Re: [squid-users] Squid 3.0.STABLE17 is available

Thomas Meier wrote:
> im also have this error:
>
> 2009/07/30 09:48:16| HttpMsg.cc(157) first line of HTTP message is invalid
> 2009/07/30 09:48:16| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:48:19| Starting Squid Cache version 3.0.STABLE17 for
> sparc-sun-sol
> aris2.9...
> 2009/07/30 09:48:19| Process ID 17345
> 2009/07/30 09:48:19| With 32768 file descriptors available
> 2009/07/30 09:48:19| Performing DNS Tests...
> 2009/07/30 09:48:19| Successful DNS name lookup tests...
>
>
> ####
>
> 2009/07/30 09:47:38| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:48:16| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:48:46| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:48:58| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:49:44| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:49:53| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:50:02| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:51:33| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:51:39| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:51:45| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:52:26| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:53:02| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:53:12| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:53:22| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:53:30| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:54:32| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:55:35| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:57:04| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:57:41| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:57:57| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:58:22| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:59:31| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:59:44| assertion failed: http.cc:738: "!eof"
> 2009/07/30 09:59:59| assertion failed: http.cc:738: "!eof"
> 2009/07/30 10:01:13| assertion failed: http.cc:738: "!eof"
> 2009/07/30 10:01:48| assertion failed: http.cc:738: "!eof"
> 2009/07/30 10:02:29| assertion failed: http.cc:738: "!eof"
> 2009/07/30 10:02:34| assertion failed: http.cc:738: "!eof"
> 2009/07/30 10:02:41| assertion failed: http.cc:738: "!eof"
> 2009/07/30 10:03:51| assertion failed: http.cc:738: "!eof"
>
> back to stable 13 the good old unfixed Problem is back:
>
> 2009/07/30 11:05:05| tunnelReadServer: FD 377: read failure: (0) Error 0
> 2009/07/30 11:05:16| tunnelReadServer: FD 211: read failure: (0) Error 0
> 2009/07/30 11:05:20| tunnelReadServer: FD 462: read failure: (0) Error 0
> 2009/07/30 11:05:32| tunnelReadServer: FD 437: read failure: (0) Error 0
> 2009/07/30 11:05:36| tunnelReadServer: FD 340: read failure: (0) Error 0
> 2009/07/30 11:05:41| tunnelReadServer: FD 432: read failure: (0) Error 0
> 2009/07/30 11:06:13| assertion failed: store_client.cc:430:
> "STORE_DISK_CLIENT =
> = getType()"
> 2009/07/30 11:06:18| Starting Squid Cache version 3.0.STABLE13 for
> sparc-sun-sol
> aris2.9...
>
>
>
>> > On Tuesday 28 July 2009 23:22:56 Amos Jeffries wrote:
>>
>>> >> The next formally bundled will be STABLE18. However the daily
>>> snapshots
>>> >> serve as intermediate updates on STABLE
>>> >> (http://www.squid-cache.org/Versions/v3/3.0/).
>>> >>
>>> >> I just have not yet had time to apply these fixes to the branch yet.
>>>
>> > > > 3.0.STABLE17-20090729 still crashing here (x86_64)...
>> > > 2009/07/29 16:07:45| ctx: enter level 0: >
>> 'http://images.windowsmedia.com/svcswitch/MG_pt-
>> > br.xml?locale=416&geoid=20&version=1
>> >
>> 1.0.6001.7004&userlocale=416'
>>
>
>> > > 2009/07/29 16:07:45| assertion failed: http.cc:738: "!eof" > > I
>> also applied (the patch from previous e-mail) against this version:
>> > > patching file src/HttpMsg.cc
>> > patching file src/HttpReply.cc
>> > patching file src/HttpRequest.cc
>> > patching file src/pconn.cc
>> > > The only solution here was downgrade to previous release...
>> > > Any clue?
>>
>
> It seems to be a fread() issue between Squid and x86_64. Disappears on
> i386/i686.
>
> Maybe *BSD specific as well, but there are a few reports without OS info
> fuzzing my info there.
> I plan on spending more time over the weekend adding good tracing info to
> find it.
>

Okay. This gets rid of the assert and adds some debug instead.
The reason for sending eof=1 when not at true EOF is not yet clear, so
use carefully, but additional debugs are added when the flag is set.
debug_options ... 11,9 for these.

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE6 or 3.0.STABLE17
   Current Beta Squid 3.1.0.12