Support PAN profile for bluedroid

Probably blocks bug 972780.

Support PAN feature on bluetoothd

It looks like we need to require the root permission (modify the "init.bluetooth.rc") to build special device nodes(e.g. dev/tun, socket, ..., etc.) for supporting PAN profile. I will file another bug to discussing whether we need to require the root permission on bluetoothd.

(In reply to Tom Tung from comment #3) > Created attachment 8692461 [details] > Bug 1227907 - Support PAN feature on bluetoothd - bluetoothd change > > It looks like we need to require the root permission (modify the > "init.bluetooth.rc") to build special device nodes(e.g. dev/tun, socket, > ..., etc.) for supporting PAN profile. I will file another bug to discussing > whether we need to require the root permission on bluetoothd. Bug 1228283

Comment on attachment 8692461 [details] Bug 1227907 - Support PAN feature on bluetoothd - bluetoothd change Thanks! Having PAN support is pretty cool. Just r- until the comments have been addresses, but most of it is trivial. I'd prefer to not land this patch until there's a patch for the Gecko side and some example code that makes use of PAN.

(In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #5) > Comment on attachment 8692461 [details] > Bug 1227907 - Support PAN feature on bluetoothd - bluetoothd change > > Thanks! Having PAN support is pretty cool. > > Just r- until the comments have been addresses, but most of it is trivial. > > I'd prefer to not land this patch until there's a patch for the Gecko side > and some example code that makes use of PAN. Hi Thomas, Thank you for your reviews and comments! I revise my code and update it to [1]. Actually I have done part of basic functions(connect/disconnect) of PAN profile for the Gecko side and uploaded them as a patch to Bug 972780. Would you like to review it? [1] https://github.com/mozilla-b2g/platform_system_bluetoothd/pull/40

Hi Tom > Actually I have done part of basic functions(connect/disconnect) of PAN > profile for the Gecko side and uploaded them as a patch to Bug 972780. Would > you like to review it? Ah, I see. WebIDL is not my domain, but I can review the IPC protocol code from that patch.

> [1] https://github.com/mozilla-b2g/platform_system_bluetoothd/pull/40 I left more comments in the pull request.

(In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #7) > Ah, I see. WebIDL is not my domain, but I can review the IPC protocol code > from that patch. Hi Thomas, Thanks! I will split my patch into two patches. One of them will be daemon part and I will ask you for review.

(In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #8) > > [1] https://github.com/mozilla-b2g/platform_system_bluetoothd/pull/40 > > I left more comments in the pull request. Thanks! I revise my code and update it again.

(In reply to Tom Tung from comment #10) > (In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #8) > > > [1] https://github.com/mozilla-b2g/platform_system_bluetoothd/pull/40 > > > > I left more comments in the pull request. > > Thanks! I revise my code and update it again. I left some comments.

(In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #11) > > > > [1] https://github.com/mozilla-b2g/platform_system_bluetoothd/pull/40 > > I left some comments. Thanks you! I revise them and update it to [1].

Comment on attachment 8692461 [details] Bug 1227907 - Support PAN feature on bluetoothd - bluetoothd change A few more nits, but it looks good now. Thanks!

(In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #13) > Comment on attachment 8692461 [details] > Bug 1227907 - Support PAN feature on bluetoothd - bluetoothd change > > A few more nits, but it looks good now. Thanks! Hi Thomas, Thank you!! But I changed a little part of code, such as (1)PAN's order in |register_func| and |unregister_service| at "src/service.c", (2)Error handling for exceeding interface name bytes in |append_bt_ifname| at "src/bt-pan-io.c", and (3) add "static" for |append_bt_ifname|. I don't change the declaration method for IFNAME_LEN to |static const|, and the reason is that C's compiler don't treats |static const|'s variable as constant. Thus, if I do so, the |padding[IFNAME_LEN]| will cause error. I just have a little afraid whether there is anything wrong. Can you review it again? Tom

Reverted in https://github.com/mozilla-b2g/platform_system_bluetoothd/commit/570531db9e81a871c421b27025ae558022b1015a due to bug 1230589.

Tom, please test the patch on JB, KK and L. The platform APIs aren't overly stable and this kind of breakage is common. I asked you to not land this patch until you have the Gecko side ready. Please don't re-land before all patches in bug 972780 have an r+.

(In reply to Thomas Zimmermann [:tzimmermann] [:tdz] from comment #17) Thomas, OK, right now I only test the patch on KK. I will test the patch on other environment. Sorry for that. I just thought you changed the "r-" to "r+" means I can land this patch, and I was wrong. I will wait until all the patches bug 972780 have an r+. Tom

(In reply to Wes Kocher (:KWierso) from comment #16) Thanks, really sorry about that.

Reset the assignee to default since I'm not working on this bug anymore.

Firefox OS is not being worked on