No. GFP_KERNEL.
Let me explain things more clearly.
I made the following change to the kernel (I've simplified the code):
So, to follow your steps, I did the following,
static void myfunc ()
{
if (jiffies > 20 * HZ) {
void *p = kmem_cache_alloc (malloc_sizes->cs_cachep, GFP_KERNEL);
if (p) {
printk ("Allocated\n");
}
kmem_cache_free (malloc_sizes->cs_cachep, p);
}
}
asmlinkage NORET_TYPE void do_exit(long code)
{
struct task_struct *tsk = current;
<snip>
if (unlikely(current->ptrace & PT_TRACE_EXIT)) {
current->ptrace_message = code;
ptrace_notify((PTRACE_EVENT_EXIT << 8) | SIGTRAP);
}
myfunc();
acct_process(code);
__exit_mm(tsk);
exit_sem(tsk);
<snip>
}
This code did not produce any problems. So I guess there is no problem
in calling kmem_cache_alloc from do_exit.
There must be some bug in your code which is manifested this way.
Regards,
Om
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
FAQ: http://kernelnewbies.org/faq/
