[TESTING] - NAND OMGB (1.2.3)

[[ACL]]

Originally Posted by detule

As far as hacks go, this does wonders for my missing ringer.

Also there's stuff up on gitorious for the mic on incoming calls issue.

Though, I guess differences from gbx0b to omgb might render these ineffectual. Plus it sounds like you are on your way to figuring out more elegant solutions.

I havent actually solved the issue and havent commited the change either until i get enough feedback from the other guys. We are a few commits behind on AudioHardware.cpp since i'm not sure if we even need them. But its simple. here is what i added.

--- a/libaudio-wince/AudioHardware.cpp
+++ b/libaudio-wince/AudioHardware.cpp
@@ -1028,6 +1028,12 @@ status_t AudioHardware::doRouting()
             audProcess = (ADRC_ENABLE | EQ_ENABLE | RX_IIR_ENABLE);
         } else {
             LOGI("Routing audio to Handset\n");
+                       if ((mMode == AudioSystem::MODE_NORMAL) && (mCurSndDevice != -1)) {
+                               LOGI("BUG: We are ignoring request to allow ring!!\n");
+                               LOGI("Mode %d out devices %x currentdevice %d)\n", mMode, outputDevices, mCurSndDevice);
+                               mCurSndDevice = sndDevice;
+                               return NO_ERROR;
+                       }
             sndDevice = SND_DEVICE_HANDSET;
         }
     }

On the cdma rhod, this condition exists where the sound is routed to handset right before it swaps to the speaker. Simply bypassing this condition helps it ring.