Thread (71 messages) 71 messages, 7 authors, 2011-07-09

Re: [PATCH 05/12] Input: synaptics - process button bits in AGM packets

From: Chase Douglas <hidden>
Date: 2011-07-05 18:20:03
Also in: lkml

On 07/05/2011 11:18 AM, Henrik Rydberg wrote:
quoted
quoted
Any particular reason to move these and leave them unassigned for clickpads?
Yes.  The current implementation incorrectly parses the x, y, z of AGM
packets and assigns junk values to the corresponding fields of the
temporary hw struct.  Luckily, this struct is then just discarded upon
return (synaptics_parse_hw_state returns 1 causing
synaptics_process_packet() to exit immediately).

Instead, this patch parses the w value first to determine the packet
type, and then use this packet type information to parse the remaining
position and pressure fields correctly...

Notice that the "else" clause is taken for SGM packets (w != 2), even
for clickpads.
Functionally, there is no difference between assigning new junk or
reusing old junk, hence that part of the patch is not strictly needed.
True, it's not necessary, but I do like the new flow better. I was
reviewing the patch myself when I started wondering why we ever assigned
junk that was incorrect. I think this will help clean up the code and
make it more readable.

-- Chase
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help