Hi Alex, On Wed, May 08, 2013, Alex Deymo wrote: > Changes from the v4 are: > * uint32_t --> unsigned int for the "attempt" argument. I think that > restrict that value to unsigned given that -1 is never used for > attempt is a good practice. > * "pincb_iter" --> "btd_adapter_pin_cb_iter". The "pin_cb" instead > of "pincb" is to match the btd_adapter_pin_cb_t callback type > already used in the code. > * Added a bool pincode_requested member to btd_adapter (see patch 4/8) > to signal if the device requested a pincode during the bonding attempt. > This is basically to fix the case were we keep retrying the bonding > and the device keeps sending an authentication error without consuming > a pin code from the pincode callback list. This ensures that either a > pincode was consumed from the list before retrying. Together with the > fact that the pin codes provided by the callback list is a finite list, > this ensures the retry mechanism will finish. > > Please consider this version for bluez. We are actively testing this > patchset in our last chromiumos release and I don't have any bug reported > regarding this feature. Thanks! > Alex. > > Alex Deymo (8): > core: Convert the pincode callback to an interable list. > plugins: Extend the pin code callback with the call number > core: Add support for retrying a bonding > core: retry bonding attempt until the iterator reaches the end. > core: Add device_get_class to the public interface. > autopair: Add the autopair plugin. > core: Expose the last bonding attempt timeout on retry > autopair: Try a fixed pincode for keyboards rejecting random codes > > Makefile.plugins | 3 + > plugins/autopair.c | 165 ++++++++++++++++++++++++++++++++++++++++++++++++++ > plugins/wiimote.c | 7 ++- > src/adapter.c | 175 ++++++++++++++++++++++++++++++++++++++++++++++------- > src/adapter.h | 11 +++- > src/device.c | 135 +++++++++++++++++++++++++++++++++++++++++ > src/device.h | 7 +++ > 7 files changed, 478 insertions(+), 25 deletions(-) > create mode 100644 plugins/autopair.c All patches in this set have been applied. Thanks! I did fix one thing in patch 4: for type bool the expected assignments are true/false and not TRUE/FALSE (which are GLib defines). I also pushed one small coding style/readability patch on top of this set. Johan -- To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
