Re: Possible race condition in usb-serial.c

[J]

> This problem will need some deeper surgery probably
> involving
> removal of the refcounting.

Refcounting may be OK if used consistently. 
It is not OK when some pointers are ref-counted, 
but other (in serial_table) are not (like it is
in the current version).

As for the deeper surgery, what do you think about my
earlier suggestion to start by rewriting
usb_serial_probe
to fully initialize usb_serial before it is added to 
serial_table? 

So, instead of the current:
1. create_serial
...
2. mutex_lock(&table_lock);
3. get_free_serial (which inserts serial to
serial_table)
...
4. initializes serial
5. mutex_unlock(&table_lock);

we will get:

1. create_serial
2. initializes serial

3. add_serial_toserial_table  (with internal mutex
lock if needed)

Similar approach should be used in other places to
minimize the code executed under the mutex.

John

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/