Daniel Kurtz wrote:

On Tue, Mar 20, 2012 at 11:38 PM, Nick Dyer [off-list ref] wrote:

quoted

mxt_make_highchg exists due to the interrupts being edge triggered. It
forces the CHG line to go high.

With this implementation, I suspect that if a new message arrives after you
have read T44 and before you finish processing messages, then the interrupt
handler will not ever be run. What testing have you done on this?

Actually, for the MXT224E on which I've been testing, the CHG line
goes high after reading the first byte of the T5 message (in other
words, the first byte of the first message, no matter how many
messages T44 reports).

The CHG line behaviour is configurable - you should find this documented in
the data sheet.

Expect an updated version of this patchset soon.

Thanks for all of your reviews!

no problem.

-- 
Nick Dyer
Software Engineer, ITDev Ltd
Hardware and Software Development Consultancy
Website: http://www.itdev.co.uk