Dear Hannes,
In message 1391444250-29430-1-git-send-email-oe5hpm@oevsv.at you wrote:
changes to omap24_i2c_write(...) for polling ARDY Bit from IRQ-Status. Otherwise on a subsequent call the transfer of last byte from the predecessor is aborted and therefore lost. For exmaple when i2c_write(...) is followed by a i2c_setspeed(...) (which has to deactivate and activate master for changing psc,...).
Signed-off-by: Hannes Petermaier oe5hpm@oevsv.at
Changes for v2:
- fixed compile error due to '= ='
- removed [PATCH 1/2]: only 1 patch is needed
- fixed omap24_i2c_write(...) for waiting until all transfer is finished.
General note: it is always a wise idea to add the responsible custodian (here Heiko) on Cc: (done here). [You can automate this by adding a Cc: line below you Signed-off-by: entry - then git-send-email will automatically do what is needed.
- /* some divisors may cause a precission loss, but shouldn't
* be a big thing, because i2c_clk is then allready very slow.*/
This is incorrect multi-line comment format; it should look like this:
/* * some divisors may cause a precission loss, but shouldn't * be a big thing, because i2c_clk is then allready very slow. */
Please fix globally.
- omap24_i2c_setspeed(adap, speed);
omap24_i2c_setspeed() returns int, and can return error codes. These should be handled - here, and everywhere else where the function is being called.
- /* poll ARDY bit for making sure that last byte really has been
* transferred on the bus.*/- do {
status = wait_for_event(adap);- } while (!(status & I2C_STAT_ARDY));
This is potentially an endless loop. Please make sure that it will time out (and then issue an error message).
Best regards,
Wolfgang Denk