Opened 6 years ago

Closed 6 years ago

#293 closed enhancement (fixed)

Prepare to work with 2.6.28 (andy-tracking) kernel

Reported by: PaulFertser Owned by: mickey
Priority: minor Milestone: milestone5
Component: framework/general Version:
Keywords: Cc: ole.d@…

Description

As this kernel will soon be declared a new official "stable", we should prepare to work with it. I've installed c59a8bdc7ca4b547 and tested how FSO works with it.

It seems to be quite stable (except don't try to press the power button unless you are sure it's fully suspended; fixed in new versions). I've tested LEDs, power supplies, brightness control, suspend/resume, GPRS, GPS, bluetooth ("hcitool dev" works as expected) and WiFi (when there're APs nearby, iwlist scan shows them, else a cryptic message is output). Everything seem to be working ok.

Two things i noticed so far: after suspend/resume bluetooth no longer works. WiFi can't be disabled even right after reboot.

Attached is a minimal patch against stabilization/milestone4 that modifies the hardcoded /sys paths to correspond with those of the new kernel. It should also apply cleanly (except the "apm -s" part that was already changed to use the new semantics) to the HEAD.

Attachments (1)

0001-Modify-sys-paths-to-work-with-andy-tracking-kernel.patch (5.1 KB) - added by PaulFertser 6 years ago.

Download all attachments as: .zip

Change History (9)

comment:1 Changed 6 years ago by mickey

  • Status changed from new to assigned

Catched up with the new sysfs paths in a slightly better way, luckily all device nodes are linked from /sys/bus/platform/devices, so we can support both kernel tracks at the same time.

The only thing missing is the dbus name of the backlight node, I want to solve this otherways (introducing a wildcard or so).

comment:2 Changed 6 years ago by mickey

  • Resolution set to fixed
  • Status changed from assigned to closed

I did settle for another way to deal with the different backlight node name (additional numerical incremental alias). FSO should now fully support both kernel tracks.

comment:3 Changed 6 years ago by odlg

  • Cc ole.d@… added
  • Resolution fixed deleted
  • Status changed from closed to reopened

Using fso-testing and andy-tracking, it seems that the GPS is never powered on.

comment:4 Changed 6 years ago by jluebbe

Which git-revision of the framework are you using? This was no problem in our tests.

comment:5 Changed 6 years ago by odlg

I am using 0.8.4.5+gitr97bd68d4a2f3681dfb35647a62ca12dd8b287f8e-r0 from fso-testing

comment:6 Changed 6 years ago by jluebbe

This is the latest revision from the stabilization/milestone4 branch, which doesn't contain our new work.

comment:7 Changed 6 years ago by odlg

OK, I just upgraded from fso-testing to fso-unstable, hoping that would give the latest. It seems it did not. I am somewhat confused regarding where to get the latest, is there a daily updated feed with frameworkd and zhone?

comment:8 Changed 6 years ago by stefan

  • Resolution set to fixed
  • Status changed from reopened to closed

You can find unstable images here: http://downloads.freesmartphone.org/fso-unstable/

It's not daily but often.

I close this bug as it should really be fixed in a newer version

Note: See TracTickets for help on using tickets.