Dear Marcelo,

This patch should fix various problems with the closing_wait and 
close_delay serial parameters, but I can only test the standard serial 
driver.

There are various scaling problems when HZ != 100:

* In several drivers, the values set by TIOCSSERIAL are scaled by 
HZ/100, but the values got by TIOCGSERIAL are not scaled back.

* Invariably, the scaled closing_wait value is compared with an unscaled 
constant ASYNC_CLOSING_WAIT_NONE or the same value by another name 
(depending on the specific driver).

My patch stores the values set by TIOCSSERIAL without scaling and scales 
the values by HZ/100 at the point of use.  (This is not as bad as it 
sounds, as each value is scaled once per 'close'.)  There seems to be a 
general consensus amongst the serial drivers that the closing_wait and 
close_delay values from the user are in units of .01 seconds.

There is another problem, not related to HZ scaling:

* In several drivers, the closing_wait and close_delay values are 
written to a struct serial_state by TIOCSSERIAL, but the values used in 
the close routine are read from a struct async_struct, with no code to 
transfer of values between the two structures.  My patch ignores the 
members in struct async_struct and uses the values from struct serial_state.

-- Ian.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>