than only when perlint.o is missing. 2001-12-05 19:43 gregor * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: * Reworked the constantize_string() stuff a little bit because it was puking on things like "\033[H\033[2J\0" (which occurs in the languages/jako/mandelzoom.jako example). 2001-12-05 19:39 gregor * languages/jako/: Makefile, mandelzoom.jako: Added 'mandelzoom' example. 2001-12-05 19:38 gregor * languages/jako/jakoc: * Added sleep() * Changed some instances of die() to confess() 2001-12-05 18:34 gregor * t/: op/pmc.t, pmc/pmc.t: s/sting->int/string->int/ 2001-12-05 18:08 gregor * Parrot/.cvsignore: Ignore PMC.pm 2001-12-05 18:05 gregor * classes/.cvsignore: Ignore the generated files. 2001-12-05 17:19 gregor * pbc2c.pl: * Reverted recent changes, since they broke the script. Real fix coming later. 2001-12-05 17:03 simon * classes/Makefile.in: Remember: change Makefile.in, not Makefile. D'oh. 2001-12-05 17:02 simon * Makefile.in: Ensure PMC.pm gets remade. 2001-12-05 16:58 simon * MANIFEST, Makefile.in, classes/Makefile.in, classes/perlundef.pmc, include/parrot/pmc.h: Add perlundef class. 2001-12-05 16:45 simon * core.ops: Document "new P0, foo" 2001-12-05 16:43 simon * MANIFEST, Makefile.in, pbc2c.pl, pmc_pm.pl, Parrot/Assembler.pm, lib/Parrot/Assembler.pm, t/op/pmc.t, t/pmc/pmc.t: Allow assembler to know about PMC class names. 2001-12-04 13:39 ajgough * MANIFEST: Added t/op/pmc.t to MANIFEST. 2001-12-04 05:02 ajgough * classes/perlnum.pmc: Picky compiler pointed out probable thinko in perlnum.c 2001-12-04 04:24 ajgough * encodings/singlebyte.c, t/op/pmc.t, t/pmc/pmc.t: A much less broken form of strtod and appropriate tests. Does not yet check for overflow. 2001-12-04 01:31 ajgough * string.c, encodings/singlebyte.c, encodings/utf16.c, encodings/utf8.c, include/parrot/encoding.h: Added length parameter to extract_num and extract_int string vtables, without this it would be impossible to have strings with nulls in while still knowing where they ended. 2001-12-03 18:40 ajgough * string.c, classes/perlstring.pmc, encodings/singlebyte.c, encodings/utf16.c, encodings/utf8.c, include/parrot/encoding.h, include/parrot/string.h: Moved pmc string to number conversion into the string encoding tables. 2001-12-03 00:21 ajgough * core.ops, t/op/pmc.t, t/pmc/pmc.t: Added set i, p and some tests for p-> integer. 2001-12-02 23:09 jgoff * classes/Makefile.in: Removed some redundant targets. 2001-12-02 19:02 ajgough * t/: op/pmc.t, pmc/pmc.t: Modified all pmc tests to make use of fp equality macro rather than using perl to generate expected output, all tests now have < 2001-11-28 16:40 simon * classes/perlnum.pmc: De-C++ 2001-11-28 15:22 simon * string.c: Fixes Alex's string concat worry. 2001-11-28 14:27 ajgough * t/op/string.t: test "null onto null" incorrectly passing as it produced no output. Output provided. 2001-11-28 11:14 simon * include/parrot/pmc.h: Placate AIX. 2001-11-26 21:03 dan * MANIFEST, hints/darwin.pl: Tweaks to build on Mac OS X 10.1 2001-11-26 17:29 simon * classes/perlint.pmc: Various subtle but non-showstopper bugs. 2001-11-26 11:23 simon * classes/pmc2c.pl: Replace dependency on Text::Balanced Courtesy of: Bart Lateur 2001-11-26 05:22 jgoff * global_setup.c: Last file that should need to be committed. I swear. 2001-11-26 05:09 jgoff * core.ops: Let's try this again... Changes to core.ops required 2001-11-26 04:58 jgoff * classes/perlint.pmc, classes/perlnum.pmc, t/op/pmc.t, t/pmc/pmc.t: Makefile.in - Added missing cleanup line for perl*.c core.ops - Added set(p,n|nc), modified print(p) global_setup.c - Uncommented PerlNum_init classes/perlint.pmc - Added set_number functions classes/perlnum.pmc - Set up basic number functions t/op/pmc.t - Added tests for set P0. and {add,mul,sub,div} tests against numbers and numbers+integers 2001-11-25 13:18 simon * Makefile.in: Windows XP make doesn't like $< 2001-11-25 12:56 simon * MANIFEST, Makefile.in, classes/genclass.pl, classes/perlint.c, classes/perlint.pmc, classes/perlnum.c, classes/perlnum.pmc, classes/perlstring.c, classes/perlstring.pmc, classes/pmc2c.pl, examples/assembly/mops_p.pasm: PMC preprocessor. Courtesy of: Angel Faus I (Simon) also added a couple more integer methods, but no tests since I'm a Bad Person. 2001-11-25 08:40 brentdax * ops2c.pl, Parrot/Op.pm, Parrot/OpsFile.pm, lib/Parrot/Op.pm, lib/Parrot/OpsFile.pm: Allow _ in op names 2001-11-25 05:54 jgoff * classes/perlint.c, t/op/pmc.t, t/pmc/pmc.t: classes/perlint.c - Added basic math ops t/op/pmc.t - Added tests of basic math ops 2001-11-23 18:27 simon * core.ops, classes/perlint.c, examples/assembly/mops_p.pasm: Add everything you need to make MOPS use PMCs. 2001-11-23 18:06 simon * core.ops: Eh, you'll need this for rewriting mops.pasm. 2001-11-23 17:55 simon * core.ops, classes/perlint.c, include/parrot/pmc.h, t/op/pmc.t, t/pmc/pmc.t: More work on integer PMCs. 2001-11-23 17:29 simon * Makefile.in: Surprise! Whitespace! HLAGHLAGHLAGH 2001-11-23 17:28 simon * Makefile.in, ops2c.pl, ops2pm.pl: Add vtables.ops to the core Courtesy of: Jeff Goff 2001-11-22 18:39 simon * core.ops, global_setup.c, docs/parrot_assembly.pod, t/op/pmc.t, t/pmc/pmc.t: We can now create PMCs. 2001-11-21 05:49 jgoff * Makefile.in: Makefile cleans up *.pasm *.pbc *.out 2001-11-20 16:00 simon * docs/vtables.pod, include/parrot/pmc.h: Reorder PMC flags so there's a fixed number of user-defined flags. Courtesy of: Dave Mitchell 2001-11-19 03:39 jgoff * Parrot/Assembler.pm, lib/Parrot/Assembler.pm, t/op/basic.t: Parrot/Assembler.pm - Allow "print 'foo'", and properly escape nested quotes. t/op/basic.t - Add tests for single-quoted strings, escaped quotes, and proper handling of newlines. 2001-11-16 22:03 simon * classes/perlint.c: More work on the Perl Integer class. More needed, though. 2001-11-16 21:57 simon * pmc.c, vtable.tbl, classes/genclass.pl, classes/perlint.c: Rename ->new to ->init, and rework pmc.c:pmc_new to fiddle it. 2001-11-16 20:49 simon * interpreter.c, runops_cores.c, trace.c, include/parrot/runops_cores.h: Kill superfluous runops cores. 2001-11-16 17:50 simon * MANIFEST, Makefile.in, classes/intclass.c, classes/perlint.c, classes/perlnum.c, classes/perlstring.c, classes/scalarclass.c, include/parrot/parrot.h, include/parrot/pmc.h: Skeleton for Perl scalar types, done right. 2001-11-16 16:15 simon * classes/genclass.pl: This tidies up genclass to deal with the new-style multimethods. 2001-11-16 15:29 simon * Makefile.in, make_vtable_ops.pl, vtable.tbl, Parrot/Vtable.pm, lib/Parrot/Vtable.pm, include/parrot/parrot.h: Rework multimethods to be *true* multimethods - that is, have different types. 2001-11-15 22:29 simon * MANIFEST, Makefile.in, pmc.c: pmc.c linkage and support. Oh, and making it compile. 2001-11-15 22:26 dan * t/op/interp.t: Forgot to tell CVS about the test... 2001-11-15 22:10 dan * MANIFEST, Types_pm.in, core.ops, vtable_h.pl, docs/parrot_assembly.pod, include/parrot/string.h: Added in support for PMC registers for the interpreter. Added in the docs for the runinterp opcode. Added in code for both the newinterp and runinterp opcodes, so we can now create new interpreters from within parrot and have them do things. (Though we can't pass data back and forth. Yet) 2001-11-15 21:24 simon * vtable.tbl, classes/scalarclass.c, docs/vtables.pod: More documentation, and other changes as discussed with Dan. 2001-11-15 19:51 simon * classes/scalarclass.c: De-cut-and-pasto. 2001-11-15 19:50 simon * classes/scalarclass.c: A little code tidiness. 2001-11-15 19:33 simon * include/parrot/pmc.h: Missed this in the previous commit - un-constantize the base for user-defined PMC flags. 2001-11-15 19:28 simon * classes/scalarclass.c: Hey look, sanity. 2001-11-15 17:09 simon * classes/scalarclass.c: It seems I have completely forgotten how to write C. 2001-11-15 16:53 simon * docs/vtables.pod: Start of the PMC classes documentation. 2001-11-15 14:46 simon * classes/scalarclass.c: Changing to use a Perl-style struct instead of the union cache. 2001-11-13 23:56 tom * string.c: Removed leading _ from string_index as function names starting with _ are reserved to implementors by the C standard. Also fixed string_index to use the encoding routines instead of assuming a single byte encoding. 2001-11-13 12:59 jgoff * string.c: Patched string_concat to accept null strings. That test was failing. Nothing to do with ord(). 2001-11-13 12:43 jgoff * string.c: string_ord function, string_concat fixes 2001-11-13 04:52 jgoff * core.ops: Added ord to core.ops 2001-11-13 04:35 jgoff * docs/strings.pod, include/parrot/exceptions.h, include/parrot/string.h, t/op/string.t: core.ops - Added ord operator, documentation string.c - Added string_ord, _string_index (_string_index is a static helper) Patched string_concat to handle null strings properly. docs/strings.pod - Description of string_ord include/parrot/exceptions.h - Added ORD_OUT_OF_STRING exception include/parrot/string.h - Added string_ord() t/op/string.t - Added full set of test for ord_i_(s|sc)(_(i|ic))? string.c contains a bit of documentation explaining the logic behind _string_index, primarily having to do with multiple encoding formats. 2001-11-12 22:30 simon * t/op/macro.t: Count 'em. 2001-11-12 22:28 simon * Makefile.in: Add a rule to compile "platform" 2001-11-12 22:26 simon * Makefile.in: Tidy up distclean target. 2001-11-12 16:54 dan * MANIFEST, Parrot/Assembler.pm, lib/Parrot/Assembler.pm, t/op/macro.t: Fixes to the assembler's macro facility and a test Courtesy of: Jeff 2001-11-12 16:50 dan * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: Allow input from stdin Courtesy of: Jeff 2001-11-12 16:36 dan * Parrot/Test.pm, lib/Parrot/Test.pm: Fix to make test work on CygWin Courtesy of: Alex Gough 2001-11-12 11:22 simon * examples/assembly/mops.pasm: Trivial tightening of mops loop. 2001-11-07 23:57 ajgough * docs/tests.pod: This is a basic quide to writing tests with assembler. 2001-11-07 22:29 dan * stacks.c: Added a little documentation on the cleanup parameter for an entry 2001-11-07 14:43 dan * MANIFEST, platforms/generic.c, platforms/linux.c, platforms/linux.h: Portability patches. Courtesy of: Andy Dougherty 2001-11-06 19:51 tom * Makefile.in: Clean up object files in the subdirectories as well as those in the top level directory. Enhanced version of patch from Daniel Grunblatt 2001-11-06 16:14 dan * Configure.pl, MANIFEST, Makefile.in, core.ops, platforms/generic.c, platforms/generic.h: Made things a bit more platform-independent. Makefile still needs work to pick up changes to the platform-specific files. 2001-11-05 21:32 dan * docs/parrot_assembly.pod: Added docs for the GC control ops 2001-11-04 02:52 gregor * platforms/: win32.c, win32.h: * win32 dynamic loading functions. Courtesy of Jason Diamond 2001-11-04 00:04 brentdax * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: Assembler fixes: allows an absolute path to be used and gives a useful error message if the original source file can't be found 2001-11-02 12:37 gregor * platforms/: linux.c, linux.h: * Beginning of dynamic loading support. Linux only for now. 2001-11-02 12:13 gregor * MANIFEST, time.c, include/parrot/time.h: * Removed time.[hc], since they have been superseded by the new platform stuff. 2001-11-02 12:11 gregor * Configure.pl, Makefile.in, core.ops, vtable.tbl, classes/intclass.c, classes/scalarclass.c, hints/mswin32.pl, include/parrot/parrot.h, platforms/linux.c, platforms/linux.h, platforms/win32.c, platforms/win32.h: Patch from Jason Diamond to make win32 platform stuff work: * Modified Configure.pl, mswin32.pl, and Makefile.in so that the platform specific files in the platforms directory gets copied to the correct directories. The Makefile will re-copy the platform.h and .c files if the original is modified. * #include platform.h in parrot.h instead of time.h. * Removed time.h and time.obj from the Makefile. time.h and time.c aren't being used anymore since that code was moved to linux.c and win32.c. * Renamed the BOOL typedef in parrot.h to BOOLVAL (inspired by INTVAL and FLOATVAL). This was conflicting with the BOOL typedef in windows.h. This required updating several files besides parrot.h. * Added a DEFAULT_OPEN_MODE to linux.h and win32.h and used that instead of the missing identifiers (on Windows) in core.ops in the calls to open. * Added Parrot_sleep and Parrot_setenv to both the linux and win32 platform files and called those from core.ops. The Linux code for these functions came from core.ops so they should hopefully work. Courtesy of Jason Diamond 2001-11-02 06:43 brentdax * Makefile.in, hints/mswin32.pl: Fix classes/intclass.obj building for Win32 (courtesy of Jason Diamond ) 2001-11-02 02:26 gregor * .cvsignore, MANIFEST, include/parrot/.cvsignore, platforms/linux.c, platforms/linux.h, platforms/win32.c, platforms/win32.h: * Beginnings of platform support files in platform/*.[hc] 2001-11-01 23:37 tom * classes/intclass.c: Added missing *'s to STRING arguments and return values to avoid compiler warnings. 2001-11-01 22:21 gregor * MANIFEST, Makefile.in, core.ops, obscure.ops, time.c, include/parrot/parrot.h, include/parrot/time.h: Initial "fix" of time(n) op for Win32: * Initial Win32 code Courtesy of Richard J Cox * time.c and include/parrot/time.h First attempt at getting this working across platforms. The macro WIN32 being tested might not be right. * New functions Parrot_intval_time() and Parrot_floatval_time() wrap the appropriate functionality depending upon platform. Changed core.ops to call these. Added obscure.ops with obscure transcendental ops. Added simple file I/O ops (close, open, read, write): * Initial versions Courtesy of Jeff * I changed some arg ordering stuff. * I added read/write of INTVALs and FLOATVALs. Added environment variable ops (getenv, setenv) Added system error ops (err(i) and err(s) for accessing errno and strerr(). Moved print ops in with other "temporary" developer ops, since they'll move to a devel.ops or temp.ops oplib once we have oplibs. 2001-10-31 22:51 tom * MANIFEST, Makefile.in, chartype.c, encoding.c, global_setup.c, packfile.c, string.c, strnative.c, strutf16.c, strutf32.c, strutf8.c, transcode.c, chartypes/unicode.c, chartypes/usascii.c, encodings/singlebyte.c, encodings/utf16.c, encodings/utf32.c, encodings/utf8.c, include/parrot/chartype.h, include/parrot/encoding.h, include/parrot/exceptions.h, include/parrot/parrot.h, include/parrot/string.h, include/parrot/strnative.h, include/parrot/strutf16.h, include/parrot/strutf32.h, include/parrot/strutf8.h, include/parrot/transcode.h: Reworked the string code in line with Dan's latest design that separates character sets from encodings. 2001-10-31 19:53 dan * docs/parrot_assembly.pod: Finished I/O section of docs 2001-10-31 16:51 dan * core.ops, docs/parrot_assembly.pod, include/parrot/parrot.h: Moved includes out of core.ops and into parrot.h where they belong (suitably guarded with #ifdefs). Also some updates to parrot_assembly.pod 2001-10-30 21:43 dan * core.ops, stacks.c, docs/parrot_assembly.pod, include/parrot/stacks.h: Added entrytype opcode to fetch the type of the entry on the stack 2001-10-29 18:30 dan * stacks.c: You can turn off stack entry type validation on pop by passing in a type of 0 now. (For those cases where you just want to throw the darned thing away) 2001-10-29 14:37 gregor * core.ops, languages/jako/jakoc: * New variants of the save ops to save constants on the stack. This is useful for saving zero arg and result counts on the stack, for example. * Modified jakoc to to adapt to the newly announced calling conventions. Even though Jako source code appears to show subroutines with fixed numbers of arguments and results, they are implemented with the variable count calling convention. Future modifications will be needed to ensure that an exception is raised if the stack frame is of the wrong size (or, at least that we pop the complete frame off). We'll also have to make more changes to support the true fixed-arg-count and fixed-result-count parts of the calling convention spec. 2001-10-28 16:53 simon * MANIFEST, stacks.c, classes/scalarclass.c, include/parrot/stacks.h: Unbelievably bad errors of mine, plus a type incompatibility of Dan's. 2001-10-28 08:30 simon * classes/scalarclass.c: Beginning of a scalar class. 2001-10-28 08:29 simon * make_vtable_ops.pl, vtable.tbl, Parrot/Vtable.pm, lib/Parrot/Vtable.pm, classes/genclass.pl, classes/intclass.c, include/parrot/pmc.h: This fixes the problem with vtables and pointer arithmetic. 2001-10-26 19:58 dan * core.ops, interpreter.c, stacks.c, include/parrot/interpreter.h, include/parrot/stacks.h: Split the generic stack into a user push stack and a control stack 2001-10-26 02:59 bdwheele * Parrot/OpsFile.pm, lib/Parrot/OpsFile.pm: [no log message] 2001-10-24 20:27 dan * languages/scheme/: Makefile, Scheme.pm, schemec, Scheme/Generator.pm, Scheme/Parser.pm, Scheme/Test.pm, Scheme/Tokenizer.pm, t/harness, t/arith/basic.t, t/arith/logic.t, t/arith/nested.t, t/io/basic.t, t/logic/basic.t: Added the mini Scheme compiler 2001-10-24 20:23 dan * MANIFEST, pdump.c, docs/parrot_assembly.pod: Update for interpreter needs in pdump.c. Added scheme bits to manifest 2001-10-24 16:43 gregor * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: * Make the assembler handle octal character code escapes in string constants. 2001-10-24 15:56 dan * MANIFEST, examples/assembly/life.pasm: Added an example in to play the game of life 2001-10-24 15:54 dan * core.ops: Added sleep. We'll need to override this to do events properly anyway, so... 2001-10-24 14:03 gregor * Makefile.in, pbc2c.pl, pdump.c: * Adapt pdump.c and pbc2c.pl to the new requirement to pass interpreter pointer into string functions. Courtesy of: Ryan O'Neil 2001-10-23 20:09 gregor * languages/jako/jakoc: * Terminological refinement in regexp capture comment. 2001-10-23 15:23 simon * classes/genclass.pl: Use a saner ANSI C construct to make MIPSPro happy. Courtesy of: Alex Gough 2001-10-23 00:34 tom * core.ops, string.c, strnative.c, strutf16.c, strutf32.c, strutf8.c, include/parrot/string.h: Make sure the interpreter structure is passed to the transcoder as necessary, including passing it to string_compare(). 2001-10-23 00:32 tom * Makefile.in: Make pdump build again by including resources.o in the link. 2001-10-22 22:43 dan * core.ops, packfile.c, string.c, strnative.c, strutf16.c, strutf32.c, strutf8.c, test_main.c, transcode.c, include/parrot/packfile.h, include/parrot/parrot.h, include/parrot/resources.h, include/parrot/string.h, include/parrot/transcode.h: Everything now uses the allocators and deallocators in resources.c. Unfortunately we lost standalone string-test mode, but I can cope with that. Please note that until we have a real allocation/deallocation/GC system, this'll leak. A *lot*. 2001-10-22 20:27 dan * MANIFEST, Makefile.in, pmc.c, resources.c, include/parrot/resources.h: First cut to handle arena allocation of PMC headers. (Amongst other things) 2001-10-22 19:02 dan * Configure.pl, config_h.in, register.c, include/parrot/register.h: Now each register type has its own mask macro, which is the right thing to do. Patch courtesy of Jason Gloudon 2001-10-21 23:40 tom * stacks.c, t/op/stacks.t: Fix bug in handling of rollover to a new stack chunk and add a test to test this rollover in future. Courtesy of Jason Gloudon . 2001-10-21 19:14 dan * include/parrot/string.h: Changed string 'unused' field to the language field. 2001-10-21 18:18 gregor * languages/jako/: Makefile, mops.jako: * Makefile: 'clean' target removes .list files, too. And, it does rm -f with wildcards rather than listing things explicitly. * mops.jako actually prints out the millions-of-operations-per-second (mops) value instead of just the operations-per-second value. 2001-10-21 18:03 simon * classes/intclass.o: *smack*! WRONG! 2001-10-21 17:47 simon * MANIFEST, Makefile.in, make_vtable_ops.pl, pmc.c, vtable_h.pl, Parrot/Vtable.pm, classes/genclass.pl, classes/intclass.c, classes/intclass.o, include/parrot/parrot.h, include/parrot/pmc.h, lib/Parrot/Vtable.pm: Initial work on PMCs. Explanation to follow 2001-10-21 15:20 gregor * examples/assembly/queens.pasm: * Modernize the calling conventions in queens.pasm. It was using swap, which now must be rotate. Also, changed to use return values on the stack instead of in I0. 2001-10-19 18:14 gregor * core.ops, stacks.c, include/parrot/stacks.h, languages/jako/jakoc: Per Dan's request, replacing 'swap' op by 'rotate'. * core.ops: Replaced 'swap()' by 'rotate(i|ic)'. * stacks.c and stacks.h: Removed swap-related functions from my prior commit, replacing them with stack_depth(), stack_entry() and rotate_entries(). NOTE: The stack_depth() and rotate_entries() routines could be made more efficient through the caching of the sum of chunk 'used' members from lower chunks in a field in the upper chunks. Also, if the interpreter's stack_base was changed to hold the pointer to the *top* chunk (since we usually come at the stack from the top down) instead of the *bottom* chunk. The code here at least works, and we can decide about bigger changes separately. For now, rotate only handles rotates in one direction and of magnituded one. A fully general rotate would do handle rotates in both directions and of arbitrary amounts. Again, this implementation is sufficient for current applications, and a feature-complete version can be developed if this turns out to be otherwise the right type of implementation. This change includes a new exception for ERROR_STACK_SHALLOW, meaning "not-empty-yet-not-deep-enough-for-what-you-asked". * jakoc: Changed the calling convention to use rotate instead of swap (of course). Also changed the convention for return value passing to use the stack as well. 2001-10-19 13:36 gregor * MANIFEST, core.ops, disassemble.pl, stacks.c, examples/assembly/queens.pasm, include/parrot/stacks.h, languages/jako/.cvsignore, languages/jako/Makefile, languages/jako/bench.jako, languages/jako/euclid.jako, languages/jako/fact.jako, languages/jako/jakoc, languages/jako/mandelbrot.jako, languages/jako/mops.jako, languages/jako/primes.jako, languages/jako/queens.jako, languages/jako/sub.jako: * New stack functions push_entry(), pop_entry() and swap_entry(). The first two exist for the support of the last one. These functions don't care what is in the stack entries. These exist to support the new 'swap' op (see next bullet). * New op 'swap' for swapping the top two elements on the stack. This is necessary for passing arguments to functions, since the return address will be on top of any arguments that have been pushed on the stack. This way, you can use a (swap, pop) pair to retrieve each argument (for subroutine linkage). * Disassembler now prints numbers in decimal by default, with a new '-x' flag to turn on hexadecimal display. Since other members of the Parrot Tool Set use decimal by default, this makes it easier to compare output from the various tools. * Jako compiler has more explanatory comments and the following new features: * Lexically scoped variables. * Subroutines that take arguments and return values: sub int foo(int bar) { if (bar == 1) { return 42; } else { return 13; } } * ... if (...) and ... unless (...) statement modifiers can be used on next, last and redo. * Bitwise and (&, &=) and or (|, |=) ops. * Type names are just 'int', 'num', 'str', and 'poly' now. The old long names 'integer', 'number', 'string', and 'polytype' are no longer valid. * Integer variable assignments can now use 0b (binary), 0 (octal) and 0x (hexadecimal) radix prefixes on literals. * New Jako and Parrot assembly examples queens.{jako,pasm} for solving the eight-queens problem. The assembly example is a hand-modified version of the assembly code produced by the Jako compiler from queens.jako. A few unnecessary ops have been removed, lots of unnecessary labels have been removed, and all the Jako source code comments have been removed. * Makefile and .cvsignore in languages/jako modified so that the assembler produces listings when assembling the programs, yet the *.list files are ignored by CVS. * Made the default target in languages/jako/Makefile 'all'. * Modified some of the Jako examples to use some of the new features of Jako, such as subroutines and lexical variable scoping. 2001-10-19 02:43 gregor * runops_cores.c, include/parrot/runops_cores.h: Bug fix (gcc doesn't complain, but MIPSpro does): * Added missing extern to runops_cores array declaration in include/parrot/runops_cores.h * Also added 'const' there and in runops_cores.c 2001-10-18 15:18 gregor * .cvsignore, include/parrot/oplib/.cvsignore: * Ignore the vtable-related generated files. 2001-10-18 15:14 gregor * core.ops: * left and right shifts now allow the 2nd argument to be a constant. 2001-10-18 14:55 simon * make_vtable_ops.pl: This is C, not Perl. How on earth did that even compile? 2001-10-18 14:54 simon * make_vtable_ops.pl: More vtable ops. 2001-10-18 12:33 simon * Makefile.in, make_vtable_ops.pl, vtable.tbl: Initial stab at vtable opcode creation stuff. 2001-10-18 02:15 gregor * MANIFEST, Makefile.in, interpreter.c, runops_cores.c, trace.c, include/parrot/interpreter.h, include/parrot/runops_cores.h, include/parrot/trace.h: Preparation for autogenerated runops cores: * Split tracing and runops core stuff out of interpreter.c into trace.c and runops_cores.c * Similar split of stuff from include/parrot/interpreter.h * Manifest and Makefile.in updates to make things still compile. 2001-10-18 01:21 gregor * Makefile.in, ops2c.pl, ops2pm.pl, pbc2c.pl: Bug fixes: * Makefile.in: examples/assembly/mops.c depends on pbc2c.pl * Makefile.in: Parrot/OpLib/core.pm depends on ops2pm.pl * ops2c.pl: Added second arg to mkdir (perl 5.005_02 complains) * ops2pm.pl: Added second arg to mkdir (perl 5.005_02 complains) * pbc2c.pl: Removed explicit include of , replacing it with copy of $ops->preamble. 2001-10-17 22:32 bdwheele * interpreter.c: Moved op dump code into trace_op_dump, per Dan's request. 2001-10-17 22:08 gregor * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: * Now handles label-and-comment-only lines. 2001-10-17 22:04 bdwheele * interpreter.c: added more comprehensive trace output 2001-10-17 16:50 tom * Makefile.in: Removed the explicit rule for building mops.o from mops.c so that the implicit rule for compiling C files is used instead. This means that the chosen C flags are used, including any optimisation options. 2001-10-17 15:24 tom * core.ops, Parrot/OpsFile.pm, lib/Parrot/OpsFile.pm: Enhanced the op explosion logic to be able to handle exploding this: sub(i, i|ic, i|ic) into these: sub(i, i, i) sub(i, i, ic) sub(i, ic, i) without also generating this unwanted version: sub(i, ic, ic) The enhanced version rejects any expansion where there is more than one expanded argument and all the expanded arguments are constants. 2001-10-17 14:20 simon * Makefile.in, ops2c.pl, ops2pm.pl, examples/assembly/Makefile: Build fixes to use correct Perl version and ensure that things are running in the right directory. Courtesy of: Andy Dougherty 2001-10-17 13:16 simon * Makefile.in: core.pm is more critical than that. 2001-10-17 13:05 simon * core.ops, Parrot/OpsFile.pm, lib/Parrot/OpsFile.pm: "Explode" ops into multiple versions for different parameters. Courtesy of: Tom Hughes 2001-10-17 12:54 simon * Makefile.in: Fix dependencies and cleaning. 2001-10-17 12:45 simon * Configure.pl, Makefile.in: Fixes to make Parrot compile without the temporary build hack Courtesy of: Jason Gloudon 2001-10-16 20:43 gregor * Makefile.in: * Fixed dependencies for building Parrot/OpLib/core.pm 2001-10-16 19:35 bdwheele * core.ops, t/op/string.t: Changed set s,sc to use string_copy & added test, thanks to Alex Gough 2001-10-16 16:49 bdwheele * core.ops: added opcode variants for constants and 3 arg concat 2001-10-16 15:12 bdwheele * core.ops: added "set Sx,Sy" as a string copy to match the other register-to-register copies available 2001-10-16 13:09 gregor * core.ops: * Added #include in the preamble for gettimeofday(). 2001-10-16 13:09 gregor * Makefile.in: * Removed cd t; make, which no longer applies 2001-10-16 03:18 bdwheele * core.ops: fix save & restore operations to push correct type onto stack. [good eye to Jason Gloudon for noticing this] 2001-10-16 01:38 bdwheele * core.ops, register.c, examples/assembly/fact.pasm, include/parrot/register.h: send in the clones: clone routines for parrot + example program 2001-10-15 22:37 gregor * MANIFEST, Makefile.in, examples/assembly/.cvsignore, examples/assembly/Makefile, examples/assembly/bsr.pasm, examples/assembly/call.pasm, examples/assembly/euclid.pasm, examples/assembly/jump.pasm, examples/assembly/local_label.pasm, examples/assembly/mops.pasm, examples/assembly/stack.pasm, examples/assembly/substr.pasm, examples/assembly/trace.pasm, t/.cvsignore, t/Makefile, t/call.pasm, t/euclid.pasm, t/jump.pasm, t/local_label.pasm, t/test.pasm, t/test2.pasm, t/test3.pasm, t/test4.pasm, t/test5.pasm, t/trace.pasm: * Moved examples from t/*.pasm to examples/assembly/*.pasm, with some renames so the names match what they demonstrate. 2001-10-15 22:18 gregor * core.ops, t/test.pasm: * time_n uses gettimeofday() to get sub-second times. * t/test.pasm uses time_n directly rather than doing int stuff and converting to num later. 2001-10-15 12:28 gregor * t/: test.pasm, test2.pasm, trace.pasm: * t/test.pasm prints using string constants instead of string regs. * t/test2.pasm assembles (using implicit op modes) * t/trace.pasm updated for new behavior of trace_ic, which does not require an 'end' after 'trace'. 2001-10-15 00:47 gregor * core.ops, interpreter.c, test_main.c, include/parrot/interpreter.h, t/test.pasm: * New interpreter flag PARROT_PROFILE_FLAG * New runops cores to handle profiling * Support for -p flag in test_prog to enable profiling * New profile_ic op to control profiling at run-time 2001-10-14 21:54 gregor * core.ops, Parrot/OpsFile.pm, docs/.cvsignore, docs/Makefile, lib/Parrot/OpsFile.pm: Integrated op documentation. * core.ops now contains POD documentation of ops, interspersed with the ops themselves * Ops are organized into POD =head2 sections * Parrot/OpsFile.pm now treats the preamble of the ops file as anything up to the first POD directive OR op definition. * docs/Makefile now generates core_ops.pod from core.ops vi perldoc -u. 2001-10-14 11:58 simon * Makefile.in: Unbreak Digital Un^W^WTru64 2001-10-14 11:00 simon * Configure.pl: Remove weird typo. 2001-10-14 10:56 tom * Configure.pl: Stop Configure trying to make op_info.c and op.h as op_info.c no longer exists and op.h is no longer a generated file. 2001-10-14 10:54 tom * Makefile.in: Added vtable.h to the list of things removed by the clean target. 2001-10-14 01:43 gregor * .cvsignore, MANIFEST, Makefile.in, basic_opcodes.ops, build_interp_starter.pl, core.ops, disassemble.pl, interpreter.c, make_op_header.pl, memory.c, opcheck.pl, opcode_table, ops2c.pl, ops2pm.pl, packfile.c, pbc2c.pl, process_opfunc.pl, Parrot/Assembler.pm, Parrot/Op.pm, Parrot/OpsFile.pm, Parrot/OpLib/.cvsignore, lib/Parrot/Assembler.pm, lib/Parrot/Op.pm, lib/Parrot/OpsFile.pm, lib/Parrot/OpLib/.cvsignore, include/parrot/.cvsignore, include/parrot/interp_guts.h, include/parrot/interpreter.h, include/parrot/op.h, include/parrot/oplib/.cvsignore, t/.cvsignore, t/Makefile, t/euclid.pasm, t/test.pasm, t/op/integer.t, t/op/number.t, t/op/stacks.t, t/op/string.t, t/op/trans.t: Moby Patch: * basic_opcodes.ops and opcode_table replaced by core.ops * process_opfunc.pl replaced by Parrot/OpsFile.pm, Parrot/Op.pm, and ops2c.pl. These convert *.ops to *_ops.c and include/parrot/oplib/*_ops.h * New ops2pm.pl converts *.ops to Parrot/OpLib/*.pm * disassemble.pl modified to use * interpreter.c modified to use structures in core_ops.c instead of the old op_info and op_func stuff. We're not yet to the point of handling multiple oplibs at a time, but we are getting there... Fingerprinting stuff disabled for now. We used to go off opcode_table. We need to rethink this. I have some ideas to make it unnecessary, but it remains to be seen whether others will like the direction they take us. Termination condition for while(){} loop in runops is now (pc) instead of (*pc). Any op can cause a HALT by returning a NULL pointer. Ops end, bounds and trace all do this. So, now the 'end' op body is actually executed, although all it does is HALT. * Makefile.in modified to perform a test build using pbc2c.pl, which is the bytecode --> C compiler I've posted to the list before. * Op full names are now autogenerated. Op root names must not contain underscores, so some op renamings have happened. Such as clear_i becoming cleari. 2001-10-13 11:49 simon * Configure.pl, MANIFEST: Make parrot actually build again. 2001-10-12 21:34 dan * docs/parrot_assembly.pod: Added docs for the new save/restore/bsr opcodes 2001-10-12 20:56 dan * MANIFEST, Makefile.in, basic_opcodes.ops, interpreter.c, opcode_table, process_opfunc.pl, stacks.c, include/parrot/stacks.h, t/test3.pasm, t/test4.pasm, t/test5.pasm: Changes to support generic stack now in. We also have subroutines, of a simple sort! 2001-10-12 19:15 tom * t/op/string.t: Improved string comparison tests try all string combinations with each of the comparison ops. 2001-10-12 19:09 dan * interpreter.c: Properly sets up generic stack now 2001-10-12 19:09 dan * include/parrot/interpreter.h: Fixed stack chunk struct type in interpreter structure 2001-10-12 18:59 dan * stacks.c, include/parrot/parrot.h, include/parrot/stacks.h: More changes to get the generic stack working 2001-10-12 18:59 dan * Configure.pl, MANIFEST, config_h.in, testparrotsizes_c.in, hints/vms.pl: Made the changes to figure out mask bits for various register and stack chunks 2001-10-11 22:09 dan * docs/parrot_assembly.pod: Tossed the parameter for return 2001-10-11 19:28 tom * strnative.c: Assignment, not comparison. (Plus formatted for coding standards) Courtesy of Bryan C. Warnock 2001-10-10 19:21 tom * basic_opcodes.ops, opcode_table, string.c, strnative.c, strutf16.c, strutf32.c, strutf8.c, docs/strings.pod, include/parrot/string.h, t/op/string.t: Added support for string comparisons and extended the comparison ops to include comparison of strings. 2001-10-10 17:58 dan * MANIFEST: Shouldn't have added vtable.h to the manifest... 2001-10-10 17:54 dan * MANIFEST, stacks.c, vtable_h.pl, Parrot/Vtable.pm, lib/Parrot/Vtable.pm: Now spits out the #defines for the various vtable type entries properly 2001-10-10 15:37 dan * Parrot/Vtable.pm, lib/Parrot/Vtable.pm: Re-enabled change to parse new vtable.tbl file now that I've actually checked *in* the new format vtable.tbl file... 2001-10-10 15:32 dan * vtable.tbl: New Column to indicate the type of the vtabel entry so we can know how many slots it takes 2001-10-10 05:26 thgibbs * Parrot/Vtable.pm, lib/Parrot/Vtable.pm, vtable_h.pl: Patch to allow parrot to compile. PACKAGE is not defined, therefore, I added a struct before it to have it not complain until we define PACKAGE. Also, there was an extra shift in Vtable.pm and a missing semicolon at the end of the struct declaration. 2001-10-10 05:09 thgibbs * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: This patch makes string constants of the form U"string" become unicode encoded parrot strings and the form N"string" become native encoded parrot strings. Courtesy of: Tom Hughes 2001-10-10 04:55 dan * vtable_h.pl, Parrot/Vtable.pm, lib/Parrot/Vtable.pm: Start of support for generating the vtable function table offset macros 2001-10-10 01:05 simon * t/op/integer.t: The test for cmod_i was using the 'mod Ix, Ix, Ix' form which compiles into the mod op, not the cmod_i op, hence wasn't testing what it thought it was. Courtesy of: Alex Gough 2001-10-09 03:45 thgibbs * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: Patch to allow assembler to generate unicode strings. Courtesy of: Tom Hughes 2001-10-08 19:30 brentdax * Configure.pl, test_c.in, hints/vms.pl: Cleaner VMS support, more info for Parrot::Config to remember, more flexible test.c stuff 2001-10-08 15:04 gregor * basic_opcodes.ops: * Removed trash from early mod_i experiments (pardon our dust) 2001-10-08 14:56 gregor * basic_opcodes.ops: * Fixed a couple of typos. 2001-10-08 14:45 gregor * basic_opcodes.ops, opcode_table, t/op/integer.t, t/op/number.t: Arithmetic enhancements: * Renamed existing mod_i as cmod_i and added "corrected" mod_i: NOTE: This "uncorrected mod" algorithm uses the C language's built-in mod operator (x % y), which is ... the remainder when x is divided by y, and thus is zero when y divides x exactly. ... The direction of truncation for / and teh sign of the result for % are machine-dependent for negative operands, as is the action taken on overflow or underflow. -- [1], page 41 Also: ... if the second operand is 0, the result is undefined. Otherwise, it is always true that (a/b)*b + a%b is equal to z. I both operands are non-negative, then teh remainder is non-negative and smaller than the divisor; if not, it is guaranteed only that the absolute value of the remainder is smaller than the absolute value of the divisor. -- [1], page 205 This op is provided for those who need it (such as speed-sensitive applications with heavy use of mod, but using it only with positive arguments), but a more mathematically useful numeric mod based on floor(x/y) and defined with y == 0 is provided by the mod_i op. [1] Brian W. Kernighan and Dennis M. Ritchie, *The C Programming Language*, Second Edition. Prentice Hall, 1988. * Added "corrected" mod_i: NOTE: This "corrected mod" algorithm is based on the C code on page 70 of [1]. Assuming correct behavior of C's built-in mod operator (%) with positive arguments, this algorithm implements a mathematically convenient version of mod, defined thus: x mod y = x - y * floor(x / y) For more information on this definition of mod, see section 3.4 of [2], pages 81-85. References: [1] Donald E. Knuth, *MMIXware: A RISC Computer for the Third Millennium* Springer, 1999. [2] Ronald L. Graham, Donald E. Knuth and Oren Patashnik, *Concrete Mathematics*, Second Edition. Addison-Wesley, 1994. * Added mod_n, using the same formula as above, but with FLOATVAL arguments. * Added cmod_n, using the C math library's fmod() function: NOTE: This "uncorrected mod" algorithm uses the built-in C math library's fmod() function, which computes ... the remainder of dividing x by y. The return value is x - n * y, where n is the quotient of x / y, rounded towards zero to an integer. -- fmod() manpage on RedHat Linux 7.0 In addition, fmod() returns the remainder, unless y is zero, when the function fails and errno is set. According to page 251 of [1], the result when y is zero is implementation- defined. This op is provided for those who need it, but a more mathematically useful numeric mod based on floor(x/y) instead of truncate(x/y) and defined with y == 0 is provided by the mod_n op. [1] Brian W. Kernighan and Dennis M. Ritchie, *The C Programming Language*, Second Edition. Prentice Hall, 1988. * Added and modified tests as appropriate for the above. 2001-10-08 08:49 thgibbs * include/parrot/exceptions.h, include/parrot/parrot.h, include/parrot/string.h, include/parrot/strutf16.h, include/parrot/strutf32.h, include/parrot/strutf8.h, include/parrot/transcode.h, include/parrot/unicode.h, MANIFEST, Makefile.in, global_setup.c, string.c, strnative.c, transcode.c, strutf16.c, strutf32.c, strutf8.c: Added code to handle unicode strings for utf8, utf16, and utf32. Courtesy of: Tom Hughes 2001-10-08 08:10 thgibbs * memory.c: Fixed uninitialized variables. Courtesy of: Bryan C. Warnock 2001-10-08 01:21 gregor * packfile.c: * Casting a return value * Potential uninitialized variable fix Courtesy of Bryan C. Warnock 2001-10-08 01:15 gregor * packfile.c, test_main.c: One of Bryan's languishing patches (with a few tweaks of my own): * Line wrapping * Printf format fixes Courtesy of: Bryan C. Warnock 2001-10-07 22:39 thgibbs * build_interp_starter.pl: No longer adds (void*) casts to interp_guts.c function table as they are not necessary in ANSI C. 2001-10-07 16:27 gregor * basic_opcodes.ops, interpreter.c, opcode_table, process_opfunc.pl, test_main.c, include/parrot/interpreter.h, t/trace.pasm: * Two new ops, trace_ic and bounds_ic that twiddle bits in interpreter->flags and set interpreter->resume_addr. * Two new runops core variants, so now we cover the relevant flag space over {trace, notrace} x {boundscheck, noboundscheck} * Runops core selection that converts the flags into an index into a core table (runops_cores) to get the core function pointer. * A new trace_op variant that doesn't do bounds checking (used by the non-bounds-checking-but-tracing runops core). * PC is passed into the cores so that they can be used when resuming (they used to assume starting at the beginning of the code, which would thwart resumption). * runops() itself is changed to have a while() loop in it so that any time we drop out of the core, we'll resume if we have an interpreter->resume_addr set (after reselecting the core, though). * process_opfunc.pl has a new "macro": RESUME(), that sets up interpreter->resume_addr. This is used by the new ops. The new ops still return the new PC as the next op (which MUST be an 'end' op to force the core loop to terminate -- *pc being false is the only termination condition we can count on for all cores). * New interpreter flag: PARROT_BOUNDS_FLAG, used to control bounds checking core selection. * test_prog now has a '-b' flag in addition to the '-t' flag. Now, bounds checking is off by default (I saw a performance boost: 6.3 Mop/s with -b, 7.0 Mop/s without). * A new sample/test program: t/trace.pasm that twiddles the trace flag while it is running. 2001-10-07 15:41 gregor * .cvsignore, Makefile.in, build_interp_starter.pl, interpreter.c, include/parrot/.cvsignore, include/parrot/interpreter.h: Now that 0.0.2 is released, I can clear out the backlog of approved but uncommitted patches. Here's one Dan approved on 2001-09-21 (Simplified DO_OP): * test_prog and pdump are linked with only the files the need, not al the files. * build_interp_starter.pl writes out op_info.[hc], too. These files contain op name, arg count and arg type information. This will be useful for multiple things in the future, including better trace output and the crystalizing loader. * The DO_OP macro is now simpler and doesn't require explicit temporaries. * interpreter.c is modified to use the new op info in tracing output. * The interpreter just points its opcode_funcs and opcode_info members at the default tables rather than allocating memory on the fly. 2001-10-07 03:58 thgibbs * include/parrot/.cvsignore, vtable_h.pl: Minor updates fixing a comment and adding vtable.h to .cvsignore as it is produced by vtable_h.in 2001-10-06 20:58 thgibbs * vtable.tbl: Added * to REGEX type. Courtesy of Bryan C. Warnock 2001-10-06 13:41 simon * MANIFEST, Makefile.in, vtable.tbl, vtable_h.pl, Parrot/Vtable.pm, include/parrot/parrot.h, lib/Parrot/Vtable.pm: Framework for vtable functions. 2001-10-06 06:52 thgibbs * make_op_header.pl: Removed confusing and unused typedef. 2001-10-06 06:31 thgibbs * Parrot/Opcode.pm, lib/Parrot/Opcode.pm: Removed variables that are no longer used. 2001-10-06 06:26 thgibbs * disassemble.pl: Changed intval to op in disassembler. 2001-10-06 06:21 thgibbs * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: Fixed typo with Assembler, local labels, and integer constants. 2001-10-06 03:11 ask * Makefile.in: make `make test` actually run the tests on darwin/macosx 2001-10-06 02:04 thgibbs * Parrot/Assembler.pm, Parrot/Opcode.pm, Parrot/PackFile.pm, Parrot/String.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, include/parrot/packfile.h, include/parrot/register.h, lib/Parrot/Assembler.pm, lib/Parrot/Opcode.pm, lib/Parrot/PackFile.pm, lib/Parrot/String.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm, t/op/integer.t: No, it didn't mess anything up. I just forgot to type make. :) 2001-10-06 02:02 thgibbs * Parrot/Assembler.pm, Parrot/Opcode.pm, Parrot/PackFile.pm, Parrot/String.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, include/parrot/packfile.h, include/parrot/register.h, lib/Parrot/Assembler.pm, lib/Parrot/Opcode.pm, lib/Parrot/PackFile.pm, lib/Parrot/String.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm, t/op/integer.t: oops...this messed something up. I must have missed something in the patch...I have reverted. 2001-10-06 01:57 thgibbs * Parrot/Assembler.pm, Parrot/Opcode.pm, Parrot/PackFile.pm, Parrot/String.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, include/parrot/packfile.h, include/parrot/register.h, lib/Parrot/Assembler.pm, lib/Parrot/Opcode.pm, lib/Parrot/PackFile.pm, lib/Parrot/String.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm, t/op/integer.t, Types_pm.in, basic_opcodes.ops, config_h.in, memory.c, packfile.c, process_opfunc.pl, strnative.c, test_main.c: This patch ensures that all bytecode is of type opcode_t except for the constant table constants. That means that almost all of the INTVALs in the code were replaced with opcode_t and the i pack type was replaced with the op pack type in calls to pack_arg and such. Also, since it has been declared that ALL integer constants are to be 32 bits, some test cases were changed to ensure that an integer constant was no bigger than 2 ** 31. process_opfunc.pl was also cleaned up to no longer look at either opcode_table or interp_guts.h but gather all of its information from Opcode.pm. The Assembler.pm module was changed to do late binding of arguments so that 0 could be treated like 0.0 without having to be so declared. Also, only strings now use the [sc:N] intermediate form as it was unnecessary for integer and float types. 2001-10-05 21:59 gregor * MANIFEST, languages/jako/Makefile: Fixed things due to the movement of little_languages/* to languages/jako. 2001-10-05 18:31 gregor * MANIFEST: Updated MANIFEST to reflect recent additions for Jako. 2001-10-05 17:54 gregor * languages/jako/: Makefile, bench.jako, euclid.jako, fact.jako, fib.jako, hello.jako, jakoc, leibniz.jako, mandelbrot.jako, mops.jako, primes.jako, sub.jako: * Compiler: * User-defined subroutines. * Generates assembly code with the Jako code in comments. * "end;" no longer required at the end of programs. * Better handling of const-reg and reg-const. * Optimizes const-const operations to compile-time evaluations. * Constants in addition to variables. * Arithmetic fixes. * Allow leading minus signs in integer and number constants (ahem). * goto's. * New sample programs: * leibniz.jako (based on Greg McCarroll's leibniz.pasm example) * mandelbrot.jako (based on Leon Brocard's translation of Glenn Rhodes' C code) * mops.jako (based on t/test.pasm) * primes.jako (based on Leon Brocard's primes.pasm example) * sub.jako (demonstrates a simple user-defined subroutine) 2001-10-05 12:54 gregor * Parrot/Opcode.pm, lib/Parrot/Opcode.pm: * Removed mysterious extra copy of code after POD. 2001-10-05 12:46 gregor * basic_opcodes.ops, t/op/string.t: * Enabled the previously SKIPped test of printing string regs after clear_s * Made print_s do nothing if the STRING pointer in the register file is NULL. 2001-10-04 21:19 dan * Configure.pl: VMS tweaks 2001-10-04 21:17 dan * hints/vms.pl: Hints file for VMS 2001-10-04 00:58 bdwheele * Parrot/Assembler.pm, lib/Parrot/Assembler.pm: fixed but in Assembler which stopped equates from working. 2001-10-03 21:50 gregor * Makefile.in, docs/.cvsignore, docs/Makefile, docs/parrotbyte.pod: * More documentation updates for the packfile stuff. 2001-10-03 21:37 gregor * docs/parrotbyte.pod: * Pointed the reader at Parrot::PackFile. 2001-10-03 21:34 gregor * Parrot/PackFile.pm, lib/Parrot/PackFile.pm: * Updated documentation to match current implementation. 2001-10-03 18:30 gregor * .cvsignore, disassemble.pl, Parrot/Assembler.pm, lib/Parrot/Assembler.pm, t/call.pasm, t/euclid.pasm, t/local_label.pasm, t/test.pasm, t/test2.pasm, t/test3.pasm, t/op/basic.t: * Added new syntax to the assembler to allow square-bracket arithmetic expressions involving (global) labels, the current instructions base address ('@'), numeric constants (not equates) and basic arithmetic operators (+, -, *, / -- although + and - are likely to be the only ones used in practice). * Fixed a bug in Parrot/Assemble.pl that caused macro expansion to fail (argument passing wasn't working). * Enabled t/op/basic.t's last test, which is actually a copy of the code from t/call.pasm. Now all tests there are enabled and working. * Added new example/test program t/call.pasm to demonstrate working macros and poor-man's subroutine calls * Added header comments to t/*.pasm 2001-10-03 18:00 simon * ChangeLog: Needed, since it's in the MANIFEST. 2001-10-03 17:56 simon * t/op/: string.t, time.t: Small fixes to help Win32 on its way. 2001-10-03 17:41 simon * MANIFEST, README: Oops. Tiny bugs. 2001-10-03 17:39 simon * MANIFEST, NEWS, README: README, NEWS and MANIFEST? Could it be time for Parrot 0.0.2? I think so! 2001-10-03 17:21 simon * basic_opcodes.ops, interpreter.c, packfile.c: Evil printf bandaid (which allows Sparc64 to pass tests) Courtesy of: Andy Dougherty 2001-10-03 16:58 simon * string.c, t/op/string.t: chop fixes for n < 0 and tests Courtesy of: Alex Gough 2001-10-03 15:51 thgibbs * interpreter.c: Changed an occurrence of INTVAL to opcode_t Courtesy of: Bryan C. Warnock 2001-10-02 17:48 dan * include/parrot/stacks.h: Fix typo in function name 2001-10-02 17:35 dan * include/parrot/stacks.h: changed to support new stack stuff coming after 0.02 2001-10-02 15:01 simon * Configure.pl, Types_pm.in, basic_opcodes.ops, config_h.in, interpreter.c, make_op_header.pl, memory.c, packfile.c, process_opfunc.pl, register.c, string.c, strnative.c, Parrot/Opcode.pm, Parrot/PackFile.pm, Parrot/String.pm, lib/Parrot/Opcode.pm, lib/Parrot/PackFile.pm, lib/Parrot/String.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, docs/overview.pod, docs/strings.pod, include/parrot/interpreter.h, include/parrot/memory.h, include/parrot/packfile.h, include/parrot/parrot.h, include/parrot/register.h, include/parrot/stacks.h, include/parrot/string.h, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm: The Great Renaming. 2001-10-02 12:49 gregor * process_opfunc.pl: * Fixed a tyop: s/string/integer/ for INT_CONST expansion 2001-10-02 12:40 simon * basic_opcodes.ops, process_opfunc.pl, register.c, include/parrot/register.h: Move preprocessing of register accessors to process_opfunc.pl Courtesy of: Jason Gloudon 2001-10-02 01:20 ask * README: add a few lines about lists. 2001-10-01 23:00 brentdax * Configure.pl, Makefile.in, hints/mswin32.pl: Make 'make clean' work on Win32 (courtesy of Mattia Barbon ) 2001-10-01 22:50 brentdax * Configure.pl, test_c.in: More changes for 64-bit compatibility 2001-10-01 21:51 simon * process_opfunc.pl: Michael Fischer's patch broke conditionals. 2001-10-01 15:38 simon * process_opfunc.pl: Small hack to keep process_opfunc.pl from generating extra return() statements. Courtesy of: Michael Fischer 2001-10-01 15:32 simon * Parrot/Opcode.pm, lib/Parrot/Opcode.pm: Documentation fix. Courtesy of: Josh Wilmes 2001-10-01 01:49 simon * packfile.c: Simon is an idiot. Don't trust him with your code. Courtesy of: Gibbs Tanton - tgibbs 2001-10-01 00:59 simon * Parrot/String.pm, lib/Parrot/String.pm: Was in manifest, but not in CVS. 2001-09-30 21:25 simon * MANIFEST, Types_pm.in, assemble.pl, basic_opcodes.ops, disassemble.pl, packfile.c, pdump.c, process_opfunc.pl, Parrot/Assembler.pm, Parrot/PackFile.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, Parrot/PackFile/FixupTable.pm, lib/Parrot/Assembler.pm, lib/Parrot/PackFile.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm, lib/Parrot/PackFile/FixupTable.pm, include/parrot/packfile.h, include/parrot/string.h: Move NVs to the constant table. * Changes to Types_pm.in to make life easier for some of the packing/etc. I'm doing elsewhere. Also, we now pack 'n' as 'i'. * Moved the guts of assemble.pl to Parrot/Assembler.pm so that eventually we can use all these tools together inside scripts rather than call out to the shell to run the programs. * As I was tracking down problems, I converted the internal documentation of the assembler to POD from comments. * Changed the numeric regexp in the assembler to detect leading signs and trailing exponents. * Updated basic_opcodes.ops to fetch numeric constants from the constant table. * Enhanced the disassembler in a number of ways to make my life easier while tracking down bugs (try it out). It now puts in labels for ops that are the destinations of flow control ops. * packfile.c now has a TRACE_PACKFILE switch that can be set to get it to print info as it processes. I needed this while debugging, but I can remove it or change it before committing if its a problem. * Updated pdump.c because it was missing a call to init_world(). * Updated process_opfunc.pl to deal with numeric constants being stored in the constant table instead of inline. * Added multiple type support to Parrot/PackFile/Constant.pm * Added Parrot/String.pm, to be used for constants. * Added packfile.[hc] functions for allocating new constants of various types. * Made string_copy visible to the world in parrot/string.h. Wow. Courtesy of: Gregor N. Purdy 2001-09-29 15:56 gregor * languages/jako/Makefile: Removed assumption that '.' is on the path. 2001-09-29 15:55 gregor * languages/jako/Makefile: Removed references to leibnitz test program. 2001-09-29 15:21 thgibbs * t/op/trans.t, basic_opcodes.ops: Added NV casts to transcendental math ops args to work on Solaris. 2001-09-27 15:15 thgibbs * assemble.pl: Added case insensitive matches to hex and binary characters Courtesy of: Johnathan Scott Duff 2001-09-27 04:46 thgibbs * assemble.pl: This patch makes integer constants of the form 0b[01]+ work with perl 5.005. The oct function in versions of perl before 5.6 did not handle the 0b numeric prefix; therefore, it was returning 0. To fix this, I added a from_binary function that would take a binary string and return its decimal representation. Now, instead of blindly calilng oct, it makes a distinction between 0/0x numbers and 0b numbers. 2001-09-26 19:30 thgibbs * assemble.pl: Error now returns a non-zero status Courtesy of: Andrew Kuchling 2001-09-26 19:13 gregor * MANIFEST, Makefile.in, basic_opcodes.ops, bytecode.c, interpreter.c, packfile.c, pdump.c, test_main.c, include/parrot/bytecode.h, include/parrot/interpreter.h, include/parrot/packfile.h, include/parrot/parrot.h, t/op/integer.t, t/op/stacks.t: * bytecode.[hc] removed. packfile.[hc] are the interface to the packfile/bytecode file from C. May be renamed in the future. * packfile.[hc] changed to use 'struct PackFile *', etc. instead of 'PackFile *', just like the interpreter, etc. * Constants are accessed via the interpreter's 'code' member (struct PackFile * code). Macros PCONST(i) and PNCONST are provided to access constants by index and the number of constants, respectively (assumes a variable 'interpreter' of the appropriate type is in scope). * Byte Code is accessed via the interpreter's 'code' member. * Changed PackFile_Constant to point to a STRING instead of having members for the string_make() arguments. Also has a 'type' IV in preparation for allowing multiple types of constants in const_table. * Removed hardcoded example program from test_main.c. We have plenty of test programs elsewhere. * test_prog now uses a PackFile to deal with the program before sending it on to the interpreter. * Added missing 'end' ops to some tests in t/op/integer.t and t/op/stacks.t. 2001-09-26 18:38 thgibbs * assemble.pl: Separated code into functions and added comments. 2001-09-26 18:29 simon * config_h.in: Fix 64-bit chunk-base issue. 2001-09-26 06:55 thgibbs * t/op/bitwise.t: Added end to bitwise tests. 2001-09-26 00:54 dan * docs/parrot_assembly.pod: Added in the base for threading, I/O, and interpreter creation. Skeleton only, but it's something to peer at. 2001-09-25 21:44 thgibbs * t/op/trans.t: Removed loops in transcendtal op test. 2001-09-25 10:12 simon * MANIFEST, t/op/basic.t, t/op/bitwise.t, t/op/integer.t, t/op/string.t, t/op/time.t: More tests Courtesy of: Alex Gough 2001-09-24 19:08 gregor * t/op/number.t: * Enabled the clear_n test. 2001-09-24 18:56 gregor * Parrot/.cvsignore: * Ignore the generated Types.pm file. 2001-09-24 18:52 gregor * languages/jako/Makefile, t/op/number.t: * No need to skip the inc and dec numeric tests anymore. I fixed the assembler earlier today. 2001-09-24 18:21 simon * MANIFEST, t/op/stacks.t: Oops. 2001-09-24 18:19 simon * build_interp_starter.pl, interpreter.c, string.c, Parrot/Test.pm, include/parrot/exceptions.h, include/parrot/interpreter.h, lib/Parrot/Test.pm, t/op/string.t: 1) captured error output, so can test errors, this might be bad, not sure. 2) basic tests for register stack frames, including some errors 3) I think we can stop skipping one of the string tests 4) I added some more substr tests, but substr was broken, so 5) changed substr to work with out of range values, or 6) throw an error if they're far too silly Courtesy of: Alex Gough 2001-09-24 17:27 gregor * basic_opcodes.ops, opcode_table, t/op/number.t: * Op set_n was missing. Added it and a test for it. We now have the same int and num ops, except mod is int-only. 2001-09-24 17:01 gregor * assemble.pl: * Fixed a typo: s/Line/line/ 2001-09-24 16:02 gregor * basic_opcodes.ops: * Fixed dec_n_nc opcode, which had += where it should have had -=. 2001-09-24 15:34 gregor * assemble.pl: * Fixed handling of constants starting with '0' that should not be sent through oct(). For example 'set N2 0.1' was being interpreted as 'set N2 0' (naughty). As reported by Leon Brocard in the comments of his mandelbrot.pasm example. 2001-09-24 02:21 bdwheele * assemble.pl: fixed dumb code when --include was not specified. 2001-09-24 02:17 simon * assemble.pl: Shit-sweeping. 2001-09-24 01:56 bdwheele * assemble.pl: * Added --include=dir option to assembler to provide a path to look for included files. * Error message of a macro is defined inside of a macro 2001-09-23 21:29 bdwheele * assemble.pl: Added __FILE__ equate, fixed __LINE__ equate. 2001-09-23 15:50 simon * Configure.pl: Fixed typo. Parrot might work again on non-8.3 filesystems. :) 2001-09-23 02:03 ask * Configure.pl, test_c.in: fix missing newline at the end of test.c Courtesy of: acme@astray.com (Leon Brocard) rename 'test' executable used in Configure process to test_siz. 2001-09-23 01:03 simon * process_opfunc.pl: iPaq fix. No, really. 2001-09-23 00:47 simon * MANIFEST: Missed several, in fact. 2001-09-23 00:46 simon * MANIFEST: Missed one. 2001-09-22 21:13 simon * assemble.pl: Assembler register numbering rigour. Courtesy of: Leon Brocard 2001-09-22 18:20 simon * Makefile.in, lib/Test/More.pm, lib/Test/Simple.pm, lib/Test/Utils.pm: I'm sick of having to install Test::More everywhere. Temporary measure until we remove this silly dependency. Also a typo in Makefile.in which suggests some things were only happening correctly by sheer chance. :) 2001-09-22 18:06 simon * README: test.pasm is dead, long like make test. 2001-09-22 17:47 simon * t/op/number.t: De-fang overzealous floating point test. 2001-09-22 17:39 simon * process_opfunc.pl: Remove hard-coded NV size assumption. Tru64 is nearly there! 2001-09-22 17:18 simon * assemble.pl, disassemble.pl: More portability fixups. 2001-09-22 16:44 simon * assemble.pl, Parrot/PackFile/Constant.pm, lib/Parrot/PackFile/Constant.pm: And I *think* this fixes Tru64. 2001-09-22 16:37 simon * Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/ConstTable.pm: Oops. 2001-09-22 16:36 simon * Parrot/PackFile.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, lib/Parrot/PackFile.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm: More removal of hard-coded pack formats. 2001-09-22 15:21 simon * assemble.pl: Grrr. Why isn't this working? 2001-09-22 15:18 simon * Parrot/PackFile.pm, Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile.pm, lib/Parrot/PackFile/ConstTable.pm: Make assemble use the PackFile modules; also fixed some bugs in them. 2001-09-22 14:38 simon * Configure.pl, MANIFEST, Types_pm.in, assemble.pl, disassemble.pl: Abstract the argument and opcode packing types and functions out into a separate module. This *should* help us maintain consistency when we change the various types. 2001-09-22 00:46 brentdax * Configure.pl: Delete test.o (or its equivalent) after Configure's test program finishes. 2001-09-21 21:24 brentdax * Parrot/Test.pm, lib/Parrot/Test.pm: Fix error with \ as path separator (courtesy of Mattia Barbon ) 2001-09-21 21:09 gregor * Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm: * Fixed pad calculation (similar to recent bug fix on C library) 2001-09-21 20:23 simon * disassemble.pl: Pack types and sizes have gone askew. 2001-09-21 20:18 brentdax * Configure.pl: Fix typo in setting stuff from the C test program 2001-09-21 10:25 simon * basic_opcodes.ops, process_opfunc.pl: Document patches. Courtesy of: Leon Brocard 2001-09-21 06:06 ask * Config_pm.in, assemble.pl: fix 5.005_03. 2001-09-21 06:03 ask * Configure.pl: make the ivsize detection work for us normal people without . in our PATH. :-) 2001-09-21 05:56 brentdax * test_main.c: Change _read to read for portability to Tru64 (Damien Neil and likely many others) 2001-09-21 05:46 ask * Makefile.in: "update" target to update from cvs. 2001-09-21 05:44 brentdax * Configure.pl, test_c.in: Fix issues with ivsize and nvsize 2001-09-21 02:43 brentdax * Configure.pl: Fix a typo that forced the -g flag on 2001-09-20 22:41 dan * packfile.c: Had some no-value returns. Naughty 2001-09-20 22:40 dan * t/test.pasm: Added labels to output 2001-09-20 15:39 simon * Parrot/Opcode.pm, Parrot/Test.pm, lib/Parrot/Opcode.pm, lib/Parrot/Test.pm: This patch does two only somewhat related things. Parrot/Opcode.pm: Change the hard-wired assumption that sizeof(nv) = 2*sizeof(iv) to a PConfig-type computed value. Parrot/Test.pm: Change it to run with the perl that was used to generate this version of Parrot. This is useful if, for example, you want to use bleadperl to build and test parrot (e.g. to try out various 64-bit options) but you still have a 5.6.x (or 5.005_xx) installed as the default 'perl' on your system. Since perl-5.7.x is installed 'version-only', a plain 'perl' is quite likely to pick up the older perl in such cases. Such an older perl might well not understand pack('q') used in the assembler. Courtesy of: Andy Dougherty 2001-09-20 15:14 gregor * packfile.c: * Fixed a typo in pad calculation that throwing off unpack and pack. 2001-09-20 15:03 thgibbs * t/op/integer.t: Removed overflow tests from integer.t. set_i_ic tests now cover all registers. Courtesy of: Damien Neil neild@misago.org 2001-09-20 14:48 thgibbs * process_opfunc.pl: Changed process_opfunc.pl to work on 64 bit machines. 2001-09-20 14:35 simon * Configure.pl, assemble.pl: Pack IVs correctly. Courtesy of: Andy Dougherty 2001-09-20 14:06 gregor * languages/jako/jakoc: * Fixed a tyop in the string interpolation regex. Courtesy of Michael L Maraist 2001-09-20 13:56 gregor * languages/jako/jakoc: * Updated jakoc to work with the single-destination comp ops. 2001-09-20 05:08 bdwheele * assemble.pl: macro expansion. Here's the test macro I was using: # this uses I32,I31 as temporaries. # macro must be preceeded by a label. testi macro num,got,want print "Test #" print num print "..." set I32,want set I31,got eq I32,I31,$ok print "NG\n" branch $end $ok: print "OK\n" $end: endm It is invoked as: Test1: testi 1,I3,I2 2001-09-20 03:52 dan * stacks.c, include/parrot/stacks.h: Preliminary files for the generic stack support 2001-09-20 02:49 bdwheele * assemble.pl: Added include processing to assemble.pl the syntax is: include 'file.to.include' 2001-09-19 23:57 thgibbs * t/op/integer.t, t/op/number.t, assemble.pl, opcode_table: Fixed opcode_table listing of if_i_ic and if_n_ic. Dan had changed the function to only use 2 args, but the opcode_table still had three listed. Fixed integer.t and number.t to use the new comparison operators. Improved assemble.pl ability to grok operators by constructing a function set and comparing operator types. 2001-09-19 22:42 dan * assemble.pl: Binmodes the output filehandles now. (No more bytecode munching...) 2001-09-19 22:32 dan * Configure.pl, basic_opcodes.ops, opcode_table, process_opfunc.pl, t/euclid.pasm, t/test.pasm, t/test2.pasm: Changed opcodes so conditional branches take a single argument, and fall through otherwise. 2001-09-19 21:19 thgibbs * memory.c: Added explanation of mem_allocate_aligned give by Dan Sugalski Courtesy of: Josh Wilmes 2001-09-19 21:04 thgibbs * include/parrot/interpreter.h, Configure.pl, build_interp_starter.pl, bytecode.c, config_h.in, interpreter.c, process_opfunc.pl: Changed opcode_t from a union to a configurable type. 2001-09-19 20:54 thgibbs * assemble.pl: Missed a couple of changes for 64 bit portability. 2001-09-19 20:37 gregor * packfile.c, include/parrot/packfile.h: * Added "methods" for setting and getting all the various attributes, probably sufficient for an assembler to use the library. 2001-09-19 20:04 thgibbs * assemble.pl: Modifications to run on 64 bit platforms. To run on a 64 bit platform manually change the setting for 'i' from 'l' to 'q'. This should be automatically done by a Configure setting in the near future. 2001-09-19 19:10 thgibbs * config_h.in: changed c++ comment to c comment. Courtesy of: Andy Dougherty 2001-09-19 17:48 gregor * .cvsignore, MANIFEST, Makefile.in, packfile.c, pdump.c, include/parrot/packfile.h: * C code for dealing with PackFiles as if they were objects. Mostly read-only interface for now, but it would not be hard to make it read-write. [packfile.c, include/parrot/packfile.h] * PackFile Dumping program that falls short of disassembling, but tests the above code and provides a hex listing of the byte code contents of the PackFile. [pdump.c] * New Makefile target 'pdump' for the above program, and related Makefile.in changes. * Updated MANIFEST to reflect the added files * Update .cvsignore to ignore pdump. 2001-09-19 14:55 thgibbs * MANIFEST: Added number.t and trans.t to MANIFEST 2001-09-19 05:40 thgibbs * process_opfunc.pl: Now taking & of opcode_t instead of i member of union opcode_t. I hope this will be more portable for set_n_nc. 2001-09-19 05:39 thgibbs * t/op/integer.t: Added end to test 1. 2001-09-19 00:09 thgibbs * t/op/number.t: Added end to each test 2001-09-19 00:00 simon * t/op/number.t: Number tests! Wooo! Courtesy of: Damien Neil 2001-09-18 23:52 thgibbs * interpreter.c: Added casts to appropriate types to make -Wall not complain on certain printf formats 2001-09-18 23:31 thgibbs * t/op/integer.t: Added end to each test. 2001-09-18 22:54 thgibbs * t/op/trans.t: Added trans.t for transcendental math function tests. 2001-09-18 22:28 simon * MANIFEST: MANIFEST updates. (And notice how I carefully snuck in the t/op/integer.t there...) Courtesy of: Andy Dougherty 2001-09-18 22:26 simon * t/: integer.t, op/integer.t: Except it doesn't go there. Duh. 2001-09-18 22:25 simon * t/integer.t: Tests for integer ops. Courtesy of: Damien Neil 2001-09-18 22:03 thgibbs * include/parrot/bytecode.h, include/parrot/interpreter.h, build_interp_starter.pl, bytecode.c, config_h.in, interpreter.c, make_op_header.pl, process_opfunc.pl, test_main.c: Changed IV to opcode_t union for operators. This required a few perl script changes to output the correct type. There were also some changes to allow -pedantic in gcc 2001-09-18 21:07 gregor * disassemble.pl, Parrot/PackFile.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, lib/Parrot/PackFile.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm: Converted the disassembler to use the Parrot::PackFile Perl classes to do its job. 2001-09-18 19:18 brentdax * Makefile.in: Fix missing functions for real this time. (I swear!) 2001-09-18 18:49 brentdax * Configure.pl: Remove unnecessary -I.. (courtesy of Andy Dougherty } 2001-09-18 18:47 brentdax * Makefile.in: Fix bugs with missing math, etc. functions 2001-09-18 14:42 thgibbs * strnative.c: Sorry Simon...forgot bustart was a void* not a char*. 2001-09-18 13:59 thgibbs * strnative.c: Removed unnecessary IV cast in concat. 2001-09-18 13:58 thgibbs * t/op/string.t: Removed unnecessary IV cast in concat. Added end to concat test in string.t 2001-09-18 12:07 gregor * languages/jako/jakoc: * var num ... no longer requires assignment * automatic casting between int and num * Fixed an error message Courtesy of: Simon Cozens 2001-09-18 11:57 simon * TODO: Gonna have to start thinking of more things to do, soon. 2001-09-18 10:42 simon * opcode_table: Introduce a little more tidiness into the opcode table. 2001-09-18 03:06 gregor * basic_opcodes.ops, opcode_table, t/euclid.pasm: Resurrected register-const ops per Dan's request. 2001-09-18 02:20 gregor * .cvsignore, Configure.pl, MANIFEST, Makefile.in, build_interp_starter.pl, bytecode.h, events.h, exceptions.h, global_setup.h, interpreter.h, io.h, memory.h, parrot.h, process_opfunc.pl, register.h, stacks.h, string.h, strnative.h: Committed some files that were missed in the last *.h moving commit. 2001-09-18 02:17 gregor * include/parrot/.cvsignore: Fixed a typo in .cvsignore. 2001-09-18 02:16 gregor * include/parrot/: .cvsignore, bytecode.h, events.h, exceptions.h, global_setup.h, interpreter.h, io.h, memory.h, parrot.h, register.h, stacks.h, string.h, strnative.h: Added .cvsignore file missed in last commit for the move of *.h to include/parrot. 2001-09-18 01:32 simon * t/Makefile: Trivial "clean" target to t/Makefile. 2001-09-18 01:32 simon * assemble.pl, build_interp_starter.pl, disassemble.pl, Parrot/Opcode.pm, lib/Parrot/Opcode.pm: Change icky Digest::MD5 for nice simple unpack() Courtesy of: Jonathan Scott Duff 2001-09-18 00:35 simon * basic_opcodes.ops, opcode_table, docs/parrot_assembly.pod, t/op/string.t: New concat op. Hey, if I'm going to talk about it in an article, better implement it... 2001-09-17 22:54 gregor * MANIFEST, Parrot/PackFile.pm, Parrot/PackFile/ConstTable.pm, Parrot/PackFile/Constant.pm, Parrot/PackFile/FixupTable.pm, lib/Parrot/PackFile.pm, lib/Parrot/PackFile/ConstTable.pm, lib/Parrot/PackFile/Constant.pm, lib/Parrot/PackFile/FixupTable.pm: Functions for reading, writing and manipulating Parrot Pack Files. NOTE: This is a proposed term based on the fact that the file contains more than just byte code, and another language (Java) calls its binary files by a broader notion of what's in them (Class Files). Since Perl has packages, and since we use pack() to write them, we'll call them PackFiles (at least for now). That does, however, beg the question of file extension. Should we name the files "foo.pack" instead of "foo.pbc"? 2001-09-17 20:23 thgibbs * interpreter.c, test_main.c: Uncuddled else statements for coding standard. 2001-09-17 19:57 gregor * Parrot/Opcode.pm, lib/Parrot/Opcode.pm: Added a $revision variable, parsing of the CVS Id line in opcode_table and a &revision() subroutine. 2001-09-17 19:53 gregor * opcode_table: Added a CVS Id tag to it. 2001-09-17 18:58 thgibbs * interpreter.c: Cleared interpreter flags on creation. 2001-09-17 18:38 gregor * build_interp_starter.pl, bytecode.c, bytecode.h, interpreter.c, interpreter.h, test_main.c: * Provides macros in interp_guts.h for setting up arrays with op names and op arg counts (done via build_interp_starter.pl). These are used in the op tracing * The functions in bytecode.[hc] pass around a pointer to the length of the bytecode so that later when we get to runops, we know how much bytecode we've got and we can detect out-of-bounds jumping. The documentation is updated, too. * runops in interpreter.c now looks at interpreter->flags to decide if the core of runops should be runops_trace_core or runops_notrace_core. These new functions contain just the while-loop portion of runops. A new function runops_generic does any other setup (such as checking the bytecode fingerprint) or wrapup (such as complaining if we ended up out-of-bounds). NOTE: I didn't know what we should do for functions in here that are not part of the api, so I gave them docs with 'TODO' marks mentioning they really aren't part of the api. Guidance appreciated. * test_main.c now checks for '-t' arg and sets the tracing flag on its interpreter instance as appropriate. 2001-09-17 17:47 bdwheele * assemble.pl: added some default equates, and if they're sub references, they are computed each time it is referenced. 2001-09-17 16:15 dan * interpreter.h: Added flags to the interpreter structure and defined a few. 2001-09-17 15:51 simon * bytecode.h, string.h: VAR_SCOPE on globals. 2001-09-17 15:20 simon * MANIFEST: Parrot/Opcode.pm, Parrot/Test.pm into MANIFEST. 2001-09-17 15:05 gregor * t/jump.pasm, MANIFEST: Added a test program to demonstrate jump_i. 2001-09-17 14:44 thgibbs * register.c: Fixed register pop not using correct offset (off by one). Courtesy of: Philip Kendall 2001-09-17 13:56 thgibbs * interpreter.c, register.c: Added code to call Parrot_clear_s during initialization and Parrot_clear_s and Parrot_clear_p when a new stack frame is pushed. 2001-09-17 13:51 gregor * languages/jako/: Makefile, bench.jako, euclid.jako, fact.jako, fib.jako, hello.jako: Added bench.pasm to "clean" 2001-09-17 13:49 gregor * languages/jako/: jakoc, Makefile: Added CVS Id tag. 2001-09-17 13:49 gregor * languages/jako/jakoc: Code streamlining. Courtesy of: Buggs 2001-09-17 13:29 gregor * MANIFEST, opcheck.pl: Added an opcode_table checker to make sure that all opcodes listed there conform to a simple set of naming rules (rules themselves, of course, are subject to change). 2001-09-17 13:23 thgibbs * string.c: Modified string.c to correctly use the return value of realloc. 2001-09-17 05:26 brentdax * Makefile.in: Making the Makefile smarter 2001-09-16 23:06 brentdax * hints/mswin32.pl: Finishing the Configure hints 2001-09-16 23:05 brentdax * Configure.pl, Makefile.in, parrot.h, test_main.c, Parrot/Opcode.pm, lib/Parrot/Opcode.pm: Win32 compatibility, Configure hints, 5.005 compatibility, MANIFEST checking 2001-09-16 19:25 bdwheele * assemble.pl: parrot assembly equates 2001-09-16 18:03 gregor * Makefile.in: Added 'test' target to run new test suite. 2001-09-16 17:33 simon * basic_opcodes.ops: 'not' was being !, rather than ~. Too many Perl programmers here, I suspect. Courtesy of: Tom Hughes 2001-09-16 17:21 simon * Parrot/Test.pm, lib/Parrot/Test.pm, t/harness, t/op/basic.t, t/op/string.t: A *real* test suite! Courtesy of: "Mattia Barbon" 2001-09-16 16:49 gregor * basic_opcodes.ops, opcode_table: At Dan's request, my new register-constant comparison ops have been removed. 2001-09-16 16:40 gregor * languages/jako/jakoc: Fixed a copy-paste error in one of jakoc's error messages. Courtesy of: Buggs 2001-09-16 16:37 gregor * languages/jako/jakoc: Modified comparison implementation to use temporary registers in the case of register-constant and constant-register comparisons. This will allow the compiler to emit code that works when the constant-comparison ops are backed out of Parrot. If the ops ever make their way back in, then jakoc can emit more efficient code again. 2001-09-16 16:19 bdwheele * assemble.pl, t/local_label.pasm: Added local labels ($whatever) ... though not subroutine based like the documentation wants, they are local to the last non-local label defined. 2001-09-16 14:02 simon * MANIFEST: And remember, when you add stuff, keep the manifest up to date. 2001-09-16 02:55 ask * Makefile.in: a "shared" target for building libparrot.so 2001-09-16 02:45 thgibbs * bytecode.c, bytecode.h, events.h, exceptions.h, global_setup.c, global_setup.h, interpreter.c, interpreter.h, io.h, make_op_header.pl, memory.c, memory.h, parrot.c, parrot.h, process_opfunc.pl, register.c, register.h, stacks.h, string.c, string.h, strnative.c, strnative.h, test_main.c: Updated files to use the coding standard PDD. Also changed Perl_Interp to Parrot_Interp. 2001-09-16 02:21 ask * README: emphasise the bits about the license not being worked out yet. 2001-09-15 22:21 simon * process_opfunc.pl: basic_opcodes.c should *scream* that it's autogenerated. 2001-09-15 21:58 gregor * languages/jako/Makefile: Fixed Makefile to use 'jakoc' instead of 'jako_compiler.pl' to compile. 2001-09-15 21:58 gregor * languages/jako/: jako_compiler.pl, jakoc: Renamed the Jako compiler to jakoc from jako_compiler.pl 2001-09-15 21:51 gregor * languages/jako/: Makefile, bench.jako, euclid.jako, jako_compiler.pl: Added a new bench.jako example based on Leon Brocard's Bench.java. Jako Compiler: * fix a bug in assignment handling * add some more conditionals to while and if * revert to using implicit opcodes in a few cases that were broken before the assembler patch went in recently. 2001-09-15 21:46 gregor * assemble.pl, basic_opcodes.ops, opcode_table: New ops for register-constant INTEGER comparisons: {eq,ne,lt,le,gt,ge}_ic_ic New ops for register-register and register-constant NUMERIC comparisons: eq_nc_ic {ne,lt,le,gt,ge}_{n,nc}_ic Assembler: Allow uppercase and underscores in labels. Tweaks to the op infer code to make it work with the examples I've been creating. 2001-09-15 19:16 thgibbs * README: Updated README to include Configure.pl 2001-09-15 17:05 simon * build_interp_starter.pl: Casting happiness Courtesy of: Jarkko Hietaniemi 2001-09-15 16:45 simon * assemble.pl: And the same again: Parrot/Opcode.pm wasn't "..-aware". 2001-09-15 16:40 simon * t/Makefile: The Parrot/Opcode.pm dependency means we need to do -I.. in the test Makefile 2001-09-15 12:26 simon * memory.c: Fixed to match the coding standards. Courtesy of: Gibbs Tanton - tgibbs 2001-09-15 01:57 ask * Configure.pl, Makefile.in, Parrot/Opcode.pm, lib/Parrot/Opcode.pm: make the Makefile use the perl that was used to run Configure.pl 2001-09-15 01:48 ask * Parrot/Opcode.pm, lib/Parrot/Opcode.pm: if we need perl 5.6, just say so. :-) 2001-09-15 01:42 gregor * languages/jako/: euclid.jako, fact.jako, fib.jako, jako_compiler.pl: Compiler: Added variable interpolation in double-quoted strings. Examples: Made use of variable interpolation. 2001-09-15 01:28 gregor * Parrot/.cvsignore: Ignore the Config.pm generated file. 2001-09-15 01:22 gregor * languages/jako/: .cvsignore, Makefile, euclid.jako, fact.jako, fib.jako, hello.jako, jako_compiler.pl: gnp: Added some more sample Jako files and added many more operators to the compiler. Preparations for supporting all the Math and bitwise ops. 2001-09-14 19:06 simon * assemble.pl, build_interp_starter.pl, disassemble.pl, make_op_header.pl, process_opfunc.pl, Parrot/Opcode.pm, lib/Parrot/Opcode.pm: Move common table reading stuff to a module. Courtesy of: Damien Neil 2001-09-14 17:42 simon * .cvsignore: More .cvsignore foo. 2001-09-14 17:40 simon * assemble.pl, config_h.in: De-coredump Courtesy of: Gibbs Tanton - tgibbs 2001-09-14 17:38 simon * assemble.pl, t/Makefile: Make assemble.pl look in its parent directory if the data files it needs aren't in the current one. Plus a Makefile that lets "make foo.pbc" work in t/ Courtesy of: Rocco Caputo 2001-09-14 16:29 simon * make_op_header.pl: Wildly experimental patch to make Parrot compile on FreeBSD. (And probably segfault everywhere else.) 2001-09-14 15:58 dan * languages/jako/: euclid.jako, jako_compiler.pl: Added in the first mini-language that targets Parrot. Courtesy of: "Gregor N. Purdy" 2001-09-14 15:20 dan * memory.c: Was misaligning returned memory. Fixed. 2001-09-14 15:08 simon * Makefile.in, basic_opcodes.ops, bytecode.c, global_setup.c, interpreter.c, interpreter.h, memory.c, parrot.c, parrot.h, register.c, register.h, string.c, string.h, strnative.c, test_main.c: Fixes to include Courtesy of: Gibbs Tanton - tgibbs 2001-09-14 12:05 simon * strnative.c: Appease Tru64. 2001-09-14 12:01 simon * memory.c: Appease Tru64. 2001-09-14 11:54 simon * interpreter.c: Remove lvalue cast. 2001-09-14 11:45 simon * Configure.pl: Be 5.005 friendly. 2001-09-14 11:39 simon * Configure.pl: Tru64 cc doesn't like -Wall 2001-09-14 10:57 simon * Config_pm.in, Configure.pl, MANIFEST, Makefile, Makefile.in, assemble.pl, build_interp_starter.pl, bytecode.c, bytecode.h, config.h.in, config_h.in, disassemble.pl, interpreter.c: Configure now generates Makefile Courtesy of: Brent Dax 2001-09-14 10:08 simon * t/test3.pasm: Add "end" Courtesy of: Gibbs Tanton - tgibbs 2001-09-14 10:07 simon * assemble.pl: * uses Getopt::Long for options. -c is now --checksyntax. I wasn't sure how to keep compatible (patches welcome!) * options include: --help --version --verbose --output=file --listing=file --checksyntax * produces verbose listing of what the assembler saw :) Only one nitpick with it: unknown symbols are given as 0xffffffff, unfortunately, this includes symbols which may be defined later in the file (i.e. forward jumps). Courtesy of: Brian Wheeler 2001-09-14 10:06 simon * strnative.c: substr growth fixup Courtesy of: Gibbs Tanton - tgibbs 2001-09-14 10:03 simon * bytecode.c, global_setup.c, interpreter.c, memory.c, memory.h, register.c, string.c, strnative.c, test_main.c: Coding standards conformance Courtesy of Gibbs Tanton - tgibbs 2001-09-14 09:46 simon * assemble.pl: Make the assembler grok transcendental maths again. Courtesy of: Gibbs Tanton - tgibbs 2001-09-13 17:29 simon * assemble.pl: Better opcode guessing, from Brian Wheeler. (Again!) 2001-09-13 17:20 dan * docs/parrot_assembly.pod: Took out duplicate entries for iton, ntoi, and tostring. 2001-09-13 17:16 simon * assemble.pl, basic_opcodes.ops, opcode_table: From Brian Wheeler: This patch gives the assembler support of '\a','\n','\r','\t', and '\\' in string constants. In addition, it changes (for all registers) "I reg %li is ..." to just the value of the register. Printing constants is also supported, but alas, you have to specify the type (print_sc, print_ic, print_nc). 2001-09-13 15:38 simon * assemble.pl: Why don't I actually *test* this stuff before applying it, eh? 2001-09-13 15:34 simon * TODO, t/euclid.pasm: From: "Gregor N. Purdy" Subject: [patch (really)] Removed opcode qualifers t/euclid.pasm to exercise opcode inference 2001-09-13 14:42 simon * assemble.pl: A new assembler, from Brian Wheeler. 2001-09-13 09:44 simon * basic_opcodes.ops, interpreter.h, parrot.h, strnative.c, test_main.c: Two more from Tom Hughes: Patch to fix C++ style comments Patch to remove use of structure constant/cast 2001-09-13 09:39 simon * TODO, docs/strings.pod: Update string docs. 2001-09-13 09:36 simon * Makefile, basic_opcodes.ops: Tom Hughes Patch to link with the maths library Tom Hughes Patch to fix += on rvalue 2001-09-13 08:57 simon * MANIFEST: I guess nobody's been using the nightly builds... 2001-09-13 08:27 simon * assemble.pl, basic_opcodes.ops, opcode_table: Bitwise ops, from Brian Wheeler. 2001-09-13 08:23 simon * .cvsignore, t/.cvsignore: .cvsignore's, from Gregor Purdy. 2001-09-13 08:22 simon * Makefile: Euclidean test and test output rearranging from Gregor Purdy. 2001-09-13 08:21 simon * assemble.pl, basic_opcodes.ops, disassemble.pl, opcode_table: Math ops from Gibbs Tanton. 2001-09-13 08:14 simon * basic_opcodes.ops, opcode_table, string.c, string.h, strnative.c, t/test2.pasm: String substr op; also rearranged "encoding" to be a vtable pointer, not an array index. Haven't changed docs to reflect this, oops. Simon 2001-09-12 19:48 dan * t/euclid.pasm: Implementation of Euclid's Algorithm, courtesy of Gregor Purdy 2001-09-12 19:39 dan * basic_opcodes.ops, opcode_table: From Gregor Purdy. Adds mod_i. 2001-09-12 19:23 dan * interpreter.h, parrot.h: Quick placeholder entry for string function tables 2001-09-12 18:58 dan * memory.c: Added in a cast to make gcc happy. 2001-09-12 16:44 dan * docs/parrot_assembly.pod: PDD describing parrot's assembly language and the opcodes in it 2001-09-12 14:20 dan * memory.c: Fixed alignment issues with allocation 2001-09-12 10:54 simon * Makefile, assemble.pl, bytecode.c, disassemble.pl, make_op_header.pl, opcode_table, process_opfunc.pl: Brian Wheeler's big patch: * Changes the opcode_table file to provide additional information about the operands. Case shouldn't be a problem since that data never becomes a C symbol [this is pretty much as before] * Padding errors solved: assemble.pl and bytecode.c were padding the constants incorrectly. It should have been 4-(size % 4), not just (size % 4). It is now fixed in both places. * assembler has less special cases, and should be easier to hang error checking on * disassembler dumps constant table and the format is a bit prettier, including register names, etc. 2001-09-11 23:21 dan * build_interp_starter.pl, interpreter.c: Abstracted the runops function core into a macro for easy replacement later. 2001-09-11 10:43 simon * Configure.pl, MANIFEST, Makefile, config.h, config.h.in: Configure system from Brent Dax 2001-09-11 09:38 simon * assemble.pl, test.pbc: From: Bryan C. Warnock Subject: Patch: assembler deferred output Also got rid of the bytecode, since that's probably way out of date. 2001-09-11 09:26 simon * basic_opcodes.ops, opcode_table, t/test2.pasm: From: Bryan C. Warnock Subject: length_s_i patch 2001-09-10 23:20 simon * MANIFEST: And that move needs a MANIFEST change. So many things to remember! 2001-09-10 23:18 simon * t/: test.pasm, test2.pasm, test3.pasm: Move tests to t/. (Oh, CVS doesn't work the way I think it does.) 2001-09-10 23:18 simon * test.pasm, test2.pasm, test3.pasm: Move tests to t/ 2001-09-10 23:04 dan * test3.pasm: Register push/pop test 2001-09-10 22:47 simon * basic_opcodes.ops, bytecode.c, parrot.h, test_main.c: Fix up format warnings, from Sam Tregar . (Modified to avoid conflict with Dan's changes.) 2001-09-10 22:45 dan * disassemble.pl: Happier about blanks in opcode_table 2001-09-10 22:40 dan * interpreter.c, register.c: Yanked out debugging code, since these work OK. The bug I want is elsewhere. 2001-09-10 22:30 simon * MANIFEST: Uhm. We want make_op_header.pl. We really, really do. 2001-09-10 22:26 simon * Makefile, TODO, assemble.pl, basic_opcodes.ops, build_interp_starter.pl, disassemble.pl, make_op_header.pl, opcode_table, process_opfunc.pl: Automatic opcode numbering, from Leon Brocard . This means we can add lots more opcodes without fiddling about with the numbers. Hooray! Don't forget that "end" needs to be op zero, else the interpreter falls off the end of the code. This is Bad, and I wonder if there's a way around it. 2001-09-10 19:49 simon * MANIFEST, make_op_table_build.pl: Hey! This doesn't do anything... 2001-09-10 19:46 simon * Makefile: "clean" target cleanup from Bryan C. Warnock . 2001-09-10 18:31 simon * MANIFEST: You know what this means... 2001-09-10 18:30 simon * assemble.pl: "use strict" patch from Leon Brocard . Hey, why wait for a release...? :) 2001-09-10 18:13 simon * interpreter.c: Bad Dan! No cookie! 2001-09-10 17:56 simon * README: A *real* README file, which doesn't claim it isn't for public consumption. 2001-09-10 17:02 simon * TODO: More things to do. 2001-09-10 16:56 dan * TODO: Added in assembler gripe. Also added a few blank lines for clarity 2001-09-10 16:49 dan * register.h, register.c: Renamed the core functions so the opcodes can find 'em OK. (There were some conflicts) 2001-09-10 16:48 dan * basic_opcodes.ops, opcode_table: Added register window ops 2001-09-10 11:07 simon * TODO: Trivial TODO list. 2001-09-10 11:05 simon * test.pbc, test2.pasm, test_main.c: Other string bits and pieces; note "-s" flag to test prog. 2001-09-10 11:01 simon * docs/strings.pod: New string documentation. 2001-09-10 11:01 simon * basic_opcodes.ops, opcode_table: String ops. 2001-09-10 10:53 simon * global_setup.c: More string fu. 2001-09-10 10:51 simon * memory.h: Just some more aliases for now. Used by string handling stuff. 2001-09-10 10:50 simon * assemble.pl, bytecode.c, bytecode.h: Temporary hack to store string constants in bytecode and recreate them at runtime. 2001-09-10 10:48 simon * op.h: op.h is now generated by the Makefile; remove it. 2001-09-10 10:47 simon * Makefile: clean and op.h Makefile targets. 2001-09-10 10:45 simon * string.c, string.h, strnative.c: More string vtable stuff. 2001-09-07 21:07 dan * assemble.pl: Much less picky about whitespace in the opcode_table now. (Doesn't whine on blank lines) 2001-09-07 16:23 dan * Makefile, assemble.pl, global_setup.h, interpreter.c, interpreter.h, opcode_table, parrot.h, test.pasm: (Here's hoping this takes) Added global_setup.h to declare init_world function Moved variable declarations out of the interpreter loop for small speed boost. Changed end opcode to now be opcode #0. A function is provided for it, but isn't ever executed. When the interpreter sees an opcode 0 in the stream it means the end of stream. Things were coring on Cygwin otherwise. (I think I was dereferencing a zero, which is bad) 2001-09-06 22:07 dan * basic_opcodes.ops, process_opfunc.pl: Minor tweaks to basic_opcodes.ops to fix some NV issues. The rest are minor changes now that things are in a fixed state. 2001-09-06 16:13 dan * basic_opcodes.c, test_opcodes.c: test_opcodes.c is now irrelevant--they're all in basic_opcodes.ops. basic_opcodes.c is now autogenerated from basic_opcodes.ops, and should be nuked. So it is. :) 2001-09-06 03:35 dan * basic_opcodes.ops, build_interp_starter.pl, process_opfunc.pl: Scripts to build the header file that holds the function table creation code, as well as the script to turn an opcode file to C source. (And the basic opcode functions turned into an opcode source file) 2001-09-06 03:31 dan * Makefile, interpreter.c: Changes to support the autogeneration of the opcode function pointer table in interpreter.c, along with the changes to the Makefile needed. 2001-09-05 13:27 simon * bytecode.c, config.h, docs/opcodes.pod, docs/overview.pod, docs/parrotbyte.pod, docs/vtables.pod: Added stubs for the documents; working on them this week. (What's left of it.) config.h change is just to type VTABLE, and the bytecode thing you might want to revoke. It's apidoc for the functions in there. 2001-09-03 18:26 simon * docs/strings.pod: A start on the string API documentation. 2001-09-03 17:43 simon * opcode_table: Since we specify the opcode number, we can rearrange to a more human-friendly ordering. 2001-09-03 17:42 simon * basic_opcodes.c: Coupl'a logic bugs. (I think) 2001-09-03 17:41 simon * string.c, string.h, strnative.c: Tidying up the string vtables. 2001-09-03 17:38 simon * make_op_header.pl: Spit out the whole of op.h, not just the bit in the middle. 2001-08-29 13:16 ask * README: it seems like we indeed can commit. What about sending commit mails? 2001-08-29 13:07 simon * Makefile, README, assemble.pl, basic_opcodes.c, bytecode.c, bytecode.h, config.h, disassemble.pl, events.h, exceptions.h, global_setup.c, interpreter.c, interpreter.h, io.h, make_op_header.pl, make_op_table_build.pl, memory.c, memory.h, op.h, opcode_table, parrot.c, parrot.h, register.c, register.h, stacks.h, string.c, string.h, strnative.c, strnative.h, test.pasm, test.pbc, test_main.c, test_opcodes.c: Initial checkin of Simon's work directory - Dan, feel free to nuke. I'm still working on the string stuff, though. 2001-08-29 12:44 ask * README: update README with list address 2001-08-29 12:36 ask * README: first readme.