| 00:02 |
jonpry |
gzip settings? |
| 00:05 |
WoZZeR |
I did gzip -9 |
| 00:09 |
ali1234 |
strip all the binaries |
| 00:09 |
WoZZeR |
doesn't it need it? |
| 00:09 |
ali1234 |
yes |
| 00:10 |
[Delenn] |
Markinus, it still needs calibration |
| 00:10 |
[Delenn] |
however |
| 00:12 |
gauner1986 |
nice to see the port going on |
| 00:12 |
gauner1986 |
if only my gf wouldn't keep me away ^^ |
| 00:13 |
[Delenn] |
heh |
| 00:13 |
[Delenn] |
learn she and look how she ports it :P |
| 00:15 |
gauner1986 |
couldnt think of anything more sexy |
| 00:15 |
gauner1986 |
:P |
| 00:15 |
WoZZeR |
jonpry: working on keyboard stuff now |
| 00:17 |
[Delenn] |
:D |
| 00:17 |
[Delenn] |
yeah fuck her while she porting :D |
| 00:17 |
gauner1986 |
i wonder how port quality would suffer from that :D |
| 00:18 |
[Delenn] |
:D |
| 00:18 |
[Delenn] |
linux fuck brain a lot however |
| 00:18 |
gauner1986 |
true |
| 00:18 |
gauner1986 |
but i guess ce does too |
| 00:18 |
gauner1986 |
if you're not used to it |
| 00:19 |
[Delenn] |
even linux style assume this |
| 00:19 |
[Delenn] |
*ever |
| 00:19 |
[Delenn] |
ce... maybe... for me it's rather logical |
| 00:19 |
[Delenn] |
at least there a lot of stuffs are binary compatible |
| 00:20 |
gauner1986 |
yeah |
| 00:20 |
gauner1986 |
bad thing about linux is that drivers become incompatible on changing minor kernel revisions |
| 00:20 |
gauner1986 |
they even become incompatible if you switch compiler |
| 00:21 |
[Delenn] |
))) |
| 00:21 |
[Delenn] |
yes |
| 00:23 |
LeTama |
gauner1986: re! too bad for ppp |
| 00:23 |
gauner1986 |
re letama :) |
| 00:23 |
gauner1986 |
maybe the new kernel does better with ppp :P |
| 00:23 |
LeTama |
gauner1986: it seems a little more stable with evo |
| 00:23 |
LeTama |
gauner1986: I saw one failure though |
| 00:24 |
LeTama |
gauner1986: but it didn't crash modem |
| 00:24 |
gauner1986 |
thats good |
| 00:24 |
gauner1986 |
so you were able to establish the connection again without rebooting? |
| 00:24 |
LeTama |
gauner1986: yes |
| 00:25 |
LeTama |
gauner1986: just unchecked/re-checked data and it went good |
| 00:25 |
gauner1986 |
sounds fine |
| 00:25 |
LeTama |
the other kernel seems to have a memory leak |
| 00:25 |
gauner1986 |
hum? |
| 00:25 |
gauner1986 |
where? |
| 00:26 |
gauner1986 |
i'm just compiling it |
| 00:26 |
LeTama |
not 100% sure, but I got some allocation errors on smd tty |
| 00:26 |
gauner1986 |
hm |
| 00:26 |
LeTama |
could be because of some debug flags in kernel or could be real |
| 00:27 |
LeTama |
that's why I switched to evo as smd is a little more evolved |
| 00:27 |
gauner1986 |
what do you mean by other kernel? |
| 00:27 |
gauner1986 |
i want to switch too now |
| 00:27 |
LeTama |
linuxtogo kernel |
| 00:27 |
gauner1986 |
ok |
| 00:27 |
gauner1986 |
^^ |
| 00:28 |
gauner1986 |
markinus: are you currently porting cotullas stuff? |
| 00:29 |
LeTama |
gauner1986: it's very functional now |
| 00:29 |
gauner1986 |
i believe that a lot of stability comes with that kernel then |
| 00:30 |
LeTama |
gauner1986: could be |
| 00:31 |
WoZZeR |
ohhh, it's the .git folder |
| 00:31 |
[Delenn] |
lol |
| 00:32 |
gauner1986 |
hm |
| 00:32 |
gauner1986 |
i got bluescreen |
| 00:32 |
gauner1986 |
do i have to change default.txt? |
| 00:32 |
LeTama |
gauner1986: ah, yes, you need usb plugged |
| 00:32 |
gauner1986 |
usb is plugged |
| 00:32 |
LeTama |
gauner1986: try again, it happens from time to time |
| 00:33 |
LeTama |
gauner1986: it misses one of the fix Cotulla made later |
| 00:33 |
gauner1986 |
hm |
| 00:33 |
Hunterkll |
harddrives aren't supposed to shoot out fire and molten silicon... eh? |
| 00:33 |
Hunterkll |
:o |
| 00:33 |
LeTama |
gauner1986: got to go, ttyl |
| 00:34 |
[Delenn] |
Hunterkll, how hand? |
| 00:34 |
Hunterkll |
fix'd |
| 00:34 |
[Delenn] |
:) |
| 00:34 |
gauner1986 |
letama: ok, bye |
| 00:34 |
[Delenn] |
bb |
| 00:34 |
gauner1986 |
lol |
| 00:34 |
gauner1986 |
even netrippers flash on kernel start is in |
| 00:34 |
gauner1986 |
:P |
| 00:34 |
Hunterkll |
i picked it out of the case.. was like |
| 00:34 |
Hunterkll |
'why aren't you spinning' |
| 00:34 |
Hunterkll |
so i shook it |
| 00:34 |
Hunterkll |
and suddenly fire :D |
| 00:35 |
gauner1986 |
[delenn]: what fix did you make against the bluescreens? :) |
| 00:36 |
[Delenn] |
hm... with evo driver it should work 100%\ |
| 00:37 |
gauner1986 |
i got instant bluescreen after haret takeover.. |
| 00:37 |
gauner1986 |
but too late to take a log now |
| 00:37 |
gauner1986 |
will grab one later |
| 00:37 |
[Delenn] |
GSOD? |
| 00:37 |
[Delenn] |
:D |
| 00:38 |
[Delenn] |
thinks gauner still gsodder#1 |
| 00:38 |
gauner1986 |
i am |
| 00:38 |
gauner1986 |
:) |
| 00:38 |
gauner1986 |
but i'm also gsodder #1 |
| 00:39 |
gauner1986 |
eh |
| 00:39 |
gauner1986 |
bsodder |
| 00:39 |
gauner1986 |
:) |
| 00:39 |
Hunterkll |
i have to scandisk |
| 00:39 |
Hunterkll |
before every android boot |
| 00:39 |
Hunterkll |
to prevent sod :o |
| 00:39 |
[Delenn] |
maybe Markinus didn't include time fix? |
| 00:39 |
gauner1986 |
markinus: ok.. leds don't work correctly for me |
| 00:39 |
gauner1986 |
time fix? |
| 00:39 |
[Delenn] |
timer |
| 00:39 |
gauner1986 |
ah |
| 00:39 |
gauner1986 |
the speedy issue |
| 00:40 |
gauner1986 |
i hope he did |
| 00:40 |
gauner1986 |
^^ |
| 00:40 |
[Delenn] |
yes speeddy (tm) |
| 00:40 |
gauner1986 |
markinus: button backlight goes on for me if the display switches off Oo |
| 00:40 |
gauner1986 |
but not if i press the buttons |
| 00:40 |
[Delenn] |
http://www.gitorious.com/linux-on-wince-htc/linux-on-wince-htc/blobs/master/arch/arm/mach-msm/timer.c |
| 00:40 |
gauner1986 |
really strange |
| 00:40 |
[Delenn] |
yes no fix :) |
| 00:40 |
gauner1986 |
ah |
| 00:40 |
gauner1986 |
ok |
| 00:40 |
gauner1986 |
then i'll just copy the file from the other repo |
| 00:40 |
gauner1986 |
:D |
| 00:42 |
Markinus |
gauner1986: the display is going dark, -> leds on-> after some seconds -> both off, it's userspace |
| 00:42 |
jonpry |
wozzer: how is it going? |
| 00:42 |
[Delenn] |
Markinus, why u exclude speedy(tm) fix>\ |
| 00:42 |
WoZZeR |
jonpry: not too good |
| 00:42 |
gauner1986 |
Markinus: hm. it did work perfectly with the linuxtogo kernel |
| 00:42 |
[Delenn] |
? |
| 00:42 |
WoZZeR |
if I bind something to /rfs/temp, and switch root to /rfs |
| 00:42 |
WoZZeR |
will the bind follow? |
| 00:43 |
jonpry |
its supposed to |
| 00:43 |
WoZZeR |
-/rfs/tmp* |
| 00:43 |
WoZZeR |
ok, I'll test this then |
| 00:43 |
Markinus |
[Delenn]: I think I overseen it . . |
| 00:43 |
Hunterkll |
WoZZeR, that's how you install gentoo! :P |
| 00:43 |
Hunterkll |
mount -o bind /dev /mnt/gentoo/dev |
| 00:43 |
[Delenn] |
http://www.gitorious.com/linux-on-wince-htc/linux-on-wince-htc/blobs/master/arch/arm/mach-msm/timer.c |
| 00:43 |
Markinus |
[Delenn]: I will integrate it |
| 00:43 |
Hunterkll |
chroot /mnt/gentoo /bin/bash |
| 00:43 |
Hunterkll |
:P |
| 00:43 |
[Delenn] |
static void __init msm_timer_init(void) |
| 00:43 |
[Delenn] |
start of it |
| 00:44 |
WoZZeR |
cool, let's see if this works |
| 00:44 |
WoZZeR |
jonpry: http://pastebin.com/Ch67dsBN |
| 00:44 |
WoZZeR |
keyboard code so far |
| 00:45 |
gauner1986 |
markinus: cant be userspace then? :) |
| 00:45 |
gauner1986 |
markinus: can you supply your liblights ? |
| 00:46 |
WoZZeR |
damn, using old keymap again |
| 00:46 |
WoZZeR |
can commands be run after exec switch_root /rfs /init |
| 00:46 |
WoZZeR |
or does that run the new init? |
| 00:47 |
Markinus |
gauner1986: I using the FroyoStone_Sense_V2 |
| 00:47 |
gauner1986 |
hm |
| 00:47 |
trav |
hey =D |
| 00:47 |
Markinus |
gauner1986: the kernel didn'T have such code . . |
| 00:48 |
gauner1986 |
markinus: hum? what code? |
| 00:51 |
Markinus |
gauner1986: what you mean what happend to the led |
| 00:52 |
gauner1986 |
must be userspace somehow yeah |
| 00:52 |
gauner1986 |
but why didnt it do that before? |
| 00:52 |
gauner1986 |
Oo |
| 00:53 |
[Delenn] |
it's so funny vibro-debugging :( |
| 00:53 |
gauner1986 |
lol |
| 00:53 |
gauner1986 |
nand? |
| 00:53 |
[Delenn] |
no |
| 00:53 |
[Delenn] |
1 or 0 always |
| 00:53 |
[Delenn] |
:( |
| 00:54 |
jonpry |
wozzer: sed -i -e "s:$: physkeyboard=${physkey}:" /tmp/cmdline, should be /rfs/tmp/cmdline |
| 00:54 |
WoZZeR |
ahh, thanks |
| 00:54 |
WoZZeR |
trying now |
| 00:55 |
jonpry |
and you might want to mount to /rfs/proc/cmdline |
| 00:55 |
jonpry |
depending on what code follows |
| 00:55 |
WoZZeR |
mount -t proc proc /proc |
| 00:55 |
gauner1986 |
ok, gotta go now |
| 00:55 |
WoZZeR |
is run at the top |
| 00:55 |
gauner1986 |
see you tomorrow |
| 00:55 |
jonpry |
yeah, but there is more code that fudges cmdline right? |
| 00:55 |
gauner1986 |
gnight |
| 00:55 |
WoZZeR |
only at end of rootfs |
| 00:56 |
WoZZeR |
where I don't care if it has physkeyboard |
| 00:56 |
jonpry |
something eventually --binds the whole of proc |
| 00:58 |
WoZZeR |
./sys/class/htc_hw/machine_variant |
| 00:58 |
WoZZeR |
that should be available to initrd, right? |
| 00:59 |
WoZZeR |
said it can't find it |
| 00:59 |
WoZZeR |
worked in android |
| 00:59 |
WoZZeR |
oh |
| 00:59 |
jonpry |
you might have to mount /sys first |
| 00:59 |
WoZZeR |
sys is umounted |
| 00:59 |
WoZZeR |
duh... |
| 01:00 |
WoZZeR |
if I umount proc, does the bind still hold? |
| 01:00 |
WoZZeR |
for when it gets remounted |
| 01:07 |
jonpry |
why umount it? |
| 01:07 |
jonpry |
did you say you don't care about cmdline after some point in initrd/ |
| 01:07 |
WoZZeR |
echo "Cleaning up..." |
| 01:07 |
WoZZeR |
#umount -l /proc |
| 01:07 |
WoZZeR |
#umount -l /sys |
| 01:07 |
WoZZeR |
no, it needs to be there in rootfs |
| 01:08 |
WoZZeR |
rootfs does the keyboard stuff |
| 01:09 |
jonpry |
i'm not sure |
| 01:09 |
Jack-E |
~seen zego |
| 01:09 |
apt |
zego <~zego@122.129.199.154> was last seen on IRC in channel #htc-linux, 5d 14h 39m 34s ago, saying: 'CM6.0.0-RC3 based ROM for HD2: http://pastebin.com/0xSN3rbH (for those who care)'. |
| 01:10 |
jonpry |
wozzer: its probably best if you don't duplicate all the code that remounts cmdline |
| 01:11 |
jonpry |
just do an extra sed on the variable. then it will work |
| 01:11 |
WoZZeR |
I'm trying to get around editing the rootfs |
| 01:11 |
WoZZeR |
physkey=${physkey,,} |
| 01:11 |
jonpry |
rootfs remounts cmdline? |
| 01:11 |
WoZZeR |
why doesn't it like that |
| 01:12 |
WoZZeR |
jonpry: yeah |
| 01:12 |
WoZZeR |
to add serialnumber |
| 01:12 |
WoZZeR |
but at the end |
| 01:12 |
WoZZeR |
actuall |
| 01:12 |
WoZZeR |
./bin/echo "Rootfs loaded." |
| 01:12 |
WoZZeR |
./bin/mount -t sysfs sys /sys |
| 01:12 |
jonpry |
mount --bind /rfs/tmp/cmdline /rfs/proc/cmdline |
| 01:12 |
WoZZeR |
./bin/mount -t proc proc /proc |
| 01:13 |
WoZZeR |
physkey=${physkey,,} <- initrd does not like that |
| 01:13 |
jonpry |
./bin/mount -t sysfs sysfs /sys |
| 01:13 |
jonpry |
what does that even do? |
| 01:13 |
WoZZeR |
lowercase the RHOD50000 |
| 01:14 |
jonpry |
where did you find that little gem |
| 01:14 |
Jack-E |
~seem cotulla |
| 01:14 |
WoZZeR |
google |
| 01:14 |
Jack-E |
~seen cotulla |
| 01:14 |
apt |
cotulla <~opera@nat004-252-205-109.tvoe.tv> was last seen on IRC in channel #htc-linux, 3d 11h 47m 14s ago, saying: 'it's my time :D'. |
| 01:14 |
WoZZeR |
works on ubuntu |
| 01:15 |
jonpry |
is this even a bash script? |
| 01:15 |
WoZZeR |
init is a bash script |
| 01:15 |
WoZZeR |
I believe |
| 01:15 |
WoZZeR |
everything else works like bash |
| 01:16 |
jonpry |
could be sh |
| 01:16 |
WoZZeR |
sed -e 'y/RHOD/rhod/' |
| 01:16 |
WoZZeR |
use that instead? |
| 01:16 |
jonpry |
you could cat /sys/mach_var | tr 'A-Za-z' |
| 01:17 |
WoZZeR |
physkey=$(cat /sys/class/htc_hw/machine_variant | tr'A-Za-z') |
| 01:17 |
WoZZeR |
like that? |
| 01:18 |
jonpry |
actually tr 'A-Z' 'a-z' |
| 01:18 |
jonpry |
physkey=$(cat /sys/class/htc_hw/machine_variant | tr 'A-Z' 'a-z') |
| 01:19 |
WoZZeR |
ok, what about only using 7 chars? |
| 01:21 |
jonpry |
maybe that line of yours works |
| 01:21 |
WoZZeR |
rfs/proc/cmdline doesn't exist |
| 01:22 |
WoZZeR |
does it need a / atht estart? |
| 01:22 |
jonpry |
it might |
| 01:22 |
jonpry |
also need to make sure rfs is mounted |
| 01:22 |
WoZZeR |
it is |
| 01:22 |
WoZZeR |
mount -t ext2 -o noatime,nodiratime,sync,ro /dev/block/loop2 /rfs |
| 01:23 |
WoZZeR |
or mount -t squashfs -o noatime,nodiratime,sync,ro /dev/block/loop2 /rfs |
| 01:23 |
WoZZeR |
one of those is run |
| 01:24 |
jonpry |
you might have to go ahead and mount proc for it too. like mount --bind /proc /rfs/proc |
| 01:24 |
stinebd |
what are you doing? |
| 01:24 |
jonpry |
when rootfs tries to mount the already mounted system, it will be fine |
| 01:25 |
WoZZeR |
trying to add physkeyboard to cmdline |
| 01:25 |
jonpry |
just don't umount /proc |
| 01:25 |
WoZZeR |
so I don't have to edit rootfs |
| 01:25 |
stinebd |
those are in rootfs |
| 01:25 |
WoZZeR |
I know |
| 01:25 |
WoZZeR |
I'm trying to add physkeyboard to cmdline in intrd |
| 01:26 |
stinebd |
the only thing that uses physkeyboard is our rootfs |
| 01:26 |
WoZZeR |
I know |
| 01:26 |
stinebd |
it's not a kernel parameter or anything |
| 01:26 |
WoZZeR |
I want 1 nbh for the nand boot |
| 01:26 |
WoZZeR |
not 14 |
| 01:26 |
jonpry |
and we can detect rhodium version now |
| 01:27 |
stinebd |
it shouldn't be done in initramfs |
| 01:28 |
WoZZeR |
It should be done correctly in rootfs |
| 01:28 |
stinebd |
then why not do that? |
| 01:28 |
WoZZeR |
I'd rather not have to edit rootfs |
| 01:29 |
stinebd |
why? |
| 01:29 |
WoZZeR |
it's not included in the xip |
| 01:29 |
stinebd |
so what does the bring up? |
| 01:30 |
WoZZeR |
the xip only has zImage and initrd.gz in it |
| 01:30 |
WoZZeR |
although it's looking like it can only be done in rootfs |
| 01:30 |
WoZZeR |
if I can't get it working here, then I'll start work on including rootfs in the xip |
| 01:31 |
WoZZeR |
this is already starting to look like a dirty hack |
| 01:31 |
stinebd |
so as you have things now, what does initramfs do and then what does the live system do? |
| 01:31 |
WoZZeR |
for what? |
| 01:31 |
stinebd |
booting |
| 01:32 |
WoZZeR |
xip loads directly into zImage. zImage loads initramfs, initramfs loads rootfs |
| 01:32 |
WoZZeR |
root then loads android I believe |
| 01:32 |
stinebd |
so you don't have any changes to our stock initramfs? |
| 01:33 |
stinebd |
(right now) |
| 01:33 |
WoZZeR |
nothing other than getting keyboard to load |
| 01:33 |
WoZZeR |
it's stock |
| 01:33 |
WoZZeR |
normal initramfs is stock |
| 01:34 |
stinebd |
is there a thread for nand development that i could look through? i don't really understand how its laid out at all |
| 01:34 |
stinebd |
if i knew what the challenges were i could probably help you do what you need to |
| 01:35 |
WoZZeR |
I'm using tinboot to load zImage, instead of haret |
| 01:35 |
WoZZeR |
it's the exact same setup as the normal XDAndroid build |
| 01:35 |
WoZZeR |
except instead of haret, it launches right into the kernel |
| 01:35 |
stinebd |
well it's not the same, without nand we have the luxury of not caring about mtd |
| 01:35 |
stinebd |
we just give it loopback filesystems and it's happy |
| 01:35 |
WoZZeR |
nand right now is only dealing as a haret replacement |
| 01:36 |
jonpry |
the only trouble with nand is that there is no practical way to set boot parms. ie no kernel cmd line |
| 01:36 |
WoZZeR |
.ifdef rhod500 |
| 01:36 |
WoZZeR |
.asciz "gsensor_axis=2,1,3 msmvkeyb_toggle=off hw3d.force=0 panic=0 pm.sleep_mode=2" |
| 01:36 |
WoZZeR |
.endif |
| 01:36 |
stinebd |
jonpry: that would normally be handled by the bootloader |
| 01:36 |
WoZZeR |
we have cmdline parms |
| 01:36 |
WoZZeR |
just I need a different nbh for each cmdline |
| 01:36 |
WoZZeR |
I already fixed the data problem |
| 01:36 |
WoZZeR |
so I don't need force_cdma |
| 01:36 |
stinebd |
ok, i see what you're trying to do then |
| 01:37 |
stinebd |
so you still have all those loopbacks on the sd card? |
| 01:37 |
WoZZeR |
yeah |
| 01:37 |
WoZZeR |
for the most part, it's the same |
| 01:37 |
stinebd |
ok |
| 01:37 |
WoZZeR |
none of the image is on nand yet |
| 01:37 |
WoZZeR |
it's just a haret subsitute for now |
| 01:38 |
stinebd |
unfortunately you can't modify cmdline the way you want to |
| 01:38 |
WoZZeR |
SERIALNO=$(cat /data/serialno) |
| 01:38 |
WoZZeR |
cp /proc/cmdline /tmp/cmdline |
| 01:38 |
WoZZeR |
sed -i -e "s:$: androidboot.serialno=${SERIALNO}:" /tmp/cmdline |
| 01:38 |
WoZZeR |
mount --bind /tmp/cmdline /proc/cmdline |
| 01:38 |
WoZZeR |
I saw rootfs do that |
| 01:38 |
WoZZeR |
which I was trying to replicate |
| 01:38 |
stinebd |
yeah, that's the problem |
| 01:38 |
stinebd |
rootfs does it already |
| 01:39 |
WoZZeR |
until we have internal storage access, I was trying to leave everything as stock as possible |
| 01:39 |
WoZZeR |
unless it's included in the 'xip' |
| 01:40 |
WoZZeR |
or atleast as few files changed as possible |
| 01:40 |
stinebd |
so is /sys/class/htc_hw/machine_variant available in xdandroid's kernel? |
| 01:40 |
jonpry |
you could always patch rootfs from initramfs |
| 01:40 |
stinebd |
which lets you detect rhod version |
| 01:40 |
WoZZeR |
it's in our nand kernel |
| 01:40 |
WoZZeR |
which is a fork |
| 01:40 |
stinebd |
that's unfortunate |
| 01:41 |
WoZZeR |
well, I don't want phh to implement it until we know it works for all devices |
| 01:41 |
WoZZeR |
or if it even works on other devices |
| 01:42 |
WoZZeR |
we only have 6 files or so changed from xdandroid kernel |
| 01:42 |
WoZZeR |
most are just additional sound stuff |
| 01:42 |
WoZZeR |
http://gitorious.org/~acl/linux-on-qualcomm-s-msm/linux-msm-rhod-nand |
| 01:42 |
WoZZeR |
there's our git |
| 01:43 |
stinebd |
ok, it looks like you can hackishly bindmount a file multiple times |
| 01:43 |
WoZZeR |
stinebd: in my initramfs? |
| 01:43 |
stinebd |
so you might want to try it on a live system and see if that method works |
| 01:44 |
jonpry |
i don't think it would fail anyways. its just going to copy the contents of tmp/cmdline over itself. then try to remount, which will fail, but since it already has the right data... |
| 01:44 |
Hunterkll |
ya'll remind me of the times when i sysgen a linux system from pure source |
| 01:45 |
Hunterkll |
shady bind mounts, contorted compiler chains, all in the elusive goal of toolchain independance :o |
| 01:45 |
stinebd |
jonpry: it will successfully remount, at the stage of rootfs init where it sets the serialno |
| 01:45 |
WoZZeR |
I really only need it to get to the keyboard part |
| 01:45 |
stinebd |
at least on my laptop that's how it behaves ;) |
| 01:45 |
WoZZeR |
if it's a giant pain in the ass, I'll just edit rootfs |
| 01:45 |
stinebd |
so what you can do... |
| 01:46 |
Hunterkll |
so what exactly is this bind mount acheiving? |
| 01:46 |
WoZZeR |
I'm a nub, esp. when it comes to linux, so I'm just trying to replicate something. But I assume that it's like a synlink? |
| 01:46 |
Hunterkll |
oh, i know what it is |
| 01:46 |
Hunterkll |
i mean, what is the goal of using bind mouns? |
| 01:46 |
stinebd |
in initramfs, after rootfs is mounted, create some temp file in rootfs (so it survives chroot) with the cmdline, modify that cmdline how you want it, mount /proc under rfs, bind mount the modified cmdline under /rfs/proc, then continue booting |
| 01:47 |
Hunterkll |
... |
| 01:47 |
Hunterkll |
what? |
| 01:47 |
WoZZeR |
isn't that what we had? |
| 01:47 |
stinebd |
yes |
| 01:47 |
Hunterkll |
my head just exploded |
| 01:47 |
stinebd |
but you do need to make sure the tmp cmdline is in rootfs |
| 01:47 |
Hunterkll |
so you're doing runtime detection of a feature |
| 01:47 |
Hunterkll |
then creating a file to flag what you are using |
| 01:47 |
Hunterkll |
that you want to be alive in chroot |
| 01:48 |
WoZZeR |
http://pastebin.com/5fK73dT7 |
| 01:48 |
WoZZeR |
that's the current state |
| 01:48 |
Hunterkll |
but it has to be there before most services come up? |
| 01:48 |
stinebd |
that should probably work |
| 01:48 |
Hunterkll |
cmdline is a file, correct? |
| 01:49 |
stinebd |
well by probably i mean possibly. no warranty |
| 01:49 |
WoZZeR |
just added some bad commands to init so it crashs |
| 01:49 |
WoZZeR |
and I can see more |
| 01:49 |
Hunterkll |
why are you using mount/bind mounts to .... relocate a file? |
| 01:49 |
Hunterkll |
do symlinks not work? |
| 01:49 |
stinebd |
no they don't |
| 01:49 |
Hunterkll |
hardlinks? |
| 01:49 |
stinebd |
1. because there's a chroot, 2. because /proc is readonly |
| 01:49 |
Hunterkll |
what filesystem are you using? |
| 01:49 |
Hunterkll |
okay |
| 01:49 |
Hunterkll |
so there's a chroot |
| 01:50 |
Hunterkll |
so you want to drop a symlink to /tmp/cmdline in /proc |
| 01:50 |
stinebd |
you can't drop symlinks in /proc |
| 01:50 |
Hunterkll |
already made inside of /proc in the rootfs image |
| 01:50 |
Hunterkll |
.... |
| 01:50 |
Hunterkll |
you are wheel. you can do anything. ;) |
| 01:50 |
stinebd |
the filesystem itself is read only |
| 01:50 |
Hunterkll |
(hint: make it before proc gets mounted and hope procfs dosn't trash it? :P) |
| 01:50 |
ali1234 |
wat |
| 01:50 |
Hunterkll |
that *sometimes* worked back in 2.4 |
| 01:50 |
Hunterkll |
:) |
| 01:51 |
stinebd |
yeah nothing under a mountpoint is accessible after something is mounted over it |
| 01:51 |
ali1234 |
/proc/cmdline? |
| 01:51 |
Hunterkll |
As it should be, but ... :P |
| 01:51 |
stinebd |
unless of course you remount it elsewhere |
| 01:51 |
Hunterkll |
ali1234, I dunno. but i'm trying to figure it for 'em :P |
| 01:51 |
stinebd |
and then we get into an infinite loop |
| 01:51 |
ali1234 |
you know it's read only yeah? |
| 01:51 |
Hunterkll |
stinebd, dosn't /proc/cmdline pull from the kernel??? |
| 01:51 |
stinebd |
yes |
| 01:52 |
stinebd |
which is the problem. they want to modify it |
| 01:52 |
ali1234 |
so change it on bootloader |
| 01:52 |
Hunterkll |
.... |
| 01:52 |
stinebd |
bootloader is static |
| 01:52 |
Hunterkll |
stinebd, this is wrong. |
| 01:52 |
ali1234 |
or recompile the kernel with a different built in command line |
| 01:52 |
Hunterkll |
stinebd, they need to have better runtime detection |
| 01:52 |
stinebd |
Hunterkll: patches welcome |
| 01:52 |
Hunterkll |
you are not going to modify cmdline in runtime |
| 01:52 |
Hunterkll |
.... unless you feel like hacking up a solution using ksplice :P |
| 01:53 |
Hunterkll |
I grok what you're trying to do now |
| 01:53 |
stinebd |
how exactly would ksplice help? |
| 01:53 |
Hunterkll |
stinebd, modify the cmdline that's loaded in ram in the kernel image :P |
| 01:53 |
Hunterkll |
runtime patching of the kernel |
| 01:53 |
stinebd |
uh |
| 01:53 |
ali1234 |
fix the damn bootloader to actually supply command line atags section, if it doesn't |
| 01:53 |
ali1234 |
if it does, then use it |
| 01:54 |
stinebd |
none of that is the problem anyway |
| 01:54 |
Hunterkll |
stinebd, er, cmdline isn't stored in kernelspace anymore? |
| 01:54 |
stinebd |
ksplice would require a new kernel image for every cmdline you want :P |
| 01:54 |
stinebd |
quite wasteful i think |
| 01:54 |
Hunterkll |
no! |
| 01:54 |
Hunterkll |
it's a delta patch! |
| 01:55 |
Hunterkll |
a binary delta patch for a runtime image! |
| 01:55 |
Hunterkll |
one kernel, many tiny patches! |
| 01:55 |
Triskelios |
uh, ksplice is overkill for this |
| 01:55 |
Hunterkll |
one for each cmdline, in your case :P |
| 01:55 |
Hunterkll |
Triskelios, it was a joke, really ;) |
| 01:55 |
ali1234 |
jeez |
| 01:55 |
Triskelios |
besides, it isn't data in the image |
| 01:55 |
stinebd |
anyway, the real problem is our upstream kernel doesn't do the job yet |
| 01:55 |
stinebd |
so everyone knows this is a dirty hack |
| 01:55 |
Hunterkll |
So you have broken hardware detection |
| 01:55 |
Hunterkll |
:/ |
| 01:55 |
stinebd |
not broken |
| 01:55 |
stinebd |
unimplemented |
| 01:55 |
Hunterkll |
Ah. |
| 01:56 |
Hunterkll |
Have you just considered distributing different nbh for each keymap? |
| 01:56 |
Hunterkll |
:( |
| 01:56 |
WoZZeR |
Hunterkll: yes |
| 01:56 |
ali1234 |
lol not keymaps again |
| 01:56 |
WoZZeR |
I do that now, I don't want to |
| 01:56 |
Hunterkll |
lol |
| 01:56 |
Hunterkll |
Okay |
| 01:56 |
Hunterkll |
here's my next suggestion |
| 01:56 |
Hunterkll |
do you have a way to detect which keyboard you have from userspace? |
| 01:56 |
ali1234 |
you can change the keymap from userspace |
| 01:56 |
Hunterkll |
if for some reason it's hardcoded in driver, then make modules and chose and load at boot. ;) |
| 01:56 |
stinebd |
it is in userspace |
| 01:57 |
ali1234 |
put up a box saying "press key X now", detect keyboard type, set keymap with setkeycodes |
| 01:57 |
stinebd |
we just use a cmdline in rootfs as init arguments |
| 01:57 |
Hunterkll |
ali1234, they want automagic :p |
| 01:57 |
WoZZeR |
I'm trying to only edit the initrd |
| 01:57 |
ali1234 |
you only need to do this once at first boot, and then just remember the keymap in a file |
| 01:57 |
WoZZeR |
ok, /rfs/proc/cmdline is correct now |
| 01:57 |
WoZZeR |
before the switch_root |
| 01:57 |
WoZZeR |
but rootfs remounts proc, will that matter? |
| 01:57 |
ali1234 |
yes |
| 01:57 |
stinebd |
possibly not? |
| 01:57 |
Hunterkll |
Will trash your binds |
| 01:58 |
Hunterkll |
mount over them |
| 01:58 |
ali1234 |
because proc is not a regular filesystem |
| 01:58 |
Hunterkll |
erase your work |
| 01:58 |
Hunterkll |
ali1234, does bind mounting a single file into /proc even work? :D |
| 01:58 |
stinebd |
well proc is already mounted by initramfs right? |
| 01:58 |
ali1234 |
you cannot create an image with a file call /proc/cmdline and expect the kernel to just use that for cmdline |
| 01:58 |
ali1234 |
this will 100% not work |
| 01:58 |
Hunterkll |
proc reads data from kernel... /sys writes data to |
| 01:58 |
stinebd |
it will 89% work |
| 01:58 |
Hunterkll |
or am i wrong again? :P |
| 01:58 |
stinebd |
let's take wagers |
| 01:58 |
stinebd |
wrong again |
| 01:58 |
Hunterkll |
haha |
| 01:58 |
stinebd |
both read and both write |
| 01:58 |
jonpry |
just don't unmount proc. then when rootfs mounts over it will fail. you know /proc already mounted |
| 01:58 |
Hunterkll |
how things change :P |
| 01:58 |
ali1234 |
Hunterkll: they both read and write |
| 01:59 |
stinebd |
jonpry wins |
| 01:59 |
ali1234 |
/sys/ replaces /proc mostly |
| 01:59 |
stinebd |
(in theory) |
| 01:59 |
Hunterkll |
shakes 2.4.1 out of his head |
| 01:59 |
Hunterkll |
I would have removed rootfs's attempt to mount /proc also, just in case |
| 01:59 |
stinebd |
no modifying of rootfs allowed |
| 01:59 |
ali1234 |
you can actually modify the command line through /proc/atags |
| 01:59 |
stinebd |
or else he'd just be modifying rootfs's init |
| 01:59 |
ali1234 |
but you would have to kexec a new kernel to make it actually use them |
| 01:59 |
Hunterkll |
why is this not allowed? :P |
| 01:59 |
Hunterkll |
ali1234, that almost sounds like something they'd be happy with |
| 01:59 |
stinebd |
our license is too restrictive |
| 02:00 |
ali1234 |
Hunterkll: but it's totally dumb |
| 02:00 |
Hunterkll |
... license? ? |
| 02:00 |
stinebd |
(joke) |
| 02:00 |
Hunterkll |
:P |
| 02:00 |
WoZZeR |
I want to keep this as stock as possible |
| 02:00 |
stinebd |
no forking where unnecessary |
| 02:00 |
Hunterkll |
ali1234, i realize that, you realize that, and we both know 101% dumb software runs the world |
| 02:00 |
Hunterkll |
stinebd, without looking at any code, i'd press forward with a quick and dirty hw detect in driver |
| 02:01 |
ali1234 |
i don't even understand why you are trying to modify the cmdline |
| 02:01 |
Hunterkll |
depends on how feasable that is though |
| 02:01 |
stinebd |
Hunterkll: look at the code and then say that |
| 02:01 |
WoZZeR |
ali1234: because rootfs doesn't have 'good' keyboard detection |
| 02:01 |
Hunterkll |
stinebd, hence my second line :P |
| 02:01 |
WoZZeR |
it's based on cmdline |
| 02:01 |
ali1234 |
WoZZeR: i don't understand what that ahs to do with cmdline |
| 02:01 |
WoZZeR |
physkeyboard |
| 02:01 |
Hunterkll |
ali1234, they set the keymap through cmdline |
| 02:01 |
stinebd |
rootfs uses cmdline switches to set the keymap in android |
| 02:01 |
Hunterkll |
and the keymap is hard coded in driver |
| 02:01 |
Hunterkll |
is what i gather |
| 02:01 |
WoZZeR |
elif /bin/grep -c 'physkeyboard=rhod500' /proc/cmdline >/dev/null ; then |
| 02:01 |
ali1234 |
WoZZeR: you can set a different keymap with setkeycodes |
| 02:02 |
stinebd |
android :( |
| 02:02 |
Hunterkll |
ali1234, what if key values are hardcoded in driver? that still work? |
| 02:02 |
ali1234 |
Hunterkll: yes |
| 02:02 |
Hunterkll |
(also yea, android will ignore bash's keycodes :( ) |
| 02:02 |
Hunterkll |
or whatever shell you set through |
| 02:03 |
ali1234 |
setkeycodes modifies the evdev output |
| 02:03 |
Hunterkll |
oh, nvm, setkeycodes modifies the kernel tablef |
| 02:03 |
ali1234 |
how does android read keycodes from kernel? |
| 02:03 |
Hunterkll |
for some reason i was thinking setkeycodes = stty :o |
| 02:03 |
Hunterkll |
stty erase ^H ... the first thing I do on a solaris box :( |
| 02:03 |
WoZZeR |
cp -f /init.etc/keymaps/rhod500_microp-keypad.kl /etc/keymaps/microp-keypad.kl |
| 02:03 |
WoZZeR |
that's what it runs |
| 02:03 |
ali1234 |
no, it even works on X nowadays because most people use evdev with X |
| 02:03 |
Hunterkll |
wicked |
| 02:04 |
Hunterkll |
does xf86 support evdev? |
| 02:04 |
WoZZeR |
ok... w/e |
| 02:04 |
WoZZeR |
it works now |
| 02:04 |
stinebd |
xorg. and yes |
| 02:04 |
WoZZeR |
:D |
| 02:04 |
Hunterkll |
forget xorg |
| 02:04 |
stinebd |
nobody uses xf86 |
| 02:04 |
Hunterkll |
i have... legacy needs |
| 02:04 |
Hunterkll |
:( |
| 02:04 |
WoZZeR |
no serial number though |
| 02:04 |
stinebd |
bah |
| 02:04 |
WoZZeR |
stinebd: how important is the serial number |
| 02:04 |
Hunterkll |
WoZZeR, your redic bind mount system worked?!?! :o |
| 02:04 |
stinebd |
somewhere between not at all and very little |
| 02:04 |
Hunterkll |
ali1234, i'm confused... :P |
| 02:05 |
WoZZeR |
Hunterkll: yes |
| 02:05 |
jonpry |
lol |
| 02:05 |
Hunterkll |
also re ali's satement : http://linux.about.com/library/cmd/blcmdl8_setkeycodes.htm |
| 02:05 |
stinebd |
it prevents running certain games or something. tying installations to serial numbers |
| 02:05 |
Hunterkll |
it might work also :P |
| 02:05 |
stinebd |
without that set, you'll look like an emulator instance and that game (or whatever it was) will lock you out |
| 02:05 |
WoZZeR |
stinebd: should I just add that code from rootfs to initrd? |
| 02:05 |
stinebd |
you can if you want i guess |
| 02:06 |
stinebd |
99% of the users won't care, it took months for someone to report it to us |
| 02:06 |
Hunterkll |
ali1234, bind mounting a file over a bind mounted procfs inside a chroot |
| 02:06 |
stinebd |
lies and slander |
| 02:06 |
Hunterkll |
this makes my head hurt more then sysgening MVS 3.6 :( |
| 02:06 |
stinebd |
procfs is a natural mount |
| 02:06 |
WoZZeR |
./data's not mounted yet though... |
| 02:06 |
Hunterkll |
mount --bind /proc /rfs/proc ... |
| 02:06 |
stinebd |
which makes it much more interesting |
| 02:06 |
jonpry |
lol |
| 02:06 |
Hunterkll |
mkdir -m 0777 /rfs/tmp |
| 02:06 |
Hunterkll |
mount /sdcard/cache /rfs/tmp |
| 02:06 |
Hunterkll |
mount --bind /proc /rfs/proc |
| 02:07 |
Hunterkll |
is in your script, i believe |
| 02:07 |
Hunterkll |
that's not a natural mount |
| 02:07 |
WoZZeR |
so? |
| 02:07 |
Hunterkll |
:P |
| 02:07 |
WoZZeR |
it works |
| 02:07 |
stinebd |
look in rootfs init |
| 02:07 |
Hunterkll |
WoZZeR, he said i was wrong when i said bind'd file over bind'd proc in chroot |
| 02:07 |
Hunterkll |
:P |
| 02:07 |
jonpry |
wozzer is just ahead of his time |
| 02:07 |
stinebd |
oh wait yeah |
| 02:07 |
stinebd |
that would probably fail anyway |
| 02:07 |
Hunterkll |
wozzer said it worked? :o |
| 02:07 |
WoZZeR |
I have keyboard |
| 02:08 |
stinebd |
the real proc mount would fail |
| 02:08 |
WoZZeR |
but I lost the serial number |
| 02:08 |
Hunterkll |
stinebd, inside rootfs, yes.... and it just skips merrily along :P |
| 02:08 |
Hunterkll |
"WHAT? I CANT MOUNT PROC? OKAY LETS GO" |
| 02:08 |
stinebd |
why should that kill init? |
| 02:08 |
Hunterkll |
It shouldn't. |
| 02:08 |
Hunterkll |
But if proc weren't actually there, it would. :P |
| 02:09 |
Hunterkll |
At least, things would die |
| 02:09 |
stinebd |
the vitriol in your typing indicated it should :P |
| 02:09 |
Hunterkll |
it should have a sanity check and throw big red warnings at yoy |
| 02:09 |
Hunterkll |
or |
| 02:09 |
Hunterkll |
just say ER IN PR BLK SPR MT |
| 02:09 |
stinebd |
red? we can barely figure out mounts and you want console colors? |
| 02:09 |
Hunterkll |
:D |
| 02:09 |
jonpry |
i like software that fails silently. like mjdroidutil.exe |
| 02:10 |
WoZZeR |
lol |
| 02:10 |
Hunterkll |
hacks git repos and replaces all debug and status messags with solaris style messages |
| 02:10 |
Hunterkll |
ERR IN SPRBLK SKN DSK C0T0D0S1 |
| 02:10 |
stinebd |
massive corn clog in port 7 |
| 02:10 |
Hunterkll |
actually, the one i just pasted is almost a real one |
| 02:10 |
Hunterkll |
mins a few letters here or there, it's saying that slice 1 (partition 2) has a bad superblock |
| 02:10 |
Hunterkll |
:P |
| 02:11 |
Hunterkll |
and it's going to run fsck on it |
| 02:11 |
Hunterkll |
I remember the sadness that filled my heart when my 40mhz sparc woldn't boot and gave that message repeatedly... do you know how hard it is to find boxed retail copies of unix from ~12 years ago? :( |
| 02:12 |
Hunterkll |
that little sun ipx was like a beloved pet to me, really... |
| 02:12 |
Hunterkll |
X11R5 |
| 02:12 |
Hunterkll |
.... |
| 02:12 |
Hunterkll |
Does anyone want to hurt me now? :( |
| 02:13 |
stinebd |
strange, serialno should still work |
| 02:13 |
Hunterkll |
WoZZeR, is it readable otherwise or is it mangled system wide |
| 02:13 |
stinebd |
even with a failed rebind, you're just modifying the source file |
| 02:13 |
WoZZeR |
Hunterkll: what? /proc/cmdline? |
| 02:13 |
Hunterkll |
eg android only can't read it, or you can't read at all... try dumping the file (cat/type) while in your booted system |
| 02:13 |
Hunterkll |
see if serial is still there |
| 02:14 |
Hunterkll |
you may have botched your sed statement |
| 02:14 |
WoZZeR |
android can read /proc/cmdline |
| 02:14 |
WoZZeR |
there's no serial |
| 02:14 |
Hunterkll |
okay. and that is where you get serial from? |
| 02:14 |
WoZZeR |
I already went over that |
| 02:14 |
WoZZeR |
./data/serialno |
| 02:14 |
WoZZeR |
but it can't edit it |
| 02:14 |
Hunterkll |
oh |
| 02:14 |
ali1234 |
so the parameter set in /proc/cmdline is not even used by kernel, it's just for android? |
| 02:14 |
Hunterkll |
how do you generate serialno file? |
| 02:14 |
stinebd |
http://gitorious.org/xdandroid-eclair/eclair-rootfs/blobs/master/init#line545 for context |
| 02:15 |
Hunterkll |
ali1234, Has anyone in this channel ever mentioned they hate android? :P |
| 02:15 |
WoZZeR |
I have |
| 02:15 |
WoZZeR |
a few times |
| 02:15 |
WoZZeR |
but I like challanges |
| 02:15 |
stinebd |
ali1234: physkeyboard is used exclusively by our rootfs. andboot.serialno is used exclusively by android |
| 02:15 |
jonpry |
wozzer: can it read /tmp/cmdline? |
| 02:15 |
Hunterkll |
ah |
| 02:15 |
WoZZeR |
I'll test in a second |
| 02:15 |
WoZZeR |
just rebooted |
| 02:15 |
ali1234 |
wow. just. wow |
| 02:15 |
Hunterkll |
stinebd, randomly generated, and only change is proc mounts.... so data is not modified. should work |
| 02:15 |
Hunterkll |
ali1234, i'd try and find a way to pull the real serial # :( |
| 02:15 |
Hunterkll |
WinMo can do it, you can too! :o |
| 02:15 |
ali1234 |
why don't you just change android to read from /data/serialno |
| 02:15 |
WoZZeR |
that's what I'd do |
| 02:16 |
David_ |
whats up with that Ctll_mach_msm_LSENSOR_SRC.tar.gz ?? |
| 02:16 |
stinebd |
ali1234: lol :( |
| 02:16 |
jonpry |
i can get the serial |
| 02:16 |
jonpry |
but who wants it |
| 02:16 |
Hunterkll |
i used to have android build sys |
| 02:16 |
Hunterkll |
jonpry, android does for copy protection in apps :P |
| 02:16 |
stinebd |
only a couple apps |
| 02:16 |
WoZZeR |
jonpry: I can easily throw it in the initrd then |
| 02:16 |
jonpry |
so why don't we all use 1 serial. then we all can have free apps |
| 02:16 |
stinebd |
market uses google account for copy protection |
| 02:16 |
jonpry |
this reminds me cue cat |
| 02:17 |
Hunterkll |
i need sleep now been awake 24 hour oops :P |
| 02:17 |
Hunterkll |
good lukc |
| 02:17 |
stinebd |
jonpry: is serial stored in mtd somewhere? |
| 02:17 |
WoZZeR |
jonpry: no /tmp/cmdline |
| 02:17 |
jonpry |
stinebd, its in spl |
| 02:17 |
WoZZeR |
brb |
| 02:17 |
stinebd |
because if we can pull the actual one with dd or some crap, that would be ideal |
| 02:17 |
jonpry |
wozzer: can you pastebin your latest script/ |
| 02:18 |
jonpry |
stinebd: its easy enough to make a driver that can serve it up on sysfs |
| 02:18 |
stinebd |
k. get crackin |
| 02:18 |
stinebd |
runs |
| 02:19 |
jonpry |
how does normal android provide serial? |
| 02:19 |
Triskelios |
if I want to modify a structure in a stock HTC kernel which doesn't appear in /proc/kallsyms, does scanning for a pc-relative reference in a function with a known address make sense? |
| 02:19 |
stinebd |
bootloader gives it that cmdline argument |
| 02:20 |
jonpry |
guess we need uboot |
| 02:20 |
stinebd |
wassat |
| 02:20 |
ali1234 |
just add it to the bootlaoder you have |
| 02:21 |
jonpry |
android device bootloader. you know with nandroid and cool stuff |
| 02:21 |
WoZZeR |
jonpry: the complete initrd? |
| 02:21 |
stinebd |
the bootloader we have is haret |
| 02:21 |
jonpry |
just this thing that mounts remounts etc |
| 02:21 |
ali1234 |
stinebd: great. haret can do anything you need... |
| 02:21 |
WoZZeR |
jonpry: that's the init -> http://pastebin.com/ChH8u5PH |
| 02:22 |
stinebd |
i'm not sure haret can do that |
| 02:22 |
stinebd |
heh |
| 02:22 |
ali1234 |
just patch it |
| 02:22 |
stinebd |
lol |
| 02:22 |
WoZZeR |
uboot is dangerous |
| 02:22 |
WoZZeR |
it's a spl replacement |
| 02:22 |
jonpry |
wozzer: you need to bind rfs/tmp |
| 02:22 |
WoZZeR |
if I'm thinking of the right thing |
| 02:22 |
ali1234 |
WoZZeR: it does not have to replace spl |
| 02:23 |
ali1234 |
you can run it from anywhere you can run a kernel |
| 02:23 |
WoZZeR |
jonpry: where? |
| 02:23 |
jonpry |
95 |
| 02:23 |
jonpry |
put --bind |
| 02:24 |
stinebd |
i must stop slowing you folks down now. goodnight and good luck |
| 02:24 |
WoZZeR |
lol, thanks |
| 02:24 |
WoZZeR |
jonpry: is serialno a pain to get? |
| 02:24 |
WoZZeR |
if this doesn't work |
| 02:25 |
jonpry |
shouldn't be |
| 02:25 |
WoZZeR |
android booting now |
| 02:25 |
WoZZeR |
I'll let you know if it fixes it |
| 02:28 |
WoZZeR |
no /tmp/cmdline |
| 02:33 |
Cass |
Markinus, you here ? |
| 02:34 |
jonpry |
argh. what do they want for serial? IMEI? |
| 02:34 |
WoZZeR |
I don't care right now |
| 02:34 |
WoZZeR |
it's not something that has to be done today |
| 02:34 |
jonpry |
nothing has to be done |
| 02:35 |
jonpry |
wozzer: what is mounted on /tmp? |
| 02:35 |
WoZZeR |
the cache folder on the sd card |
| 02:35 |
WoZZeR |
which is seems to format |
| 02:35 |
jonpry |
so it is clearing temp on boot? |
| 02:36 |
WoZZeR |
I think so |
| 02:36 |
WoZZeR |
all I see is lost+found |
| 02:36 |
WoZZeR |
let me look at rootfs init |
| 02:36 |
WoZZeR |
maybe android is formatting it |
| 02:38 |
WoZZeR |
brb, making some food |
| 02:40 |
jonpry |
ok. i have an idea. lol |
| 02:41 |
jonpry |
make it one level removed. so save to /somesameplace/cmdline. bind that to /rfs/tmp/cmdline, then bind /rfs/proc/cmdline. then when init does something silly like rm -R -f /tmp/*, it will not clear the bind. hence an irremovable file |
| 02:45 |
WoZZeR |
I don't know if that's the issue though |
| 02:45 |
WoZZeR |
becuase /proc/cmdline is correct |
| 02:45 |
WoZZeR |
it just doesn't add the serialno in rootfs |
| 02:45 |
WoZZeR |
cp /proc/cmdline /tmp/cmdline |
| 02:45 |
WoZZeR |
that might be the issue |
| 02:45 |
WoZZeR |
you can't copy a file onto it'self? |
| 02:47 |
jonpry |
well, it can't remount the file. so you need to make sure that it is copying onto itself |
| 02:50 |
jonpry |
tis weird that there is nothing in tmp though. should at least be whatever rootfs made even if it doesn't match /proc/cmdline |
| 02:50 |
WoZZeR |
I don't see anything that would change /tmp |
| 02:50 |
WoZZeR |
http://pastebin.com/Z09cKG4a |
| 02:50 |
WoZZeR |
there's the rootfs init |
| 02:52 |
jonpry |
its chroot'ing again |
| 02:54 |
WoZZeR |
where? |
| 02:55 |
WoZZeR |
oh crap |
| 02:55 |
WoZZeR |
wrong file |
| 02:55 |
WoZZeR |
whoops, |
| 02:55 |
WoZZeR |
http://pastebin.com/GvqYwiyH |
| 03:00 |
jonpry |
you don't have a /tmp/build.prop? |
| 03:00 |
WoZZeR |
mount --bind /tmp/build.prop /system/build.prop |
| 03:01 |
WoZZeR |
temp is empty |
| 03:01 |
WoZZeR |
so it must get deleted later |
| 03:01 |
jonpry |
yeah, or there is another chroot later on. either way doesn't help us with what is happening at this point |
| 03:02 |
jonpry |
what is the contents of /proc/cmdline when it finishes booting? |
| 03:02 |
WoZZeR |
gsensor_axis=2,1,3 msmvkeyb_toggle=off hw3d.force=0 panic=0 pm.sleep_mode=2 physkeyboard=rhod500 |
| 03:02 |
jonpry |
is that the real cmdline or one made by your hack? |
| 03:03 |
WoZZeR |
the one from my hack. asciz "gsensor_axis=2,1,3 msmvkeyb_toggle=off hw3d.force=0 panic=0 pm.sleep_mode=2" is the real one |
| 03:04 |
jonpry |
you changed root fs to do the init chardev? |
| 03:04 |
WoZZeR |
no |
| 03:04 |
WoZZeR |
that's just the one mounted |
| 03:04 |
WoZZeR |
not the one I'm using |
| 03:06 |
jonpry |
hmm |
| 03:06 |
jonpry |
its hard to tell what is happening there. maybe add some prints and such |
| 03:07 |
WoZZeR |
it's hard to see because it boots into android shortly |
| 03:08 |
WoZZeR |
I can try a sleep |
| 03:08 |
WoZZeR |
but we'll see |
| 03:09 |
jonpry |
the other option is to use your own rootfs init and pack it into initrd. then after mounting rootfs, you just patch rootfs with the new init |
| 03:09 |
WoZZeR |
yeah, I was trying not to do that |
| 03:09 |
jonpry |
lol |
| 03:13 |
jonpry |
who cares about apps anyways. i can't even get marketplace to start |
| 03:13 |
WoZZeR |
lol |
| 03:14 |
WoZZeR |
I may just create serialno on sd card in andboot |
| 03:15 |
jonpry |
you could just copy those 2 lines of code that generate it and add it to cmdline |
| 03:15 |
WoZZeR |
I think it needs to be the same for apps to work |
| 03:15 |
WoZZeR |
not just randomly generated each time |
| 03:16 |
jonpry |
the generator checks to see if the file exists first |
| 03:17 |
jonpry |
[ -f "/data/serialno" ] || echo -e `cat /dev/urandom | tr -dc 'A-Z0-9' | fold -w 12 | head -n 1` >/data/serialno |
| 03:17 |
WoZZeR |
yeah |
| 03:17 |
WoZZeR |
testing now |
| 03:17 |
WoZZeR |
[ -f "/sdcard/andboot/serialno" ] || echo -e `cat /dev/urandom | tr -dc 'A-Z0-9' | fold -w 12 | head -n 1` >/sdcard/andboot/serialno |
| 03:17 |
WoZZeR |
SERIALNO=$(cat /sdcard/andboot/serialno) |
| 03:18 |
WoZZeR |
crap |
| 03:18 |
WoZZeR |
sed -i -e "s:$: androidboot.serialno=${SERIALNO}:" /rfs/tmp/cmdline |
| 03:18 |
WoZZeR |
forgot to add that... lol |
| 03:18 |
jonpry |
might be important |
| 03:19 |
WoZZeR |
any word from wistilt2 on keyboard light? |
| 03:23 |
jonpry |
haven't seen him today |
| 03:23 |
WoZZeR |
hmm, didn't add the number |
| 03:24 |
WoZZeR |
hmmm, blank serial no |
| 03:24 |
jonpry |
paste? |
| 03:24 |
WoZZeR |
ahh |
| 03:24 |
WoZZeR |
./dev/urandom |
| 03:24 |
WoZZeR |
should be rfs/? |
| 03:24 |
jonpry |
no |
| 03:24 |
WoZZeR |
cp -a /dev/* /rfs/dev |
| 03:25 |
jonpry |
it is in both |
| 03:25 |
WoZZeR |
it created a blank serial number |
| 03:25 |
WoZZeR |
the file has 2 blank lines |
| 03:25 |
WoZZeR |
no urandom in initrd? |
| 03:25 |
WoZZeR |
I don't see it in the bin |
| 03:26 |
jonpry |
its a device file |
| 03:26 |
jonpry |
like chardev |
| 03:26 |
WoZZeR |
oh |
| 03:26 |
WoZZeR |
[ -f "/sdcard/andboot/serialno" ] || echo -e `cat /dev/urandom | tr -dc 'A-Z0-9' | fold -w 12 | head -n 1` >/sdcard/andboot/serialno |
| 03:26 |
WoZZeR |
SERIALNO=$(cat /sdcard/andboot/serialno) |
| 03:26 |
WoZZeR |
that's what I have |
| 03:26 |
WoZZeR |
and it created a blank file |
| 03:27 |
jonpry |
your initrf may not have all the commands |
| 03:27 |
jonpry |
fold and head |
| 03:27 |
WoZZeR |
I see them |
| 03:27 |
WoZZeR |
lrwxrwxrwx 1 wozzer wozzer 7 2010-08-25 14:14 fold -> busybox |
| 03:27 |
WoZZeR |
lrwxrwxrwx 1 wozzer wozzer 7 2010-08-25 14:14 head -> busybox |
| 03:28 |
jonpry |
what if you run that command after boot? |
| 03:28 |
WoZZeR |
build now |
| 03:28 |
jonpry |
probably have to delete serialno |
| 03:28 |
WoZZeR |
yeah |
| 03:29 |
WoZZeR |
cat /dev/urandom |
| 03:29 |
WoZZeR |
sleep 20 |
| 03:29 |
WoZZeR |
added that to initrd |
| 03:29 |
WoZZeR |
to see if it spits anything out |
| 03:29 |
jonpry |
omg |
| 03:30 |
jonpry |
that will spit some stuff out alright |
| 03:30 |
jonpry |
it just won't ever stop |
| 03:33 |
WoZZeR |
does not exist |
| 03:34 |
jonpry |
mknod it i guess |
| 03:34 |
WoZZeR |
mknod /dev/urandom -? |
| 03:34 |
WoZZeR |
what else do I need |
| 03:34 |
jonpry |
mknod -m 644 /dev/urandom c 1 9 |
| 03:36 |
WoZZeR |
flashing now |
| 03:39 |
WoZZeR |
w00t, serial number |
| 03:40 |
WoZZeR |
good work! |
| 03:41 |
jonpry |
cool. so only one nbh now? |
| 03:41 |
WoZZeR |
yup |
| 03:41 |
WoZZeR |
need data! |
| 03:41 |
WoZZeR |
where's the testers when I need them |
| 03:41 |
jonpry |
they start getting online about now. |
| 03:42 |
WoZZeR |
it's a saturday though |
| 03:43 |
jonpry |
thats a good day for people willing to flash their phones |
| 03:43 |
WoZZeR |
I need some rhod100 variants |
| 03:44 |
jonpry |
so say somebody uses it. and keymap is wrong. how are they going to collect info? |
| 03:45 |
WoZZeR |
softkeyboard |
| 03:45 |
WoZZeR |
I can't make a ">" key |
| 03:45 |
WoZZeR |
so I need to pop open the softkeyboard for that |
| 03:46 |
jonpry |
i see |
| 03:47 |
jonpry |
wouldn't it be great if rhod had camera |
| 03:47 |
WoZZeR |
do we know why camera doesn't work? |
| 03:48 |
jonpry |
well, the last time anybody worked on it. the trouble was with adsp |
| 03:48 |
jonpry |
it might work now. but the code is in cam branch. and it might require some userspace settings or something. not sure where to get started |
| 03:50 |
WoZZeR |
hmmm, seems like a pain |
| 03:50 |
WoZZeR |
just up your ally :P |
| 03:51 |
jonpry |
if we could use that stupid wavedev loader to get the adsp tables. probably could get them 100%. then qcam would work |
| 03:51 |
WoZZeR |
alley |
| 03:51 |
[acl] |
sup fellas |
| 03:51 |
WoZZeR |
[acl]: yo |
| 03:51 |
[acl] |
whats the good news ? |
| 03:51 |
WoZZeR |
we got single nbh now |
| 03:52 |
WoZZeR |
it's quite a dirty looking hack |
| 03:52 |
[acl] |
i didnt know we had multiple |
| 03:52 |
WoZZeR |
one for each keyboard |
| 03:52 |
[acl] |
ahh |
| 03:52 |
[acl] |
ok |
| 03:52 |
WoZZeR |
rhod400, rhod500 |
| 03:52 |
[acl] |
how yall do it? |
| 03:52 |
WoZZeR |
jonpry made some code to get the model number |
| 03:53 |
[acl] |
ahh |
| 03:53 |
[acl] |
nice |
| 03:53 |
jonpry |
hi [acl] |
| 03:53 |
WoZZeR |
then we hackishly edited the cmdline in initrd |
| 03:53 |
[acl] |
we needed that anyways |
| 03:53 |
[acl] |
jonpry: sup bro.. been having issues with my connection here.. havent read logs in a while |
| 03:53 |
jonpry |
ah yes island internet |
| 03:54 |
[acl] |
WoZZeR: thats good.. should work fine for now |
| 03:54 |
WoZZeR |
yeah, until rootfs does it for us |
| 03:54 |
jonpry |
its good lots of stuff is working now. nand seems better than msm-2.6.27 for the time being |
| 03:55 |
WoZZeR |
jonpry: only because we can commit |
| 03:55 |
[acl] |
jonpry: woah.. awesome... what changes do we have that main dont? |
| 03:55 |
[acl] |
is phh around to make commits to the main ? |
| 03:55 |
WoZZeR |
incall volume changing |
| 03:55 |
[acl] |
i think he is out till tues if i remember correctly |
| 03:55 |
WoZZeR |
I think jonpry said that it could cause problems for other devices, don't remember though |
| 03:55 |
[acl] |
WoZZeR:incall vol? you guys trying to make me proud ? yall must be.. there is a tear in my eye |
| 03:56 |
WoZZeR |
[acl]: we get a lot done when you are on vacation, when is your next one? |
| 03:56 |
jonpry |
speaker/mic seem to work well. volume control. pohket can't get it to work though. he thinks it is because he has gsm phone. i think nand is just better :p |
| 03:56 |
WoZZeR |
:P |
| 03:56 |
[acl] |
WoZZeR: LOL .. in dec :-) |
| 03:57 |
WoZZeR |
Pohket had some changes that didn't work for me either |
| 03:57 |
[acl] |
jonpry: how did you implement it ? |
| 03:57 |
WoZZeR |
it's dealing with calls in a weird way or something |
| 03:57 |
jonpry |
the speaker management call vol stuff is probably good for other devices. htc_acoustic is rhod only for now |
| 03:58 |
[acl] |
jonpry: in that case phh can pick and rip whatever is needed frmo our code and push himself. This way we dont break other devices. |
| 03:58 |
jonpry |
[acl] its just a matter of loading the right stuff in the ADIE tables. we have all this crap in AudioParams.c. you just have to use the right one at the right time |
| 03:58 |
jonpry |
pohket made us up a nice state machine that calculates when the speaker is supposed to be on and such |
| 03:59 |
[acl] |
ahh ic |
| 03:59 |
[acl] |
did wistilt also provide his changes? |
| 03:59 |
[acl] |
for kb lights ? |
| 03:59 |
jonpry |
no. he got them on. but that seems to be a bit of a herring. now he can't turn them off. ever, at all |
| 04:00 |
[acl] |
jonpry: ahh.. ok.. |
| 04:01 |
[acl] |
i spoke to him 2 nights ago about what else he might tackle.. i think pm might be calling his name and that horrible sleep issue where our lcd seems to be on |
| 04:02 |
jonpry |
i want to get usb working. bzo says he got adb with an old kernel |
| 04:03 |
jonpry |
i looked at it a bit and that sounds possible. there is nothing wrong with the code. same code should work on all msm7k and even qsd |
| 04:03 |
[acl] |
its easier with haret version. with nand we had to do some prep work to get the irqs working. We should be able to make those bzo changes now |
| 04:03 |
jonpry |
i tried on haret, and it doesn't work |
| 04:03 |
[acl] |
with his changes? |
| 04:03 |
jonpry |
ulpi write fail. same as on nand |
| 04:03 |
jonpry |
bzo changes? |
| 04:04 |
[acl] |
didnt u say bzo got adb working |
| 04:04 |
jonpry |
by using an old kernel. but i don't know which version |
| 04:04 |
[acl] |
old kernel as in haret rite? |
| 04:04 |
jonpry |
yes |
| 04:09 |
[acl] |
whats this irq 26 issue that was posted on ppcg.. im so behind |
| 04:13 |
[acl] |
WoZZeR: u modded irq.h ? |
| 04:13 |
[acl] |
int_usb_hs |
| 04:13 |
WoZZeR |
[acl]: yeah, there were some usb tests |
| 04:14 |
WoZZeR |
usb got hits from dex or something |
| 04:14 |
[acl] |
you think the irq numbers are different ? |
| 04:14 |
WoZZeR |
jonpry knows more |
| 04:14 |
WoZZeR |
wistilt2 wanted me to change to 6 |
| 04:14 |
jonpry |
the irq number is fine. wistilt2 was getting out of control |
| 04:14 |
WoZZeR |
I'll change back |
| 04:23 |
[acl] |
Its gonna take some work to update the main..lol.. looks like alot of changes were made |
| 04:23 |
[acl] |
i cant seem em all since this connection is torturing me |
| 04:45 |
[acl] |
aite fellas.. i gotta jet.. thanks for the update on the project.. ill ttyl |
| 04:46 |
[acl] |
WoZZeR: enjoy ur few days before work starts |
| 04:46 |
WoZZeR |
thanks, later |
| 05:16 |
rafyvitto |
hey hd2 devotes =] |
| 05:17 |
rafyvitto |
hats down to cotulla i coldn belive it until i tryed an android build for the hd2, its pretty amazing, it almost feels like a native device "almost" |
| 05:17 |
rafyvitto |
and decoders etc etc all the devs |
| 05:18 |
rafyvitto |
dcordes* |
| 05:37 |
bmnbmn |
WoZZeR: tester is here! |
| 05:37 |
bmnbmn |
a few hours too late |
| 05:41 |
bmnbmn |
Though I wish you could find out why slower microsd cards don't work :( |
| 05:46 |
jonpry |
they probably set the pll frequency wrong |
| 05:46 |
WoZZeR |
jonpry: it happens with both clocks |
| 05:46 |
WoZZeR |
old and new |
| 05:47 |
jonpry |
doesn't mean it wasn't wrong before |
| 05:47 |
WoZZeR |
I've never had a problem with it |
| 05:47 |
WoZZeR |
something with all the audio changes |
| 05:47 |
jonpry |
afaict most sd cards will actually work with 50mhz speed |
| 05:47 |
WoZZeR |
no reports before sound worked |
| 05:47 |
WoZZeR |
could be rpcrouter |
| 05:47 |
WoZZeR |
could be almost anything |
| 05:48 |
jonpry |
but others are definitely limited to 25mhz |
| 05:48 |
jonpry |
and some require a special cmd code to change the io buffer timing |
| 05:49 |
jonpry |
can't really tell until somebody can reproduce |
| 05:49 |
bmnbmn |
im gonan flash your new test |
| 05:49 |
bmnbmn |
and see what happens |
| 05:49 |
WoZZeR |
ok |
| 05:49 |
bmnbmn |
i have nothing else to do |
| 05:51 |
jonpry |
you'll notice that in clocks-wince.c. they use same pll settings for sd cards |
| 05:51 |
jonpry |
on both cdma and gsm. but one is 25% faster |
| 05:52 |
WoZZeR |
jonpry: that's only in the new clocks. The old clocks they are different |
| 05:52 |
jonpry |
i'm pretty sure i have old clocks |
| 05:52 |
WoZZeR |
jonpry: but it doesn't effect my RHOD500 |
| 05:52 |
WoZZeR |
as far as I can tell |
| 05:53 |
jonpry |
depends on how fast your sd card is |
| 05:53 |
bmnbmn |
whoaa |
| 05:53 |
WoZZeR |
which is why I just say get a class 4 |
| 05:53 |
jonpry |
lol |
| 05:53 |
WoZZeR |
if you are using a class 2, you should just return your phone and get an lg dumbass phone |
| 05:53 |
WoZZeR |
:P |
| 05:53 |
bmnbmn |
keyboard stuff is broke for me, maybe |
| 05:54 |
WoZZeR |
bmnbmn: how so? |
| 05:54 |
bmnbmn |
mkdir: cannot create directory /fs/tmp': File exists |
| 05:54 |
jonpry |
class 2 isn't really going to tell you anything about the bus timing. we are way out of spec with wrong clock speed |
| 05:54 |
WoZZeR |
no, just let it go |
| 05:54 |
bmnbmn |
oh ok |
| 05:54 |
bmnbmn |
just lots of failures |
| 05:54 |
WoZZeR |
bmnbmn: there's lots of crap in there that doesn't need to be |
| 05:54 |
WoZZeR |
I wouldn't say it's clean code |
| 05:54 |
bmnbmn |
my keyboard wasnt in the echo'd cmdline though |
| 05:55 |
WoZZeR |
?? |
| 05:55 |
jonpry |
although in practice bus timing does have some correlation with the technology of the card |
| 05:55 |
WoZZeR |
what did it say |
| 05:55 |
bmnbmn |
let me get back there and find out |
| 05:56 |
WoZZeR |
bmnbmn: which echo'd cmd line? |
| 05:56 |
WoZZeR |
are you in android now? |
| 05:56 |
bmnbmn |
nope, same irq 26 problem |
| 05:57 |
jonpry |
bmnbmn: you have gsm? |
| 05:57 |
bmnbmn |
cdma, rhod400 |
| 05:57 |
jonpry |
there goes that theory |
| 05:57 |
bmnbmn |
i noticed when i get the irq 26 error, that my touch screen calibration file does not get written to |
| 05:57 |
bmnbmn |
or i guess, anything to my sd card doesnt get written |
| 05:57 |
WoZZeR |
bmnbmn: because it can't read the card |
| 05:57 |
bmnbmn |
its bootting off of it, how can it not read it |
| 05:58 |
WoZZeR |
it stops readying it, I guess |
| 05:58 |
WoZZeR |
actually, it depends on where it dies |
| 05:58 |
WoZZeR |
it may not be reading it at all |
| 05:58 |
WoZZeR |
it may detect that it's there |
| 05:58 |
WoZZeR |
but it may not be reading it |
| 05:59 |
bmnbmn |
INFO: task rpcrouter:15 blocked for more than 120 seconds. |
| 06:00 |
bmnbmn |
how about this one |
| 06:00 |
bmnbmn |
sdslot_switchvdd: Invalid VDD 17 specified |
| 06:01 |
bmnbmn |
sdslot_switchvdd: Disabling SD slot power |
| 06:01 |
bmnbmn |
that doesnt sound good |
| 06:02 |
bmnbmn |
looks like you already had a pastebin entry with that in it anyways |
| 06:02 |
bmnbmn |
thanks to oogle |
| 06:05 |
WoZZeR |
something not right, I think it should work fine |
| 06:05 |
WoZZeR |
I'm assuming something in all the audio changes made something go wrong |
| 06:06 |
jonpry |
switchvdd sounds bad |
| 06:07 |
bmnbmn |
jonpry: must not be that bad, wozzer had a dmesg with it |
| 06:07 |
bmnbmn |
so im assuming everyones getting that |
| 06:08 |
bmnbmn |
WoZZeR: the closest i got to booting was when i used the 8/19 system.ext2 with your test |
| 06:08 |
bmnbmn |
it actually booted to the splash screen and froze there |
| 06:09 |
jonpry |
wozzer: do you know if gsm people are using your nbh's. and having stuff work? |
| 06:09 |
WoZZeR |
no idea |
| 06:09 |
WoZZeR |
no repsonces |
| 06:13 |
bmnbmn |
well, its got my keyboard in the cmdline |
| 06:13 |
bmnbmn |
so that works |
| 06:13 |
sykosoft |
was reading the logs |
| 06:14 |
sykosoft |
saw a mention of is the nbh working |
| 06:14 |
sykosoft |
yes, wozzer's latest test 5 nbh for sprint rhod400 is working |
| 06:14 |
sykosoft |
8/30 build |
| 06:14 |
bmnbmn |
we're up to test 7 now |
| 06:14 |
sykosoft |
I did have a problem where it was completely fresh, and wouldn't start withotu a data.img already in place |
| 06:14 |
bmnbmn |
or he is |
| 06:14 |
sykosoft |
seriously? I downloaded test5 last night |
| 06:15 |
bmnbmn |
he made a unified nbh a bit ago that autodetects rhodium versions |
| 06:15 |
sykosoft |
all right, haven't tested that one yet |
| 06:16 |
sykosoft |
my bad, I'll test that one in 1 hour or so |
| 06:17 |
sykosoft |
ppcgeeks thread still says test5, where should I get the test7? |
| 06:17 |
bmnbmn |
it should be in that thread |
| 06:17 |
bmnbmn |
maybe its test 6 :O |
| 06:18 |
jonpry |
wozzer, i dare you to set sd card to 50mhz |
| 06:18 |
sykosoft |
ok, I'll test 6 when I get back |
| 06:18 |
WoZZeR |
jonpry: how do I do that? |
| 06:19 |
tids2k |
darky u thr' |
| 06:20 |
bmnbmn |
clock_late_init(): disabled 16 unused clocks |
| 06:20 |
bmnbmn |
a few FIXMEs |
| 06:27 |
WoZZeR |
sykosoft: rhod 400 or 500? |
| 06:31 |
bmnbmn |
i will be right back |
| 06:33 |
jonpry |
wozzer: i am not sure |
| 06:34 |
WoZZeR |
jonpry: [ physkey == "rhod300" ] || physkey="tilt2" |
| 06:34 |
WoZZeR |
is that correct? |
| 06:36 |
WoZZeR |
[ "$physkey" == "rhod300" ] || physkey="tilt2" |
| 06:36 |
WoZZeR |
or is it that? |
| 06:37 |
jonpry |
what are you trying to do? |
| 06:37 |
WoZZeR |
change physkey from rhod300 to tilt2 |
| 06:45 |
jonpry |
physkey="rhod300"; |
| 06:45 |
jonpry |
if [ $physkey = "rhod300" ]; then |
| 06:45 |
jonpry |
physkey="tilt2" |
| 06:45 |
jonpry |
fi |
| 06:46 |
jonpry |
i couldn't get anything else to work |
| 06:49 |
jonpry |
bzo: hi |
| 06:50 |
WoZZeR |
hmm |
| 06:50 |
bzo |
hey |
| 06:50 |
WoZZeR |
hey |
| 06:50 |
jonpry |
you got usb to work? |
| 06:51 |
bzo |
hasn't usb always worked? I though just adb was the problem |
| 06:51 |
jonpry |
http://www.androidonhtc.com/wiki/Port_Status |
| 06:52 |
bzo |
right, with the caveat "2 Only works if USB is connected at boot time" |
| 06:53 |
jonpry |
thats raph and diam |
| 06:53 |
bzo |
hmm, I suspect the wiki is incorrect |
| 06:54 |
bzo |
it's hard to say though because adb has been the only thing that has worked on usb for me regardless of device |
| 06:55 |
jonpry |
i will try while usb connected |
| 06:56 |
jonpry |
don't think i have done that yet |
| 07:01 |
WoZZeR |
any of my nand testers here? |
| 07:02 |
jonpry |
hmm. usb |
| 07:02 |
bzo |
working? |
| 07:03 |
NeoMatrixJR_Mobi |
Hey WoZZeR, when u guys go back to winmo what do u use to backup/restore. Your stuff all the time? |
| 07:03 |
jonpry |
seems to |
| 07:03 |
bzo |
adb? |
| 07:03 |
WoZZeR |
NeoMatrixJR_Mobi: I make my own roms |
| 07:03 |
WoZZeR |
no restoring/backingup |
| 07:03 |
jonpry |
i'm running rhobuntu |
| 07:03 |
NeoMatrixJR_Mobi |
Jonpry: works when u boot w/ USB plugged in or in general? |
| 07:04 |
jonpry |
boot w/ USB active |
| 07:04 |
WoZZeR |
I remember having usb options when it was plugged in |
| 07:04 |
WoZZeR |
but mass storage didn't work |
| 07:04 |
bzo |
mass storage can't work until everything is moved to nand |
| 07:04 |
bzo |
because it requires exclusive access to the sdcard |
| 07:05 |
jonpry |
or a partition |
| 07:05 |
bzo |
hmm, true. I guess a partitioned setup should be able to support it even w/o nand |
| 07:06 |
NeoMatrixJR_Mobi |
But I thought they took out partitioned sd support a while back.... |
| 07:06 |
bzo |
I think the partitioned stuff still works, its just no encouraged |
| 07:07 |
bzo |
s/no/not |
| 07:08 |
bzo |
the wiki on port status has some questionable status aside from the rhodium usb |
| 07:09 |
jonpry |
shouldn't be too hard to trace what happens when usb is plugged in |
| 07:09 |
bzo |
how is 3d listed as being partially working? lol |
| 07:09 |
jonpry |
software |
| 07:09 |
bzo |
also, I think the power savings/suspend works equally bad on all devices :) |
| 07:10 |
bzo |
not sure why the rhodium/topaz is called out as being worse |
| 07:11 |
NeoMatrixJR_Mobi |
WoZZer/Jonpry/acl: do your kernel changes still work properly on SD as well as nand? |
| 07:12 |
WoZZeR |
some do, some don't |
| 07:12 |
WoZZeR |
afiak |
| 07:12 |
bmnbmn |
WoZZeR: did you just post that test 7? |
| 07:12 |
bmnbmn |
with the *may* fix IRQ 26 errors? |
| 07:12 |
jonpry |
[101392.363744] usb-storage: device found at 23 |
| 07:12 |
jonpry |
[101392.363749] usb-storage: waiting for device to settle before scanning |
| 07:12 |
jonpry |
[101397.360492] usb-storage: device scan complete |
| 07:12 |
jonpry |
[101397.362988] scsi 15:0:0:0: Direct-Access HTC Android Phone 0100 PQ: 0 ANSI: 2 |
| 07:12 |
jonpry |
[101397.410823] sd 15:0:0:0: [sdb] Attached SCSI removable disk |
| 07:12 |
jonpry |
[101397.411009] sd 15:0:0:0: Attached scsi generic sg2 type 0 |
| 07:13 |
WoZZeR |
bmnbmn: yes |
| 07:13 |
SleekX |
hello guys |
| 07:13 |
bmnbmn |
WoZZeR: i'll give it a shot right nowq |
| 07:13 |
SleekX |
good job getting sound on rhodium |
| 07:14 |
bzo |
WoZZeR: does the backlight turn off with nand boot? |
| 07:14 |
WoZZeR |
no |
| 07:14 |
bzo |
damn, guess the control to toggle the automatic backlight stuff has to be traced |
| 07:15 |
bmnbmn |
maybe one day jonpry will receive that broken tp2 |
| 07:15 |
jonpry |
i checked the mail on thursday. still no sign |
| 07:20 |
jonpry |
bzo: gsm/cdma? |
| 07:20 |
bzo |
my tp2? cdma |
| 07:21 |
jonpry |
i updated the port status |
| 07:22 |
bzo |
it must be maddening to realize that usb has been working all this time :) |
| 07:23 |
jonpry |
i only looked at it yesterday, so i'm not too upset |
| 07:28 |
jonpry |
were almost certainly just not turning on a clk/pll/vreg |
| 07:30 |
bzo |
you mean on usb insert detection? |
| 07:31 |
jonpry |
i think that works |
| 07:31 |
jonpry |
even when nothing is inserted, we should be able to configure the hardware. we get all these ulpi_write failures very early on |
| 07:32 |
bzo |
err, or better phrased: in response to usb insertion? |
| 07:32 |
bzo |
wince seems to be initing something after usb is inserted |
| 07:33 |
bzo |
which lets it work on haret boot |
| 07:33 |
jonpry |
its hard to tell what wince is configuring when. but at least for nand. we don't even have a configuration for a usb pll |
| 07:33 |
jonpry |
and there are 3 usb clocks. of which we only enable 2 |
| 07:35 |
bzo |
you think usb has its own pll, and doesn't derive from msm plls? |
| 07:35 |
bmnbmn |
WoZZeR: no irq 26 error, at the splash screen now |
| 07:35 |
WoZZeR |
we'll see if it works any better |
| 07:35 |
WoZZeR |
you may want a few reboots |
| 07:38 |
jonpry |
bzo: i'm pretty sure |
| 07:38 |
bmnbmn |
shit, frozen |
| 07:38 |
WoZZeR |
bmnbmn: give it time |
| 07:38 |
WoZZeR |
it can take a while |
| 07:38 |
bmnbmn |
ok |
| 07:38 |
WoZZeR |
especially on your slow ass card |
| 07:38 |
bmnbmn |
:P |
| 07:39 |
bmnbmn |
are you saying its gonna boot even with the splash animation frozen? |
| 07:39 |
WoZZeR |
is it completly frozen? |
| 07:39 |
bmnbmn |
yes |
| 07:40 |
WoZZeR |
when did it stop? |
| 07:40 |
bmnbmn |
uhh, like 3 minutes after it got to it? |
| 07:40 |
WoZZeR |
press the end key |
| 07:40 |
bmnbmn |
nothing |
| 07:40 |
WoZZeR |
ok, the it's probably not fixed |
| 07:40 |
jonpry |
bzo: there is stuff in upstream about changing ebi clock to 128mhz when enabling usb |
| 07:41 |
bmnbmn |
also i think something is wrong with your sed, because i didnt get that cmdline echo this time |
| 07:41 |
WoZZeR |
jonpry: I just realized that when I plug the usb in, the screen turns on |
| 07:41 |
WoZZeR |
but no hits on usb |
| 07:42 |
bzo |
jonpry: I don't think ebi clock is something we know how to control right now (or are certain about the relationship with the plls) |
| 07:42 |
jonpry |
until we get it to work in haret. i don't think it will work for you. maybe if you go into spl bootloader, connect usb, and then continue? |
| 07:43 |
jonpry |
wistilt2 said usb has a pll. but i'm not certain he looked it up |
| 07:44 |
bmnbmn |
WoZZeR: yeah got irq 26 error on 2nd boot |
| 07:44 |
WoZZeR |
ok |
| 07:44 |
bmnbmn |
so close |
| 07:44 |
WoZZeR |
then it didn't fix it |
| 07:44 |
bmnbmn |
yet so far |
| 07:44 |
jonpry |
bzo: doesn't this stuff work on kaiser/vogue? |
| 07:45 |
bzo |
I wouldn't think ebi has a direct relationship to usb. It's just one of the busses |
| 07:45 |
bzo |
yes, I'm pretty sure that stuff works on the vogue |
| 07:45 |
bzo |
now that I think about it, I seem to recall ebi being the external bus to the peripherals and such |
| 07:46 |
bzo |
perhaps it needs to be set to some minimum speed to insure there is enough bandwidth for usb |
| 07:47 |
bmnbmn |
WoZZeR: i noticed your modules is missing some stuff |
| 07:47 |
bmnbmn |
that the regular kernel modules has |
| 07:47 |
bmnbmn |
is that supposed to be that way? |
| 07:48 |
jonpry |
128 is probably slower than normal |
| 07:48 |
WoZZeR |
like what? |
| 07:48 |
bmnbmn |
like ramzswap |
| 07:48 |
WoZZeR |
it doesn't normally have that |
| 07:48 |
bmnbmn |
your ipv6 is like 4mb too |
| 07:49 |
bmnbmn |
while the original is like 300kb |
| 07:49 |
WoZZeR |
where's the source? |
| 07:49 |
bmnbmn |
http://zimages.googlecode.com/files/htc-msm-linux-20100830_123544-package.tar.bz2 |
| 07:49 |
WoZZeR |
no |
| 07:49 |
WoZZeR |
source |
| 07:49 |
bzo |
jonpry: I think 128 is the max, according to the acpuclock tables (though defined, I don't know that they actually get changed for us) |
| 07:50 |
jonpry |
this all seems silly. shouldn't hero have a working usb pll configuration for 7201a? |
| 07:50 |
bmnbmn |
WoZZeR: i have no clue |
| 07:51 |
bzo |
jonpry: the configuration is fine, it's the low level commands to effect the changes |
| 07:51 |
bzo |
the android amss seem to hide this kind of stuff in the a9 behind rpc calls |
| 07:52 |
bzo |
that's why the htc source releases have been of limited use in stuff like this |
| 07:52 |
jonpry |
but they would indicate if the pll needed to be configured or not |
| 07:53 |
bzo |
quite possibly - I don't think anyone has dug into that specifically |
| 07:54 |
bzo |
jonpry: are you aware of the msm7200 docs floating out there? |
| 07:55 |
jonpry |
i have seen a few. there is stuff on baidu |
| 07:55 |
jonpry |
seems really cool and all except that half the registers aren't visible from a11 |
| 07:56 |
bzo |
yeah and of course the registers are mapped slightly differently :/ |
| 07:56 |
bmnbmn |
WoZZeR: when im booting up, i do get a complaint about ramzswap, should it really not be there? |
| 07:56 |
WoZZeR |
I don't have issues with it |
| 07:56 |
bmnbmn |
original modules 2mb, yours are 17mb, thats quite a jump |
| 07:57 |
WoZZeR |
then build your own kernel and modules |
| 07:58 |
bmnbmn |
whoa whoa whoa |
| 07:58 |
WoZZeR |
http://balsat.hopto.org/linuxmsm.php |
| 07:58 |
WoZZeR |
those are just as big |
| 08:00 |
bmnbmn |
indeed |
| 08:22 |
bmnbmn |
gsensor_axis=2,1,3 msmvkeyb_toggle=off hw3d.force=0 panic=0 pm.sleep_mode=2 physkeyboard=rhod400 androidboot.serialno=R9PK6NCQR9T8 androidboot.serialno=HFMOXJTNS95J |
| 08:22 |
bmnbmn |
two serial numbers |
| 08:56 |
sykosoft |
wozzer: rhod400 on test5 |
| 09:56 |
DomZ |
~seen zego |
| 09:56 |
apt |
zego <~zego@122.129.199.154> was last seen on IRC in channel #htc-linux, 5d 23h 26m 29s ago, saying: 'CM6.0.0-RC3 based ROM for HD2: http://pastebin.com/0xSN3rbH (for those who care)'. |
| 10:50 |
prdiii |
seen darkstone |
| 10:51 |
prdiii |
~seen darkstone |
| 10:51 |
apt |
darkstone <dtrack@kimoshi.net> was last seen on IRC in channel #htc-linux, 1d 17h 38m 44s ago, saying: 'bbl D:'. |
| 10:55 |
Stianxxs |
will this work for HTC Touch Elfin? |
| 11:04 |
johnb81 |
~seen cotulla |
| 11:04 |
apt |
cotulla <~opera@nat004-252-205-109.tvoe.tv> was last seen on IRC in channel #htc-linux, 3d 21h 36m 45s ago, saying: 'it's my time :D'. |
| 12:07 |
camro |
phh: still here ? |
| 12:36 |
[Delenn] |
hey |
| 13:01 |
GNUtoo|laptop |
hi, |
| 13:02 |
[Delenn] |
ðø |
| 13:02 |
[Delenn] |
hi |
| 13:04 |
GNUtoo|laptop |
with the htcdream, under FSO or under android(refence ril based ril available here: http://gitorious.org/replicant/ril/trees/replicant/reference-ril ) we've got some CME error 30, but only in US or AU.so: |
| 13:04 |
GNUtoo|laptop |
*did other free ril had this issue |
| 13:05 |
GNUtoo|laptop |
on other phones too |
| 13:05 |
[Delenn] |
30 is NO NETWORK SERVICE? |
| 13:05 |
GNUtoo|laptop |
I'm aware only of the vilord's ril |
| 13:05 |
GNUtoo|laptop |
I know |
| 13:05 |
GNUtoo|laptop |
but that's not normal |
| 13:06 |
[Delenn] |
is it response to some command? |
| 13:07 |
GNUtoo|laptop |
yes |
| 13:07 |
[Delenn] |
which command? |
| 13:07 |
GNUtoo|laptop |
I think AT+COPS |
| 13:07 |
[Delenn] |
if u are using original htc ril, u got this also? |
| 13:07 |
GNUtoo|laptop |
no |
| 13:07 |
[Delenn] |
hm intersting |
| 13:08 |
GNUtoo|laptop |
and AT+CREG seem to have succedded |
| 13:09 |
[Delenn] |
hm |
| 13:09 |
GNUtoo|laptop |
I'll try to find the ril log,for now I looked at the fso log |
| 13:12 |
GNUtoo|laptop |
[Delenn], do you have a clue on what I should do now? |
| 13:13 |
[Delenn] |
hm wait |
| 13:13 |
GNUtoo|laptop |
ok |
| 13:13 |
GNUtoo|laptop |
mickeyl, is there at the moment so it might be a good oportunity |
| 13:13 |
[Delenn] |
maybe look to AT commands in logs which can change AT+COPS response |
| 13:14 |
[Delenn] |
also compare htc specific commands... if they exist there... |
| 13:16 |
GNUtoo|laptop |
ok |
| 13:17 |
[Delenn] |
there should be AT+COPS=XX l... compare them |
| 13:17 |
[Delenn] |
or maybe u send it too early? |
| 13:17 |
[Delenn] |
show log.. |
| 13:18 |
GNUtoo|laptop |
ok |
| 13:18 |
GNUtoo|laptop |
one sec |
| 13:19 |
GNUtoo|laptop |
http://pastebin.com/e0ArzDmV |
| 13:20 |
GNUtoo|laptop |
I can't find the ril log because it's old and now I use fso |
| 13:20 |
mickeyl |
this log does not show the actual AT communication |
| 13:21 |
GNUtoo|laptop |
ok |
| 13:21 |
mickeyl |
either you disabled the libfsotransport INFO logs or piping them elsewhere |
| 13:21 |
GNUtoo|laptop |
mickeyl, is there a way to be more verbose? |
| 13:21 |
GNUtoo|laptop |
because that is DEBUG I think |
| 13:21 |
[Delenn] |
look |
| 13:21 |
mickeyl |
[libfsotransport] |
| 13:21 |
mickeyl |
log_level = INFO |
| 13:21 |
GNUtoo|laptop |
ok |
| 13:21 |
[Delenn] |
2010-08-29T04:42:02.798309Z [DEBUG] QualcommHtcUnsolicitedResponseHandler : Dispatching AT unsolicited '+CREG', '2' |
| 13:21 |
[Delenn] |
2010-08-29T04:42:02.801391Z [DEBUG] QualcommHtcUnsolicitedResponseHandler : Dispatching AT unsolicited '+CGREG', '2' |
| 13:21 |
mickeyl |
etc. |
| 13:22 |
GNUtoo|laptop |
ok thanks a lot |
| 13:22 |
[Delenn] |
seems it's ready only at this moment |
| 13:22 |
GNUtoo|laptop |
I'll write a mail to the person |
| 13:22 |
[Delenn] |
it send response to AT+CREG |
| 13:22 |
[Delenn] |
while we ask for state before... |
| 13:22 |
[Delenn] |
I think it should wait until +CREG response appear in queue |
| 13:23 |
[Delenn] |
try to insert delay |
| 13:23 |
[Delenn] |
after CREG |
| 13:23 |
[Delenn] |
registration |
| 13:23 |
[Delenn] |
500 ms for example |
| 13:24 |
GNUtoo|laptop |
ok |
| 13:24 |
GNUtoo|laptop |
I got the log via mail and the tester in AU is not at the same timezone than me |
| 13:24 |
GNUtoo|laptop |
so sould we send him a binary? a patch? |
| 13:25 |
GNUtoo|laptop |
or commit the delay and see |
| 13:25 |
[Delenn] |
dunno.. for tester I think better binary.. |
| 13:25 |
GNUtoo|laptop |
anyway I'll ask him for more details |
| 13:25 |
[Delenn] |
because he will just install it and do work :) |
| 13:25 |
[Delenn] |
no f with compiling and etc |
| 13:25 |
GNUtoo|laptop |
ok |
| 13:26 |
[Delenn] |
different operators may need different registration time I guess |
| 13:27 |
GNUtoo|laptop |
mickeyl, where is libfsotransport? |
| 13:27 |
GNUtoo|laptop |
I do not have it in /etc/conf/freesmartphone/trout/ |
| 13:28 |
GNUtoo|laptop |
strange |
| 13:28 |
GNUtoo|laptop |
normally it should be in fsogsmd |
| 13:29 |
mickeyl |
it's in fsogsmd.conf |
| 13:29 |
mickeyl |
as section |
| 13:34 |
[Delenn] |
are 0x8000 before kernel busy by PPT? |
| 13:35 |
GNUtoo|laptop |
mickeyl, yes but it's not there in the htcdream's fsogsmd right? |
| 13:35 |
GNUtoo|laptop |
should the tester add it |
| 13:35 |
GNUtoo|laptop |
it's in default tough |
| 13:44 |
mickeyl |
if it's missing in htcdream's fsogsmd conf we should add it there |
| 13:45 |
GNUtoo|htcdream |
ok |
| 13:45 |
mickeyl |
done |
| 13:47 |
GNUtoo|laptop |
thanks |
| 13:47 |
GNUtoo|laptop |
my phone doesn't last a lot |
| 13:47 |
GNUtoo|laptop |
maybe less than 10min |
| 13:49 |
GNUtoo|laptop |
damn battery |
| 13:49 |
GNUtoo|laptop |
or whatever |
| 13:51 |
[Delenn] |
who remember leo mtype? |
| 13:51 |
[Delenn] |
oh it should be in startup.txt |
| 14:06 |
[Delenn] |
~seen NetRipper |
| 14:06 |
apt |
netripper is currently on #htc-linux. Has said a total of 65 messages. Is idling for 4d 1h 2m 7s, last said: 'well if you pass a nullpointer for the first atag structure, it will ignore all atags.. so it could boot without (but you'd need to at least set memory bank information in fixup())'. |
| 14:06 |
LeTama |
hi all |
| 14:06 |
[Delenn] |
ping MEM_LINUX_KERNEL_BASE |
| 14:06 |
[Delenn] |
hi |
| 14:07 |
[Delenn] |
ping NetRipper |
| 14:11 |
luc_ |
Guys what's that http://cotulla.pp.ru/leo/Android/Ctll_mach_msm_LSENSOR_SRC.tar.gz |
| 14:12 |
[Delenn] |
for devs only seems |
| 14:12 |
luc_ |
yes but what does it fix? |
| 14:20 |
gauner1986 |
hi |
| 14:20 |
LeTama |
hello gauner1986 |
| 14:20 |
[Delenn] |
hey\ |
| 14:23 |
gauner1986 |
news? |
| 14:23 |
gauner1986 |
:) |
| 14:24 |
camro |
hi gauner1986 |
| 14:25 |
LeTama |
not from me :) |
| 14:25 |
gauner1986 |
:) |
| 14:28 |
camro |
gauner1986: whichb dvice are you using ? |
| 14:28 |
gauner1986 |
hi camro |
| 14:28 |
gauner1986 |
leo |
| 14:28 |
camro |
any battery problems ? |
| 14:28 |
gauner1986 |
hm.. not |
| 14:28 |
gauner1986 |
no |
| 14:28 |
gauner1986 |
nothing serious |
| 14:29 |
camro |
so, not going down very fast ... |
| 14:29 |
camro |
ex: i tried now my blackstone: |
| 14:31 |
camro |
12:30 to 13:00 -> from 60% to 17% |
| 14:31 |
camro |
13:00 to 13:30 -> from 17% to 12% |
| 14:32 |
camro |
13:30 to 14:00 -> from 12% to 11% |
| 14:32 |
gauner1986 |
and what eats cpu? |
| 14:32 |
camro |
14:00 to 14:30 -> from 11% to 8% |
| 14:33 |
camro |
well - i use the systempaneel - for plotting (which needs a lot of power ..) |
| 14:33 |
camro |
and all the time - display on (so, that the power consume is normally high..) |
| 14:34 |
Hunterkll |
I asked a customer if he would prefer paper or plastic. He said "I'd like a little bit of paper and a little bit of plastic. I guess you could say I'm bi-sackual." |
| 14:34 |
camro |
the real thing, what i'm wondering ...the mV show me: 3690 on 8% ... |
| 14:34 |
camro |
i mean, that the display is wrong ... |
| 14:34 |
camro |
now - i'm waiting to see how it go'on .. |
| 14:35 |
camro |
then if it is at 5% - i will reboot to windows and will look what it shows me ... |
| 14:35 |
camro |
i'M sure, this is not normal .. |
| 14:36 |
camro |
must look, how i can ask the battery on console, that i can make ascript to write all 5min the battery + voltage in linux ... |
| 14:36 |
Hunterkll |
heh |
| 14:36 |
Hunterkll |
my HD2 shuts off at 4% in android |
| 14:36 |
camro |
this seems to be also not fully fixed ... |
| 14:36 |
Hunterkll |
and won't boot back up into winmo cuz it's really 0% :) |
| 14:36 |
camro |
well, i saw, that my device shutdow on 5% and on wm i have then ~40% left .. |
| 14:37 |
camro |
all guys are thinking, that the battery consume is high - i don'T think so ... |
| 14:37 |
camro |
the display or reading from driver is maybe wrong ... |
| 14:37 |
camro |
now it is 14:35 here - and battery is also 8% |
| 14:38 |
camro |
i must test this in windows with a battery plotter - and then again on android ... |
| 14:38 |
camro |
i'm sure, this is not normal |
| 14:39 |
camro |
research necceary for it ... |
| 14:41 |
Ondalf |
wish research done on kaiser too... it also suffers calibration problems and funny readings |
| 14:42 |
Hunterkll |
hd2 just off 4% for me heh |
| 14:44 |
camro |
well, will be the next in my list - the battery ... |
| 14:44 |
camro |
hopefully, phh will be back soon to implement all the fixes which he got ... |
| 14:46 |
Ondalf |
mines only 15% off |
| 14:47 |
camro |
5% now - and 3679mV |
| 14:47 |
[Delenn] |
not bad |
| 14:47 |
ninstaah |
hey guys, any of you got the msm72k datasheet (not the msm7200) ? |
| 14:47 |
[Delenn] |
but it depends on load very |
| 14:47 |
camro |
well, for nwo not the consume make me crazy .. .the curve, that is what is strange .. |
| 14:48 |
Ondalf |
72000? |
| 14:48 |
ninstaah |
ye Ondalf |
| 14:48 |
camro |
CPU ~25% |
| 14:48 |
Ondalf |
oh theres one like that too |
| 14:48 |
ninstaah |
I guess that msm72k = msm72000 right? |
| 14:49 |
camro |
nistaah: should be .. k=1000 |
| 14:49 |
Ondalf |
afaik yes, but where you need one? |
| 14:50 |
ninstaah |
get SET_ADDRESS and GET_.. to work right. |
| 14:51 |
[Delenn] |
what is it? |
| 14:51 |
ninstaah |
it's going to be a port of psfreedom for the desire (rly all htc androids with msm72k) |
| 14:52 |
LordCraigus |
bah |
| 14:52 |
LordCraigus |
tmo sux ass |
| 14:52 |
LordCraigus |
been on the phone with them for like 3 days straight |
| 14:52 |
gauner1986 |
markinus: current git doesnt compile |
| 14:52 |
LordCraigus |
only for them to find out that its a bad sim |
| 14:52 |
gauner1986 |
make[1]: *** No rule to make target `arch/arm/mach-msm/board-htcleo-ls.o', needed by `arch/arm/mach-msm/built-in.o'. Stop. |
| 14:53 |
[Delenn] |
heh |
| 14:54 |
gauner1986 |
obj-$(CONFIG_MACH_HTCLEO) += board-htcleo-proximity.o board-htcleo-leds.o board-htcleo-ls.o |
| 14:54 |
gauner1986 |
mhm |
| 14:54 |
gauner1986 |
strange |
| 14:58 |
DomZ |
can I know guys which android rom you are using atm ? |
| 14:59 |
DomZ |
on leo |
| 14:59 |
[Delenn] |
seems it's not finished |
| 15:02 |
[Delenn] |
ow |
| 15:10 |
[Delenn] |
somebody ever get such errors? |
| 15:10 |
[Delenn] |
http://pastebin.com/RNZV816T |
| 15:16 |
LeTama |
Delenn, nice! never had that. |
| 15:20 |
gauner1986 |
also havent seen that |
| 15:20 |
[Delenn] |
seems it received twice RPCROUTER_CTRL_CMD_HELLO |
| 15:28 |
[Delenn] |
it's not funny.. in linux any wrong action cause system crash :D |
| 15:28 |
DomZ |
LeTama does ppp works for you ? |
| 15:28 |
[Delenn] |
for example if u insert item in klist and something wrong, system will die |
| 15:33 |
[Delenn] |
[ 17.942145] VFS: Cannot open root device "<NULL>" or unknown-block(0,0) |
| 15:33 |
[Delenn] |
is it means that it can't find rootfs? |
| 15:33 |
gauner1986 |
yes |
| 15:35 |
gauner1986 |
ah ok |
| 15:35 |
gauner1986 |
there is no board-htcleo-ls |
| 15:35 |
gauner1986 |
.c |
| 15:35 |
gauner1986 |
;D |
| 15:35 |
[Delenn] |
;D |
| 15:35 |
[Delenn] |
:D |
| 15:36 |
[Delenn] |
hm [ 17.572549] mmc0: Slot eject status = 1 |
| 15:36 |
[Delenn] |
is 1 - missing? |
| 15:38 |
gauner1986 |
good question |
| 15:38 |
gauner1986 |
sounds logical |
| 15:38 |
gauner1986 |
:P |
| 15:38 |
gauner1986 |
but who knows |
| 15:38 |
[Delenn] |
seems yes |
| 15:39 |
DomZ |
ppp is worst then rmnet here :s |
| 15:40 |
DomZ |
than |
| 15:41 |
[Delenn] |
why? |
| 15:41 |
[Delenn] |
O_o |
| 15:41 |
gauner1986 |
" /home/gauner/hd2linux/markinus/linux-on-wince-htc/arch/arm/mach-msm/board-htcleo-ls.c:112: undefined reference to `microp_read_lightsensor_value' |
| 15:41 |
DomZ |
it is not working at all while rmnet was working without doing anuthing |
| 15:42 |
[Delenn] |
:D |
| 15:42 |
[Delenn] |
but RMNET lost upload packets |
| 15:42 |
gauner1986 |
then you're doing sth wrong |
| 15:43 |
Hunterkll |
wtf |
| 15:43 |
Hunterkll |
this laptop is a 1.3ghz pentium m |
| 15:43 |
Hunterkll |
the other low end laptop i used winxp on was a 366 pent 2 mobile (celeron with a different name) |
| 15:44 |
Hunterkll |
why is the 1.3 with almost 4x ram slower then the 366mhz? :( |
| 15:44 |
DomZ |
I've activate data and roaming set APN, is there something elso to do ? |
| 15:46 |
DomZ |
lol I've activate BT and my phone is stuck (I can't do nothing : buttons and touch screen is no more working :s) :> reboot |
| 15:48 |
gauner1986 |
arch/arm/mach-msm/board-htcleo-panel.c:517: error: static declaration of 'config_gpio_table' follows non-static declaration |
| 15:48 |
gauner1986 |
wtf |
| 15:48 |
gauner1986 |
^^ |
| 15:48 |
StummeJ |
Hey do you guys know what combiniation of haret will display gpio watch data? |
| 15:48 |
epsas |
hello |
| 15:49 |
epsas |
what graphical environment are people here using? |
| 15:49 |
gauner1986 |
mac os x |
| 15:49 |
epsas |
ahh - for their htc's, i meant. |
| 15:50 |
gauner1986 |
android |
| 15:50 |
gauner1986 |
:P |
| 15:51 |
epsas |
how does android draw? I assume that it doesn't use an X protocol |
| 15:52 |
epsas |
just curious -- a friend gave me a verizon xv6500, and I am going to put linux on it |
| 15:52 |
[Delenn] |
DomZ, make dump |
| 15:55 |
gauner1986 |
okay |
| 15:55 |
gauner1986 |
integrating lightsensor is somewhat more complex than i thought |
| 15:55 |
gauner1986 |
*damn* |
| 15:56 |
darkstone |
looks like the psfreedom port is coming along nicely |
| 15:56 |
darkstone |
:D |
| 15:56 |
gauner1986 |
darkstone: really? |
| 15:56 |
darkstone |
for MSM72k USB devices |
| 15:56 |
gauner1986 |
port to leo? |
| 15:56 |
gauner1986 |
:D |
| 15:56 |
darkstone |
which ours uses |
| 15:56 |
darkstone |
:D |
| 15:56 |
gauner1986 |
who is doing it? |
| 15:56 |
darkstone |
someone names DrMon |
| 15:56 |
darkstone |
#psfreedom |
| 15:56 |
gauner1986 |
link? |
| 15:56 |
gauner1986 |
ah |
| 15:56 |
Markinus |
Hi |
| 15:56 |
gauner1986 |
hey markinus |
| 15:57 |
Markinus |
I forgott to add the ls file to git |
| 15:57 |
gauner1986 |
there's some error in latest git |
| 15:57 |
gauner1986 |
yeah |
| 15:57 |
Markinus |
I'm not at home, I push it later, you can remove it from makefile |
| 15:57 |
gauner1986 |
ok |
| 15:58 |
[Delenn] |
hi Markinus |
| 15:58 |
Markinus |
[Delenn]: hi |
| 15:59 |
Markinus |
[Delenn]: I integrated all your changes in this repo |
| 15:59 |
[Delenn] |
but not lsensor? |
| 15:59 |
gauner1986 |
did you do big changes to the board-htcleo-ls.c file? |
| 15:59 |
gauner1986 |
i want to test it |
| 15:59 |
gauner1986 |
:) |
| 15:59 |
Markinus |
[Delenn]: I did it, but fogott to add on commit :( |
| 15:59 |
[Delenn] |
there still power problems I believe |
| 16:00 |
[Delenn] |
or u fixed it? |
| 16:00 |
gauner1986 |
power problems? |
| 16:00 |
[Delenn] |
lsensor power |
| 16:00 |
[Delenn] |
I didn't finish it |
| 16:01 |
Markinus |
[Delenn]: yes, I saw it, but we didn'T have a GPIO for this |
| 16:01 |
[Delenn] |
there no GPIO |
| 16:01 |
Markinus |
[Delenn]: is there maybe a vreg? |
| 16:01 |
[Delenn] |
microp power |
| 16:01 |
MN |
hi |
| 16:01 |
Markinus |
[Delenn]: ahh, ok |
| 16:01 |
[Delenn] |
seems shared with proximity |
| 16:01 |
[Delenn] |
also we need setup bkl right at system start |
| 16:01 |
Markinus |
[Delenn]: like evo then |
| 16:02 |
[Delenn] |
in my last tests u need to run kernel with android auto bkl off |
| 16:02 |
[Delenn] |
otherwise u will get shit |
| 16:02 |
Markinus |
ok, |
| 16:02 |
gauner1986 |
[delenn] last time i tested it was the other way round |
| 16:02 |
gauner1986 |
it needed to be enabled |
| 16:03 |
Markinus |
see you later, have o write on my masterthesis |
| 16:03 |
[Delenn] |
okbb |
| 16:03 |
[Delenn] |
http://htc-linux.org/wiki/index.php?title=Leo/Status |
| 16:03 |
Markinus |
bye |
| 16:03 |
[Delenn] |
is it outdated? |
| 16:04 |
MN |
just wondering has pm been completed? |
| 16:05 |
gauner1986 |
with the evo tree it mostly is i believe |
| 16:06 |
MN |
okay would that mean the device wouldnt heat up whilst using the phone |
| 16:06 |
MN |
with the evo tree |
| 16:06 |
[Delenn] |
Indicator LEDs Yes Yes, but not in Sense Builds |
| 16:06 |
Hunterkll |
MN: No |
| 16:06 |
[Delenn] |
is problem with sense fixed? |
| 16:06 |
Hunterkll |
It'll still heat up under load |
| 16:06 |
Hunterkll |
that's how computers do :) |
| 16:06 |
gauner1986 |
[delenn] in evo tree it is back again for some strange reason |
| 16:06 |
gauner1986 |
markinus doesnt have it |
| 16:07 |
[Delenn] |
as well |
| 16:07 |
[Delenn] |
camera is working fully? |
| 16:09 |
MN_ |
okay |
| 16:10 |
[Delenn] |
shit still [ 17.572335] mmc0: Slot eject status = 1 |
| 16:11 |
gauner1986 |
what are testing? |
| 16:11 |
[Delenn] |
mmc0 is sd card |
| 16:11 |
[Delenn] |
or wlan? |
| 16:11 |
gauner1986 |
isnt it mmcblk0 or sth? |
| 16:11 |
gauner1986 |
for sd card? |
| 16:12 |
[Delenn] |
FUCK |
| 16:13 |
gauner1986 |
? |
| 16:13 |
[Delenn] |
forgot one thing :D |
| 16:15 |
dcordes |
what's the status with 4g kernel ? |
| 16:15 |
dcordes |
gauner1986: what's missing 'feature' wise ? |
| 16:15 |
gauner1986 |
lsensor |
| 16:16 |
gauner1986 |
cotulla released the source |
| 16:16 |
dcordes |
gauner1986: did you put the master mode wifi ? |
| 16:16 |
gauner1986 |
no, i dont have commit rights |
| 16:16 |
gauner1986 |
it's working very good except of lsensor |
| 16:16 |
dcordes |
gauner1986: ok I think it will be nice to put it in linuxtogo |
| 16:16 |
gauner1986 |
have to do further tests though |
| 16:17 |
dcordes |
gauner1986: is anything working better than it does in htc-msm-2.6.32 ? |
| 16:17 |
gauner1986 |
hm |
| 16:17 |
gauner1986 |
did get bluescreen after boot again |
| 16:18 |
gauner1986 |
dcordes_ cant say for sure right now |
| 16:18 |
dcordes |
gauner1986: ok. whata bout cotulla ? |
| 16:18 |
gauner1986 |
dan said pm is way better |
| 16:18 |
dcordes |
despite from it being city in texas |
| 16:18 |
dcordes |
where's the code |
| 16:19 |
[Delenn] |
:D |
| 16:19 |
gauner1986 |
:D |
| 16:19 |
gauner1986 |
on cotullas site |
| 16:20 |
gauner1986 |
timer fix doesnt seem to be fully committed |
| 16:20 |
dcordes |
gauner1986: we should include the useful bits in htc-msm-2.6.32 |
| 16:21 |
dcordes |
gauner1986:but my suspicion he did not rebase again removes any motivation forme tm erge |
| 16:21 |
dcordes |
to merge |
| 16:21 |
dcordes |
and I won't have the time next weeks |
| 16:21 |
gauner1986 |
yeah.. there's some work to do |
| 16:22 |
[Delenn] |
dcordes, why you want to continue work under htc-msm-2.6.32? |
| 16:22 |
gauner1986 |
markinus already integrated some things |
| 16:22 |
dcordes |
[Delenn]: I do. in fact I didn't clone the evo branch yet. I will pull it once it is in linuxtogo kernel.git . |
| 16:23 |
[Delenn] |
gauner1986, which kernel you are using now? |
| 16:23 |
dcordes |
ali1234: ping |
| 16:24 |
gauner1986 |
[delenn] markinus latest git with the timer.c file from ltg |
| 16:24 |
gauner1986 |
:P |
| 16:24 |
dcordes |
gauner1986: what's the evo .git address and branchname ? |
| 16:24 |
[Delenn] |
can you make dmesg to pasterbin? |
| 16:24 |
gauner1986 |
[delenn] of something specific? |
| 16:24 |
[Delenn] |
just bootinh |
| 16:24 |
[Delenn] |
booting |
| 16:24 |
gauner1986 |
http://www.gitorious.com/linux-on-wince-htc/linux-on-wince-htc |
| 16:24 |
gauner1986 |
[delenn] sure |
| 16:25 |
gauner1986 |
master branch @ dcordes |
| 16:25 |
dcordes |
gauner1986: git:// ? |
| 16:25 |
gauner1986 |
git://gitorious.org/linux-on-wince-htc/linux-on-wince-htc.git |
| 16:25 |
dcordes |
gauner1986: thanks |
| 16:25 |
dcordes |
[[Kernel]] |
| 16:25 |
unilinky |
http://htc-linux.org/wiki/index.php?title=Kernel |
| 16:26 |
dcordes |
ali1234: I would like to copy git://gitorious.org/linux-on-wince-htc/linux-on-wince-htc.git master to git://git.linuxtogo.org/home/groups/mobile-linux/kernel.git |
| 16:26 |
[Delenn] |
I still have some unknown problems with mmc |
| 16:26 |
[Delenn] |
so your log gauner1986 may help me |
| 16:27 |
gauner1986 |
ok |
| 16:28 |
gauner1986 |
i'm just putting a new kernel and booting |
| 16:29 |
dcordes |
ali1234: but don't know the commands. can you help with that ? |
| 16:33 |
[Delenn] |
hm |
| 16:33 |
[Delenn] |
or |
| 16:33 |
[Delenn] |
[ 17.940911] VFS: Cannot open root device "<NULL>" or unknown-block(0,0) |
| 16:33 |
[Delenn] |
[ 17.947757] Please append a correct "root=" boot option; here are the available partitions: |
| 16:33 |
[Delenn] |
[ 17.954500] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) |
| 16:33 |
[Delenn] |
means that now initrd? |
| 16:33 |
[Delenn] |
*no |
| 16:34 |
LeTama |
Delenn, usually yes |
| 16:34 |
[Delenn] |
shit... why I mess with mmc all this time %) |
| 16:34 |
gauner1986 |
it means that it couldnt open the root device specified in root= |
| 16:35 |
LeTama |
Delenn: for initrd, you usually have a message earlier during boot |
| 16:35 |
[Delenn] |
gauner1986, so what about dmesg log? |
| 16:35 |
gauner1986 |
currently stuck at haret handover |
| 16:36 |
gauner1986 |
for whatever reason.. |
| 16:36 |
gauner1986 |
i'm putting old zimage from linuxtogo and see |
| 16:37 |
[Delenn] |
hm |
| 16:37 |
[Delenn] |
is kernel decompressed inplace? |
| 16:40 |
[Delenn] |
lol working |
| 16:40 |
[Delenn] |
was too small initrd offset |
| 16:43 |
[Delenn] |
lol audio died |
| 16:43 |
gauner1986 |
no its zImage |
| 16:44 |
gauner1986 |
ah ok |
| 16:44 |
gauner1986 |
:P |
| 16:44 |
gauner1986 |
i had to move it to 00c |
| 16:44 |
gauner1986 |
to have it working |
| 16:46 |
[Delenn] |
so markinus kernel is good? |
| 16:48 |
gauner1986 |
can tell if i get it to boot again |
| 16:48 |
gauner1986 |
:D |
| 16:49 |
[Delenn] |
oki |
| 16:51 |
gauner1986 |
ok, up again |
| 16:51 |
gauner1986 |
do you still need dmesg? |
| 16:51 |
[Delenn] |
no |
| 16:51 |
[Delenn] |
working already |
| 16:52 |
gauner1986 |
ok |
| 16:57 |
gauner1986 |
cam works |
| 17:05 |
[Delenn] |
really time to abandon this shit |
| 17:06 |
gauner1986 |
everything is working great so far |
| 17:07 |
gauner1986 |
except the leds |
| 17:07 |
gauner1986 |
for some strange reason |
| 17:08 |
[Delenn] |
oh |
| 17:08 |
gauner1986 |
someone here on darkstones sense v2? |
| 17:09 |
gauner1986 |
or is there a recent version of liblights somewhere? |
| 17:09 |
[Delenn] |
again not compatibles? |
| 17:09 |
[Delenn] |
:( |
| 17:09 |
darkstone |
liblights is very odd on my v2 version xD |
| 17:09 |
darkstone |
touch the status bar |
| 17:09 |
darkstone |
and the orange led will flash |
| 17:10 |
DuperMan |
that's kinda nifty oO |
| 17:10 |
[Delenn] |
lol |
| 17:10 |
gauner1986 |
darkstone |
| 17:10 |
gauner1986 |
can you upload me the version of liblights youre using? |
| 17:10 |
darkstone |
in v2? |
| 17:10 |
darkstone |
ok |
| 17:11 |
gauner1986 |
ah |
| 17:11 |
gauner1986 |
got a sleep of death |
| 17:11 |
gauner1986 |
:( |
| 17:11 |
darkstone |
i think this is related to kernel |
| 17:11 |
darkstone |
i need to revert wifi changes |
| 17:11 |
darkstone |
and test |
| 17:11 |
gauner1986 |
ah no |
| 17:11 |
gauner1986 |
it just took long to wake up again |
| 17:11 |
darkstone |
ah |
| 17:12 |
darkstone |
could be my 245mhz script |
| 17:12 |
darkstone |
xd |
| 17:12 |
gauner1986 |
i'm not using v2 |
| 17:12 |
gauner1986 |
:) |
| 17:12 |
DuperMan |
that's a common new thing, delayed wake ups/ts freeze on wakeup |
| 17:12 |
[Delenn] |
shit phone not working |
| 17:12 |
gauner1986 |
ok |
| 17:12 |
darkstone |
nvm then |
| 17:12 |
darkstone |
xD |
| 17:12 |
gauner1986 |
yeah |
| 17:12 |
gauner1986 |
ts freezes |
| 17:13 |
noellenchris |
hello all |
| 17:13 |
darkstone |
gauner1986: http://www.multiupload.com/R9FMXSGT84 |
| 17:13 |
DuperMan |
gauner1986: hardly dev talk, but using interactive governor "healed" it for me:) |
| 17:13 |
gauner1986 |
<3>[ 614.043356] msm_i2c msm_i2c.0: error, status 63c8 (5C) |
| 17:13 |
gauner1986 |
<3>[ 614.043425] msm_i2c msm_i2c.0: Error during data xfer (-5) (5C) |
| 17:13 |
gauner1986 |
<3>[ 614.043574] msm_i2c msm_i2c.0: error, status 43c8 (34) |
| 17:13 |
gauner1986 |
<3>[ 614.043624] msm_i2c msm_i2c.0: Error during data xfer (-5) (34) |
| 17:13 |
gauner1986 |
<6>[ 614.044900] [R] late_resume end |
| 17:14 |
gauner1986 |
wtf? |
| 17:14 |
gauner1986 |
^^ |
| 17:14 |
gauner1986 |
interactive governor? |
| 17:14 |
gauner1986 |
thanks darkstone |
| 17:14 |
darkstone |
np |
| 17:15 |
gauner1986 |
wtf |
| 17:15 |
DuperMan |
gauner1986: in setCPU. |
| 17:15 |
gauner1986 |
you have lights.qsd8k |
| 17:15 |
gauner1986 |
and lights.htcleo |
| 17:15 |
gauner1986 |
which one to use? |
| 17:15 |
noellenchris |
quick ? for gauner1986: PPP and Wifi, on the different builds PPP works great until I turn on Wifi, then when data reconnects it locks android. I reboot and eventually corrupts data.img. Is this known already? thanks.... |
| 17:16 |
darkstone |
gauner1986: qsd8k |
| 17:16 |
gauner1986 |
ok |
| 17:17 |
darkstone |
need to clean up my builds |
| 17:17 |
darkstone |
but im lazy :D |
| 17:17 |
darkstone |
too busy watching psfreedom progress |
| 17:17 |
darkstone |
hurr |
| 17:17 |
gauner1986 |
noellenchris: yes.. there are some problems with ppp crashes atm |
| 17:17 |
gauner1986 |
theres not much in channel now |
| 17:17 |
gauner1986 |
but drmon told me that theres some glitch |
| 17:17 |
gauner1986 |
android usb gadget needs to be disabled |
| 17:17 |
darkstone |
yeah |
| 17:17 |
gauner1986 |
so its not usable |
| 17:18 |
gauner1986 |
as everyday kernel |
| 17:18 |
darkstone |
https://spreadsheets.google.com/ccc?key=tarAC83Wjsq2QZvP4iR5mlw&hl=en#gid=0 |
| 17:18 |
darkstone |
yup |
| 17:18 |
DuperMan |
the mixed open sourced and closed source nature of android's giving me a headache:( |
| 17:19 |
[Delenn] |
so PPP still sucks? |
| 17:19 |
[Delenn] |
:D |
| 17:19 |
DuperMan |
[Delenn]: you're being overly technical, but yes. :-P |
| 17:19 |
[Delenn] |
))) |
| 17:19 |
gauner1986 |
have to watch how it is on the new kernel |
| 17:19 |
noellenchris |
thanks for the info, i was going nuts at least till i figured out it happens when I toggled wifi, love the ppp , and miss it. I've been playing with the FRG33 , it's not bad....w/out ppp, slow uploads..:( |
| 17:21 |
gauner1986 |
still lights not working |
| 17:21 |
noellenchris |
stepped out of the Ferrari and back onto the horse...lol, that's what it feels like going from ppp to rmnet....hehe |
| 17:21 |
gauner1986 |
strange |
| 17:21 |
DuperMan |
darkstone, gauner1986: may I be privy to how the new Evo based branch is expected to supercede current kernel trees? |
| 17:22 |
gauner1986 |
it only works on device wakeup |
| 17:23 |
darkstone |
i won't be able to answer that |
| 17:23 |
gauner1986 |
[delenn] is it possible to modularize the msm usb stuff? instead of keeping it as embedded kernel module? |
| 17:23 |
darkstone |
D: |
| 17:23 |
[Delenn] |
yes maybe... |
| 17:24 |
[Delenn] |
we can build all as modules in theory |
| 17:24 |
domi007 |
hi guys, no sleep-wake up for rhobuntu yet? |
| 17:24 |
gauner1986 |
we're facing this problem with the ps3 mod |
| 17:24 |
[Delenn] |
which problem? |
| 17:24 |
gauner1986 |
android usb stuff has to be disabled |
| 17:24 |
gauner1986 |
for the mod to work |
| 17:25 |
[Delenn] |
hm libhtc_ril won't load without qmi drivers? |
| 17:25 |
gauner1986 |
it loads without rmnet support |
| 17:26 |
[Delenn] |
but not without qmi? |
| 17:26 |
gauner1986 |
dunno |
| 17:27 |
gauner1986 |
we'd need to modularize the android usb gadget |
| 17:27 |
gauner1986 |
currently it doesnt allow that |
| 17:39 |
gauner1986 |
g sensor doesnt work |
| 17:56 |
dcordes |
gauner1986: does it have ehci_hcd, backlight API backlight driver, etc ? |
| 17:59 |
gauner1986 |
dunno.. dont think it has everything yet |
| 18:04 |
gauner1986 |
markinus: there seems to be a problem with sensors |
| 18:05 |
gauner1986 |
maybe its microp or sth |
| 18:05 |
gauner1986 |
g sensor doesnt work |
| 19:01 |
emwe |
hiho |
| 19:01 |
camro |
hi |
| 19:02 |
emwe |
rrew, around? |
| 19:11 |
jonpry |
~seen psycho_ |
| 19:11 |
apt |
psycho_ <~psycho@2001:470:c25d:4:4a5b:39ff:fe62:13fd> was last seen on IRC in channel #htc-linux, 8d 16h 20m 46s ago, saying: 'there is actually broadcom code in there'. |
| 19:11 |
WoZZeR_ |
hey jonpry |
| 19:11 |
jonpry |
hi wozzer |
| 19:12 |
jonpry |
~seen psycho |
| 19:12 |
apt |
psycho <~psycho@2001:470:c25d:4:4a5b:39ff:fe62:13fd> was last seen on IRC in channel #htc-linux, 6d 15h 58m 2s ago, saying: 'gentoofr34k: I will be on here tomorrow after work, hop on if you get a chance and I will let you know'. |
| 19:12 |
jonpry |
whats up? |
| 19:12 |
WoZZeR_ |
not a whole lot |
| 19:12 |
jonpry |
stoked for the job? |
| 19:12 |
WoZZeR_ |
one ob bmn's last statement's was that he got 2 serialno's |
| 19:13 |
WoZZeR_ |
yeah, I'm excited, don't start till tues though |
| 19:13 |
jonpry |
i saw that. weird maybe the plan works on his rootfs |
| 19:13 |
WoZZeR |
but he should be using a stock one |
| 19:13 |
jonpry |
what company are you working for? |
| 19:13 |
WoZZeR |
Market Strategies |
| 19:15 |
WoZZeR |
jonpry: can you think of any reason that 2 of the exact same files would be different on different phones? |
| 19:15 |
WoZZeR |
if he was using the correct rootfs? |
| 19:15 |
jonpry |
not really |
| 19:15 |
jonpry |
ever figure out the sd problem? |
| 19:16 |
WoZZeR |
also, tilt2's get stuck at SMD ) |
| 19:16 |
WoZZeR |
SMD 0 |
| 19:16 |
WoZZeR |
jonpry: no, not really |
| 19:17 |
jonpry |
pohket tried? |
| 19:17 |
WoZZeR |
new vzw phone pictured, I hope they change the keyboard |
| 19:17 |
WoZZeR |
http://www.engadget.com/2010/09/04/htcs-dual-mode-cdma-gsm-android-slider-hits-the-fcc-decked-out/ |
| 19:17 |
WoZZeR |
yeah |
| 19:17 |
WoZZeR |
any someone else in the forums |
| 19:17 |
WoZZeR |
http://forum.ppcgeeks.com/showpost.php?p=1935697&postcount=138 |
| 19:19 |
jonpry |
how does keyboard detection work if it freezes on smd0? |
| 19:20 |
WoZZeR |
no idea if it does |
| 19:20 |
WoZZeR |
but it doesn't crash there |
| 19:22 |
jonpry |
thats right before checking for cdma |
| 19:23 |
jonpry |
er nm, right after |
| 19:23 |
jonpry |
then it does rpc. that is basically first use of rpc |
| 19:23 |
WoZZeR |
so it could be CDMA detection? |
| 19:25 |
jonpry |
i really don't know. have to add printk's to code in that vicinity until you can actually figure out where it is freezing |
| 19:33 |
camro |
hi jonpry ... |
| 19:34 |
camro |
how is going - and whats about the sound stuff ? any news - are you almost ready with the changes ? |
| 19:34 |
jonpry |
hi camro. did you fix any of your issues? |
| 19:35 |
camro |
no, as i told you .. if you are changing some parts on sound ... i'll be waiting that your are finished with it ... |
| 19:35 |
Hunterkll |
I demand fish. |
| 19:35 |
camro |
well, for now testing with the battery ... because it shows only bullsh..t |
| 19:35 |
Hunterkll |
SUshi and kefir dispenser on HD2 |
| 19:37 |
camro |
jonpry: but maybe you can push me in the right direction ... for battery - there are some parts which confusing me ... |
| 19:39 |
camro |
jonpry - do you know the mem where the kernel read the battery values ? |
| 19:41 |
jonpry |
camro: i have a couple of ideas about battery |
| 19:41 |
camro |
yes ? |
| 19:41 |
jonpry |
+0xfc110 2b batt ID |
| 19:41 |
jonpry |
+0xfc112 2b batt Temperature |
| 19:41 |
jonpry |
+0xfc114 2b batt Voltage |
| 19:41 |
jonpry |
+0xfc116 2b batt Charge |
| 19:41 |
jonpry |
+0xfc118 2b batt Discharge |
| 19:41 |
camro |
example: = |
| 19:42 |
camro |
this is on rhod ? |
| 19:42 |
jonpry |
yeah. smem_base + those values |
| 19:42 |
camro |
http://www.ibt-power.com/Battery_packs/Li_Ion/Li_Ion_DiscGph.JPG |
| 19:42 |
jonpry |
look at the code for gnome batter applet |
| 19:42 |
jonpry |
it is much different than our stuff. it doesn't assume that battery state of charge is monotonic with battery voltage |
| 19:43 |
jonpry |
it calculates all kinds of cool stuff. like certainty of state of charge |
| 19:43 |
camro |
the charge and discharge also ? |
| 19:43 |
camro |
so, the battery volume is not in the smem ? |
| 19:44 |
jonpry |
battery volume? |
| 19:44 |
camro |
+0xfc116 and +0xfc118 <---- is really the consume in mA ? |
| 19:45 |
camro |
how is now the calc of battery % ? |
| 19:45 |
jonpry |
i don't know what the units are. probably fixed point milliamps |
| 19:45 |
camro |
i can't find anything in smem ? |
| 19:45 |
camro |
(i mean for the %) |
| 19:46 |
camro |
so, is this maybe a constant calc, which is wrong ? |
| 19:46 |
jonpry |
like i was saying, the state of charge is a really complicated subject. you need to record history on what happens to the battery under different charge/discharge scenarios in order to develop curves about the battery |
| 19:46 |
camro |
on wm also ? |
| 19:47 |
jonpry |
i don't know what wm does. but seeing as how the meter seems to work. it probably does what needs to be done |
| 19:47 |
bzo |
there's been a bunch of research on battery charge in the past |
| 19:47 |
camro |
can't belive - if i load my battery to 50% and flash image, then i also have 50% in wm |
| 19:47 |
bzo |
wm does have a stateful battery calculation |
| 19:47 |
camro |
or is this only be calc by the mV ? |
| 19:48 |
camro |
bzo: what happen with it ? |
| 19:48 |
bzo |
someone tried to disassemble the wm code, but it has a lot of math so it was difficult |
| 19:48 |
jonpry |
afaik the math is in gnome battery applet |
| 19:49 |
bzo |
jonpry: yeah, math is probably similar. that could be a good find |
| 19:49 |
bzo |
does it use current over time combined with the voltage calcs? |
| 19:49 |
Hunterkll |
bzo it needs to be done |
| 19:49 |
camro |
hmm - ex: on android batt is 5% at 3720mV <- can only be a mistake .. |
| 19:50 |
Hunterkll |
if you want accuracy, you need to do the wm code |
| 19:51 |
jonpry |
i think there are a lot of integrals like that. in general you need to calculate the battery impedance. which itself is a nonlinear function over the state of charge and temperature |
| 19:51 |
camro |
i would say: ex: 4,2 full (should be in a range of 100 - 90%) -> 3,2 1% |
| 19:51 |
jonpry |
but if you can plot some points with certainty, then you can interpolate with some efficacy |
| 19:51 |
kenkern |
hi |
| 19:52 |
bzo |
problem with li-ion is that the discharge curve is so flat for most of the capacity |
| 19:52 |
bzo |
that's why the integral calc is needed |
| 19:52 |
camro |
don't know the logic now, but it is not calc in reation to this table ...http://www.ibt-power.com/Battery_packs/Li_Ion/Li_Ion_DiscGph.JPG |
| 19:52 |
bzo |
also, voltage fluctuates a lot based on load |
| 19:52 |
Hunterkll |
at least it's not like a damn mercury battery |
| 19:52 |
kenkern |
is it possible to build a kernel for the hero with usb_gadget ? |
| 19:53 |
Hunterkll |
which maintains almost perfect voltage regulation until it drops to 0 |
| 19:53 |
camro |
but the logic for it should be in the kernel - correct ? |
| 19:53 |
Hunterkll |
correct, so it reports the correct data via /sys |
| 19:53 |
bzo |
one thing that does need to be reversed or calculated though are the battery charactoristics |
| 19:53 |
michyprima |
camro: investingating when I developed battery driver patch for supporting 2300mAh on leo told me that not always battery gets at 4,2v when full |
| 19:53 |
bzo |
Captnoord found those battery stats for the diamond/raph, but it has not been done for the rhodium |
| 19:54 |
camro |
well, yes, google is full of a lot engineers which are did a lot of tests ... |
| 19:54 |
camro |
the voltage is onyl 1 component ... |
| 19:54 |
Hunterkll |
reverse the wm code |
| 19:55 |
camro |
but - we must live with the values what we get and build a ogic for it ... |
| 19:55 |
Hunterkll |
as aggrivating as it is, it is the easiest way compared to the others |
| 19:55 |
camro |
well, a small problem is: if you have big battery packs ... |
| 19:55 |
camro |
is it possible to read the batt size ? smem ? |
| 19:55 |
bzo |
yeah, reversing is the best way to get the data, since it is probably the manufacturer stats |
| 19:55 |
michyprima |
you have battery problems on rhodium if I understood correctly? |
| 19:55 |
jonpry |
i don't see why kernel doesn't report the measured values. do everything else in userland |
| 19:55 |
camro |
but what i readed .. there is no access to this infos |
| 19:56 |
jonpry |
thats how it works on pc |
| 19:58 |
camro |
example: my battery was for 1hr: > 88% full, and had 4,225mV |
| 19:59 |
camro |
now it has: 92% and is at 4,222 :) |
| 19:59 |
camro |
(so the logic to wm is also not the best :) |
| 20:00 |
jonpry |
lol |
| 20:00 |
camro |
anyhow ... |
| 20:00 |
bzo |
no battery calc is perfect, wm is probably as good as it gets |
| 20:00 |
jonpry |
that is the whole point of this conversation. mv is not one to one with state of charge |
| 20:00 |
camro |
so: we get only the voltage and temp ... |
| 20:00 |
camro |
also we can see it is charging or discharging ... |
| 20:01 |
camro |
do you know, when does the battery stop with charging = |
| 20:01 |
camro |
is this electric regulation in battery when it stop with the charging ? |
| 20:01 |
bzo |
yes, the battery has its own circuit to control charging |
| 20:02 |
bzo |
that is the most reliable way to keep it from exploding |
| 20:02 |
camro |
so, do we see, when the battery tell to system: hey i'm full ? smem ? |
| 20:03 |
camro |
maybe the charging on adr: +0xfc116 2b batt Charge |
| 20:03 |
bzo |
the battery doesn't tell the phone. It just stops accepting a charge current |
| 20:04 |
bzo |
the phone will see that the charge current is a trickle state and can report that the battery is full |
| 20:04 |
camro |
so, in my example: the battery seems to befull (voltage 4,225), but wm mean it is not full and load it (related to the wm logic) |
| 20:04 |
camro |
in realitiy - the battery doesn't charge ... this could be possible - correct ? |
| 20:05 |
jonpry |
voltage can / will fall during charge |
| 20:05 |
camro |
bzo: trickle state ? |
| 20:05 |
bzo |
small amount of current to maintain charge |
| 20:06 |
camro |
so, where do the system know if the battery doesn't accept charging ? |
| 20:07 |
bzo |
look, the phone doesn't definately know what the battery is doing. It just observes voltage and current over time and gives an estimate of status |
| 20:07 |
camro |
(maybe only a smal logic like: if it the smem shows that it does charging and the volateg doesn't change -> then it could be full |
| 20:08 |
camro |
well, today i made a lot of tests ... |
| 20:08 |
jonpry |
one would think that if the battery turns itself off, it must be a smart battery. and should be able to report SOC |
| 20:09 |
camro |
ex: in -1hrs from 100 - 5% / 1 hr later 15% / 1 hr later 6% / 1hr later 5% |
| 20:09 |
camro |
ups |
| 20:10 |
camro |
ex: in -1hrs from 100 - 55% / 1 hr later 15% / 1 hr later 6% / 1hr later 5% |
| 20:10 |
camro |
do you know where the logic is in kernel ? |
| 20:11 |
bzo |
jonpry: i don't think the battery charge circuit is that sophisticated |
| 20:12 |
bzo |
camro: htc_battery_smem.c |
| 20:12 |
jonpry |
turning off charging is about as complicated as it gets |
| 20:12 |
camro |
there i am ... |
| 20:13 |
camro |
but haven't seen there the logic calc for capacity |
| 20:13 |
bzo |
I have a feeling that detecting full charge is a simpler than capacity. Probably a lookup table with voltage/current/temp |
| 20:14 |
camro |
i would also think so |
| 20:14 |
bzo |
from what I recall those charge ICs cost only like a quarter |
| 20:15 |
camro |
well, what we would like to have is another story ... |
| 20:15 |
camro |
let us do as good as possible ... |
| 20:17 |
jonpry |
bzo: if they use such a dumb controller. why use one at all? just use circuits on the phone |
| 20:17 |
camro |
bzo: thanks found it ... |
| 20:17 |
camro |
static int htc_get_batt_info(struct battery_info_reply *buffer) |
| 20:18 |
camro |
i see, there are a lot of things missing which are not think about ... |
| 20:18 |
camro |
hehe |
| 20:18 |
bzo |
jonpry: so they have full control over it and it is tightly coupled with the battery charactoristics |
| 20:18 |
bzo |
a lot of bad press about exploding batteries :) |
| 20:21 |
camro |
why - the lock logic is in the battery included ... |
| 20:22 |
bzo |
which lock? |
| 20:22 |
camro |
(closing the charge process) |
| 20:23 |
camro |
good - at first i will make a list for charging (ex: 4,2 = 95 - 100 /......./ 3,2 = 1%) |
| 20:23 |
camro |
then i will look to implement such kind aof curve .. |
| 20:23 |
[Delenn] |
what's up? |
| 20:24 |
camro |
[2010-09-05 19:41] <jonpry> +0xfc110 2b batt ID << this is the tope of battery ... |
| 20:24 |
camro |
and they have different curve models ? correct ? |
| 20:25 |
rrew |
emwe: hey m8 :), i need a little bit of support again :( |
| 20:25 |
MN |
Hi |
| 20:25 |
emwe |
rrew, then we rather talk in private ;) |
| 20:25 |
eugenb |
does already someone implementing the changes of cotulla's tree?^^ |
| 20:26 |
jonpry |
you could probably do something with the battery capacity. but imho it is a waste of time. best to calculate the capacity because they are all going to be different anyways. you might get a 1300 with way bigger capacity than an 1800 |
| 20:26 |
MN |
what are the changes in cotulla's latest tree |
| 20:27 |
eugenb |
lsensor xD |
| 20:27 |
eugenb |
other changes i don't know :D |
| 20:28 |
MN |
okay |
| 20:29 |
[1]Hunterkll |
can you build a bootable kernel using cygwin |
| 20:29 |
[1]Hunterkll |
i do not remember this info |
| 20:30 |
camro |
jonpry, the main problem for me is: its hsows 15 -5% , but if i go to check the voltage - it shows me: 3,7, and then it con't be, that the battery is almost empty and i get warnings - or maybe it shutdown the system ...it is complete |
| 20:30 |
camro |
then i boot to WM, and i have there 35% left ... |
| 20:31 |
jonpry |
what is your point. if you don't want to write and algorithm to calculate the state of charge, then you aren't going to make it work |
| 20:32 |
Hunterkll |
camro, reverse the winmo driver |
| 20:32 |
Hunterkll |
most painfree way of doing it |
| 20:32 |
camro |
sure, you must sit down at logic for it ... |
| 20:32 |
WoZZeR |
if you can get a hold of the nuerom guy, he may have some code for TP2 battery, may be similar to all |
| 20:32 |
camro |
hunterkill: do you have the wm driver ? |
| 20:33 |
[Delenn] |
are there ds2746? |
| 20:34 |
[Delenn] |
or delta-adc method? |
| 20:34 |
jonpry |
wozzer: who is nuerom |
| 20:34 |
WoZZeR |
http://www.nuerom.com/BlogEngine/ |
| 20:34 |
camro |
nuerom - also never heard before ... |
| 20:34 |
WoZZeR |
no activity for a while |
| 20:34 |
WoZZeR |
Open Sourcing, nueBattery2 For TP2 CDMA, v1.2 Build 122 |
| 20:35 |
bzo |
he's more of a wince guy, though he did hang out here for a while a few months ago |
| 20:35 |
jonpry |
there is always that weird dex command |
| 20:35 |
Hunterkll |
camro, i assumed if you had phone you had driver ;) |
| 20:35 |
MN |
~seen Cotullav20 |
| 20:35 |
apt |
cotullav20 <~opera@nat004-252-205-109.tvoe.tv> was last seen on IRC in channel #htc-linux, 2d 7h 35m 49s ago, saying: 'hey'. |
| 20:35 |
bzo |
jonpry: ? |
| 20:35 |
camro |
don'T know which dll is related for it ... |
| 20:36 |
camro |
but the math logic to get .. maybe is not so easy ... |
| 20:36 |
flagada |
cotulla remove last file ? |
| 20:36 |
jonpry |
0x8a get battery data |
| 20:36 |
bzo |
I believe all that does is update battery smem |
| 20:37 |
jonpry |
could be |
| 20:37 |
eugenb |
jonpry: yeah, he removed it |
| 20:38 |
eugenb |
jonpry: sry, wrong person selected :D @flagada, yes he removed it xD |
| 20:38 |
camro |
well, that the values are correct - i belive it ... |
| 20:38 |
camro |
there is only the voltage and temp which is the base .. |
| 20:39 |
camro |
then ...battery_table* |
| 20:39 |
camro |
anyhow ... for now i log the next day in winmo all data, then i will try to find alogic for it ... |
| 20:44 |
Microang |
~seen cotulla |
| 20:44 |
apt |
cotulla <~opera@nat004-252-205-109.tvoe.tv> was last seen on IRC in channel #htc-linux, 4d 7h 17m ago, saying: 'it's my time :D'. |
| 20:47 |
Jellenl |
Do I have to format my SD-Card before installing the .cab? |
| 20:47 |
Hunterkll |
lol |
| 20:50 |
Hunterkll |
i love how people come looking for cotulla and leave when they realize he's gone 'awol' lol |
| 20:51 |
Jamesws |
Hello |
| 20:53 |
MN |
Hi |
| 20:53 |
Jamesws |
I've recently ordered a Htc uni |
| 20:54 |
Jamesws |
Looking forward to doing some development here |
| 21:01 |
DuperMan |
Hunterkll: away while on lunch?:D |
| 21:04 |
Hunterkll |
lol |
| 21:04 |
Hunterkll |
away without leave |
| 21:04 |
DuperMan |
hehehe... SIT! CODE! |
| 21:04 |
DuperMan |
SCHNELL! |
| 21:05 |
DuperMan |
j/k |
| 21:18 |
camro |
[2010-09-05 19:53] <michyprima> camro: investingating when I developed battery driver patch for supporting 2300mAh on leo told me that not always battery gets at 4,2v when full << COULD THIS BE TRUE ? |
| 21:25 |
Hunterkll |
camro, it can be true |
| 21:25 |
Hunterkll |
not every battery will stabilize at or above 4.2 |
| 21:27 |
camro |
hm ...well, but the logic can be: if it charge with 100ma and the voltage will not been go higher in 5min, then the battery can be set to full .. |
| 21:27 |
Hunterkll |
no |
| 21:27 |
Hunterkll |
get the logic from winmo |
| 21:27 |
Hunterkll |
to prevent damage |
| 21:27 |
camro |
if there are no values left in the database, it can be set only on voltage |
| 21:27 |
camro |
winmo logic is a prevent to damage *G |
| 21:28 |
camro |
what should be happen ? |
| 21:28 |
camro |
if i log the whole battery values in winmo, then build a logic on it .. |
| 21:28 |
camro |
thats it |
| 21:29 |
camro |
for now - i would need a lot of logs from winmo side .. |
| 21:30 |
Hunterkll |
why bother? just reverse the winmo code |
| 21:30 |
Hunterkll |
that's how the HD2 bat driver came around |
| 21:31 |
Jamesws |
How did you guys impelement power management in your ports? |
| 21:32 |
Jamesws |
Turning the screen off, cpu scaling ect. |
| 21:34 |
camro |
there is a tool for winmo .. -> battlog |
| 21:40 |
Hunterkll |
Jamesws, same way it works on other linux devices? |
| 21:40 |
MN |
Hello |
| 21:41 |
Hunterkll |
camro: I just feel like it'd be easier to trace assembly then do all the work you are doing... we have some very good arm asm gurus here |
| 21:41 |
Hunterkll |
toss it at cotulla lol |
| 21:41 |
Jamesws |
Meh, I've been porting Ubuntu to a device and the battery lasts like an hour :/ |
| 21:41 |
Hunterkll |
well |
| 21:41 |
Jamesws |
I guess I'll have to find out how others do it |
| 21:41 |
Hunterkll |
get cpu scaling working |
| 21:41 |
Hunterkll |
backlight control |
| 21:41 |
Hunterkll |
etc :) |
| 21:41 |
Hunterkll |
it's all linux kernel drivers |
| 21:41 |
Jamesws |
I thought it would just work |
| 21:41 |
Jamesws |
Must be kernel related |
| 21:41 |
Hunterkll |
... what device? |
| 21:42 |
Hunterkll |
if the kernel dosn't support it on your hardware, it won't work :P |
| 21:42 |
Jamesws |
Please don't kick/ban me but an iPhone |
| 21:42 |
Jamesws |
I've just bought a htc uni to play with though |
| 21:42 |
camro |
ida is also my friend ... but not sure, if that make sense .. |
| 21:42 |
MN |
lol really? |
| 21:42 |
Hunterkll |
if you have all android drivers and not standard linux drivers, it won't work outright |
| 21:42 |
Jamesws |
Hunterkll: I'm not using Android drivers |
| 21:42 |
Hunterkll |
the drivers need to be modified to be more standard linuxy |
| 21:42 |
Hunterkll |
Jamesws, the 'android' drivers i'm talking about are in the kernel as well |
| 21:42 |
Jamesws |
Hmm |
| 21:43 |
Hunterkll |
they just put out data/controls different then what regular linux expects |
| 21:43 |
Hunterkll |
android is a hackjob and a half |
| 21:43 |
Jamesws |
The kernel was originally created for Debian though |
| 21:43 |
Hunterkll |
oh |
| 21:43 |
Hunterkll |
well in that case |
| 21:43 |
Jamesws |
I have the device up and running with touch and stuff |
| 21:43 |
Hunterkll |
wtf |
| 21:44 |
Hunterkll |
How familiar are you with linux? |
| 21:44 |
Hunterkll |
Because you just jumped down the rabbit hole with no way back out.. ;) |
| 21:45 |
Jamesws |
I'm pretty good with everything except the kernel :p |
| 21:45 |
Hunterkll |
does regular debian work with kde/gnome etc? |
| 21:45 |
Jamesws |
Hunterkll: Yeh I've released an image |
| 21:45 |
Hunterkll |
.... everything else isn't linux |
| 21:45 |
Jamesws |
Erm |
| 21:45 |
Hunterkll |
linux is just the kernel :P |
| 21:45 |
Jamesws |
ixproject.org |
| 21:45 |
Jamesws |
I know I know xD |
| 21:45 |
Hunterkll |
everything else is gnu userland :D |
| 21:45 |
Hunterkll |
so in other words you suck at linux |
| 21:45 |
Hunterkll |
:P |
| 21:45 |
Hunterkll |
lol |
| 21:45 |
Jamesws |
xD |
| 21:46 |
Hunterkll |
so, there's a fully working debian that lasts more than an hour? |
| 21:46 |
Hunterkll |
it sounds to me like you have zero cpu scaling zero power management and everything is going beserk |
| 21:46 |
Jamesws |
No the debian port was never released |
| 21:46 |
Jamesws |
So I picked it up and did a debian and ubuntu port |
| 21:46 |
Hunterkll |
okay, so neither has working power management of any kind |
| 21:46 |
Hunterkll |
:P |
| 21:47 |
Jamesws |
Non at all :( |
| 21:47 |
Hunterkll |
are there any other 'vanillia' linux ports to iphone |
| 21:47 |
Hunterkll |
or is it just android |
| 21:47 |
Hunterkll |
if it's just android, down the kernel you go, because none of those drivers (for the most part) will work very well / at all with linux :P |
| 21:47 |
Jamesws |
Yeh the creator went straight ahead and did Android |
| 21:47 |
Hunterkll |
though cpu scaling should be done from cpufreq in the kernel |
| 21:47 |
Hunterkll |
perhaps you have the performance governor loaded as the default? |
| 21:47 |
Jamesws |
But he created the kernel, then ran Debian to port Android |
| 21:47 |
Hunterkll |
debian userland, perhaps |
| 21:48 |
Hunterkll |
but not with any decent pmu |
| 21:48 |
Hunterkll |
pm, as it were |
| 21:48 |
Hunterkll |
he ran debian as a base to get stuff start working then switched over to full and once he was in deep enough he could interact directly |
| 21:48 |
Jamesws |
Pretty much |
| 21:48 |
Jamesws |
He used Debian to compile Android afaik |
| 21:49 |
Hunterkll |
lol, not close to what i said, but easy enough |
| 21:49 |
Hunterkll |
okay |
| 21:49 |
Hunterkll |
so here's what i want you to investigate: GPE and Opie |
| 21:49 |
Hunterkll |
before you try and do this crazy XFCE and gnome and crap, look up two user interfaces designed for handheld devices :P |
| 21:49 |
Hunterkll |
GPE is X11 based, Opie is pure framebuffer based |
| 21:49 |
Jamesws |
Hunterkll: I'm using X11 with E17 and Illume at the moment |
| 21:50 |
Hunterkll |
You may want to find a kernel guru to help you with power management... but it sounds like your cpu is running full tilt the whole time, at the very least |
| 21:50 |
Jamesws |
Illume's a module for e for mobile devices, it's used with OpenMOko |
| 21:50 |
Hunterkll |
well, whatever works for you :P |
| 21:50 |
Hunterkll |
wifi works.. that's good |
| 21:51 |
Jamesws |
So I have the Interface, wifi and touch done, I guess I'll go hunting for a kernel guru :) |
| 21:51 |
Hunterkll |
but that is usually working from android -> regular linux |
| 21:51 |
Hunterkll |
touch is easy usually |
| 21:51 |
Jamesws |
It was |
| 21:51 |
Hunterkll |
it looks like who ever made the touch driver must have had tslib in mind when coding it |
| 21:51 |
Jamesws |
Only hard part was messing around xorg.conf for calibration |
| 21:51 |
Jamesws |
I used evdev |
| 21:51 |
Hunterkll |
calibration? we don't have no stinkin' calibration :P |
| 21:51 |
Hunterkll |
oi, still factoring in tslib :P |
| 21:52 |
Hunterkll |
http://tslib.berlios.de/ |
| 21:52 |
Jamesws |
I know, I've read all that |
| 21:52 |
Hunterkll |
is the best |
| 21:52 |
Jamesws |
I also used evtouch for a while |
| 21:52 |
Jamesws |
Had trouble with it in a landscape orientation though |
| 21:54 |
Jamesws |
I instead patched the evdev driver with a hacktastic patch I found on redhat mailing lists |
| 21:54 |
Jamesws |
For right click support |
| 21:56 |
Hunterkll |
heh |
| 21:58 |
Hunterkll |
first thing you should really investigate is backlight control (dimming/off etc) (same for the screen panel itself) and cpu scaling |
| 21:58 |
Jamesws |
I've managed to get it to dim |
| 21:58 |
Jamesws |
That works out of the box |
| 21:58 |
Jamesws |
It also blanks |
| 21:58 |
Jamesws |
But I can't figure out how to make it turn fully off |
| 21:59 |
Jamesws |
But It sounds like a kernel problem, so I'll search for a kernel guru with some spare time :p |
| 21:59 |
Hunterkll |
try setting zero to the ts value? |
| 21:59 |
Hunterkll |
er, bkl value |
| 22:00 |
Jamesws |
xD I've done a search on bkl value and I'm now read about the bigbang :p |
| 22:00 |
Jamesws |
I think I took a wrong turn at google |
| 22:04 |
Jamesws |
Hunterkll: I think I'm far out of my depth here :p |
| 22:04 |
Jamesws |
I'm happy with the port up to now, I'll just have to keep it on charge xD |
| 22:06 |
Jamesws |
I get a universal in the next couple of days, i think it'll be easier for me to play with than the iPhone |
| 22:07 |
camro |
jonpry: still here ? |
| 22:07 |
jonpry |
maybe |
| 22:07 |
camro |
lol |
| 22:08 |
camro |
the adresses - from where dave you found this ? |
| 22:08 |
camro |
smem |
| 22:08 |
jonpry |
wiki |
| 22:08 |
camro |
maybe i'm blind .. |
| 22:08 |
camro |
ah - not from kernel ..oki |
| 22:08 |
camro |
because it doesn'T change there anything .. |
| 22:09 |
camro |
HaRET(10)# pd 0xfc110 0x0080 |
| 22:09 |
camro |
000fc110 | 00000000 01080e06 0024003f 00500470 | ........?.$.p.P. |
| 22:09 |
camro |
000fc120 | 098a06b8 003c00a0 04ca0024 06c70050 | ......<.$...P... |
| 22:09 |
camro |
000fc130 | 00d01360 0024003c 005004ca 148c06c7 | `...<.$...P..... |
| 22:09 |
camro |
in c -> values_32 = (void *)(MSM_SHARED_RAM_BASE + htc_batt_info.resources->smem_offset); |
| 22:10 |
camro |
#define MSM_SHARED_RAM_BASE IOMEM(0xE0100000) |
| 22:10 |
jonpry |
pd 0x1ffc110 |
| 22:10 |
camro |
then i know what to do :) |
| 22:10 |
zagan13 |
hey |
| 22:11 |
jonpry |
if anything is true, that is not |
| 22:11 |
camro |
of course .. was thinking your values are the final addr :) |
| 22:11 |
camro |
lol |
| 22:11 |
zagan13 |
yea |
| 22:11 |
camro |
thanks jonpry ... |
| 22:11 |
zagan13 |
oihnio |
| 22:12 |
camro |
hi zagan |
| 22:12 |
zagan13 |
hey wat up |
| 22:12 |
camro |
nothing special .. |
| 22:12 |
zagan13 |
oh cool. |
| 22:29 |
WoZZeR |
What's the difference between INT_SDC2_0 and INT_SDC2_1? |
| 22:30 |
gauner1986 |
hey |
| 22:31 |
gauner1986 |
Markinus: g-sensor doesnt work |
| 22:34 |
Markinus |
gauner1986: hi, in the kernel yes, but not in userspace |
| 22:34 |
Markinus |
gauner1986: I'm looking to it |
| 22:34 |
gauner1986 |
hm |
| 22:35 |
Markinus |
gauner1986: I found already something . . |
| 22:35 |
gauner1986 |
markinus: and also leds dont work for sure |
| 22:35 |
gauner1986 |
i tested it |
| 22:35 |
gauner1986 |
on the linuxtogo kernel it works |
| 22:35 |
gauner1986 |
and on yours not |
| 22:35 |
gauner1986 |
i took darkstones liblights |
| 22:35 |
gauner1986 |
from v2 |
| 22:35 |
Markinus |
gauner1986: step by step :) |
| 22:35 |
gauner1986 |
okay |
| 22:35 |
gauner1986 |
just wanted to let you know |
| 22:35 |
gauner1986 |
;) |
| 22:36 |
S-Stream |
so, what are you guys doing anyway? :o |
| 22:36 |
S-Stream |
evo tree? |
| 22:36 |
Markinus |
gauner1986: if I control the led devices directly, it works perfect . . |
| 22:36 |
Markinus |
strange thing |
| 22:36 |
gauner1986 |
hm |
| 22:37 |
gauner1986 |
yeah |
| 22:37 |
gauner1986 |
the library doesnt do anything else |
| 22:37 |
S-Stream |
gauner1986: are you using a sense or non-sense build? |
| 22:37 |
gauner1986 |
sense |
| 22:37 |
S-Stream |
ah |
| 22:37 |
S-Stream |
sense builds have issue with liblights |
| 22:37 |
S-Stream |
shrugs |
| 22:37 |
S-Stream |
try AOSP build |
| 22:37 |
gauner1986 |
mine doesnt have issues with linuxtogo builds |
| 22:38 |
S-Stream |
:o ok |
| 22:42 |
gauner1986 |
markinus: btw.. did you notice the screen hangs when trying to switch it on? |
| 22:42 |
gauner1986 |
sometimes it lags for ages |
| 22:48 |
camro |
jonpry: how is going with audio fixes ? |
| 22:51 |
Markinus |
gauner1986: yes, but I didn'T look for it yet |
| 22:51 |
gauner1986 |
ok |
| 22:52 |
gauner1986 |
someone said that using the interactive governor fixes the problem |
| 22:52 |
gauner1986 |
which is of course not a solution |
| 22:52 |
gauner1986 |
i guess it uses fixed cpu speeds |
| 22:53 |
Markinus |
could you test as default the ondemand gov on linuxtogo? this is what I have here as active |
| 22:53 |
Markinus |
gauner1986: leds and screen |
| 22:54 |
gauner1986 |
on linuxtogo? |
| 22:54 |
gauner1986 |
OO |
| 22:54 |
gauner1986 |
Oo |
| 22:54 |
gauner1986 |
or on evo? |
| 22:54 |
gauner1986 |
i'm already on ondemand |
| 22:54 |
Markinus |
linuxtogo, there is userscpace gov active |
| 22:54 |
gauner1986 |
it is set by setcpu on boot |
| 22:56 |
Markinus |
gauner1986: ok. |
| 22:56 |
Markinus |
gauner1986: I have the gsensor |
| 22:56 |
Markinus |
but a bit strange, could you test something? |
| 22:57 |
gauner1986 |
ok |
| 22:57 |
Markinus |
it's a small fix |
| 22:57 |
gauner1986 |
did you pull? |
| 22:57 |
Markinus |
gauner1986:no, test it without. Edit include/linux/bma150.h |
| 22:57 |
gauner1986 |
ok |
| 22:57 |
Markinus |
change |
| 22:57 |
Markinus |
#define BMA150_G_SENSOR_NAME "bma150_uP_spi" |
| 22:57 |
Markinus |
to |
| 22:57 |
Markinus |
#define BMA150_G_SENSOR_NAME "bma150" |
| 22:57 |
Markinus |
and test it please |
| 22:58 |
gauner1986 |
k |
| 23:03 |
gauner1986 |
ahrg |
| 23:03 |
gauner1986 |
that kernel doesnt boot at all oO |
| 23:03 |
Markinus |
?! |
| 23:03 |
gauner1986 |
dunno.. |
| 23:03 |
Markinus |
blue? |
| 23:04 |
gauner1986 |
no |
| 23:04 |
gauner1986 |
haret init |
| 23:04 |
gauner1986 |
Oo |
| 23:04 |
gauner1986 |
could you upload your zImage? |
| 23:04 |
Markinus |
never had this problem |
| 23:04 |
gauner1986 |
i just have to transfer every file from my home pc |
| 23:04 |
gauner1986 |
which is a very slow connection |
| 23:04 |
Markinus |
moment |
| 23:04 |
gauner1986 |
dunno.. maybe it's garbled |
| 23:07 |
Markinus |
www.markinus.de/zImage |
| 23:09 |
gauner1986 |
thanks |
| 23:10 |
gauner1986 |
not passing haret init |
| 23:10 |
gauner1986 |
Oo |
| 23:10 |
gauner1986 |
what do you have in default.txt? |
| 23:10 |
gauner1986 |
i didnt change mine |
| 23:10 |
LordDeath |
what kind of problems does thi august 31st wifi driver solve? http://htc-linux.org/wiki/index.php?title=LeoWifi |
| 23:11 |
gauner1986 |
wifi tethering added |
| 23:11 |
gauner1986 |
nothing else |
| 23:11 |
LordDeath |
thx |
| 23:11 |
Markinus |
gauner1986: ehhh, the default vals |
| 23:11 |
Markinus |
gauner1986: maybe your sd card has a problem |
| 23:12 |
gauner1986 |
hm |
| 23:13 |
gauner1986 |
hm |
| 23:13 |
gauner1986 |
cant get it to boot |
| 23:14 |
gauner1986 |
ill retry linuxtogo kernel |
| 23:16 |
gauner1986 |
linuxtogo kernel works |
| 23:16 |
gauner1986 |
cant be sdcard then |
| 23:16 |
Markinus |
gauner1986: it's exactly same default.txt |
| 23:17 |
gauner1986 |
very strange |
| 23:17 |
gauner1986 |
i need to leave now |
| 23:17 |
Markinus |
yes |
| 23:17 |
gauner1986 |
girlfriend stresses me |
| 23:17 |
gauner1986 |
:P |
| 23:17 |
Markinus |
my wife is sleeping :) |
| 23:17 |
gauner1986 |
be back tomorrow morning |
| 23:17 |
Markinus |
bye |
| 23:17 |
gauner1986 |
lucky guy |
| 23:17 |
gauner1986 |
:P |
| 23:17 |
gauner1986 |
she calls me an asshole |
| 23:17 |
gauner1986 |
:D |
| 23:17 |
Markinus |
ouch |
| 23:18 |
Markinus |
buy roses :) |
| 23:18 |
gauner1986 |
next time |
| 23:18 |
gauner1986 |
:P |
| 23:18 |
gauner1986 |
goodnight |
| 23:18 |
epsas |
Is anyone using an Apache? |
| 23:18 |
fakker |
i hit mine, she no longer speaks |
| 23:18 |
Markinus |
night |
| 23:18 |
fakker |
nn |
| 23:20 |
dcordes |
[[Apache]] |
| 23:20 |
MN |
Hi |
| 23:21 |
MN |
Markinus: hows it going |
| 23:21 |
dcordes |
[Apache] |
| 23:21 |
dcordes |
what's up with the wiki bot |
| 23:21 |
Markinus |
MN: it's ok |
| 23:22 |
MN |
good evo tree going well |
| 23:25 |
Markinus |
I found the gsensor problem, it's only a a different device name in this kernel. I think this could be fixed in userspace not in kernel. The driver is evos one. |
| 23:26 |
Markinus |
dcordes: we could port the stuff now |
| 23:26 |
Markinus |
to linuxtogo |
| 23:28 |
dcordes |
Markinus: I don't remember the commands :( |
| 23:29 |
Markinus |
dcordes: ok, I try it, same name? Should be ok |
| 23:29 |
Markinus |
(like GT) |
| 23:30 |
dcordes |
GT ? |
| 23:30 |
dcordes |
same name as gitorious branch would reduce confusion |
| 23:30 |
MN |
dcordes what is needed for 3d acceleration in meego i read on xda that its only usable in android atm |
| 23:30 |
dcordes |
MN: that's right. it is a proprietary interface |
| 23:31 |
dcordes |
MN: what's needed... qualcomm must provide us with a driver |
| 23:31 |
MN |
ok i see now thanks for the info |
| 23:33 |
MN |
hows the new ubuntu build going |
| 23:34 |
dcordes |
MN: going good. I will prepare a new ready to go rootfilesystem once back from vacation |
| 23:35 |
MN |
okay enjoy your vacation |
| 23:35 |
dcordes |
thanks |
| 23:35 |
Markinus |
dcordes: pushing . . |
| 23:36 |
Markinus |
dcordes: I hope it does what I want :) |
| 23:36 |
dcordes |
^^ |
| 23:39 |
MN |
Markinus what are you currently working on |
| 23:40 |
Markinus |
copying repo from gitorious to linuxtogo |
| 23:44 |
MN |
okay going well? |
| 23:44 |
Markinus |
I know this in some minutes :) |
| 23:44 |
HardDisk_WP |
dcordes, wohin gehts? |
| 23:45 |
jonpry |
wozzer: there are multiple SD controllers |
| 23:47 |
Markinus |
dcordes: http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog;h=refs/heads/linux-on-wince-htc |
| 23:49 |
dcordes |
Markinus: did it say 'sending updated to cia' ? |
| 23:49 |
dcordes |
HardDisk_WP: south france |
| 23:49 |
HardDisk_WP |
have fun |
| 23:49 |
Markinus |
Sending updates to CIA.. fatal: Invalid revision range 0000000000000000000000000000000000000000..7677794c9be14576b03dca46ec710103fba466ca |
| 23:49 |
Markinus |
done. |
| 23:50 |
dcordes |
lol |
| 23:50 |
Markinus |
dcordes: seems he had a problem :) |
| 23:51 |
Markinus |
dcordes: I will check something new in soon, then we can see it |
| 23:52 |
flagada |
who is in south france because me to ? :) |
| 23:54 |
GNUtoo|laptop |
flagada, dcordes is |
| 23:54 |
flagada |
oh yeah :) |
| 23:55 |
flagada |
me toulouse and dcordes |
| 23:55 |
flagada |
? |
| 23:56 |
tmzt_ |
dcordes: qualcomm is trying, linux people don't want it |
| 23:57 |
dcordes |
flagada: ouf that's far. I start off in Nice :> |
| 23:58 |
flagada |
ok good |
| 23:58 |
flagada |
lol |
