codeworkx changed the topic of #teamhacksung to: HACKSUNG IS DEAD, HAIL HACKSUNG!
<geecko> IbrahimA|znc, are u there?
<geecko> IbrahimA|znc, it's a kernel with new ION. if you flash this your phone will very likely NOT boot but i need to see what's going on in there.
<geecko> i just hope it's not crashing before i can get logcat.
<geecko> the omni server is so slow :(
<geecko> xplodwild, can you put a working build of omni rom n7000 on your server?
<geecko> hmm, 15 min left, not that bad after all
<geecko> ok i flashed the kernel on cm10.2 even though it's meant for android 4.4 :P
<geecko> beautiful crashing loop ^^
<geecko> "hwcomposer module not found" what ??
g_byers is now known as gbyers[Away]
<geecko> WHAT !! cm10.2 is booting with new ion haha!!
<geecko> let's see how the camera behaves
gbyers[Away] is now known as g_byers
<geecko> .. it works? whaaat ?
<geecko> did i flash the right kernel??
<geecko> kernel version: Tue Feb 18 00:26:51
<geecko> right one
<geecko> Entropy512, explain this shit
<Entropy512> geecko: odd... that's including your new mali????????
<geecko> Entropy512, nope only ION
<Entropy512> well, "new" as in stuff you put into omni months ago
<Entropy512> cm10.2 never transitioned to r3p2 to my knowledge
<geecko> Entropy512, oh that's because i flashed my r3p2 test kernel for cm10.2 before ;)
<geecko> so the blobs are here :)
<Entropy512> ah
<geecko> you're right, it would have never worked without that
<Entropy512> working omni for n7000 is at dl.omnirom.org
<geecko> Entropy512, i know, it was just slow.
<geecko> but i won't need it for now after all
<geecko> Entropy512, it really looks like ION is not used at all x) i don't understand
<geecko> btw i'm using the latest omnirom codebase with my raw patches to develop
<geecko> so i can build the kernel standalone this way
<Entropy512> hmm maybe it's only used internally within the kernel?
<Entropy512> or the userspace interface is backwards-compatible for the version you transitioned to (unlike msm_ion...)
<geecko> Entropy512, paulk told me that the exynos camera module is using libsecion
<geecko> the only problem is that it allocates from a pool that doesn't exists anymore, lol
<Entropy512> must be backwards compatible or has a fallback
<geecko> ION_HEAP_TYPE_EXYNOS_CONTIG has been wiped
<geecko> if the camera isn't even breaking??? what will i do ??? :P
<geecko> oh
<geecko> ION_HEAP_EXYNOS_CONTIG_MASK != ION_HEAP_TYPE_EXYNOS_CONTIG
<geecko> or maybe it's just another name for the same value?
<geecko> is the camera module built with the EXYNOS_ION flag?
<geecko> if not, ion isn't used at all, it seems
<Entropy512> hmm
<Entropy512> adodd@susquehanna:~/android_work/omni44/device/samsung/galaxys2-common/camera$ grep EXYNOS_ION *
<Entropy512> adodd@susquehanna:~/android_work/omni44/device/samsung/galaxys2-common/camera$
<geecko> Entropy512, haha :P
<Entropy512> gs2 camera doesn't seem to use ion at all
<geecko> that's why it's not derping.
<geecko> nice.
<Entropy512> 4412 camera does
<geecko> i don't care about 4412 hahahahah /P
<geecko> :D
<geecko> Entropy512, so next step : disable UMP, replace binaries to r3p2-01rel0 with dma-buf, eventually revert changes in the Mali kernel driver between rel3 and rel0
<geecko> it _will_ break
<Entropy512> heh. :)
<geecko> Entropy512, at first i will only manually change the API version in the header and see what happens :P too lazy
<geecko> to avoid api mismatch
<Entropy512> heh
<Entropy512> ok afk a bit
<geecko> Entropy512, it's really late for me, i should be already sleeping, but i can't :P
<Entropy512> i know that feeling
IbrahimA|znc is now known as IbrahimA
<IbrahimA> Entropy512: cm11 switched, but probably not 10.2 yeah
<IbrahimA> geecko: could i rebuild with a different defconfig for my device so that it has a higher chance of actually doing something? lol
<IbrahimA> or is it too early for it to actually do anything, but you just want debug output
<geecko> IbrahimA, i'm pushing the patch to gerrit
<IbrahimA> omni gerrit?
<IbrahimA> neat
<geecko> IbrahimA, actually it works...
<IbrahimA> oh ok
<geecko> IbrahimA, but it doesn't improve anything at the moment. it's just running.
<IbrahimA> hmm probably should not be testing anything this week, got a tough exam to study for
<geecko> ok
<IbrahimA> but maybe on friday if you need testing on other devices i'll be free
IbrahimA is now known as IbrahimA|znc
<geecko> Entropy512, xplodwild, IbrahimA|znc, https://gerrit.omnirom.org/#/c/5663/ did my duty for today, now going to sleep! :)
geecko has quit [Ping timeout: 245 seconds]
arcus has quit [Disconnected by services]
arcus_ has joined #teamhacksung
arcus_ is now known as arcus
IbrahimA|znc is now known as IbrahimA
IbrahimA is now known as IbrahimA|znc
<RudyValencia> does the modem firmware on Galaxy S3 control bluetooth as well?
IbrahimA|znc is now known as IbrahimA
IbrahimA is now known as IbrahimA|znc
hmmwhatsthisdo has quit [Ping timeout: 248 seconds]
kzard has joined #teamhacksung
evcz has quit [Read error: Connection reset by peer]
evcz has joined #teamhacksung
evcz has quit [Read error: Operation timed out]
evcz has joined #teamhacksung
geecko has joined #teamhacksung
<geecko> Entropy512, rz2k from #linux-sunxi just told me that the Insignal blobs should run just by changing the API version in the kernel driver, which i did
<geecko> Entropy512, i get zero errors in logcat, but i also see that the graphics stack isn't initializing at all, which is weird
<geecko> so it's a kernel with new ION + UMP disabled + Mali API version 19 and corresponding blobs
<geecko> display not working at all isn't weird, it's expected, but i wanted to see a bit of logging :P
<geecko> craaap that's because i removed the wrong blobs
<geecko> lol
<geecko> i removed libEGL.so, libGLESv1_CM.so and libGLESv2.so from /system/lib
<geecko> i wanted to push libEGL.so, libGLESv1_CM_mali.so and libGLESv2_mali.so
<geecko> but they belong in /system/lib/egl
<geecko> logcat, rel0 dma-buf blobs with UMP left activated http://pastebin.com/xaesur5q
<geecko> without UMP : http://pastebin.com/qe6ngYe4
<geecko> looks good to me. it's crashing correctly, haha
Turtuga has quit [Ping timeout: 252 seconds]
evcz has quit [Read error: Connection reset by peer]
evcz has joined #teamhacksung
evcz has quit [Read error: Connection reset by peer]
<geecko> ok i pushed the proprietary gralloc.so from insignal
<geecko> it's complaining about libion.so missing
<geecko> kthanks
<geecko> downloading the factory image..
evcz has joined #teamhacksung
jok_r has joined #teamhacksung
jok_r has quit [Remote host closed the connection]
jok_r has joined #teamhacksung
<geecko> when renaming libsecion.so to libion.so :
<geecko> cannot locate symbol "ion_sync_fd" referenced by "gralloc.exynos4.so"
<geecko> ok it's just a missing function from libsecion.so
<geecko> it should be easy to implement
<geecko> ok i pushed libion.so from manta :P
<geecko> (old) hwcomposer is doing some weird shit, i will try to remove it and see
<geecko> I/[Gralloc]( 1877): using (fd=16)
<geecko> I/[Gralloc]( 1877): id = s3cfb
<geecko> I/[Gralloc]( 1877): xres_virtual = 800 px
<geecko> I/[Gralloc]( 1877): xres = 800 px
<geecko> I/[Gralloc]( 1877): yres = 1280 px
<geecko> I/[Gralloc]( 1877): yres_virtual = 3840 px
<geecko> I/[Gralloc]( 1877): bpp = 32
<geecko> I/[Gralloc]( 1877): r = 16:8
<geecko> I/[Gralloc]( 1877): g = 8:8
<geecko> I/[Gralloc]( 1877): b = 0:8
<geecko> I/[Gralloc]( 1877): width = 64 mm (317.500000 dpi)
<geecko> I/[Gralloc]( 1877): height = 106 mm (306.716980 dpi)
<geecko> I/[Gralloc]( 1877): refresh rate = 57.20 Hz
<geecko> E/[Gralloc-ERROR]( 1877): int init_frame_buffer_locked(private_module_t*):356 Error mapping the framebuffer (Invalid argument)
<geecko> E/SurfaceFlinger( 1877): hwcomposer module not found
<geecko> E/SurfaceFlinger( 1877): ERROR: failed to open framebuffer (Invalid argument), aborting
<geecko> it's proprietary gralloc from Insignal and libion from manta
<geecko> Entropy512.
<geecko> i guess it shouldn't happen
jok_r has quit [Quit: Ex-Chat]
<geecko> hmmm maybe it's because s3cfb isn't updated yet
<geecko> that's a lot of incertitudes :(
evcz has quit [Ping timeout: 248 seconds]
<Entropy512> geecko: looks like an s3cfb incompatibility
<geecko> Entropy512, i hope so. that's the last thing i have to update
evcz has joined #teamhacksung
plastikman has joined #teamhacksung
<geecko> Entropy512, yep you're right :)
<geecko> i begin to merge s3c-fb in a few minutes
kzard has quit [Ping timeout: 252 seconds]
Turtuga has joined #teamhacksung
evcz has quit [Ping timeout: 248 seconds]
<geecko> let's merge this shit :D
IbrahimA|znc is now known as IbrahimA
<IbrahimA> eeeeenteresting
IbrahimA is now known as IbrahimA|znc
davi has joined #teamhacksung
<geecko> linking errors
<geecko> arch/arm/mach-exynos/built-in.o:sec_thermistor.c:function sec_debug_check_crash_key: error: undefined reference to 'read_lcd_register'
<geecko> drivers/built-in.o:bmp180.c:function s3cfb_set_clock: error: undefined reference to 'to_fb_plat'
<geecko> drivers/built-in.o:bmp180.c:function s3cfb_channel_localpath_off: error: undefined reference to 'to_fb_plat'
<geecko> drivers/built-in.o:bmp180.c:function s3cfb_channel_localpath_on: error: undefined reference to 'to_fb_plat'
<geecko> drivers/built-in.o:bmp180.c:function s3cfb_window_on: error: undefined reference to 'to_fb_plat'
<geecko> drivers/built-in.o:s3cfb_s6e8aa0.c:function s6e8ax0_probe: error: undefined reference to 'lcd_early_suspend'
<geecko> drivers/built-in.o:s3cfb_s6e8aa0.c:function s6e8ax0_probe: error: undefined reference to 'lcd_late_resume'
<geecko> drivers/built-in.o:s5p-dsim.c:function s5p_dsim_isr: error: undefined reference to 's3cfb_vsync_status_check'
<geecko> drivers/built-in.o:fimc_output.c:function fimc_qbuf_output_dma_auto: error: undefined reference to 's3cfb_direct_ioctl'
<geecko> drivers/built-in.o:fimc_output.c:function fimc_qbuf_output_dma_auto: error: undefined reference to 's3cfb_direct_ioctl'
<geecko> drivers/built-in.o:fimc_output.c:function fimc_qbuf_output_dma_auto: error: undefined reference to 's3cfb_direct_ioctl'
<geecko> drivers/built-in.o:fimc_output.c:function fimc_qbuf_output_dma_auto: error: undefined reference to 's3cfb_direct_ioctl'
<geecko> drivers/built-in.o:fimc_dev_u1.c:function fimc_open: error: undefined reference to 's3cfb_open_fifo'
<geecko> drivers/built-in.o:fimc_dev_u1.c:function fimc_open: error: undefined reference to 's3cfb_close_fifo'
<geecko> woops.. next time i'll use pastebin :P
kossy has quit [Excess Flood]
kossy has joined #teamhacksung
<geecko> wow, it should be building. i commented a lot of stuff that is miscellaneous regarding my mental health
<geecko> nope, not building :(
arcus has quit [Disconnected by services]
arcus_ has joined #teamhacksung
arcus_ is now known as arcus
davi has quit [Ping timeout: 245 seconds]
plastikman has quit [Ping timeout: 260 seconds]
evcz has joined #teamhacksung
paulk-collins has joined #teamhacksung
plastikman has joined #teamhacksung
plastikman has quit [Ping timeout: 240 seconds]
evcz has quit [Quit: evcz]
plastikman has joined #teamhacksung
evcz has joined #teamhacksung
plastikman has quit [Ping timeout: 272 seconds]
plastikman has joined #teamhacksung
evcz has quit [Ping timeout: 264 seconds]
plastikman has quit [Client Quit]
IbrahimA|znc is now known as IbrahimA
<IbrahimA> geecko: lolz, so when you do this stuff, do you basically bring in updated files, then try to make it compile?
<IbrahimA> or do you actually read through all the code, understand it, then try to backport changes bit by bit?
<IbrahimA> it seems like a pain in the ass to do the second
evcz has joined #teamhacksung
jamestunnicliffe has quit [Remote host closed the connection]
jamestunnicliffe has joined #teamhacksung
jamestunnicliffe has quit [Changing host]
jamestunnicliffe has joined #teamhacksung
Cristian_Silaghi has joined #teamhacksung
Cristian_Silaghi is now known as sn0w
sn0w is now known as sn0w14
paulk-collins has quit [Remote host closed the connection]
plastikman has joined #teamhacksung
kzard has joined #teamhacksung
IbrahimA is now known as IbrahimA|znc
<geecko> IbrahimA|znc, 1st method of course
evcz has quit [Ping timeout: 265 seconds]
kzard has quit [Ping timeout: 252 seconds]
plastikman has quit [Read error: Operation timed out]
sn0w14 has quit [*.net *.split]
jamestunnicliffe has quit [*.net *.split]
jamestunnicliffe has joined #teamhacksung
sn0w14 has joined #teamhacksung
<sn0w14> geecko any news? :) I really appreciate your hard work sir. :)
<geecko> sn0w14, workin on it
<geecko> thanks btw
<sn0w14> you can post here some progress or anything you consider important. :)
<sn0w14> Many important info were written here by you. And I like this. :D
<geecko> sn0w14, yeah i know, it's really difficult to get started, i had to document a lot by myself
<geecko> and i still think that having a central place (e.g. a wiki) with all our findings would be awesome
<geecko> but it takes time, and we're lazy...
<sn0w14> Yes... I know that feel bro :)
<geecko> i'm merging the s3c-fb driver right now
<geecko> trying to get this shit compiling
<geecko> i have the feeling that it will break a lot of stuff
<sn0w14> You will make it, I feel this :D
<geecko> i hope that as much as you, because the note is my only phone
<sn0w14> And my only phone is GT-i9100 :D
<geecko> well the screen is smaller so it's not as painful as the note :P
<sn0w14> I don't understand something. If SAMSUNG released newer r3p2 driver in i9300-Update12, why there's older driver in Insignal? Weird :P
<sn0w14> Yeah, I know that. :)
<geecko> sn0w14, insignal drop is from months ago
<geecko> +6months
<sn0w14> Ok. :)
<geecko> we have to use these blobs with the big update because sammy's blobs are allocating memory using UMP, while Insignal's blobs are allocating memory using dma-buf which is required by the new hwc, fb driver
<geecko> and also ion
<geecko> we have to update everything at once, because it 100% incompatible with what we have right now
<geecko> sn0w14, take a look at http://git.insignal.co.kr
<sn0w14> Yes, I know the reason why you use r3p2-01rel0 :D I read last night :)
<geecko> that's all the userspace stuff (including hwc)
<sn0w14> And if hwc is gonna work, Project Butter will become reality for our devices? :)
<geecko> sn0w14, yes.
<geecko> sn0w14, to me Project Butter only means "actually working" :P
<sn0w14> geecko, yeah... :P
arcus has quit [Disconnected by services]
arcus_ has joined #teamhacksung
arcus_ is now known as arcus
<geecko> sn0w14, right now we have graphical glitches and inconsistent framerate
<sn0w14> Future looks bright :D
<geecko> sn0w14, actually merging s3c-fb isn't as easy as I thought...
<geecko> but i'm doing my best, i just need some time.
IbrahimA|znc is now known as IbrahimA
<geecko> Entropy512, compilation errors :P http://pastebin.com/9Q9aZKb4
<sn0w14> geecko, there is always rom for improvement, right? Take it easy. :)
<Entropy512> fun
<Entropy512> i'm still trying to figure out how to backport fuse without breaking shit
<IbrahimA> so geecko right now are you updating stuff from update 12 or insignal?
<geecko> Entropy512, i already solved these errors before, but i lost my work, it was hosted on vpn :P
<Entropy512> :(
jamestunnicliffe has quit [Remote host closed the connection]
<geecko> Entropy512, not a big deal
<geecko> Entropy512, it's mostly header updates
<geecko> Entropy512, but there's a lot of shit breaking in s3cfb _panel_ drivers
<Entropy512> crap, just what I was afraid of
<geecko> everything is tightly coupled, FUCK YOU samsung :D
<geecko> pure mess
<geecko> the i9300 kernel is just a pile of crap updated since 4+ years
jamestunnicliffe has joined #teamhacksung
<geecko> Entropy512, i mean... there's THREE mali drivers in update12
<Entropy512> yup
<Entropy512> typical samsung
<IbrahimA> lol
<geecko> Entropy512, r2p4 and r3p1 in /drivers/media/video/samsung, r3p2 in /drivers/gpu/mali400. why ? why ??
<geecko> there's also some bits of exynos 5 support, panel drivers for unknown boards, etc etc
Fire-Balls has joined #teamhacksung
<geecko> IbrahimA, that's actually sad
<IbrahimA> they probably just have a poor release/merge process
<Entropy512> look at superbrick
<IbrahimA> lol yea
<geecko> IbrahimA, they have so many devices to maintain, they can't work properly
<Entropy512> they had no clue which devices had MMC_CAP_ERASE enabled or disabled or why
<Fire-Balls> is the newly anounced kitkat update for the s3 and note 2 going to be of any help?
<geecko> Fire-Balls, that's a mystery for now.
IbrahimA is now known as IbrahimA|znc
<geecko> Fire-Balls, 1% chances it will bring a 3.4 kernel? :D
<geecko> 3.4 kernel = all they legacy crap removed
<geecko> that's terribly unlikely
<Fire-Balls> Who do we need to bribe to make that happen?
<geecko> their*
<geecko> Fire-Balls, no one, just wait
<Fire-Balls> we will see when it gets here.
<geecko> Fire-Balls, don't expect too much
<geecko> Fire-Balls, 4.3 brought us mali r3p2 already
<geecko> Fire-Balls, but if my work is giving results, their blobs will be 100% useless because they all use UMP for memory allocation, unlike Insignal one
IbrahimA|znc is now known as IbrahimA
<IbrahimA> im kinda surprised they mentioned the note 10.1
<IbrahimA> sorta figured they abandoned it after 4.1
<Fire-Balls> Maybe updated Bluetooth stuff so my headset volume controls actually work.
<IbrahimA> Entropy512: uhh... im confused, but it looks to me like n80xx-common is in omni.dependencies twice
<IbrahimA> i think you said it was due to gerrit showing things weird when you compare from a non-base revision, but
<IbrahimA> looking at github it's definitely there twice
<Entropy512> ooooh, I think we had a gerrit derp
<Entropy512> your last patch somehow duped things
<Entropy512> split out the removal into a separate patch
<IbrahimA> oh, the original n801x rename?
<IbrahimA> yeah, i guess so
<IbrahimA> sure thing
<Entropy512> yeah