2019-06-05, 12:01
(2019-06-04, 23:00)Milhouse Wrote:Hello, yes I just tested every kernel upgrade from 0426 and indeed the issue appears in #0512 and working fine in #0511(2019-06-04, 17:27)Milhouse Wrote: @asavah Thanks, but the commit you suggest reverting first appeared in 5.1.1 and isn't in 5.0.y until 5.0.15 (the last 5.0.y kernel for these builds is 5.0.13) but the problem being reported by @SirMacke is apparently with 5.0.10 (#0427). Unfortunately it's a vague confirmation so I can't be sure as he's not sure, and I don't have the time to look into vague reports.
If the bad build number is confirmed/certain then sure, I can spend some time trawling through whatever changed.
@SirMackie all you need to do is test #0426 and #0427 - if you confirm #0426 works and #0427 does not then that's something for me/others to go on, but unfortunately if you're not sure of the build numbers I don't have the time for a potentially wild goose chase.
This RPi post suggests that Bluetooth issues were first seen in #0514 which - in the case of RPi - corresponds with the introduction of kernel 5.1.1 (updating from 5.1.0). This is why I'm a bit anal about bug reports being accurate rather than vague with regard to working/not working build numbers!
In kernel 5.1.1 there are four new Bluetooth related commits:
https://github.com/torvalds/linux/commit...228e50cce7
https://github.com/torvalds/linux/commit...0c11a503a1
https://github.com/torvalds/linux/commit...4522ae806b (as suggested by @asavah)
https://github.com/torvalds/linux/commit...b44ec4d3d3
For x86_64 we switched to 5.1.1 with #0512.
@SirMacke it would be worth testing #0511 and #0512, and confirming if the same Bluetooth working/not working behaviour is seen there.
I'll upload a build later this evening with the commit suggested by @asavah reverted.