[TESTING] - NAND OMGB (1.2.3)

[[ACL]]

Originally Posted by vzihome

Muziling had mentioned to ACL about a battery hack he found, ACL said that he would release it on next build don't know if he did on this build...correct me if I'm wrong!

Sent from my Touch Pro 2 using Tapatalk

Ahh, you are correct. Well the hack was to avoid these nasty div by 0 we used to get in the logs.

diff --git a/arch/arm/mach-msm/htc_battery_smem.c b/arch/arm/mach-msm/htc_battery_smem.c
index 1d9e750..7967b28 100644
--- a/arch/arm/mach-msm/htc_battery_smem.c
+++ b/arch/arm/mach-msm/htc_battery_smem.c
@@ -1371,7 +1371,12 @@ static int htc_rhod_batt_corr( struct battery_info_reply *buffer )
        
         /* TODO: temp algo is not ok..if someone have the timte -> must look into the driver */
        buffer->batt_tempRAW = ( buffer->batt_tempRAW * 5200 ) / htc_adc_range;
-       av_index = ( buffer->batt_tempRAW * 18 ) / ( 2600 - buffer->batt_tempRAW );
+       /* HACK: This is a giant hack since if batt_tempRAW = 2600 we get a div by 0 */
+       if (buffer->batt_tempRAW < 2600) {
+               av_index = ( buffer->batt_tempRAW * 18 ) / ( 2600 - buffer->batt_tempRAW );
+       } else { 
+               av_index =1347;
+       }
        buffer->batt_tempRAW = htc_battery_temperature_lut( av_index );
        if ( buffer->batt_tempRAW < 250 ) /* fixed til so long we can get constant values - otherwise the algo wouln'T work good */
                buffer->batt_tempRAW = 250;

I'll prioritise this if everyone is getting less and less batt time now.

[anish88]

Originally Posted by [ACL]

Ahh, you are correct. Well the hack was to avoid these nasty div by 0 we used to get in the logs.

diff --git a/arch/arm/mach-msm/htc_battery_smem.c b/arch/arm/mach-msm/htc_battery_smem.c
index 1d9e750..7967b28 100644
--- a/arch/arm/mach-msm/htc_battery_smem.c
+++ b/arch/arm/mach-msm/htc_battery_smem.c
@@ -1371,7 +1371,12 @@ static int htc_rhod_batt_corr( struct battery_info_reply *buffer )
        
         /* TODO: temp algo is not ok..if someone have the timte -> must look into the driver */
        buffer->batt_tempRAW = ( buffer->batt_tempRAW * 5200 ) / htc_adc_range;
-       av_index = ( buffer->batt_tempRAW * 18 ) / ( 2600 - buffer->batt_tempRAW );
+       /* HACK: This is a giant hack since if batt_tempRAW = 2600 we get a div by 0 */
+       if (buffer->batt_tempRAW < 2600) {
+               av_index = ( buffer->batt_tempRAW * 18 ) / ( 2600 - buffer->batt_tempRAW );
+       } else { 
+               av_index =1347;
+       }
        buffer->batt_tempRAW = htc_battery_temperature_lut( av_index );
        if ( buffer->batt_tempRAW < 250 ) /* fixed til so long we can get constant values - otherwise the algo wouln'T work good */
                buffer->batt_tempRAW = 250;

I'll prioritise this if everyone is getting less and less batt time now.

I'd be interested for sure. I have never gotten good battery life.

[sykosoft]

In reviewing the last several pages of posts, did I miss the release notes for the 01/16/12 release?

Michael

[arrrghhh]

Originally Posted by sykosoft

In reviewing the last several pages of posts, did I miss the release notes for the 01/16/12 release?

Michael

Yes..

http://forum.ppcgeeks.com/2164598-post993.html

[sykosoft]

Originally Posted by arrrghhh

Yes..

http://forum.ppcgeeks.com/2164598-post993.html

Thank you, I had indeed missed that.

Michael

[mbellot]

Originally Posted by [ACL]

I'll prioritise this if everyone is getting less and less batt time now.

I'll let you know tomorrow, I'm running my two-fer battery calibration now ("Battery Calibration" and "Battery Left").

Last time I ran it I was going 38 hours before the battery cried uncle.

I've pushed gps from 1.2.2 into this build but have it turned off for now. Once things appear stable I'll start mucking around with it again.

Updated the keymaps to get long press recent apps. Still need to try updating the APN to see if Boost MMS will work.

I'm hoping LK doesn't need another update for a while, repeating today's pain is not high on my list of fun things to do.

[milda25]

Originally Posted by [ACL]

Ahh, you are correct. Well the hack was to avoid these nasty div by 0 we used to get in the logs.

diff --git a/arch/arm/mach-msm/htc_battery_smem.c b/arch/arm/mach-msm/htc_battery_smem.c
index 1d9e750..7967b28 100644
--- a/arch/arm/mach-msm/htc_battery_smem.c
+++ b/arch/arm/mach-msm/htc_battery_smem.c
@@ -1371,7 +1371,12 @@ static int htc_rhod_batt_corr( struct battery_info_reply *buffer )
        
         /* TODO: temp algo is not ok..if someone have the timte -> must look into the driver */
        buffer->batt_tempRAW = ( buffer->batt_tempRAW * 5200 ) / htc_adc_range;
-       av_index = ( buffer->batt_tempRAW * 18 ) / ( 2600 - buffer->batt_tempRAW );
+       /* HACK: This is a giant hack since if batt_tempRAW = 2600 we get a div by 0 */
+       if (buffer->batt_tempRAW < 2600) {
+               av_index = ( buffer->batt_tempRAW * 18 ) / ( 2600 - buffer->batt_tempRAW );
+       } else { 
+               av_index =1347;
+       }
        buffer->batt_tempRAW = htc_battery_temperature_lut( av_index );
        if ( buffer->batt_tempRAW < 250 ) /* fixed til so long we can get constant values - otherwise the algo wouln'T work good */
                buffer->batt_tempRAW = 250;

I'll prioritise this if everyone is getting less and less batt time now.

Arrrggghh, I wrote about this post (concerning battery life). BTW: what about your battery life? Am I alone with this problem or has someone same problem?
And about GPS - my GPS doesn't see satelites. I tested it in Google Maps and Sygic. GPS finding, but with no success (no signal from universe ) Tried stock radio 4.49.25.91 and latest 4.49.25.95. Is this problem in my hand or is there any bug? Of course, zip from ACL released a few days ago I have installed.

[arrrghhh]

Originally Posted by milda25

Arrrggghh, I wrote about this post (concerning battery life). BTW: what about your battery life? Am I alone with this problem or has someone same problem?
And about GPS - my GPS doesn't see satelites. I tested it in Google Maps and Sygic. GPS finding, but with no success (no signal from universe ) Tried stock radio 4.49.25.91 and latest 4.49.25.95. Is this problem in my hand or is there any bug? Of course, zip from ACL released a few days ago I have installed.

Battery life - not so good compared to HaRET. ACL's post does seem to indicate that it would effect battery life - so perhaps I have misunderstood. Wouldn't be the first time .

I'll have to play with GPS more. I thought it worked, but I guess I've never tried to get a lock on OMGB - I was recalling my testing with GBX0B. Sorry for the mixup. I'm trying GPS inside right now, so far no luck - but that usually doesn't work unless I do QuickGPS in WinMo first, which isn't an option here .

[mbellot]

Originally Posted by arrrghhh

Battery life - not so good compared to HaRET. ACL's post does seem to indicate that it would effect battery life - so perhaps I have misunderstood. Wouldn't be the first time .

I'm getting 24+ hours of light use with OMGB. Not quite as good as FRX07.1, but darn close.

Originally Posted by arrrghhh

I'll have to play with GPS more. I thought it worked, but I guess I've never tried to get a lock on OMGB - I was recalling my testing with GBX0B. Sorry for the mixup. I'm trying GPS inside right now, so far no luck - but that usually doesn't work unless I do QuickGPS in WinMo first, which isn't an option here .

GPS worked on the previous OMGB (12-21 + acoustic) when I added the 1.2.2 gps.rhodium.so library into the build. Lock was not super fast, but not bad.

It never locked on a location inside for me. I don't ever remember being able to get a GPS lock indoors with WinMo, so I'm guessing it's a hardware limitation.

One problem I saw (more inferred) was that the GPS wasn't being fully turned off because battery life would go to he!! until I rebooted.

It was also fairly certain to get a hard lock-up (requiring reset button) if I was actively using the GPS for more than 20-30 minutes (Maps while driving, c:geo for geocaching, etc).

[arrrghhh]

Originally Posted by mbellot

I'm getting 24+ hours of light use with OMGB. Not quite as good as FRX07.1, but darn close.

My battery could be close to death, but I don't get nearly that...

Originally Posted by mbellot

GPS worked on the previous OMGB (12-21 + acoustic) when I added the 1.2.2 gps.rhodium.so library into the build. Lock was not super fast, but not bad.

It never locked on a location inside for me. I don't ever remember being able to get a GPS lock indoors with WinMo, so I'm guessing it's a hardware limitation.

One problem I saw (more inferred) was that the GPS wasn't being fully turned off because battery life would go to he!! until I rebooted.

It was also fairly certain to get a hard lock-up (requiring reset button) if I was actively using the GPS for more than 20-30 minutes (Maps while driving, c:geo for geocaching, etc).

Well it definitely worked with the flashed zip. Took a little while to lock, and certainly did take going outside. For the record mbellot, I have definitely gotten locks inside - but again, only after using QuickGPS. With no assistance, the radio takes quite a long time to get a fix.

The other bugs you mention are pretty much status quo - although GPS doesn't seem to kill sleep, you think it still does despite the indicator going green? I haven't had a problem with the phone sleeping, but I haven't used GPS a whole lot. Just got a speedo app to see how accurate my speedo is in my truck on the way home from work... worked great! I don't think I used it for even 20 minutes.... maybe 10. That lockup bug sounds like the same thing that happens in HaRET... Not sure anyone has even captured logs from such a crash.