[android-beginners] Re: emulator trace wanted
I've come as far as realizing that every process forked by init will segfault early, and I likely need to debug the linux kernel to see where the fault gets generated, in order to better pinpoint the point where qemu misbehaves. So I attached to qemu's gdb server, but then realized that the prebuilt kernel is not of a file format I know. Anyone know what kind of operation I need to do to extract the kernel image so I get symbols? Or where I can find build information for cross-building the kernel, that may be the way I need to go anyhow. Niklas Hallqvist wrote: Ah nice, after fixing a 64-bit issue with the actual logging in qemu I got to see the linux console, and it turns out that the servicemanager respawns repeatedly, and finally init gives up on it: ... init: cannot open '/initlogo.rle' yaffs: dev is 32505856 name is mtdblock0 yaffs: passed flags yaffs: Attempting MTD mount on 31.0, mtdblock0 yaffs: dev is 32505857 name is mtdblock1 yaffs: passed flags yaffs: Attempting MTD mount on 31.1, mtdblock1 yaffs: dev is 32505858 name is mtdblock2 yaffs: passed flags yaffs: Attempting MTD mount on 31.2, mtdblock2 init: cannot find '/system/bin/playmp3', disabling 'bootsound' init: untracked pid 22 exited init: untracked pid 23 exited init: untracked pid 48 exited init: untracked pid 49 exited init: critical process 'servicemanager' exited 4 times in 4 minutes; rebooting into recovery mode Restarting system with command 'recovery'. Reboot failed -- System halted I did compile qemu with gcc3.3, since that is what is default on OpenBSD, and qemu is reported o not misbehave until after 4.2.something. There are some 64 bit issues in qemu, most of them in slirp, but they look a bit atypical, almost intentional, and more modern qemus does not have that code changed. I will investigate some more. I just wanted to give show what happened here, and if anyone has any ideas on how to better debug my problem, feel free to tell me. I don't really think this is the right group I am using, it's just that I am a beginner on android and don't really know what is supposed to happen on boot. Otherwise I'm a longtimer in opensource (20 years or so), and operating systems development, so I'm not a beginner on the technical issues. What group should I discuss porting issues in, when it comes to host system porting (as opposed to target)? Niklas David Turner wrote: hmmm, try to invoke it with: emulator -show-kernel -logcat '*:v' for the record, the emulator is a fork of QEMU 0.8.2 and, as such, can only be compiled into a correct binary with specific versions of gcc (see the README.ANDROID for details in the sources), and only 32-bit builds are supported. This issue will be fixed in a later release, though I can't specify which. On Sat, Dec 20, 2008 at 12:24 AM, niklas niklas+goo...@appli.seniklas%2bgoo...@appli.se wrote: I have just finished the first port of the android dev system to OpenBSD/amd64, surely to still contain quite a few bugs, but it has gone through the compilation phase ok. When I start the emulator, I have just guessed at appropriate arguments, it fires up the gui, and it looks nice, but then it sits with ANDROID in the window, just highlighting gui elements as the mouse passes over them, and having the cursor blink for about ten seconds, then it stops. I have waited quite a while for a potential initialization of the disk images, but nothing seems to happen, although CPU is chewing. If someone could send me a trace with -debug all, it could help me find out where it ends working... or any hints on what part I may have problem would be appriciated of course. I supect a 64bit issue, saw quite a few warnings during the builds with pointers casted to longs.. felt really nineties actually. Well, it's fun to do porting work like this... --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Android Beginners group. To post to this group, send email to android-beginners@googlegroups.com To unsubscribe from this group, send email to android-beginners-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/android-beginners?hl=en -~--~~~~--~~--~--~---
[android-beginners] Re: emulator trace wanted
I don't need the info on cross-compiling the kernel, I found the PREBUILT file telling how to do it. sorry for the bw waste. Niklas Hallqvist wrote: I've come as far as realizing that every process forked by init will segfault early, and I likely need to debug the linux kernel to see where the fault gets generated, in order to better pinpoint the point where qemu misbehaves. So I attached to qemu's gdb server, but then realized that the prebuilt kernel is not of a file format I know. Anyone know what kind of operation I need to do to extract the kernel image so I get symbols? Or where I can find build information for cross-building the kernel, that may be the way I need to go anyhow. Niklas Hallqvist wrote: Ah nice, after fixing a 64-bit issue with the actual logging in qemu I got to see the linux console, and it turns out that the servicemanager respawns repeatedly, and finally init gives up on it: ... init: cannot open '/initlogo.rle' yaffs: dev is 32505856 name is mtdblock0 yaffs: passed flags yaffs: Attempting MTD mount on 31.0, mtdblock0 yaffs: dev is 32505857 name is mtdblock1 yaffs: passed flags yaffs: Attempting MTD mount on 31.1, mtdblock1 yaffs: dev is 32505858 name is mtdblock2 yaffs: passed flags yaffs: Attempting MTD mount on 31.2, mtdblock2 init: cannot find '/system/bin/playmp3', disabling 'bootsound' init: untracked pid 22 exited init: untracked pid 23 exited init: untracked pid 48 exited init: untracked pid 49 exited init: critical process 'servicemanager' exited 4 times in 4 minutes; rebooting into recovery mode Restarting system with command 'recovery'. Reboot failed -- System halted I did compile qemu with gcc3.3, since that is what is default on OpenBSD, and qemu is reported o not misbehave until after 4.2.something. There are some 64 bit issues in qemu, most of them in slirp, but they look a bit atypical, almost intentional, and more modern qemus does not have that code changed. I will investigate some more. I just wanted to give show what happened here, and if anyone has any ideas on how to better debug my problem, feel free to tell me. I don't really think this is the right group I am using, it's just that I am a beginner on android and don't really know what is supposed to happen on boot. Otherwise I'm a longtimer in opensource (20 years or so), and operating systems development, so I'm not a beginner on the technical issues. What group should I discuss porting issues in, when it comes to host system porting (as opposed to target)? Niklas David Turner wrote: hmmm, try to invoke it with: emulator -show-kernel -logcat '*:v' for the record, the emulator is a fork of QEMU 0.8.2 and, as such, can only be compiled into a correct binary with specific versions of gcc (see the README.ANDROID for details in the sources), and only 32-bit builds are supported. This issue will be fixed in a later release, though I can't specify which. On Sat, Dec 20, 2008 at 12:24 AM, niklas niklas+goo...@appli.seniklas%2bgoo...@appli.se wrote: I have just finished the first port of the android dev system to OpenBSD/amd64, surely to still contain quite a few bugs, but it has gone through the compilation phase ok. When I start the emulator, I have just guessed at appropriate arguments, it fires up the gui, and it looks nice, but then it sits with ANDROID in the window, just highlighting gui elements as the mouse passes over them, and having the cursor blink for about ten seconds, then it stops. I have waited quite a while for a potential initialization of the disk images, but nothing seems to happen, although CPU is chewing. If someone could send me a trace with -debug all, it could help me find out where it ends working... or any hints on what part I may have problem would be appriciated of course. I supect a 64bit issue, saw quite a few warnings during the builds with pointers casted to longs.. felt really nineties actually. Well, it's fun to do porting work like this... --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Android Beginners group. To post to this group, send email to android-beginners@googlegroups.com To unsubscribe from this group, send email to android-beginners-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/android-beginners?hl=en -~--~~~~--~~--~--~---
[android-beginners] Re: emulator trace wanted
Ah nice, after fixing a 64-bit issue with the actual logging in qemu I got to see the linux console, and it turns out that the servicemanager respawns repeatedly, and finally init gives up on it: ... init: cannot open '/initlogo.rle' yaffs: dev is 32505856 name is mtdblock0 yaffs: passed flags yaffs: Attempting MTD mount on 31.0, mtdblock0 yaffs: dev is 32505857 name is mtdblock1 yaffs: passed flags yaffs: Attempting MTD mount on 31.1, mtdblock1 yaffs: dev is 32505858 name is mtdblock2 yaffs: passed flags yaffs: Attempting MTD mount on 31.2, mtdblock2 init: cannot find '/system/bin/playmp3', disabling 'bootsound' init: untracked pid 22 exited init: untracked pid 23 exited init: untracked pid 48 exited init: untracked pid 49 exited init: critical process 'servicemanager' exited 4 times in 4 minutes; rebooting into recovery mode Restarting system with command 'recovery'. Reboot failed -- System halted I did compile qemu with gcc3.3, since that is what is default on OpenBSD, and qemu is reported o not misbehave until after 4.2.something. There are some 64 bit issues in qemu, most of them in slirp, but they look a bit atypical, almost intentional, and more modern qemus does not have that code changed. I will investigate some more. I just wanted to give show what happened here, and if anyone has any ideas on how to better debug my problem, feel free to tell me. I don't really think this is the right group I am using, it's just that I am a beginner on android and don't really know what is supposed to happen on boot. Otherwise I'm a longtimer in opensource (20 years or so), and operating systems development, so I'm not a beginner on the technical issues. What group should I discuss porting issues in, when it comes to host system porting (as opposed to target)? Niklas David Turner wrote: hmmm, try to invoke it with: emulator -show-kernel -logcat '*:v' for the record, the emulator is a fork of QEMU 0.8.2 and, as such, can only be compiled into a correct binary with specific versions of gcc (see the README.ANDROID for details in the sources), and only 32-bit builds are supported. This issue will be fixed in a later release, though I can't specify which. On Sat, Dec 20, 2008 at 12:24 AM, niklas niklas+goo...@appli.seniklas%2bgoo...@appli.se wrote: I have just finished the first port of the android dev system to OpenBSD/amd64, surely to still contain quite a few bugs, but it has gone through the compilation phase ok. When I start the emulator, I have just guessed at appropriate arguments, it fires up the gui, and it looks nice, but then it sits with ANDROID in the window, just highlighting gui elements as the mouse passes over them, and having the cursor blink for about ten seconds, then it stops. I have waited quite a while for a potential initialization of the disk images, but nothing seems to happen, although CPU is chewing. If someone could send me a trace with -debug all, it could help me find out where it ends working... or any hints on what part I may have problem would be appriciated of course. I supect a 64bit issue, saw quite a few warnings during the builds with pointers casted to longs.. felt really nineties actually. Well, it's fun to do porting work like this... --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Android Beginners group. To post to this group, send email to android-beginners@googlegroups.com To unsubscribe from this group, send email to android-beginners-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/android-beginners?hl=en -~--~~~~--~~--~--~---
[android-beginners] Re: emulator trace wanted
hmmm, try to invoke it with: emulator -show-kernel -logcat '*:v' for the record, the emulator is a fork of QEMU 0.8.2 and, as such, can only be compiled into a correct binary with specific versions of gcc (see the README.ANDROID for details in the sources), and only 32-bit builds are supported. This issue will be fixed in a later release, though I can't specify which. On Sat, Dec 20, 2008 at 12:24 AM, niklas niklas+goo...@appli.seniklas%2bgoo...@appli.se wrote: I have just finished the first port of the android dev system to OpenBSD/amd64, surely to still contain quite a few bugs, but it has gone through the compilation phase ok. When I start the emulator, I have just guessed at appropriate arguments, it fires up the gui, and it looks nice, but then it sits with ANDROID in the window, just highlighting gui elements as the mouse passes over them, and having the cursor blink for about ten seconds, then it stops. I have waited quite a while for a potential initialization of the disk images, but nothing seems to happen, although CPU is chewing. If someone could send me a trace with -debug all, it could help me find out where it ends working... or any hints on what part I may have problem would be appriciated of course. I supect a 64bit issue, saw quite a few warnings during the builds with pointers casted to longs.. felt really nineties actually. Well, it's fun to do porting work like this... --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Android Beginners group. To post to this group, send email to android-beginners@googlegroups.com To unsubscribe from this group, send email to android-beginners-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/android-beginners?hl=en -~--~~~~--~~--~--~---
