curl-loader-devel

curl-loader 0.43 Segmentation fault

[Iliyan P. <ili...@gm...>]

Hello,

I am using curl-loader-0.43
First, I did one modification in fgets buffer, because its default value=
  =

(1024) was too short and curl-loader failed to parse my config file:
file: parse_conf.c
function: parse_config_file
row: 2638
change:
-char fgets_buff[1024];
+char fgets_buff[2048];

After this change, curl-loader started successfully.

I tested my web server several times successfully. Unfortunately,  =

curl-loader crashed with segmentation fault when I started it with "-d" =
 =

option:
# curl-loader -d -f testWGM.cnf

Below is gdb's output. I can provide core file or whatelse you would ask=
.


=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D  loading batch is: Test        =

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
  Operations:             Success                 Failed                =
  =

Timed out
URL0:GetBestMap         0      0                0      0                =
 =

0      0
URL1:GetMapObject       0      0                0      0                =
 =

0      0
URL2:GetMapImage        0      0                0      0                =
 =

0      0
URL3:GetObjectTra       0      0                0      0                =
 =

0      0
URL4:GetObjectToo       0      0                0      0                =
 =

0      0
URL5:GetObjectsSt       0      0                0      0                =
 =

0      0
URL6:GetObjectDat       0      0                0      0                =
 =

0      0
URL7:GetObjectCom       0      0                0      0                =
 =

0      0
------------------------------------------------------------------------=
--------
Interval stats (latest:0 sec, clients:10, CAPS-curr:10000):
H/F    =

Req:10,1xx:0,2xx:0,3xx:0,4xx:0,5xx:0,Err:0,T-Err:0,D:0ms,D-2xx:0ms,Ti:0B=
/s,To:1700B/s
H/F/S  =

Req:0,1xx:0,2xx:0,3xx:0,4xx:0,5xx:0,Err:0,T-Err:0,D:0ms,D-2xx:0ms,Ti:0B/=
s,To:0B/s
------------------------------------------------------------------------=
--------
Summary stats (runs:1 secs, CAPS-average:10):
H/F    =

Req:10,1xx:0,2xx:0,3xx:0,4xx:0,5xx:0,Err:0,T-Err:0,D:0ms,D-2xx:0ms,Ti:0B=
/s,To:1700B/s
H/F/S  =

Req:0,1xx:0,2xx:0,3xx:0,4xx:0,5xx:0,Err:0,T-Err:0,D:0ms,D-2xx:0ms,Ti:0B/=
s,To:0B/s
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
  Automatic: adding 10 clients/sec. Stop inc and manual [M].
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
*** stack smashing detected ***:  =

/install/data/curl-loader/curl-loader-0.43/curl-loader terminated

Program received signal SIGABRT, Aborted.
[Switching to Thread -1212275008 (LWP 1095)]
0xffffe410 in __kernel_vsyscall ()
(gdb) bt
#0  0xffffe410 in __kernel_vsyscall ()
#1  0xb7c0cdf0 in raise () from /lib/tls/i686/cmov/libc.so.6
#2  0xb7c0e641 in abort () from /lib/tls/i686/cmov/libc.so.6
#3  0xb7c429bb in ?? () from /lib/tls/i686/cmov/libc.so.6
#4  0x080992ac in ?? ()
#5  0x00000000 in ?? ()
(gdb)

Regards,
Iliyan Peychev

Re: curl-loader 0.43 Segmentation fault

[Robert I. <cor...@gm...>]

Iliyan,

On 10/12/07, Iliyan Peychev <ili...@gm...> wrote:
> First, I did one modification in fgets buffer, because its default value
> (1024) was too short and curl-loader failed to parse my config file:
> file: parse_conf.c
> function: parse_config_file
> row: 2638
> change:
> -char fgets_buff[1024];
> +char fgets_buff[2048];

I will increase it to some 4K.
Thank you very much for your pointer.

> After this change, curl-loader started successfully.
> I tested my web server several times successfully.

Good.

> Unfortunately, curl-loader crashed with segmentation fault when I started it with "-d"
> option:
> # curl-loader -d -f testWGM.cnf

-d is a freshly added option, which is not well-debugged.

If logging of only server responses is enough, you might try adding to
you batch config file the relevant logging tags.

Please, kindly, fill the PROBLEM-REPORTING form in the tarball
with more data.

> Below is gdb's output. I can provide core file or whatelse you would ask.
The trace you sent was not informative.
Could you try to run it just from gdb or any other debugger?
Thank you.

Sincerely,
Robert Iakobashvili,
coroberti %x40 gmail %x2e com
...........................................................
http://curl-loader.sourceforge.net
A web testing and traffic generation tool.

Re: curl-loader 0.43 Segmentation fault

[Iliyan P. <ili...@gm...>]

Robert,

Here you are filled problem report file. I also recompiled curl-loader  =

with the following options:
make optimize=3D0 debug=3D1

Regards,
Iliyan Peychev

On Fri, 12 Oct 2007 11:35:54 +0300, Robert Iakobashvili  =

<cor...@gm...> wrote:

> Iliyan,
>
> On 10/12/07, Iliyan Peychev <ili...@gm...> wrote:
>> First, I did one modification in fgets buffer, because its default va=
lue
>> (1024) was too short and curl-loader failed to parse my config file:
>> file: parse_conf.c
>> function: parse_config_file
>> row: 2638
>> change:
>> -char fgets_buff[1024];
>> +char fgets_buff[2048];
>
> I will increase it to some 4K.
> Thank you very much for your pointer.
>
>> After this change, curl-loader started successfully.
>> I tested my web server several times successfully.
>
> Good.
>
>> Unfortunately, curl-loader crashed with segmentation fault when I  =

>> started it with "-d"
>> option:
>> # curl-loader -d -f testWGM.cnf
>
> -d is a freshly added option, which is not well-debugged.
>
> If logging of only server responses is enough, you might try adding to=

> you batch config file the relevant logging tags.
>
> Please, kindly, fill the PROBLEM-REPORTING form in the tarball
> with more data.
>
>> Below is gdb's output. I can provide core file or whatelse you would =
 =

>> ask.
> The trace you sent was not informative.
> Could you try to run it just from gdb or any other debugger?
> Thank you.
>
> Sincerely,
> Robert Iakobashvili,
> coroberti %x40 gmail %x2e com
> ...........................................................
> http://curl-loader.sourceforge.net
> A web testing and traffic generation tool.
>
> ----------------------------------------------------------------------=
---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser=
.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
> _______________________________________________
> curl-loader-devel mailing list
> cur...@li...
> https://lists.sourceforge.net/lists/listinfo/curl-loader-devel

Re: curl-loader 0.43 Segmentation fault

[Robert I. <cor...@gm...>]

Iliyan,

On 10/12/07, Iliyan Peychev <ili...@gm...> wrote:
> Here you are filled problem report file. I also recompiled curl-loader
> make optimize=0 debug=1

Thank you for using PROBLEM-REPORTING form, I will look into it on Sunday.

Although the backtrace it shows some libevent problems, I am suspecting
memory issues in client_tracing_function, loader.c, line 1087.


In meanwhile, please, try to use the tags:

LOG_RESP_HEADERS, when defined as 1, the program logs response headers
to a file. Directory <batch-name> is created with subdirs url0, url1,
url<n> Headers of responses are logged to the files named:
cl-<client-num>-cycle-<cycle-num>.hdr

LOG_RESP_BODIES, when defined as 1, the program logs response headers
to a file. Directory <batch-name> is created with subdirs url0,
url1... url<n> Headers of responses are logged to the files named:
cl-<client-num>-cycle-<cycle-num>.body



-- 
Sincerely,
Robert Iakobashvili,
coroberti %x40 gmail %x2e com
...........................................................
http://curl-loader.sourceforge.net
A web testing and traffic generation tool.

Re: curl-loader 0.43 Segmentation fault

[Robert I. <cor...@gm...>]

Iliyan,

On 10/12/07, Iliyan Peychev <ili...@gm...> wrote:
> Here you are filled problem report file. I also recompiled curl-loader
> with the following options:
> make optimize=0 debug=1

Thank you for the PRF.
Please, take the latest version from svn, which is supposed to
fix the -d option bug.

The size of fgets_buff increased.
Y have been added to our THANKS list. Thank you.

Sincerely,
Robert Iakobashvili,
coroberti %x40 gmail %x2e com
...........................................................
http://curl-loader.sourceforge.net
A web testing and traffic generation tool.