Heureka! (Eureka)

During the weekend I started to read this years #asteroidos irc logs and while doing so porbably (almost certainly) found the solution to our longterm outstanding bluetooth-sleep-bug! Well it is 'kido', the lead developer of asteroidos, that found the solution... (See... 14:34:12 kido:holy guacamole!!)

So they had a huge problem with bluetooth-sleep on Qualcomm based SoC's... Sounds familiar?

After hours, even weeks of research there was a solution in the end: Turns out that the bluetooth stack never got a sleep signal from the system and therefore never went in suspend state.

If I got his solution right, he patched a small tool they use (brcm-patchram-plus) to configure the UART/IRQ for specific Broadcom bluetooth chips.The commit can be found here. Specially have a look at lines 213-216. Looks like he even included the values needed by our beloved Nexus5!


@mautz

I really think that this is the solution to our problem. I guess we should add that 'brcm-patchram-plus' tool and start it as a service after having started the bluetooth service. Looks like they do it that way in asteroidos. https://github.com/AsteroidOS/meta-a...e3c3d4fed6bc93