Re: mem_allocate_aligned

2001-12-11 Thread Jonathan Stowe
On Mon, 10 Dec 2001, Dan Sugalski wrote: We need an allocate zeroed call too. We can get zeroed pages faster in some cases than we can allocate pages and zero them ourselves. Isn't that calloc() ? /J\ -- Jonathan Stowe | http://www.gellyfish.com | This

Just-in-Time Predereferencing Patch

2001-12-11 Thread Gregor N. Purdy
All -- I am attaching a .patch file that contains the patch, and a .txt file that describes the patch. Have fun, but be sure you are home by 11. Regards, -- Gregor _ /Inspiration Innovation Excellence (TM)

Re: Just-in-Time Predereferencing Patch

2001-12-11 Thread Simon Cozens
On Tue, Dec 11, 2001 at 10:09:07AM -0500, Gregor N. Purdy wrote: -runops_generic(core, interpreter, pc); +if ((interpreter-flags PARROT_PREDEREF_FLAG) != 0) { + size_t offset = pc - (opcode_t *)interpreter-code-byte_code; + + if (!interpreter-prederef_code)

Re: Just-in-Time Predereferencing Patch

2001-12-11 Thread Gregor N. Purdy
Simon -- I haven't tried this, so feel free to shoot me down, but is your 22% speed-up based on an unpatched Parrot, or a patched Parrot without the -P flag? I'm trying to work out whether or not this patch actually slows down the general case, which wouldn't be pleasant. Based on patched

Re: Any last requests?

2001-12-11 Thread Andy Dougherty
In perl.perl6.internals, you wrote: Unless I hear howls of pain to the contrary, Parrot 0.0.3 will be released in the next few hours. So if there's anything life-threatening that needs changing, let me know now. Oh, and I renamed test_prog, to see who's watching. Anyone on a Case-insensitive

Default blocksize

2001-12-11 Thread melvin.smith
On most UNIX flavors I can use the stat() system call and the {struct stat}.st_blksize member to get an idea of the blocksize of a regular file. I don't currently have access to anything but Linux, Solaris and AIX to check this one, and I'm especially not sure about VMS, but for now I'm using

Re: Any last requests?

2001-12-11 Thread Gregor N. Purdy
All -- I suspect the fix is to mkdir lib mv Parrot Test lib (with the appropriate MANIFEST corrections and CVS incantations) If this counts as a motion, I second the motion, assuming that counts. Regards, -- Gregor

Re: gc object destruction

2001-12-11 Thread Benoit Cerrina
just to make a point what you are talking about is not destruction (as in c++) but finalization (as in any gced language) the difference being in the determinism. Destruction is almost always necessary in c++ since the memory allocated need to be deallocated. Like you justly specify finalization

Re: mem_allocate_aligned

2001-12-11 Thread Dan Sugalski
At 08:40 AM 12/11/2001 +, Jonathan Stowe wrote: On Mon, 10 Dec 2001, Dan Sugalski wrote: We need an allocate zeroed call too. We can get zeroed pages faster in some cases than we can allocate pages and zero them ourselves. Isn't that calloc() ? Yep, if we're using the malloc family

Re: Any last requests?

2001-12-11 Thread Simon Cozens
On Tue, Dec 11, 2001 at 02:10:45PM -0500, Gregor N. Purdy wrote: mkdir lib mv Parrot Test lib (with the appropriate MANIFEST corrections and CVS incantations) If this counts as a motion, I second the motion, assuming that counts. *nodnod*. Do it. Or I'll do it in the morning. Up

untested ops

2001-12-11 Thread Alex Gough
I've added a lot of missing tests to the test suite which has improved coverage. We now have 205 tests which ought to run. gripe type=we've done this beforeuse $(PERL) in Makefiles/gripe Oh, and Windows is ON FIRE! The following ops do not yet have tests: close_i close_ic err_i err_s