<hampsterblade>
So on our phone we have really bad signal in cyanogenmod. are there any known fixes in other phones you guys have worked on?
<Entropy512>
no clue what phone you have, and signal is the same as it is in Touchwizz with the same modem firmware
onethird has joined #teamhacksung
<hampsterblade>
Galaxy S 4G, and basically on stock I get kind of poor service where I am, but on CM7 it's really slow to connect and a lot of the time doesn't connect at all
<hampsterblade>
I'm working on the kernel, but don't really know where to look to troubleshoot
<hampsterblade>
interesting
<hampsterblade>
of fuck I forgot to install the efs
<roniez>
Cora and Aschuan makes up for all of that.
rtfpessoa has quit [Quit: Instantbird 1.1]
xplodwild_work has quit [Ping timeout: 245 seconds]
xplodwild_work has joined #teamhacksung
xplodwild_work has quit [Quit: Page closed]
lidroid has joined #teamhacksung
<ohanar>
pershoot: nvm, I should really learn to ignore case when greping through code :)
Kaik541 has quit [Ping timeout: 260 seconds]
<atinm|away>
oh man .. guild wars 2 is in beta..
<atinm|away>
the only video game I've ever played was Guild Wars... I was playing that the night I ended up in the hospital for my liver transplant and a lot of my hallucinations were Guild Wars NPCs chasing me around
<atinm|away>
going to have to find time for this after I find time to start coding android again :-P
Kaik541 has joined #teamhacksung
antiochasylum has joined #teamhacksung
<antiochasylum>
@codework, is the n7000 still on official cm repo?
<antiochasylum>
or was it moved back to teamhacksung?
Kaik541 has quit [Ping timeout: 245 seconds]
Kaik541 has joined #teamhacksung
<antiochasylum>
anyone aliver?
<antiochasylum>
alive*
Kaik541 has quit [Ping timeout: 245 seconds]
<lolwat>
why does dsp manager not work?
<lolwat>
(galaxy s2, cm9 nightly)
ExtremeDevilz has joined #teamhacksung
sbrissen_gone is now known as sbrissen
lodder has quit [Ping timeout: 244 seconds]
xplodwild_work has joined #teamhacksung
lodder has joined #teamhacksung
mobilx has joined #teamhacksung
Kaik541 has joined #teamhacksung
<antiochasylum>
damn, anyone alive?
<antiochasylum>
damn, anyone alive? 2 days ago the n7000 repo showed up on official CM repo, but now its gone, was it removed I supposed?
<xplodwild_work>
antiochasylum: I answered you on XDA
<antiochasylum>
Yea, its under teamhacksung
<antiochasylum>
but he told me to use offical cm repo, but I cant find it under the official repo
<xplodwild_work>
he's using CM9 kernel so nothing there
<xplodwild_work>
so I'm almost sure it's mfc_fw issue
<xplodwild_work>
can't check as I'm at work
ExtremeDevilz has joined #teamhacksung
<stickyboy>
Entropy512: Did you see my message last night about in-call audio on Tab 7+? Speakerphone b0rked like on Note (remote can't hear you), but mic is ok.
<Entropy512>
stickyboy: yeah
<Entropy512>
interesting
<Entropy512>
didn't think you guys had anything other than speakerphone for calls
<Entropy512>
xplodwild_work: it may be failing due to resolution issues, not firmware - I just tested MFC and it's working fine, but it may be that's the sort of errors you get if you don't have the camera resolution fixups that recently got moved
<stickyboy>
Just so we're clear: you need the exynos4tab-common props AND device tree. As well as the p6210 props AND device tree. :)
<stickyboy>
They're all on mine and or ohanar's github.
Rebellos has joined #teamhacksung
<Hero_Over>
k thank you
<Rebellos>
Hey there! Do you think that it matters if I execute "gpio_direction_output(GPIO_CAM_VGA_nRST, 1);" instead of "gpio_direction_output(GPIO_CAM_VGA_nRST, 0); gpio_set_value(GPIO_CAM_VGA_nRST, 1); " ?
<Hero_Over>
there is no overlapping between exynos tab and our specific props?
<Rebellos>
...Except that second case setting 1 will for sure generate rising edge?
<Hero_Over>
you may run into some drain with leaving the value static
<Hero_Over>
I've updated your manifest and am anxious to get home
<Hero_Over>
not enough time at work now to wubi and fork trees
<stickyboy>
Hero_Over: Overlapping? They inherit from each other.
<stickyboy>
Err, the devices call the exynos4tab-common files, both the device trees and the props.
<Hero_Over>
only an i7 can carry workload like that
<stickyboy>
16 GB RAM, though. Linux md RAID1.
<Hero_Over>
oh shit
<Hero_Over>
1600mhz?
<stickyboy>
It's not SSD, but it's not bad.
<Hero_Over>
im speaking of your ddr3 mods?
<stickyboy>
Yah, I'm not sure. Probably 1333?
<Hero_Over>
i have ocz vertex 3 500mb/s read/write
<Hero_Over>
and 6gb 2000mhz
<Hero_Over>
but 16gb is mice for building
<stickyboy>
Lemme get dmidecode to check my RAM speed.
<Hero_Over>
i need to upgrade
<Hero_Over>
i have my i7 950 @ 4.0 on water i love it
<stickyboy>
RAM is 1333 MHz.
<stickyboy>
Cool, SSD is too expensive around here. Otherwise I'd have one already.
<Hero_Over>
1333 isnt bad at all way better than ddr2
<stickyboy>
Same with 2600k. Just not worth it for me.
<Hero_Over>
yeah the k is fancy though :)
<stickyboy>
Actually, 16GB is a bit overkill. I don't think the build ever really NEEDS more than ~4.
<Hero_Over>
and my ssd was only like 200
<Hero_Over>
hmmmm while handlin the OS and building i think 8 sounds about right
<stickyboy>
I wanted to make a RAM disk, but I fucked up and erased my home folder last weekend due to a dmsetup mixup.
<Hero_Over>
unless you dont touch PC while building
<Hero_Over>
oucchhh
<Hero_Over>
what distro u on
<stickyboy>
Arch.
<Hero_Over>
u like?
<Hero_Over>
ive been a fed/ub kid for a while and looking for something trimmed
<stickyboy>
I was on Ubuntu for awhile for the ease, but I started having problems with their GCC 4.6.x creating shitty binaries.
<Hero_Over>
why not just use new gcc?
<Hero_Over>
or even older
<stickyboy>
Heh, I did. GCC 4.4.x is in the repos. I guess I'd just had enough of Ubuntu.
<Hero_Over>
yeah i just hate how over time they bloat it more and more with featureless code lol
<Hero_Over>
im a code nazi
<stickyboy>
They're going a direction which kinda gets under my skin. Like they wrote "Ubuntu Kernel" a whole bunch of times in their release notes for 12.04.
<stickyboy>
It's Linux. Period.
<Hero_Over>
yeah canonical is going the way of the fruit
<Hero_Over>
trying to own linux
<Hero_Over>
when the best part of linux is we own it and contribute
<stickyboy>
Testing off-mode charging right now...
<stickyboy>
ohanar just submitted some changes which should fix it.
<stickyboy>
ohanar == locerra, btw.
<Hero_Over>
kk
<Hero_Over>
does it use proper method or is it screen off only
<stickyboy>
ohanar == IRC and github, loceraa == XDA.
<Hero_Over>
routing to the right partition
<stickyboy>
Proper Samsung low-power mode.
<Hero_Over>
nice
<Hero_Over>
is it working?
<stickyboy>
Ah, it works... but it's funny. Lemme show you a pic.
<Hero_Over>
ha ok
<Hero_Over>
im gunna have a smoke real quick be back in 10...
<Hero_Over>
well it is not my forte but i am efficient with proper trail and error
noobnl has joined #teamhacksung
<Hero_Over>
is current codition causing wakelock ?
<stickyboy>
ohanar: Are those ~8 LPM images the only ones 7.7 has?
<ohanar>
yeah
<stickyboy>
Ok. I see about 20 for us.
<ohanar>
same files as the tab 10.1
<ohanar>
(well at least in name)
<ohanar>
sounds like the galaxys2 for you then
<stickyboy>
Seems I have roughly double; two for each state.
<stickyboy>
battery_batteryerror_180.qmg and battery_batteryerror.qmg
datagutt has quit [Quit: kthxbai]
<stickyboy>
Lemme whip something up real quick...
<ohanar>
we might need to move lpm.rc out to the device trees then, since the configuration might need to be different
nebkat|beer is now known as nebkat
nebkat is now known as nebkat|beer
<stickyboy>
Hopefully not. Maybe the binary's just calling certain images?
<ohanar>
Hero_Over: if you are playing with wifi, you can try the compat-wireless 3.4 or the compat-wireless 3.3 based stuff that was in the galaxys2 update4
<ohanar>
Hero_Over: as for the gpio/sdio stuff, you can either go about trying to do it the way it is done for the origen board, or you can try to complete the placeholder functions that were in the c210.c file from update4
<Hero_Over>
i will attempt the latter first
<stickyboy>
ohanar: Yah, the P6200 playlpm binary references 6 more images than 6800's.
<ohanar>
stickyboy: yeah, we need to seperate out the props
<Hero_Over>
where can i pull the repo from update4?
<ohanar>
stickyboy: can you post the lpm.rc from the stock rom somewhere?
Turtuga_ is now known as Turtuga
noobnl has quit [Quit: noobnl]
<stickyboy>
ohanar: Ok. Lemme get it working in exynos4tab-common first.
<stickyboy>
ohanar: Yeah, I've got a bunch of stock initramfs on my github.
<Hero_Over>
hacksung?
<stickyboy>
alanorth
<stickyboy>
Hero_Over: We're not hacksung... yet. We're still groupies.
<ohanar>
Hero_Over: I'm pretty sure that there is a branch with the update4 sources
<Hero_Over>
ok ill see what i can find
<Hero_Over>
and i thought you mentioned you were maintainers?
<ohanar>
Hero_Over: samsung-ics-3.0.15 is the branch
<Hero_Over>
perfect thank you
<stickyboy>
Hero_Over: There are no maintainers. It's just us. :)
<Hero_Over>
sorry for all the ? i just havent gone deep into sgs 2 ics
<ohanar>
the hope is that we eventually get to a stable enough point that we will be an official port
<ohanar>
but we aren't that close yet :)
<Hero_Over>
i think it can happen
<Hero_Over>
and a 2 man team is almost trivial to get that goal
<Hero_Over>
thats why i wanna tackle what i can. i dont plan to maintain but to help where i can and kernels is where my heart lies
<ohanar>
hopefully you can get something going with the wifi, I'm sure that annoys you just as much as me (being on wifi only devices)
<Hero_Over>
yeah exactly
<Hero_Over>
i think also there seems to be a voltage issue... i havent dug in and debugged it but it seems the reason for low strentght
<stickyboy>
ohanar: Regarding your vibration problems... where did this immvibed binary in the exynos4tab-common props come from? Maybe its from P6200 or something?
<Hero_Over>
any comment on module voltage?
noobnl has joined #teamhacksung
<ohanar>
Hero_Over: nope, but if you want, I also have our old kernel (based on the stock sources) on my github, which may be of help (specifically for mach-px.c)
<Hero_Over>
yeah that would be great reference
<ohanar>
stickyboy: don't remember anymore :p
<Hero_Over>
unfamiliar with this atheros. used to broadcom
<stickyboy>
ohanar: Just a hunch, seeing as some of these binaries actually are device specific...?
<ohanar>
well, the lpm stuff we now know is
<ohanar>
Hero_Over: the android_kernel_samsung_c210 is the old kernel
<ohanar>
Hero_Over: although I was able to get to more or less the same bad signal situation with that kernel by using compat-wireless
<ohanar>
stickyboy: good news, looks like the lpm.rc can still be shared
<ohanar>
we just need different props
<Hero_Over>
sorry had an executive meeting. i love my job lol
<stickyboy>
ohanar: w00t
<Hero_Over>
so the issue is definitely in the driver
<ohanar>
Hero_Over: yeah, although I'm pretty sure it is related to the sdio/gpio stuff (not possitive though)
<ohanar>
stickyboy: also, I've figured out how to do lpm the proper way, so I'm going to revert my hacks and use the correct way instead (just verifying they work right now)
<stickyboy>
Ok.
<Hero_Over>
be back in 20. (work)
<stickyboy>
ohanar: Just finished modding the makefile for lpm... flashing test build now.
<ohanar>
stickboy: success?
<stickyboy>
ohanar: Fail!
<stickyboy>
Now it just pauses at the initial, centered battery with the clock thingy.
<stickyboy>
But there's only one hahaa ;)
<ohanar>
this is with the props from your stock?
<stickyboy>
Yah. I only pushed a new playlpm, maybe I need all the others too.
<stickyboy>
Like libQmageDecoder.so
<ohanar>
yeah, i'm pretty sure you want all the lpm stuff to match
<ohanar>
ok, i just pushed the revert + correct lpm changeset
sbrissen is now known as sbrissen_gone
<stickyboy>
ohanar: You might wanna check that immvibed. The md5 matches the one from my stock ROM.
<ohanar>
ok, i'm going to remove the lpm stuff from the common repo
<stickyboy>
ok
<stickyboy>
Hmm, the immvibed and charging_mode stuff seems to match the ones in the common props.
<stickyboy>
Not sure what I'm missing.
noobnl has quit [Quit: noobnl]
<ohanar>
stickyboy: look at the galaxys2 props, it sounds like your lpm is more similar to theirs than to mine (which is very similar to the galaxy tab 10.1)
noobnl has joined #teamhacksung
noobnl has quit [Quit: noobnl]
<ohanar>
stickyboy: you should update the p62x0 vendor trees with the lpm stuff
<stickyboy>
ohanar: Will do once I solve this stupid problem heeh
<stickyboy>
Might be my fault... I only see 10 .qmg files in my zip.
<stickyboy>
I'm trying another build.
<ohanar>
ok, well everything is working dandy on the tab7.7, so I can't really help you :-/
<stickyboy>
ehehe yeah
<stickyboy>
Well I have vibration, so neener neener.
<stickyboy>
Hmm, looks like I need to clobber.
<stickyboy>
Not sure what I have to delete in out/ to get it to stop using some old shit.
<ohanar>
well probably just the prop files
<ohanar>
since you don't need to link to any of the lpm files
<stickyboy>
Does it cache the make files in out?
<stickyboy>
I deleted all of system/{media,lib,bin} in out and it was still only getting half of my qmg files.
<stickyboy>
I just clobbered, so we'll see in a few minutes hhaha
<stickyboy>
ohanar: Hah, yeah... I had a syntax error in my makefile and I didn't realize it til I clobbered; it apparently does cache those make files...
<stickyboy>
This next build should be much better.
<stickyboy>
ohanar: Fuck, this auto dependencies shit is annoying.
<stickyboy>
I think it just went and pulled your changes. Now I'm getting build errors about busybox in my ramdisk.
<ohanar>
stickyboy: everything i'm building on is pushed up to my github
<ohanar>
(well in regards to lpm stuff that is)
stickyboy is now known as sticky|away
sticky|away is now known as stickyboy
<Entropy512>
<stickyboy> Entropy512: But yeah, that's the problem you guys have right? - yup, mic no worky in speakerphone mode
<hampsterblade>
woo for fixing bluetooth
<hampsterblade>
Hoping this fix works
<Entropy512>
<Rebellos> Hey there! Do you think that it matters if I execute "gpio_direction_output(GPIO_CAM_VGA_nRST, 1);" instead of "gpio_direction_output(GPIO_CAM_VGA_nRST, 0); gpio_set_value(GPIO_CAM_VGA_nRST, 1); " ? - hmm good question... but at first glance, I'd think the first would set it to be an output,
<Entropy512>
the second would set it to be an input with a pullup
<Entropy512>
but I'd need to look at the code and the exynos TRM
<hampsterblade>
E/bluedroid( 2502): bt_enable: Timeout waiting for HCI device to come up this is the issue we have
Hero_Over has quit [Quit: Page closed]
stickyboy is now known as sticky|away
noobnl has joined #teamhacksung
<hampsterblade>
Dang, the number of commits are insane
<Entropy512>
heheh
<Entropy512>
and it might be in the aries kernel repo instead...
<hampsterblade>
yeah, I can't find any commits relevant to the issue
<methril>
codeworkx, in the teamhacksung webpage, it takes so long to receive the reset password e-mail?
<Rebellos>
@Entropy512 Thanks for answer. I'm afraid you are wrong. I'm still working on PV210 arch but guess it does use same GPIO driver. I'm rewriting CE147 and S5KA~~ drivers basing on Bada kernel disassembly cuz appearantly Samsung messed up something with power lines between Aries and Wave boards and even with *theoretically* same connection mapping and chipsets it just doesn't work.
Turl has joined #teamhacksung
<Turl>
hi all
<Turl>
is andrew dodd in here?
<Rebellos>
@Entropy512 you might want to have a look at lxr links I've posted + some S5~~ CPU manual to see what these functions actually do.
<Entropy512>
ok, will revisit the TRM, I haven't looked at it in way too long
<Entropy512>
fix it for one device, breaks another. :)
<Turl>
Entropy512: the original patch it's a huge kludge :)
<Turl>
why don't you guys fix it in kernel?
<Entropy512>
you mean place it in the same location in sysfs as something else that already exists?
<Entropy512>
that'll just put you back where you were before that patch
<Entropy512>
e.g. it will again behave the same as the other three dock devices
<Turl>
Entropy512: I mean, revert all these fixes and the original hack
<Turl>
Entropy512: then implement the usb_audio switch on your kernel
<Turl>
and send a status 1 when you plug a dock
<Entropy512>
at which point - all of steven's settings items are still needed
<Entropy512>
this isn't like HTC (or was it motorola) docks where the dock tells the device that there's a cable in the port
<Entropy512>
all the device knows is that it's in a dock
<Entropy512>
so there needs to be a setting to override it in the UI
<Entropy512>
and as I said before
<Turl>
Entropy512: you can have the setting still
<Turl>
and poke another sysfs to enable-disable the auto usbaudio uevent generation
<Entropy512>
making the kernel put a status 1 into the usb_audio sysfs file
<Entropy512>
will just put you in the exact same state as before the latest patch
<Turl>
Entropy512: no >.<
<Entropy512>
since the existing 3 dock types are still treated the same
<Turl>
because it'll be just on your kernel
<Turl>
the problem is, your hack watches an extra sysfs that doesn't need to be watched
<Turl>
on samsungs, it might not do a thing
<Entropy512>
so you are saying that on your device, /sys/class/switch/dock is there but unused?
<Turl>
no
<Turl>
Entropy512: what I'm suggesting is
<Entropy512>
then what is the harm in looking at it?
<Turl>
Entropy512: 1) revert everything and leave it as on AOSP
<Entropy512>
change it once in frameworks/base
<Entropy512>
instead of changing 5-10 or more device kernels
<Turl>
Entropy512: the harm is that on other phones, none of the conditions match (the name is not "dock", usually it's descriptive text like CARDOCK or DESK or MEDIA or whatever) so it falls on the default condition
<Turl>
forces a reroute
<Parth>
trollworkz has two jobs? :O
<Turl>
derps bluetooth audio, already working dock or whatever
<Entropy512>
but on other phones, "dock" doesn't exist
<Entropy512>
therefore no problem
<Turl>
it does
<Turl>
it does on most if not all of the phones with docks
<Entropy512>
you just said it didn't
<Entropy512>
<Entropy512> so you are saying that on your device, /sys/class/switch/dock is there but unused?
<Turl>
it's a standard on the framework
<Entropy512>
<Turl> no
<Turl>
Entropy512: "no, I'm not saying that"
<Entropy512>
if it's a standard on the framework, why didn't the framework look at it in any way, shape, or form before?
<Turl>
Entropy512: it does...
<Turl>
and it generates intents
<Entropy512>
where else does the framework look at /sys/class/switch/dock?