summaryrefslogtreecommitdiffstats
path: root/Documentation
diff options
context:
space:
mode:
authorJean Delvare <khali@linux-fr.org>2012-10-05 22:23:54 +0200
committerJean Delvare <khali@endymion.delvare>2012-10-05 22:23:54 +0200
commite7ee51405835cac72e7b6e0ff26dba608cf186cc (patch)
treecbc011e79a70e32475c3f7f4d91e5a6739eedfd0 /Documentation
parent2614a8594152cad49047ed02255a3d9fbbbea8dd (diff)
downloadblackbird-op-linux-e7ee51405835cac72e7b6e0ff26dba608cf186cc.tar.gz
blackbird-op-linux-e7ee51405835cac72e7b6e0ff26dba608cf186cc.zip
i2c-mux-gpio: Add support for dynamically allocated GPIO pins
The code instantiating an i2c-mux-gpio platform device doesn't necessarily know in advance the GPIO pin numbers it wants to use. If pins are on a GPIO device which gets its base GPIO number assigned dynamically at run-time, the values can't be hard-coded. In that case, let the caller tell i2c-mux-gpio the name of the GPIO chip and the (relative) GPIO pin numbers to use. At probe time, the i2c-mux-gpio driver will look for the chip and apply the proper offset to turn relative GPIO pin numbers to absolute GPIO pin numbers. The same could be (and was so far) done on the caller's end, however doing it in i2c-mux-gpio has two benefits: * It avoids duplicating the code on every caller's side (about 30 lines of code.) * It allows for deferred probing for the muxed part of the I2C bus only. If finding the GPIO chip is the caller's responsibility, then deferred probing (if the GPIO chip isn't there yet) will not only affect the mux and the I2C bus segments behind it, but also the I2C bus trunk. Signed-off-by: Jean Delvare <khali@linux-fr.org> Cc: Peter Korsgaard <peter.korsgaard@barco.com>
Diffstat (limited to 'Documentation')
0 files changed, 0 insertions, 0 deletions
OpenPOWER on IntegriCloud