Dan Sugalski wrote:
We can't devolve to isa checking under the hood, because there are
cases where a class can assert that it has a role without pulling in
the role externally. (Storable, for example, will be a likely thing
here as classes assert they do Storable without pulling in an external
On Wed, May 05, 2004 at 08:12:24PM -0400, Austin Hastings wrote:
: Apparently the AUTOMETH code will be invoked to determine what code matches
: a request. AUTOMETH will return a reference to a matching method. This
: implies that AUTOMETH is expected to resolve dispatch within the class --
: Here
-Original Message-
From: Larry Wall [mailto:[EMAIL PROTECTED]
Sent: Friday, 07 May, 2004 12:40 PM
To: Perl6 Language
Subject: Re: Specifying class interfaces with AUTOMETH
: This seems an awkward way to implement dispatch, since the
: price of using AUTOMETH is being able to write
On Fri, May 07, 2004 at 01:39:43PM -0400, Austin Hastings wrote:
: I agree: classes have a chance to pre-specify the class
: signatures, and objects get a chance to dynamically agree to
: undertake method calls.
:
: The timeframe I was thinking of was object construction time.
I don't see the
sometime ago Andy Lester wrote:
It's been a while since I've heard from any hoplites. Where are you on
your individual Phalanx modules?
recently Tim Bunce wrote:
Encouraging simple status reports, say monthly, may help ensure progress.
ok, I have written 8 tests scripts for a total of 227
On Fri, May 07, 2004 at 08:28:36AM +0200, Dominic Letarte ([EMAIL PROTECTED]) wrote:
ok, I have written 8 tests scripts for a total of 227 tests cases (11
tests cases are presents in the current distribution of PRD). My test
code is covering all the things said in the first half of the pod
On Fri, May 07, 2004 at 09:46:02AM -0500, Andy Lester wrote:
On Fri, May 07, 2004 at 08:28:36AM +0200, Dominic Letarte ([EMAIL PROTECTED]) wrote:
ok, I have written 8 tests scripts for a total of 227 tests cases (11
tests cases are presents in the current distribution of PRD). My test
On Thu, May 06, 2004 at 05:09:03PM -0600, Jim Cromie wrote:
stevan little wrote:
I looked on the site (http://qa.perl.org/phalanx/), but there was
nothing said about how to get involved in this project. Are you
looking for help? And if so, how can I help?
sure. Pick a module from
Encouraging simple status reports, say monthly, may help
ensure progress.
Or...we could set some deadlines...like have 50% of your module tested by
YAPC, and/or possibly have some incentives.
~~Andrew
Encouraging simple status reports, say monthly, may help
ensure progress.
Or...we could set some deadlines...like have 50% of your module tested
by YAPC, and/or possibly have some incentives.
Say ..50%.. of modules tested by YAPC or we cancel the conference?
How about asking YAS if they
Encouraging simple status reports, say monthly, may help ensure
progress.
Or...we could set some deadlines...like have 50% of your
module tested
by YAPC, and/or possibly have some incentives.
Say ..50%.. of modules tested by YAPC or we cancel the conference?
Are you
On Fri, May 07, 2004 at 08:12:32AM -0500, James.FitzGibbon ([EMAIL PROTECTED]) wrote:
Or...we could set some deadlines...like have 50% of your module tested
by YAPC, and/or possibly have some incentives.
How about asking YAS if they would offer reduced price admission if you
flesh out the
Potozniak, Andrew wrote:
Encouraging simple status reports, say monthly, may help ensure
progress.
Or...we could set some deadlines...like have 50% of your
module tested
by YAPC, and/or possibly have some incentives.
Say ..50%.. of modules tested by YAPC or we cancel the
Maybe daily or weekly reminders would be helpful, because I know most of the
time when I go home from work I rarely to almost never thing about the
Phalanx project. I just think there is a huge lack of communication between
Hoplites. It would be nice to have a system where you can ask other
hey all,
Ive been doing hoplite type stuff on B::Concise,
so I figured Id come here, shield spear in hand, to seek some free help.
If you build bleadperl, please apply this patch, and maybe rebuild with
-Uuseperlio.
(sh Configure -des -O -Uuseperlio) I think.
For me, the patch fixes this:
linux
On May 7, 2004, at 1:37 PM, Tim Bunce wrote:
But still no hoplite has actually adopted the DBI...
Since I would like to get involved here, and (with the exception of
HTML::Template, which is already taken) the module I know best in the
100 is DBI. Besides, it seems like Tim really want's someone
stevan little wrote:
On May 7, 2004, at 1:37 PM, Tim Bunce wrote:
But still no hoplite has actually adopted the DBI...
Since I would like to get involved here, and (with the exception of
HTML::Template, which is already taken) the module I know best in the
100 is DBI. Besides, it seems like Tim
On Fri, May 07, 2004 at 03:03:10PM -0400, stevan little wrote:
On May 7, 2004, at 1:37 PM, Tim Bunce wrote:
But still no hoplite has actually adopted the DBI...
Since I would like to get involved here, and (with the exception of
HTML::Template, which is already taken) the module I know
Tim,
I subscribed to dbi-dev and dbi-users both. I will download and install
subversion this weekend (i have been meaning to do this, but putting it
off). And then I will get to work on #1 on your list (convert all tests
to Test::More). I will also keep in mind your larger goal of developing
On Fri, May 07, 2004 at 10:31:49PM +0100, Tim Bunce ([EMAIL PROTECTED]) wrote:
Here's what I'd like to see done soonish:
1. Convert all exisiting test files to Test::More
2. Parts of t/10examp.t should be broken out to separate test scripts
3. Start increasing coverage, especially prior
Jim Cromie wrote:
If you build bleadperl, please apply this patch, and maybe rebuild
with -Uuseperlio.
(sh Configure -des -O -Uuseperlio) I think.
never mind testing this - its now in as 22801
the rest still is open tho.
Andy,
Devel::Cover and I are already very well aquainted, its one of my
favorite tools actually.
steve
On May 7, 2004, at 6:14 PM, Andy Lester wrote:
On Fri, May 07, 2004 at 10:31:49PM +0100, Tim Bunce
([EMAIL PROTECTED]) wrote:
Here's what I'd like to see done soonish:
1. Convert all
On May 6, 2004, at 6:58 AM, Nicholas Clark wrote:
On Tue, May 04, 2004 at 02:55:25PM +0200, Leopold Toetsch wrote:
Bernhard Schmalhofer [EMAIL PROTECTED] wrote:
'libnci.so' is used for testing the native call interface. However I
noticed
that the tests in t/pmc/nci.t were skipped, because
Adam Thomason [EMAIL PROTECTED] wrote:
amd64 breaks. Shared libraries must be compiled with -fPIC; see
http://www.x86-64.org/lists/discuss/msg02621.html for why. LD_SHARED
has the flag, but nci_test.o isn't compiled with it.
Well, then we have to add $(LD_SHARED).
needs to be a CC_SHARED
Joshua Gatcomb [EMAIL PROTECTED] wrote:
Since I am unable to help with the C part of things, I
figured I could at least track
down where the problem was for everyone. I have
isolated the ARENA_DOD_FLAGS problem
on Cygwin to a change committed to CVS between
2004-04-15 10:00 and 10:15 EDT.
Nicholas Clark [EMAIL PROTECTED] wrote:
So I don't know what to do next.
Difficult. Set a breakpoint (after the Cdlfunc) has executed at
nci_test.c:nci_pi, check any used pointer addresses from C code if they
match somewhere in that PMC.
And: Does it run with:
parrot -G t/pmc/nci_23.pasm
Andrew Dougherty [EMAIL PROTECTED] wrote:
Off the top of my head, I can think of three instances where building
without dynamic loading is relevant:
Ok. We need some more Configure support for this. First: if
--disable-shared or such is set on Configure., we should disable dynamic
targets in
Jeff Clites [EMAIL PROTECTED] wrote:
1707# if ! DISABLE_GC_DEBUG
1708/* It's easy to forget that string comparison can trigger
GC */
1709if (GC_DEBUG(interpreter))
1710Parrot_do_dod_run(interpreter, DOD_trace_stack_FLAG);
1711# endif
Basically, when
# New Ticket Created by Adam Thomason
# Please include the string: [perl #29402]
# in the subject line of all future correspondence about this issue.
# URL: http://rt.perl.org:80/rt3/Ticket/Display.html?id=29402
The recently posted sha1 and md5 programs produce incorrect answers on amd64,
Nicholas Clark [EMAIL PROTECTED] wrote:
0x0002e354 in pobject_lives (interpreter=0x1000200, obj=0xd13a68)
Ok, that's exactly:
#5 0x001ae474 in Parrot_NCI_set_string_keyed (interpreter=0x1000200, \
pmc=0x100f6a0,
Nicholas Clark [EMAIL PROTECTED] wrote:
#5 0x001ae474 in Parrot_NCI_set_string_keyed (interpreter=0x1000200, pmc=0x100f6a0,
func=0xd13a68, value=0x1001ba8) at classes/nci.c:86
Ok. Fixed. And this test is removed. It was bogus anyway. The next test
has the right way to attach a callback info.
Adam Thomason [EMAIL PROTECTED] wrote:
The recently posted sha1 and md5 programs produce incorrect answers on
amd64, at least in part because bitops on 64-bit values are screwy.
One particularly odd case:
_main:
set I1, 0x
shl I1, I1, 32
set I2, 0x
Joshua Gatcomb [EMAIL PROTECTED] wrote:
#4 0x00423977 in enter_nci_method
(interpreter=0x100d1d68, type=26, func=0x481670,
name=0x53f8d9 thread1, proto=0x53f8d4 vIOP) at
Does the recent change related to NCI cure the problem?
leo
Leopold Toetsch [EMAIL PROTECTED] wrote:
Dan Sugalski [EMAIL PROTECTED] wrote:
... And... we
move *all* the operator functions out of the vtable and into the MMD
system. All of it.
This *all* includes vtable functions like add_int() or add_float() too,
I presume. For these we have left
On Fri, May 07, 2004 at 11:56:10AM +0200, Leopold Toetsch wrote:
Nicholas Clark [EMAIL PROTECTED] wrote:
0x0002e354 in pobject_lives (interpreter=0x1000200, obj=0xd13a68)
Ok, that's exactly:
#5 0x001ae474 in
Nicholas Clark [EMAIL PROTECTED] wrote:
Yes, the fix you committed works. Thanks:
Welcome
All tests successful, 1 test and 54 subtests skipped.
Fine
However, I still have to make libnci.dylib by hand. I'm not sure of the best
way to integrate things for that into the Configure system.
Leopold Toetsch [EMAIL PROTECTED] writes:
Dan Sugalski [EMAIL PROTECTED] wrote:
At 11:35 AM +0200 4/30/04, Leopold Toetsch wrote:
Dan Sugalski [EMAIL PROTECTED] wrote:
If we go MMD all the way, we can skip the bytecode-C-bytecode
transition for MMD functions that are written in parrot
Joshua Gatcomb wrote:
I am not sure where to go from here. Any suggestions?
Ok, here is a sample debugging session:
$ cat hello.pasm
print hello\n
end
$ parrot hello.pasm
hello
$ gdb parrot
...
(gdb) b new_pmc_header
(gdb) r hello.pasm
Breakpoint 1, new_pmc_header (interpreter=0x82e5668,
Piers Cawley [EMAIL PROTECTED] wrote:
Leopold Toetsch [EMAIL PROTECTED] writes:
- if it calls a PASM routine, registers have to be preserved. Which
registers depend on the subroutine that actually gets called (ok, this
information - which registers are changed by the sub - can be attached
Joshua Gatcomb [EMAIL PROTECTED] wrote:
Program received signal SIGSEGV, Segmentation fault.
0x00419f6c in new_pmc_header (interpreter=0x100d1d68,
flags=1024) at src/headers.c:251
251 *((Dead_PObj*)pmc)-arena_dod_flag_ptr
One more idea: When you set a breakpoint at
--- Leopold Toetsch [EMAIL PROTECTED] wrote:
#4 0x00423977 in enter_nci_method
(interpreter=0x100d1d68, type=26, func=0x481670,
name=0x53f8d9 thread1, proto=0x53f8d4
vIOP) at
Does the recent change related to NCI cure the
problem?
leo
Except for a few of the numbers changing, I
--- Leopold Toetsch [EMAIL PROTECTED] wrote:
...
So the problem should just be in two files:
headers.c
pmc.c
The patch moved PMC creation code from pmc.c to
headers.c.
The Carena_dod_flag_ptr is set in
Cadd_to_free_list. I don't see a
way, why it can be NULL then.
I must miss
--- Leopold Toetsch [EMAIL PROTECTED] wrote:
Joshua Gatcomb wrote:
I am not sure where to go from here. Any
suggestions?
Ok, here is a sample debugging session:
$ cat hello.pasm
print hello\n
end
$ parrot hello.pasm
hello
$ gdb parrot
...
(gdb) b new_pmc_header
(gdb)
Apparently it's not happy with things of the form
foo = bar * .95
where the RHS of the binary operation is a floating point constant
with no integer portion. Changing it to 0.95 works, so I assume the
grammar just needs a tweak.
--
Dan
Joshua Gatcomb wrote:
(gdb) p *pool
$1 = {last_Arena = 0x1020, object_size = 32,
objects_per_alloc = 16382, total_objects = 2048,
num_free_objects = 2047, skip = 0, replenish_level =
614, free_list = 0x1021, align_1 = 0,
add_free_object = 0x4700f0 add_free_object,
get_free_object =
Dan Sugalski [EMAIL PROTECTED] wrote:
Apparently it's not happy with things of the form
foo = bar * .95
where the RHS of the binary operation is a floating point constant
with no integer portion. Changing it to 0.95 works, so I assume the
grammar just needs a tweak.
Yep. The lexer
Joshua Gatcomb [EMAIL PROTECTED] wrote:
--- Leopold Toetsch [EMAIL PROTECTED] wrote:
Seems that you got sizeof(UINTVAL) == 8 ???
(gdb) p sizeof(PMC)
$1 = 32
(gdb) p sizeof(Dead_PObj)
$2 = 40
(gdb) p sizeof(void *)
$3 = 4
(gdb) p sizeof(UINTVAL)
$4 = 8
Aargh yes. So here we go:
$ make
--- Leopold Toetsch [EMAIL PROTECTED] wrote:
Aargh yes. So here we go:
$ make realclean
$ perl Configure.pl --intval=long --opcode=long
This was the ticket Thanks. For my own
edification, can you explain to me what in the code I
isolated that changed on the 15th accounts for this?
$ make realclean
$ cvs update -dP
$ perl Configure.pl --intval=long --opcode=long
--icuheaders=/usr/local/include --icushared='-licuuc
-licudt'
$ make
$ make test
All tests successful, 2 tests and 57 subtests skipped.
Files=102, Tests=1522, 395 wallclock secs (230.74 cusr
+ 149.40 csys =
I'm trying to find documentation on how to write PMCs. I can find lots of
PMCs in the source tree, but I can't seem to find documentation on how to
go about writing them, or what the various fields and flags do.
For example, the string PMC_data occurs in many source files, but as far
as I can
On May 7, 2004, at 1:13 AM, Leopold Toetsch wrote:
Jeff Clites [EMAIL PROTECTED] wrote:
... which is trashing the strings we've just made (and are
in the middle of using) in build_call_func:
2925if (!string_compare(interpreter, signature,
2926 string_from_cstring(interpreter,
In addition to now passing all tests, parrot is
lightening fast in comparison to what it was only a
few short hours ago. What a difference after:
--intval=long --opcode=long
$ time parrot primes2.pasm
N primes calculated to 5000 is 670
last is: 4999
real0m4.148s
user0m3.905s
sys
With all the perl scalars now indirecting through PMCs, Ponie's performance
has dropped. Not surprising, in the general case, but the performance of
lib/locale.t is vile, and I'm wondering if that regression test hits
something pathological. It slows to a crawl after subtest 99 (line 240 in
On Fri, 7 May 2004, Nicholas Clark wrote:
I'm trying to find documentation on how to write PMCs. I can find lots of
PMCs in the source tree, but I can't seem to find documentation on how to
go about writing them, or what the various fields and flags do.
For example, the string PMC_data
On May 7, 2004, at 5:04 AM, Leopold Toetsch wrote:
Nicholas Clark [EMAIL PROTECTED] wrote:
However, I still have to make libnci.dylib by hand. I'm not sure of
the best
way to integrate things for that into the Configure system.
Your patch is almost ok. But I think dependencies are still wrong -
On Fri, 7 May 2004, Simon Glover wrote:
On Fri, 7 May 2004, Nicholas Clark wrote:
I'm trying to find documentation on how to write PMCs. I can find lots of
PMCs in the source tree, but I can't seem to find documentation on how to
go about writing them, or what the various fields and flags
Joshua Gatcomb [EMAIL PROTECTED] wrote:
--- Leopold Toetsch [EMAIL PROTECTED] wrote:
$ make realclean
$ perl Configure.pl --intval=long --opcode=long
This was the ticket Thanks.
Great.
For my own
edification, can you explain to me what in the code I
isolated that changed on the
There are entries shown, which are resolved and applied, e.g.
#29034. Why is that line still there?
leo
At 12:54 PM +0200 5/7/04, Leopold Toetsch wrote:
Leopold Toetsch [EMAIL PROTECTED] wrote:
Dan Sugalski [EMAIL PROTECTED] wrote:
... And... we
move *all* the operator functions out of the vtable and into the MMD
system. All of it.
This *all* includes vtable functions like add_int() or
For some reason this change to src/headers.c breaks on OS X. I'm not sure
why:
@@ -244,6 +245,7 @@
pool = flags PObj_constant_FLAG ?
interpreter-arena_base-constant_pmc_pool :
interpreter-arena_base-pmc_pool;
+assert(sizeof(Dead_PObj) = sizeof(PMC));
pmc =
Nicholas Clark [EMAIL PROTECTED] wrote:
With all the perl scalars now indirecting through PMCs, Ponie's performance
has dropped. Not surprising, in the general case, but the performance of
lib/locale.t is vile, and I'm wondering if that regression test hits
something pathological. It slows to
Nicholas Clark [EMAIL PROTECTED] wrote:
I'm trying to find documentation on how to write PMCs. I can find lots of
PMCs in the source tree, but I can't seem to find documentation on how to
go about writing them, or what the various fields and flags do.
Well, there isn't too much documentation
At 4:58 PM +0200 5/7/04, Leopold Toetsch wrote:
Dan Sugalski [EMAIL PROTECTED] wrote:
Apparently it's not happy with things of the form
foo = bar * .95
where the RHS of the binary operation is a floating point constant
with no integer portion. Changing it to 0.95 works, so I assume the
At 5:16 PM +0100 5/7/04, Nicholas Clark wrote:
I built ponie with profiling, and I find that it's spending an inordinate
amount of time in trace_children (half a second per call):
Wow, that's... bad. It seems like you've an insane number of PMCs,
though, since there were over 1.3 billion calls to
Dan Sugalski [EMAIL PROTECTED] wrote:
It'd help if I could actually rebuild this so it could be tested...
Works. Applied. Tests ok. Thanks,
leo
Larry Wall [EMAIL PROTECTED] wrote:
[I'm assuming that these internal signatures contain actual type
No. They are really internal, denoting C data types: 'i' = int, 'd' =
double and so on. These signature just happen to live in STRINGs so that
we can pass them around e.g. to VTABLEs.
Larry
Nicholas Clark wrote:
I built ponie with profiling, and I find that it's spending an inordinate
amount of time in trace_children (half a second per call):
One more question: Looking at the test, there seems to be a lot of stuff
collected in @Locale: How do these PMCs look like? And how many
Nicholas Clark [EMAIL PROTECTED] wrote:
For some reason this change to src/headers.c breaks on OS X. I'm not sure
why:
Ah yes. Needs
#if ARENA_DOD_FLAGS
around. Fixed
Thanks,
leo
On Fri, 2004-05-07 at 11:13, Leopold Toetsch wrote:
Larry Wall [EMAIL PROTECTED] wrote:
[I'm assuming that these internal signatures contain actual type
No. They are really internal, denoting C data types: 'i' = int, 'd' =
double and so on. These signature just happen to live in STRINGs
On Fri, May 07, 2004 at 06:57:24PM +0200, Leopold Toetsch wrote:
Nicholas Clark [EMAIL PROTECTED] wrote:
With all the perl scalars now indirecting through PMCs, Ponie's performance
has dropped. Not surprising, in the general case, but the performance of
lib/locale.t is vile, and I'm
On May 7, 2004, at 11:48 AM, chromatic wrote:
On Fri, 2004-05-07 at 11:13, Leopold Toetsch wrote:
Larry Wall [EMAIL PROTECTED] wrote:
[I'm assuming that these internal signatures contain actual type
No. They are really internal, denoting C data types: 'i' = int, 'd'
=
double and so on. These
On Fri, May 07, 2004 at 08:17:05PM +0200, Leopold Toetsch wrote:
Ah yes. Needs
#if ARENA_DOD_FLAGS
around. Fixed
Thanks. Everything compiles again now, and all tests pass
(after I do the little .so - .dylib dance)
Nicholas Clark
As Nick seems to be hitting some limits, I've tried this:
stuff a differing amount of PMCs into an Array and time a DOD run:
$ parrot -j 10m.pasm
create 10 PerlInts 0.064826
DOD sweeps: 4 one is0.013948
$ parrot -j 10m.pasm
create 50 PerlInts 0.698430
DOD sweeps: 12 one is 0.070052
I apoligize if this is a duplicate, at least 3
messages I have sent to the list have not made it to
the archive after about 6 hours.
1. I reported that extend test 13 was hanging just
like test 12 was previously requiring a kill for the
test suite to continue. You can disregard - see point
2.
-Original Message-
From: Leopold Toetsch [mailto:[EMAIL PROTECTED]
Sent: Friday, May 07, 2004 3:59 AM
To: [EMAIL PROTECTED]
Subject: Re: [perl #29402] [BUG] 64-bit bitops
One particularly odd case:
_main:
set I1, 0x
shl I1, I1, 32
set I2,
On Mon, May 03, 2004 at 10:46:28AM -0400, Dan Sugalski wrote:
Okay, here's the rules for PMCs that live outside parrot, and calling
into parrot from the outside.
1) *ALL* PMCs which are created outside parrot must be registered
(unless they're otherwise anchored)
2) No call into parrot
At 9:42 PM +0200 5/7/04, Leopold Toetsch wrote:
As Nick seems to be hitting some limits, I've tried this:
stuff a differing amount of PMCs into an Array and time a DOD run:
[Progressively nastier numbers snipped]
Athlon 800, optimized build. These 78 DOD runs all taking half a
second are of
Nicholas Clark writes:
Thanks. Everything compiles again now, and all tests pass
(after I do the little .so - .dylib dance)
I'm new to OS X. Might you describe said dance?
Luke
On Fri, May 07, 2004 at 03:11:54PM -0600, Luke Palmer wrote:
Nicholas Clark writes:
Thanks. Everything compiles again now, and all tests pass
(after I do the little .so - .dylib dance)
I'm new to OS X. Might you describe said dance?
We haven't resolved how to best patch the Makefile
Nicholas Clark [EMAIL PROTECTED] wrote:
On Fri, May 07, 2004 at 06:57:24PM +0200, Leopold Toetsch wrote:
I tried building with -DDISABLE_GC_DEBUG=1 and the build fails:
./parrot parrot-config.imc VERSION DEVEL
parrot: src/platform.c:665: Parrot_memcpy_aligned_sse_debug: Assertion `((unsigned
On Fri, May 07, 2004 at 10:09:10PM +0200, Leopold Toetsch wrote:
Nicholas Clark [EMAIL PROTECTED] wrote:
The PMC only used the PMC_data member to store a pointer on to a perl5
SV body.
That's suboptimal :) Please have a look at the timings I posted a minute
ago. Stuff the pointer into
81 matches
Mail list logo
