Subject: escapism - msg#00022

Previous Message by Date: click to view message preview

Re: external headset mic

Thanks Anton, On Tue, 2007-04-10 at 15:28 +0400, Anton Vorontsov wrote: > Hello Travis, > > On Mon, Apr 09, 2007 at 09:08:07PM -0500, Travis Stratman wrote: > > Hello all, > > > > It's been a couple of weeks since I brought this up, but I finally got a > > chance to look at the code and work something out. > > > > Here's where we left off: > > > The headset parts should work on the 4700. The headset > > > microphone is > > > wired into the EXT (external microphone) and MDT (microphone detect) > > > pins on the AK4641, so I'd start by tracking down the 4641 > > > driver to see > > > if we're paying attention to the microphone detect signal. > > > > I did this, and yes, the driver was properly paying attention to the mic > > detect signal. However, I think that there was an issue with timing. I > > built 2.6.20-hh4 and installed it and tested the headset again -- no > > game. So, I went into the code and added around 10 printk's to track > > what was going on and print the register values that were being set. To > > my surprise, when I loaded the new module, it worked like a charm. > > Since I didn't really change anything in the code, my best guess is some > > sort of timing issue, right? > > > > My strategy now is to start going through and replacing the prints that > > seem to be in a place where a delay would make a difference w/ micro > > sleeps (about 5000 as a value to replace a printk?) until I get it > > optimized and then I'll throw a patch out there (might be another week > > or two). If anyone has some better suggestions, I'm definitely open. > > Could you please post current patch, with CC: > kernel-discuss-CN5wO63fgwogsBAKwltoeQ@xxxxxxxxxxxxxxxx > > I do not promise, but maybe other folks will have some idea about > delay issue. The patch is pasted below. If anyone has any suggestions about the delay or other ways to fix this please let me know. Otherwise, I will work on optimizing it by hand. My best guess is to insert a delay to replace one of the first two printk() in the select_mic function. The other thing that I suspect is possibly the printk that reads the value of the registers in the select_mic function. Thanks, -Travis ------------------------------- --- sound/i2c/ak4641.c.orig 2007-04-05 10:28:37.000000000 -0500 +++ sound/i2c/ak4641.c 2007-04-05 11:23:04.000000000 -0500 @@ -367,23 +367,31 @@ void snd_ak4641_hp_connected(struct snd_ak4641 *ak, int connected) { snd_ak4641_lock(ak); + /* Some debug statements */ + printk(KERN_DEBUG "snd_ak4641_hp_connected: detecting headphone\n"); if (connected != ak->hp_connected) { ak->hp_connected = connected; + printk(KERN_DEBUG "Headphone connected has changed to %s\n", connected ? "true" : "false"); /* headphone or speaker, on playback */ if (ak->playback_on) { if (connected) { snd_ak4641_headphone_on(ak, 1); snd_ak4641_speaker_on(ak, 0); + printk(KERN_DEBUG "Turned headphone on, speaker off\n"); } else { snd_ak4641_speaker_on(ak, 1); snd_ak4641_headphone_on(ak, 0); + printk(KERN_DEBUG "Turned speaker on, headphone off\n"); } } /* headset or internal mic, on capture */ - if (ak->capture_on) + if (ak->capture_on) { + printk(KERN_DEBUG "Selecting mic\n"); snd_ak4641_select_mic(ak); + + } } snd_ak4641_unlock(ak); } @@ -532,9 +540,12 @@ { int mic = 0; u8 r_mic; + + printk(KERN_DEBUG "snd_ak4641_select_mic\n"); if (ak->hp_connected) { /* check headset mic */ + printk(KERN_DEBUG "Checking headset mic\n"); ak->regs[R_MIC] = WRITE_MASK(ak->regs[R_MIC], R_MIC_MPWRE, R_MIC_MPWRE); snd_ak4641_hwsync_write(ak, R_MIC); snd_ak4641_hwsync_read(ak, R_STATUS); @@ -544,9 +555,11 @@ (mic ? R_MIC_MSEL : 0x00) | (ak->capture_on ? (mic ? R_MIC_MPWRE : R_MIC_MPWRI) : 0x00), R_MIC_MSEL | R_MIC_MPWRI | R_MIC_MPWRE); + printk(KERN_DEBUG "r_mic = 0x%X, ak->regs[R_MIC] = 0x%X\n", r_mic, ak->regs[R_MIC]); if (r_mic != ak->regs[R_MIC]) { ak->regs[R_MIC] = r_mic; snd_ak4641_hwsync_write(ak, R_MIC); + printk(KERN_DEBUG "synched to hardware, exiting select_mic\n"); } } ----------------------------

Next Message by Date: click to view message preview

Re: Trying to get Open Embedded Working

alright cool thanks. Downgrading to the 1.6.6 bitbake version seemed to help, but now I'm getting this error when I do bitbake Mplayer . . . . NOTE: package qemu-native-0.7.0-r0: task do_fetch: completed NOTE: package qemu-native-0.7.0-r0: task do_unpack: started NOTE: Unpacking /stuff/build/tmp/downloads/qemu-0.7.0.tar.gz to /stuff/build/tmp/work/i686-linux/qemu-native-0.7.0-r0/ NOTE: package qemu-native-0.7.0-r0: task do_unpack: completed NOTE: package qemu-native-0.7.0-r0: task do_patch: started NOTE: package qemu-native-0.7.0-r0: task do_patch: completed NOTE: package qemu-native-0.7.0-r0: task do_configure: started NOTE: package qemu-native-0.7.0-r0: task do_configure: completed NOTE: package qemu-native-0.7.0-r0: task do_compile: started ERROR: function do_compile failed ERROR: see log in /stuff/build/tmp/work/i686-linux/qemu-native-0.7.0-r0/temp/log.do_compile.11071 NOTE: Task failed: /stuff/build/tmp/work/i686-linux/qemu-native-0.7.0-r0/temp/log.do_compile.11071 NOTE: package qemu-native-0.7.0-r0: task do_compile: failed ERROR: TaskFailed event exception, aborting NOTE: package qemu-native-0.7.0: failed ERROR: Build of mplayer failed Any ideas about how to fix this? Is this because of a problem with my setup? I've attached at copy of the log file the error output mentioned. Thanks, Jared On Tue, 2007-04-10 at 13:50 +0200, Rene Wagner wrote: > On Tue, 2007-04-10 at 13:21 +0200, Rene Wagner wrote: > > On Mon, 2007-04-09 at 19:44 -0400, Jared Stemen wrote: > > > I have no idea what it is or how to fix it? Any Ideas? I followed this > > > guide http://www.openembedded.org/wiki/GettingStarted . But I'm using > > > org.openembedded.oz354fam083 branch instead of the org.openembedded.dev > > > branch. > > > > If you intend to build stuff for Familiar 0.8.4 I suggest using our OE > > tree instead [1] which comes with a compatible version of bitbake. > > > > The error you pasted looks a lot like you're trying to use a more recent > > version of bitbake which appears to be incompatible. Before you proceed, > > make sure to wipe any traces of it in case you installed it > > (/usr/lib/python* or something). > > Err, sorry. Don't wipe /usr/lib/python*! Just look for anything bitbake > in there and remove it. > > > I should add that you may need gcc 3.3 or 3.4 to compile a few native > > tools which won't compile with 4.x. > > > > Regards, > > > > Rene > > > > [1] http://familiar.handhelds.org/#build > log.do_configure.11071 Description: Text document _______________________________________________ Hx4700-port mailing list Hx4700-port-CN5wO63fgwogsBAKwltoeQ@xxxxxxxxxxxxxxxx https://www.handhelds.org/mailman/listinfo/hx4700-port

Previous Message by Thread: click to view message preview

RE: external headset mic

Hello all, It's been a couple of weeks since I brought this up, but I finally got a chance to look at the code and work something out. Here's where we left off: > The headset parts should work on the 4700. The headset > microphone is > wired into the EXT (external microphone) and MDT (microphone detect) > pins on the AK4641, so I'd start by tracking down the 4641 > driver to see > if we're paying attention to the microphone detect signal. I did this, and yes, the driver was properly paying attention to the mic detect signal. However, I think that there was an issue with timing. I built 2.6.20-hh4 and installed it and tested the headset again -- no game. So, I went into the code and added around 10 printk's to track what was going on and print the register values that were being set. To my surprise, when I loaded the new module, it worked like a charm. Since I didn't really change anything in the code, my best guess is some sort of timing issue, right? My strategy now is to start going through and replacing the prints that seem to be in a place where a delay would make a difference w/ micro sleeps (about 5000 as a value to replace a printk?) until I get it optimized and then I'll throw a patch out there (might be another week or two). If anyone has some better suggestions, I'm definitely open. Thanks, Travis

Next Message by Thread: click to view message preview

What does everyone user their linux-running-hx4700s for?

I was just curious; what does everyone use their hx4700s for? I'd like to use mine to listen to mp3s, browser the Internet, pim applications, and maybe as a vnc client. However, the more I look into it, the more I think it may not be possible. Ps does anyone know of an mp3 player with a gui that can easily be built with Open Embedded for gpe? -Thanks, Jared