Raspberry Pi doesn't have a modem or BT, so the code devoted to managing them is just wasting our precious CPU cycles.
Created attachment 8494048 [details] [diff] [review] Add configure options for force-disabling ril and bluetooth
(In reply to Chris Jones [:cjones] mostly inactive; ni?/f?/r? if you need me from comment #0) > Raspberry Pi doesn't have a modem or BT, so the code devoted to managing > them is just wasting our precious CPU cycles. Hum, surely, if there's no model or BT, and the code to manage them *run*, there's something very wrong. I can understand that you'd want to save the disk space (although, how much does that strip? a few kB?), but I wouldn't expect CPU cycles to be wasted. If there are, I'd say there's something wrong that should be fixed.
Sorry, missed this comment. I don't care about the disk space. Enabling MOZ_BT and MOZ_RIL tell gecko there's supposed to be a bluetooth chip and modem. When they fail to connect, they try to reconnect because there are race conditions in the initialization of those devices. That's where the wasted cycles and noise come from. (Bluetooth is particularly aggressive with retries.) I don't see any point in wasting engineer time in trying to support a BT/RIL-enabled-but-no-chip/modem code path. What's the benefit for that new code? On the other hand, the code paths in which BT/RIL are *disabled*, *are* supported already, because that's the case on every platform besides Gonk-on-phone. Why not ride on that existing long-term-committed maintenance burden?
Does passing --disable-b2g-ril not work? That would surprise me.
(Without your patch, that is).
Not according to autoconf.mk. I double-check again later this afternoon.
OK! I swear I tested --disable-b2g-ril and --disable-b2g-bt before, but however I flubbed it then they seem to be working now. One less patch to land.
Status: NEW → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → INVALID
4 years ago
You need to log in before you can comment on or make changes to this bug.