From normalperson at yhbt.net  Sat Mar  1 16:36:58 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 1 Mar 2008 13:36:58 -0800
Subject: [Mongrel-development] [PATCH] avoid needless syscall when
	num_processors limit is reached
In-Reply-To: <b6f68fc60802291441k5a414ae6ve96dfbc88c9da090@mail.gmail.com>
References: <20080225080111.GA11848@soma>
	<b6f68fc60802291441k5a414ae6ve96dfbc88c9da090@mail.gmail.com>
Message-ID: <20080301213658.GA28988@mayonaise>

Evan Weaver <evan at cloudbur.st> wrote:
> Is there already a ticket for this? I'm not seeing it.

No, not for this patch in particular.  I was just browsing the mongrel
source and this patch just seems like the right thing to do when the
num_processors limit is reached.

I was referring to another patch in that email (ticket/14, --num-procs
not being accepted by mongrel_rails).

They're related, but this is just a nitpicky patch and the other one
(ticket/14) is a real bug).

> On Mon, Feb 25, 2008 at 3:01 AM, Eric Wong <normalperson at yhbt.net> wrote:
> > Since we're going to close the socket immediately after the
> >  num_processors limit is reached, there's no point in calling
> >  setsockopt(2) to enable TCP_CORK on it.  Instead, only enable
> >  TCP_CORK for connections we are able to handle.
> >
> >  While we're at it, avoid calling worker_list.length twice in the
> >  connection rejected case and instead just set num_workers to
> >  @workers.list.length once.  I'm assuming the original caching of
> >  worker_list = @workers.list to avoid having the log message
> >  display a different number due to a race condition; and this
> >  preserves that functionality while avoiding an extra method call.
> >
> >  ---
> >   This patch is against
> >   svn://rubyforge.org/var/svn/mongrel/branches/stable_1-1 r981
> >
> >   lib/mongrel.rb |   16 +++++++---------
> >   1 files changed, 7 insertions(+), 9 deletions(-)
> >
> >  diff --git a/lib/mongrel.rb b/lib/mongrel.rb
> >  index d99c56d..cfdfc49 100644
> >  --- a/lib/mongrel.rb
> >  +++ b/lib/mongrel.rb
> >  @@ -270,18 +270,16 @@ module Mongrel
> >            while true
> >              begin
> >                client = @socket.accept
> >  -
> >  -              if defined?($tcp_cork_opts) and $tcp_cork_opts
> >  -                client.setsockopt(*$tcp_cork_opts) rescue nil
> >  -              end
> >  -
> >  -              worker_list = @workers.list
> >  -
> >  -              if worker_list.length >= @num_processors
> >  -                STDERR.puts "Server overloaded with #{worker_list.length} processors (#@num_processors max). Dropping connection."
> >  +
> >  +              num_workers = @workers.list.length
> >  +              if num_workers >= @num_processors
> >  +                STDERR.puts "Server overloaded with #{num_workers} processors (#@num_processors max). Dropping connection."
> >                  client.close rescue nil
> >                  reap_dead_workers("max processors")
> >                else
> >  +                if defined?($tcp_cork_opts) and $tcp_cork_opts
> >  +                  client.setsockopt(*$tcp_cork_opts) rescue nil
> >  +                end
> >                  thread = Thread.new(client) {|c| process_client(c) }
> >                  thread[:started_on] = Time.now
> >                  @workers.add(thread)
> >  --
> >  Eric Wong

From normalperson at yhbt.net  Sat Mar  1 16:38:53 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 1 Mar 2008 13:38:53 -0800
Subject: [Mongrel-development] [PATCH (trivial)] test/unit/test_ws: fix
	require after 'testhelp' got renamed to 'test_helper'
In-Reply-To: <b6f68fc60802282029r647c6022p8c63b4a2ef8f9f3f@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<Pine.LNX.4.64.0802282330000.3722@sophie.icewall.org>
	<b6f68fc60802282029r647c6022p8c63b4a2ef8f9f3f@mail.gmail.com>
Message-ID: <20080301213853.GB28988@mayonaise>

Tests all run correctly on the stable_1-2 branch now.
---

  Since I'm looking at the tests... :)

  Evan Weaver <evan at cloudbur.st> wrote:
  > That would be great; I haven't gotten to it. Can you merge it against
  > branches/stable_1-2?
  > 
  > Also make sure to add a test if it changes any API or adds any functionality.
  > 
  > Thanks
  > 
  > Evan

 test/unit/test_ws.rb |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/test/unit/test_ws.rb b/test/unit/test_ws.rb
index d239e18..68a79a8 100644
--- a/test/unit/test_ws.rb
+++ b/test/unit/test_ws.rb
@@ -4,7 +4,7 @@
 # Additional work donated by contributors.  See http://mongrel.rubyforge.org/attributions.html 
 # for more information.
 
-require 'test/testhelp'
+require 'test/test_helper'
 
 include Mongrel
 
-- 
Eric Wong

From evan at cloudbur.st  Sat Mar  1 16:48:05 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 1 Mar 2008 16:48:05 -0500
Subject: [Mongrel-development] [PATCH (trivial)] test/unit/test_ws: fix
	require after 'testhelp' got renamed to 'test_helper'
In-Reply-To: <20080301213853.GB28988@mayonaise>
References: <20080229015309.GA9080@untitled>
	<Pine.LNX.4.64.0802282330000.3722@sophie.icewall.org>
	<b6f68fc60802282029r647c6022p8c63b4a2ef8f9f3f@mail.gmail.com>
	<20080301213853.GB28988@mayonaise>
Message-ID: <b6f68fc60803011348t799ef67aldeed20d1d8e47e98@mail.gmail.com>

Applied, thanks.

Do you want a commit bit? What's your RubyForge name?

Evan

On Sat, Mar 1, 2008 at 4:38 PM, Eric Wong <normalperson at yhbt.net> wrote:
> Tests all run correctly on the stable_1-2 branch now.
>  ---
>
>   Since I'm looking at the tests... :)
>
>   Evan Weaver <evan at cloudbur.st> wrote:
>   > That would be great; I haven't gotten to it. Can you merge it against
>   > branches/stable_1-2?
>   >
>   > Also make sure to add a test if it changes any API or adds any functionality.
>   >
>   > Thanks
>   >
>   > Evan
>
>   test/unit/test_ws.rb |    2 +-
>   1 files changed, 1 insertions(+), 1 deletions(-)
>
>  diff --git a/test/unit/test_ws.rb b/test/unit/test_ws.rb
>  index d239e18..68a79a8 100644
>  --- a/test/unit/test_ws.rb
>  +++ b/test/unit/test_ws.rb
>  @@ -4,7 +4,7 @@
>   # Additional work donated by contributors.  See http://mongrel.rubyforge.org/attributions.html
>   # for more information.
>
>  -require 'test/testhelp'
>  +require 'test/test_helper'
>
>   include Mongrel
>
>  --
>  Eric Wong
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From normalperson at yhbt.net  Sat Mar  1 17:24:19 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 1 Mar 2008 14:24:19 -0800
Subject: [Mongrel-development] [PATCH (trivial)] test/unit/test_ws: fix
	require after 'testhelp' got renamed to 'test_helper'
In-Reply-To: <b6f68fc60803011348t799ef67aldeed20d1d8e47e98@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<Pine.LNX.4.64.0802282330000.3722@sophie.icewall.org>
	<b6f68fc60802282029r647c6022p8c63b4a2ef8f9f3f@mail.gmail.com>
	<20080301213853.GB28988@mayonaise>
	<b6f68fc60803011348t799ef67aldeed20d1d8e47e98@mail.gmail.com>
Message-ID: <20080301222419.GA9727@soma>

Evan Weaver <evan at cloudbur.st> wrote:
> Applied, thanks.
> 
> Do you want a commit bit? What's your RubyForge name?

Sure, I'm normalperson: http://rubyforge.org/users/normalperson/

Thanks!

> Evan
> 
> On Sat, Mar 1, 2008 at 4:38 PM, Eric Wong <normalperson at yhbt.net> wrote:
> > Tests all run correctly on the stable_1-2 branch now.
> >  ---
> >
> >   Since I'm looking at the tests... :)
> >
> >   Evan Weaver <evan at cloudbur.st> wrote:
> >   > That would be great; I haven't gotten to it. Can you merge it against
> >   > branches/stable_1-2?
> >   >
> >   > Also make sure to add a test if it changes any API or adds any functionality.
> >   >
> >   > Thanks
> >   >
> >   > Evan
> >
> >   test/unit/test_ws.rb |    2 +-
> >   1 files changed, 1 insertions(+), 1 deletions(-)
> >
> >  diff --git a/test/unit/test_ws.rb b/test/unit/test_ws.rb
> >  index d239e18..68a79a8 100644
> >  --- a/test/unit/test_ws.rb
> >  +++ b/test/unit/test_ws.rb
> >  @@ -4,7 +4,7 @@
> >   # Additional work donated by contributors.  See http://mongrel.rubyforge.org/attributions.html
> >   # for more information.
> >
> >  -require 'test/testhelp'
> >  +require 'test/test_helper'
> >
> >   include Mongrel
> >
> >  --
> >  Eric Wong

From evan at cloudbur.st  Sat Mar  1 18:41:44 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 1 Mar 2008 18:41:44 -0500
Subject: [Mongrel-development] [PATCH (trivial)] test/unit/test_ws: fix
	require after 'testhelp' got renamed to 'test_helper'
In-Reply-To: <20080301222419.GA9727@soma>
References: <20080229015309.GA9080@untitled>
	<Pine.LNX.4.64.0802282330000.3722@sophie.icewall.org>
	<b6f68fc60802282029r647c6022p8c63b4a2ef8f9f3f@mail.gmail.com>
	<20080301213853.GB28988@mayonaise>
	<b6f68fc60803011348t799ef67aldeed20d1d8e47e98@mail.gmail.com>
	<20080301222419.GA9727@soma>
Message-ID: <b6f68fc60803011541t410b2250q26b2b67c73fad3c9@mail.gmail.com>

Added you.

Evan

On Sat, Mar 1, 2008 at 5:24 PM, Eric Wong <normalperson at yhbt.net> wrote:
> Evan Weaver <evan at cloudbur.st> wrote:
>
> > Applied, thanks.
>  >
>  > Do you want a commit bit? What's your RubyForge name?
>
>  Sure, I'm normalperson: http://rubyforge.org/users/normalperson/
>
>  Thanks!
>
>
>
>  > Evan
>  >
>  > On Sat, Mar 1, 2008 at 4:38 PM, Eric Wong <normalperson at yhbt.net> wrote:
>  > > Tests all run correctly on the stable_1-2 branch now.
>  > >  ---
>  > >
>  > >   Since I'm looking at the tests... :)
>  > >
>  > >   Evan Weaver <evan at cloudbur.st> wrote:
>  > >   > That would be great; I haven't gotten to it. Can you merge it against
>  > >   > branches/stable_1-2?
>  > >   >
>  > >   > Also make sure to add a test if it changes any API or adds any functionality.
>  > >   >
>  > >   > Thanks
>  > >   >
>  > >   > Evan
>  > >
>  > >   test/unit/test_ws.rb |    2 +-
>  > >   1 files changed, 1 insertions(+), 1 deletions(-)
>  > >
>  > >  diff --git a/test/unit/test_ws.rb b/test/unit/test_ws.rb
>  > >  index d239e18..68a79a8 100644
>  > >  --- a/test/unit/test_ws.rb
>  > >  +++ b/test/unit/test_ws.rb
>  > >  @@ -4,7 +4,7 @@
>  > >   # Additional work donated by contributors.  See http://mongrel.rubyforge.org/attributions.html
>  > >   # for more information.
>  > >
>  > >  -require 'test/testhelp'
>  > >  +require 'test/test_helper'
>  > >
>  > >   include Mongrel
>  > >
>  > >  --
>  > >  Eric Wong
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From normalperson at yhbt.net  Sat Mar  1 18:52:46 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 1 Mar 2008 15:52:46 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080229222731.GA13406@untitled>
References: <20080229015309.GA9080@untitled>
	<3ae7f4480802290203v50023b57se867e531c162f066@mail.gmail.com>
	<3ae7f4480802290221o2929dd9ev154859dd5475808c@mail.gmail.com>
	<20080229222731.GA13406@untitled>
Message-ID: <20080301235246.GA30656@mayonaise>

Eric Wong <normalperson at yhbt.net> wrote:
> ry dahl <ry at tinyclouds.org> wrote:
> > while i'm nitpicking...:P
> > 
> > #define ASCII_UPPER(ch) ('a' <= ch && ch <= 'z' ? ch - 'a' + 'A' : ch)
> > 
> >     memcpy( RSTRING_PTR(f)
> >           , RSTRING_PTR(global_http_prefix)
> >           , RSTRING_LEN(global_http_prefix)
> >           );
> >     for(i = 0; i < length; i++) {
> >       ch = RSTRING_PTR(f) + RSTRING_LEN(global_http_prefix) + i;
> >       if(field[i] == '-') {
> >         *ch = '_';
> >       } else {
> >         *ch = ASCII_UPPER(field[i]);
> >       }
> >     }
> 
> The ASCII_UPPER helps readability, thanks.
> 
> I haven't had a chance to benchmark your for loop vs my while loop I
> wrote, but given the wide variety of compilers, versions and options, I
> expect my while loop to be slightly faster in general.

My version is about 0.3% faster :)

> Additionally, reordering the if/else condition to put the more common
> case first didn't seem to help in my case (gcc 4.1.1-21 from Debian,
> -O2), so I left it alone.
> 
> However, I haven't tried __builtin_expect() here, either (fast_xs did
> benefit from __builtin_expect() in my tests).

__builtin_expect didn't help, probably because the common cases
(lowercase chars) for HTTP headers don't overwhelmingly favor the less
common cases (non-lowercase chars).

> Perhaps rolling the checks together would be fastest, eliminating
> a redundant check for '-' if it's lowercase, still:
> 
>   *ch = ('a' <= *fch && *fch <= 'z') ?
>         (*fch - 'a' + 'A') :
>         (*fch != '-' ? *fch : '_');
 
> Not that other inefficiencies drastically overshadow the increases
> we get here :)

Here's my updated version, I've removed the extra 'i' variable and made
the code a little more compact (and hopefully more readable) (with no
measurable performance impact from my original).

I kept the bitwise operation for uppercasing instead of relying on
subtraction since it's typically faster (no sign/overflow checks)
although hard to measure in real-world performance here (my compiler
(gcc 4.1.1-21 -O2 on Debian/x86) preserved the subb instruction).

diff --git a/ext/http11/ext_help.h b/ext/http11/ext_help.h
index 8b4d754..1017c64 100644
--- a/ext/http11/ext_help.h
+++ b/ext/http11/ext_help.h
@@ -4,6 +4,8 @@
 #define RAISE_NOT_NULL(T) if(T == NULL) rb_raise(rb_eArgError, "NULL found for " # T " when shouldn't be.");
 #define DATA_GET(from,type,name) Data_Get_Struct(from,type,name); RAISE_NOT_NULL(name);
 #define REQUIRE_TYPE(V, T) if(TYPE(V) != T) rb_raise(rb_eTypeError, "Wrong argument type for " # V " required " # T);
+#define ASCII_UPCASE_CHAR(ch) (ch & ~0x20)
+
 
 #ifdef DEBUG
 #define TRACE()  fprintf(stderr, "> %s:%d:%s\n", __FILE__, __LINE__, __FUNCTION__)
diff --git a/ext/http11/http11.c b/ext/http11/http11.c
index 2982467..8e46e4c 100644
--- a/ext/http11/http11.c
+++ b/ext/http11/http11.c
@@ -7,7 +7,6 @@
 #include <assert.h>
 #include <string.h>
 #include "http11_parser.h"
-#include <ctype.h>
 
 static VALUE mMongrel;
 static VALUE cHttpParser;
@@ -62,7 +61,8 @@ DEF_MAX_LENGTH(HEADER, (1024 * (80 + 32)));
 
 void http_field(void *data, const char *field, size_t flen, const char *value, size_t vlen)
 {
-  char *ch, *end;
+  char *ch;
+  const char *fch;
   VALUE req = (VALUE)data;
   VALUE v = Qnil;
   VALUE f = Qnil;
@@ -71,15 +71,24 @@ void http_field(void *data, const char *field, size_t flen, const char *value, s
   VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE);
 
   v = rb_str_new(value, vlen);
-  f = rb_str_dup(global_http_prefix);
-  f = rb_str_buf_cat(f, field, flen); 
 
-  for(ch = RSTRING(f)->ptr, end = ch + RSTRING(f)->len; ch < end; ch++) {
-    if(*ch == '-') {
-      *ch = '_';
-    } else {
-      *ch = toupper(*ch);
-    }
+  /*
+   * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len)
+   * in my testing, because: there's no minimum allocation length (and
+   * no check for it, either), RSTRING(f)->len does not need to be
+   * written twice, and and RSTRING(f)->ptr[len] will already be
+   * null-terminated for us.
+   */
+  f = rb_str_new(NULL, RSTRING(global_http_prefix)->len + flen);
+  memcpy(RSTRING(f)->ptr,
+         RSTRING(global_http_prefix)->ptr,
+         RSTRING(global_http_prefix)->len);
+
+  ch = RSTRING(f)->ptr + RSTRING(global_http_prefix)->len;
+  for(fch = field; flen-- != 0; ++fch) {
+    *ch++ = (*fch >= 'a' && *fch <= 'z') ?
+            ASCII_UPCASE_CHAR(*fch) :
+            (*fch == '-' ? '_' : *fch);
   }
 
   rb_hash_aset(req, f, v);
-- 
Eric Wong

From filipe at icewall.org  Sat Mar  1 20:43:28 2008
From: filipe at icewall.org (Filipe)
Date: Sat, 1 Mar 2008 22:43:28 -0300 (BRT)
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
 in header parsing
In-Reply-To: <20080301235246.GA30656@mayonaise>
References: <20080229015309.GA9080@untitled>
	<3ae7f4480802290203v50023b57se867e531c162f066@mail.gmail.com>
	<3ae7f4480802290221o2929dd9ev154859dd5475808c@mail.gmail.com>
	<20080229222731.GA13406@untitled> <20080301235246.GA30656@mayonaise>
Message-ID: <Pine.LNX.4.64.0803012238090.3722@sophie.icewall.org>


Hey Eric! as now you have commit permission, please apply the patches
yourself to head. after testing it on head we can add this to branches,
ok?

Cheers,

filipe {
  @              icewall.org
  GPG         1024D/A6BA423E
  http://filipe.icewall.org/
}


On Sat, 1 Mar 2008, Eric Wong wrote:

> Eric Wong <normalperson at yhbt.net> wrote:
>> ry dahl <ry at tinyclouds.org> wrote:
>>> while i'm nitpicking...:P
>>>
>>> #define ASCII_UPPER(ch) ('a' <= ch && ch <= 'z' ? ch - 'a' + 'A' : ch)
>>>
>>>     memcpy( RSTRING_PTR(f)
>>>           , RSTRING_PTR(global_http_prefix)
>>>           , RSTRING_LEN(global_http_prefix)
>>>           );
>>>     for(i = 0; i < length; i++) {
>>>       ch = RSTRING_PTR(f) + RSTRING_LEN(global_http_prefix) + i;
>>>       if(field[i] == '-') {
>>>         *ch = '_';
>>>       } else {
>>>         *ch = ASCII_UPPER(field[i]);
>>>       }
>>>     }
>>
>> The ASCII_UPPER helps readability, thanks.
>>
>> I haven't had a chance to benchmark your for loop vs my while loop I
>> wrote, but given the wide variety of compilers, versions and options, I
>> expect my while loop to be slightly faster in general.
>
> My version is about 0.3% faster :)
>
>> Additionally, reordering the if/else condition to put the more common
>> case first didn't seem to help in my case (gcc 4.1.1-21 from Debian,
>> -O2), so I left it alone.
>>
>> However, I haven't tried __builtin_expect() here, either (fast_xs did
>> benefit from __builtin_expect() in my tests).
>
> __builtin_expect didn't help, probably because the common cases
> (lowercase chars) for HTTP headers don't overwhelmingly favor the less
> common cases (non-lowercase chars).
>
>> Perhaps rolling the checks together would be fastest, eliminating
>> a redundant check for '-' if it's lowercase, still:
>>
>>   *ch = ('a' <= *fch && *fch <= 'z') ?
>>         (*fch - 'a' + 'A') :
>>         (*fch != '-' ? *fch : '_');
>
>> Not that other inefficiencies drastically overshadow the increases
>> we get here :)
>
> Here's my updated version, I've removed the extra 'i' variable and made
> the code a little more compact (and hopefully more readable) (with no
> measurable performance impact from my original).
>
> I kept the bitwise operation for uppercasing instead of relying on
> subtraction since it's typically faster (no sign/overflow checks)
> although hard to measure in real-world performance here (my compiler
> (gcc 4.1.1-21 -O2 on Debian/x86) preserved the subb instruction).
>
> diff --git a/ext/http11/ext_help.h b/ext/http11/ext_help.h
> index 8b4d754..1017c64 100644
> --- a/ext/http11/ext_help.h
> +++ b/ext/http11/ext_help.h
> @@ -4,6 +4,8 @@
> #define RAISE_NOT_NULL(T) if(T == NULL) rb_raise(rb_eArgError, "NULL found for " # T " when shouldn't be.");
> #define DATA_GET(from,type,name) Data_Get_Struct(from,type,name); RAISE_NOT_NULL(name);
> #define REQUIRE_TYPE(V, T) if(TYPE(V) != T) rb_raise(rb_eTypeError, "Wrong argument type for " # V " required " # T);
> +#define ASCII_UPCASE_CHAR(ch) (ch & ~0x20)
> +
>
> #ifdef DEBUG
> #define TRACE()  fprintf(stderr, "> %s:%d:%s\n", __FILE__, __LINE__, __FUNCTION__)
> diff --git a/ext/http11/http11.c b/ext/http11/http11.c
> index 2982467..8e46e4c 100644
> --- a/ext/http11/http11.c
> +++ b/ext/http11/http11.c
> @@ -7,7 +7,6 @@
> #include <assert.h>
> #include <string.h>
> #include "http11_parser.h"
> -#include <ctype.h>
>
> static VALUE mMongrel;
> static VALUE cHttpParser;
> @@ -62,7 +61,8 @@ DEF_MAX_LENGTH(HEADER, (1024 * (80 + 32)));
>
> void http_field(void *data, const char *field, size_t flen, const char *value, size_t vlen)
> {
> -  char *ch, *end;
> +  char *ch;
> +  const char *fch;
>   VALUE req = (VALUE)data;
>   VALUE v = Qnil;
>   VALUE f = Qnil;
> @@ -71,15 +71,24 @@ void http_field(void *data, const char *field, size_t flen, const char *value, s
>   VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE);
>
>   v = rb_str_new(value, vlen);
> -  f = rb_str_dup(global_http_prefix);
> -  f = rb_str_buf_cat(f, field, flen);
>
> -  for(ch = RSTRING(f)->ptr, end = ch + RSTRING(f)->len; ch < end; ch++) {
> -    if(*ch == '-') {
> -      *ch = '_';
> -    } else {
> -      *ch = toupper(*ch);
> -    }
> +  /*
> +   * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len)
> +   * in my testing, because: there's no minimum allocation length (and
> +   * no check for it, either), RSTRING(f)->len does not need to be
> +   * written twice, and and RSTRING(f)->ptr[len] will already be
> +   * null-terminated for us.
> +   */
> +  f = rb_str_new(NULL, RSTRING(global_http_prefix)->len + flen);
> +  memcpy(RSTRING(f)->ptr,
> +         RSTRING(global_http_prefix)->ptr,
> +         RSTRING(global_http_prefix)->len);
> +
> +  ch = RSTRING(f)->ptr + RSTRING(global_http_prefix)->len;
> +  for(fch = field; flen-- != 0; ++fch) {
> +    *ch++ = (*fch >= 'a' && *fch <= 'z') ?
> +            ASCII_UPCASE_CHAR(*fch) :
> +            (*fch == '-' ? '_' : *fch);
>   }
>
>   rb_hash_aset(req, f, v);
> -- 
> Eric Wong
> _______________________________________________
> Mongrel-development mailing list
> Mongrel-development at rubyforge.org
> http://rubyforge.org/mailman/listinfo/mongrel-development
>

From normalperson at yhbt.net  Sat Mar  1 21:39:21 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 1 Mar 2008 18:39:21 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <Pine.LNX.4.64.0803012238090.3722@sophie.icewall.org>
References: <20080229015309.GA9080@untitled>
	<3ae7f4480802290203v50023b57se867e531c162f066@mail.gmail.com>
	<3ae7f4480802290221o2929dd9ev154859dd5475808c@mail.gmail.com>
	<20080229222731.GA13406@untitled>
	<20080301235246.GA30656@mayonaise>
	<Pine.LNX.4.64.0803012238090.3722@sophie.icewall.org>
Message-ID: <20080302023921.GA8733@mayonaise>

Filipe <filipe at icewall.org> wrote:
> 
> Hey Eric! as now you have commit permission, please apply the patches
> yourself to head. after testing it on head we can add this to branches,
> ok?

All three applied to trunk:

r989 | http11: ~6% performance increase in header parsing
r988 | mongrel: avoid needless syscall when num_processors limit is reached
r987 | mongrel_rails: support -n/--num-procs command-line option

I also noticed there was a general performance decrease in the trunk
HTTP parser versus stable-1-1 and 1-2.  I haven't had time to dig
into this, yet.  r989 seems to bring it inline with the performance
of the unpatched stable-1-2 branch...

> On Sat, 1 Mar 2008, Eric Wong wrote:
> 
> > Eric Wong <normalperson at yhbt.net> wrote:
> >> ry dahl <ry at tinyclouds.org> wrote:
> >>> while i'm nitpicking...:P
> >>>
> >>> #define ASCII_UPPER(ch) ('a' <= ch && ch <= 'z' ? ch - 'a' + 'A' : ch)
> >>>
> >>>     memcpy( RSTRING_PTR(f)
> >>>           , RSTRING_PTR(global_http_prefix)
> >>>           , RSTRING_LEN(global_http_prefix)
> >>>           );
> >>>     for(i = 0; i < length; i++) {
> >>>       ch = RSTRING_PTR(f) + RSTRING_LEN(global_http_prefix) + i;
> >>>       if(field[i] == '-') {
> >>>         *ch = '_';
> >>>       } else {
> >>>         *ch = ASCII_UPPER(field[i]);
> >>>       }
> >>>     }
> >>
> >> The ASCII_UPPER helps readability, thanks.
> >>
> >> I haven't had a chance to benchmark your for loop vs my while loop I
> >> wrote, but given the wide variety of compilers, versions and options, I
> >> expect my while loop to be slightly faster in general.
> >
> > My version is about 0.3% faster :)
> >
> >> Additionally, reordering the if/else condition to put the more common
> >> case first didn't seem to help in my case (gcc 4.1.1-21 from Debian,
> >> -O2), so I left it alone.
> >>
> >> However, I haven't tried __builtin_expect() here, either (fast_xs did
> >> benefit from __builtin_expect() in my tests).
> >
> > __builtin_expect didn't help, probably because the common cases
> > (lowercase chars) for HTTP headers don't overwhelmingly favor the less
> > common cases (non-lowercase chars).
> >
> >> Perhaps rolling the checks together would be fastest, eliminating
> >> a redundant check for '-' if it's lowercase, still:
> >>
> >>   *ch = ('a' <= *fch && *fch <= 'z') ?
> >>         (*fch - 'a' + 'A') :
> >>         (*fch != '-' ? *fch : '_');
> >
> >> Not that other inefficiencies drastically overshadow the increases
> >> we get here :)
> >
> > Here's my updated version, I've removed the extra 'i' variable and made
> > the code a little more compact (and hopefully more readable) (with no
> > measurable performance impact from my original).
> >
> > I kept the bitwise operation for uppercasing instead of relying on
> > subtraction since it's typically faster (no sign/overflow checks)
> > although hard to measure in real-world performance here (my compiler
> > (gcc 4.1.1-21 -O2 on Debian/x86) preserved the subb instruction).
> >
> > diff --git a/ext/http11/ext_help.h b/ext/http11/ext_help.h
> > index 8b4d754..1017c64 100644
> > --- a/ext/http11/ext_help.h
> > +++ b/ext/http11/ext_help.h
> > @@ -4,6 +4,8 @@
> > #define RAISE_NOT_NULL(T) if(T == NULL) rb_raise(rb_eArgError, "NULL found for " # T " when shouldn't be.");
> > #define DATA_GET(from,type,name) Data_Get_Struct(from,type,name); RAISE_NOT_NULL(name);
> > #define REQUIRE_TYPE(V, T) if(TYPE(V) != T) rb_raise(rb_eTypeError, "Wrong argument type for " # V " required " # T);
> > +#define ASCII_UPCASE_CHAR(ch) (ch & ~0x20)
> > +
> >
> > #ifdef DEBUG
> > #define TRACE()  fprintf(stderr, "> %s:%d:%s\n", __FILE__, __LINE__, __FUNCTION__)
> > diff --git a/ext/http11/http11.c b/ext/http11/http11.c
> > index 2982467..8e46e4c 100644
> > --- a/ext/http11/http11.c
> > +++ b/ext/http11/http11.c
> > @@ -7,7 +7,6 @@
> > #include <assert.h>
> > #include <string.h>
> > #include "http11_parser.h"
> > -#include <ctype.h>
> >
> > static VALUE mMongrel;
> > static VALUE cHttpParser;
> > @@ -62,7 +61,8 @@ DEF_MAX_LENGTH(HEADER, (1024 * (80 + 32)));
> >
> > void http_field(void *data, const char *field, size_t flen, const char *value, size_t vlen)
> > {
> > -  char *ch, *end;
> > +  char *ch;
> > +  const char *fch;
> >   VALUE req = (VALUE)data;
> >   VALUE v = Qnil;
> >   VALUE f = Qnil;
> > @@ -71,15 +71,24 @@ void http_field(void *data, const char *field, size_t flen, const char *value, s
> >   VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE);
> >
> >   v = rb_str_new(value, vlen);
> > -  f = rb_str_dup(global_http_prefix);
> > -  f = rb_str_buf_cat(f, field, flen);
> >
> > -  for(ch = RSTRING(f)->ptr, end = ch + RSTRING(f)->len; ch < end; ch++) {
> > -    if(*ch == '-') {
> > -      *ch = '_';
> > -    } else {
> > -      *ch = toupper(*ch);
> > -    }
> > +  /*
> > +   * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len)
> > +   * in my testing, because: there's no minimum allocation length (and
> > +   * no check for it, either), RSTRING(f)->len does not need to be
> > +   * written twice, and and RSTRING(f)->ptr[len] will already be
> > +   * null-terminated for us.
> > +   */
> > +  f = rb_str_new(NULL, RSTRING(global_http_prefix)->len + flen);
> > +  memcpy(RSTRING(f)->ptr,
> > +         RSTRING(global_http_prefix)->ptr,
> > +         RSTRING(global_http_prefix)->len);
> > +
> > +  ch = RSTRING(f)->ptr + RSTRING(global_http_prefix)->len;
> > +  for(fch = field; flen-- != 0; ++fch) {
> > +    *ch++ = (*fch >= 'a' && *fch <= 'z') ?
> > +            ASCII_UPCASE_CHAR(*fch) :
> > +            (*fch == '-' ? '_' : *fch);
> >   }
> >
> >   rb_hash_aset(req, f, v);
> > -- 
> > Eric Wong

From evan at cloudbur.st  Sat Mar  1 22:15:01 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 1 Mar 2008 22:15:01 -0500
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080302023921.GA8733@mayonaise>
References: <20080229015309.GA9080@untitled>
	<3ae7f4480802290203v50023b57se867e531c162f066@mail.gmail.com>
	<3ae7f4480802290221o2929dd9ev154859dd5475808c@mail.gmail.com>
	<20080229222731.GA13406@untitled> <20080301235246.GA30656@mayonaise>
	<Pine.LNX.4.64.0803012238090.3722@sophie.icewall.org>
	<20080302023921.GA8733@mayonaise>
Message-ID: <b6f68fc60803011915m568606d4yc54bdb056322538c@mail.gmail.com>

Trunk has the Ruby 1.9 compatibility changes, which might have
something to do with it. Those need to get backported to stable_1-2 at
some point. Trunk also has a new logger that Wayne wrote. That could
be degrading the parser performance somehow.

Evan

On Sat, Mar 1, 2008 at 9:39 PM, Eric Wong <normalperson at yhbt.net> wrote:
> Filipe <filipe at icewall.org> wrote:
>  >
>
> > Hey Eric! as now you have commit permission, please apply the patches
>  > yourself to head. after testing it on head we can add this to branches,
>  > ok?
>
>  All three applied to trunk:
>
>  r989 | http11: ~6% performance increase in header parsing
>  r988 | mongrel: avoid needless syscall when num_processors limit is reached
>  r987 | mongrel_rails: support -n/--num-procs command-line option
>
>  I also noticed there was a general performance decrease in the trunk
>  HTTP parser versus stable-1-1 and 1-2.  I haven't had time to dig
>  into this, yet.  r989 seems to bring it inline with the performance
>  of the unpatched stable-1-2 branch...
>
>
>
>  > On Sat, 1 Mar 2008, Eric Wong wrote:
>  >
>  > > Eric Wong <normalperson at yhbt.net> wrote:
>  > >> ry dahl <ry at tinyclouds.org> wrote:
>  > >>> while i'm nitpicking...:P
>  > >>>
>  > >>> #define ASCII_UPPER(ch) ('a' <= ch && ch <= 'z' ? ch - 'a' + 'A' : ch)
>  > >>>
>  > >>>     memcpy( RSTRING_PTR(f)
>  > >>>           , RSTRING_PTR(global_http_prefix)
>  > >>>           , RSTRING_LEN(global_http_prefix)
>  > >>>           );
>  > >>>     for(i = 0; i < length; i++) {
>  > >>>       ch = RSTRING_PTR(f) + RSTRING_LEN(global_http_prefix) + i;
>  > >>>       if(field[i] == '-') {
>  > >>>         *ch = '_';
>  > >>>       } else {
>  > >>>         *ch = ASCII_UPPER(field[i]);
>  > >>>       }
>  > >>>     }
>  > >>
>  > >> The ASCII_UPPER helps readability, thanks.
>  > >>
>  > >> I haven't had a chance to benchmark your for loop vs my while loop I
>  > >> wrote, but given the wide variety of compilers, versions and options, I
>  > >> expect my while loop to be slightly faster in general.
>  > >
>  > > My version is about 0.3% faster :)
>  > >
>  > >> Additionally, reordering the if/else condition to put the more common
>  > >> case first didn't seem to help in my case (gcc 4.1.1-21 from Debian,
>  > >> -O2), so I left it alone.
>  > >>
>  > >> However, I haven't tried __builtin_expect() here, either (fast_xs did
>  > >> benefit from __builtin_expect() in my tests).
>  > >
>  > > __builtin_expect didn't help, probably because the common cases
>  > > (lowercase chars) for HTTP headers don't overwhelmingly favor the less
>  > > common cases (non-lowercase chars).
>  > >
>  > >> Perhaps rolling the checks together would be fastest, eliminating
>  > >> a redundant check for '-' if it's lowercase, still:
>  > >>
>  > >>   *ch = ('a' <= *fch && *fch <= 'z') ?
>  > >>         (*fch - 'a' + 'A') :
>  > >>         (*fch != '-' ? *fch : '_');
>  > >
>  > >> Not that other inefficiencies drastically overshadow the increases
>  > >> we get here :)
>  > >
>  > > Here's my updated version, I've removed the extra 'i' variable and made
>  > > the code a little more compact (and hopefully more readable) (with no
>  > > measurable performance impact from my original).
>  > >
>  > > I kept the bitwise operation for uppercasing instead of relying on
>  > > subtraction since it's typically faster (no sign/overflow checks)
>  > > although hard to measure in real-world performance here (my compiler
>  > > (gcc 4.1.1-21 -O2 on Debian/x86) preserved the subb instruction).
>  > >
>  > > diff --git a/ext/http11/ext_help.h b/ext/http11/ext_help.h
>  > > index 8b4d754..1017c64 100644
>  > > --- a/ext/http11/ext_help.h
>  > > +++ b/ext/http11/ext_help.h
>  > > @@ -4,6 +4,8 @@
>  > > #define RAISE_NOT_NULL(T) if(T == NULL) rb_raise(rb_eArgError, "NULL found for " # T " when shouldn't be.");
>  > > #define DATA_GET(from,type,name) Data_Get_Struct(from,type,name); RAISE_NOT_NULL(name);
>  > > #define REQUIRE_TYPE(V, T) if(TYPE(V) != T) rb_raise(rb_eTypeError, "Wrong argument type for " # V " required " # T);
>  > > +#define ASCII_UPCASE_CHAR(ch) (ch & ~0x20)
>  > > +
>  > >
>  > > #ifdef DEBUG
>  > > #define TRACE()  fprintf(stderr, "> %s:%d:%s\n", __FILE__, __LINE__, __FUNCTION__)
>  > > diff --git a/ext/http11/http11.c b/ext/http11/http11.c
>  > > index 2982467..8e46e4c 100644
>  > > --- a/ext/http11/http11.c
>  > > +++ b/ext/http11/http11.c
>  > > @@ -7,7 +7,6 @@
>  > > #include <assert.h>
>  > > #include <string.h>
>  > > #include "http11_parser.h"
>  > > -#include <ctype.h>
>  > >
>  > > static VALUE mMongrel;
>  > > static VALUE cHttpParser;
>  > > @@ -62,7 +61,8 @@ DEF_MAX_LENGTH(HEADER, (1024 * (80 + 32)));
>  > >
>  > > void http_field(void *data, const char *field, size_t flen, const char *value, size_t vlen)
>  > > {
>  > > -  char *ch, *end;
>  > > +  char *ch;
>  > > +  const char *fch;
>  > >   VALUE req = (VALUE)data;
>  > >   VALUE v = Qnil;
>  > >   VALUE f = Qnil;
>  > > @@ -71,15 +71,24 @@ void http_field(void *data, const char *field, size_t flen, const char *value, s
>  > >   VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE);
>  > >
>  > >   v = rb_str_new(value, vlen);
>  > > -  f = rb_str_dup(global_http_prefix);
>  > > -  f = rb_str_buf_cat(f, field, flen);
>  > >
>  > > -  for(ch = RSTRING(f)->ptr, end = ch + RSTRING(f)->len; ch < end; ch++) {
>  > > -    if(*ch == '-') {
>  > > -      *ch = '_';
>  > > -    } else {
>  > > -      *ch = toupper(*ch);
>  > > -    }
>  > > +  /*
>  > > +   * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len)
>  > > +   * in my testing, because: there's no minimum allocation length (and
>  > > +   * no check for it, either), RSTRING(f)->len does not need to be
>  > > +   * written twice, and and RSTRING(f)->ptr[len] will already be
>  > > +   * null-terminated for us.
>  > > +   */
>  > > +  f = rb_str_new(NULL, RSTRING(global_http_prefix)->len + flen);
>  > > +  memcpy(RSTRING(f)->ptr,
>  > > +         RSTRING(global_http_prefix)->ptr,
>  > > +         RSTRING(global_http_prefix)->len);
>  > > +
>  > > +  ch = RSTRING(f)->ptr + RSTRING(global_http_prefix)->len;
>  > > +  for(fch = field; flen-- != 0; ++fch) {
>  > > +    *ch++ = (*fch >= 'a' && *fch <= 'z') ?
>  > > +            ASCII_UPCASE_CHAR(*fch) :
>  > > +            (*fch == '-' ? '_' : *fch);
>  > >   }
>  > >
>  > >   rb_hash_aset(req, f, v);
>  > > --
>  > > Eric Wong
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From zedshaw at zedshaw.com  Mon Mar  3 04:46:59 2008
From: zedshaw at zedshaw.com (Zed A. Shaw)
Date: Mon, 3 Mar 2008 04:46:59 -0500
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
 in header parsing
In-Reply-To: <20080229015309.GA9080@untitled>
References: <20080229015309.GA9080@untitled>
Message-ID: <20080303044659.5a550c19.zedshaw@zedshaw.com>

On Thu, 28 Feb 2008 17:53:09 -0800
Eric Wong <normalperson at yhbt.net> wrote:

> Allocate one string object and avoid appending to it causing it
> to be resized.  Additionally, optimize the string toupper copy
> so that it's done in a single pass.

Not bad, but I have some code style comments on your C code:

* I think the RSTRING macro is going away in the new ruby.  Ran into
that recently.
* Your inner loop isn't that readable and too clever.  The original
for loop is clear, and should have the same or more speed when
handed to a modern compiler.  Since you either already know the length
of the target string, or it's a ruby string that ends in \0, just use a
regular pointer loop to go through the string.  No point starting at
the back and decrementing a counter that also increment other pointers.
Additionally, the compiler should make it faster with pointers directly
instead and there's CPU operations that do this natively anyway, but
nothing using the non-standard backward-increment-forward-dual-pointers
thing you have.
* After your memcpy you don't clamp the string with a \0 since you're
relying on the backwards loop to do it.  This is kind of too clever to
be useful, so I'd do it the regular way and memcpy then set '\0'.
That's also safer.
* You
should also double check the code in ruby to make sure that
any other RString internals don't need to be updated.  I believe
there's a few other things that have to be touched to keep ruby sane
for some operations.  Look at how the ruby dup operations are done.
* You should get some real HTTP test samples, the worse the better.
All you've done is shown that it's marginally faster for a single run
of a simple set of headers.  I'd validate other combinations before
putting in code this complex for just a small boost like that.
* Also, now that I think about it, if you don't care that the original
string is modified in place then you can just have ragel do all of this
as it goes.  Simply modify the parser to have it do this transform on
the header chars using the existing pointer.  That'd probably be
alright since people don't usually keep the input headers around when
using the mongrel parser.

Give those suggestions a try and see what you get.  Thanks for helping
out Eric.

Zed

>  void http_field(void *data, const char *field, size_t flen, const char *value, size_t vlen)
>  {
> -  char *ch, *end;
> +  char *ch;
> +  const char *fch;
> +  int i;
>    VALUE req = (VALUE)data;
>    VALUE v = Qnil;
>    VALUE f = Qnil;
> @@ -71,15 +72,30 @@ void http_field(void *data, const char *field, size_t flen, const char *value, s
>    VALIDATE_MAX_LENGTH(vlen, FIELD_VALUE);
>  
>    v = rb_str_new(value, vlen);
> -  f = rb_str_dup(global_http_prefix);
> -  f = rb_str_buf_cat(f, field, flen); 
>  
> -  for(ch = RSTRING(f)->ptr, end = ch + RSTRING(f)->len; ch < end; ch++) {
> -    if(*ch == '-') {
> +  /*
> +   * using rb_str_new(NULL, len) here is faster than rb_str_buf_new(len)
> +   * in my testing, because: there's no minimum allocation length (and
> +   * no check for it, either), RSTRING(f)->len does not need to be
> +   * written twice, and and RSTRING(f)->ptr[len] will already be
> +   * null-terminated for us.
> +   */
> +  f = rb_str_new(NULL, RSTRING(global_http_prefix)->len + flen);
> +  memcpy(RSTRING(f)->ptr,
> +         RSTRING(global_http_prefix)->ptr,
> +         RSTRING(global_http_prefix)->len);
> +
> +  i = (int)flen;
> +  fch = field;
> +  ch = RSTRING(f)->ptr + RSTRING(global_http_prefix)->len;
> +  while(--i >= 0) {
> +    if(*fch == '-') {
>        *ch = '_';
>      } else {
> -      *ch = toupper(*ch);
> +      *ch = (*fch >= 'a' && *fch <= 'z') ? (*fch & ~0x20) : *fch;
>      }
> +    ++ch;
> +    ++fch;
>    }
>  
>    rb_hash_aset(req, f, v);
> -- 
> Eric Wong
> _______________________________________________
> Mongrel-development mailing list
> Mongrel-development at rubyforge.org
> http://rubyforge.org/mailman/listinfo/mongrel-development


-- 
Zed A. Shaw
- Hate: http://savingtheinternetwithhate.com/
- Good: http://www.zedshaw.com/
- Evil: http://yearofevil.com/

From normalperson at yhbt.net  Sun Mar  2 07:37:12 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sun, 2 Mar 2008 04:37:12 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080303044659.5a550c19.zedshaw@zedshaw.com>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
Message-ID: <20080302123712.GA13979@hand.yhbt.net>

"Zed A. Shaw" <zedshaw at zedshaw.com> wrote:
> On Thu, 28 Feb 2008 17:53:09 -0800
> Eric Wong <normalperson at yhbt.net> wrote:
> 
> > Allocate one string object and avoid appending to it causing it
> > to be resized.  Additionally, optimize the string toupper copy
> > so that it's done in a single pass.
> 
> Not bad, but I have some code style comments on your C code:
> 
> * I think the RSTRING macro is going away in the new ruby.  Ran into
> that recently.

Yup.  The version I committed to trunk uses the 1.9-compatible
RSTRING_{PTR,LEN} macros

> * Your inner loop isn't that readable and too clever.  The original
> for loop is clear, and should have the same or more speed when
> handed to a modern compiler.  Since you either already know the length
> of the target string, or it's a ruby string that ends in \0, just use a
> regular pointer loop to go through the string.  No point starting at
> the back and decrementing a counter that also increment other pointers.
> Additionally, the compiler should make it faster with pointers directly
> instead and there's CPU operations that do this natively anyway, but
> nothing using the non-standard backward-increment-forward-dual-pointers
> thing you have.

I actually shortened the one committed to trunk a bit (but used nested
ternary operators, so I may rewrite that for better readability).

It modifies flen directly instead of copying it to a temporary variable.
I don't think Ragel actually terminates field with '\0', but I'll double
check later.

> * After your memcpy you don't clamp the string with a \0 since you're
> relying on the backwards loop to do it.  This is kind of too clever to
> be useful, so I'd do it the regular way and memcpy then set '\0'.
> That's also safer.

rb_str_new() actually does it, the loop shouldn't overwrite it.  But I
understand if you want me to clamp it again manually for easier
auditing.

> * You
> should also double check the code in ruby to make sure that
> any other RString internals don't need to be updated.  I believe
> there's a few other things that have to be touched to keep ruby sane
> for some operations.  Look at how the ruby dup operations are done.
> * You should get some real HTTP test samples, the worse the better.
> All you've done is shown that it's marginally faster for a single run
> of a simple set of headers.  I'd validate other combinations before
> putting in code this complex for just a small boost like that.
> * Also, now that I think about it, if you don't care that the original
> string is modified in place then you can just have ragel do all of this
> as it goes.  Simply modify the parser to have it do this transform on
> the header chars using the existing pointer.  That'd probably be
> alright since people don't usually keep the input headers around when
> using the mongrel parser.

I'll look into all of those three.  I think I saw rb_str_modify() but
either forgot to use it or figured I didn't need it :X

> Give those suggestions a try and see what you get.  Thanks for helping
> out Eric.

Thanks for the feedback, Zed.

There's also something else that occured to me is why we always create a
new RString object in the first place if it's getting set into a Hash
(and frozen).  I suspect we'll get a much bigger performance increase by
avoiding redundant object creation when we get common header names and
directly using global objects to represent them (like global_http_host).


Additionally, my understanding is that frozen objects never get GC-ed,
and since strings get frozen when being put into a hash, this could
be a memory leak if somebody unleashes a random header generator.

Maybe we should keep a global store of predefined HTTP headers we always
accept (all those mentioned in HTTP RFCs), and leave 100 or so slots
open for some random headers clients may put in;
letting those weird clients fill those slots on a first-come basis.

Afterwards, any new headers we haven't seen and don't know about would
either raise an exception or get silently dropped ...

This would prevent memory leaks if my understanding of frozen objects
is correct...


Way past my bedtime to test these things now, though.

-- 
Eric Wong

From evan at cloudbur.st  Sun Mar  2 15:31:44 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sun, 2 Mar 2008 15:31:44 -0500
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080302123712.GA13979@hand.yhbt.net>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
Message-ID: <b6f68fc60803021231x316d5f0fn8718142b6ab2cb02@mail.gmail.com>

> This would prevent memory leaks if my understanding of frozen objects
> is correct...

I don't think the GC treats frozen objects in any special way.

Evan

On Sun, Mar 2, 2008 at 7:37 AM, Eric Wong <normalperson at yhbt.net> wrote:
> "Zed A. Shaw" <zedshaw at zedshaw.com> wrote:
>  > On Thu, 28 Feb 2008 17:53:09 -0800
>  > Eric Wong <normalperson at yhbt.net> wrote:
>  >
>  > > Allocate one string object and avoid appending to it causing it
>  > > to be resized.  Additionally, optimize the string toupper copy
>  > > so that it's done in a single pass.
>  >
>  > Not bad, but I have some code style comments on your C code:
>  >
>  > * I think the RSTRING macro is going away in the new ruby.  Ran into
>  > that recently.
>
>  Yup.  The version I committed to trunk uses the 1.9-compatible
>  RSTRING_{PTR,LEN} macros
>
>
>  > * Your inner loop isn't that readable and too clever.  The original
>  > for loop is clear, and should have the same or more speed when
>  > handed to a modern compiler.  Since you either already know the length
>  > of the target string, or it's a ruby string that ends in \0, just use a
>  > regular pointer loop to go through the string.  No point starting at
>  > the back and decrementing a counter that also increment other pointers.
>  > Additionally, the compiler should make it faster with pointers directly
>  > instead and there's CPU operations that do this natively anyway, but
>  > nothing using the non-standard backward-increment-forward-dual-pointers
>  > thing you have.
>
>  I actually shortened the one committed to trunk a bit (but used nested
>  ternary operators, so I may rewrite that for better readability).
>
>  It modifies flen directly instead of copying it to a temporary variable.
>  I don't think Ragel actually terminates field with '\0', but I'll double
>  check later.
>
>
>  > * After your memcpy you don't clamp the string with a \0 since you're
>  > relying on the backwards loop to do it.  This is kind of too clever to
>  > be useful, so I'd do it the regular way and memcpy then set '\0'.
>  > That's also safer.
>
>  rb_str_new() actually does it, the loop shouldn't overwrite it.  But I
>  understand if you want me to clamp it again manually for easier
>  auditing.
>
>
>  > * You
>  > should also double check the code in ruby to make sure that
>  > any other RString internals don't need to be updated.  I believe
>  > there's a few other things that have to be touched to keep ruby sane
>  > for some operations.  Look at how the ruby dup operations are done.
>  > * You should get some real HTTP test samples, the worse the better.
>  > All you've done is shown that it's marginally faster for a single run
>  > of a simple set of headers.  I'd validate other combinations before
>  > putting in code this complex for just a small boost like that.
>  > * Also, now that I think about it, if you don't care that the original
>  > string is modified in place then you can just have ragel do all of this
>  > as it goes.  Simply modify the parser to have it do this transform on
>  > the header chars using the existing pointer.  That'd probably be
>  > alright since people don't usually keep the input headers around when
>  > using the mongrel parser.
>
>  I'll look into all of those three.  I think I saw rb_str_modify() but
>  either forgot to use it or figured I didn't need it :X
>
>
>  > Give those suggestions a try and see what you get.  Thanks for helping
>  > out Eric.
>
>  Thanks for the feedback, Zed.
>
>  There's also something else that occured to me is why we always create a
>  new RString object in the first place if it's getting set into a Hash
>  (and frozen).  I suspect we'll get a much bigger performance increase by
>  avoiding redundant object creation when we get common header names and
>  directly using global objects to represent them (like global_http_host).
>
>
>  Additionally, my understanding is that frozen objects never get GC-ed,
>  and since strings get frozen when being put into a hash, this could
>  be a memory leak if somebody unleashes a random header generator.
>
>  Maybe we should keep a global store of predefined HTTP headers we always
>  accept (all those mentioned in HTTP RFCs), and leave 100 or so slots
>  open for some random headers clients may put in;
>  letting those weird clients fill those slots on a first-come basis.
>
>  Afterwards, any new headers we haven't seen and don't know about would
>  either raise an exception or get silently dropped ...
>
>  This would prevent memory leaks if my understanding of frozen objects
>  is correct...
>
>
>  Way past my bedtime to test these things now, though.
>
>
>
>  --
>  Eric Wong
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From normalperson at yhbt.net  Sun Mar  2 17:35:19 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sun, 2 Mar 2008 14:35:19 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <b6f68fc60803021231x316d5f0fn8718142b6ab2cb02@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<b6f68fc60803021231x316d5f0fn8718142b6ab2cb02@mail.gmail.com>
Message-ID: <20080302223519.GB13979@hand.yhbt.net>

Evan Weaver <evan at cloudbur.st> wrote:
> > This would prevent memory leaks if my understanding of frozen objects
> > is correct...
> 
> I don't think the GC treats frozen objects in any special way.

Oops, it was symbols that leaked, not frozen objects in general.  Phew.

-- 
Eric Wong

From normalperson at yhbt.net  Thu Mar  6 02:54:21 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Wed, 5 Mar 2008 23:54:21 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080302123712.GA13979@hand.yhbt.net>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
Message-ID: <20080306075421.GA1583@hand.yhbt.net>

Eric Wong <normalperson at yhbt.net> wrote:
> "Zed A. Shaw" <zedshaw at zedshaw.com> wrote:
> > On Thu, 28 Feb 2008 17:53:09 -0800
> > Eric Wong <normalperson at yhbt.net> wrote:
> > * You
> > should also double check the code in ruby to make sure that
> > any other RString internals don't need to be updated.  I believe
> > there's a few other things that have to be touched to keep ruby sane
> > for some operations.  Look at how the ruby dup operations are done.

rb_str_modify() only needs to be called for duplicated strings.
For new strings, it's safe to write directly to the pointer after
rb_str_new(0, len) is called (the 1.8.6.111 source has several
examples of this).

> > * You should get some real HTTP test samples, the worse the better.
> > All you've done is shown that it's marginally faster for a single run
> > of a simple set of headers.  I'd validate other combinations before
> > putting in code this complex for just a small boost like that.

I still need to get to that Friday or this weekend and test with Rfuzz,
too.  I've added some more unit tests to ensure my code works for now.

r992 is a fairly big change (with pretty good benefits, see below), but
I've coded that in as fool-proof a way as I possibly could.

> > * Also, now that I think about it, if you don't care that the original
> > string is modified in place then you can just have ragel do all of this
> > as it goes.  Simply modify the parser to have it do this transform on
> > the header chars using the existing pointer.  That'd probably be
> > alright since people don't usually keep the input headers around when
> > using the mongrel parser.

done in r990

> I'll look into all of those three.  I think I saw rb_str_modify() but
> either forgot to use it or figured I didn't need it :X
> 
> > Give those suggestions a try and see what you get.  Thanks for helping
> > out Eric.
> 
> Thanks for the feedback, Zed.
> 
> There's also something else that occured to me is why we always create a
> new RString object in the first place if it's getting set into a Hash
> (and frozen).  I suspect we'll get a much bigger performance increase by
> avoiding redundant object creation when we get common header names and
> directly using global objects to represent them (like global_http_host).

I'm getting an even bigger (~22%) performance improvement by predefining
some common HTTP headers as global frozen strings upfront (r992)

> Additionally, my understanding is that frozen objects never get GC-ed,
> and since strings get frozen when being put into a hash, this could
> be a memory leak if somebody unleashes a random header generator.

Thankfully, I was wrong about frozen strings :)

-- 
Eric Wong

From ry at tinyclouds.org  Thu Mar  6 04:53:34 2008
From: ry at tinyclouds.org (ry dahl)
Date: Thu, 6 Mar 2008 10:53:34 +0100
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080306075421.GA1583@hand.yhbt.net>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
Message-ID: <3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>

Hi Eric,

>  I'm getting an even bigger (~22%) performance
> improvement by predefining
>  some common HTTP headers as global frozen
> strings upfront (r992)

Why don't you do this in Ragel? It will be faster and you don't need
to depend on bsearch. I pull out content-length header in ragel in
ebb:
 http://github.com/ry/ebb/tree/master/src/parser.rl
(although there is a bug with this because (content_length |
message_header ) needs some priorities set so message_header isn't
matching content-length too. should be
   (content_length >2 | message_header >1 )
or something).

Also - the state machine should be upgraded to compile with ragel 6.
this basically involves removing %write eof;

ry

From normalperson at yhbt.net  Sat Mar  8 03:12:10 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 8 Mar 2008 00:12:10 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
Message-ID: <20080308081210.GA30702@hand.yhbt.net>

ry dahl <ry at tinyclouds.org> wrote:
> Hi Eric,
> 
> >  I'm getting an even bigger (~22%) performance
> > improvement by predefining
> >  some common HTTP headers as global frozen
> > strings upfront (r992)
> 
> Why don't you do this in Ragel? It will be faster and you don't need
> to depend on bsearch. I pull out content-length header in ragel in
> ebb:
>  http://github.com/ry/ebb/tree/master/src/parser.rl
> (although there is a bug with this because (content_length |
> message_header ) needs some priorities set so message_header isn't
> matching content-length too. should be
>    (content_length >2 | message_header >1 )
> or something).

Then I'd have to define a new C function for every header I wanted to
optimize for, and then also point to that function inside the Ragel file
for each header.

Unless we use something like ERB to generate this code for both Ragel
and C, I'm not sure it's worth the effort to go through with all the
extra code.

Currently, with my C/CPP code, I can add or remove headers to memoize
strings for by adding or removing one line per header in the C file.
Pretty much as easy as it gets maintenance-wise.

> Also - the state machine should be upgraded to compile with ragel 6.
> this basically involves removing %write eof;

Yes.  At the same time, I'm not sure if I should force every other
developer to upgrade... Evan?  Zed?

-- 
Eric Wong

From evan at cloudbur.st  Sat Mar  8 03:14:40 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 8 Mar 2008 03:14:40 -0500
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080308081210.GA30702@hand.yhbt.net>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
Message-ID: <b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>

Upgrading ragel is fine with me.

Evan

On Sat, Mar 8, 2008 at 3:12 AM, Eric Wong <normalperson at yhbt.net> wrote:
> ry dahl <ry at tinyclouds.org> wrote:
>
> > Hi Eric,
>  >
>  > >  I'm getting an even bigger (~22%) performance
>  > > improvement by predefining
>  > >  some common HTTP headers as global frozen
>  > > strings upfront (r992)
>  >
>  > Why don't you do this in Ragel? It will be faster and you don't need
>  > to depend on bsearch. I pull out content-length header in ragel in
>  > ebb:
>  >  http://github.com/ry/ebb/tree/master/src/parser.rl
>  > (although there is a bug with this because (content_length |
>  > message_header ) needs some priorities set so message_header isn't
>  > matching content-length too. should be
>  >    (content_length >2 | message_header >1 )
>  > or something).
>
>  Then I'd have to define a new C function for every header I wanted to
>  optimize for, and then also point to that function inside the Ragel file
>  for each header.
>
>  Unless we use something like ERB to generate this code for both Ragel
>  and C, I'm not sure it's worth the effort to go through with all the
>  extra code.
>
>  Currently, with my C/CPP code, I can add or remove headers to memoize
>  strings for by adding or removing one line per header in the C file.
>  Pretty much as easy as it gets maintenance-wise.
>
>
>  > Also - the state machine should be upgraded to compile with ragel 6.
>  > this basically involves removing %write eof;
>
>  Yes.  At the same time, I'm not sure if I should force every other
>  developer to upgrade... Evan?  Zed?
>
>  --
>  Eric Wong
>
>
> _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar  8 07:37:53 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 8 Mar 2008 10:37:53 -0200
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080308081210.GA30702@hand.yhbt.net>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
Message-ID: <71166b3b0803080437w2809929ajc5bf9ec022c1512c@mail.gmail.com>

On Sat, Mar 8, 2008 at 6:12 AM, Eric Wong <normalperson at yhbt.net> wrote:
>
>  Unless we use something like ERB to generate this code for both Ragel
>  and C, I'm not sure it's worth the effort to go through with all the
>  extra code.
>

I think is not worth the effort, and also add complexity to the code
generation, which should be the simpler as possible.

>  Yes.  At the same time, I'm not sure if I should force every other
>  developer to upgrade... Evan?  Zed?

Well, the Ragel compilation is not run every time, but only when the
parser get updated, so I can say will be fine the most common cases
(developers playing with mongrel code already handle the ragel
compiled code).

-- 
Luis Lavena
Multimedia systems
-
A common mistake that people make when trying to design
something completely foolproof is to underestimate
the ingenuity of complete fools.
Douglas Adams

From ry at tinyclouds.org  Sat Mar  8 12:37:53 2008
From: ry at tinyclouds.org (ry dahl)
Date: Sat, 8 Mar 2008 18:37:53 +0100
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080308081210.GA30702@hand.yhbt.net>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
Message-ID: <3ae7f4480803080937y20b729e0vb9b549cb8184118f@mail.gmail.com>

>  Then I'd have to define a new C function for every header I wanted to
>  optimize for, and then also point to that function inside the Ragel file
>  for each header.

Nah, you can just define

enum header_types {
 MONGREL_CONTENT_LENGTH,
 MONGREL_CONTENT_TYPE,
 etc
}

and define a single callback to handle them

void (*field_cb)(http_parser*, int header_type, char *at, int len)

ry

From normalperson at yhbt.net  Sat Mar  8 17:02:20 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Sat, 8 Mar 2008 14:02:20 -0800
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
	<b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>
Message-ID: <20080308220220.GA27749@mayonaise>

Evan Weaver <evan at cloudbur.st> wrote:
> Upgrading ragel is fine with me.

Here's my work-in-progress.  The changes below are pretty much copied
from ry's changes in ebb.

I haven't been able to figure out why
test_horrible_queries(HttpParserTest) is failing, so any help here would
be appreciated.

Thanks.

diff --git a/Rakefile b/Rakefile
index f47d7a2..07a89c2 100644
--- a/Rakefile
+++ b/Rakefile
@@ -55,13 +55,15 @@ task :ragel do
   Dir.chdir "ext/http11" do
     target = "http11_parser.c"
     File.unlink target if File.exist? target
-    sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}"
+    # sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}" # ragel 5.x
+    sh "ragel -G2 http11_parser.rl" # ragel 6.0
     raise "Failed to build C source" unless File.exist? target
   end
   Dir.chdir "ext/http11" do
     target = "../../ext/http11_java/org/jruby/mongrel/Http11Parser.java"
     File.unlink target if File.exist? target
-    sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}"
+    # sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}" # ragel 5.x
+    sh "ragel -J -o #{target} http11_parser.java.rl" # ragel 6.0
     raise "Failed to build Java source" unless File.exist? target
   end
 end
diff --git a/ext/http11/http11_parser.rl b/ext/http11/http11_parser.rl
index a418605..0c4e2d4 100644
--- a/ext/http11/http11_parser.rl
+++ b/ext/http11/http11_parser.rl
@@ -114,7 +114,7 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
   p = buffer+off;
   pe = buffer+len;
 
-  assert(*pe == '\0' && "pointer does not end on NUL");
+  /* assert(*pe == '\0' && "pointer does not end on NUL"); */
   assert(pe - p == len - off && "pointers aren't same distance");
 
 
@@ -130,23 +130,11 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
   assert(parser->field_len <= len && "field has length longer than whole buffer");
   assert(parser->field_start < len && "field starts after buffer end");
 
-  if(parser->body_start) {
-    /* final \r\n combo encountered so stop right here */
-    %%write eof;
-    parser->nread++;
-  }
-
   return(parser->nread);
 }
 
 int http_parser_finish(http_parser *parser)
 {
-  int cs = parser->cs;
-
-  %%write eof;
-
-  parser->cs = cs;
-
   if (http_parser_has_error(parser) ) {
     return -1;
   } else if (http_parser_is_finished(parser) ) {
@@ -161,5 +149,5 @@ int http_parser_has_error(http_parser *parser) {
 }
 
 int http_parser_is_finished(http_parser *parser) {
-  return parser->cs == http_parser_first_final;
+  return parser->cs >= http_parser_first_final;
 }
-- 
Eric Wong

From evan at cloudbur.st  Mon Mar 24 14:39:05 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Mon, 24 Mar 2008 14:39:05 -0400
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080308220220.GA27749@mayonaise>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
	<b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>
	<20080308220220.GA27749@mayonaise>
Message-ID: <b6f68fc60803241139g70b2d8a3v5d01e7472367fb74@mail.gmail.com>

Hi all,

I backported the trunk parser changes and applied the Ragel fix to
branches/stable_1-2 . Two tests are currently failing... could someone
(Eric, Ry?) take a look at these? I am not good enough with Ragel to
be able to say whether or not the eof change is related to the test
failures.

Thanks,

Evan

On Sat, Mar 8, 2008 at 6:02 PM, Eric Wong <normalperson at yhbt.net> wrote:
> Evan Weaver <evan at cloudbur.st> wrote:
>
> > Upgrading ragel is fine with me.
>
>  Here's my work-in-progress.  The changes below are pretty much copied
>  from ry's changes in ebb.
>
>  I haven't been able to figure out why
>  test_horrible_queries(HttpParserTest) is failing, so any help here would
>  be appreciated.
>
>  Thanks.
>
>  diff --git a/Rakefile b/Rakefile
>  index f47d7a2..07a89c2 100644
>  --- a/Rakefile
>  +++ b/Rakefile
>  @@ -55,13 +55,15 @@ task :ragel do
>    Dir.chdir "ext/http11" do
>      target = "http11_parser.c"
>      File.unlink target if File.exist? target
>  -    sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}"
>  +    # sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}" # ragel 5.x
>  +    sh "ragel -G2 http11_parser.rl" # ragel 6.0
>      raise "Failed to build C source" unless File.exist? target
>    end
>    Dir.chdir "ext/http11" do
>      target = "../../ext/http11_java/org/jruby/mongrel/Http11Parser.java"
>      File.unlink target if File.exist? target
>  -    sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}"
>  +    # sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}" # ragel 5.x
>  +    sh "ragel -J -o #{target} http11_parser.java.rl" # ragel 6.0
>      raise "Failed to build Java source" unless File.exist? target
>    end
>   end
>  diff --git a/ext/http11/http11_parser.rl b/ext/http11/http11_parser.rl
>  index a418605..0c4e2d4 100644
>  --- a/ext/http11/http11_parser.rl
>  +++ b/ext/http11/http11_parser.rl
>  @@ -114,7 +114,7 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
>    p = buffer+off;
>    pe = buffer+len;
>
>  -  assert(*pe == '\0' && "pointer does not end on NUL");
>  +  /* assert(*pe == '\0' && "pointer does not end on NUL"); */
>    assert(pe - p == len - off && "pointers aren't same distance");
>
>
>  @@ -130,23 +130,11 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
>    assert(parser->field_len <= len && "field has length longer than whole buffer");
>    assert(parser->field_start < len && "field starts after buffer end");
>
>  -  if(parser->body_start) {
>  -    /* final \r\n combo encountered so stop right here */
>  -    %%write eof;
>  -    parser->nread++;
>  -  }
>  -
>    return(parser->nread);
>   }
>
>   int http_parser_finish(http_parser *parser)
>   {
>  -  int cs = parser->cs;
>  -
>  -  %%write eof;
>  -
>  -  parser->cs = cs;
>  -
>    if (http_parser_has_error(parser) ) {
>      return -1;
>    } else if (http_parser_is_finished(parser) ) {
>  @@ -161,5 +149,5 @@ int http_parser_has_error(http_parser *parser) {
>   }
>
>   int http_parser_is_finished(http_parser *parser) {
>  -  return parser->cs == http_parser_first_final;
>  +  return parser->cs >= http_parser_first_final;
>   }
>  --
>
>
> Eric Wong
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From normalperson at yhbt.net  Mon Mar 24 23:43:07 2008
From: normalperson at yhbt.net (Eric Wong)
Date: Mon, 24 Mar 2008 20:43:07 -0700
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <b6f68fc60803241139g70b2d8a3v5d01e7472367fb74@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
	<b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>
	<20080308220220.GA27749@mayonaise>
	<b6f68fc60803241139g70b2d8a3v5d01e7472367fb74@mail.gmail.com>
Message-ID: <20080325034307.GA13379@soma>

Evan Weaver <evan at cloudbur.st> wrote:
> Hi all,
> 
> I backported the trunk parser changes and applied the Ragel fix to
> branches/stable_1-2 . Two tests are currently failing... could someone
> (Eric, Ry?) take a look at these? I am not good enough with Ragel to
> be able to say whether or not the eof change is related to the test
> failures.

Hi Evan,

I never committed the Ragel 6 upgrade to trunk because I couldn't figure
out why it was broken.  I was hoping someone else would step in and fix
it for us :)

Reverting my attempt at upgrading to Ragel 6 fixes one test...



As for the other test that fails, you uncommented the nasty_pound_header
test in the commit for multi-line.  This test doesn't work in
stable-1_1 nor trunk, and I've never made any effort into getting it
working (nor was it my fault it's broken in the first place :)

Digging into the revision history (with git :), the nasty_pound_header
test was committed in its commented form in r361 and has never changed
until now.

The Mongrel HTTP parser code does not appear to handle multi-line
HTTP headers at all, so I'm not surprised that test fails.  I would
suggest disabling the nasty_pound_header test again until we can
successfully parse multi-line headers.

> Thanks,
> 
> Evan
> 
> On Sat, Mar 8, 2008 at 6:02 PM, Eric Wong <normalperson at yhbt.net> wrote:
> > Evan Weaver <evan at cloudbur.st> wrote:
> >
> > > Upgrading ragel is fine with me.
> >
> >  Here's my work-in-progress.  The changes below are pretty much copied
> >  from ry's changes in ebb.
> >
> >  I haven't been able to figure out why
> >  test_horrible_queries(HttpParserTest) is failing, so any help here would
> >  be appreciated.
> >
> >  Thanks.
> >
> >  diff --git a/Rakefile b/Rakefile
> >  index f47d7a2..07a89c2 100644
> >  --- a/Rakefile
> >  +++ b/Rakefile
> >  @@ -55,13 +55,15 @@ task :ragel do
> >    Dir.chdir "ext/http11" do
> >      target = "http11_parser.c"
> >      File.unlink target if File.exist? target
> >  -    sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}"
> >  +    # sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}" # ragel 5.x
> >  +    sh "ragel -G2 http11_parser.rl" # ragel 6.0
> >      raise "Failed to build C source" unless File.exist? target
> >    end
> >    Dir.chdir "ext/http11" do
> >      target = "../../ext/http11_java/org/jruby/mongrel/Http11Parser.java"
> >      File.unlink target if File.exist? target
> >  -    sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}"
> >  +    # sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}" # ragel 5.x
> >  +    sh "ragel -J -o #{target} http11_parser.java.rl" # ragel 6.0
> >      raise "Failed to build Java source" unless File.exist? target
> >    end
> >   end
> >  diff --git a/ext/http11/http11_parser.rl b/ext/http11/http11_parser.rl
> >  index a418605..0c4e2d4 100644
> >  --- a/ext/http11/http11_parser.rl
> >  +++ b/ext/http11/http11_parser.rl
> >  @@ -114,7 +114,7 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
> >    p = buffer+off;
> >    pe = buffer+len;
> >
> >  -  assert(*pe == '\0' && "pointer does not end on NUL");
> >  +  /* assert(*pe == '\0' && "pointer does not end on NUL"); */
> >    assert(pe - p == len - off && "pointers aren't same distance");
> >
> >
> >  @@ -130,23 +130,11 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
> >    assert(parser->field_len <= len && "field has length longer than whole buffer");
> >    assert(parser->field_start < len && "field starts after buffer end");
> >
> >  -  if(parser->body_start) {
> >  -    /* final \r\n combo encountered so stop right here */
> >  -    %%write eof;
> >  -    parser->nread++;
> >  -  }
> >  -
> >    return(parser->nread);
> >   }
> >
> >   int http_parser_finish(http_parser *parser)
> >   {
> >  -  int cs = parser->cs;
> >  -
> >  -  %%write eof;
> >  -
> >  -  parser->cs = cs;
> >  -
> >    if (http_parser_has_error(parser) ) {
> >      return -1;
> >    } else if (http_parser_is_finished(parser) ) {
> >  @@ -161,5 +149,5 @@ int http_parser_has_error(http_parser *parser) {
> >   }
> >
> >   int http_parser_is_finished(http_parser *parser) {
> >  -  return parser->cs == http_parser_first_final;
> >  +  return parser->cs >= http_parser_first_final;
> >   }
> >  --
> >
> >
> > Eric Wong
> -- 
> Evan Weaver
> Cloudburst, LLC

From evan at cloudbur.st  Tue Mar 25 00:53:16 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Tue, 25 Mar 2008 00:53:16 -0400
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <20080325034307.GA13379@soma>
References: <20080229015309.GA9080@untitled>
	<20080303044659.5a550c19.zedshaw@zedshaw.com>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
	<b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>
	<20080308220220.GA27749@mayonaise>
	<b6f68fc60803241139g70b2d8a3v5d01e7472367fb74@mail.gmail.com>
	<20080325034307.GA13379@soma>
Message-ID: <b6f68fc60803242153i65dbb67re6bc9457060ee03f@mail.gmail.com>

>  I never committed the Ragel 6 upgrade to trunk because I couldn't figure
>  out why it was broken.  I was hoping someone else would step in and fix
>  it for us :)

Maybe Ry knows. Otherwise, I'll try to hack through it. I think
something might have to happen in the state transitions to make up for
the inability to explicitly close the output (is that even remotely
close?).

>  Digging into the revision history (with git :), the nasty_pound_header
>  test was committed in its commented form in r361 and has never changed
>  until now.

I'm not surprised, but there was no comment as to why it was disabled,
so I enabled it and let it fail. We should maybe remove that test
entirely and document multiline headers as not supported.

Thanks for looking in to this.

Evan



>  >
>  > On Sat, Mar 8, 2008 at 6:02 PM, Eric Wong <normalperson at yhbt.net> wrote:
>  > > Evan Weaver <evan at cloudbur.st> wrote:
>  > >
>  > > > Upgrading ragel is fine with me.
>  > >
>  > >  Here's my work-in-progress.  The changes below are pretty much copied
>  > >  from ry's changes in ebb.
>  > >
>  > >  I haven't been able to figure out why
>  > >  test_horrible_queries(HttpParserTest) is failing, so any help here would
>  > >  be appreciated.
>  > >
>  > >  Thanks.
>  > >
>  > >  diff --git a/Rakefile b/Rakefile
>  > >  index f47d7a2..07a89c2 100644
>  > >  --- a/Rakefile
>  > >  +++ b/Rakefile
>  > >  @@ -55,13 +55,15 @@ task :ragel do
>  > >    Dir.chdir "ext/http11" do
>  > >      target = "http11_parser.c"
>  > >      File.unlink target if File.exist? target
>  > >  -    sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}"
>  > >  +    # sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}" # ragel 5.x
>  > >  +    sh "ragel -G2 http11_parser.rl" # ragel 6.0
>  > >      raise "Failed to build C source" unless File.exist? target
>  > >    end
>  > >    Dir.chdir "ext/http11" do
>  > >      target = "../../ext/http11_java/org/jruby/mongrel/Http11Parser.java"
>  > >      File.unlink target if File.exist? target
>  > >  -    sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}"
>  > >  +    # sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}" # ragel 5.x
>  > >  +    sh "ragel -J -o #{target} http11_parser.java.rl" # ragel 6.0
>  > >      raise "Failed to build Java source" unless File.exist? target
>  > >    end
>  > >   end
>  > >  diff --git a/ext/http11/http11_parser.rl b/ext/http11/http11_parser.rl
>  > >  index a418605..0c4e2d4 100644
>  > >  --- a/ext/http11/http11_parser.rl
>  > >  +++ b/ext/http11/http11_parser.rl
>  > >  @@ -114,7 +114,7 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
>  > >    p = buffer+off;
>  > >    pe = buffer+len;
>  > >
>  > >  -  assert(*pe == '\0' && "pointer does not end on NUL");
>  > >  +  /* assert(*pe == '\0' && "pointer does not end on NUL"); */
>  > >    assert(pe - p == len - off && "pointers aren't same distance");
>  > >
>  > >
>  > >  @@ -130,23 +130,11 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
>  > >    assert(parser->field_len <= len && "field has length longer than whole buffer");
>  > >    assert(parser->field_start < len && "field starts after buffer end");
>  > >
>  > >  -  if(parser->body_start) {
>  > >  -    /* final \r\n combo encountered so stop right here */
>  > >  -    %%write eof;
>  > >  -    parser->nread++;
>  > >  -  }
>  > >  -
>  > >    return(parser->nread);
>  > >   }
>  > >
>  > >   int http_parser_finish(http_parser *parser)
>  > >   {
>  > >  -  int cs = parser->cs;
>  > >  -
>  > >  -  %%write eof;
>  > >  -
>  > >  -  parser->cs = cs;
>  > >  -
>  > >    if (http_parser_has_error(parser) ) {
>  > >      return -1;
>  > >    } else if (http_parser_is_finished(parser) ) {
>  > >  @@ -161,5 +149,5 @@ int http_parser_has_error(http_parser *parser) {
>  > >   }
>  > >
>  > >   int http_parser_is_finished(http_parser *parser) {
>  > >  -  return parser->cs == http_parser_first_final;
>  > >  +  return parser->cs >= http_parser_first_final;
>  > >   }
>  > >  --
>  > >
>  > >
>  > > Eric Wong
>  > --
>  > Evan Weaver
>  > Cloudburst, LLC
>
>
> _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From ry at tinyclouds.org  Tue Mar 25 04:13:24 2008
From: ry at tinyclouds.org (ry dahl)
Date: Tue, 25 Mar 2008 09:13:24 +0100
Subject: [Mongrel-development] [PATCH] http11: ~6% performance increase
	in header parsing
In-Reply-To: <b6f68fc60803242153i65dbb67re6bc9457060ee03f@mail.gmail.com>
References: <20080229015309.GA9080@untitled>
	<20080302123712.GA13979@hand.yhbt.net>
	<20080306075421.GA1583@hand.yhbt.net>
	<3ae7f4480803060153v18e955a5j9038f93e558f81d1@mail.gmail.com>
	<20080308081210.GA30702@hand.yhbt.net>
	<b6f68fc60803080014x3241d9b0n6430c5ec93f7ea2d@mail.gmail.com>
	<20080308220220.GA27749@mayonaise>
	<b6f68fc60803241139g70b2d8a3v5d01e7472367fb74@mail.gmail.com>
	<20080325034307.GA13379@soma>
	<b6f68fc60803242153i65dbb67re6bc9457060ee03f@mail.gmail.com>
Message-ID: <3ae7f4480803250113m4aefb748o308e841cb2abec7e@mail.gmail.com>

This is broken in Ebb too. I'll put some time into it again today and
see if I can fix it.
ry

On Tue, Mar 25, 2008 at 5:53 AM, Evan Weaver <evan at cloudbur.st> wrote:
> >  I never committed the Ragel 6 upgrade to trunk because I couldn't figure
>  >  out why it was broken.  I was hoping someone else would step in and fix
>  >  it for us :)
>
>  Maybe Ry knows. Otherwise, I'll try to hack through it. I think
>  something might have to happen in the state transitions to make up for
>  the inability to explicitly close the output (is that even remotely
>  close?).
>
>
>  >  Digging into the revision history (with git :), the nasty_pound_header
>  >  test was committed in its commented form in r361 and has never changed
>  >  until now.
>
>  I'm not surprised, but there was no comment as to why it was disabled,
>  so I enabled it and let it fail. We should maybe remove that test
>  entirely and document multiline headers as not supported.
>
>  Thanks for looking in to this.
>
>
>
>  Evan
>
>
>
>  >  >
>  >  > On Sat, Mar 8, 2008 at 6:02 PM, Eric Wong <normalperson at yhbt.net> wrote:
>  >  > > Evan Weaver <evan at cloudbur.st> wrote:
>  >  > >
>  >  > > > Upgrading ragel is fine with me.
>  >  > >
>  >  > >  Here's my work-in-progress.  The changes below are pretty much copied
>  >  > >  from ry's changes in ebb.
>  >  > >
>  >  > >  I haven't been able to figure out why
>  >  > >  test_horrible_queries(HttpParserTest) is failing, so any help here would
>  >  > >  be appreciated.
>  >  > >
>  >  > >  Thanks.
>  >  > >
>  >  > >  diff --git a/Rakefile b/Rakefile
>  >  > >  index f47d7a2..07a89c2 100644
>  >  > >  --- a/Rakefile
>  >  > >  +++ b/Rakefile
>  >  > >  @@ -55,13 +55,15 @@ task :ragel do
>  >  > >    Dir.chdir "ext/http11" do
>  >  > >      target = "http11_parser.c"
>  >  > >      File.unlink target if File.exist? target
>  >  > >  -    sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}"
>  >  > >  +    # sh "ragel http11_parser.rl | rlgen-cd -G2 -o #{target}" # ragel 5.x
>  >  > >  +    sh "ragel -G2 http11_parser.rl" # ragel 6.0
>  >  > >      raise "Failed to build C source" unless File.exist? target
>  >  > >    end
>  >  > >    Dir.chdir "ext/http11" do
>  >  > >      target = "../../ext/http11_java/org/jruby/mongrel/Http11Parser.java"
>  >  > >      File.unlink target if File.exist? target
>  >  > >  -    sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}"
>  >  > >  +    # sh "ragel -J http11_parser.java.rl | rlgen-java -o #{target}" # ragel 5.x
>  >  > >  +    sh "ragel -J -o #{target} http11_parser.java.rl" # ragel 6.0
>  >  > >      raise "Failed to build Java source" unless File.exist? target
>  >  > >    end
>  >  > >   end
>  >  > >  diff --git a/ext/http11/http11_parser.rl b/ext/http11/http11_parser.rl
>  >  > >  index a418605..0c4e2d4 100644
>  >  > >  --- a/ext/http11/http11_parser.rl
>  >  > >  +++ b/ext/http11/http11_parser.rl
>  >  > >  @@ -114,7 +114,7 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
>  >  > >    p = buffer+off;
>  >  > >    pe = buffer+len;
>  >  > >
>  >  > >  -  assert(*pe == '\0' && "pointer does not end on NUL");
>  >  > >  +  /* assert(*pe == '\0' && "pointer does not end on NUL"); */
>  >  > >    assert(pe - p == len - off && "pointers aren't same distance");
>  >  > >
>  >  > >
>  >  > >  @@ -130,23 +130,11 @@ size_t http_parser_execute(http_parser *parser, const char *buffer, size_t len,
>  >  > >    assert(parser->field_len <= len && "field has length longer than whole buffer");
>  >  > >    assert(parser->field_start < len && "field starts after buffer end");
>  >  > >
>  >  > >  -  if(parser->body_start) {
>  >  > >  -    /* final \r\n combo encountered so stop right here */
>  >  > >  -    %%write eof;
>  >  > >  -    parser->nread++;
>  >  > >  -  }
>  >  > >  -
>  >  > >    return(parser->nread);
>  >  > >   }
>  >  > >
>  >  > >   int http_parser_finish(http_parser *parser)
>  >  > >   {
>  >  > >  -  int cs = parser->cs;
>  >  > >  -
>  >  > >  -  %%write eof;
>  >  > >  -
>  >  > >  -  parser->cs = cs;
>  >  > >  -
>  >  > >    if (http_parser_has_error(parser) ) {
>  >  > >      return -1;
>  >  > >    } else if (http_parser_is_finished(parser) ) {
>  >  > >  @@ -161,5 +149,5 @@ int http_parser_has_error(http_parser *parser) {
>  >  > >   }
>  >  > >
>  >  > >   int http_parser_is_finished(http_parser *parser) {
>  >  > >  -  return parser->cs == http_parser_first_final;
>  >  > >  +  return parser->cs >= http_parser_first_final;
>  >  > >   }
>  >  > >  --
>  >  > >
>  >  > >
>  >  > > Eric Wong
>  >  > --
>  >  > Evan Weaver
>  >  > Cloudburst, LLC
>  >
>  >
>  > _______________________________________________
>  >  Mongrel-development mailing list
>  >  Mongrel-development at rubyforge.org
>  >  http://rubyforge.org/mailman/listinfo/mongrel-development
>  >
>
>
>
>  --
>  Evan Weaver
>  Cloudburst, LLC
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>

From luislavena at gmail.com  Sat Mar 29 09:24:45 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 10:24:45 -0300
Subject: [Mongrel-development] Can no longer checkout svn as anonymous
Message-ID: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>

Hello guys,

After the move to Trac, my CI can no longer checkout from the
repository. Been using the http path stated here:

http://rubyforge.org/scm/?group_id=1306

using this command:

svn checkout http://mongrel.rubyforge.org/svn/

which no longer works, since mongrel.rubyforge now resolves to the
trac instance.

Any workaround this? Maybe is a Rubyforge thing that RF staff can solve?

For the time being, I'm switching to svn:// protocol.

Just FYI.
-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Sat Mar 29 10:59:01 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 29 Mar 2008 10:59:01 -0400
Subject: [Mongrel-development] Can no longer checkout svn as anonymous
In-Reply-To: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>
References: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>
Message-ID: <b6f68fc60803290759y69ce4fan59e06cf5bc2b8a21@mail.gmail.com>

svn:// works, or you can point mongrel.rubyforge.org to the
Rubyforge.org IP in your /etc/hosts file.

If there's some Apache rewrite I can add so that that URL still gets
forwarded I will do it, but I wasn't quite sure how.

Evan

On Sat, Mar 29, 2008 at 9:24 AM, Luis Lavena <luislavena at gmail.com> wrote:
> Hello guys,
>
>  After the move to Trac, my CI can no longer checkout from the
>  repository. Been using the http path stated here:
>
>  http://rubyforge.org/scm/?group_id=1306
>
>  using this command:
>
>  svn checkout http://mongrel.rubyforge.org/svn/
>
>  which no longer works, since mongrel.rubyforge now resolves to the
>  trac instance.
>
>  Any workaround this? Maybe is a Rubyforge thing that RF staff can solve?
>
>  For the time being, I'm switching to svn:// protocol.
>
>  Just FYI.
>  --
>  Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar 29 11:03:02 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 12:03:02 -0300
Subject: [Mongrel-development] Can no longer checkout svn as anonymous
In-Reply-To: <b6f68fc60803290759y69ce4fan59e06cf5bc2b8a21@mail.gmail.com>
References: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>
	<b6f68fc60803290759y69ce4fan59e06cf5bc2b8a21@mail.gmail.com>
Message-ID: <71166b3b0803290803s792fd3ffk22fa4a9f12f89da4@mail.gmail.com>

On Sat, Mar 29, 2008 at 11:59 AM, Evan Weaver <evan at cloudbur.st> wrote:
> svn:// works, or you can point mongrel.rubyforge.org to the
>  Rubyforge.org IP in your /etc/hosts file.
>

Started to use svn:// for the CI, thank you.

>  If there's some Apache rewrite I can add so that that URL still gets
>  forwarded I will do it, but I wasn't quite sure how.
>

I don't know. Last time I played with apache was in the 90's and with
1.x version I guess.

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Sat Mar 29 11:10:04 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 29 Mar 2008 11:10:04 -0400
Subject: [Mongrel-development] Can no longer checkout svn as anonymous
In-Reply-To: <71166b3b0803290803s792fd3ffk22fa4a9f12f89da4@mail.gmail.com>
References: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>
	<b6f68fc60803290759y69ce4fan59e06cf5bc2b8a21@mail.gmail.com>
	<71166b3b0803290803s792fd3ffk22fa4a9f12f89da4@mail.gmail.com>
Message-ID: <b6f68fc60803290810v50948afco159909ae336f79d7@mail.gmail.com>

I guess I could just try it. I wasn't sure if you could forward to a
remote server in a sane way.

Evan

On Sat, Mar 29, 2008 at 11:03 AM, Luis Lavena <luislavena at gmail.com> wrote:
> On Sat, Mar 29, 2008 at 11:59 AM, Evan Weaver <evan at cloudbur.st> wrote:
>  > svn:// works, or you can point mongrel.rubyforge.org to the
>  >  Rubyforge.org IP in your /etc/hosts file.
>  >
>
>  Started to use svn:// for the CI, thank you.
>
>
>  >  If there's some Apache rewrite I can add so that that URL still gets
>  >  forwarded I will do it, but I wasn't quite sure how.
>  >
>
>  I don't know. Last time I played with apache was in the 90's and with
>  1.x version I guess.
>
>  --
>
>
> Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar 29 11:15:05 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 12:15:05 -0300
Subject: [Mongrel-development] Can no longer checkout svn as anonymous
In-Reply-To: <b6f68fc60803290810v50948afco159909ae336f79d7@mail.gmail.com>
References: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>
	<b6f68fc60803290759y69ce4fan59e06cf5bc2b8a21@mail.gmail.com>
	<71166b3b0803290803s792fd3ffk22fa4a9f12f89da4@mail.gmail.com>
	<b6f68fc60803290810v50948afco159909ae336f79d7@mail.gmail.com>
Message-ID: <71166b3b0803290815v38b41433oa83a38560ea08769@mail.gmail.com>

On Sat, Mar 29, 2008 at 12:10 PM, Evan Weaver <evan at cloudbur.st> wrote:
> I guess I could just try it. I wasn't sure if you could forward to a
>  remote server in a sane way.

svn will follow redirects, by default,
http://mongrel.rubyforge.org/svn/ was redirecting to itself... (at
least that was bzr pulling reported).

Regards and thank you,
-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Sat Mar 29 11:23:47 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 29 Mar 2008 11:23:47 -0400
Subject: [Mongrel-development] Can no longer checkout svn as anonymous
In-Reply-To: <71166b3b0803290815v38b41433oa83a38560ea08769@mail.gmail.com>
References: <71166b3b0803290624h4b77a7ffu4a1b3dcc38e8ec94@mail.gmail.com>
	<b6f68fc60803290759y69ce4fan59e06cf5bc2b8a21@mail.gmail.com>
	<71166b3b0803290803s792fd3ffk22fa4a9f12f89da4@mail.gmail.com>
	<b6f68fc60803290810v50948afco159909ae336f79d7@mail.gmail.com>
	<71166b3b0803290815v38b41433oa83a38560ea08769@mail.gmail.com>
Message-ID: <b6f68fc60803290823ma763a11kea0961f82796eb45@mail.gmail.com>

Rubyforge needs to see the Host: header to know which repository to
access, so just a plain 30x isn't good enough.

Evan

On Sat, Mar 29, 2008 at 11:15 AM, Luis Lavena <luislavena at gmail.com> wrote:
> On Sat, Mar 29, 2008 at 12:10 PM, Evan Weaver <evan at cloudbur.st> wrote:
>  > I guess I could just try it. I wasn't sure if you could forward to a
>  >  remote server in a sane way.
>
>  svn will follow redirects, by default,
>  http://mongrel.rubyforge.org/svn/ was redirecting to itself... (at
>  least that was bzr pulling reported).
>
>  Regards and thank you,
>  --
>
>
> Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar 29 17:56:12 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 18:56:12 -0300
Subject: [Mongrel-development] Stats of latest commits
Message-ID: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>

Hello guys,

Looks like latests commits (either to trunk or stable_1-2) broke the
build completely (at least, for me).

trunk: 996 (normalperson)
ruby 1.8.6 (2007-09-24 patchlevel 111) [i386-mswin32]

61 tests, 494 assertions, 2 failures, 1 errors

===

  1) Failure:
test_base_handler_config(ConfiguratorTest) [./test/test_configurator.rb:82]:
<[Errno::EBADF, Errno::ECONNREFUSED]> exception expected but was
Class: <Errno::ECONNRESET>
Message: <"An existing connection was forcibly closed by the remote host.">
---Backtrace---
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/timeout.rb:56:in `timeout'
C:/Ruby/lib/ruby/1.8/timeout.rb:76:in `timeout'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
C:/Ruby/lib/ruby/1.8/net/http.rb:2029:in `read_status_line'
C:/Ruby/lib/ruby/1.8/net/http.rb:2018:in `read_new'
C:/Ruby/lib/ruby/1.8/net/http.rb:1059:in `request'
C:/Ruby/lib/ruby/1.8/net/http.rb:957:in `request_get'
C:/Ruby/lib/ruby/1.8/net/http.rb:380:in `get_response'
C:/Ruby/lib/ruby/1.8/net/http.rb:547:in `start'
C:/Ruby/lib/ruby/1.8/net/http.rb:379:in `get_response'
C:/Ruby/lib/ruby/1.8/net/http.rb:356:in `get'
./test/test_configurator.rb:83:in `test_base_handler_config'
./test/test_configurator.rb:82:in `test_base_handler_config'
---------------

  2) Failure:
test_horrible_queries(HttpParserTest)
    [./test/test_http11.rb:221:in `test_horrible_queries'
     ./test/test_http11.rb:219:in `times'
     ./test/test_http11.rb:219:in `test_horrible_queries']:
<Mongrel::HttpParserError> exception expected but none was thrown.

  3) Error:
test_parse_ie6_urls(HttpParserTest):
Mongrel::HttpParserError: Invalid HTTP format, parsing fails.
    ./test/test_http11.rb:67:in `execute'
    ./test/test_http11.rb:67:in `test_parse_ie6_urls'
    ./test/test_http11.rb:58:in `each'
    ./test/test_http11.rb:58:in `test_parse_ie6_urls'

===

"stable", on the other hand, is doom, pastie follows:

http://pastie.caboo.se/172564


Don't know what happened, or maybe it's me :-P

regards,
-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Sat Mar 29 18:02:31 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 29 Mar 2008 18:02:31 -0400
Subject: [Mongrel-development] Stats of latest commits
In-Reply-To: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
References: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
Message-ID: <b6f68fc60803291502t61cae96cg5bc69b1405595473@mail.gmail.com>

Port already in use?

Evan

On Sat, Mar 29, 2008 at 5:56 PM, Luis Lavena <luislavena at gmail.com> wrote:
> Hello guys,
>
>  Looks like latests commits (either to trunk or stable_1-2) broke the
>  build completely (at least, for me).
>
>  trunk: 996 (normalperson)
>  ruby 1.8.6 (2007-09-24 patchlevel 111) [i386-mswin32]
>
>  61 tests, 494 assertions, 2 failures, 1 errors
>
>  ===
>
>   1) Failure:
>  test_base_handler_config(ConfiguratorTest) [./test/test_configurator.rb:82]:
>  <[Errno::EBADF, Errno::ECONNREFUSED]> exception expected but was
>  Class: <Errno::ECONNRESET>
>  Message: <"An existing connection was forcibly closed by the remote host.">
>  ---Backtrace---
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
>  C:/Ruby/lib/ruby/1.8/timeout.rb:56:in `timeout'
>  C:/Ruby/lib/ruby/1.8/timeout.rb:76:in `timeout'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:2029:in `read_status_line'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:2018:in `read_new'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:1059:in `request'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:957:in `request_get'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:380:in `get_response'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:547:in `start'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:379:in `get_response'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:356:in `get'
>  ./test/test_configurator.rb:83:in `test_base_handler_config'
>  ./test/test_configurator.rb:82:in `test_base_handler_config'
>  ---------------
>
>   2) Failure:
>  test_horrible_queries(HttpParserTest)
>     [./test/test_http11.rb:221:in `test_horrible_queries'
>      ./test/test_http11.rb:219:in `times'
>      ./test/test_http11.rb:219:in `test_horrible_queries']:
>  <Mongrel::HttpParserError> exception expected but none was thrown.
>
>   3) Error:
>  test_parse_ie6_urls(HttpParserTest):
>  Mongrel::HttpParserError: Invalid HTTP format, parsing fails.
>     ./test/test_http11.rb:67:in `execute'
>     ./test/test_http11.rb:67:in `test_parse_ie6_urls'
>     ./test/test_http11.rb:58:in `each'
>     ./test/test_http11.rb:58:in `test_parse_ie6_urls'
>
>  ===
>
>  "stable", on the other hand, is doom, pastie follows:
>
>  http://pastie.caboo.se/172564
>
>
>  Don't know what happened, or maybe it's me :-P
>
>  regards,
>  --
>  Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar 29 18:03:24 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 19:03:24 -0300
Subject: [Mongrel-development] Stats of latest commits
In-Reply-To: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
References: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
Message-ID: <71166b3b0803291503r6dd6e335o8b98240432d4be5@mail.gmail.com>

On Sat, Mar 29, 2008 at 6:56 PM, Luis Lavena <luislavena at gmail.com> wrote:
> Hello guys,
>
>  Looks like latests commits (either to trunk or stable_1-2) broke the
>  build completely (at least, for me).


Updated, it seems the sequence of rake tasks matters:

rake clobber compile test
(trunk)

  1) Failure:
test_base_handler_config(ConfiguratorTest) [./test/test_configurator.rb:82]:
<[Errno::EBADF, Errno::ECONNREFUSED]> exception expected but was
Class: <Errno::ECONNRESET>
Message: <"An existing connection was forcibly closed by the remote host.">
---Backtrace---
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/timeout.rb:56:in `timeout'
C:/Ruby/lib/ruby/1.8/timeout.rb:76:in `timeout'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
C:/Ruby/lib/ruby/1.8/net/http.rb:2029:in `read_status_line'
C:/Ruby/lib/ruby/1.8/net/http.rb:2018:in `read_new'
C:/Ruby/lib/ruby/1.8/net/http.rb:1059:in `request'
C:/Ruby/lib/ruby/1.8/net/http.rb:957:in `request_get'
C:/Ruby/lib/ruby/1.8/net/http.rb:380:in `get_response'
C:/Ruby/lib/ruby/1.8/net/http.rb:547:in `start'
C:/Ruby/lib/ruby/1.8/net/http.rb:379:in `get_response'
C:/Ruby/lib/ruby/1.8/net/http.rb:356:in `get'
./test/test_configurator.rb:83:in `test_base_handler_config'
./test/test_configurator.rb:82:in `test_base_handler_config'
---------------

61 tests, 541 assertions, 1 failures, 0 errors

===

stable_1-2, on the other hand, still highly broken:

59 tests, 416 assertions, 2 failures, 35 errors

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Sat Mar 29 18:06:23 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 29 Mar 2008 18:06:23 -0400
Subject: [Mongrel-development] Stats of latest commits
In-Reply-To: <71166b3b0803291503r6dd6e335o8b98240432d4be5@mail.gmail.com>
References: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
	<71166b3b0803291503r6dd6e335o8b98240432d4be5@mail.gmail.com>
Message-ID: <b6f68fc60803291506te584459xb9142398012191aa@mail.gmail.com>

Can you try binary searching stable to see when it broke? (Probably 995 though.)

Evan

On Sat, Mar 29, 2008 at 6:03 PM, Luis Lavena <luislavena at gmail.com> wrote:
> On Sat, Mar 29, 2008 at 6:56 PM, Luis Lavena <luislavena at gmail.com> wrote:
>
> > Hello guys,
>  >
>  >  Looks like latests commits (either to trunk or stable_1-2) broke the
>  >  build completely (at least, for me).
>
>
>  Updated, it seems the sequence of rake tasks matters:
>
>  rake clobber compile test
>  (trunk)
>
>
>   1) Failure:
>  test_base_handler_config(ConfiguratorTest) [./test/test_configurator.rb:82]:
>  <[Errno::EBADF, Errno::ECONNREFUSED]> exception expected but was
>  Class: <Errno::ECONNRESET>
>  Message: <"An existing connection was forcibly closed by the remote host.">
>  ---Backtrace---
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
>  C:/Ruby/lib/ruby/1.8/timeout.rb:56:in `timeout'
>  C:/Ruby/lib/ruby/1.8/timeout.rb:76:in `timeout'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
>  C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:2029:in `read_status_line'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:2018:in `read_new'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:1059:in `request'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:957:in `request_get'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:380:in `get_response'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:547:in `start'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:379:in `get_response'
>  C:/Ruby/lib/ruby/1.8/net/http.rb:356:in `get'
>  ./test/test_configurator.rb:83:in `test_base_handler_config'
>  ./test/test_configurator.rb:82:in `test_base_handler_config'
>  ---------------
>
>  61 tests, 541 assertions, 1 failures, 0 errors
>
>  ===
>
>  stable_1-2, on the other hand, still highly broken:
>
>  59 tests, 416 assertions, 2 failures, 35 errors
>
>
>
>  --
>  Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar 29 18:07:02 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 19:07:02 -0300
Subject: [Mongrel-development] Stats of latest commits
In-Reply-To: <b6f68fc60803291502t61cae96cg5bc69b1405595473@mail.gmail.com>
References: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
	<b6f68fc60803291502t61cae96cg5bc69b1405595473@mail.gmail.com>
Message-ID: <71166b3b0803291507q6f40848cnda017bec2861c0cf@mail.gmail.com>

On Sat, Mar 29, 2008 at 7:02 PM, Evan Weaver <evan at cloudbur.st> wrote:
> Port already in use?
>

Yep, looks like one test isn't teardown properly, but that only
happens on stable branch, not in trunk (on trunk only one failure
remains).

It seems the merge didn't happen properly? (or something is missing).

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Sat Mar 29 18:12:27 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Sat, 29 Mar 2008 18:12:27 -0400
Subject: [Mongrel-development] Stats of latest commits
In-Reply-To: <71166b3b0803291507q6f40848cnda017bec2861c0cf@mail.gmail.com>
References: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
	<b6f68fc60803291502t61cae96cg5bc69b1405595473@mail.gmail.com>
	<71166b3b0803291507q6f40848cnda017bec2861c0cf@mail.gmail.com>
Message-ID: <b6f68fc60803291512y75ce4f7bx2df0638cdeb65673@mail.gmail.com>

It is possible that I just plain broke it--maybe related to the order
of the tests, which could be different for you than for me.

Try reverting the change around the method mentioned in
http://mongrel.rubyforge.org/ticket/19 . The solution in the ticket
didn't really make sense and I need someone else to verify my change.

Evan

On Sat, Mar 29, 2008 at 6:07 PM, Luis Lavena <luislavena at gmail.com> wrote:
> On Sat, Mar 29, 2008 at 7:02 PM, Evan Weaver <evan at cloudbur.st> wrote:
>  > Port already in use?
>  >
>
>  Yep, looks like one test isn't teardown properly, but that only
>  happens on stable branch, not in trunk (on trunk only one failure
>  remains).
>
>  It seems the merge didn't happen properly? (or something is missing).
>
>  --
>
>
> Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Sat Mar 29 18:15:37 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Sat, 29 Mar 2008 19:15:37 -0300
Subject: [Mongrel-development] Stats of latest commits
In-Reply-To: <b6f68fc60803291512y75ce4f7bx2df0638cdeb65673@mail.gmail.com>
References: <71166b3b0803291456j3e40b6a3oe796400ef8e83bc3@mail.gmail.com>
	<b6f68fc60803291502t61cae96cg5bc69b1405595473@mail.gmail.com>
	<71166b3b0803291507q6f40848cnda017bec2861c0cf@mail.gmail.com>
	<b6f68fc60803291512y75ce4f7bx2df0638cdeb65673@mail.gmail.com>
Message-ID: <71166b3b0803291515v5b197286n7bffa838c88eeab1@mail.gmail.com>

On Sat, Mar 29, 2008 at 7:12 PM, Evan Weaver <evan at cloudbur.st> wrote:
> It is possible that I just plain broke it--maybe related to the order
>  of the tests, which could be different for you than for me.
>
>  Try reverting the change around the method mentioned in
>  http://mongrel.rubyforge.org/ticket/19 . The solution in the ticket
>  didn't really make sense and I need someone else to verify my change.
>

Can you try this in latest version?

ruby -Ilib;ext;bin;test
"C:/Ruby/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb"
"test/test_helper.rb" "test/unit/test_cgi_wrapper.rb"
"test/unit/test_command.rb" "test/unit/test_conditional.rb"
"test/unit/test_configurator.rb" "test/unit/test_debug.rb"
"test/unit/test_handlers.rb" "test/unit/test_http_parser.rb"
"test/unit/test_redirect_handler.rb"
"test/unit/test_request_progress.rb" "test/unit/test_response.rb"
"test/unit/test_stats.rb" "test/unit/test_uriclassifier.rb"
"test/unit/test_ws.rb"

(only change the rake_test_loader.rb location).

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From builder at mmediasys.com  Sun Mar 30 14:03:27 2008
From: builder at mmediasys.com (builder at mmediasys.com)
Date: Sun, 30 Mar 2008 15:03:27 -0300
Subject: [Mongrel-development] [CruiseControl] mongrel-trunk-mswin32 build
	996 failed
Message-ID: <47efd5ef6ab08_c4813de3a61a6@KEORE.tmail>

The build failed.

  Note: if you set Configuration.dashboard_url in config/site_config.rb, you'd see a link to the build page here.

CHANGES
-------
This is the first build


BUILD LOG
---------
D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"  
rake aborted!
Could not find RubyGem echoe (>= 2.7.5)
D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work/rakefile:3
(See full trace by running task with --trace)
(in D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work)

dir : D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work
command : ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"
executed command : echo D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/build-996/build.log && ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:\Users\Luis\.cruise\projects\mongrel-trunk-mswin32\build-996\build.log 2>&1
exitstatus: 1
STDERR TAIL START

STDERR TAIL END


PROJECT SETTINGS
----------------
# Project-specific configuration for CruiseControl.rb

Project.configure do |project|

  project.email_notifier.emails = ['mongrel-development at rubyforge.org', 'rubyinstaller-devel at rubyforge.org']

  project.rake_task = 'clobber compile test'
  
  project.scheduler.polling_interval = 5.minutes

end


From builder at mmediasys.com  Sun Mar 30 14:16:23 2008
From: builder at mmediasys.com (builder at mmediasys.com)
Date: Sun, 30 Mar 2008 15:16:23 -0300
Subject: [Mongrel-development] [CruiseControl] mongrel-trunk-mswin32 build
	997 failed
Message-ID: <47efd8f798968_d5013de3a6164@KEORE.tmail>

The build failed.

  Note: if you set Configuration.dashboard_url in config/site_config.rb, you'd see a link to the build page here.

CHANGES
-------
New revision 997 detected
Revision 997 committed by luislavena on 2008-03-30 18:11:47
Remove deprecated WIN32 and use Gem::Platform::CURRENT instead.

  M /trunk
  M /trunk/Rakefile



BUILD LOG
---------
D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"  
C:/Ruby/bin/ruby extconf.rb
(in D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work)
[CruiseControl] Invoking Rake task "clobber"
Cleaning
- ext/http11/http11-i386-mswin32.def
- ext/http11/http11-i386-mswin32.exp
- ext/http11/http11-i386-mswin32.lib
- ext/http11/http11-i386-mswin32.pdb
- ext/http11/http11.obj
- ext/http11/http11_parser.obj
- ext/http11/Makefile
- ext/http11/vc60.pdb
- lib/http11.so
[CruiseControl] Invoking Rake task "compile"
checking for main() in c.lib... no
creating Makefile

Microsoft (R) Program Maintenance Utility   Version 6.00.9782.0
Copyright (C) Microsoft Corp 1988-1998. All rights reserved.

	C:\Ruby\bin\ruby -e "puts 'EXPORTS', 'Init_http11'"  > http11-i386-mswin32.def
	cl -nologo -I. -I. -IC:/Ruby/lib/ruby/1.8/i386-mswin32 -I. -MD -Zi -O2b2xg- -G6  -c -Tchttp11.c
http11.c
	cl -nologo -I. -I. -IC:/Ruby/lib/ruby/1.8/i386-mswin32 -I. -MD -Zi -O2b2xg- -G6  -c -Tchttp11_parser.c
http11_parser.c
	cl -nologo -LD -Fehttp11.so http11.obj http11_parser.obj msvcrt-ruby18.lib  oldnames.lib user32.lib advapi32.lib ws2_32.lib  -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"." -libpath:"C:/Ruby/lib" -def:http11-i386-mswin32.def -implib:http11-i386-mswin32.lib -pdb:http11-i386-mswin32.pdb
   Creating library http11-i386-mswin32.lib and object http11-i386-mswin32.exp
C:/Ruby/bin/ruby -Ilib;ext;bin;test "C:/Ruby/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb" "test/test_cgi_wrapper.rb" "test/test_command.rb" "test/test_conditional.rb" "test/test_configurator.rb" "test/test_debug.rb" "test/test_handlers.rb" "test/test_http11.rb" "test/test_redirect_handler.rb" "test/test_request_progress.rb" "test/test_response.rb" "test/test_stats.rb" "test/test_uriclassifier.rb" "test/test_ws.rb" 
[CruiseControl] Invoking Rake task "test"
Loaded suite C:/Ruby/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader
Started
...............F.............................................
Finished in 30.797 seconds.

  1) Failure:
test_base_handler_config(ConfiguratorTest) [./test/test_configurator.rb:82]:
<[Errno::EBADF, Errno::ECONNREFUSED]> exception expected but was
Class: <Errno::ECONNRESET>
Message: <"An existing connection was forcibly closed by the remote host.">
---Backtrace---
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/timeout.rb:56:in `timeout'
C:/Ruby/lib/ruby/1.8/timeout.rb:76:in `timeout'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
C:/Ruby/lib/ruby/1.8/net/http.rb:2029:in `read_status_line'
C:/Ruby/lib/ruby/1.8/net/http.rb:2018:in `read_new'
C:/Ruby/lib/ruby/1.8/net/http.rb:1059:in `request'
C:/Ruby/lib/ruby/1.8/net/http.rb:957:in `request_get'
C:/Ruby/lib/ruby/1.8/net/http.rb:380:in `get_response'
C:/Ruby/lib/ruby/1.8/net/http.rb:547:in `start'
C:/Ruby/lib/ruby/1.8/net/http.rb:379:in `get_response'
C:/Ruby/lib/ruby/1.8/net/http.rb:356:in `get'
./test/test_configurator.rb:83:in `test_base_handler_config'
./test/test_configurator.rb:82:in `test_base_handler_config'
---------------

61 tests, 541 assertions, 1 failures, 0 errors
rake aborted!
Command failed with status (1): [C:/Ruby/bin/ruby -Ilib;ext;bin;test "C:/Ru...]

(See full trace by running task with --trace)

dir : D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work
command : ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"
executed command : echo D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:/Users/Luis/.cruise/projects/mongrel-trunk-mswin32/build-997/build.log && ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:\Users\Luis\.cruise\projects\mongrel-trunk-mswin32\build-997\build.log 2>&1
exitstatus: 1
STDERR TAIL START

STDERR TAIL END


PROJECT SETTINGS
----------------
# Project-specific configuration for CruiseControl.rb

Project.configure do |project|

  project.email_notifier.emails = ['mongrel-development at rubyforge.org', 'rubyinstaller-devel at rubyforge.org']

  project.rake_task = 'clobber compile test'
  
  project.scheduler.polling_interval = 5.minutes

end


From builder at mmediasys.com  Sun Mar 30 22:51:09 2008
From: builder at mmediasys.com (Integration Builder)
Date: Sun, 30 Mar 2008 23:51:09 -0300
Subject: [Mongrel-development] [CruiseControl] mongrel-trunk-mingw32 build
	997.1 failed
Message-ID: <47f0519d53bd8_de01462a8819a@edicion01.tmail>

The build failed.

  Note: if you set Configuration.dashboard_url in config/site_config.rb, you'd see a link to the build page here.

CHANGES
-------
Build was manually requested


BUILD LOG
---------
D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"  
rake clean
(in D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work)
[CruiseControl] Invoking Rake task "clobber"
Cleaning
rake aborted!
undefined method `exitstatus' for nil:NilClass
D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work/rakefile:127:in `sub_project'
(See full trace by running task with --trace)

dir : D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work
command : ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"
executed command : echo D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/build-997.1/build.log && ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:\Users\Luis\.cruise\projects\mongrel-trunk-mingw32\build-997.1\build.log 2>&1
exitstatus: 1
STDERR TAIL START

STDERR TAIL END


PROJECT SETTINGS
----------------
# Project-specific configuration for CruiseControl.rb

Project.configure do |project|

  project.email_notifier.emails = ['mongrel-development at rubyforge.org', 'rubyinstaller-devel at rubyforge.org']

  project.rake_task = 'clobber compile test'
  
  project.scheduler.polling_interval = 5.minutes

end


From luislavena at gmail.com  Sun Mar 30 23:57:06 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Mon, 31 Mar 2008 00:57:06 -0300
Subject: [Mongrel-development] Mongrel and Windows: I was so dumb
Message-ID: <71166b3b0803302057k7ade0b07r71cf8929c4fe1bcf@mail.gmail.com>

Hello guys,

I seems that I messed with my environment a bit, and that's why some
of the test failed.

The truth is that I switched from desktop to notebook a few weeks
back, and most of the development was done on desktop until two weeks
back.

Since my notebook is loaded with NOD32 v3, versus my desktop with
version 2.7, it seems they have a "web filter" that was blocking
ruby.exe bad HTTP requests.

Yeah, SO DUMB!

Sorry I made you guys waste time checking for this...

It wouldn't happen again :-D

Also, I'm working on a series of patches that take care of mingw as
platform. I'm just tired of RUBY_PLATFORM regex. Can we add
Mongrel.windows? and Mongrel.java?

Regards guys.

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Mon Mar 31 00:04:51 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Mon, 31 Mar 2008 00:04:51 -0400
Subject: [Mongrel-development] Mongrel and Windows: I was so dumb
In-Reply-To: <71166b3b0803302057k7ade0b07r71cf8929c4fe1bcf@mail.gmail.com>
References: <71166b3b0803302057k7ade0b07r71cf8929c4fe1bcf@mail.gmail.com>
Message-ID: <b6f68fc60803302104k54416616tee31f8770af5543c@mail.gmail.com>

Wow, that is bizarre.

The platform checks are fine. Do they belong in Mongrel, or do they
belong in the build chain?

Evan

On Sun, Mar 30, 2008 at 11:57 PM, Luis Lavena <luislavena at gmail.com> wrote:
> Hello guys,
>
>  I seems that I messed with my environment a bit, and that's why some
>  of the test failed.
>
>  The truth is that I switched from desktop to notebook a few weeks
>  back, and most of the development was done on desktop until two weeks
>  back.
>
>  Since my notebook is loaded with NOD32 v3, versus my desktop with
>  version 2.7, it seems they have a "web filter" that was blocking
>  ruby.exe bad HTTP requests.
>
>  Yeah, SO DUMB!
>
>  Sorry I made you guys waste time checking for this...
>
>  It wouldn't happen again :-D
>
>  Also, I'm working on a series of patches that take care of mingw as
>  platform. I'm just tired of RUBY_PLATFORM regex. Can we add
>  Mongrel.windows? and Mongrel.java?
>
>  Regards guys.
>
>  --
>  Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Mon Mar 31 00:25:34 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Mon, 31 Mar 2008 01:25:34 -0300
Subject: [Mongrel-development] Mongrel and Windows: I was so dumb
In-Reply-To: <b6f68fc60803302104k54416616tee31f8770af5543c@mail.gmail.com>
References: <71166b3b0803302057k7ade0b07r71cf8929c4fe1bcf@mail.gmail.com>
	<b6f68fc60803302104k54416616tee31f8770af5543c@mail.gmail.com>
Message-ID: <71166b3b0803302125m1db26236m25ba80c317fb0c3b@mail.gmail.com>

On Mon, Mar 31, 2008 at 1:04 AM, Evan Weaver <evan at cloudbur.st> wrote:
> Wow, that is bizarre.
>
>  The platform checks are fine. Do they belong in Mongrel, or do they
>  belong in the build chain?
>

Specific platform usage inside mongrel:

- To writer or not the pid file:

http://mongrel.rubyforge.org/browser/trunk/lib/mongrel/configurator.rb#L83

(which is pointless, it should generate the pid file on every
platform, even Windows).

- Allow or not daemonize:

http://mongrel.rubyforge.org/browser/trunk/lib/mongrel/configurator.rb#L189

Which is fair, it should remain there

- Setup signals:

http://mongrel.rubyforge.org/browser/trunk/lib/mongrel/configurator.rb#L364

Which is also good.

Other places:

http://mongrel.rubyforge.org/browser/trunk/bin/mongrel_rails#L139

the usage of restart on Windows, seems logical, since there are no
signals to handle it.

Oh, a good catch:

http://mongrel.rubyforge.org/browser/trunk/bin/mongrel_rails#L85

Mongrel should validate pre-existing pids always, not only in
daemonize mode. Thoughts?

To use or not fastthread and cgi eof fix.

http://mongrel.rubyforge.org/browser/trunk/lib/mongrel.rb#L13

Of course, a few more check on the new Logger (buffered logger) and
cgi_multipart eof subproject. Fastthread, GemPlugin and even
mongrel_service (yeah, I plain suck at planing on the long run).

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From evan at cloudbur.st  Mon Mar 31 00:27:25 2008
From: evan at cloudbur.st (Evan Weaver)
Date: Mon, 31 Mar 2008 00:27:25 -0400
Subject: [Mongrel-development] Mongrel and Windows: I was so dumb
In-Reply-To: <71166b3b0803302125m1db26236m25ba80c317fb0c3b@mail.gmail.com>
References: <71166b3b0803302057k7ade0b07r71cf8929c4fe1bcf@mail.gmail.com>
	<b6f68fc60803302104k54416616tee31f8770af5543c@mail.gmail.com>
	<71166b3b0803302125m1db26236m25ba80c317fb0c3b@mail.gmail.com>
Message-ID: <b6f68fc60803302127i5af6b85cm48725c273411a785@mail.gmail.com>

All sounds good.

Fastthread and cgi_multipart are going away so there is no sense
worrying about them. In 1.2 they get required if they are installed,
but they are not gem dependencies for any platform.

Evan

On Mon, Mar 31, 2008 at 12:25 AM, Luis Lavena <luislavena at gmail.com> wrote:
> On Mon, Mar 31, 2008 at 1:04 AM, Evan Weaver <evan at cloudbur.st> wrote:
>  > Wow, that is bizarre.
>  >
>  >  The platform checks are fine. Do they belong in Mongrel, or do they
>  >  belong in the build chain?
>  >
>
>  Specific platform usage inside mongrel:
>
>  - To writer or not the pid file:
>
>  http://mongrel.rubyforge.org/browser/trunk/lib/mongrel/configurator.rb#L83
>
>  (which is pointless, it should generate the pid file on every
>  platform, even Windows).
>
>  - Allow or not daemonize:
>
>  http://mongrel.rubyforge.org/browser/trunk/lib/mongrel/configurator.rb#L189
>
>  Which is fair, it should remain there
>
>  - Setup signals:
>
>  http://mongrel.rubyforge.org/browser/trunk/lib/mongrel/configurator.rb#L364
>
>  Which is also good.
>
>  Other places:
>
>  http://mongrel.rubyforge.org/browser/trunk/bin/mongrel_rails#L139
>
>  the usage of restart on Windows, seems logical, since there are no
>  signals to handle it.
>
>  Oh, a good catch:
>
>  http://mongrel.rubyforge.org/browser/trunk/bin/mongrel_rails#L85
>
>  Mongrel should validate pre-existing pids always, not only in
>  daemonize mode. Thoughts?
>
>  To use or not fastthread and cgi eof fix.
>
>  http://mongrel.rubyforge.org/browser/trunk/lib/mongrel.rb#L13
>
>  Of course, a few more check on the new Logger (buffered logger) and
>  cgi_multipart eof subproject. Fastthread, GemPlugin and even
>  mongrel_service (yeah, I plain suck at planing on the long run).
>
>  --
>
>
> Luis Lavena
>  Multimedia systems
>  -
>  Human beings, who are almost unique in having the ability to learn from
>  the experience of others, are also remarkable for their apparent
>  disinclination to do so.
>  Douglas Adams
>  _______________________________________________
>  Mongrel-development mailing list
>  Mongrel-development at rubyforge.org
>  http://rubyforge.org/mailman/listinfo/mongrel-development
>



-- 
Evan Weaver
Cloudburst, LLC

From luislavena at gmail.com  Mon Mar 31 00:29:07 2008
From: luislavena at gmail.com (Luis Lavena)
Date: Mon, 31 Mar 2008 01:29:07 -0300
Subject: [Mongrel-development] Mongrel and Windows: I was so dumb
In-Reply-To: <b6f68fc60803302104k54416616tee31f8770af5543c@mail.gmail.com>
References: <71166b3b0803302057k7ade0b07r71cf8929c4fe1bcf@mail.gmail.com>
	<b6f68fc60803302104k54416616tee31f8770af5543c@mail.gmail.com>
Message-ID: <71166b3b0803302129r2afc0a4ctadd9359bade2bdad@mail.gmail.com>

On Mon, Mar 31, 2008 at 1:04 AM, Evan Weaver <evan at cloudbur.st> wrote:
> Wow, that is bizarre.
>

yeah, also if two builders run at the same time, they have a port
collition. I'm adding process_based_port to testhelp.rb to provide a
port, based on 5000 and the process number:


irb(main):001:0> 5000 + $$ % 1000
=> 5656

That will make my CI happy :-)

(I'm doing the changes right now for rubygems too, found a clash due
that during setup).

Oh, and forgot to make it generate the artifacts, so we can grab the
pre-compiled gem from there ;-)

Regards,

-- 
Luis Lavena
Multimedia systems
-
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so.
Douglas Adams

From builder at mmediasys.com  Mon Mar 31 01:30:22 2008
From: builder at mmediasys.com (Integration Builder)
Date: Mon, 31 Mar 2008 02:30:22 -0300
Subject: [Mongrel-development] [CruiseControl] mongrel-trunk-mingw32 build
	998 failed
Message-ID: <47f076ee66c88_de01462a88286@edicion01.tmail>

The build failed.

  Note: if you set Configuration.dashboard_url in config/site_config.rb, you'd see a link to the build page here.

CHANGES
-------
New revision 998 detected
Revision 998 committed by luislavena on 2008-03-31 05:25:39
Remove fixed port numbers used in tests, make tests more friendly to
CI tools.
Use of #process_based_port as port number.

  M /trunk/test/test_redirect_handler.rb
  M /trunk/test/test_handlers.rb
  M /trunk
  M /trunk/test/test_configurator.rb
  M /trunk/test/test_conditional.rb
  M /trunk/test/test_ws.rb
  M /trunk/test/testhelp.rb
  M /trunk/test/test_request_progress.rb



BUILD LOG
---------
D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"  
rake clean
(in D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work)
[CruiseControl] Invoking Rake task "clobber"
Cleaning
rake aborted!
undefined method `exitstatus' for nil:NilClass
D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work/rakefile:127:in `sub_project'
(See full trace by running task with --trace)

dir : D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work
command : ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"
executed command : echo D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:/Users/Luis/.cruise/projects/mongrel-trunk-mingw32/build-998/build.log && ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:\Users\Luis\.cruise\projects\mongrel-trunk-mingw32\build-998\build.log 2>&1
exitstatus: 1
STDERR TAIL START

STDERR TAIL END


PROJECT SETTINGS
----------------
# Project-specific configuration for CruiseControl.rb

Project.configure do |project|

  project.email_notifier.emails = ['mongrel-development at rubyforge.org', 'rubyinstaller-devel at rubyforge.org']

  project.rake_task = 'clobber compile test'
  
  project.scheduler.polling_interval = 5.minutes

end


From builder at mmediasys.com  Mon Mar 31 04:30:03 2008
From: builder at mmediasys.com (Integration Builder)
Date: Mon, 31 Mar 2008 05:30:03 -0300
Subject: [Mongrel-development] [CruiseControl] mongrel-stable_1-2-mswin32
	build 1001 failed
Message-ID: <47f0a10bbeac8_ee013de3a612@edicion01.tmail>

The build failed.

  Note: if you set Configuration.dashboard_url in config/site_config.rb, you'd see a link to the build page here.

CHANGES
-------
This is the first build


BUILD LOG
---------
D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mswin32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"  
D:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/bin/ruby extconf.rb
(in D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mswin32/work)
[CruiseControl] Invoking Rake task "clobber"
Cleaning
[CruiseControl] Invoking Rake task "compile"
checking for main() in c.lib... no
creating Makefile

Microsoft (R) Program Maintenance Utility   Version 6.00.9782.0
Copyright (C) Microsoft Corp 1988-1998. All rights reserved.

	D:\Users\Luis\ruby\ruby-186-p114-VC6-rubygems-110\bin\ruby -e "puts 'EXPORTS', 'Init_http11'"  > http11-i386-mswin32.def
	cl -nologo -I. -I. -ID:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/lib/ruby/1.8/i386-mswin32 -I. -MD -Zi -O2b2xg- -G6  -c -Tchttp11.c
http11.c
	cl -nologo -I. -I. -ID:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/lib/ruby/1.8/i386-mswin32 -I. -MD -Zi -O2b2xg- -G6  -c -Tchttp11_parser.c
http11_parser.c
	cl -nologo -LD -Fehttp11.so http11.obj http11_parser.obj msvcrt-ruby18.lib  oldnames.lib user32.lib advapi32.lib ws2_32.lib  -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"." -libpath:"D:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/lib" -def:http11-i386-mswin32.def -implib:http11-i386-mswin32.lib -pdb:http11-i386-mswin32.pdb
   Creating library http11-i386-mswin32.lib and object http11-i386-mswin32.exp
D:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/bin/ruby -Ilib;ext;bin;test "D:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb" "test/test_helper.rb" "test/unit/test_cgi_wrapper.rb" "test/unit/test_command.rb" "test/unit/test_conditional.rb" "test/unit/test_configurator.rb" "test/unit/test_debug.rb" "test/unit/test_handlers.rb" "test/unit/test_http_parser.rb" "test/unit/test_redirect_handler.rb" "test/unit/test_request_progress.rb" "test/unit/test_response.rb" "test/unit/test_stats.rb" "test/unit/test_uriclassifier.rb" "test/unit/test_ws.rb" 
[CruiseControl] Invoking Rake task "test"
./test/test_helper.rb:8: warning: already initialized constant HERE
Loaded suite D:/Users/Luis/ruby/ruby-186-p114-VC6-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader
Started
.......................F..E................................
Finished in 32.297 seconds.

  1) Failure:
test_horrible_queries(HttpParserTest)
    [./test/unit/test_http_parser.rb:105:in `test_horrible_queries'
     ./test/unit/test_http_parser.rb:103:in `times'
     ./test/unit/test_http_parser.rb:103:in `test_horrible_queries']:
<Mongrel::HttpParserError> exception expected but none was thrown.

  2) Error:
test_parse_strange_headers(HttpParserTest):
Mongrel::HttpParserError: Invalid HTTP format, parsing fails.
    ./test/unit/test_http_parser.rb:49:in `execute'
    ./test/unit/test_http_parser.rb:49:in `test_parse_strange_headers'

59 tests, 458 assertions, 1 failures, 1 errors
rake aborted!
Command failed with status (1): [D:/Users/Luis/ruby/ruby-186-p114-VC6-rubyg...]

(See full trace by running task with --trace)

dir : D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mswin32/work
command : ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"
executed command : echo D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mswin32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mswin32/build-1001/build.log && ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:\Users\Luis\.cruise\projects\mongrel-stable_1-2-mswin32\build-1001\build.log 2>&1
exitstatus: 1
STDERR TAIL START

STDERR TAIL END


PROJECT SETTINGS
----------------
# Project-specific configuration for CruiseControl.rb

Project.configure do |project|

  project.email_notifier.emails = ['mongrel-development at rubyforge.org', 'rubyinstaller-devel at rubyforge.org']

  project.rake_task = 'clobber compile test package'
  
  project.scheduler.polling_interval = 5.minutes

end


From builder at mmediasys.com  Mon Mar 31 04:30:26 2008
From: builder at mmediasys.com (Integration Builder)
Date: Mon, 31 Mar 2008 05:30:26 -0300
Subject: [Mongrel-development] [CruiseControl] mongrel-stable_1-2-mingw32
	build 1001 failed
Message-ID: <47f0a122222e0_bd41462a8815d@edicion01.tmail>

The build failed.

  Note: if you set Configuration.dashboard_url in config/site_config.rb, you'd see a link to the build page here.

CHANGES
-------
This is the first build


BUILD LOG
---------
D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mingw32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"  
D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/bin/ruby extconf.rb
(in D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mingw32/work)
[CruiseControl] Invoking Rake task "clobber"
Cleaning
[CruiseControl] Invoking Rake task "compile"
checking for main() in -lc... no
creating Makefile
gcc -I. -I. -ID:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/1.8/i386-mingw32 -I.  -g -O2   -c http11.c
gcc -I. -I. -ID:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/1.8/i386-mingw32 -I.  -g -O2   -c http11_parser.c
gcc -shared -s -o http11.so http11.o http11_parser.o -L. -LD:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib -L.  -Wl,--enable-auto-image-base,--enable-auto-import,--export-all   -lmsvcrt-ruby18  -lshell32 -lws2_32  
D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/bin/ruby -Ilib;ext;bin;test "D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb" "test/test_helper.rb" "test/unit/test_cgi_wrapper.rb" "test/unit/test_command.rb" "test/unit/test_conditional.rb" "test/unit/test_configurator.rb" "test/unit/test_debug.rb" "test/unit/test_handlers.rb" "test/unit/test_http_parser.rb" "test/unit/test_redirect_handler.rb" "test/unit/test_request_progress.rb" "test/unit/test_response.rb" "test/unit/test_stats.rb" "test/unit/test_uriclassifier.rb" "test/unit/test_ws.rb" 
[CruiseControl] Invoking Rake task "test"
./test/test_helper.rb:8: warning: already initialized constant HERE
./test/unit/test_command.rb:9: uninitialized constant GemPlugin (NameError)
	from D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb:5:in `load'
	from D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb:5
	from D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb:5:in `each'
	from D:/Users/Luis/ruby/ruby-186-svn-MINGW-rubygems-110/lib/ruby/gems/1.8/gems/rake-0.8.1/lib/rake/rake_test_loader.rb:5
rake aborted!
Command failed with status (1): [D:/Users/Luis/ruby/ruby-186-svn-MINGW-ruby...]

(See full trace by running task with --trace)

dir : D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mingw32/work
command : ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear"
executed command : echo D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mingw32/work Luis$ ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:/Users/Luis/.cruise/projects/mongrel-stable_1-2-mingw32/build-1001/build.log && ruby -e "require 'rubygems' rescue nil; require 'rake'; load 'D:/Users/Luis/tools/ccrb/tasks/cc_build.rake'; ARGV << '--nosearch' << 'cc:build'; Rake.application.run; ARGV.clear" >> D:\Users\Luis\.cruise\projects\mongrel-stable_1-2-mingw32\build-1001\build.log 2>&1
exitstatus: 1
STDERR TAIL START

STDERR TAIL END


PROJECT SETTINGS
----------------
# Project-specific configuration for CruiseControl.rb

Project.configure do |project|

  project.email_notifier.emails = ['mongrel-development at rubyforge.org', 'rubyinstaller-devel at rubyforge.org']

  project.rake_task = 'clobber compile test package'
  
  project.scheduler.polling_interval = 5.minutes

end