dtrace discuss - Dec 2005 - ustack() issues && correlating SIGSEGV activity?

Howdy,

I was playing around with the malloc/free D script provided by 
Philip Beevers:

http://www.opensolaris.org/jive/thread.jspa?threadID=4224&tstart=15

And decided to change the free:entry probe from:

pid24169::free:entry
/ sz[arg0] /
{
     printf("Freeing %p (size %d)\n", arg0, sz[arg0]);
     sz[arg0] = 0;
}

to:

pid24169::free:entry
/ ! sz[arg0] /
{
     printf("[ * ] A bad function is Freeing memory at address
%p\n",arg0);
     ustack();
}

When I created a test program [1] and attempted to run the new script, I 
got the following output:

$  dtrace -q -s foo.d -c ./fluffy
fluff: 20a30
[ * ] A bad function is Freeing memory at address ff3f1298n
               ld.so.1`free
               ld.so.1`setup+0x13ac
               ld.so.1`_setup+0x3b0
               ld.so.1`_rt_boot+0x88

Does anyone happen to know why the ustack() doesn''t show lex, yak, main
and _start? Also -- is there a way to use DTrace to trace access to memory 
and  correlate that back to the instruction and function that accessed it 
(e.g., finding the culprit of a SIGSEGV without gdb)?

Thanks for any insight,
- Ryan
--
UNIX Administrator
http://daemons.net/~matty

[1]

$ cat fluffy.c

void yak(char *fluff)
{
    free(fluff);
    lex(fluff);
}

void lex(char *fluff)
{
     free(fluff);
}

int main(int argc, char**argv)
{
    char *fluff  = malloc(1);
    printf("fluff: %0x\n",fluff);
    yak(fluff);
}

On Wed, Dec 21, 2005 at 12:19:13PM -0500, Matty wrote:
> 
> Howdy,
> 
> I was playing around with the malloc/free D script provided by 
> Philip Beevers:
> 
> http://www.opensolaris.org/jive/thread.jspa?threadID=4224&tstart=15
> 
> And decided to change the free:entry probe from:
> 
> pid24169::free:entry
> / sz[arg0] /
> {
>     printf("Freeing %p (size %d)\n", arg0, sz[arg0]);
>     sz[arg0] = 0;
> }
> 
> to:
> 
> pid24169::free:entry
> / ! sz[arg0] /
> {
>     printf("[ * ] A bad function is Freeing memory at address
%p\n",arg0);
>     ustack();
> }
> 
> When I created a test program [1] and attempted to run the new script, I 
> got the following output:
> 
> $  dtrace -q -s foo.d -c ./fluffy
> fluff: 20a30
> [ * ] A bad function is Freeing memory at address ff3f1298n
>               ld.so.1`free
>               ld.so.1`setup+0x13ac
>               ld.so.1`_setup+0x3b0
>               ld.so.1`_rt_boot+0x88
> 
> Does anyone happen to know why the ustack() doesn''t show lex, yak,
main
> and _start?
Looks like tail-call optimization to me.  Try:

int yak(char *fluff)
{
    free(fluff);
    (void) lex(fluff);
    return (0);
}

int lex(char *fluff)
{
    free(fluff);
    return (0);
}

int main(int argc, char**argv)
{
   char *fluff  = malloc(1);
   printf("fluff: %0x\n",fluff);
   (void) yak(fluff);
   return (0);
}
> Also -- is there a way to use DTrace to trace access to memory 
> and  correlate that back to the instruction and function that accessed it 
> (e.g., finding the culprit of a SIGSEGV without gdb)?
Not at the moment, but a "watchpoint" provider has been bandied about.

Cheers,
- jonathan

-- 
Jonathan Adams, Solaris Kernel Development