Please don't apply this patch, which will cause compilation failure on
IA-64. I'll re-sent it shortly.

CAI Qian

From: CAI Qian <[email protected]>
Subject: [LTP] [PATCH] Crash02: Deal with SIGSTOP and SIGSEGV
Date: Tue, 19 May 2009 20:14:43 +0800 (CST)

> Hi,
> 
> We have seen crash02 test can not finish due to the child process got
> SIGSTOP or SIGSEGV when calling random syscalls like this,
> 
> # ps aux
> ...
> 168      20121  0.0  0.0   3956   336 ?        Ds   15:10   0:00 ./crash02 -e 
> -v 100
> 168      20122  0.0  0.0   3956   336 ?        T    15:10   0:00 ./crash02 -e 
> -v 100
> 
> 
> crash02       X ffff81007fb127a0     0 21272  21271         21273       
> (L-TLB)
>  ffff81004c185d58 0000000000000046 ffff81007b494e08 0000000000000002
>  0000000000000002 0000000000000002 ffff81007c5dc860 ffff81007fb127a0
>  00000d8ebb800862 000000000000c315 ffff81007c5dca48 0000000300000001
> Call Trace:
>  [<ffffffff800235a7>] filp_close+0x5c/0x64
>  [<ffffffff8001595a>] do_exit+0x913/0x91f
>  [<ffffffff80048c18>] cpuset_exit+0x0/0x6c
>  [<ffffffff8002ad05>] get_signal_to_deliver+0x42c/0x45a
>  [<ffffffff8005a837>] do_notify_resume+0x9c/0x7af
>  [<ffffffff80096e2b>] specific_send_sig_info+0x44/0xac
>  [<ffffffff8009710e>] force_sig_info+0xae/0xb9
>  [<ffffffff80066eed>] do_page_fault+0x81e/0x830
>  [<ffffffff8005d6dc>] retint_signal+0x3d/0x79
> 
> crash02       X ffff810002376400     0 21273  21271               21272 
> (L-TLB)
>  ffff81004d201f18 0000000000000046 ffff81007b494e08 0000000000000046
>  0000000000000046 0000000000000002 ffff810078d3a040 ffffffff802eeae0
>  00000d8ebb803430 000000000001a60c ffff810078d3a228 0000000000000000
> Call Trace:
>  [<ffffffff8001595a>] do_exit+0x913/0x91f
>  [<ffffffff80048c18>] cpuset_exit+0x0/0x6c
>  [<ffffffff8005d28d>] tracesys+0xd5/0xe0
> 
> So we use WUNTRACED option. Also, we found out that it picks up vfork(),
> it will generate SIGSEGV. I am not sure what else can generate the same,
> so only blacklist vfork() at the moment.
> 
> Signed-off-by: CAI Qian <[email protected]>
> 
> --- testcases/misc/crash/crash02.c.orig       2009-05-19 19:47:24.000000000 
> +0800
> +++ testcases/misc/crash/crash02.c    2009-05-19 20:03:51.000000000 +0800
> @@ -51,6 +51,8 @@
>  */
>  
>  
> +#define _GNU_SOURCE
> +#include <sys/syscall.h>
>  #include <stdio.h>
>  #include <stdlib.h>
>  #include <string.h>
> @@ -241,7 +243,7 @@
>        my_signal (SIGALRM, monitor_fcn);
>        alarm (ntries*MAX_TRY_TIME);
>  
> -       pid = wait (&status);
> +       pid = waitpid (-1, &status, WUNTRACED);
>         if (pid <= 0)
>         {
>                 perror ("wait");
> @@ -453,7 +455,10 @@
>  {
>    long int sysno, arg1, arg2, arg3, arg4, arg5, arg6, arg7;
>  
> -  sysno = rand()%sysno_max;
> +  do {
> +    sysno = rand()%sysno_max;
> +  } while (sysno == SYS_vfork);
> +
>    arg1 = rand_long();
>    arg2 = rand_long();
>    arg3 = rand_long();
> 
> ------------------------------------------------------------------------------
> Crystal Reports - New Free Runtime and 30 Day Trial
> Check out the new simplified licensing option that enables 
> unlimited royalty-free distribution of the report engine 
> for externally facing server and web deployment. 
> http://p.sf.net/sfu/businessobjects
> _______________________________________________
> Ltp-list mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/ltp-list

------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, & 
iPhoneDevCamp asthey present alongside digital heavyweights like Barbarian
Group, R/GA, & Big Spaceship. http://www.creativitycat.com 
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to