From usenet at andreas-s.net  Fri Feb  4 14:08:09 2005
From: usenet at andreas-s.net (Andreas Schwarz)
Date: Fri Feb  4 14:05:10 2005
Subject: [Yarv-devel] Status of YARV?
Message-ID: <4203C819.3040003@andreas-s.net>

Hi,

does YARV support
- C extensions
- (native) threads
- file IO
- sockets?

Andreas

From ko1 at atdot.net  Fri Feb  4 14:20:31 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Fri Feb  4 14:19:25 2005
Subject: [Yarv-devel] Status of YARV?
In-Reply-To: (Your message of "Fri, 04 Feb 2005 20:08:09 +0100")
	<4203C819.3040003@andreas-s.net>
References: <4203C819.3040003@andreas-s.net>
Message-ID: <20050205042031.21e8742a%ko1@atdot.net>

  Andreas Schwarz <usenet@andreas-s.net> wrote :
    [ [Yarv-devel] Status of YARV? ]
    at Fri, 04 Feb 2005 20:08:09 +0100

Hi,

> does YARV support
> - C extensions

Yes.

> - (native) threads

No. I want to support it sometime.

> - file IO

Yes. (same as C extensions)

> - sockets?

Yes. (same as C extensions)

Regards,
-- 
// SASADA Koichi at atdot dot net
//
From usenet at andreas-s.net  Sat Feb  5 12:20:02 2005
From: usenet at andreas-s.net (Andreas Schwarz)
Date: Sat Feb  5 12:28:49 2005
Subject: [Yarv-devel] Benchmark results on PPC
Message-ID: <cu2v6a$486$1@sea.gmane.org>

Powerbook G4 1.33 GHz

ruby 1.9.0 powerpc-darwin7.7.0(2005-02-05)
YARVCore 0.1.0 rev: 126 (2005-01-25)
[direct threaded code] [optimize basic operation] [optimize regexp 
match] [stack caching] [inline method cache]

-----------------------------------------------------------
array:
i=0
while i<10000000
i+=1
a = [1,2,3,4,5,6,7,8,9,10]
end
--
       user     system      total        real
ruby 50.000000   0.030000  50.030000 ( 51.988689)
yarv 12.620000   0.070000  12.690000 ( 13.058565)
-----------------------------------------------------------
block:
def m
yield
end

i=0
while i<10000000
i+=1
m{
}
end
--
       user     system      total        real
ruby 35.700000   0.020000  35.720000 ( 36.538383)
yarv  4.210000   0.000000   4.210000 (  4.318696)
-----------------------------------------------------------
const:
Const = 1

i = 0
while i < 10000000
i+= 1
j = Const
end
--
       user     system      total        real
ruby 22.440000   0.020000  22.460000 ( 22.997802)
yarv  1.510000   0.000000   1.510000 (  1.549931)
-----------------------------------------------------------
ensure:
i=0
while i<1000000
i+=1
begin
begin
ensure
end
ensure
end
end

--
       user     system      total        real
ruby  1.930000   0.010000   1.940000 (  2.012598)
yarv  0.120000   0.000000   0.120000 (  0.123234)
-----------------------------------------------------------
factorial:
def fact(n)
if(n > 1)
n * fact(n-1)
else
1
end
end

fact(7300)
--
       user     system      total        real
ruby    stack level too deep
yarv  0.240000   0.030000   0.270000 (  0.274964)
-----------------------------------------------------------
fib:
def fib n
if n < 3
1
else
fib(n-1) + fib(n-2)
end
end

fib(32)

--
       user     system      total        real
ruby 11.960000   0.010000  11.970000 ( 12.257111)
yarv  1.250000   0.000000   1.250000 (  1.277800)
-----------------------------------------------------------
lists:
#from http://www.bagley.org/~doug/shootout/bench/lists/lists.ruby

NUM = 10
SIZE = 10000

def test_lists()
# create a list of integers (Li1) from 1 to SIZE
li1 = (1..SIZE).to_a
# copy the list to li2 (not by individual items)
li2 = li1.dup
# remove each individual item from left side of li2 and
# append to right side of li3 (preserving order)
li3 = Array.new
while (not li2.empty?)
li3.push(li2.shift)
end
# li2 must now be empty
# remove each individual item from right side of li3 and
# append to right side of li2 (reversing list)
while (not li3.empty?)
li2.push(li3.pop)
end
# li3 must now be empty
# reverse li1 in place
li1.reverse!
# check that first item is now SIZE
if li1[0] != SIZE then
p "not SIZE"
0
else
# compare li1 and li2 for equality
if li1 != li2 then
return(0)
else
# return the length of the list
li1.length
end
end
end

i = 0
while i<NUM
i+=1
result = test_lists()
end

result
--
       user     system      total        real
ruby  0.500000   0.000000   0.500000 (  0.516968)
yarv  0.190000   0.020000   0.210000 (  0.225273)
-----------------------------------------------------------
method:
def m
end

i=0
while i<10000000
i+=1
m
end
--
       user     system      total        real
ruby 26.470000   0.030000  26.500000 ( 27.273037)
yarv  2.910000   0.000000   2.910000 (  3.000279)
-----------------------------------------------------------
poly_method:
class C1
def m
1
end
end
class C2
def m
2
end
end

o1 = C1.new
o2 = C2.new

i=0
while i< 10000000
o = (i % 2 == 0) ? o1 : o2
o.m
i+=1
end
--
       user     system      total        real
ruby 47.140000   0.050000  47.190000 ( 48.338530)
yarv  7.130000   0.000000   7.130000 (  7.164417)
-----------------------------------------------------------
proc:
def m &b
b
end

pr = m{
a = 1
}

i=0
while i<1000000
i+=1
pr.call
end

--
       user     system      total        real
ruby  5.210000   0.000000   5.210000 (  5.268402)
yarv  0.920000   0.000000   0.920000 (  0.935314)
-----------------------------------------------------------
reccount:
def reccount n
if n > 1
1 + reccount(n-1)
else
1
end
end

reccount 7000

--
       user     system      total        real
ruby    stack level too deep
yarv  0.010000   0.000000   0.010000 (  0.005373)
-----------------------------------------------------------
regexp:
i=0
while i<1000000
/hoge/ =~ 'xxxhogexxx'
i+=1
end
--
       user     system      total        real
ruby  3.340000   0.000000   3.340000 (  3.363574)
yarv  1.280000   0.000000   1.280000 (  1.295496)
-----------------------------------------------------------
rescue:
i=0
while i<10000000
i+=1
begin
rescue
end
end
--
       user     system      total        real
ruby 23.150000   0.010000  23.160000 ( 23.330190)
yarv  1.490000   0.010000   1.500000 (  1.509364)
-----------------------------------------------------------
rescue2:
i=0
while i<100000
i+=1
begin
raise
rescue
end
end
--
       user     system      total        real
ruby  2.530000   0.190000   2.720000 (  2.743173)
yarv  1.810000   0.220000   2.030000 (  2.049670)
-----------------------------------------------------------
simpleiter:
1000000.times{|simpleiter|
simpleiter
}
--
       user     system      total        real
ruby  1.130000   0.000000   1.130000 (  1.122475)
yarv  0.470000   0.000000   0.470000 (  0.473542)
-----------------------------------------------------------
simplereturn:
def m
return 1
end
i=0
while i<1000000
i+=1
m
end

--
       user     system      total        real
ruby  2.720000   0.000000   2.720000 (  2.728355)
yarv  0.300000   0.000000   0.300000 (  0.299973)
-----------------------------------------------------------
so_ackermann:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: ackermann-ruby.code,v 1.4 2004/11/13 07:40:41 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/

def ack(m, n)
if m == 0 then
n + 1
elsif n == 0 then
ack(m - 1, 1)
else
ack(m - 1, ack(m, n - 1))
end
end

NUM = 7
ack(3, NUM)


--
       user     system      total        real
ruby    stack level too deep
yarv  0.340000   0.000000   0.340000 (  0.338907)
-----------------------------------------------------------
so_array:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: ary-ruby.code,v 1.4 2004/11/13 07:41:27 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/
# with help from Paul Brannan and Mark Hubbart

n = 9000 # Integer(ARGV.shift || 1)

x = Array.new(n)
y = Array.new(n, 0)

n.times{|bi|
x[bi] = bi + 1
}

(0 .. 999).each do |e|
(n-1).step(0,-1) do |bi|
y[bi] += x.at(bi)
end
end
# puts "#{y.first} #{y.last}"


--
       user     system      total        real
ruby 30.210000   0.040000  30.250000 ( 30.566711)
yarv 11.950000   0.010000  11.960000 ( 12.050104)
-----------------------------------------------------------
so_concatenate:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: strcat-ruby.code,v 1.4 2004/11/13 07:43:28 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/
# based on code from Aristarkh A Zagorodnikov and Dat Nguyen

STUFF = "hello\n"
hello = ''
400000.times do |e|
hello << STUFF
end
# puts hello.length


--
       user     system      total        real
ruby  0.860000   0.040000   0.900000 (  0.908787)
yarv  0.350000   0.000000   0.350000 (  0.344135)
-----------------------------------------------------------
so_count_words:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: wc-ruby.code,v 1.4 2004/11/13 07:43:32 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/
# with help from Paul Brannan

require 'stringio'

input = StringIO.new
data = File.read($DRIVER_PATH + '/wc.input')

5000.times{|i|
input.write data
}
input.seek 0

nl = nw = nc = 0
while true
data = (input.read(4096) or break) << (input.gets || "")

nc += data.length
nl += data.count("\n")
((data.strip! || data).tr!("\n", " ") || data).squeeze!
nw += data.count(" ") + 1
end
# puts "#{nl} #{nw} #{nc}"

--
       user     system      total        real
ruby  0.700000   0.160000   0.860000 (  0.897124)
yarv  0.750000   0.110000   0.860000 (  0.869444)
-----------------------------------------------------------
so_exception:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: except-ruby.code,v 1.4 2004/11/13 07:41:33 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/

$HI = 0
$LO = 0
NUM = 250000 # Integer(ARGV[0] || 1)


class Lo_Exception < Exception
def initialize(num)
@value = num
end
end

class Hi_Exception < Exception
def initialize(num)
@value = num
end
end

def some_function(num)
begin
hi_function(num)
rescue
print "We shouldn't get here, exception is: #{$!.type}\n"
end
end

def hi_function(num)
begin
lo_function(num)
rescue Hi_Exception
$HI = $HI + 1
end
end

def lo_function(num)
begin
blowup(num)
rescue Lo_Exception
$LO = $LO + 1
end
end

def blowup(num)
if num % 2 == 0
raise Lo_Exception.new(num)
else
raise Hi_Exception.new(num)
end
end


i = 1
max = NUM+1
while i < max
i+=1
some_function(i+1)
end
--
       user     system      total        real
ruby 11.790000   0.620000  12.410000 ( 12.541421)
yarv  7.810000   0.760000   8.570000 (  8.711128)
-----------------------------------------------------------
so_matrix:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: matrix-ruby.code,v 1.4 2004/11/13 07:42:14 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/

n = 60 #Integer(ARGV.shift || 1)

size = 30

def mkmatrix(rows, cols)
count = 1
mx = Array.new(rows)
(0 .. (rows - 1)).each do |bi|
row = Array.new(cols, 0)
(0 .. (cols - 1)).each do |j|
row[j] = count
count += 1
end
mx[bi] = row
end
mx
end

def mmult(rows, cols, m1, m2)
m3 = Array.new(rows)
(0 .. (rows - 1)).each do |bi|
row = Array.new(cols, 0)
(0 .. (cols - 1)).each do |j|
val = 0
(0 .. (cols - 1)).each do |k|
val += m1.at(bi).at(k) * m2.at(k).at(j)
end
row[j] = val
end
m3[bi] = row
end
m3
end

m1 = mkmatrix(size, size)
m2 = mkmatrix(size, size)
mm = Array.new
n.times do
mm = mmult(size, size, m1, m2)
end
# puts "#{mm[0][0]} #{mm[2][3]} #{mm[3][2]} #{mm[4][4]}"


--
       user     system      total        real
ruby  8.880000   0.010000   8.890000 (  8.950790)
yarv  3.120000   0.010000   3.130000 (  3.160729)
-----------------------------------------------------------
so_nested_loop:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: nestedloop-ruby.code,v 1.4 2004/11/13 07:42:22 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/
# from Avi Bryant

n = 16 # Integer(ARGV.shift || 1)
x = 0
n.times do
n.times do
n.times do
n.times do
n.times do
n.times do
x += 1
end
end
end
end
end
end
# puts x


--
       user     system      total        real
ruby 26.630000   0.010000  26.640000 ( 26.883816)
yarv  9.760000   0.000000   9.760000 (  9.875278)
-----------------------------------------------------------
so_object:
#!/usr/bin/ruby
# -*- mode: ruby -*-
# $Id: objinst-ruby.code,v 1.4 2004/11/13 07:42:25 bfulgham Exp $
# http://www.bagley.org/~doug/shootout/
# with help from Aristarkh Zagorodnikov

class Toggle
def initialize(start_state)
@bool = start_state
end

def value
@bool
end

def activate
@bool = !@bool
self
end
end

class NthToggle < Toggle
def initialize(start_state, max_counter)
super start_state
@count_max = max_counter
@counter = 0
end

def activate
@counter += 1
if @counter >= @count_max
@bool = !@bool
@counter = 0
end
self
end
end

n = 1500000 # (ARGV.shift || 1).to_i

toggle = Toggle.new 1
5.times do
toggle.activate.value ? 'true' : 'false'
end
n.times do
toggle = Toggle.new 1
end

ntoggle = NthToggle.new 1, 3
8.times do
ntoggle.activate.value ? 'true' : 'false'
end
n.times do
ntoggle = NthToggle.new 1, 3
end

--
       user     system      total        real
ruby 22.170000   0.040000  22.210000 ( 22.529004)
yarv 17.690000   0.020000  17.710000 ( 17.964235)
-----------------------------------------------------------
so_random:
# from http://www.bagley.org/~doug/shootout/bench/random/random.ruby

IM = 139968
IA = 3877
IC = 29573

$last = 42.0

def gen_random(max)
(max * ($last = ($last * IA + IC) % IM)) / IM
end

N = 1000000

i=0
while i<N
i+=1
gen_random(100.0)
end
# "%.9f" % gen_random(100.0)
--
       user     system      total        real
ruby  9.100000   0.020000   9.120000 (  9.227853)
yarv  3.760000   0.030000   3.790000 (  3.828192)
-----------------------------------------------------------
so_sieve:
# from http://www.bagley.org/~doug/shootout/bench/sieve/sieve.ruby
num = 100
count = i = j = 0
flags0 = Array.new(8192,1)
k = 0
while k < num
k+=1
count = 0
flags = flags0.dup
i = 2
while i<8192
i+=1
if flags[i]
# remove all multiples of prime: i
j = i*i
while j < 8192
j += i
flags[j] = nil
end
count += 1
end
end
end
count
--
       user     system      total        real
ruby  8.830000   0.000000   8.830000 (  8.912284)
yarv  0.990000   0.030000   1.020000 (  1.043040)
-----------------------------------------------------------
strconcat:
i=0
while i<1000000
"#{1+1} #{1+1} #{1+1}"
i+=1
end
--
       user     system      total        real
ruby 11.880000   0.010000  11.890000 ( 12.065036)
yarv  7.850000   0.030000   7.880000 (  7.975811)
-----------------------------------------------------------
super:

class C
def m
1
end
end

class CC < C
def m
super()
end
end

obj = CC.new

i = 0
while i<10000000
obj.m
i+=1
end
--
       user     system      total        real
ruby 38.610000   0.020000  38.630000 ( 38.937559)
yarv  6.760000   0.010000   6.770000 (  6.828927)
-----------------------------------------------------------
swap:
a = 1
b = 2
i=0
while i<10000000
i+=1
a, b = b, a
end

--
       user     system      total        real
ruby 35.970000   0.070000  36.040000 ( 36.653012)
yarv  1.690000   0.010000   1.700000 (  1.715293)
-----------------------------------------------------------
tak:

def tak x, y, z
unless y < x
z
else
tak( tak(x-1, y, z),
tak(y-1, z, x),
tak(z-1, x, y))
end
end

tak(18, 9, 0)

--
       user     system      total        real
ruby 42.320000   0.010000  42.330000 ( 42.909755)
yarv  4.550000   0.010000   4.560000 (  4.662077)
-----------------------------------------------------------
tarai:
def tarai( x, y, z )
if x <= y
then y
else tarai(tarai(x-1, y, z),
tarai(y-1, z, x),
tarai(z-1, x, y))
end
end

tarai(12, 6, 0)
--
       user     system      total        real
ruby 32.840000   0.010000  32.850000 ( 33.302277)
yarv  5.020000   0.000000   5.020000 (  5.075307)
-----------------------------------------------------------
times:
10000000.times{|i|
}
--
       user     system      total        real
ruby  6.920000   0.010000   6.930000 (  6.951283)
yarv  4.680000   0.010000   4.690000 (  4.717908)
-----------------------------------------------------------
whileloop:
i = 0
while i<10000000
i+=1
end
--
       user     system      total        real
ruby 18.780000   0.050000  18.830000 ( 18.944069)
yarv  1.210000   0.000000   1.210000 (  1.215278)
-----------------------------------------------------------
zsuper:
i = 0

class C
def m a
1
end
end

class CC < C
def m a
super
end
end

obj = CC.new

while i<10000000
obj.m 10
i+=1
end
--
       user     system      total        real
ruby 41.410000   0.040000  41.450000 ( 41.589805)
yarv  6.850000   0.000000   6.850000 (  6.885806)

From ko1 at atdot.net  Tue Feb  8 10:37:10 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Tue Feb  8 10:35:16 2005
Subject: [Yarv-devel] Benchmark results on PPC
In-Reply-To: <cu2v6a$486$1@sea.gmane.org>
References: <cu2v6a$486$1@sea.gmane.org>
Message-ID: <4208DCA6.4090005@atdot.net>

Hi,

Andreas Schwarz wrote:

 > Powerbook G4 1.33 GHz

Thanks you :)
From ko1 at atdot.net  Tue Feb  8 12:33:04 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Tue Feb  8 12:31:02 2005
Subject: [Yarv-devel] YARV new feature
Message-ID: <20050209023304.273304d%ko1@atdot.net>

Hi,
I added new feature to YARV.


* goto/label statement

I supported goto/label statement.
Using this scheme, you can do more flexible programming on Ruby.


like this:
	i=0
	__label__(:start)

	__goto__(:label)
	p "skipped"
	__label__(:label)
	p "end"

	__goto__(:start) if (i+=1) < 3




Enjoy Happy Ruby Hacking!



... of course, only joke :-P

But really implemented. You can use "goto/label" statement when you
add option to build yarv:

	$ ruby extconf.rb --enable-support-goto


Regards,
-- 
// SASADA Koichi at atdot dot net
//

From surrender_it at yahoo.it  Fri Feb 11 13:12:04 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Fri Feb 11 13:08:42 2005
Subject: [Yarv-devel] YARV new feature
In-Reply-To: <20050209023304.273304d%ko1@atdot.net>
Message-ID: <20050211181205.67916.qmail@web50102.mail.yahoo.com>

 --- SASADA Koichi <ko1@atdot.net> ha scritto: 
> Hi,
> I added new feature to YARV.
> 
> 
> * goto/label statement

Sorry for the dumb qu4estion: why do you support this?
:)

Another thing, I notice that since some revisions my
mingw (gcc version 3.4.2 (mingw-special) ) raise a
warning since thread_call0() calls call_cfunc().

I think that in vm.c, line 424
 
    VALUE *argv;
should be 
    const VALUE *argv


(notice I am compiling withouth warning flags or
warning would be thousands :)

=====
icq #69488917


		
___________________________________ 
Nuovo Yahoo! Messenger: E' molto pi? divertente: Audibles, Avatar, Webcam, Giochi, Rubrica
 Scaricalo ora! 
http://it.messenger.yahoo.it
From ko1 at atdot.net  Fri Feb 11 16:42:11 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Fri Feb 11 16:40:14 2005
Subject: [Yarv-devel] YARV new feature
In-Reply-To: <20050211181205.67916.qmail@web50102.mail.yahoo.com>
References: <20050211181205.67916.qmail@web50102.mail.yahoo.com>
Message-ID: <420D26B3.3030508@atdot.net>

Hi,

gabriele renzi wrote:
>  --- SASADA Koichi <ko1@atdot.net> ha scritto: 
> 
>>Hi,
>>I added new feature to YARV.
>>
>>
>>* goto/label statement
> 
> 
> Sorry for the dumb qu4estion: why do you support this?
> :)

only joke (as I wrote in original post).


> 
> Another thing, I notice that since some revisions my
> mingw (gcc version 3.4.2 (mingw-special) ) raise a
> warning since thread_call0() calls call_cfunc().
> 
> I think that in vm.c, line 424
>  
>     VALUE *argv;
> should be 
>     const VALUE *argv

Thank you.


Regards,
SASADA Koichi
From ko1 at atdot.net  Fri Feb 18 14:14:52 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Fri Feb 18 14:13:45 2005
Subject: [Yarv-devel] [ANN] YARV: Yet Another RubyVM 0.1.1
Message-ID: <20050219041452.1cb796d%ko1@atdot.net>

Hi,

I'm happy to announce that I released YARV: Yet Another RubyVM 0.1.1.

Project Home page: http://www.atdot.net/yarv/
YARV 0.1.1       : http://www.atdot.net/yarv/yarv-0.1.1.tar.gz
0.1.1 souces     : http://www.atdot.net/viewcvs.cgi/yarv/tags/0.1.1/

Changes (between 0.1.0):

* some bug fixes
* refoctoring

* New Features
  * support goto (*1)
  * adding Ruby option "-rite" to work ruby script on YARV
    instead of current ruby interpreter (*2)
  * defined? support
  * permit block parameter as local variables or so

* Optimization
  * Instructions unification (aka super instruction) is supported
    tentatively.
  * JIT Compiler (Ruby script -> Native machine code), very experimental
    implementation
  * AOT Compiler support method dispatch

If you want to know moreover, please subscribe yarv-devel
(you can know about this ML in YARV Home page).


This project is supported by IPA (Information-technology
Promotion Agency, Japan) "Exploratory Software Project (youth)".


-- 
// SASADA Koichi at atdot dot net
//

(*1) it's only joke
(*2) using -r (require) option with ite.rb :)


From surrender_it at yahoo.it  Fri Feb 18 15:48:04 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Fri Feb 18 15:44:26 2005
Subject: [Yarv-devel] [ANN] YARV: Yet Another RubyVM 0.1.1
In-Reply-To: <20050219041452.1cb796d%ko1@atdot.net>
Message-ID: <20050218204805.72253.qmail@web50104.mail.yahoo.com>

 --- SASADA Koichi <ko1@atdot.net> ha scritto: 
> Hi,
> 
> I'm happy to announce that I released YARV: Yet
> Another RubyVM 0.1.1.

great news


> * Optimization
>   * Instructions unification (aka super instruction)
> is supported
>     tentatively.
>   * JIT Compiler (Ruby script -> Native machine
> code), very experimental
>     implementation
>   * AOT Compiler support method dispatch

just wanted to understand: is there a way we can test
JIT or AOT? 
And, btw, I think in jitcompile.c:
 #if __GNUC__ && __i386__ && OPT_JIT_COMPILE 
 /* this part may be separeted to another file */

 #include <sys/mman.h>
is a little weak since it would allow gcc on win32 to
try loading mman.h so it would need a 
&& !defined(WIN32)

=====
icq #69488917


		
___________________________________ 
Nuovo Yahoo! Messenger: E' molto pi? divertente: Audibles, Avatar, Webcam, Giochi, Rubrica
 Scaricalo ora! 
http://it.messenger.yahoo.it
From ko1 at atdot.net  Mon Feb 21 00:10:51 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon Feb 21 00:09:43 2005
Subject: [Yarv-devel] [ANN] YARV: Yet Another RubyVM 0.1.1
In-Reply-To: (Your message of "Fri, 18 Feb 2005 21:48:04 +0100 (CET)
	")  <20050218204805.72253.qmail@web50104.mail.yahoo.com>
References: <20050218204805.72253.qmail@web50104.mail.yahoo.com>
Message-ID: <20050221141051.c01dd2%ko1@atdot.net>

  gabriele renzi <surrender_it@yahoo.it> wrote :
    [ Re: [Yarv-devel] [ANN] YARV: Yet Another RubyVM 0.1.1 ]
    at Fri, 18 Feb 2005 21:48:04 +0100 (CET)

Hi,

> just wanted to understand: is there a way we can test
> JIT or AOT? 

JIT:

1. extconf.rb --enable-opt-jit-compile
2. make some method

ex)
  def compiled
    1+1 # yarv can't do jit compile complex scripts
  end

3. compile it

  jitcompile(self, :compiled) # this method jit compile self#compiled

4. call that method

  compiled()


----
  script will #=>
  
	  def compiled
	    1+1 # yarv can't do jit compile complex scripts
	  end

	  jitcompile(self, :compiled) # this method jit compile self#compiled

	  compiled()

5. run this script


AOT:

1. extconf.rb --enable-test-aot-compile

2. modify some file. note that default target is rb/aottest.rb

3. do compile with rb/aotcompile.rb
   you can do aot compile with "make aotc"

   compiled.inc will be created (ruby -> c compiled c source file).

4. call "compiled()" method

5. run this script




> And, btw, I think in jitcompile.c:
>  #if __GNUC__ && __i386__ && OPT_JIT_COMPILE 
>  /* this part may be separeted to another file */
> 
>  #include <sys/mman.h>
> is a little weak since it would allow gcc on win32 to
> try loading mman.h so it would need a 
> && !defined(WIN32)

OK. Thank you.


Regards,
-- 
// SASADA Koichi at atdot dot net
//
From vincent.isambart at gmail.com  Wed Feb 23 04:12:30 2005
From: vincent.isambart at gmail.com (Vincent Isambart)
Date: Wed Feb 23 04:08:41 2005
Subject: [Yarv-devel] "a = {}; a[0] || = 1" crashes
Message-ID: <7d9a1f5305022301121e20f43c@mail.gmail.com>

Hello,

I have a little problem with YARV. Note that my tests were done with
the version from one week and a half ago so if the bug has been
corrected you can trash this mail ;)

YARV crashes with the following simple Ruby script:
a = {}; a[0] || = 1

I checked it a little and it seems the error is during the compilation
step. I do not know YARV or the Ruby interpreter very well so I may be
mistaken.
The problem seems to be in compile.c, and there search for NODE_OP_ASGN1.
The error seems to be the "ADD_SEND (ret, nd_line(node),
ID2SYM(node->nd_mid), I2F(1));"
For the || and && operators, node->nd_mid's value is 0 and 1, it's not an ID...
If you try a parse and disassemble the "a = {}; a[0] || = 1" script,
you'll see that YARV will try to invoque null...

I hope it's of any help.
Keep up the good work :)

Kind regards,
Vincent Isambart
From vincent.isambart at gmail.com  Wed Feb 23 04:38:15 2005
From: vincent.isambart at gmail.com (Vincent Isambart)
Date: Wed Feb 23 04:34:25 2005
Subject: [Yarv-devel] Mistake on the web page
Message-ID: <7d9a1f5305022301383998ff11@mail.gmail.com>

Hi,

There is an error an the YARV web page: the 0.1.1 link points to
yarv-0.1.0.tar.gz instead of yarv-0.1.1.tar.gz...

Kind regards,
Vincent Isambart
From ko1 at atdot.net  Wed Feb 23 10:48:28 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Wed Feb 23 10:47:00 2005
Subject: [Yarv-devel] "a = {}; a[0] || = 1" crashes
In-Reply-To: (Your message of "Wed, 23 Feb 2005 10:12:30 +0100")
	<7d9a1f5305022301121e20f43c@mail.gmail.com>
References: <7d9a1f5305022301121e20f43c@mail.gmail.com>
Message-ID: <20050224004828.2834742%ko1@atdot.net>

  Vincent Isambart <vincent.isambart@gmail.com> wrote :
    [ [Yarv-devel] "a = {}; a[0] || = 1" crashes ]
    at Wed, 23 Feb 2005 10:12:30 +0100

Hi,

> YARV crashes with the following simple Ruby script:
> a = {}; a[0] || = 1
> 
> I checked it a little and it seems the error is during the compilation
> step. I do not know YARV or the Ruby interpreter very well so I may be
> mistaken.
> The problem seems to be in compile.c, and there search for NODE_OP_ASGN1.
> The error seems to be the "ADD_SEND (ret, nd_line(node),
> ID2SYM(node->nd_mid), I2F(1));"
> For the || and && operators, node->nd_mid's value is 0 and 1, it's not an ID...
> If you try a parse and disassemble the "a = {}; a[0] || = 1" script,
> you'll see that YARV will try to invoque null...

Thank you.
I missed it! I fixed it.

Regards,
-- 
// SASADA Koichi at atdot dot net
//
From ko1 at atdot.net  Wed Feb 23 10:49:38 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Wed Feb 23 10:48:09 2005
Subject: [Yarv-devel] Mistake on the web page
In-Reply-To: (Your message of "Wed, 23 Feb 2005 10:38:15 +0100")
	<7d9a1f5305022301383998ff11@mail.gmail.com>
References: <7d9a1f5305022301383998ff11@mail.gmail.com>
Message-ID: <20050224004938.2845ac5%ko1@atdot.net>

  Vincent Isambart <vincent.isambart@gmail.com> wrote :
    [ [Yarv-devel] Mistake on the web page ]
    at Wed, 23 Feb 2005 10:38:15 +0100

Hi,

> There is an error an the YARV web page: the 0.1.1 link points to
> yarv-0.1.0.tar.gz instead of yarv-0.1.1.tar.gz...

Thanks also. I fixed it. (as you know, it's miss at copy and paste)

Regrads,
-- 
// SASADA Koichi at atdot dot net
//
From vincent.isambart at gmail.com  Wed Feb 23 11:03:22 2005
From: vincent.isambart at gmail.com (Vincent Isambart)
Date: Wed Feb 23 10:59:35 2005
Subject: [Yarv-devel] "a = {}; a[0] || = 1" crashes
In-Reply-To: <20050224004828.2834742%ko1@atdot.net>
References: <7d9a1f5305022301121e20f43c@mail.gmail.com>
	<20050224004828.2834742%ko1@atdot.net>
Message-ID: <7d9a1f5305022308031768464a@mail.gmail.com>

??????

> > YARV crashes with the following simple Ruby script:
> > a = {}; a[0] || = 1
> >
> > I checked it a little and it seems the error is during the compilation
> > step. I do not know YARV or the Ruby interpreter very well so I may be
> > mistaken.
> > The problem seems to be in compile.c, and there search for NODE_OP_ASGN1.
> > The error seems to be the "ADD_SEND (ret, nd_line(node),
> > ID2SYM(node->nd_mid), I2F(1));"
> > For the || and && operators, node->nd_mid's value is 0 and 1, it's not an ID...
> > If you try a parse and disassemble the "a = {}; a[0] || = 1" script,
> > you'll see that YARV will try to invoque null...
> 
> Thank you.
> I missed it! I fixed it.

Thanks for the fast correction.

But I checked your correction and there is something odd. An extract
of what you did:

    if(id == 0 || id == 1){
(...)
    }
    else if(id == 1){
      /* and */
      VALUE label = NEW_LABEL(nd_line(node));
      
      
    }
    else{
      ADD_SEQ  (ret, COMPILE("NODE_OP_ASGN1 args->head: ",
node->nd_args->nd_head));
      ADD_SEND (ret, nd_line(node), ID2SYM(id), I2F(1));
      ADD_SEND (ret, nd_line(node), ID2SYM(idASET), I2F(argc+1));
    }

I think the "else if(id == 1)" is useless... This branch will never be
executed :P

Kind regards,
????????
Vincent Isambart

From ko1 at atdot.net  Wed Feb 23 11:15:37 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Wed Feb 23 11:14:08 2005
Subject: [Yarv-devel] "a = {}; a[0] || = 1" crashes
In-Reply-To: <7d9a1f5305022308031768464a@mail.gmail.com>
References: <7d9a1f5305022301121e20f43c@mail.gmail.com>	<20050224004828.2834742%ko1@atdot.net>
	<7d9a1f5305022308031768464a@mail.gmail.com>
Message-ID: <421CAC29.3010303@atdot.net>

Hi,

Vincent Isambart wrote:
> ??????

????? :)

> 
> I think the "else if(id == 1)" is useless... This branch will never be
> executed :P
> 
Ya! I removed it.

> ????????

Good night :)
-- 
// SASADA Koichi at atdot dot net
//
From cwillia1 at rochester.rr.com  Thu Feb 24 19:18:44 2005
From: cwillia1 at rochester.rr.com (Chris Williams)
Date: Thu Feb 24 19:14:53 2005
Subject: [Yarv-devel] How to compile Ruby 1.9 snapshots in Windows(MinGW)?
Message-ID: <000001c51acf$91e343f0$6400a8c0@ChrisLaptop>

Hi,
            I'd love to play with YARV, but currently the only way to do
that is to compile Ruby 1.9 and YARV yourself. I'm on Win XP and don't
have MS VC++ anywhere. So I've been trying to compile with MinGW (as
well as MYSYS and MYSYS DTK).
 
I run:
autoconf
./configure -enable-shared
make
 
And make fails with the following:
 
ar rcu libmsvcrt-ruby18-static.a array.o bignum.o class.o compar.o dir.o
dln.o enum.o error.o eval.o file.o gc.o hash.o inits.o io.o marshal.o
math.o numeric.o object.o pack.o parse.o process.o prec.o random.o
range.o re.o regex.o ruby.o signal.o sprintf.o st.o string.o struct.o
time.o util.o variable.o version.o fileblocks.o crypt.o flock.o acosh.o
win32.o  dmyext.o
gcc -g -O2    -I. -I.  -c main.c
gcc -g -O2      main.o dmyext.o libmsvcrt-ruby18-static.a -lwsock32   -o
miniruby.exe
c:\ruby19\build\ruby\miniruby.exe: No such file to load -- ubygems
(LoadError)
make: *** [rbconfig.rb] Error 1
 
Anyone out there who's compiled 1.9 with MinGW and can give me some
pointers?
 
Thanks,
Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/yarv-devel/attachments/20050224/4b8e17e6/attachment-0001.htm
From ko1 at atdot.net  Thu Feb 24 21:09:04 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Thu Feb 24 21:06:47 2005
Subject: [Yarv-devel] How to compile Ruby 1.9 snapshots in Windows(MinGW)?
In-Reply-To: <000001c51acf$91e343f0$6400a8c0@ChrisLaptop>
References: <000001c51acf$91e343f0$6400a8c0@ChrisLaptop>
Message-ID: <421E88C0.3080103@atdot.net>

Hi,

 > gcc -g -O2      main.o dmyext.o libmsvcrt-ruby18-static.a -lwsock32 
   -o
 > miniruby.exe
 >
 > c:\ruby19\build\ruby\miniruby.exe: No such file to load -- ubygems
 > (LoadError)

At first, you need ruby 1.9. (libmsvcrt-ruby18-static.a should not 
appear above)

And you must remove RUBYOPT environment variable (Maybe your RUBYOPT 
includes "-rubygems")


BTW, you can also use yarv with cygwin.

Cheers,
SASADA Koichi
From cwillia1 at rochester.rr.com  Thu Feb 24 22:45:32 2005
From: cwillia1 at rochester.rr.com (Chris Williams)
Date: Thu Feb 24 22:41:49 2005
Subject: [Yarv-devel] How to compile Ruby 1.9 snapshots in Windows(MinGW)?
In-Reply-To: <421E88C0.3080103@atdot.net>
Message-ID: <000001c51aec$7aad50f0$6400a8c0@ChrisLaptop>

Koichi-san,

(Or should it be Sasad-san?)

>  > gcc -g -O2      main.o dmyext.o libmsvcrt-ruby18-static.a -lwsock32
>    -o
>  > miniruby.exe
>  >
>  > c:\ruby19\build\ruby\miniruby.exe: No such file to load -- ubygems
>  > (LoadError)
> 
> At first, you need ruby 1.9. (libmsvcrt-ruby18-static.a should not
> appear above)
> 
> And you must remove RUBYOPT environment variable (Maybe your RUBYOPT
> includes "-rubygems")
> 
> BTW, you can also use yarv with cygwin.
> 
> Cheers,
> SASADA Koichi

You are correct, my RUBYOPT environment variable was set to "rubygems".
After removing the variable from my system Ruby properly built. I'm
running the YARV benchmarks now!

Thanks,
Chris

From cwillia1 at rochester.rr.com  Fri Feb 25 00:11:15 2005
From: cwillia1 at rochester.rr.com (Chris Williams)
Date: Fri Feb 25 00:07:24 2005
Subject: [Yarv-devel] Unable to complete benchmarks
Message-ID: <000001c51af8$6e970610$6400a8c0@ChrisLaptop>

Koichi-san, all,
            After your help, I was able to compile and install Ruby 1.9
and also patch it and try out YARV. I'm having problems running the
benchmarks either in my command prompt or through the MSYS shell with
'make benchmark'.
 
What seems to be happening is that the sub-process used to run the
benchmark will run and peg my CPU at 99% usage, then finish and
disappear but the result isn't joined back to the main thread.
 
The popen block appears to be stuck on 'puts io.gets' - and when I watch
my processes in the Task Manager I see a subprocess is spawned, runs and
disappears with no results. What is even stranger is that it seems to be
random as to when it happens. Sometimes I can't complete the first
benchmark in ruby, sometimes not in YARV, sometimes it happens on the
second benchmark, etc.
 
I'm using:
ruby 1.9.0 (2005-02-23) [i386-mingw32]
YARVCore 0.1.1 rev: 140 (2005-02-18)
[direct threaded code] [optimize basic operation] [optimize regexp
match] [stack caching] [inline method cache]
WinXP SP2
Intel P4 2.8 GHz
 
Has anyone else seen this? This seems like it would be a Ruby bug.
 
Thanks,
Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/yarv-devel/attachments/20050225/2600c97c/attachment.htm
From surrender_it at yahoo.it  Fri Feb 25 11:20:02 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Fri Feb 25 11:16:10 2005
Subject: [Yarv-devel] Unable to complete benchmarks
In-Reply-To: <000001c51af8$6e970610$6400a8c0@ChrisLaptop>
Message-ID: <20050225162002.4370.qmail@web50108.mail.yahoo.com>

 --- Chris Williams <cwillia1@rochester.rr.com> ha
scritto: 

> I'm using:
> ruby 1.9.0 (2005-02-23) [i386-mingw32]
> YARVCore 0.1.1 rev: 140 (2005-02-18)
> [direct threaded code] [optimize basic operation]
> [optimize regexp
> match] [stack caching] [inline method cache]
> WinXP SP2
> Intel P4 2.8 GHz
>  
> Has anyone else seen this? This seems like it would
> be a Ruby bug.

I remember this some time ago,. but it was related to
an alrady fixed bug in IO.read, IIRC.
Using your same ruby/os on a athlonxp 2000+ I can run
tests and benchmarks fine with both yarv revision 140
and 148 from the command prompt. 



=====
icq #69488917


		
___________________________________ 
Nuovo Yahoo! Messenger: E' molto pi? divertente: Audibles, Avatar, Webcam, Giochi, Rubrica
 Scaricalo ora! 
http://it.messenger.yahoo.it
From surrender_it at yahoo.it  Sat Feb 26 03:03:29 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Sat Feb 26 02:59:33 2005
Subject: [Yarv-devel] some fun bugs
Message-ID: <20050226080329.87296.qmail@web50103.mail.yahoo.com>

since 0.0.1 was released I tried running a tiny script
which stresses callcc a lot, by solving the "use just
4 colours in a map" problem. It always segfaulted.

Yesterday I was trying to reduce the script to a
minimum to find what was causing the segfault.. and I
found 4 reasons :)

1: using callcc 
callcc do |k| 
 k
end

gives:
./benchmark/bm_4col.rb:6: [BUG] Segmentation fault
Actually, it seem that just

 callcc
causes a segfault.

2: using infinite recursion
def callcc # forget the name, just infinite recursion
 callcc 
end

callcc 

gives:
./benchmark/bm_4col.rb:8: [BUG] stack overflow


3: calling a method with wrong number of args:

def foo

end
foo(1)

gives:
./benchmark/bm_4col.rb:6: [BUG] thread_backtrace:
unkown instruction (-10741130)

notice even that "unknown" is misspelled :)

4: return from block
 def foo
  3.timed do 
   return
  end
 end
 
 foo
./benchmark/bm_4col.rb:15: [BUG] thread_backtrace:
unkown instruction (-8024)


Sidenote: it seem that a paper on the implementation
of Lua 5.0 has recently been released, maybe of
interest:
http://www.tecgraf.puc-rio.br/~lhf/ftp/doc/sblp2005.pdf

=====
icq #69488917


		
___________________________________ 
Nuovo Yahoo! Messenger: E' molto pi? divertente: Audibles, Avatar, Webcam, Giochi, Rubrica
 Scaricalo ora! 
http://it.messenger.yahoo.it
From ko1 at atdot.net  Sat Feb 26 03:54:38 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Sat Feb 26 03:48:30 2005
Subject: [Yarv-devel] some fun bugs
In-Reply-To: <20050226080329.87296.qmail@web50103.mail.yahoo.com>
References: <20050226080329.87296.qmail@web50103.mail.yahoo.com>
Message-ID: <4220394E.1050209@atdot.net>

Hi,

gabriele renzi wrote:
> since 0.0.1 was released I tried running a tiny script
> which stresses callcc a lot, by solving the "use just
> 4 colours in a map" problem. It always segfaulted.

0.1.1
> 
> 1: using callcc 
> callcc do |k| 
>  k
> end
> 
> gives:
> ./benchmark/bm_4col.rb:6: [BUG] Segmentation fault
> Actually, it seem that just
> 
>  callcc
> causes a segfault.

callcc is not supported by YARV.


> 2: using infinite recursion
> def callcc # forget the name, just infinite recursion
>  callcc 
> end
> 
> callcc 
> 
> gives:
> ./benchmark/bm_4col.rb:8: [BUG] stack overflow

This means YARV can check stack overflow. (now, only calling rb_bug)


> 3: calling a method with wrong number of args:
> 
> def foo
> 
> end
> foo(1)
> 
> gives:
> ./benchmark/bm_4col.rb:6: [BUG] thread_backtrace:
> unkown instruction (-10741130)
> 
> notice even that "unknown" is misspelled :)
> 
mmm... i'll fix it. This is caused by bugs of creating backtrace process.

> 4: return from block
>  def foo
>   3.timed do 
>    return
>   end
>  end
>  
>  foo
> ./benchmark/bm_4col.rb:15: [BUG] thread_backtrace:
> unkown instruction (-8024)
> 
i'll also fix it.

> 
> Sidenote: it seem that a paper on the implementation
> of Lua 5.0 has recently been released, maybe of
> interest:
> http://www.tecgraf.puc-rio.br/~lhf/ftp/doc/sblp2005.pdf

I'll check it as soon as possible. Thank you.


Thanks,
SASADA Koichi
From vincent.isambart at gmail.com  Sat Feb 26 04:41:15 2005
From: vincent.isambart at gmail.com (Vincent Isambart)
Date: Sat Feb 26 04:37:20 2005
Subject: [Yarv-devel] Little Crash
Message-ID: <7d9a1f5305022601411cf66e11@mail.gmail.com>

Hi,

I got a little program crashing, and I do not know if it's a know bug:
    def foo(&block)
      $a = block
    end

    foo { puts "yo" }

    loop {
      $a.call
    }

What's weird is that it is not crashing anymore if I remove the loop
{} around the $a.call...

When trying to reduce the size of the previous program I found
something else that crashes:
    a = proc { puts "yo" }
    a.call
but I suppose it's because proc is not yet supported...

A little mistake in tmpl/insns.inc.tmpl: it should be DO NOT TOUCH
instead of DO NOT TOUGH :)

Keep up the good work :)

Cheers,
Vincent Isambart
From ko1 at atdot.net  Sun Feb 27 01:13:08 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Sun Feb 27 01:11:45 2005
Subject: [Yarv-devel] Little Crash
In-Reply-To: (Your message of "Sat, 26 Feb 2005 10:41:15 +0100")
	<7d9a1f5305022601411cf66e11@mail.gmail.com>
References: <7d9a1f5305022601411cf66e11@mail.gmail.com>
Message-ID: <20050227151308.66d5dd%ko1@atdot.net>

  Vincent Isambart <vincent.isambart@gmail.com> wrote :
    [ [Yarv-devel] Little Crash ]
    at Sat, 26 Feb 2005 10:41:15 +0100

Hi,

> I got a little program crashing, and I do not know if it's a know bug:
>     def foo(&block)
>       $a = block
>     end
> 
>     foo { puts "yo" }
> 
>     loop {
>       $a.call
>     }
> 
> What's weird is that it is not crashing anymore if I remove the loop
> {} around the $a.call...

loop{} is not supported. you can use loop{} which is redefined
in ruby script suche as:

	def loop
	  while true
	    yield
	  end
	end


> When trying to reduce the size of the previous program I found
> something else that crashes:
>     a = proc { puts "yo" }
>     a.call
> but I suppose it's because proc is not yet supported...

yes, "proc" is not supported. also this case is solved with redefined
method:

	def proc &b
	  b
	end

> 
> A little mistake in tmpl/insns.inc.tmpl: it should be DO NOT TOUCH
> instead of DO NOT TOUGH :)

orz


Thanks,
-- 
// SASADA Koichi at atdot dot net
//