FC/crash on android, with signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 00000068 #643
Labels
Comments
|
For debugging this, the "adb logcat" output would be interesting. You need a bit of developer experience in order to do that, though. |
|
Thank your advice very much! I did it as your method, and got the result as following, but I don't know how to do later. For debugging this, the "adb logcat" output would be interesting. You need a bit of developer experience in order to do that, though. |
|
Sorry I didn't get back earlier. It segfaults. Unfortunately that is an error I did not yet track down. I have seen the same issue (well, or a similar issue, it is hard to say exactly where it happens) on my Phone, too. It will take some days, even a week, until I will have time to investigate this further. |
|
Thank you very much! If you know the solution, please tell me. And I'll try to solve it. |
chrox
changed the title
|
Did you try rebooting your device after installing Koreader? Some users reported that it solved the force close problem on some Android phones. |
|
Thanks for your reply! I tried it many times, but it didn't work. And on my another device, the situation was better, but it crashed when it worked some time. ------------------ Original message ------------------ Did you try rebooting your device after installing Koreader? Some users reported that it solved the force close problem on some Android phones. ¡ª |
|
I wonder if the compiling flags for K2 compatibility have any effects on this while actually we have no support on K2 at all for now. And the same with arm arch flags. These tweaks ought to be tries by those who can reproduce the FC on their Android devices. But "unfortunately" both of my Android phones cannot give any FC at this moment. |
|
@chrox: L72: Unless the target device is armv5, that should be pretty much safe for anything armv6+, unless the target device doesn't support any kind of vfp. I don't think anything ever runs in Thumb1 only, so the forced Arm mode shouldn't hurt either. In any case, if it were to cause any issue, it would hopefully throw a SIGILL, and not SIGSEGV, but the JIT stuff might make matters more complicated on that front. L70 is all about glibc backward compat, so it should pretty much have no impact on bionic. |
|
And apparently the C65 is running on a 1GHz A8 anyway, so, pretty much like the Kobo Aura HD, IIRC. And very close to what the K5 uses. |
|
@chrox : Thank you very much! I tried your method, maybe I didn't understand rightly. I just invalidated ARM_BACKWARD_COMPAT_CFLAGS and ARM_ARCH. But it didn't work. And I tried to change "./mk-luajit.sh armeabi-v7a" to |
|
These days, I tried my best to fix these problems. Even if there is a little progress, but these problems are not completely resolved. From the log I thought the "koreader-g3f1827f.7z" file couldn't be unziped, so I unziped the file and copied them to /data/data/org.koreader.launcher/files. Luckly, it can run. However, the new issue came out. The errors are "Failed to receive dispatch signal. status=-11" and "signal 4 (SIGILL), code 1 (ILL_ILLOPC), fault addr 83609c10", the others are:(https://cloud.githubusercontent.com/assets/7880266/3311239/ce26d966-f6c1-11e3-8805-328dab4aa186.png) |
|
Huh. Not familiar enough with the Androic TC to figure out why it would be doing something stupid with binaries. Another very remote possibility is that the C65 runs a CPU with a weird VFP variant... Could you post the output of
|
|
Thanks for your reply. The CPU's info is these as following: Hardware : RK29board Are there any problems? Moreover, is there a solution? |
Closed
chrox
changed the title
|
FWIW, on a K5 (A8) So, besides the deprecated ThumbEE/Jazelle stuff, the C65 runs on pretty much the exact same thing. TL;DR: No idea why it goes boom. |
|
Thanks very much! |
|
This issue also happons on my xperia sp, running carbon rom (4.4.4). Using both a self compiled package (code version of 24-8-2014) and a downloaded apk koreader-android-arm-linux-androideabi-v2014.04-630-gbd2fa87.apk. logcat below is from the downloaded apk part of the adb locat is below, hopefully a solution can be found: |
|
Ah yes, that dreaded segfault when using fwrite(). I've seen this pretty often now - and still can't make much sense from it. I need to do an intensive debugging session, but will not have free time for that the next 10 days, unfortunately. |
|
by any chance did you figure out what was happening with your native crash? I am experiencing a similar problem with my app |
|
Nope. It seems to have vanished without us really tackling it... My guess is on an API/ABI mismatch for the code we compile from external libraries, whose build systems are not easily to bring under proper control, so that they may be including (potentially incompatible) system headers. But this is just speculation. |
Closed
|
nexus 7，2012，4.4.4. |
|
I installed release 127 on Boyue T62 and while it resolves the full screen issue, it looks like no progress on the crashes. KR not just crashes while reading Pdfs it also crashes when clicking on menus and such. It is pretty crashes, (it just vanishes) I cant have it run more than 30 sec. |
|
Btw I do not have Xposed or Xprivacy or similar apps installed on my Boyue T62. |
|
Use this patch, crash seems become less: |
|
Same issue with my Mega 6.3" stock Touchwiz Kitkat 4.4.2. Meat of the logcat: I have Xposed :) |
|
Same issue, on my Sony Z3C, 5.1.1 HongKong stock rom. |
This was referenced Jun 25, 2016
|
I'm going to close this because this particular problem doesn't really occur anymore, even if nothing was ever fixed as such. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
My device is an e-ink device of android 2.3.1. The problem is that the app can't run in the device, and the platform is android-9. I think it can be executed. And are there any good methods to improve?
The text was updated successfully, but these errors were encountered: