"Eric Pouech" eric.pouech@wanadoo.fr wrote:
Freeing the header while unmapping the 16/32-bit message is an alien task for mapping functions IMO.
since the buffer (the 16 bit segmented to be used instead of the 32bit linear) has been allocated in the mapping function, I do think it's a bad idea to free it outside of the unmapping function
It seems to me that it's an unnecessary complexity to introduce extra parameter for unmapping function exclusively for xxxx_UNPREPARE messages. Could you please show what you have in mind for the single mapping/unmapping function pair?