I have submitted patches to Google which adds support of a
/system/etc/pointercal (aka. tslib) file for translating coordinates. I
didn't add ts_calibrate into the phone, however. That would be nice to
have so you can calibrate the screen on 1st install.
Anson wrote:
Hi all:
We use a
/automatically), it complained about SD card not
mounted.
Misa
2008/11/11 Sean McNeil [EMAIL PROTECTED]:
mvniekerk wrote:
Well, I can help you with this much - jffs2 + Android = No Go. It
will
Not true. Android runs perfectly
something else apart from creating/formatting the
partitions? I tried the
same but Music application never accepted the card (even it the fat
partition is
mounted manually/automatically), it complained about SD card not
mounted.
Misa
2008/11/11 Sean McNeil [EMAIL PROTECTED
. That is exactly the patch that I submitted and linked to, and is
what I meant by Removing lines 257-259. If this is a well known problem,
btw, why wasn't it already fixed? ;P -J
--
From: Sean McNeil [EMAIL PROTECTED]
Sent: Wednesday, November 19
It is most likely that you do not have the Android device drivers
enabled in your kernel. You really need them to be present. Another
possibility is you are using a newer kernel which places devices in a
different location in the /sys tree. In that case, you'll need to add an
option to your
You've already figured most of it out. You'll need to check pv.com and
how Packet Video does things. I think they have an SDK to develop
hardware codecs. Then you'll want to modify external/opencore
accordingly to use your hardware codecs.
Pivotian wrote:
I suppose that Android uses the
To get a different compiler, you'll need to set in some place like your
build/target/product the location of your compiler:
override TARGET_TOOLS_PREFIX := $(TOPDIR)../cross/bin/arm-eabi-
Then you'll need to use different compiler settings in
build/core/combo/linux-arm.mk. I don't know if my
that needs to be open sourced by the owners (which they fully
intend to do).
And then there are non-Android commitments I have...
Sean
Davide wrote:
If you like, I can help you with code
or some other stuff if you wish ...
2008/11/26 Sean McNeil [EMAIL PROTECTED]:
Davide,
There are a few
Maxime Petazzoni wrote:
Hi Sean,
On Nov 27, 11:01 am, Sean McNeil [EMAIL PROTECTED] wrote:
There are a few things you can check:
1) Make sure you have an ALSA audio device enabled in your kernel.
2) Check from a shell if the directory /dev/snd is created and if there
is anything
What version of the kernel are you using? Newer versions of the linux
kernel will pass VM_EXEC on mmap which will cause the binder to fail.
You might need to fix the binder so that it is no longer a
FORBIDDEN_MMAP_FLAGS.
lqoyzy wrote:
Hi Ben,
I'm trying to port android to a s3c2410 develop
I don't think it is going to work that way. You have a different
linker/loader and all that which has to be in specific directories from
root. I think you'll have to do a chroot. That is what I do:
chroot /root /sbin/init
(I've moved init to the more appropriate /sbin directory).
Either
compilation).
Francois Minaud
On Dec 11, 5:58 pm, Sean McNeil [EMAIL PROTECTED] wrote:
Android uses a completely different linker/loader. You cannot run
Android directly from your davinci filesystem. Very important shared
libraries will not be in the correct place on the filesystem. Just
. I really appreciate
Francois Minaud
On Dec 11, 6:45 pm, Sean McNeil [EMAIL PROTECTED] wrote:
You could try to kick android off in the background by appending an
to the chroot command. Then you could check for messages with dmesg.
Sounds like you may be having linker
for your fast replies and advices. I really appreciate
Francois Minaud
On Dec 11, 6:45 pm, Sean McNeil [EMAIL PROTECTED] wrote:
You could try to kick android off in the background by appending an
to the chroot command. Then you could check for messages with dmesg.
Sounds like you may
Are you even certain you have the wpa_supplicant binary? Can you run the
command from the shell?
Richard Zhao wrote:
It still don't work.
2008/12/10 nagamatu [EMAIL PROTECTED]:
Can you start wpa_supplicant by start command like:
# start wpa_supplicant
This start command just do
Hi,
Android uses the standard power interface, but it expects that uevents
are generated when the values change. It does not poll the values, but
listens for the uevents instead.
zhqzh1982 wrote:
hi all:
is anyone know the android charging detect and battery level
detect interface
the POWER_SUPPLY_PROP_STATUS ,but i do not know the
android's battery log changed ,always display ?,now where is the
problems?
I am linux driver and kernel engineer,thanks
On Mon, Jan 5, 2009 at 12:15 PM, Sean McNeil seanmcne...@gmail.com
mailto:seanmcne...@gmail.com wrote:
Hi
? When
was this made?
Please let me know
On Jan 8, 10:18 am, Sean McNeil seanmcne...@gmail.com wrote:
Google has made some minor changes to the audio interface and Wind River
is aware of them. There is a patch pending and it will hopefully be
committed soon.
Sorry for the inconvenience
Gujare rupeshguj...@gmail.com
mailto:rupeshguj...@gmail.com wrote:
Thanks Sean,
We tested it and found that, it is working nicely. Now we are
able to compile build and play sound.
Regards,
Rupesh Gujare
On Thu, Jan 8, 2009 at 11:03 AM, Sean McNeil
seanmcne
What you are trying to do cannot be done. Accelerometers give you a
vector to GC whereas the G1 has a compass which gives you yaw/pitch/roll
to north.
伊泽 wrote:
hi pramod gurav,
I think we have the same problem,the lis302/lis331 sensor cannot provide
yaw/pitch/roll data while we feel hard to
Xiaoyang Yu (Max) wrote:
Current Android code for backlight adjustment do not work properly for
systems other than G1. This is mainly because:
1) The corresponding sysfs file changed. In mydroid/hardware/
libhardware/power/power.c, LCD_BACKLIGHT is hard coded to /sys/class/
Yes, I noticed this as well. Seems like the patch I sent you was a
little messed up. Just copy over the files from
./frameworks/base/libs/audioflinger/AudioHardwareStub.*
and you should be OK.
Sean
Marco Trevisan (Treviño) wrote:
On 8 Gen, 06:33, Sean McNeil seanmcne...@gmail.com wrote
eth1 isn't a valid device for tracking. Take a look at
frameworks/base/core/java/android/net/MobileDataStateTracker.java.
You'll see that it really only looks at rmnet0, eth0, and gprs. For DNS
to work, the net.dns# properties need to get set. You'll either have to
add eth1 into the
Huh? I do this all the time. I setup ethernet over USB and then connect
to the phone with:
ADBHOST=192.168.0.202 adb kill-server
ADBHOST=192.168.0.202 adb command
I don't know what you mean that it isn't possible. I guess I'm doing the
impossible, then. ;)
David Turner wrote:
This is not
The soft keypad in cupcake is a great step in supporting keypad-less
phones like the Openmoko. I've found the following minor issues that I
don't know if Google engineers have seen:
1) At times, the cursor will remain in front of a character that is
typed. It happens to me mostly when typing
It is very simple:
First, make sure you have connectivity to the zoom. Ping the ip address.
Next, type the command as follows (with appropriate changes):
ADBHOST=ip_address adb kill-server
ADBHOST=ip_address adb shell
If using ethernet over usb, the zoom/OMAP USB gadget driver has an odd
It looks like I'm missing a change even after repo sync. I get:
development/apps/Development/AndroidManifest.xml:74: ERROR Error: String
types not allowed (at 'configChanges' with value
'keyboard|keyboardHidden|navigation|orientation').
when building. Searching for configChanges shows several
I've been looking at iproute2 in conjunction with iptables. I got
iproute2 to compile, but need to study it more to see exactly what
should be done.
Hugo Darwin wrote:
I am thinking this problem also. Now I am porting Android in our
company x86 netbook with HSDPA.
1. ethernet
2. wifi
3.
A solution for using a calibration file /system/etc/pointercal was provided
ages ago when Android was first open sourced. It was part of the OpenMoko
support I released. You can look at that for reference.
On Fri, Feb 13, 2009 at 2:46 AM, Dianne Hackborn hack...@android.comwrote:
You'll need to
I've found that if your code is written in C++ - like the overlay
example - you need to add 'extern C' to the beginning of the const
definition of the module header. Otherwise the compiler/binutils will
end up stripping the symbols. Try using -objdump -T on the .so to make
sure you have the
I want to provide a difference of my repo and apply it to a clean copy.
How is this done? The repo command doesn't appear to support an apply
command. Is this a planned feature, is there another means to do this
easily?
--~--~-~--~~~---~--~~
unsubscribe:
Maybe try a \n at the end of the string?
forest wrote:
for example:
void fun()
{
static int c = 0;
LOGI(call func() %d, ++c);
}
the log sometimes display
call func() 1
call func() 5
call func() 8
..
when debug PVMFStatus AndroidAudioInput::DoRead() int the file
:
add \n has no effect
On 2月17日, 上午11时36分, Sean McNeil seanmcne...@gmail.com wrote:
Maybe try a \n at the end of the string?
forest wrote:
for example:
void fun()
{
static int c = 0;
LOGI(call func() %d, ++c);
}
the log sometimes display
call func() 1
call func
Hi,
Did you look in both /dev and /dev/block?
ebmajor wrote:
Hi all,
Here is my situation.
Kernel command line: root=/dev/mtdblock2 rw rootfstype=jffs2 init=/
init console=ttyS0,38400 mem=64M
/ # cat /proc/
partitions
major minor #blocks
name
31 0256
mtdblock0
31
/mtdblock2.
On Feb 18, 5:49 pm, Sean McNeil seanmcne...@gmail.com wrote:
Hi,
Did you look in both /dev and /dev/block?
ebmajor wrote:
Hi all,
Here is my situation.
Kernel command line: root=/dev/mtdblock2 rw rootfstype=jffs2 init=/
init console=ttyS0,38400 mem
Android must have a writable area that it calls /data. It places all
kinds of information in there including user preferences, dalvik cache,
etc. If you are mounting everything as / over nfs, then you probably
don't have a separate /data partition somewhere that is mounted rw.
You'll have to
We have updated the repository now. Just perform a repo sync and you
should be fine.
Jack stone wrote:
you should search in this group. There is a patch :
http://groups.google.com/group/android-porting/browse_thread/thread/b66db688d1630382/80fdb626e3194c09?#80fdb626e3194c09
On Mar 6, 8:34
ext3 ?
On Tue, Mar 3, 2009 at 7:51 AM, Sean McNeil seanmcne...@gmail.com wrote:
Yes. I use ext3 all the time, though, for the journaling. Nothing
prevents you from using ext2.
gulshan karmani wrote:
Hello All,
We have eMMC as a flash on our board and we use EXT2
filesystem
/product/eee_701/obj/
STATIC_LIBRARIES/libasound_intermediates/libasound.a', needed by `out/
target/product/eee_701/obj/SHARED_LIBRARIES/libaudio_intermediates/
LINKED/libaudio.so'. Stop.
Any clue?
On Mar 9, 2:03 am, Sean McNeil seanmcne...@gmail.com wrote:
We have updated the repository now
Those are different modes within the system. You are usually in
MODE_NORMAL. When someone calls you, it enters MODE_RINGTONE. When you
answer a call or you initiated a call then you enter MODE_IN_CALL. I
should think it rather obvious from the names as to what they do.
Typically in
in the Android.mk, but it doesn't seem to
actually build it.
On Mar 10, 3:00 pm, Androidphan niels.kee...@tass.nl wrote:
I got both alsa-lib as alsa_sound.
On Mar 10, 2:14 pm, Sean McNeil seanmcne...@gmail.com wrote:
You need both platform/hardware/alsa_audio and
platform/external/alsa-lib
I'm not sure, but could this also be related to short enums? I seem to
remember seeing something like this when I had a mixture of old code
compiled with and new code compiled without short enums.
On Thu, Mar 12, 2009 at 11:53 PM, Dan Bornstein danf...@android.com wrote:
On Thu, Mar 12, 2009
Actually, you don't know if the chip is powered and the uart is working just
because bttest passed. It just means that he started the hcid and hciattach
services. Most likely his bluetooth device is not powered up and hciattach
is simply failing.
On Fri, Mar 13, 2009 at 12:39 AM, Nick Pelly
There are major updates being merged into cupcake. You'll have to wait
until they are completed before ALSA support is updated to match.
virstud wrote:
mkdir cupcake
cd cupcake
git clone git://android.git.kernel.org/platform/external/alsa-lib.git
git clone
Hi Pavan,
While testing, try turning off
CONFIG_ANDROID_PARANOID_NETWORK
in the kernel. You are probably trying to access the socket in a process
that doesn't have network permission.
Sean
pavan savoy wrote:
Nope, Still get the error, tried out lot of things actually, fcntl,
umask, still
Did you try adding group inet to hciattach? Is that the one that kicks
off texas_post?
pavan savoy wrote:
yeah, Sean, actually I have the hciattach working when run as root,
I just want possibly the proper way that's all.
regards,
Pavan
On Fri, Mar 20, 2009 at 6:08 PM, Sean McNeil
hciattach_legacy to
more groups, I think we should figure out a way so that hci0 is opened
up for user bluetooth, isn't it ?
I suppose Nick would agree with that ?
What you say Nick ?
regards,
Pavan
On Fri, Mar 20, 2009 at 6:16 PM, Sean McNeil seanmcne...@gmail.com
mailto:seanmcne...@gmail.com
In fact all of the routines are gone such as shmat(). How come?
--~--~-~--~~~---~--~~
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
-~--~~~~--~~--~--~---
You need to create an appropriate asound.conf. It goes into /system/etc/.
liaojunsh...@gmail.com wrote:
Hi, all,
in AudioHardwareALSA.cpp, look at the function ALSAStreamOps::open(int
mode, int device) , when i need to do routing to earpiece, headset, or
other output device.
function
Currently you cannot rely on repo or git to build Android. Period.
susanner wrote:
So, currently, I can not rely on repo or git to build alsa , right? if
it's ok, would you please let me know ? Actually I am porting
mediaplayer but there is no sound .
在2009-03-23，Sean McNeil seanmcne
There is no libasound.so. It is built as libasound.a and included whole
into libaudio.so.
susanner wrote:
how can libasound be .so not .a? after make , I can't find libasound
in system/lib
在2009-03-23，Sean McNeil seanmcne...@gmail.com 写道：
Currently you cannot rely on repo or git
Hi,
I am 95% sure that the answer to all your questions below is no.
adumitru007 wrote:
I have a lot of questions about alsa, here i go..
I am trying to use alsa api in a native C application to play some
sound on the HTC Dream. If i cross-compile the application for ARM
with the tools
archerypp wrote:
I have updated my Android to the latest one using 'repo sync'.
That was a mistake. The main Android repo is currently broken. ALSA will
not be updated until main can be compiled without ALSA.
But I met this error while compiling it.
The log is :
target arm C++:
The code below assumes that the change in the display base address
happens immediately. If you have an architecture that provides shadow
registers, then you want to set them in pan_display and not in the isr
and just wait for the VSYNC.
Manish wrote:
Hi Jay,
I missed one thing in the
I submitted a patch that needed a change.
I make the change and commit.
I do a repo upload --replace, but the replaced patch is against the
previous one I just replaced.
How do I get the change to be based on the original?
I'm going to attempt an upload where I put the old number in both the
will be submitted.
Sean
Jean-Baptiste Queru wrote:
I typically do a cherry-pick for those:
repo start pick .
git cherry-pick sha of the commit you want
repo upload --replace .
repo abandon pick .
JBQ
On Thu, Apr 2, 2009 at 7:45 PM, Sean McNeil seanmcne...@gmail.com wrote:
I have a follow-up
You do realize that the phone would have to answer the call before it
could play the music, right? This would mean you will start incurring
minutes usage.
I don't think that service plans that offer music or other personal ring
play count that time towards your minute allocation.
dingerjun
Hi Fredrik,
IMO, external/* should not depend on bionic.
Fredrik Markström wrote:
What is the general opinion about the code in external (for example
skia, libjpg), can it depend on bionic-specifics like cpu-features.h
or should we try to keep the external stuff independent ?
The specific
-specific APIs (look
at web kit for example).
On Mon, Apr 6, 2009 at 6:26 PM, Sean McNeil seanmcne...@gmail.com
mailto:seanmcne...@gmail.com wrote:
Hi Fredrik,
IMO, external/* should not depend on bionic.
Fredrik Markström wrote:
What is the general opinion about
I think it is also a good idea not to eliminate an architecture, but to
include the ones that are supported. For instance, it would probably be
better to say
#if defined(__arm__) !defined(__thumb__) (__ARM_ARCH__ = 5)
if this code is supported on architectures 5 and above.
David Turner
I used to run Android from chroot under a minimal filesystem all the
time. It helps to debug the init process.
David Turner wrote:
in short, no, Android requires its own init system, specific kernel
drivers that are probably not in the moblin/ubuntu one, etc...
there are probably some
The problem is that the secondary buffer is being written to before it
has been completely removed from the screen. The video driver should be
waiting for the blanking interval before returning as that OMAP video
interface has shadow registers that only take effect during the blanking
period.
yaffs2 is only for NAND. Do not use with a MicroSD card. They have their
own bad-block managers. Use ext2/3 or vfat.
Elvis Dowson wrote:
Hi Edward,
How can I specify that the user data partition
must be yaffs2?
Let's take an example.
NAND Memory:
/dev/mtdblock3 is
You need to read frameworks/base/data/sounds/AudioPackage2.mk.
Basically, there are 2 sets of audio files now. One for limited flash
devices, one for those that have more room. So you have to include the
.mk for the set that fits your environment. This would be done in your
Hi Michael,
I'm not certain I follow your issue. What is your device subtype? I've
taken out the -fno-short-enums as that is default on the newer baseline.
Perhaps you could check make hide= and make sure that option is passed
to the compiler.
Michael Trimarchi wrote:
Hi,
Sean McNeil
IMHO this would be a really bad feature and a severe security hole.
Android should never allow this. This is one reason I've always opposed
a separate multiplexer daemon as it would allow more than just the RIL
to talk directly to the modem.
fgilani wrote:
Is it possible to send AT commands
I have noticed that there is a lot of discussion on the android-porting
group that really has nothing to do with porting. More like how do I
use this port? questions. Particularly the x86 versions like EeePC, but
I've noticed others as well.
Should some new groups be created to help reduce
This is NOT a bug in the alsa code. This is a bug in your alsa kernel
driver. We could, however, make this a non-fatal condition.
xro wrote:
better with all the method... sorry
// Stop the transfer when the buffer is full.
err = snd_pcm_sw_params_set_stop_threshold(mHandle,
there are 2 of these structures defined in
hardware/alsa_sound/AudioHardwareALSA.cpp. Look for
static StreamDefaults _defaults
one is for playback, and one for record. You are interested in the one
for playback. There are 2 things that are meaningful for this issue in
this structure: latency
FPGA stands for Field Programmable Gate Array. It is essentially
programmable hardware. The answer to such a query is no, Android cannot
be ported to an FPGA. You would basically need a 32-bit cpu core placed
on the FPGA along with supporting hardware. Then port Linux onto it,
then Android.
Which OMAP are you using? Are you aware of TIs omapzoom.org? The
OpenCore codecs are already ported to the OMAP3 3430 DSP.
Elvis Dowson wrote:
Hi Vince,
On Mon, Aug 3, 2009 at 11:04 PM, Vince claesvinc...@gmail.com
mailto:claesvinc...@gmail.com wrote:
Please keep me informed on
The ALSA interface conforms to the master branch. You appear to be
building against cupcake. There isn't a cupcake branch of the ALSA
interface at this time.
Demetris wrote:
Hi Guys, couldn’t not notice that you are trying to make Alsa work.
I have done:
1.
git clone
On 01/18/2010 05:27 PM, ani wrote:
The pcm_dmix.c is not compiled in the Android alsa lib.
I have compiled this c file and DMIX is working for me.
My questions are still the same:
a.Can we use the ALSA mixer(DMIX-i enabled in my system) for mixing
streams instead of android mixer?If yes
It would have been a lot easier to just run
stop media
before alsa_aplay.
Cheers,
Sean
On 01/29/2010 01:56 PM, Srikant wrote:
Hi,
I modified the permissions of the /system folder in the init.rc, then
I could able to kill the mediaserver process and able to run the
alsa_aplay.
thanks,
/init.rc is not a bash script. It is an init script and it is
interpreted by the init process.
On 06/28/2010 05:01 PM, fpirou wrote:
Hi,
I am newbies at android porting.
I compiled 2.6.31 linux kernel and Android 2.1 (eclair) for an iMX25
(ARM926EJ-S) target.
But, when I mount Android file
Are there any plans to support these apps anymore? Try adding
BUILD_PV_TEST_APPS := 1
to your BuildConfig.mk and you'll see that they don't compile.
Sean
--
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
My guess would be something regarding nfs options that make the file not
interpreted as a unix file. Either that, or the init.rc is improperly
formed.
On 06/29/2010 10:32 PM, fpirou wrote:
Do you have any idea why the init process does not interpret the
init.rc command file (invalid option)?
Take a look at the system property ro.sf.lcd_density. For a QVGA
screen, you want to set it as
ro.sf.lcd_density = 120
to get the proper scaling. IMHO, the density should not be necessary and
scaling could be done via the dpi. But for some reason it isn't done.
Sean
On 9/14/2010 4:54 PM,
This is all handled by the init program. Check it out. It even loads
binary blobs.
On 10/4/2010 11:02 PM, Great White Pin Geek wrote:
I'm wondering what the proper way is to dynamically load/integrate
device drivers for hardware that becomes available when a device is
docked.
For example,
Yes, I didn't include any intrusive changes. Since the code base
already had checks for version, I provided only the changes in those
areas. I am aware of the problem with video. That is why we moved to
v2.50 where it works. Here is a simple howto for resolving compilation
issues:
Create
Most likely, you are attempting to do them in the init.rc before the
/system volume is mounted.
On 3/4/2011 3:23 PM, Amit Pundir wrote:
On Fri, Mar 4, 2011 at 11:48 AM, Laylightai...@gmail.com wrote:
Hi Android Porters,
I would like to insmod 2 driver modules during boot-up. I tried adding
On 3/9/2011 8:57 AM, kismeter wrote:
On Mar 8, 10:44 pm, Joerie de Gramj.de.g...@gmail.com wrote:
You can use c++filt to demangle C++ symbol names. The method you're
looking for is android::createOMXPlugin().
Joerie
thanks
but why add _ZN7* or when add _ZN
You should never be declaring
I use eclipse to compile Android. It works pretty well, but I've never
found a means to get the parser to work with both C/C++ and Java. I use
a C/C++ project.
On 3/9/2011 11:25 AM, RD wrote:
Thanks for your reply. I have already built android for ARM and for
X86 using the command line. I
It looks like the new JB branch of external/guava does not compile for
me with the sun java on Ubuntu 12.04. I get the following error:
external/guava/guava/src/com/google/common/collect/Maps.java:542:
incompatible types; no instance(s) of type variable(s) K,V exist so that
84 matches
Mail list logo
