Re: [android-porting] Build Android Source Code On Windows Platform
On Thu, Oct 11, 2012 at 9:18 AM, Shivanagouda Biradar shivanagouda.bira...@asmaitha.com wrote: Hi All, I would like to build Android source code , having linux drivers, on windows platform. Please do suggest . Simply speaking, building directly from Windows is not supported (it's far too problematic). The alternative is to build on a virtual Linux machine instead. I.e. install VirtualBox / VMWare, create a virtual machine, install Ubuntu on it, then follow the typical Linux build instructions. Be aware that you'll need a _lot_ of disk space and probably memory though. Regards. Shivanagouda -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: Why packagemanager drops the vendor apps (Nativity Activity)'s so file.
+cc android-porting, +bcc android-ndk That's not an NDK question, I'm redirecting you to android-porting. On Sat, Apr 21, 2012 at 5:33 PM, chinlin@tw-roc chinlin0...@gmail.comwrote: If you put your Native Activity apk in /vendor, the .so file will not be extracted and dropped by package manager due to security reason? NativityActivity should be treated as same as JNI apps. There is also some problem, if you put your Native code in /system/lib, the NativeActivity can not load it. It looks like Native Activity forgot to handle this cases, it should work the same way that JNI does (it loads .so file from apps's libs folder, or /system/lib folder) Can google team put this into consideration? There maybe more native activity developers want to deploy their apps into /vendor on device ROM. Regards, Chinlin -- You received this message because you are subscribed to the Google Groups android-ndk group. To view this discussion on the web visit https://groups.google.com/d/msg/android-ndk/-/0hfrSWuUrQIJ. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: permissions failing from init on first boot after flashing device
I'm redirecting you to android-porting@, this is not related to the NDK. On Wed, Apr 18, 2012 at 8:30 PM, girlCoder best.car...@gmail.com wrote: I am working on a project that creates custom images for an Android device. We edit the kernel and other os files. Our system init calls init.herring.rc which is where we have our 'module' starting from. service ourapp /data/local/ourapp --daemon class core user root group root oneshot On the first boot after flashing our images onto the device 'ourapp' always fails due to permissions. However, after we reboot once the init.herring.rc starts up 'ourapp' fine. What could be causing this? Why would a reboot be necessary for the init.rc file to start a service... Thanks!! -- You received this message because you are subscribed to the Google Groups android-ndk group. To view this discussion on the web visit https://groups.google.com/d/msg/android-ndk/-/1mWjaSQZb_sJ. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: Static lib creation issue
Hello, First of all, this is not a question related to the NDK, so I'm redirecting this to android-porting instead. On Mon, Dec 19, 2011 at 8:39 AM, Meraj Ahmad Ansari meraj.ahmad.ans...@hotmail.com wrote: Hi List, I have 2 questions - Q1- - I am trying to build static lib with android source code by using mmm command and it is creating static lib in - target StaticLib: nativestaticlibdemo (out/target/product/generic/obj/ STATIC_LIBRARIES/nativestaticlibdemo_intermediates/ nativestaticlibdemo.a) It seams that there is no nativestaticlibdemo.a created in - out/ target/product/generic/system/lib folder as like shared lib. Is it valid behavior or i am missing something? No, this is perfectly valid behaviour. Only stuff that gets installed to the final partition image go into $ANDROID_PRODUCT_OUT/system A static library is neither an executable or loadable file, so has no point in being there. I am trying to compile only one .C file and my Android.mk is - LOCAL_PATH := $(call my-dir) include $(CLEAR_VARS) LOCAL_MODULE:= nativestaticlibdemo LOCAL_SRC_FILES := nativeTest.c include $(BUILD_STATIC_LIBRARY) Q-2 - When i use ndk-build command of Android NDK then it is not creating static lib(creation of shared lib is fine) List your library in your APP_MODULES definition. By default, the NDK only builds the executable and shared libraries in your project, as well as all their dependencies (which can be static libraries too). So if you have a shared library that depends on a static library, the latter will be automatically built. If you only have a static library, you need to list it in your APP_MODULES to force its compilation. Following is the log- rm -f /home/administrator/Development/android-ndk-r6b/samples/ nativeStaticLibDemo/libs/armeabi/lib*.so /home/administrator/ Development/android-ndk-r6b/samples/nativeStaticLibDemo/libs/armeabi- v7a/lib*.so /home/administrator/Development/android-ndk-r6b/samples/ nativeStaticLibDemo/libs/x86/lib*.so rm -f /home/administrator/Development/android-ndk-r6b/samples/ nativeStaticLibDemo/libs/armeabi/gdbserver /home/administrator/ Development/android-ndk-r6b/samples/nativeStaticLibDemo/libs/armeabi- v7a/gdbserver /home/administrator/Development/android-ndk-r6b/samples/ nativeStaticLibDemo/libs/x86/gdbserver rm -f /home/administrator/Development/android-ndk-r6b/samples/ nativeStaticLibDemo/libs/armeabi/gdb.setup /home/administrator/ Development/android-ndk-r6b/samples/nativeStaticLibDemo/libs/armeabi- v7a/gdb.setup /home/administrator/Development/android-ndk-r6b/samples/ nativeStaticLibDemo/libs/x86/gdb.setup Any pointer is really helpful. Regards, Meraj -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: What permissions does AndroidManifest.xml need to access a device driver?
Hello, I think this is not related to the NDK, try android-porting@ instead, where people with similar problems will likely be able to help you Regards On Fri, Oct 14, 2011 at 7:08 PM, Anil anil.phi...@motorolasolutions.comwrote: What permissions does AndroidManifest.xml need to access a device driver? The device driver file has 777 permissions. I am using Froyo 2.2 on the OMAP4 Blaze board. When I try to call the device driver through JNI, I get a write error and errno is Bad File Number. I used permissions uses-permission android:name=android.permission.FACTORY_TEST / uses-permission android:name=android.permission.HARDWARE_TEST / However when I look in Settings, I do not see these permissions. Anil -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] VMGL integration with android emulator
On Tue, Oct 11, 2011 at 6:54 AM, Inder inderjeet.tha...@gmail.com wrote: Hi All, I have build VMGL for host (ubuntu 10.04). Can anybody let me know how I can integrate VMGL in android emulator? or How to build VMGL for Android emulator? I'm afraid this is not possible, nor desirable: - VMGL only works for X11-based Linux systems, so it can't easily be ported to Android (which uses a completely different graphics system) - VMGL only works when emulating x86, not ARM, as far as I know - VMGL requires a X11 host, and thus is useless for developers using Windows or OS X. - VMGL doesn't handle OpenGLES, only desktop GL commands. We have a much better GLES emulation solution in the works. More news about it when it'll be ready. Regards Please repley asap. Thanks in advance.. -Inder -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: How to debug Android boot on emulator?
This is not related to the NDK, try android-porting instead. On Wed, Oct 5, 2011 at 2:51 AM, Dmitry divis1...@gmail.com wrote: Hi, I'm trying to run custom Android build on emulator. This build is created for Cortex-a9, so I had selected the -armv7 kernel image (from prebuilt/android-arm/kernel) and use -qemu -cpu cortex-a9 option. (I had cloned the latest external/qemu) Unfortunately (well, as expected), this configuration does not work. Kernel is starting, init process is started, init launches some services, but they are terminated after some time, and finally system halts. One more issue is that I cannot connect to the emulator with adb. (What is interesting, adbd is started by init but is not terminated. I can see emulator-5554, but it is offline). I had turned off most of the services (including critical, like servicemanager) in a hope that I just too slow in connecting to emulator with adb. No luck. I'm suspecting that cortex-a9 emulation is a problem and would like to debug it, but do not know how to do this. I can debug the kernel itself (using qemu -s -S debugging), but this seems to me not optimal (mostly because I do not know the kernel so good). I know that logcat can produce the crash stack of user-space application. Since I do not have adb, I do not have this stack. It seems that bionic intercepts aborting signals and directs the output to debuggerd (which is also crashing in my emulator setup). Any guess are welcome. P.S. Not sure this is correct group for this discussion... Dmitry -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: CAN interface missing in NDK platforms
I'm redirecting this thread to android-porting because it no longer has anything to do with the NDK. Maybe some folks here will better be able to help you. On Thu, Sep 1, 2011 at 5:20 PM, Palsson palsson...@gmail.com wrote: Hello David, Thanks man. Its my bad.. android-3 and android-8 got mixed up. Now it compiles. Would you know, how to include the ip tool for net interface config. It seems like the netcfg tool is only available, and the problem is that the bitrate for the can0 interface can't be configured using netcfg. When you enter netcfg can0 up, it complains that the bitrate is not set. This can be done very smoothly using the ip tool, ... # ip link set can0 type can bitrate 125000 # ip link set can0 up It comes with the linux distribution for my AM3517 nut kind of frustrating that it is not in Android. Regards Palsson On Aug 31, 3:10 pm, David Turner di...@android.com wrote: On Wed, Aug 31, 2011 at 3:07 PM, Palsson palsson...@gmail.com wrote: Hello again, Well not working. If I copy my missing header files to what I think is the proper location.. $NDK/platforms/android-8/arch-arm/usr/include/linux the build system is still not finding them. I might be missing something, but do I have to make some sort of ndk- build clean, for the build-system to search again for header files from that location? Sorry if I'm lacking some basics here. No, are you targetting API level 8, i.e. do you use APP_PLATFORM := android-8 in your Application.mk, or does your manifest states a minimum SDK version of 8? If not, you're targetting android-3 by default (which is 1.5 a.k.a. Cupcake). /Palsson On Aug 31, 12:07 pm, Palsson palsson...@gmail.com wrote: Thanks David, Is it that simple :-) I'll try that.. Regards Palsson On Aug 31, 11:04 am, David Turner di...@android.com wrote: Hello, On Wed, Aug 31, 2011 at 10:01 AM, Palsson palsson...@gmail.com wrote: Hello, Hardware: TI's AM3517 evaluation kit. OS Release: Android Froyo The socketCAN (berlios) interface is included in the Android Linux kernel (Froyo release). But when browsing the kernel header files for the NDK platforms it is missing and following includes in my native code creates errors: #include linux/can.h #include linux/can/raw.h The error is simply, file not found. As stated in the NDK documentation (see docs/STABLE-APIS.html), the kernel interface headers are not considered part of the stable APIs provided by the NDK. This also mean there is no guarantee at all that the type of socket interface you're interested in will be supported by each Android device's kernel, or even that you can access it from an application (due to lack of adequate permissions). I need to be able to access the canbus interface on my platform. It would be nice to be able to use some of the can-utils within the socketcan project. Therefore I would like to test it, by compiling the utilities with NDK. Is it possible to add the socketCAN interface to a custom NDK platform? If you're building your own platform, and still want to use the NDK to build your utilities, simply copy the relevant kernel headers files to the appropriate location (e.g. $NDK/platforms/android-8/arch-arm/usr/include if you're targetting Froyo). Regards Palsson -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] muti-threaded native code using JNI
Why are you calling pthread_exit(0) on the main thread, this will terminate
your application immediately.
You can't block on the main UI thread (which calls your JNI callback).
You'll need to find a different synchonization primitive (see AsyncHandlers
in the SDK doc)
On Wed, Aug 24, 2011 at 5:41 PM, arif arifo...@gmail.com wrote:
Hi All,
I ported one muti-threaded native code (binary) to android platform,
and is working fine,
Now i am trying to run this code by creating apk, created my native
code as shared library using NDK and
i am calling this shared library by using JNI, below is the sample
code,
But it is not working, i am new to java and jni, i did some web
search, and i found that i need to attache/detach threads in jni,
It will be very helpful if someone help me to handle this native
threads in jni.
Thanks
/** apk code
***/
public class HelloJni extends Activity
{
/** Called when the activity is first created. */
@Override
public void onCreate(Bundle savedInstanceState)
{
super.onCreate(savedInstanceState);
TextView tv = new TextView(this);
tv.setText( stringFromJNI() );
setContentView(tv);
}
public native String stringFromJNI();
static {
System.loadLibrary(my_lib);
}
}
/** JNI function
/
jstring
Java_com_example_hellojni_HelloJni_stringFromJNI( JNIEnv* env, jobject
thiz)
{
int rc = native_main();
if (rc == 0)
return (*env)-NewStringUTF(env,
Successful!);
else
return (*env)-NewStringUTF(env, Failure!);
}
/* Native code
***/
int
native_main()
{
int cnt;
pthread_t thread;
for (cnt =0; cnt 5; cnt++)
{
if ( 0 != pthread_create (thread1, NULL, (void *)
my_function,
(void *) cnt))
{
return -1;
}
}
pthread_exit(0); /* exit */
return 0;
}
void my_function(void* cnt)
{
int c = (int) cnt;
switch (c)
{
case 0:
/* calling native funtion1 */
break;
case 1:
/* calling native funtion2 */
break;
case 2:
/* calling native funtion3 */
break;
case 3:
/* calling native funtion4 */
break;
case 4:
/* calling native funtion5 */
break;
}
}
/**/
--
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
--
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
Re: [android-porting] Gingerbread linker crash on Linux kernel 3.0
It's the function that calls a shared-library or executable list of constructor functions. It is likely a constructor crashing. Unfortunately, the shared library's mapping is not available to gdb until after the constructors have run It's a know bug, you could try changing the lines in linker.c that read like: call_constructors(si); notify_gdb_of_load(si); and change them to: notify_gdb_of_load(si); call_constructors(si); However, I don't guarantee that this will work. Let me know if this helps though. On Wed, Aug 24, 2011 at 3:00 PM, Yan Weichuan yan.weich...@gmail.comwrote: I am working on Gingerbread 2.3.5 + Linux kernel 3.0, but I found all dynamic execute application will crash. And with the gdb, it show the application crash at call_arry function at linker.c. My ARM cpu is arm- v7a. Does anyone have some idea about this? (gdb) bt #0 0xafd14e44 in ?? () #1 0xb0003820 in call_array (ctor=0xafd40004, count=value optimized out, reverse=0) at bionic/linker/linker.c:1615 #2 0xb000a2f4 in ?? () from /system/bin/linker #3 0xb000a2f4 in ?? () from /system/bin/linker Backtrace stopped: previous frame identical to this frame (corrupt stack?) -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] why __func__ variable contains the type signature of the function except its bare name?
Use __FUNCTION__ instead of __func__, see http://gcc.gnu.org/onlinedocs/gcc-4.1.2/gcc/Function-Names.html On Thu, Aug 4, 2011 at 8:04 AM, Simon Chan simonchan1...@gmail.com wrote: Hi all, Now I add debug info for my overlay hal module. #define LOG_FUNC_NAME LOGD(%s, __func__) the output is as seen below, D/Overlay ( 966): int overlay_setParameter(overlay_control_device_t*, overlay_t*, int, int) D/Overlay ( 966): int overlay_commit(overlay_control_device_t*, overlay_t*) It's certainly not the thing I expect. I don't know whether it's the intended behavior or I have made something wrong? Thanks! Regards, Simon -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Ginger Bread build errors on ubuntu 10.10 64 bit.
This should be fixed now (but you may need to make clean build). Sorry, you were probably caught in the middle of various subtle C library changes, one of them introduced the issue, but another one fixed it. On Thu, May 26, 2011 at 5:27 AM, maliha malihamari...@gmail.com wrote: PLATFORM_VERSION_CODENAME=AOSP PLATFORM_VERSION=3.1.4.1.5.9.2.6.5 TARGET_PRODUCT=full TARGET_BUILD_VARIANT=eng TARGET_SIMULATOR=false TARGET_BUILD_TYPE=release TARGET_BUILD_APPS= TARGET_ARCH=arm TARGET_ARCH_VARIANT=armv5te HOST_ARCH=x86 HOST_OS=linux HOST_BUILD_TYPE=release BUILD_ID=OPENMASTER Errors: target thumb C: libstlport = external/stlport/src/cxa.c target SharedLib: libstlport (out/target/product/generic/obj/SHARED_LIBRARIES/ libstlport_intermediates/LINKED/libstlport.so) prebuilt/linux-x86/toolchain/arm-eabi-4.4.3/bin/../lib/gcc/arm-eabi/ 4.4.3/../../../../arm-eabi/bin/ld: out/target/product/generic/obj/SHARED_LIBRARIES/ libstlport_intermediates/src/monetary.o: in function _Z41__static_initialization_and_destruction_0ii.clone.0:external/ stlport/src/monetary.cpp:148: error: undefined reference to '__dso_handle' prebuilt/linux-x86/toolchain/arm-eabi-4.4.3/bin/../lib/gcc/arm-eabi/ 4.4.3/../../../../arm-eabi/bin/ld: out/target/product/generic/obj/SHARED_LIBRARIES/ libstlport_intermediates/src/locale.o: in function _Z41__static_initialization_and_destruction_0ii.clone.0:external/ stlport/src/locale.cpp:444: error: undefined reference to '__dso_handle' prebuilt/linux-x86/toolchain/arm-eabi-4.4.3/bin/../lib/gcc/arm-eabi/ 4.4.3/../../../../arm-eabi/bin/ld: out/target/product/generic/obj/SHARED_LIBRARIES/ libstlport_intermediates/src/locale_impl.o: in function std::_Locale_impl::Init::_M_count() const:external/stlport/src/locale_impl.cpp:72: error: undefined reference to '__dso_handle' prebuilt/linux-x86/toolchain/arm-eabi-4.4.3/bin/../lib/gcc/arm-eabi/ 4.4.3/../../../../arm-eabi/bin/ld: out/target/product/generic/obj/SHARED_LIBRARIES/ libstlport_intermediates/src/locale_impl.o: in function std::_Locale_impl::make_classic_locale():external/stlport/src/ locale_impl.cpp:672: error: undefined reference to '__dso_handle' collect2: ld returned 1 exit status make: *** [out/target/product/generic/obj/SHARED_LIBRARIES/ libstlport_intermediates/LINKED/libstlport.so] Error 1 Has anyone encountered the same, any thoughts on a fix? Cheerio! - Maliha. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Where to send kernel patches?
Hello, On Wed, Jun 22, 2011 at 1:30 AM, Jason killert...@gmail.com wrote: I'm wondering if anyone can point me in the right direction to submitting a kernel patches for Android. I'm working on improving support for the w8001 (see my original mail at http://groups.google.com/group/android-discuss/browse_thread/thread/18a3fc968ac72397 ) and can't seem to figure out who I hand my code off to. If my patches weren't Android-specific, I'd just submit them upstream. The Submitting Patches page (http://source.android.com/source/submit- patches.html) seems to have the instructions I need, but the repo command doesn't bring in the kernel sources. I see that there are a half-dozen Android kernel trees at http://android.git.kernel.org/ but (assuming I can figure out hot to get repo to pull down the repository) which one would I submit my patches to? If the patch isn't specific to a given chipset, you should send to it android.git.kernel.org/kernel/common.git The Android kernel team uses gerrit, but not repo, so to send a patch, you need to do something like: 1/ Select a remote branch to work on, e.g. android-2.6.39 2/ Make your patch a commit on your local git repository on top of this branch 3/ Upload the patch to gerrit with: git push ssh://r.android.com:29418/kernel.common.git+HEAD:refs/for/android-2.6.39 (this assumes you have a registered account on r.android.com and loaded your SSH keys to your ssh-agent). Try looking at @android.com or @google.com addresses in the git log that correspond to people that have submitted stuff in parts of the kernel that relate to your patch, and add them as reviewers for your patch on gerrit. Hope this helps - David Jason -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Is there OSAL layer
On Fri, Jun 3, 2011 at 11:01 AM, Vikas KM vikki...@gmail.com wrote: Hi, Is there any OSAL layer in Android Build System..? What do you mean by Android Build System? The OSAL within the platform is called the kernel interface :-) -- Regards Vikas KM -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: User input, getevent and keydown events
This is not related to the NDK, you should try android-porting instead. Briefly speaking, getevent() returns the events sent from the kernel to the framework. Since everything seems good here, you can forget about a driver/hardware failure. It looks like the framework input handling code is repeating the events for some reason. Which exact version are you using? (input handling has evolved a lot between Android releases). The no keyboard for id 0 message probably means that your event kernel driver is reporting a name that doesn't have a corresponding .kcm / .kl file combination, i.e. you're likely missing some configuration files, or your kernel driver is not complete / configured properly. On Wed, May 25, 2011 at 4:47 PM, Samuh samuh.va...@gmail.com wrote: I am working on an Android application that runs on custom hardware. This hardware allows text entry/navigation using two input devices - 1. keyboard and 2. remote. When the user enters the text/performs navigation using the keyboard every thing behaves normal. However, when the user uses the remote sometimes, a single key hit is getting repeated multiple times. For e.g: below logs capture output of getevent utility and onKeyDown() callback on an Activity. User Input: letter P and Up /Down navigation keys // Logs from getevent - please note scan codes are replaced by their mappings /dev/input/event3: 0001 P /dev/input/event3: /dev/input/event3: 0001 DPAD_DOWN 0001 /dev/input/event3: /dev/input/event3: 0001 DPAD_DOWN /dev/input/event3: /dev/input/event3: 0001 DPAD_UP 0001 /dev/input/event3: /dev/input/event3: 0001 DPAD_UP /dev/input/event3: /dev/input/event3: 0001 DPAD_DOWN 0001 /dev/input/event3: /dev/input/event3: 0001 DPAD_DOWN /dev/input/event3: /dev/input/event3: 0001 DPAD_UP 0001 /dev/input/event3: /dev/input/event3: 0001 DPAD_UP /dev/input/event3: /dev/input/event3: 0001 DPAD_DOWN 0001 /dev/input/event3: /dev/input/event3: 0001 DPAD_DOWN /dev/input/event3: /dev/input/event3: 0001 ENTER 0001 /dev/input/event3: /dev/input/event3: 0001 ENTER /dev/input/event3: /dev/input/event3: 0001 ENTER 0001 /dev/input/event3: /dev/input/event3: 0001 ENTER /dev/input/event3: /dev/input/event3: 0001 ENTER 0001 /dev/input/event3: /dev/input/event3: 0001 ENTER /dev/input/event3: /dev/input/event3: 0001 P 0001 /dev/input/event3: Logs from onKeyDown(): KeyDown( 1495): Keycode: 19 .. repetitions of 19 KeyDown( 1495): Keycode: 19 KeyDown( 1495): Keycode: 20 .. repetitions of 20 KeyDown( 1495): Keycode: 20 KeyDown( 1495): Keycode: 66 .. repetitions of 66 KeyDown( 1495): Keycode: 66 Until eventually it spits out: I/InputQueue-JNI( 1495): Sending finished signal for input channel '408b8088 XXX (client)' since it is being unregistered while an input message is still in progress. I/InputQueue-JNI( 1495): Ignoring finish signal on channel that is no longer registered. W/KeyCharacterMap( 1495): No keyboard for id 0 W/KeyCharacterMap( 1495): Using default keymap: /system/usr/keychars/ qwerty.kcm.bin Can someone please guide me as to what could be going wrong here? Thanks. -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] GL acceleration in Emulator - demo - at Google IO
On Thu, May 19, 2011 at 7:45 PM, Suman Saraf sumansa...@gmail.com wrote: Thanks David. I checked out the latest code and have started to fiddle with it. I am trying to run this on x86 inside VirtualBox. I have a few questions and would be glad if you could help: 1. I built everything inside shared and system and tests/gles_android_wrapper. That produced the EGL, GLES_v1 and GLES_v2 libraries. I copied these to my device along with the egl.cfg and gles_emul.cfg sounds about right. 2. do I need to build gralloc also? I did build it and planted it in system/lib/gralloc.default.so; overwriting the gralloc.default.so for the software only renderer. nope, this doesn't work yet. 3. now I need to build a renderer application which can render it remotely. I tried building host/renderer on Linux. Ideally I want to build this on Windows? Is it possible? the resulting emulator_renderer binary (on Linux) failed to initialize frame-buffer because of unsupported EGL_RENDERABLE_TYPE in eglChooseConfig. the only one that works is ut_renderer for now. And usage is a bit specific. forget about emulator_renderer for now. To reiterate, I ideally want to build the emulator_renderer on Windows. Is it possible? Is there an alternate way to render this remotely on Windows? How? Am I completely off track or is this really the way to get this thing to run? This work is in progress and *very very* experimental. I would like to ask you to be patient before trying to use it. usage instructions are going to change a lot in the coming weeks so I prefer not to document these officially. Thanks for your comprehension - David Thanks Suman On 16-May-2011, at 9:09 PM, David Turner wrote: On Fri, May 13, 2011 at 3:58 PM, Suman Saraf sumansa...@gmail.com wrote: Hello, I just saw this demo on youtube. Looks very cool. I have been working independently on the same problem trying to remote GL calls to the host. This basically means mapping EGL and GL calls to GLX/GL on Linux (which is my host right now). I am using Chromium to pack/remote/unpack and manage state. Does the google implementation also use Chromium? Not that I know of. It's a custom wire protocol. Are there some code bits which have been checked into the AOSP tree which I could look at? Yes, look at development/tools/emulator/opengl/ (warning: work in progress) Thanks Suman -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Whether NeofreeRunner is compatible with android 2.3
On Tue, May 17, 2011 at 2:38 PM, vishnu vardhanraj4...@gmail.com wrote: this is Ravulapati.Iam new to android porting.we r intersted to port android 2.3 version (ginger bread) on Neofree Runner whether it supports or not.If it supports can any body send links to get source code and all stuff and also steps to follow. The FreeRunner has an ARMv4T CPU, while Android requires ARMv5TE or higher by default. So Gingerbread will not run on it without specific modifications, some of them might not be trivial (e.g. the Dalvik JIT or V8 one). -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] GL acceleration in Emulator - demo - at Google IO
On Fri, May 13, 2011 at 3:58 PM, Suman Saraf sumansa...@gmail.com wrote: Hello, I just saw this demo on youtube. Looks very cool. I have been working independently on the same problem trying to remote GL calls to the host. This basically means mapping EGL and GL calls to GLX/GL on Linux (which is my host right now). I am using Chromium to pack/remote/unpack and manage state. Does the google implementation also use Chromium? Not that I know of. It's a custom wire protocol. Are there some code bits which have been checked into the AOSP tree which I could look at? Yes, look at development/tools/emulator/opengl/ (warning: work in progress) Thanks Suman -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Info required on Goldfish Kernel
On Fri, May 13, 2011 at 7:56 AM, Srinidhi K V srinidh...@yahoo.co.inwrote: Hi All, I was wondering why there is no Goldfish kernel after 2.6.29 . Is it google stopped forward porting it to the subsequent versions or is there something new in subsequent kernels (2.6.30 onwards) that wont support goldfish? A rebase to 2.6.36 is planned for the future, though is of low priority. There is no other reason than inertia (and the fact that we didn't find an issue that forced us to upgrade). I had an experimental rebase to 2.6.35 working perfectly fine a few months ago, but the final version will be on 2.6.36. Thanks in Advance Srinidhi -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Upgrading Goldfish Kernel
On Thu, May 12, 2011 at 5:00 PM, Srinidhi K V srinidh...@yahoo.co.inwrote: Hi, I have set of device drivers which works on 2.6.35 kernel. I am developing a module which will integrated with Android. During the development stage I want to test the module on SDK before doing it on real hardware. But my drivers doesn't work on 2.6.29 Goldfish Kernel. So I wanted to upgrade Goldfish Kernel to 2.6.35. To my knowledge the latest kernel for qemu is 2.6.29. So, I would like to know what would be the best way to upgrade Kernel. Any pointers will be of great help. For now, you will have to rebase the goldfish kernel sources manually. This requires adding a few missing includes here and there though, so won't be fully automatic. We may have a 2.6.36 system in the future, but no promises here. Thanks in Advance Srinidhi KV -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Running my own libgps.so on emulator
IIRC, the hardware abstraction library will try to load a library named /system/lib/hw/libgps.hardware.so first before any system wide libgps.so Where hardware corresponds to the hardware name, and is goldfish when running under emulation. See development/tools/emulator/system/gps/Android.mk for the details on how to force your module to go to /system/lib/hw/, instead of the default /system/lib. You probably want to replace the emulator version with your own sources during testing. On Mon, May 9, 2011 at 10:07 AM, Itai steiner.i...@gmail.com wrote: Hello, I wrote and built my own libgps.so in my directory on hardware/ vendor/gps Afterwards, I re-built the entire android project. The make completed successfully and libgps was placed in system/lib However, when I load the emulator, with : out/host/linux-x86/bin/emulator -sysdir out/target/product/generic/ - kernel prebuilt/android-arm/kernel/kernel-qemu -avd And2.2 and run my test application (via Eclipse), I see that it still uses the emulated library and not mine. How can I direct the emulator to use my implementation? -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Emulator:how are add-on platform images merged ?
They are not merge, they are simply searched in numerical order (so sysdir.1 first, then sysdir.2, etc...) A platform add-on typically provides a full system.img if it provides new system libraries / frameworks. There is no way to provide partial system images that get merged into the main one. On Wed, May 4, 2011 at 9:53 AM, freakingtux kees.jongenbur...@gmail.comwrote: Hello I am trying to understand what can be put in a platform add-ons and therefore need to understand how some parts of the emulator work. When a new avd is created for an add-on a config.ini file is put in the avd's directory. This ini files contains multiple entries to a sysdir image.sysdir.2=platforms/android-8/images/ image.sysdir.1=add-ons/example_addon-ver670881-user-linux-x86/images/ When the emulator is started it looks like the different sysdir's are merges somehow. How does this work? Can I override anything? Greetings -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: android-ndk,handler usbfs read write
On Fri, Mar 18, 2011 at 3:26 AM, WuZhenda wuzhe...@gmail.com wrote: hi,all I have ported the libusb hidapi to the android, but when I use ndk-build to operate the usb device on java platform,I always return false. I doubt you're using ndk-build to operate anything on the device :-) so I do chmod 666 /proc/bus/usb/002/001 cmd manually to make the device can be read write , then it return OK. Not NDK related. Normal applications can't modify the system. Moving this forum to bcc. but I think when I mount the usbfs,I have made the usb can be read write already by the cmd mount usbfs none /proc/bus/usb -o devmode=0666. why? -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Android boot process deadlock
Not NDK-related. Don't cross-post! On Fri, Mar 11, 2011 at 1:40 PM, Amit Pundir pundira...@gmail.com wrote: Hi, I'm trying to create a surface while handling the InputReader. I followed this test given here frameworks/base/services/surfaceflinger/tests/surface/surface.cpp to create a surface in libui. After doing that my Android is not booting up. From logcat I found that Android run-time is not starting up. # logcat - beginning of /dev/log/main I/DEBUG ( 870): debuggerd: Jan 25 2011 16:09:01 I/Netd( 869): Netd 1.0 starting - beginning of /dev/log/system I/Vold( 868): Vold 2.1 (the revenge) firing up D/Vold( 868): USB mass storage support is not enabled in the kernel D/Vold( 868): usb_configuration switch is not enabled in the kernel D/Vold( 868): Volume sdcard state changing -1 (Initializing) - 0 (No-Media) D/Vold( 868): Volume sdcard state changing 0 (No-Media) - 2 (Pending) D/Vold( 868): Volume sdcard state changing 2 (Pending) - 1 (Idle-Unmounted) request_suspend_state: wakeup (0-0) at 9747131339 (2000-01-01 00:00:07.839813222 UTC) request_suspend_state: wakeup (0-0) at 14498565684 (2000-01-01 00:00:12.591217049 UTC) request_suspend_state: wakeup (0-0) at 20147918738 (2000-01-01 00:00:18.240600621 UTC) request_suspend_state: wakeup (0-0) at 24564239174 (2000-01-01 00:00:22.656921057 UTC) .. Seems to me that boot process is stuck in some kind of a deadlock. Is there anything fundamentally wrong in this approach? Any special lock(s) I need to take care of? Regards, Amit Pundir PS. Excuse me for cross-posting. I did that to reach a wider audience. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: init.architecture.rc
On Fri, Feb 11, 2011 at 10:19 AM, Stefan sgo...@dh-electronics.de wrote: It's not architecture, but board specific. So, I have to delete the goldfish-init and write my board specific file. You don't have to delete it technically, it will just be ignored. What do I have to write into e.g. init.myOwnBoard.rc if I only want to start Android as I can see it running on the emulator? That's quite an open question. The point of this file is to perform any board-specific initialization. It could defining new services, mounting partitions, defining system properties, whatever. Apart from that, consider it an extension of init.rc It really depends on the customization you have done. Do I have to edit the normal init.rc too for this? The idea is that init.rc is platform-specific, while init.board.rc is board-specific. I.e. if you modify the platform in significant ways, you would modify init.rc to reflect that. You could then use the same platform to build various product configurations (each corresponding to a different board), and put their own customizations in init.board1.rc, init.board2.rc, etc... Of course, if all you do is developing a single product/board, you can just modifiy init.rc instead. Would you be so kind and give me an example for a running init.XXX.rc- File? -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Android NDK: x86-4.2.1 libsupc++ support
On Wed, Feb 2, 2011 at 3:08 PM, Subramani Venkatesh subbu31m...@gmail.comwrote: David, Thanks for the info.. I tried for quite sometime but not succesful in building libstc++, here is the log of failure.. Configure: x86-4.2.1 toolchain build Building : x86-4.2.1 toolchain [this can take a long time]. Install : x86-4.2.1 toolchain binaries. Copying libstdc++ prebuild binaries. ERROR: ERROR: Missing toolchain-specific include directory: /home/svenkat0/Projects/NDK/android-ndk-r5b/toolchains/x86-4.2.1/prebuilt/linux-x86/i686-android-linux-gnu/include/c++/4.2.1 May I understand what is the real issue in here. The real issue is that the libstdc++ sources were probably not built by the GCC build script. I don't really know why though at this point. Try using the --verbose option (once or twice) to see what's happening. Sorry if I can't help. -Subbu On Tue, Feb 1, 2011 at 3:46 PM, David Turner di...@android.com wrote: On Tue, Feb 1, 2011 at 4:13 PM, Subramani Venkatesh subbu31m...@gmail.com wrote: Hi, I am porting an application to Android x86 platform There is no such thing for now, though we're working on it :-) , while porting I understood that my application depends on Android libsupc++ support for exception handling, I verified for this library in NDK prebuilt directory, I did not find it, may I know the procedure to build it using Android NDK? See docs/DEVELOPMENT.html, otherwise try the following (not tested in a long time though) build/tools/download-toolchain-sources.sh /tmp/ndk/src build/tools/build-gcc.sh --copy-libstdxx /tmp/ndk/src `pwd` x86-4.2.1 Regards, Subbu -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Android NDK: x86-4.2.1 libsupc++ support
On Tue, Feb 1, 2011 at 4:13 PM, Subramani Venkatesh subbu31m...@gmail.comwrote: Hi, I am porting an application to Android x86 platform There is no such thing for now, though we're working on it :-) , while porting I understood that my application depends on Android libsupc++ support for exception handling, I verified for this library in NDK prebuilt directory, I did not find it, may I know the procedure to build it using Android NDK? See docs/DEVELOPMENT.html, otherwise try the following (not tested in a long time though) build/tools/download-toolchain-sources.sh /tmp/ndk/src build/tools/build-gcc.sh --copy-libstdxx /tmp/ndk/src `pwd` x86-4.2.1 Regards, Subbu -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: using the camera service from android native code
Redirecting to android-porting, this is not NDK-related. On Fri, Jan 21, 2011 at 8:45 PM, ajay mna...@gmail.com wrote: hi, we are trying to modify the android filesystem.we are trying to use the camera service from a c++ code.we are using the cameralistner class present in /root/Desktop/froyo/frameworks/base/include/camera/ Camera.h . the cameralistner class consists of only 3 pure virtual functions when i try compiling the the .so i get the following error : defined reference to `typeinfo for android::RefBase' when investigated we got know that cameralistner inherits from refbase (cameralistner:virtual public refbase). we also got to that the refbase is part libutils.so please help! thanks and regards, Ajay -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.comandroid-ndk%2bunsubscr...@googlegroups.com . For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] SDK v.s. SDK-addon v.s. hardware build.
On Wed, Jan 19, 2011 at 7:06 AM, freakingtux kees.jongenbur...@gmail.comwrote: Hello, I am trying to understand how to deploy and use the different Android platform make targets. I am working with the following cases: Case-1) I will need to make small modifications to the standard Android APIs or code Case-2) I will want to add new APIs next to the existing Android APIs Case-3) I will want third parties to be able to develop using those APIs under Windows and Linux Case-4) I will want to be able to build targeting real hardware. Android globally supports the following targets Target-1) Creating a SDK. In This contains the basic tools like the emulator binary and the SDK updater Target-2) Creating a platform-x delivery (android.jar + emulator image and such) (in 2.2. this is part of the make sdk target) Target-3) Creating a SDK-addon (vendor.jar + and emulator image) Target-4) Creating a build targeting real hardware. The questions I really have is about the relationship between the SDK-addon and a platform-x delivery. When I create an sdk-addon it gets delivered with full emulator images but no android.jar. How does that work? add-ons are for stuff that add to the platform (hence the name). They correspond to case-2 and case-3 in your list. They don't come with android.jar, which is specific to a given platform API. Should I run CTS on the sdk-addon? The CTS is supposed to perform platform-specific tests, and ignore additionnal vendor-specific APIs. However, it is run against a working system image (which may include add-ons, but that's not the point). Note: If you provide a custom android.jar, chances are that you will not pass the CTS (depending on the modifications you made to the platform). What is the real difference between this sdk-addon and a platform? The platform is what _every_ Android application can expect to use. If you break the API, some applications will not be able to run properly. Is it correct I can not create a AVD with multiple sdk-addons combined? An AVD provides a system image. Said system image could provide anything, including the APIs of multiple add-ons. However, on the development side, I don't think you can build against multiple-addons at once. If like expected in Case-1 and Case-2 I make changes to the Android platform do I need to delivery a new SDK so people can develop it or can I somehow decouple the delivery of a platform and the sdk? I think it is possible to make the SDK tools point to a custom repository of platform images, but I don't remember the details. How can I ensure the different target builds match? I figured out how to do most things here but I am still struggling on how to use them. Greetings -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: Enable and disable ADB Driver initialization dynamically
nto android-ndk related, try android-porting On Thu, Jan 6, 2011 at 3:42 PM, b_hirawat b.hirawa...@gmail.com wrote: HI, I am trying to enable and disable ADB driver initialization dynamically in my android device. I have tried to block the ADB initialization in adb_function_init(void) function in '/drivers/usb/ gadget/f_adb ' file but after flashing, the device is not booting up. Can anybody tell me whether my approach is correct or I need to take any other approach for this implementation. Any Inputs is appreciable. Thanks in advanced. B. Hirawat -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.comandroid-ndk%2bunsubscr...@googlegroups.com . For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] How to permanently mount a SD card in android-x86 OS?
Please use the android-x86 forums instead, not the official Android ones, for such questions. On Sat, Dec 18, 2010 at 5:22 PM, 捷超 王 ee07b...@gmail.com wrote: The SDCARD variable methods provided on http://www.android-x86.org/documents/sdcardhowto seems only suitable for temporarily mount a a SD card, i.e., SD card will be lost due to the OS reboot. Anyone knows way to permanently mount a SD card in android-x86 OS? Thank you again! -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: Native code segfault on Tegra2
On Wed, Dec 15, 2010 at 5:42 PM, Howard M. Harte hhar...@gmail.com wrote: I don't know where it's crashing, tried compiling for armv5 with - msoft-float, and still see the segfault, but no core is dumped, so no stack trace. The device is not rooted so I can't change the ulimit to get a core dump. Any ideas on how to get a backtrace, or even the faulting IP? At least if I had the IP, I could look at the disassembly. There should be a stacktrace in your logcat, at a minimum, which will give you the IP. If not, this either isn't a stackfault, or there is something fishy in your system image. Howe do you run the code exactly? Is it through JNI or calling a standalone executable from the adb shell? If all else fails, I will root it (Viewsonic gTablet) and see if changing the ulimit will produce a core dump. Thanks, Howard On Dec 14, 2:14 am, David Turner di...@android.com wrote: First, you should try android-porting, not android-ndk Do you know what the code that crashes does exactly? You should have a stack trace that tells you where the problem is located, which might give you some ideas. What compiler flags are you using when building with the gcc-4.4.0 toolchain. Not using the correct one is going to produce invalid machine code (that's why there is an NDK in the first place). Also, keep in mind that Tegra2 doesn't support NEON or VFPv3-D32 (but you would get a SIGILL fault instead, so it's probably unrelated). On Tue, Dec 14, 2010 at 3:46 AM, Howard M. Harte hhar...@gmail.com wrote: I have some native Linux code that I built with gcc-4.4.0 from the Android source repository. I build it for ARMv6, so it can run on all ARM-based Android platforms, ie: ARMv6 (MSM7K, etc), ARMv7 (OMAP, MSM8K, Hummingbird.) When I try to run this code on a Tegra2 CPU, I get a segmentation fault. If I compile these binaries along with Android for the Tegra2 (I'm using the harmony build in CM6.1) then it runs fine. Unfortunately, these same binaries will segfault on non- Tegra2 ARMv7 platforms such as OMAP. Both Tegra2 and OMAP are reported as ARMv7. I don't mind having separate binaries, but I'd like to understand what is going on in case there is some way to have unified binaries. In any case, if Tegra2 is special, then I need to find a way to differentiate it from the other ARMv7 CPUs. I know this is not strictly related to the Android NDK, but figured this would be the best place to find people who might know what the issue is here. Also, I'm wondering if I might face a similar issue with shared libraries that I've built using the NDK for my application? Thanks, Howard -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.comandroid-ndk%2bunsubscr...@googlegroups.com android-ndk%2bunsubscr...@googlegr oups.com . For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] libgps.so on emulator?
The SDK system images contain /system/lib/hw/libgps.goldfish.so which implements the GPS implementation for the emulated session. Source code is available under sdk/emulator/gps/ On Fri, Nov 26, 2010 at 2:59 PM, Dudero sinfanh...@googlemail.com wrote: Hello, I do not understand why I can not find the libgps.so at /system/ lib in the emulator (Eclair 2.1 with GPS support enabled). The gps test app GPSStatus works well and receives the NMEA-Messages via Telnet... Does the emulator not require this libary? Best Regards, Andreas -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Android build issue
That's right, the emulator is built as a 32-bit binary, so you need 32-bit development libraries for it to work. On Tue, Oct 26, 2010 at 9:38 AM, Przemyslaw Wegrzyn pwegr...@codepainters.com wrote: On 10/26/10 4:37 PM, Butchi Peddi wrote: I am trying to build android source on Ubuntu(9.10) and getting compilation errors as below. I uninstalled jdk6 jre but no luck. The error is related to stdc++ lib. I checked in synaptic package manager and it shows libstdc++6-4.4-dev got installed. I guess that you have 64-bit Ubuntu installed? If so, try installing lib32stdc++6 http://packages.ubuntu.com/jaunty/lib32stdc++6 BR, Przemek -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] CANNOT LINK EXECUTABLE
Probably a problem in the dynamic linker. You should try to enable debug traces to get more information. See bionic/linker/linker_debug.h and the end of bionic/linker/linker.c for more details. On Tue, Oct 26, 2010 at 12:38 PM, jakiee janghwa...@gmail.com wrote: I ran the Eclair code on superH (sh4) architecture based machine and am trying to run Froyo code on same platform. But I have the issue to run core components such as app-process, mediaserver from init process. If I manually run those binaries in /system/bin folder, I can see this message, ./mediaserver link_image[1995]: failed to link ./mediaserver CANNOT LINK EXECUTABLE I can see same kind of message for app-process, mediaserver, system_server, dalvikvm and run-core-tests script while it is ok to run other executables such as logcat, ping, etc. ./run-core-tests link_image[1995]: failed to link dalvikvm CANNOT LINK EXECUTABLE Could anybody to let me know what the issue might be and what/how to check? Thanks in advance, -Jake -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Is it or will it be possible to install a driver through an app?
Installing a kernel driver requires root access (for good reasons), applications are not allowed to do that. On Thu, Oct 7, 2010 at 7:29 PM, Andres Buritica and...@thelinuxkid.comwrote: Hi, I saw the release of the Netflix video out adapter for the iPhone and it made me wonder if something similar to this is possible for Android devices. Assuming that the driver is installed when the Netflix application is updated and has not been baked into iOS, does the App API allow the installation of drivers (kernel modules)? Here's the news link for reference: http://news.cnet.com/8301-13506_3-20018573-17.html Thanks, -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: [Error ] annot find symbol symbol : method log(java.lang.String)
Please avoid cross-posting !! this should go into android-developers only. On Wed, Sep 22, 2010 at 6:57 PM, Atul Raut atul.dr...@gmail.com wrote: Hi I have imported import android.os.Message; still giving error. On Wed, Sep 22, 2010 at 10:24 PM, Atul Raut atul.dr...@gmail.com wrote: Hi am getting this error : cannot find symbol symbol : method log(java.lang.String) when using method : log(some test, +test); I have imported import android.text.TextUtils; import android.util.Log; Still its throwing same error why it is so ? -Atul -- You received this message because you are subscribed to the Google Groups android-ndk group. To post to this group, send email to android-...@googlegroups.com. To unsubscribe from this group, send email to android-ndk+unsubscr...@googlegroups.comandroid-ndk%2bunsubscr...@googlegroups.com . For more options, visit this group at http://groups.google.com/group/android-ndk?hl=en. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Error when build Froyo for Emulator
You should select 1 (Device); the simulator is something different than the emulator, and I'm not sure it's supported anymore anyway. hope this helps On Sat, Aug 28, 2010 at 4:13 PM, Tung Dang tung...@gmail.com wrote: Hi all , I build the Android system image (Froyo) for emulator mode and get error, the configuration is as follow : t...@ubuntu:/media/ANDROID/Froyo$ choosecombo Build for the simulator or the device? 1. Device 2. Simulator Which would you like? [1] 2 Build type choices are: 1. release 2. debug Which would you like? [2] Which product would you like? [sim] Variant choices are: 1. user 2. userdebug 3. eng Which would you like? [eng] PLATFORM_VERSION_CODENAME=REL PLATFORM_VERSION=2.2 TARGET_PRODUCT=sim TARGET_BUILD_VARIANT=eng TARGET_SIMULATOR=true TARGET_BUILD_TYPE=debug TARGET_BUILD_APPS= TARGET_ARCH=x86 HOST_ARCH=x86 HOST_OS=linux HOST_BUILD_TYPE=release BUILD_ID=MASTER The errors is this : target thumb C++: libandroid_runtime = frameworks/base/core/jni/android_server_BluetoothA2dpService.cpp target thumb C++: libandroid_runtime = frameworks/base/core/jni/android_server_Watchdog.cpp frameworks/base/core/jni/android_server_Watchdog.cpp: In function ‘void dumpOneStack(int, int)’: frameworks/base/core/jni/android_server_Watchdog.cpp:38: warning: ignoring return value of ‘ssize_t write(int, const void*, size_t)’, declared with attribute warn_unused_result frameworks/base/core/jni/android_server_Watchdog.cpp:43: warning: ignoring return value of ‘ssize_t write(int, const void*, size_t)’, declared with attribute warn_unused_result frameworks/base/core/jni/android_server_Watchdog.cpp:47: warning: ignoring return value of ‘ssize_t write(int, const void*, size_t)’, declared with attribute warn_unused_result frameworks/base/core/jni/android_server_Watchdog.cpp: In function ‘void dumpKernelStacks(JNIEnv*, _jobject*, _jstring*)’: frameworks/base/core/jni/android_server_Watchdog.cpp:73: warning: ignoring return value of ‘ssize_t write(int, const void*, size_t)’, declared with attribute warn_unused_result frameworks/base/core/jni/android_server_Watchdog.cpp:91: warning: ignoring return value of ‘ssize_t write(int, const void*, size_t)’, declared with attribute warn_unused_result In function ‘int open(const char*, int, ...)’, inlined from ‘void dumpKernelStacks(JNIEnv*, _jobject*, _jstring*)’ at frameworks/base/core/jni/android_server_Watchdog.cpp:66: /usr/include/bits/fcntl2.h:51: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments make: *** [out/debug/host/linux-x86/pr/sim/obj/SHARED_LIBRARIES/libandroid_runtime_intermediates/android_server_Watchdog.o] Error 1 t...@ubuntu:/media/ANDROID/Froyo$ Any body know how to fix this bug ? Thanks -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] libc in dynamic linker
The dynamic linker uses a special version of the static c library. The shared c library is used by other shared libraries and dynamic executables. Prelinking is only an optimization used to speed up shared library loading and reducing their sizes. It is not required. Regards. On Friday, July 16, 2010, Liu Xin navy.x...@gmail.com wrote: Hi Developers, i can't work out this problem. how linker uses libc? dynamic liner in android(bionic/linker) seems to link bionic(libc) without abnormal things. is the libc static or dynamic? if it is static, then why we bother build it to a shared library? or if it is static, can normal program reuse the static libc later on? btw, is prelink(apriorio) optional? if i globally forbids it in build, can android system still start up? Thanks, --lx -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Android Kernel Vs Linux Kernel.
On Wed, Jul 7, 2010 at 5:57 AM, sanjeev sharma sanjeevsharmae...@gmail.comwrote: Hi, Let me tell you that to run Android on Linux kernel you need Android specific driver to Linux kernel like. Belows is feature of an Android based kernel. 01). GoldFish : Qemutrace , FB Driver/RTC/Power/Nand/MMC/Events/tty/ audio for godlfish For the record, Goldfish is the name of the virtual ARM board that is emulated by the Android emulator. You don't need any goldfish-specific stuff to run on a real device. Hope this helsp. 02). pmem : Android pmem allocator 03). yaffs : mtdif2 04). kernel debugger 05). android USB : USB Gadget driver for android. 06). HAL : TI Wifi Control Functions , Bluetooth (RFCOMM-L2CAP Socket) 07). Android Shared Memory (ASHMEM) 08). Paranoid Network 09). Binder: mmap fix , more offset validation 10). Power : User Wake Lock , Wake Lock , Early Suspend , FB Eearlysuspend 11). Logger Subsystem 12). Low Memory Killer: /sys/module/lowmemorykiller/parameters/adj and minfree 11). Etc : Timed GPIO / Timed Output Above is an difference between Normal Linux kernel and Android Kernel.Above Driver need to integrate to run Android on Linux kernel. Sanjeev sharma On Mon, Jul 5, 2010 at 3:49 PM, padmapriyak padmapri...@tataelxsi.co.inwrote: Hi, I want to port Android Fryo(linux version 2.6.32) on platform which is also using modified version of linux kernel(2.6.32). I compared 2.6.32.9 kernel and android kernel(modified kernel 2.6.32). and found changes.Some of the changes includes 1.Support of Numeric and alpha numeric password. 2.YAFFS support. 3.Power Management 4.Memory Management to enhance shared memory 5.Touchscreen. My query is, *Is there anything that is specific to android in kernel level. (i.e)any android specific code.If so,tell me what are all the changes mandate to bring android kernel?If not,then how android kernel differs from Linux Kernel?* Note: * ** *Many post says Compare linux kernel and bring android specific change.What does that android specific change mean?. Regards, Padma. -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] boot speed -- scan package, IO buffering, and prefetch
The problem is that you didn't count the time to do the cat in your overall benchmark. Hence, the savings you present are purely hyptohetical, even if you implement prefetching. On Sun, Jun 20, 2010 at 4:31 PM, Tomei tomei.ninge...@gmail.com wrote: Hello, It seems package scanning speed is greatly influenced by IO buffering: I did some simple experiment. I am testing with a fresh build so there are NO apps under /data/app. First boot up the phone. Then do either (A) or (B): (A): # stop # sync; echo 3 /proc/sys/vm/drop_caches # start (B): # stop # sync; echo 3 /proc/sys/vm/drop_caches # cat /system/app/*.apk /dev/null # start case (A): I/PackageManager( 3114): Time to scan packages: 12.668 seconds case (B): I/PackageManager( 3189): Time to scan packages: 7.032 seconds Can PackageManagerService be easily reworked so that it pre-fetches the buffers for the next app to scan in a separate thread? I can't just blindly run a cat /system/app/*.apk /dev/null before package scanning happens, because doing so would cost more time than the saving indicated from above Thanks! -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Android Openssl library - libcrypto.so and libssl.so
Remove libdl from your LOCAL_SHARED_LIBRARIES, and add LOCAL_LDLIBS += -ldl instead in your Android.mk The Android.mk is for a full system/firmware build, and assumes that libdl is going to be built at the same time than libssl. For the NDK, you need to tell your libcrypto/libssl to rely on the version that is already installed on the target system. On Mon, Jun 21, 2010 at 10:45 PM, Abhyudai Shanker abhyudai.shan...@gmail.com wrote: I am tryin to 'make' the code I have downloaded from the link : http://www.google.com/url?sa=Dq=http://android.git.kernel.org/%3Fp%3Dplatform/external/openssl.git%3Ba%3Dsummaryusg=AFQjCNHSVU0QjNdHMsLAuZX-ln7YKqEmCA This code provides libcrypto.so and libssl.so which I am interested in. I am basically trying to get *libcrypto.so and libssl.so* which would required to run my openssl based code on my htc nexus one phone. However, when i try to run the above downloaded code, it compiles all the files required for libcrypto.so but gives the error: make: *** No rule to make target `/cygdrive/c/android-ndk-r4/samples/DMVPNC/bin/ ndk/local/armeabi/libdl.so', needed by `/cygdrive/c/android-ndk-r4/samples/DMVPN C/bin/ndk/local/armeabi/libcrypto.so'. Stop. However, other projects which require the same libdl.so are getting it. I am not able to understand why this is happening. Please guide me as to how i should proceed Looking forward to hear from you Abhyudai Shanker -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Building android without debug symbols
I'm not sure to understand what you want, but the debug symbols are removed from the system images by default by the build. On Mon, Jun 14, 2010 at 4:21 AM, yogesh embedded.yog...@gmail.com wrote: Hi, I am trying to build android eclair without debug symbols for hawkboard, so that its rootfs size is less and takes less time to execute . Can any one know how to disable all the debug symbols from android eclair. Regard Yogesh -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] How to make the emulator work with the Custom Android IMage
use -show-kernel to get kernel messages, and -logcat '*:v' to get logcat traces, that should tell you a lot about what's going wrong. On Thu, May 27, 2010 at 11:32 PM, gopu gopuraj...@gmail.com wrote: Hi , Created a custom Android Image Its working in real hardware.But in emulator its not working. How to make it work in emulator ?.I just replaced the system.img ramdisk.img in the SDK/platforms/platfo../image/ Is their a way to debug in emulator env? Thanks Regards Gopu -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Regarding Toolchain
The toolchain sources live at kernel.org, to get them try: repo init -u android.git.kernel.org/toolchain/manifest.git -b master repo sync Alternatively, NDK r4 comes with several scripts that can help your download / build / install them. See ndk/build/tools/rebuild-all-prebuilt.sh On Fri, May 21, 2010 at 12:10 AM, patel akhipa...@gmail.com wrote: Hi All, I am very new in Android and i want to compile android toolchain . can some one guide from where i can download android toolchain and supported library. Thanks, Ak -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: Emulator's /dev/ttyS0
On Fri, Apr 30, 2010 at 12:41 AM, ari das.ar...@gmail.com wrote: Hi Coolio, Your input helped a lot.I had the similar kind of issue.I have an application which has to be run from the android shell.The requests from the application will go to my library and it sends it to the /dev/ttyS0.I have a portlistener in the host windows machine which will listen to the port given as emulator option . the command is emulator.exe -avd androidavd -verbose -qemu -serial COM9 But after that there was coming to the portlistener. Then I changed by library port options to /dev/ttyS1, but that also didnt work.After that I changed /dev/ttyS2 than i was able to see something coming into the portlistener. I want to ask coolio that how did u come to know that it will be /dev/ ttys2?? Because the Android emulator already uses /dev/ttyS0 and /dev/ttyS1 in the guest system for its own usage (/dev/ttyS0 is to get kernel messages, like thoses displayed with -show-kernel, and /dev/ttyS1 is to implement the qemud multiplexing daemon, an internal implementation detail you should not care about). So when you do -qemu -serial COM9, it really means allocate the next available emulated serial port device and connect it to the host COM9, hence the /dev/ttyS2 Please let me know a about your findings. On Jan 23 2009, 5:20 am, coolio justinca...@gmail.com wrote: Thanks, What is the default configuration for /dev/ttyS2 and what is the command to change the configuration? Justin On Jan 22, 1:51 pm, David Turner di...@android.com wrote: if you're using the -qemu -serial device option when starting the emulator, it will probably be mapped to /dev/ttyS2 or ttyS3 within the emulator. that's because the emulator uses a couple of serial ports for its own needs: one is to get kernel messages, and the other for the qemud multiplexing daemon which manages a communication channel between the emulator program and various Android sub-systems (e.g. the GSM stack, the GPS support, etc...) you can use emulator -verbose to see which -serial dev are inserted before yours when starting the emulator, if that can help On Thu, Jan 22, 2009 at 11:52 AM, coolio justinca...@gmail.com wrote: Hi, I run SDK's emulator and opened /dev/ttyS0, when I write to that device , it didn't show up in host's serial port. Does anybody konw what /dev/ttyS0 in emulator is mapped to? Is there any possibility that I can map emulator's /dev/ttyS0 to /dev/ttyS0 of the linux host that is running emulator. If I add a serial port on linux host, can I add/map the serial port to emulator? Thanks- Hide quoted text - - Show quoted text - -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Downloading android source through git
android-x86.org has nothing to do with the official Android open-source tree. I recommend you ask on their forums instead, we won't be able to help you there. On Thu, Apr 15, 2010 at 11:21 PM, suhas suvisu...@gmail.com wrote: Hi, I want to download the android source for x86. I tried git and it always gives curl: (6) Couldn't resolve host 'android.git.kernel.org' . So i am thinking of downloading it through windows through mysgit installer. Could you please tell me how to configure the installer and fully download the android source code at git://git.android-x86.org/platform/manifest.git. or 'git://android- x86.git.sf.net/gitroot/android-x86/manifest.git -b donut-x86' Thanks all -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: Question on building modified framework code + windows emulator
The problem is that the Cygwin gcc doesn't support the -mno-cygwin flag anymore. Try downgrading gcc to an earlier version instead, this should fix it. On Wed, Mar 24, 2010 at 11:48 AM, MBethDev mbeth...@gmail.com wrote: Hi Jett, thanks for replying. Tried following your suggestion but ran into issues. Are you able to build successfully? Anyway, here's the errors I get: 1) After running make_windows_sdk.sh, I get what looks like an error from find-jdk-tools-jar.sh: build/core/find-jdk-tools-jar.sh: line 5: [: /c/Program: binary operator expected The problem being my javac is installed in Program Files and cygwin wasn't able to parse the whitespace properly. Yet the build continued. I'm not sure if this is critical though, but I let the build continue anyway 2) After some time I get the following build error: gcc: gcc: gcc: gcc: The -mno-cygwin flag has been removed; use a mingw- targeted cross-compiler. The -mno-cygwin flag has been removed; use a mingw-targeted cross- compiler. The -mno-cygwin flag has been removed; use a mingw-targeted cross- compiler.The -mno-cygwin flag has been removed; use a mingw-targeted cross-compiler. make: *** [out/host/windows-x86/obj/EXECUTABLES/emulator_intermediates/ distrib/zlib-1.2.3/adler 32.o] Error 1 make: *** Waiting for unfinished jobs make: *** [out/host/windows-x86/obj/EXECUTABLES/emulator_intermediates/ distrib/zlib-1.2.3/defla te.o] Error 1 make: *** [out/host/windows-x86/obj/EXECUTABLES/emulator_intermediates/ distrib/zlib-1.2.3/crc32 .o] Error 1 make: *** [out/host/windows-x86/obj/EXECUTABLES/emulator_intermediates/ distrib/zlib-1.2.3/compr ess.o] Error 1 Error: Build failed Aborting How do I set the -mno-cygwin flag (what I mean is, where in the android build scripts and how to set it there?) I also checked out the development/tools/eclipse/README_WINDOWS.txt and it says to set the Default Text File to DOS/text. How do I set this? On Mar 17, 7:33 pm, Jett jett...@gmail.com wrote: Hi MBethDev, The sdk should build under windows system. You should build sdk under Ubuntu first(which you already did) and copy all files into Cygwin(in Windows) system. Then use /development/build/tools/make_windows_sdk.sh to accomplish it. On 3月16日, 上午10時34分, MBethDev mbeth...@gmail.com wrote: Hello, I'm not sure if this is the right forum to ask, but I'll ask anyway in case someone knowledgeable might know this: I recently added some functionality in the android framework code (more specifically I *added* a new class that encapsulates this functionality that I'm trying to write, and I'm using android-1.6), built it on Ubuntu, and was able to run an application that uses that functionality on the Ubuntu emulator. However, when I tried to import the framework images as well as the android.jar file over to my windows environment (where I code most of the time), the compilation failed (I'm using Eclipse on Windows to build my application). I'm getting the following errors: [2010-03-15 19:13:54 - MyApp]C:\MyApp\res\values\styles.xml:6: ERROR Error retrieving parent for item: No resource found that matches the given name 'android:style/Theme.Translucent'. [2010-03-15 19:13:54 - MyApp]C:\MyApp\res\values\styles.xml:9: ERROR Error: No resource found that matches the given name: attr 'android:colorForeground'. [2010-03-15 19:13:54 - MyApp]C:\MyApp\res\values\styles.xml:10: ERROR Error: No resource found that matches the given name: attr 'android:windowAnimationStyle'. [2010-03-15 19:13:54 - MyApp]C:\MyApp\res\values\styles.xml:7: ERROR Error: No resource found that matches the given name: attr 'android:windowBackground'. [2010-03-15 19:13:54 - MyApp]C:\MyApp\res\values\styles.xml:8: ERROR Error: No resource found that matches the given name: attr 'android:windowNoTitle'. I basically copied the three *.img files to the platform img folder on Windows, copied the android.jar file as well as the platform/ android-1.6/data directory that was created when I ran make sdk on Ubuntu, and still get these errors. Am I missing something here? If anyone can help me I will truly appreciate it. MBeth- Hide quoted text - - Show quoted text - -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting To unsubscribe from this group, send email to android-porting+ unsubscribegooglegroups.com or reply to this email with the words REMOVE ME as the subject. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting To unsubscribe, reply using remove me as the subject.
Re: [android-porting] Re: Why Android 2.1 works so slow
On Sat, Apr 10, 2010 at 8:19 AM, supermaximus79 andruschenk...@mail.ruwrote:
You mean hardware OpenGL drivers?
Because from log file i see that Android uses software OpenGL:
libGLES_android.so
Yes, Android provides a software implementation has a fallback, but
performance is not its second name.
I/SurfaceFlinger( 1731): SurfaceFlinger is starting
I/SurfaceFlinger( 1731): SurfaceFlinger's main thread ready to run.
Initializing graphics H/W...
init: event { 'add', '/class/vc/vcs7', 'vc', '', 7, 7 }
init: event { 'add', '/class/vc/vcsa7', 'vc', '', 7, 135 }
E/SurfaceFlinger( 1731): Couldn't open /sys/power/wait_for_fb_sleep
or /sys/power/wait_for_fb_wake
I/gralloc ( 1731): using (fd=22)
I/gralloc ( 1731): id = s3c-fb
I/gralloc ( 1731): xres = 800 px
I/gralloc ( 1731): yres = 480 px
I/gralloc ( 1731): xres_virtual = 800 px
I/gralloc ( 1731): yres_virtual = 960 px
I/gralloc ( 1731): bpp = 16
I/gralloc ( 1731): r= 11:5
I/gralloc ( 1731): g= 5:6
I/gralloc ( 1731): b= 0:5
I/gralloc ( 1731): width= 127 mm (160.00 dpi)
I/gralloc ( 1731): height = 76 mm (160.421051 dpi)
I/gralloc ( 1731): refresh rate = 60.16 Hz
E/FramebufferNativeWindow( 1731): xDpi -2046820352
E/FramebufferNativeWindow( 1731): yDpi -2046820352
D/libEGL ( 1731): egl.cfg not found, using default config
D/libEGL ( 1731): loaded /system/lib/egl/libGLES_android.so
I/SurfaceFlinger( 1731): EGL informations:
I/SurfaceFlinger( 1731): # of configs : 8
I/SurfaceFlinger( 1731): vendor: Android
I/SurfaceFlinger( 1731): version : 1.4 Android META-EGL
I/SurfaceFlinger( 1731): extensions: EGL_KHR_image
EGL_KHR_image_base
EGL_KHR_image_pixmap EGL_ANDROID_image_init: no such service
'bootanim'
native_buffer EGL_ANDROID_swap_rectangle
EGL_ANDROID_get_render_buffer
I/SurfaceFlinger( 1731): Client API: OpenGL ES
I/SurfaceFlinger( 1731): EGLSurface: 5-6-5-0, config=0x100
I/SurfaceFlinger( 1731): flags : 001c
W/SurfaceFlinger( 1731): ro.sf.lcd_density not defined, using 160
dpi by default.
I/SurfaceFlinger( 1731): OpenGL informations:
I/SurfaceFlinger( 1731): vendor: Android
I/SurfaceFlinger( 1731): renderer : Android PixelFlinger 1.1
I/SurfaceFlinger( 1731): version : OpenGL ES-CM 1.0
I/SurfaceFlinger( 1731): extensions: GL_OES_byte_coordinates
GL_OES_fixed_point GL_OES_single_precision GL_OES_read_format
GL_OES_compressed_paletted_textu
On 10 апр, 16:23, David Turner di...@android.com wrote:
2010/4/9 supermaximus79 andruschenk...@mail.ru
I have built in chip s3c6410 2d and 3d hardware graphics accelerators,
but i don't use them with Android.
I just implemented in my framebuffer driver double buffering and pan
function, which change offset from frame buffer start.
I this not enough to make Android drawing fast?
Probably, software composition is quite CPU intensive, especially if you
have large amounts of pixels
(800x480 should be considered large, especially for an ARMv6)
I know that it's possible to use hardware accelerators with Android,
but i didn't find an examples.
I downloaded Samsung i5700 mobile phone Android sources, this phone
based on chip i use - s3c6410, but found nothing about graphics
accelerators in them.
I don't know what are the Samsung i5700 Android sources, but you're
probably missing OpenGL ES drivers.
Could you help me with information about hardware accelerations with
Android.
Where from should i start?
Thanks!
On Apr 9, 8:22 pm, David Turner di...@android.com wrote:
do you have hardware graphics acceleration or are you relying solely
on
software rendering ?
2010/4/9 Максим Андрущенко andruschenk...@mail.ru
Hi!
On my target platform, based on s3c6410 (armv6) Android works very
slow!
I use 16 bpp with resolution 800x480.
Screen updates slow, seldom i see some blinking, when i quickly
type on
soft keyboard, or if to change desktop.
Is there anybody has the same problem, is it possible to fix it?
--
unsubscribe:
android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com
android-porting%2bunsubscr...@googlegroups.comandroid-porting%252bunsubscr...@googlegroups.com
android-porting%2bunsubscr...@googlegroups.comandroid-porting%252bunsubscr...@googlegroups.com
android-porting%252bunsubscr...@googlegroups.comandroid-porting%25252bunsubscr...@googlegroups.com
website:http://groups.google.com/group/android-porting
To unsubscribe, reply using remove me as the subject.
--
unsubscribe:
android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com
android-porting%2bunsubscr...@googlegroups.comandroid-porting%252bunsubscr...@googlegroups.com
website:http://groups.google.com/group/android
Re: [android-porting] Why Android 2.1 works so slow
do you have hardware graphics acceleration or are you relying solely on software rendering ? 2010/4/9 Максим Андрущенко andruschenk...@mail.ru Hi! On my target platform, based on s3c6410 (armv6) Android works very slow! I use 16 bpp with resolution 800x480. Screen updates slow, seldom i see some blinking, when i quickly type on soft keyboard, or if to change desktop. Is there anybody has the same problem, is it possible to fix it? -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting To unsubscribe, reply using remove me as the subject. -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Running emulator from SDK built on android source
emulator -verbose -show-kernel -logcat '*:v' -avd avd1 that might help you understand what's happening here On Thu, Mar 11, 2010 at 11:12 PM, Anandi heyan...@gmail.com wrote: HI all, I have built the android source and SDK on my Ubuntu machine successfully. I am now trying to run the emulator generated in this sdk. But the mobile screen on emulator appears blank and adb commands executed keeping this emulator running give device offline error. To launch emulator i followed these steps: 1. android create avd -n avd1 -t 1 -sdcard 256M 2. emulator -avd avd1 3. adb devices Can anyone please suggest the solution to this problem? Thanks, Anu -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: Set system time from radio
The Android code only assumes UTC, except that you can use a small extension to the NITZ string to encode the timezone too. See the comment and implementation in handleOfInit() in external/qemu/telephony/android_modem.c here: http://android.git.kernel.org/?p=platform/external/qemu.git;a=blob;f=telephony/android_modem.c;h=6e73411d9fdd706b8f369684c03451e6e8d7221a;hb=master On Fri, Mar 5, 2010 at 2:47 PM, ms12 mstic...@gmail.com wrote: Ok, got that to work, thanks. Discovered that the time comes back in local time zone, not UTC. Android code assumes UTC. The timezone comes with it of course, but if you don't know if it has already been applied, I would think it difficult to determine in the code. Is there a standard for this? Mike On Mar 3, 5:22 pm, David Turner di...@android.com wrote: Yes, you can use a NITZ unsolicited message. See hardware/ril/reference-ril/reference-ril.c:onUnsolicited() On Wed, Mar 3, 2010 at 2:25 PM, ms12 mstic...@gmail.com wrote: Is there a way to set the system time from the cell phone? I've searched ril.h and ril_commands.h and there seems to be no way to request the time from the cell modem. Am I missing something? Mike -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com android-porting%2bunsubscr...@googlegroups.comandroid-porting%252bunsubscr...@googlegroups.com website:http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] cupcake emulator crash on Windows
From the error message, it looks like the latest versions of Cygwin have removed support for the -mno-cygwin flag. I believe we're working on a fix, in the meantime, try to install an older Cygwin version (if that is possible, I sincerely don't know yet) On Wed, Mar 3, 2010 at 1:27 AM, Santosh Singh santo...@googlemail.comwrote: Hi There, I am getting following error when building Android Windows SDK in emulator. Can you help me please? $ development/build/tools/make_windows_sdk.sh out/host/linux-x86/sdk/android-sdk_eng.root_linux-x86.zip ../mysdk/ Using SDK 1.5 Current values: - Input SDK: out/host/linux-x86/sdk/android-sdk_eng.root_linux-x86.zip - Output dir: ../mysdk/ - Temp dir: /cygdrive/c/DOCUME~1/SANTOS~1.SIN/LOCALS~1/Temp Building... including vendor/aosp/vendorsetup.sh PLATFORM_VERSION_CODENAME=REL PLATFORM_VERSION=1.6 TARGET_PRODUCT=generic TARGET_BUILD_VARIANT=eng TARGET_SIMULATOR= TARGET_BUILD_TYPE=release TARGET_ARCH=arm HOST_ARCH=x86 HOST_OS=windows HOST_BUILD_TYPE=release BUILD_ID=Donut build/core/main.mk:360: sdk-only: javac available. build/core/main.mk:381: sdk-only: javac not available. host StaticLib: emulator-hw (out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/emulator-hw.a) host StaticLib: emulator-arm (out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/emulator-arm.a) host StaticLib: emulator-tcg (out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-tcg_intermediates/emulator-tcg.a) host C++: libhost = build/libs/host/pseudolocalize.cpp ar.exe crs out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/emulator-arm.a out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/exec.o out/host /windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/cpu-exec.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/target-arm/op_helper.o out/host/windows -x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/target-arm/iwmmxt_helper.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/target-arm/neon_helper.o out/ho st/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/target-arm/helper.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/target-arm/translate.o out/h ost/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/target-arm/machine.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/translate-all.o out/host/w indows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/hw/armv7m.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/hw/armv7m_nvic.o out/host/windows-x86/ob j/STATIC_LIBRARIES/emulator-arm_intermediates/arm-semi.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/trace.o out/host/windows-x86/obj/STATIC_LIBRARIES/emul ator-arm_intermediates/varint.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/dcache.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-arm_intermediates/f pu/softfloat.o ar.exe crs out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-tcg_intermediates/emulator-tcg.a out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-tcg_intermediates/tcg/tcg.o out/h ost/windows-x86/obj/STATIC_LIBRARIES/emulator-tcg_intermediates/tcg/tcg-dyngen.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-tcg_intermediates/tcg/tcg-runtime.o ar.exe crs out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/emulator-hw.a out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/android_arm.o o ut/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/arm_pic.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/cdrom.o out/host/windows-x86/ obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/dma.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/irq.o out/host/windows-x86/obj/STATIC_LIBRARIES/emula tor-hw_intermediates/hw/goldfish_audio.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/goldfish_battery.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulato r-hw_intermediates/hw/goldfish_device.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/goldfish_events_device.o out/host/windows-x86/obj/STATIC_LIBRARIES/em ulator-hw_intermediates/hw/goldfish_fb.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/goldfish_interrupt.o out/host/windows-x86/obj/STATIC_LIBRARIES/emula tor-hw_intermediates/hw/goldfish_memlog.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/goldfish_mmc.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-h w_intermediates/hw/goldfish_nand.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_intermediates/hw/goldfish_switch.o out/host/windows-x86/obj/STATIC_LIBRARIES/emulator-hw_in
Re: [android-porting] executable directories under /system partition
Yes, it's an Android-specific security feature. Look at system/core/include/private/android_filesystem_config.h On Sun, Jan 24, 2010 at 8:46 PM, Jun Sun monkey.j...@gmail.com wrote: This might be a dumb question, but I can't seem to figure it out. In my donut build tree, I added some additional directories to out/target/product/dream-open/system/. Just as I expected, next re-make actually included the new directories in the system.img file. However, during run-time, I found that all the executables I put under the new directories have their x bit stripped and hence they cannot execute any more. Is this a security feature of Android? Where is this feature implemented? I can't find anything in the build log that shows the execution bit stripping. Or is it implemented at run-time? I traced /init.rc and could not find anything there either. I am just confused at this point. Cheers. Jun -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: running emulator build on actual platform
On Thu, Feb 25, 2010 at 9:49 PM, Ajay meajaypil...@gmail.com wrote: Thanks. So you mean to say that when i compile a native code using the prebuilt toolchain of Android for emulator, the ARM instructions generated is in no way limited by some limitations of the emulator(if any). Yes, that's what I'm saying :-) ajay On Feb 26, 12:55 am, David Turner di...@android.com wrote: As far ar I know, the system images that are provided with the SDK are built in the same way than the ones that run on retail Dream/G1 and Magic/MyTouch devices. In other words, they are not tuned specifically for the emulator (but they contain emulator-specific code that is not part of the system image of production devices). Also, the kernel images must be different between the emulator and a real device, but you probably already knew that :-) On Tue, Feb 23, 2010 at 7:55 AM, Ajay meajaypil...@gmail.com wrote: Hi, I am running an Android image built for the emulator on an actual ARMv6 platform. Although it does work perfectly well, i am curious if i am taking a performance hit due to that. I wonder if the android emulator has a incomplete support for the ARM instruction set and the prebuilt toolchain is tuned for that constraint of the emulator. thank you ajay -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com android-porting%2bunsubscr...@googlegroups.comandroid-porting%252bunsubscr...@googlegroups.com website:http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Set system time from radio
Yes, you can use a NITZ unsolicited message. See hardware/ril/reference-ril/reference-ril.c:onUnsolicited() On Wed, Mar 3, 2010 at 2:25 PM, ms12 mstic...@gmail.com wrote: Is there a way to set the system time from the cell phone? I've searched ril.h and ril_commands.h and there seems to be no way to request the time from the cell modem. Am I missing something? Mike -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] running emulator build on actual platform
As far ar I know, the system images that are provided with the SDK are built in the same way than the ones that run on retail Dream/G1 and Magic/MyTouch devices. In other words, they are not tuned specifically for the emulator (but they contain emulator-specific code that is not part of the system image of production devices). Also, the kernel images must be different between the emulator and a real device, but you probably already knew that :-) On Tue, Feb 23, 2010 at 7:55 AM, Ajay meajaypil...@gmail.com wrote: Hi, I am running an Android image built for the emulator on an actual ARMv6 platform. Although it does work perfectly well, i am curious if i am taking a performance hit due to that. I wonder if the android emulator has a incomplete support for the ARM instruction set and the prebuilt toolchain is tuned for that constraint of the emulator. thank you ajay -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: gdbserver - illegal instruction
Please start here: http://android.git.kernel.org/?p=toolchain/gdb.git;a=summary On Tue, Feb 9, 2010 at 7:30 AM, Jim Ancona scarh...@gmail.com wrote: On Feb 9, 6:34 am, Mobile Enthusiast mobenthusi...@gmail.com wrote: Hi, I'm trying to run gdbserver on the Neo Freerunner, but I keep getting an illegal instruction error. It might be because the gdbserver executable i'm using is the one from the prebuilt/android-arm/ directory and that's armv5 whereas I need armv4. Question: How do I confirm whether this is indeed the case? Where can I get a gdbserver executable that is armv4? I tried cross-compiling, but that didn't work. May have done something wrong with my cross-compile, but I'm not sure. If suggestion is to cross-compile for v4, can anyone point me to a resource? They've been scarce. In particular, where is the source for the the gdbserver executable included in prebuilt/android-arm/gdbserver/ ? The README.txt in that directory says: To compile a new gdbserver, see //toolchain/gdb/src/gdb/gdbserver/README_ANDROID.txt But there is no toolchain/gdb/src/ directory in the source tree. Presumably if we had the source we could build an executable with our armv4t toolchain. Jim -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Accessing the NFS server from Android : RPC: Unknown protocol issue
Well, the system is probably simply missing /etc/protocol and/or /etc/services since that's what GLibc does. Did you try copying a version from your Linux machine to a system image to see if this solves the issue ? On Wed, Jan 20, 2010 at 7:35 AM, Kishor krish2...@gmail.com wrote: Hi David, Thanks for your reply. I have compiled the client application statically with arm-linux-gcc tool chain. The executable is linked to glibc statically. So here when i run the application, glibc part should search the protocol file in /etc folder? Does the android not allowing this? Is there any alternative? The reason behind compiling application statically is that bionic doesn't support all required function/library. Also for accessing the NFS server the required services(NFS UTILS, PORTMAP) are compiled statically and running fine on target. Thanks, krish On Wed, Jan 20, 2010 at 5:34 AM, David Turner di...@android.com wrote: And /etc/protocols is not supported either. See bionic/libc/bionic/stubs.c for the getprotobyname() implementation. On Tue, Jan 19, 2010 at 4:03 PM, David Turner di...@android.com wrote: /etc/services is not supported in Android. Instead the C library provides a hard-coded list. See bionic/libc/netbsd/net/services.h for the current list, and bionic/libc/tools/genserv.py for a Python script to update it. Hope this helps. On Tue, Jan 19, 2010 at 7:34 AM, krish24 krish2...@gmail.com wrote: Hi All, I am running a client application in target which receives the read/ write request from Modem side. To serve the read/write request for files, application access the NFS file system(Local NFS). The client uses the RPC node to receive the Modem read/write request. It access the NFS file system on Host side through the SUNRPC. Issue: When we run the application on target , we receive “127.0.0.1: RPC: Unknown protocol”. Here the Client application is calling the “clnt_create” which gets the handle on a remote host where the local server is located. The one of the parameter of “clnt_create” is “protocol” and the value of which is “UDP”. I tested the same application at Kernel Level rootfs and it works fine. There is “protocol” file present in Kernel rootfs /etc folder. This protocol file contains the list of protocols and the protocol numbers understood by the IP layer. I copied the protocol file in Android rootfs /etc folder but application still gives the same warning. Where i should define the protocols in Android? Is that warning coming because of protocol file not recognized by Android or something else needs to be done for setting the NFS on Android? Will appreciate your reply and suggestions. Thanks, krish -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Accessing the NFS server from Android : RPC: Unknown protocol issue
/etc/services is not supported in Android. Instead the C library provides a hard-coded list. See bionic/libc/netbsd/net/services.h for the current list, and bionic/libc/tools/genserv.py for a Python script to update it. Hope this helps. On Tue, Jan 19, 2010 at 7:34 AM, krish24 krish2...@gmail.com wrote: Hi All, I am running a client application in target which receives the read/ write request from Modem side. To serve the read/write request for files, application access the NFS file system(Local NFS). The client uses the RPC node to receive the Modem read/write request. It access the NFS file system on Host side through the SUNRPC. Issue: When we run the application on target , we receive “127.0.0.1: RPC: Unknown protocol”. Here the Client application is calling the “clnt_create” which gets the handle on a remote host where the local server is located. The one of the parameter of “clnt_create” is “protocol” and the value of which is “UDP”. I tested the same application at Kernel Level rootfs and it works fine. There is “protocol” file present in Kernel rootfs /etc folder. This protocol file contains the list of protocols and the protocol numbers understood by the IP layer. I copied the protocol file in Android rootfs /etc folder but application still gives the same warning. Where i should define the protocols in Android? Is that warning coming because of protocol file not recognized by Android or something else needs to be done for setting the NFS on Android? Will appreciate your reply and suggestions. Thanks, krish -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Accessing the NFS server from Android : RPC: Unknown protocol issue
And /etc/protocols is not supported either. See bionic/libc/bionic/stubs.c for the getprotobyname() implementation. On Tue, Jan 19, 2010 at 4:03 PM, David Turner di...@android.com wrote: /etc/services is not supported in Android. Instead the C library provides a hard-coded list. See bionic/libc/netbsd/net/services.h for the current list, and bionic/libc/tools/genserv.py for a Python script to update it. Hope this helps. On Tue, Jan 19, 2010 at 7:34 AM, krish24 krish2...@gmail.com wrote: Hi All, I am running a client application in target which receives the read/ write request from Modem side. To serve the read/write request for files, application access the NFS file system(Local NFS). The client uses the RPC node to receive the Modem read/write request. It access the NFS file system on Host side through the SUNRPC. Issue: When we run the application on target , we receive “127.0.0.1: RPC: Unknown protocol”. Here the Client application is calling the “clnt_create” which gets the handle on a remote host where the local server is located. The one of the parameter of “clnt_create” is “protocol” and the value of which is “UDP”. I tested the same application at Kernel Level rootfs and it works fine. There is “protocol” file present in Kernel rootfs /etc folder. This protocol file contains the list of protocols and the protocol numbers understood by the IP layer. I copied the protocol file in Android rootfs /etc folder but application still gives the same warning. Where i should define the protocols in Android? Is that warning coming because of protocol file not recognized by Android or something else needs to be done for setting the NFS on Android? Will appreciate your reply and suggestions. Thanks, krish -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Simulating accelerometer events natively?
Hello, It would be easier to modify the emulator source code instead to read data directly from the OpenIntents simulator (as an example). See external/qemu/android/hw-sensors.c for the current sensor hardware emulation code, which is limited to rotating the virtual device at the moment. Hope this helps On Wed, Jan 13, 2010 at 8:10 AM, Bill Gatliff bgatl...@gmail.com wrote: Guys: I'm trying to come up with a way to simulate an accelerometer under Goldfish. I'm familiar with the OpenIntents sensor simulator, but I want something that doesn't require me to modify my source code. The file development/emulator/sensors/sensors_qemu.c looks like it might be the source code for sensors.goldfish.so. If that's the case, then can I use the data_poll() function in the former to fake an accelerometer event? That's assuming I link the .so file into my own code, build with the Android cross-compiler, etc. etc. etc. of course. Regards, b.g. -- Bill Gatliff Embedded Linux development, training and consulting http://billgatliff.com b...@billgatliff.com -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] ADBD not binding to port 5555
Port is only used for local transports to emulator instances, not real devices. Iirc, the adbd daemon running on a device is listening on local port 5037. On Fri, Dec 18, 2009 at 4:26 AM, Hamilton Vera hamilton.lis...@gmail.comwrote: Greetings everyone. I am trying to deploy android 1.6 in an iMX51 based device. Almost everything is running ok. I am also trying to use adb over ethernet, unfortunately it is not working, it looks like adbd is not binding to port . I saw the same problem here in the list: http://www.mail-archive.com/android-porting@googlegroups.com/msg07198.html Here is what I got from a netstat (I am using a serial cable to connect to the device): # netstat Proto Recv-Q Send-Q Local Address Foreign AddressState tcp0 0 127.0.0.1:5037 0.0.0.0:* LISTEN tcp0 0 192.168.0.113:667 192.168.0.199:2049 ESTABLISHEDNFS I would appreciate any help, I am googling a lot about this but no answer yet. -- Hamilton Vera int Administrator (char Network[],char ComputationalSystems[]) http://hvera.wordpress.com -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Negative if else condition in Android.mk
if*n*eq ($(TARGET_BUILD_VARIANT),user) ... endif On Tue, Nov 24, 2009 at 8:40 AM, vj vjvdixit.l...@gmail.com wrote: Hi, is there a way to implement a negative if/else statement within Android.mk. Example: I do not want the build to be of type user: So I would like something like this: ifeq ( ! ($(TARGET_BUILD_VARIANT), user) LOCAL_SHARED_LIBRARIES += liblog endif basically, I want liblog to be included in only USERDEBUG and ENG builds. Any help would be appreciated. Thanks, Vijay -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
Re: [android-porting] Re: sigsuspend() hang-up
Probably because Linux essentially delivers signals to random threads. So the signal handler is called but not in the thread that is doing the sigsuspend. You probably need to mask SIGIO in all other threads of your process, which sucks. A better alternative is to try something different. For example, create a pipe() and have the signal handler send a single byte through write() into it. Your other thread can be blocked on read()-ing the byte, and this would work independently of the thread receiving the signal. Be sure to catch for EINTR during the write() and read() calls though :-) On Tue, Nov 10, 2009 at 12:27 PM, fadden fad...@android.com wrote: On Nov 9, 10:18 pm, Shankar Ganesh shankargane...@gmail.com wrote: Here, SIGIO is received by signal handler, but my user-space process (open core decoder node process) is hanged and never wakes-up and hence blocked within sigsuspend() . Can anyone throw some ideas if you have faced similar problem ? I don't quite understand -- you say the process receives the SIGIO, but it's hung and never wakes up. If the process is hung, how do you know it received the signal? Are you sure the signal is being delivered to the correct thread within the process? FWIW, the VM uses sigwait() to do something similar. -- unsubscribe: android-porting+unsubscr...@googlegroups.comandroid-porting%2bunsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -- unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting
[android-porting] Re: How to build Android emulator for special platform?
goldfish has nothing to do with HTC. It's the name of the Linux kernel architecture/machine configuration used to build the ARM-based kernel image specifically used by the emulator. The goldfish hardware is very unlike the one of a real device. the goldfish specific bits live in a branch on kernel-common.git, look for android-goldfish-2.6.29 irrc you should be able to rebase the corresponding changes to another kernel branch if you want to. On Tue, Oct 27, 2009 at 7:42 PM, qitao song qitaos...@gmail.com wrote: Hi,ALL: As I known, the Android SDK's emulator just be used for HTC Goldfish. But I want to build a Android emulator for special platform, for example TI' OMAP platform. I want to use it for emulating the kernel (zImag) and android file system. I'm looking forward to your help, Thank you very much. --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Closing TCP socket problem
try using SO_REUSEADDR. Also using shutdown() before closing the socket might help. On Wed, Oct 14, 2009 at 2:52 AM, Yogi yogesh.agrawal2...@gmail.com wrote: Hi All, I have a server-client pogram on android where server will be listening on 3000 port and client will send request on the same port. I am able to establish the connect successfully but when we are trying to close the socket, its going in CLOSE_WAIT state and sometime server is going in FIN_WAIT2 state.We are checking the socket state with netstat command in android. If we try to establish the connection again,server bind() fails with error Address already in use. But if i wait for some 2-3minutes and then check the socket status its get closed successfully.Now, again able to establish the connect successfully. Can you please help me to understand, what is going wrong? Is there any way to force close the sockets? Thanks in advance, Yogi --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Run Android directly on Qemu
On Tue, Oct 13, 2009 at 10:32 AM, Xuan Zhang allsy1...@gmail.com wrote: Hi, Recently, I plan to do some experiment about virtual machine running mobile os. So I intend to run Android on Qemu rather than the android emulator. I have seen android has four images: kernel-qemu,system.img, userdata.img and ramdisk.img. However, I still can't figure out how to pass them to qemu. use the -verbose flag, it will print the internal command-line being passed to QEMU. BTW, it seems android emulator defines goldfish. Is it possible to port it to qemu? Actually, I wonder whether one of the following can work: 1. Get the hw part of goldfish, and compile it with original qemu. Then run qemu with proper argument passing the images. 2. Modify the android emulator such that its interfaces works exactly as a specalized qemu rather than a emulator. I am really new to android, thanks very much. In theory, it should be possible to do that by: 1/ integrating the hw/goldfish_*.[hc] sources into the upstream qemu sources, they correspond to goldfish-specific hardware emulation that simply must be implemented in QEMU. Goldfish is the name of the virtual ARM machine/platform for the Linux kernel variant used to generate the kernel-qemu image. 2/ finding the right mix of command-line options when invoking qemu to duplicate what the normal Android emulator does (see output of -verbose option for examples) Note that the result will probably have no telephony emulation, not proper network connectivity plus a bunch of missing features (no skins, no proper keyboard emulation, etc...). However, chances are that it will run. Good luck Best, Xuan --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Memory overcommit
On Mon, Sep 14, 2009 at 3:53 PM, David Given d...@cowlark.com wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Dianne Hackborn wrote: [...] If you turn off over-commit, I believe Android won't even boot on a G1, because Linux would need to assume that RAM is needed for every mmapped() thing, and we run out of it well before the system is fully up. Not necessarily --- my intuitive understanding would be that overcommit would only applies to anonymous storage and not to file storage, as mmapped() files are backed by actual *files* and not by the pool of swap pages. (Besides, most of those mmapped() files that Android's using are read-only and their pages can be silently discarded at any point.) That can only be said for files that are mapped read-only. It doesn't apply for anything that is shared copy-on-write (e.g. the Zygote process pages, as well as initial heap) and non-file mappings. This would increase the memory pressure tremendously, given that every VM process has several megabytes of shared copy-on-write pages. (What we've actually got is a portable API implementation for native apps. The memory allocation function is defined to return NULL on allocation failure. It had never occurred to us that *making* it return NULL on allocation failure could be hard.) But it *is* hard to define what allocation failure really means if you have memory overcommit enabled. This behaviour also happens on all modern operating systems these days. If you have a swap file, you will simply make it happen later, after much much disk chugging that will make the user totally nuts anyway. - -- ┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ─ http://www.cowlark.com ─ │ │ People who think they know everything really annoy those of us who │ know we don't. --- Bjarne Stroustrup -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iD8DBQFKrslpf9E0noFvlzgRAuuOAKCAEARJSEPcHxcrEowsxwcMBvkxGwCcC1Jz kuny8oLQqn1S8mBDmqp9Bwk= =iCwo -END PGP SIGNATURE- --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Memory overcommit
On Sun, Sep 13, 2009 at 4:34 AM, David Given d...@cowlark.com wrote: Hello, I was rather startled recently to notice that the standard Android kernel appears to have the memory overcommit setting set to 1. This is --- as far as I can tell, the numbering got changed not long ago and not all the documentation has been updated --- these means 'allow all memory allocations even if no RAM+swap is available'. The knock on effect of this is that running out of memory will cause either a memory trap in the application or else the OOM killer will nuke your entire process without warning. Can anyone comment on this? I know, for example, that the Android patch has modified the OOM killer. I would have thought that memory overcommit should be disabled on this kind of embedded device? memory overcommit is used by design, there is no plan to remove it. The fact that the OOM killer will nuke processes to make room for others is part of the platform's design. As far as I know, the OOM changes were to make the killer a bit smarter about what kind of processes it would try to kill first. -- ┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ─ http://www.cowlark.com ─ │ │ They laughed at Newton. They laughed at Einstein. Of course, they │ also laughed at Bozo the Clown. --- Carl Sagan --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Memory overcommit
On Sun, Sep 13, 2009 at 4:34 AM, David Given d...@cowlark.com wrote: Hello, I was rather startled recently to notice that the standard Android kernel appears to have the memory overcommit setting set to 1. This is --- as far as I can tell, the numbering got changed not long ago and not all the documentation has been updated --- these means 'allow all memory allocations even if no RAM+swap is available'. The knock on effect of this is that running out of memory will cause either a memory trap in the application or else the OOM killer will nuke your entire process without warning. Can anyone comment on this? I know, for example, that the Android patch has modified the OOM killer. I would have thought that memory overcommit should be disabled on this kind of embedded device? To make things cleared, memory overcommit is a good thing in environments where there are tons of copy-on-write shared pages between processes (which is typical of Android), with very few copies in practice. If we were to disable it, the system would barely run with the current amount of memory. -- ┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ─ http://www.cowlark.com ─ │ │ They laughed at Newton. They laughed at Einstein. Of course, they │ also laughed at Bozo the Clown. --- Carl Sagan --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: porting OpenMP : problem with using TLS keys from constructor routines
I believe this bug has been fixed in the donut dynamic linker / C library.
Yes, the first commit was not ok, was reverted, and another fix was
submitted.
Can you reproduce the problem with the current sources?
On Mon, Sep 7, 2009 at 3:10 AM, Surinder srpl...@gmail.com wrote:
Hi,
I would like to highlight an issue related to calling functions that
create application specific TLS keys from within functions marked as
__attribute__ ((constructor)).
For e.g. consider the following routine:
static void __attribute__ ((constructor)) foo(void)
{
...
pthread_key_create(..);
...
}
bionic/linker/linker.c:__linker_init() calls foo() and the
pthread_key_create() would use the temporary TLS bionic/linker/
linker.c:__tls_area which is OK, until the main thread's TLS is
initialized in bionic/libc/bionic/libc_init_common.c and thus the key
created in the old TLS is no longer accessible resulting in the app
segfaulting.
This situation actually occurs when trying to port the gcc OpenMP
runtime. Although for the moment, I've worked around the problem among
others and am able to successfully execute OpenMP apps under Android.
In fact, I've seen a commit
http://android.git.kernel.org/?p=platform/bionic.git;a=commit;h=03eabfe65e1e2c36f4d26c78a730fa19a3bdada3
which, by re-using the temporary TLS for the main thread would
actually fix the problem I've described but I noticed that this commit
has been reverted and a different solution was committed later.
Cheers,
surinder
--~--~-~--~~~---~--~~
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
-~--~~~~--~~--~--~---
[android-porting] Re: Accessing my host devices on Android- Emulator
no, it's not possible directly. Even if the emulator supported this, the Android kernel that runs under emulation doesn't support USB Host mode anyway. 2009/8/27 தங்கமணி அருண் thangam.ar...@gmail.com Hi All, I want to know, is it possible to access my USB pen drive on Android Emulator ?? I looking for any kind of clues. Thanks in advance. -- அன்புடன் அருண் -- http://ubuntu-tam.org http://lists.ubuntu.com/ubuntu-l10n-tam http://lists.ubuntu.com/ubuntu-tam -- --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Emulator does not show up when using self-build kernel
Please follows the instructions here to rebuild an emulator-specific kernel: http://android.git.kernel.org/?p=platform/external/qemu.git;a=blob;f=docs/KERNEL.TXT;h=7387e55d5362be8e2dc00e634277a9fdbdd66db6;hb=ceb00f8c44e26d948e28b7a3ccef8fd2068589d9 Anything else is likely to fail. 2009/8/26 马不停蹄的猪 sunrui...@gmail.com Hi all, I download the android kernel and build by myself. But when I use emulator to load my kernel, the emulator does not show up. Below are the steps what I do. 1, download android source code of Cupcake. 2, make (build the source code) 3, make sdk (build the sdk) 4, run emulator(generated in cupcake/out/host/linux-x86/sdk/android- sdk_eng.srj_linux-x86/tools) and pull out config.gz by adb. 5, git clone git://android.git.kernel.org/kernel/common.git git checkout -b goldfish Now the kernel is in common directory 6, gunzip config.gz and copy to common directory and rename it to .config 7, set environment variable to the correct value: export ARCH=arm export CROSS_COMPILE=arm-eabi- 7, run make Then there are some options need to choose manually and some are choosed automatically. I list the manually choosed options as below Restart config... Support Versatile/PB platform (ARCH_VERSATILE_PB) [Y/n/?] (NEW) Y Support Versatile/AB platform (MACH_VERSATILE_AB) [N/y/?] (NEW) N PCI support (PCI) [N/y/?] (NEW) N Low address space to protect from user allocation (DEFAULT_MMAP_MIN_ADDR) [4096] (NEW) (Just 'Enter' for this option) Timer and CPU usage LEDs (LEDS) [N/y/?] (NEW) N AMBA KMI keyboard controller (SERIO_AMBAKMI) [N/y] (NEW) N ARM AMBA PL011 serial port support (SERIAL_AMBA_PL011) [N/y/?] (NEW) N ARM PrimeCell PL110 support (FB_ARMCLCD) [N/y/?] (NEW) N ARM AMBA Multimedia Card Interface support (MMC_ARMMMCI) [N/y/?] (NEW) N ARM AMBA PL030 RTC (RTC_DRV_PL030) [N/y/?] (NEW) N ARM AMBA PL031 RTC (RTC_DRV_PL031) [N/y/?] (NEW) n Finally, kernel is built successfully. And zImage is under common/arch/ arm/boot. I use this image to load emulator like below: emulator -avd cupcake -kernel /home/cupcake/common/arch/arm/boot/ zImage -sysdir /home/cupcake/out/target/product/generic The emulator does not show up. Does any one know the rootcause? Thanks in advance. --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Failed to run cupcake on emulator
The cupcake is not very indicative of what you're doing. You should at least try to explain exactly which branch/repository you're usingand how you configured your build. Also, using both a custom kernel and custom system image is fraught with peril. Couldn't you try one without the other first to try to isolate the issues? On Thu, Aug 20, 2009 at 5:53 AM, Miller Fu miller.fu...@gmail.com wrote: The cupcake can't boot up on the emulator. Can anybody help me check what is the problem? please see the Logcat I built cupcake images and kernel by myself. mille...@millerfu-desktop:~/cupcake$emulator -kernel ~/kernel/arch/arm/ boot/zImage -system out/target/product/generic/system.img -ramdisk out/ target/product/generic/ramdisk.img -initdata out/target/product/ generic/userdata.img -avd AndroidSDK 08-19 03:39:07.593: INFO/DEBUG(539): debuggerd: Aug 13 2009 14:28:58 08-19 03:39:07.613: INFO/vold(538): Android Volume Daemon version 2.0 08-19 03:39:07.724: ERROR/vold(538): Error opening switch name path '/ sys/class/switch/test2' (No such file or directory) 08-19 03:39:07.724: ERROR/vold(538): Error bootstrapping switch '/sys/ class/switch/test2' (No such file or directory) 08-19 03:39:07.724: ERROR/vold(538): Error opening switch name path '/ sys/class/switch/test' (No such file or directory) 08-19 03:39:07.724: ERROR/vold(538): Error bootstrapping switch '/sys/ class/switch/test' (No such file or directory) 08-19 03:39:07.724: DEBUG/vold(538): Bootstrapping complete 08-19 03:39:08.092: DEBUG/qemud(547): entering main loop 08-19 03:39:08.702: DEBUG/qemud(547): fdhandler_accept_event: accepting on fd 10 08-19 03:39:08.712: DEBUG/qemud(547): created client 0xe078 listening on fd 8 08-19 03:39:08.732: DEBUG/qemud(547): client_fd_receive: attempting registration for service 'boot-properties' 08-19 03:39:08.732: DEBUG/qemud(547): client_fd_receive:- received channel id 1 08-19 03:39:08.746: DEBUG/qemud(547): client_registration: registration succeeded for client 1 08-19 03:39:09.283: DEBUG/qemud(547): fdhandler_accept_event: accepting on fd 10 08-19 03:39:09.283: DEBUG/qemud(547): created client 0x10028 listening on fd 11 08-19 03:39:09.283: DEBUG/qemud(547): fdhandler_event: disconnect on fd 11 08-19 03:39:09.892: INFO/qemu-props(557): connected to 'boot- properties' qemud service. 08-19 03:39:09.912: INFO/qemu-props(557): exiting (0 properties set). 08-19 03:39:10.883: DEBUG/qemud(547): fdhandler_accept_event: accepting on fd 10 08-19 03:39:10.883: DEBUG/qemud(547): created client 0xe078 listening on fd 8 08-19 03:39:10.883: DEBUG/qemud(547): client_fd_receive: attempting registration for service 'gsm' 08-19 03:39:10.883: DEBUG/qemud(547): client_fd_receive:- received channel id 2 08-19 03:39:10.892: DEBUG/qemud(547): client_registration: registration succeeded for client 2 08-19 03:39:11.032: DEBUG/AndroidRuntime(541): AndroidRuntime START 08-19 03:39:11.032: DEBUG/AndroidRuntime(541): CheckJNI is ON 08-19 03:39:11.243: DEBUG/dalvikvm(541): DexOpt: incorrect opt magic number (0xff ff ff ff) 08-19 03:39:11.243: DEBUG/dalvikvm(541): Stale deps in cache file; removing and retrying 08-19 03:39:11.583: INFO/(542): ServiceManager: 0xac38 08-19 03:39:11.633: INFO/AudioFlinger(542): AudioFlinger's thread ready to run for output 0 08-19 03:39:11.663: DEBUG/dalvikvm(541): DexOpt: --- BEGIN 'core.jar' (bootstrap=1) --- 08-19 03:39:11.703: INFO/CameraService(542): CameraService started: pid=542 08-19 03:39:12.213: ERROR/dalvikvm(575): Too many exceptions during init (failed on 'Ljava/lang/NoClassDefFoundError;' 'java.lang.NoClassDefFoundError') 08-19 03:39:12.222: ERROR/dalvikvm(575): VM aborting 08-19 03:39:12.354: INFO/DEBUG(539): *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** 08-19 03:39:12.354: INFO/DEBUG(539): Build fingerprint: 'generic/ generic/generic/:Donut/Donut/eng.millerfu.20090813.134359:eng/test- keys' 08-19 03:39:12.363: INFO/DEBUG(539): pid: 575, tid: 575 /system/ bin/dexopt 08-19 03:39:12.363: INFO/DEBUG(539): signal 11 (SIGSEGV), fault addr deadd00d 08-19 03:39:12.373: INFO/DEBUG(539): r0 0330 r1 000c r2 000c r3 0026 08-19 03:39:12.385: INFO/DEBUG(539): r4 deadd00d r5 00020228 r6 ad06eba4 r7 08-19 03:39:12.385: INFO/DEBUG(539): r8 r9 10 fp 08-19 03:39:12.385: INFO/DEBUG(539): ip ad083ef8 sp be947aa0 lr afe13f0d pc ad03b88a cpsr 2030 08-19 03:39:12.413: INFO/DEBUG(539): #00 pc 0003b88a / system/lib/libdvm.so 08-19 03:39:12.413: INFO/DEBUG(539): #01 pc 0003af48 / system/lib/libdvm.so 08-19 03:39:12.423: INFO/DEBUG(539): #02 pc 0003b1c4 / system/lib/libdvm.so 08-19 03:39:12.423: INFO/DEBUG(539): #03 pc 0005ced6 / system/lib/libdvm.so 08-19 03:39:12.423: INFO/DEBUG(539): #04 pc 0005d110 / system/lib/libdvm.so 08-19 03:39:12.433: INFO/DEBUG(539):
[android-porting] Re: [mips] Fail at build bionic
Unfortunately, the people at mipsandroid.org have not sent any patches to
the upstream repository, so we can't really comment on this.Try on their
support forums instead.
On Wed, Aug 19, 2009 at 3:22 AM, Dennis.Yxun dennis.y...@gmail.com wrote:
HI folks:
I'm trying to build mips-cupcake, all the source code is from
www.mipsandroid.org
previous build instructions described here [1]
Slightly different is that I use my own toolchains which generated by
gentoo's crossdev script.
Here is error messages[2], can't find __floatsidf, refer attached file
for more information.
It seems that linker can't find some symbols, while following shows that
libgcc_s.so do have those functions.
How can I narrow down the problem, to make the compiling process more
verbose?
$mipsel-softfloat-linux-gnu-objdump -t
/usr/lib/gcc/mipsel-softfloat-linux-gnu/4.3.4/libgcc_s.so.1 |grep floatsidf
a820 g F .text0108 __floatsidf
$mipsel-softfloat-linux-gnu-objdump -t
/usr/lib/gcc/mipsel-softfloat-linux-gnu/4.3.4/libgcc.a |grep floatsidf
O *UND* __floatsidf
O *UND* __floatsidf
g F .text0108 .hidden __floatsidf
$ mipsel-softfloat-linux-gnu-gcc -v
Using built-in specs.
Target: mipsel-softfloat-linux-gnu
Configured with:
/var/build/cross/mipsel-softfloat-linux-gnu/portage/cross-mipsel-softfloat-linux-gnu/gcc-4.4.0/work/gcc-4.4.0/configure
--prefix=/usr
--bindir=/usr/i686-pc-linux-gnu/mipsel-softfloat-linux-gnu/gcc-bin/4.4.0
--includedir=/usr/lib/gcc/mipsel-softfloat-linux-gnu/4.4.0/include
--datadir=/usr/share/gcc-data/mipsel-softfloat-linux-gnu/4.4.0
--mandir=/usr/share/gcc-data/mipsel-softfloat-linux-gnu/4.4.0/man
--infodir=/usr/share/gcc-data/mipsel-softfloat-linux-gnu/4.4.0/info
--with-gxx-include-dir=/usr/lib/gcc/mipsel-softfloat-linux-gnu/4.4.0/include/g++-v4
--host=i686-pc-linux-gnu --target=mipsel-softfloat-linux-gnu
--build=i686-pc-linux-gnu --disable-altivec --disable-fixed-point
--without-ppl --without-cloog --with-float=soft --enable-nls
--without-included-gettext --with-system-zlib --disable-checking
--disable-werror --enable-secureplt --disable-multilib --disable-libmudflap
--disable-libssp --disable-libgomp --enable-cld --disable-libgcj
--enable-languages=c,c++,fortran
--with-sysroot=/usr/mipsel-softfloat-linux-gnu --disable-bootstrap
--disable-libgomp --enable-__cxa_atexit --enable-clocale=gnu --with-bugurl=
http://bugs.gentoo.org/ --with-pkgversion='Gentoo 4.4.0 p1.1'
Thread model: posix
gcc version 4.4.0 (Gentoo 4.4.0 p1.1)
$mipsel-softfloat-linux-gnu-ld --version
GNU ld (Linux/GNU Binutils) 2.19.51.0.9.20090606
Copyright 2008 Free Software Foundation, Inc.
This program is free software; you may redistribute it under the terms of
the GNU General Public License version 3 or (at your option) a later
version.
This program has absolutely no warranty.
[1]
http://groups.google.com/group/android-porting/browse_thread/thread/3371c6b25a8f3a07?pli=1
[2] target SharedLib: libc
(out/target/product/hmp10/obj/SHARED_LIBRARIES/libc_intermediates/LINKED/libc.so)
out/target/product/hmp10/obj/STATIC_LIBRARIES/libc_common_intermediates/libc_common.a(time.o):
In function `clock':
/mipsdroid/bionic/libc/unistd/time.c:55: undefined reference to
`__floatsidf'
/mipsdroid/bionic/libc/unistd/time.c:55: undefined reference to
`__floatsidf'
--~--~-~--~~~---~--~~
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
-~--~~~~--~~--~--~---
[android-porting] Re: ping DNS issue! ping returns unknown host www.yahoo.com but ping to IP works fine
The Android C library contains a DNS resolver that doesn't use /etc/resolv.conf or /etc/nsswitch.conf Instead, to it uses a list of system properties instead: net.dns1 - IP address of first DNS server net.dns2 - IP address of second DNS server ... Additionally, the Donut platform (not Cupcake) also supports net.dns.search - space separated domain search suffixes (used e.g. by VPN) On Wed, Aug 12, 2009 at 11:58 AM, jolly1312 jolly1...@gmail.com wrote: I am trying to get the ping working in Android running on ARM. Ping to IP addresses works fine but when I ping to name I get ping: unknown host www.yahoo.com I have already setup the default gateway: route add default gw 192.168.0.17 Also /etc/resolv.conf is also there: echo nameserver 4.2.2.2 /etc/resolv.conf And ping to nameserver works fine! Also I have /etc/nsswitch.conf as: echo hosts:files dns /etc/nsswitch.conf If I add entries in /etc/hosts then the ping with name works fine. Has anyone seen this problem before? --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: tessereact compile in android ndk
You probably didn't setup your include paths properly. Did you read the section about that in $NDK/docs/HOWTO.TXT ? On Tue, Aug 11, 2009 at 2:03 AM, gatza gat...@gmail.com wrote: Hello! me and my friend trying to port tesseract to android. Tesseract is an open source optical character recognition software first developed by HP now by Google. http://code.google.com/p/tesseract-ocr This tesseract project is an external android project http://android.git.kernel.org/?p=platform/external/tesseract.git;a=summarybut We try to compile it with the android NDK. We try to compile all the modules as a static library, expect libtesseract which we try as shared library. But we got this compile error message included at the end of the letter. Could anybody help what to do? Is it possible to compile tesseract with the android NDK? Or is it impossible, or does it need some code to change? thanks Tom In file included from sources/tesseract/dict/ patternfst.h:10, from sources/tesseract/dict/fstmodel.h:8, from sources/tesseract/dict/permute.cpp:36: sources/tesseract/dict/wordfst.h:10:36: error: nlp/fst/lib/fstlib- inl.h: No such file or directory sources/tesseract/dict/wordfst.h:12:43: error: third_party/libidn/ stringprep.h: No such file or directory sources/tesseract/dict/wordfst.h:13:36: error: util/utf8/ unicodeprops.h: No such file or directory sources/tesseract/dict/wordfst.h:14:51: error: third_party/icu/current/ unicode/uchar.h: No such file or directory sources/tesseract/dict/wordfst.h:15:39: error: i18n/utf8/ char_properties.h: No such file or directory In file included from sources/tesseract/dict/permute.cpp:39: sources/tesseract/cutil/const.h:52:1: warning: MAXFLOAT redefined In file included from sources/tesseract/dict/permute.cpp:33: build/platforms/android-1.5/arch-arm/usr/include/math.h:164:1: warning: this is the location of the previous definition In file included from sources/tesseract/dict/patternfst.h:10, from sources/tesseract/dict/fstmodel.h:8, from sources/tesseract/dict/permute.cpp:36: sources/tesseract/dict/wordfst.h:18: error: expected initializer before '' token sources/tesseract/dict/wordfst.h:22: error: expected ',' or '...' before '' token sources/tesseract/dict/wordfst.h:22: error: ISO C++ forbids declaration of 'String' with no type sources/tesseract/dict/wordfst.h: In member function 'bool ltstr::operator()(int) const': sources/tesseract/dict/wordfst.h:23: error: 's1' was not declared in this scope sources/tesseract/dict/wordfst.h:23: error: 's2' was not declared in this scope sources/tesseract/dict/wordfst.h: At global scope: sources/tesseract/dict/wordfst.h:44: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:45: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:46: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:47: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:48: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:49: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:50: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:51: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:52: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:59: error: expected ',' or '...' before '' token sources/tesseract/dict/wordfst.h:59: error: ISO C++ forbids declaration of 'String' with no type sources/tesseract/dict/wordfst.h:64: error: expected ',' or '...' before '' token sources/tesseract/dict/wordfst.h:64: error: ISO C++ forbids declaration of 'String' with no type sources/tesseract/dict/wordfst.h:66: error: ISO C++ forbids declaration of 'StdVectorFst' with no type sources/tesseract/dict/wordfst.h:66: error: expected ';' before '*' token sources/tesseract/dict/wordfst.h:67: error: 'StdVectorFst' has not been declared sources/tesseract/dict/wordfst.h:69: error: 'list' has not been declared sources/tesseract/dict/wordfst.h:69: error: expected ',' or '...' before '' token sources/tesseract/dict/wordfst.h:75: error: 'nlp_fst' has not been declared sources/tesseract/dict/wordfst.h:75: error: ISO C++ forbids declaration of 'StdVectorFst' with no type sources/tesseract/dict/wordfst.h:75: error: expected ';' before 'fst_' sources/tesseract/dict/wordfst.h:79: error: 'list' has not been declared sources/tesseract/dict/wordfst.h:79: error: expected ',' or '...' before '' token In file included from sources/tesseract/dict/fstmodel.h:8, from sources/tesseract/dict/permute.cpp:36: sources/tesseract/dict/patternfst.h:19: error: expected ',' or '...' before '' token sources/tesseract/dict/patternfst.h:19: error: ISO C++ forbids declaration of 'string' with no type sources/tesseract/dict/patternfst.h:24: error: expected ',' or '...' before
[android-porting] Re: where is the arm implemention of tls
pthread.c.arm in Android.mk means compile pthread.c in ARM mode so you're really looking for pthread.c the TLS stuff is implemented in arch-arm/bionic/__get_tls.S On Tue, Aug 4, 2009 at 5:17 AM, Ouyang Jiannan jallen@gmail.com wrote: hi, everyone, In bionic/libc/Android.mk we can see that x86 uses files pthread.c, pthread-timers.c ptrace.c and __get_tls.c in arm, it uses pthread.c.arm , pthread-timers.c.arm, ptrace.c.arm But I can't find these files or places which generate it So what's the arm support of TLS? Is there a reg uses as gs in x86 as a pointer to tls? --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: where is the arm implemention of tls
On Tue, Aug 4, 2009 at 12:24 PM, Neo zingafri...@yahoo.com wrote: On Aug 4, 8:17 am, Ouyang Jiannan jallen@gmail.com wrote: hi, everyone, In bionic/libc/Android.mk we can see that x86 uses files pthread.c, pthread-timers.c ptrace.c and __get_tls.c in arm, it uses pthread.c.arm , pthread-timers.c.arm, ptrace.c.arm But I can't find these files or places which generate it So what's the arm support of TLS? Is there a reg uses as gs in x86 as a pointer to tls? TLS is not supported in android though it is present in armv6 and later. TLS is supported in Android (there wouldn't be any pthread implementation anyway). The current ARM implementation doesn't use an ARMv6-specific register but instead relies on a small kernel trick (which updates a very specific user virtual address on each context switch instead). --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: where is the arm implemention of tls
On Wed, Aug 5, 2009 at 3:47 AM, Ouyang Jiannan jallen@gmail.com wrote: Thanks I've noticed that bionic uses address 0x0ff0 to implement TLS in arm But isn't this address a space of kernel? How can a user space program access that address without a soft-interupt... Or the memory map in arm is different from x86, so the high 1G mem is not kernel used only? I don't know the details, but I know it relies on a weird kernel tricks. Be it a special page fault, or a special page mapping. Actually, another similar address contains the address of a function to be called by user code to get the same TLS pointer. This is supposedly more portable (it is used by the ARMv4T port, if I remember correctly), but can also be slower. The ARMv6 way of using a dedicated register is supposed to get much better performance, but of course can't be used by NDK-generated binaries which are built for ARMv5TE. On Wed, Aug 5, 2009 at 3:06 AM, David Turner di...@android.com wrote: On Tue, Aug 4, 2009 at 12:24 PM, Neo zingafri...@yahoo.com wrote: On Aug 4, 8:17 am, Ouyang Jiannan jallen@gmail.com wrote: hi, everyone, In bionic/libc/Android.mk we can see that x86 uses files pthread.c, pthread-timers.c ptrace.c and __get_tls.c in arm, it uses pthread.c.arm , pthread-timers.c.arm, ptrace.c.arm But I can't find these files or places which generate it So what's the arm support of TLS? Is there a reg uses as gs in x86 as a pointer to tls? TLS is not supported in android though it is present in armv6 and later. TLS is supported in Android (there wouldn't be any pthread implementation anyway). The current ARM implementation doesn't use an ARMv6-specific register but instead relies on a small kernel trick (which updates a very specific user virtual address on each context switch instead). --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Android on solaris
At the very minimum, you will need to write Solaris kernel drivers for Android-specific features like Binder or ashmem. No small feat indeed. After that, you will need to change *any* code that touches the hardware and kernel (e.g. anything in /dev or /proc on Linux), probably make *huge* changes to the C library. I'm not even sure that Solaris has all the features requested for this kind of work. Or are you talking about the ability to build Android on Solaris instead? On Mon, Aug 3, 2009 at 9:28 AM, Rakesh Kumar rakesh.y...@gmail.com wrote: Hi friends, I want to port android on Solaris operating system. Please help me, What are the changes I need to make in android?. And what should be my approach? Regards Rakesh --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: pthread lib problem
They are not available at the moment. You should be able to implement them through the use of several mutexes and condition variables though. There are several implementations available on the web (search for implement read write lock for example). Feel free to submit a patch to the platform once you have something that works. Regards On Thu, Jul 30, 2009 at 1:50 PM, Rakesh Kumar rakesh.y...@gmail.com wrote: Hi, Is there any alternative of pthread_rwlock_*** in android. Does android have any api which does the same work? Regards Rakesh On Tue, Jul 28, 2009 at 1:54 PM, GeunSik Lim (invain) leem...@gmail.comwrote: Hi Rakesh, Android platform support posix thread api limitally. ( http://blogfiles12.naver.net/20090728_91/invain_1248769404396_kpQWX0_PNG/android-pthread-api-list.PNG ) In summary, Android's posix compatible api don't support below list that POSIX 1003.1j-2000 specified. 1. pthread_rwlock_*** 2. pthread_rwlockattr_*** 3. pthread_barrior_*** 4. pthread_barrior_attr_*** 5. pthread_spin_*** Hope it helps. Thks, GeunSik Lim. On 7월28일, 오후3시08분, Rakesh Kumar rakesh.y...@gmail.com wrote: Hi, I want t use pthread raead/write lock functions like pthread_rwlock_rdlock , pthread_rwlock_destroy etc in my code. But android pthread lib does not support them, in my view. Please tell me is there any alternative to this or not? Thanks Rakesh --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: [bionic c header] dirent64 is missing in bionic/libc/kernel/common/linux/dirent.h
that's because the struct dirent defined in Bionic is the kernel's stuct dirent64. (look at the comment in sys/dirent.h that comes with Bionic). This is done intentionally. dirent64 is not portable, if your code depends on it, use something like: #define dirent64 dirent and it should work. If your code depends on struct dirent and struct dirent64 being different, then you have a problem, but I would qualify that as a self-inflicted one :-) On Mon, Jul 27, 2009 at 7:33 AM, louis dmq...@gmail.com wrote: Hi It's my first posting. I am porting a ntive c application to android phone to share with windows machine. My compile error messages show that the source code needs struct dirent64 but it is missing in bionic/libc/include/dirent.h which is the only one included when compiled but in bionic/libc/kernel/common/ linux/dirent.h which is not included but contains struct dirent64. I have double checked for both case of compiling against glibc and bionic and noticed that there is no error in case of glibc. Would someone kindly give me a reasonable direction to get rid of compile error. Should I add '#include' in the bionic/libc/kernel/common/linux/ dirent.h to include bionic/libc/kernel/common/linux/dirent.h? --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Is arm-eabi-gcc support iwmmxt?
the -mabi option might generate code that is not compatible with the rest of the system, I would definitely not recommend using this. also info gcc might refer to a different version than the arm-eabi-gcc binaries themselves so this gives no guarantee about the result. On Fri, Jul 24, 2009 at 10:16 PM, fadden fad...@android.com wrote: On Jul 23, 7:02 pm, xuxiake2012 xuxiake2...@gmail.com wrote: I have some source code in hand which can use hardware acceleration just like MMX,IWMMXT,SSE and so on. As we known, android use arm-eabi- gcc to compile source code.so my question come : Is arm-eabi-gcc support iwmmxt? Or what should I do to make is support iwmmxt? Thanks in advance. % info gcc [Ctrl-S]iwmmxt 3.17.2 ARM Options -- These `-m' options are defined for Advanced RISC Machines (ARM) architectures: `-mabi=NAME' Generate code for the specified ABI. Permissible values are: `apcs-gnu', `atpcs', `aapcs', `aapcs-linux' and `iwmmxt'. [...] --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: undefined reference to __stack_chk_fail error
These __stack_chk_ references are created by the compiler when you the -fstack-protector option is enabled. It is likely enabled by the build system, but sometimes even prebuilt compiler binaries have it enabled by default. The corresponding functions are normally in libgcc.a, provided with your toolchain. If this is not the case, you will need to add something like -fno-stack-protector when compiling sources. On Fri, Jul 24, 2009 at 7:48 AM, Ouyang Jiannan jallen@gmail.comwrote: I'm porting bionic to our cpu I've edited makefiles to enable the compiling process for our arch make TARGET_ARCH=myarch BUILD_TINY_ANDROID=true HOST_C_INCLUDES= HOST_CUSTOM_LD_COMMAND=false If the last two options are not added, bionic will be included into host_c_includes, and host_custom_ld_command will be true I don't know why, and I haven't found anyplace which assign value to these variables... Now, during the compiling of acp and aprior, things went wrong. The compiler complain about undefind reference to __stack_chk_fail and __stack_chk_guide I've compared the g++ options between arm and our arch, only libelf_arm.a was not generated in my arch I've also check the symbols defined in libebl.a libelf.a libelfcopy.a, I found that in my *.a symbol __stack_chk_fail is undefined, but *.a of arm symbol __stack_chk_fail_local is undefined, which is defined in the toolchain in prebuilt I've been trapped here for a long time, and i still can't figure out a solution P.S. I'm using the toolchain of our own arch. And my target is to port bionic to our arch... --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: [Help] Failed to launch the android under the emul
How did you configure your build exactly, which branch are you running ? Also did you use lunch of choosecombo to setup the build. If so, you simply need to type emulator to start the emulator. On Thu, Jul 16, 2009 at 2:21 PM, Tiger Yu tigerflying...@gmail.com wrote: Hi all, I have built android project successfully with default configuration. Then I tried to start the android project with qemu(emulator) using the following command below: *../../../host/linux-x86/bin/emulator -kernel /work/maxwit/work/android/prebuilt/android-arm/kernel/kernel-qemu -system system.img -data userdata.img -ramdisk ramdisk.img -sysdir ./system -shell-serial stdio -show-kernel* It then opened a window which only displayed Android word. And the console was reported much error information. I list them below here: *request_suspend_state: wakeup (0-0) at 307918679246 (2009-07-17 00:10:05.342175542 UTC) request_suspend_state: wakeup (0-0) at 313753667910 (2009-07-17 00:10:11.177164346 UTC) request_suspend_state: wakeup (0-0) at 318601462812 (2009-07-17 00:10:16.024956245 UTC) ...* I think the system was started, for i can use all commands which android supported. But It always reported error, and never go into the android desktop system. Can someone help me? Thanks a lot. tiger --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: [android-developers] error: cloning the kernel from git
Use git://android.git.kernel.org instead of git://git.android.kernel.org On Fri, Jul 17, 2009 at 9:06 PM, krishna devarapalli krishna.andr...@gmail.com wrote: git clone git://git.android.kernel.org/kernel/common.git kernel-common Initialized empty Git repository in /home/sreevani/mydroidkernel/kernel-common/.git/ fatal: Unable to look up git.android.kernel.org (port 9418) (Name or service not known) fetch-pack from 'git://git.android.kernel.org/kernel/common.git' failed. Please suggest. Thanks, Krishna --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: where is bionic/libc/kernel/original??
the headers are not available yet. We're working on putting them in the open-source repository. (for the record, it's a mix of old and new headers, for various historical reasons). On Fri, Jun 19, 2009 at 4:54 PM, Jimmy-Creupagu jamai5...@gmail.com wrote: Hi Android guys, i'm trying to configure my source code to add a new arch. I'm following the basic steps from http://groups.google.com/group/android-porting/browse_thread/thread/6633bf1e3397efaf/a96c9a4bd570d097 but I haven't got any directory with the name original where it should be... (bionic/libc/kernel/original) Should i create it? or is something missing in my source code? thanks for any help --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Add gprof support to bionic
You can submit the patch to r.android.com, the Gerrit tool is perfect to make this kind of reviews and host a discussion about it too. On Thu, Jun 25, 2009 at 4:30 PM, Leix xie.xiao...@gmail.com wrote: The patch spans over 2000 lines. Does this group have something for sharing patches? It appears I can't attach anything to a message. On Jun 25, 2:30 am, David Turner di...@android.com wrote: no, but do you have patches to share so we can look at them? On Wed, Jun 24, 2009 at 3:36 PM, Leix xie.xiao...@gmail.com wrote: Hello group, I ported the gprof files found in glic-2.9 to android and made the system build, but when I flashed it onto a phone, which had an ARM CPU and ran on Linux 2.6.28, the libraries built with the -pg flag crashed due to a SIGBUS signal. Do you guys have any idea about why this happened? Thanks, Leix --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Add gprof support to bionic
no, but do you have patches to share so we can look at them? On Wed, Jun 24, 2009 at 3:36 PM, Leix xie.xiao...@gmail.com wrote: Hello group, I ported the gprof files found in glic-2.9 to android and made the system build, but when I flashed it onto a phone, which had an ARM CPU and ran on Linux 2.6.28, the libraries built with the -pg flag crashed due to a SIGBUS signal. Do you guys have any idea about why this happened? Thanks, Leix --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Something goes wrong on vold randomly or due to wifi
do you have details on how to reproduce this. What hardware platform are you using, which system image are you building, etc... On Mon, Jun 15, 2009 at 12:30 PM, Luca Belluccini lucabellucc...@gmail.comwrote: Vold crashes and/or starts notifying phone storage space low randomly or seems that wifi is implied on that. Anyone experienced something similar? --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: pre compiler macro for andriod
it's declared by the build system. You can also try HAVE_ANDROID_OS which is declared in AndroidConfig.h which is automatically included by the build system too before any other file. On Mon, Jun 15, 2009 at 1:14 PM, shiva dudeofin...@gmail.com wrote: Hmm Ok Thanks for your reply. In which file, it will be declared?? On Jun 15, 12:50 am, David Turner di...@android.com wrote: Use ANDROID:-) On Sat, Jun 13, 2009 at 6:28 AM, shiva dudeofin...@gmail.com wrote: Hi all... I have a C++ code, which i am porting on Android. This code actually made for Windows system. So, there are many Windows specific header files are used. So, i want to build same code on Android with conditional compilation using OS bases pre processor macros. As windows have WIN32 and Symbain nave __SYMBIAN32__. For Android, what i can use --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: pre compiler macro for andriod
Use ANDROID:-) On Sat, Jun 13, 2009 at 6:28 AM, shiva dudeofin...@gmail.com wrote: Hi all... I have a C++ code, which i am porting on Android. This code actually made for Windows system. So, there are many Windows specific header files are used. So, i want to build same code on Android with conditional compilation using OS bases pre processor macros. As windows have WIN32 and Symbain nave __SYMBIAN32__. For Android, what i can use --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: telnetd problem on android
See http://developer.android.com/guide/developing/tools/emulator.html#emulatornetworking especially the section titled Setting up Redirections through the Emulator Console On Thu, Jun 11, 2009 at 2:03 PM, nearfuture yaoming@gmail.com wrote: I want to open the telnet port on android, so I run the telnetd on android with nothing print and can find the telnetd work with ps. and run the netstat, get the following message Proto Recv-Q Send-Q Local Address Foreign Address State tcp0 0 127.0.0.1:5037 0.0.0.0:* LISTEN tcp0 0 0.0.0.0: 0.0.0.0:* LISTEN tcp0 0 0.0.0.0:23 0.0.0.0:* LISTEN udp0 0 0.0.0.0:9780.0.0.0:* and I do the telnet on the host pc telnet 10.192.225.163 Trying 10.192.225.163... Connected to 10.192.225.163. Escape character is '^]'. Connection closed by foreign host. Anyone can tell me how to enable the telnet on android or ssh Thanks --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Can't adjust backlight in our built emulator from source?
backlight emulation works as follows in 1.5 (this may change in future system images though): 1/ hardware/qemu/liblights/lights.c is compiled into /system/lib/hw/ liblights.goldfish.so 2/ libhardware recognize at runtime that it is running in the emulator (goldfish) system, then loads it 3/ the code in lights.c will be used by the system due to this, and will send messages to the emulator through the qemud channel to the hw-control service. 4/ the emulator receives it and emulates a change of backlight If you don't know what QEMUD and QEMUD services are, see: http://android.git.kernel.org/?p=platform/external/qemu.git;a=blob;f=docs/ANDROID-QEMUD.TXT http://android.git.kernel.org/?p=platform/external/qemu.git;a=blob;f=docs/ANDROID-QEMUD-SERVICES.TXT you can debug that things are working as intended with: emulator -debug-hw-control = dumps all messages sent to the hw-control QEMUD service received by the emulator. emulator -debug-qemud = dumps all traffic on the QEMUD serial port multiplexing channel. I can imagine several reasons why this is not working for you: - the system image doesn't have emulator-specific binaries in it (this is the case for user builds, you will need userdebug or eng instead) - you don't have /system/etc/init.goldfish.rc and/or /system/etc/ init.goldfish.sh and the emulation support is not started properly - you don't have /system/lib/hw/liblights.goldfish.so for some reason - you're not running under a goldfish kernel in the emulator, so liblights.goldfish.so is never loaded/used. Hope this helps On Wed, Jun 10, 2009 at 11:42 AM, ming mingderw...@gmail.com wrote: Dear All, Before we make any real platform system images, we try to build a generic version of emulator from source. But the emulator don't support backlight (we can't turn off or adjust the backlight for the emualtor) as the emulator in SDK 1.5 does. Why? Any clue? Ming- --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: How to use external modem in emulator
On Wed, Jun 3, 2009 at 11:24 AM, sram sayiram.koth...@gmail.com wrote: Hi i am trying to connect USB external modem(my Motorola Phone) on to Android emulator. i have tried the following command emulator -radio ttyACM0 -verbose -debug all and from the ADB shell tried echo ATD1234 ttyS1 but there is nothing happened. ttyS1 is not connected to the emulated modem at all. Actually, none of the emulated serial ports are, so it's not surprising that this doesn't work. as well itried using the local host 5554 and then gsm call 12344555 but got the rror message KO: Modem Emulation not started can any one help me in this. Probably because the GSM stack tried to talk with your ttyACM0 device and got answers it could not parse properly, for some reason. Try adb logcat -b radio to display the radio log, which includes the AT commands exchanged between the modem and the stack. Sram --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: How to package other data files with JNI library?
There is no other way at the moment, however you can simply put the file into res/raw and copy them to /data/data/appname when your app starts for the first time. On Fri, May 29, 2009 at 11:18 PM, efu ericf...@gmail.com wrote: I am building an app using JNI library. The JNI library needs to read some input data from the file system in runtime. To package the JNI lib into the package, I created a /libs/armeabi directory under the root tree of my project and copied libMyLib.so into it. The apk package was successfully created and installed and I can see libMyLib.so was placed correctly under /data/data/ com.mycompany.myapp/lib directory. The question is how I can pack the user data in a similar way and have the package installer place them into the same folder? I tried to put them under the same libs/armeabi directory under project root but it was not successful. --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: sd card not working :(
you probably need a custom /system/etc/vold.conf file that tells the mound daemon (vold) which device/partition to mount at boot time On Wed, Jun 3, 2009 at 12:21 AM, hairypothead...@googlemail.com hairypothead...@googlemail.com wrote: hey! iv got android running in virtualbox on my advent 4490 laptop. every time i try access the sdcard its not there.. is it because its running in virtualbox or coz its meant for the eee or for some other reason? also is there a way to make a image of my phone with all the apps and everything to run on my laptop? --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
[android-porting] Re: Everyone in google IO gets a free phone --Ha ha ha ha..... -eom
it's short for end of message, which really means that everything is in the title, and nothing in the message body. I find it confusing myself, but hey... On Thu, May 28, 2009 at 5:24 AM, liuhy liuhuaiy...@gmail.com wrote: eom? what is it? On Thu, May 28, 2009 at 1:49 AM, Yi Sun beyo...@gmail.com wrote: --~--~-~--~~~---~--~~ unsubscribe: android-porting+unsubscr...@googlegroups.com website: http://groups.google.com/group/android-porting -~--~~~~--~~--~--~---
