-
-
Notifications
You must be signed in to change notification settings - Fork 344
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
LocationBarPhone app crash at startup #565
Comments
+1 Bromite version Version: v83.0.4103.46 No Is the bug reproducible with latest version? Yes Can the bug be reproduced with corresponding Chromium version? No Is the bug a crash? Yes Describe the bug Tap on Bromite application Expected behaviour It opens with no crash Additional Information Bug is reproducible, and app (bromite) basically unusable. Will have to roll back until fixed |
Came here to post this so I confirm, but my behavior is different. I have to start the app twice to get it to run. It crashes to homescreen the first time, then opens up just fine the second time. v83.0.4103.46 |
@MGislv please fill in the version. Can you post the crash dump via I noticed a crash about |
@uazo @nikolowry @wchen342 if you experience this crash, any idea what could be the cause? |
I don't know whether it's the same problem as there's no log but since |
@csagan5 https://privatebin.net/?aee7cde0dcbac9a5#6fMZpm3Y5AoScSY76LGkJj3fEtqwGDmM52vaHW7oMKjU Make sure to download a copy if you need to further debug, as this paste will expire 23 hours from this writing |
I'll report back after I do a build today |
@mcmurphy8097234789 you need to find the lines of the crash and post them here |
@wchen342 unlikely. There must be a crash log any time you get the message "This app keeps stopping" |
bromite.log |
Later we get another line Thats as far as I can tell. It seems to mention that locationbarphone thing you mentioned earlier To be clear, re-installing the app does not appear to fix the error |
I've downloaded and installed the latest release from bromite.org and had no issues on my Essential PH-1 (arm64) running on LineageOS 16 with MicroG. Edit: Local arm64 build is working as expected for me. Sidebar: @csagan5
|
I have a fix for this bug which will be released with next version; meanwhile you can fetch the fixed version from this pre-release: https://github.com/bromite/bromite/releases/tag/83.0.4103.53
Yes, my bad - it should work on top of
@MGislv you have the issue reported here, same as others.
@nikolowry not crashing also after closing the app? By the way, thanks for testing! |
Just came here say I am facing the same issue on my Realme X (Color OS7) when using the latest Bromite. For me, I have to launch the app twice 9 out of 10 times in order to open it. First time opening crashes. It happens on homescreen as well as on app drawer. I thought the issue was on my device software. Switched back to v81.0.4044.127 for the time being. |
@TheSoulKeeper which is the "latest" version you tested? |
As a quick fix, clearing "Unimportant storage" fixed the crashes in my case. I.E. Bromite app details > Storage > Clear Storage > Free up space |
@csagan5 didn't experience any crashes |
This works for me! |
I was facing the same issue and seeing the same logcat output as posted here. I confirm that the pre-release version fixed it for me. Thanks, @csagan5. EDIT: ok, it's still not fixed. When I kill Bromite, then it won't start when trying to start it from the app drawer on homescreen. It opens though when I open a link from another app (e.g. when clicking on Source of Bromite in Aurora Droid). When crashing I see this logcat error:
Bromite version |
That pre-release .53 crash exactly like the old one to me. |
.53 also still crashes for me on cm-14.1 branch (lineage). I also tried clearing storage and that did not help either. |
|
Tested the latest pre release version and even this one crashes some times and I need to relaunch it again in order to open, |
Hi, I was talking about the latest release and pre release both. |
This issue seems to be back with It might be due to the Omnibox patch, I am running some tests. Reinstalling seems to still be a valid workaround (no data loss). |
with a java decompiler, J.N.M09VlOh_ is and dk0 is
but who are they? https://sites.google.com/a/chromium.org/dev/developers/gn-build-configuration |
@uazo I recently found out that you can un-alias them by looking into In the specific case for public static boolean org_chromium_chrome_browser_flags_ChromeFeatureList_isEnabled(String featureName) {
return J.N.M09VlOh_(featureName);
} If I remember correctly this crash happens because of a race condition where the cached features for the Java side are not yet loaded. |
race condition seems happen here but I don't know why... |
There are multiple threads, is this the right thread you are looking at? |
no, I think it's a single thread, with cascading exceptions (i don't know in java, i'm a c # developer, they are InnerExceptions). I suggest using a try-catch-log, do I fix it or do @csagan5 ?
I am still checking, it might be some inlined cctor() |
I think we could also log ChromeFeatureList_isEnabled, only when it passes by that runnable, by exposing the feature string maybe we can understand which component starts) but I wouldn't put it into production before testing it well |
I've found another path of dk0 used at initializationtime It's It use a PRIORITIZE_BOOTSTRAP_TASKS flag default true... |
If it were a single thread then there would be no race condition.
The rest of the crash log must be in the minidump on the device, they can be downloaded and extracted.
Is this involved in the crash? We must inspect the rest of the log first.
No, Edit: this part of code cannot be trivially removed |
there are multiple threads but i think that is the stack trace of one of them.
we don't know what dk0 it is.
no, no crashes from me |
Yes, but I was wondering if you were looking at the right one when you initially wrote in your comment about I cannot reproduce this at the moment, it looks like rebooting the device makes it easier to reproduce. I can see if I can find the dump after a crash. |
why take it off?? i saw the android source, with maybe you can try to disable that flag, so queue order |
Because there is no account management in Bromite, the signin part is already disabled but the component is not completely removed (the patch would be hard to maintain as well). Feel free to give it a try if you want; see also: ungoogled-software/ungoogled-chromium#520 Note: the crash is also unrelated to the account management, it happens as I explained in previous comment because the dynamic class loader has not finished in time.
We should not touch the |
See https://stackoverflow.com/a/7841448 Last time I had this problem I rebuilt it without Proguard but I cannot recall what was exactly the cause/solution. The crash log I could retrieve is:
|
ok, that's clear, so from those missing lines we don't have any other extractable information
it's from https://github.com/bromite/bromite/releases/download/85.0.4183.84/arm64_ChromeModernPublic.apk ?
|
so, we trap the error and if there is, the default is false alternatively move to |
or alternatively remove the patch completely? EDIT: simply because having the possibility to copy the url or share it seems interesting to me and it could also be developed #661 maybe we find some java open source project ready to use, i saw something in f-droid |
Nice catch! Never thought to decompile the APK to troubleshoot these. I have been using
Sounds sensible.
It makes editing the URL one tap more away; this is of course subjective, but when it was introduced it felt like moving the user one step further away from the URL.
It's a revert of the cleanup.
This has already been discussed, I am not in favor of #661 or adding anything big like that |
I have moved the initialization to |
I've found the complete stack trace in debug!
|
so
is nice to know EDIT: |
Fixed in |
Bromite version
Version: 83.0.4103.46
Arch: arm
Android version: 9.0
Device model: redmi 7a
Is this bug about the SystemWebView?
No
Is the bug reproducible with latest version?
Yes
Can the bug be reproduced with corresponding Chromium version?
No
Is the bug a crash?
Yes
Describe the bug
Steps to reproduce the bug
Steps to reproduce the bug:
"Opening the app causes a crash, same as @Pentaphon"
Expected behavior
It just opens
Screenshots
The text was updated successfully, but these errors were encountered: