Ping.
The implementation is pretty straightforward, and doesn't yet match
perl5's implementation.
Inserting a key:
1) Hash the key to a number, and fold it to between 0 and NUM_BUCKETS.
2) Create a new bucket, and save both the value and key in it.
3) Tack the bucket onto the appropriate hash index.
On Mon, Jan 07, 2002 at 11:03:05PM +, Nicholas Clark wrote:
> --- include/parrot/parrot.h~ Sat Dec 15 14:21:03 2001
> +++ include/parrot/parrot.h Mon Jan 7 21:56:40 2002
> @@ -68,6 +68,11 @@
> typedef void* BIGINT;
> typedef void* BIGFLOAT;
>
> +/* define a macro to acknowledge an unus
The patch below adds the ability to say:
../test_parrot -f - < foo.pbc
(Yes, an interactive bytecode interpreter, just what you've been
waiting for)
and
../test_parrot -f foo.pbc
And have them both do the same thing.
This seems to work fine on FreeBSD and Irix, but might be very wrong
somewh
At 09:34 PM 1/7/2002 +, Nicholas Clark wrote:
>I believe that this patch gets all of these:
>
>include/parrot/key.h:51: warning: declaration of `index' shadows global
>declaration
Applied, thanks.
Dan
--"it's like
If I've got this right, then this patch silences all warnings about unused
variables, except for those from the ops files.
This may not be the preferred way to do it, but it does flag where all the
unused parameters currently are.
Nicholas Clark
--- include/parrot/parrot.h~Sat Dec 15 14:21:
At 02:06 PM 1/7/2002 -0800, Brent Dax wrote:
>stdio is pretty std, right? That ought to be enough for Configure--we
>can compile a stdio.ops into miniparrot. I doubt we'll need the full
>ParrotIO subsystem just for building. The most complicated thing we'll
>have to do is probably backticks, an
Dave Mitchell:
# Simon Cozens <[EMAIL PROTECTED]> wrote:
# > IANDan, but here's the plan as I see it.
# >
# > 1) Move to using make.pl instead of system's make
# > 2) Compile both make.pl and Configure.pl to Parrot bytecode
# > 3) Create a miniparrot which builds anywhere and can
# run
Hi.
I read vtables.pod about making your own classes. It explains the 7 steps
neccessary to get your new class compiled in after you've written the
..pmc, .c and .h files--if it is a build-in class.
I am just trying to implement a gsl_complex class for complex numbers, but
I don't know how to ma
I believe that this patch gets all of these:
include/parrot/key.h:51: warning: declaration of `index' shadows global declaration
include/parrot/key.h:52: warning: declaration of `index' shadows global declaration
include/parrot/key.h:53: warning: declaration of `index' shadows global declaration
At 12:39 PM 1/7/2002 -0800, Steve Fink wrote:
>There was a bug in that recent patch to allow different .ops files to
>be included while Configuring. Makefile.in has:
>
>OPS_FILES = ${ops} $(GEN_OPSFILES)
>
>where GEN_OPSFILES=vtable.ops, and Configure.pl fills in ${ops} with a
>list which also inc
At 01:25 PM 1/7/2002 +, Dave Mitchell wrote:
>Simon Cozens <[EMAIL PROTECTED]> wrote:
> > IANDan, but here's the plan as I see it.
> >
> > 1) Move to using make.pl instead of system's make
> > 2) Compile both make.pl and Configure.pl to Parrot bytecode
> > 3) Create a miniparrot wh
cc -Wall -I./include -DHAS_JIT -o global_setup.o -c global_setup.c
global_setup.c: In function `init_world':
global_setup.c:29: warning: implicit declaration of function
`Parrot_PerlUndef_class_init'
no warning with appended
Nicholas Clark
--- global_setup.c~ Sat Jan 5 12:49:15 2002
++
At 06:39 PM 1/7/2002 +, Simon Glover wrote:
> Currently, whenever we create a new string with a call to string_make,
> we allocate the memory for it by a simple call to malloc (via
> mem_sys_allocate in memory.c). However, we don't check the return value
> of this call, so if the malloc f
At 02:53 AM 1/7/2002 -0800, Brent Dax wrote:
>This patch modifies Parrot::Test to goto the Test::More subs, thereby
>giving you _useful_ information on where your problem was.
>Unfortunately, this precludes unlinking the files only if the test was
>unsuccessful; if you need to see those files, you
At 09:57 PM 1/6/2002 -0800, Boris Tschirschwitz wrote:
>Hi.
>
>I just implemented some opcodes for the 'GNU Scientific Library'.
>The opcodes are in the file 'gsl.ops', and tests are in 'gsl.t'
>These are very few and rather uninteresting ops, I was mostly checking if
>I know how to do it.
>I subm
At 05:09 PM 1/6/2002 -0500, coral wrote:
>I.. think.. that you would want to insert a comma directly after '(more
>on that later)'. I think. Or maybe it's directly before the
>parentheses. Never did get that part of grammar down. But it needs a
>comma! :)
Added, thanks.
At 08:12 PM 1/7/2002 +, Simon Glover wrote:
> Subject line says it all.
Applied, thanks.
Dan
--"it's like this"---
Dan Sugalski even samurai
[EMAIL PROTECTED]
On Mon, Jan 07, 2002 at 09:27:50AM +, Simon Cozens wrote:
> On Sun, Jan 06, 2002 at 03:20:27PM -1000, David & Lisa Jacobs wrote:
> > I recently did a build from scratch and noticed that vtable.ops,
> > vtable_ops.c etc are not being generated. Is there something new I need to
> > do in the co
Subject line says it all.
Simon
--- global_setup.c.old Sat Jan 5 14:00:01 2002
+++ global_setup.c Mon Jan 7 20:07:20 2002
@@ -15,6 +15,7 @@
#include "parrot/parrot.h"
/* Needed because this might get compiled before pmcs have been built */
+void Parrot_PerlUndef_class_init(void);
On Sun, 6 Jan 2002, Bryan C. Warnock wrote:
> New File: KNOWN_ISSUES
>
> Here's the list of things that we know are broken, or
> are admittedly hackish stopgap measures.
>
> - 'long doubles' on x86 don't work. It's a packfile
Currently, whenever we create a new string with a call to string_make,
we allocate the memory for it by a simple call to malloc (via
mem_sys_allocate in memory.c). However, we don't check the return value
of this call, so if the malloc fails, Parrot promptly segfaults
when we try to make the
On Mon, Jan 07, 2002 at 01:43:49PM +, Alex Gough wrote:
> I've found a different problem with hashing and \0 in strings
Fixed.
--
God gave man two ears and one tongue so that we listen twice as much as
we speak.
-- Arab proverb
On Mon, Jan 07, 2002 at 01:43:49PM +, Alex Gough wrote:
> I tried to make this segfault, but couldn't quite manage
You won't. Electric Fence deliberately segfaults as early as it
can to stop silent corruption of the malloc arena later on.
--
"He was a modest, good-humored boy. It was Oxfor
On Mon, 7 Jan 2002, Simon Cozens wrote:
> PMC test 57 seems to work only by sheer dumb luck, and running
> it through Electric Fence highlights a mallocing problem.
I tried to make this segfault, but couldn't quite manage, in the
process I've found a different problem with hashing and \0 in
stri
Simon Cozens <[EMAIL PROTECTED]> wrote:
> IANDan, but here's the plan as I see it.
>
> 1) Move to using make.pl instead of system's make
> 2) Compile both make.pl and Configure.pl to Parrot bytecode
> 3) Create a miniparrot which builds anywhere and can run the above
>programs
PMC test 57 seems to work only by sheer dumb luck, and running
it through Electric Fence highlights a mallocing problem.
In some of the the key_element_* functions in key.c, we have
code like this:
if(key != NULL) {
INTVAL hash = key_hash(interpreter,index);
hash = hash % NUM_BUCKETS;
This patch modifies Parrot::Test to goto the Test::More subs, thereby
giving you _useful_ information on where your problem was.
Unfortunately, this precludes unlinking the files only if the test was
unsuccessful; if you need to see those files, you can just set
$ENV{POSTMORTEM} and they'll be lef
On Sun, Jan 06, 2002 at 03:20:27PM -1000, David & Lisa Jacobs wrote:
> I recently did a build from scratch and noticed that vtable.ops,
> vtable_ops.c etc are not being generated. Is there something new I need to
> do in the config or is it broken.
I thought these days it gets automatically fold
Just to let you guys know, 'nmake test' still causes a crash in test
t/op/pmc:44 here on Win32.
Details at http://www.sebastian-bergmann.de/parrot.txt
--
Sebastian Bergmann
http://sebastian-bergmann.de/ http://phpOpenTracker.de/
Did I help you? Consider a gift: http
30 matches
Mail list logo
