"Unfortunately, Firefox has stopped."

[polyzen]

logcat

[vanitasvitae]

Same for Orfox.

[vanitasvitae]

Same for Orfox.

[thestinger]

That's strange. It appears to be the same GeckoLinker issues as before but the workaround is still in place.

[thestinger]

That's strange. It appears to be the same GeckoLinker issues as before but the workaround is still in place.

[thestinger]

12-20 11:24:25.586 7325 7340 D GeckoLoader: Gecko environment env0: null

[thestinger]

12-20 11:24:25.586 7325 7340 D GeckoLoader: Gecko environment env0: null

[thestinger]

Looks like the environment variables aren't being passed along. I can't look into it right now but that's probably the problem (if anyone wants to fiddle with platform_frameworks_base).

[thestinger]

Looks like the environment variables aren't being passed along. I can't look into it right now but that's probably the problem (if anyone wants to fiddle with platform_frameworks_base).

[vanitasvitae]

I found that nearly all apps that display html content (All browsers, K9 Mail, apps that use WebView) are not working for me. Has that to do with #96?

PS: I should add, that at least "Browser" worked for a short while after the installation, but now crashes.

[vanitasvitae]

I found that nearly all apps that display html content (All browsers, K9 Mail, apps that use WebView) are not working for me. Has that to do with #96?

PS: I should add, that at least "Browser" worked for a short while after the installation, but now crashes.

[thestinger]

No, nothing to do with that. I don't know what's wrong and I'm working on it. I think there's more than one issue and the Firefox stuff isn't related.

[thestinger]

No, nothing to do with that. I don't know what's wrong and I'm working on it. I think there's more than one issue and the Firefox stuff isn't related.

[thestinger]

Issues like this are bottlenecked by build time. Need better hardware to get through this stuff faster. I've been building all day.

[thestinger]

Issues like this are bottlenecked by build time. Need better hardware to get through this stuff faster. I've been building all day.

[thestinger]

I think I've narrowed down the issue in non-Firefox browsers to the set of recent Bionic commits. I'm not entirely sure yet and it needs to be narrowed down to a specific one. I'm still working on narrowing down the cause of the Firefox issue. Just need to keep reverting / unreverting changes and building again until the cause is known.

[thestinger]

I think I've narrowed down the issue in non-Firefox browsers to the set of recent Bionic commits. I'm not entirely sure yet and it needs to be narrowed down to a specific one. I'm still working on narrowing down the cause of the Firefox issue. Just need to keep reverting / unreverting changes and building again until the cause is known.

[thestinger]

The Firefox issue is at least narrowed down to either being caused by user builds or the Bionic changes.

[thestinger]

The Firefox issue is at least narrowed down to either being caused by user builds or the Bionic changes.

[thestinger]

The non-Firefox issues might be fixed in the upcoming release in a few hours. Firefox issues are still there and need to be narrowed down.

[thestinger]

The non-Firefox issues might be fixed in the upcoming release in a few hours. Firefox issues are still there and need to be narrowed down.

[polyzen]

@vanitasvitae, K-9 works fine here

@thestinger could there be a bigger issue with permissions than #104? The storage permission for Spotify also had to be manually set.

Edit: vanitasvitae, have you manually enabled any permissions? To match cm, I enabled all but location for Play Services. The Play Store and Google Services permissions already matched.

[polyzen]

@vanitasvitae, K-9 works fine here

@thestinger could there be a bigger issue with permissions than #104? The storage permission for Spotify also had to be manually set.

Edit: vanitasvitae, have you manually enabled any permissions? To match cm, I enabled all but location for Play Services. The Play Store and Google Services permissions already matched.

[thestinger]

The storage permission for Spotify also had to be manually set.

Doubt it. CopperheadOS doesn't change any of that. Lots of apps have problems with the dynamic permissions in Android 6.0. The Google Play exceptions are certainly not a bug.

[thestinger]

The storage permission for Spotify also had to be manually set.

Doubt it. CopperheadOS doesn't change any of that. Lots of apps have problems with the dynamic permissions in Android 6.0. The Google Play exceptions are certainly not a bug.

[thestinger]

I think the Chromium/WebView issue is caused by overly aggressive dynamic fortify changes introduced recently and if that's the case then it's fixed now. The Firefox problem is unrelated and is likely related to the problems from before. It's probably caused by OpenBSD malloc but it's not clear how.

[thestinger]

I think the Chromium/WebView issue is caused by overly aggressive dynamic fortify changes introduced recently and if that's the case then it's fixed now. The Firefox problem is unrelated and is likely related to the problems from before. It's probably caused by OpenBSD malloc but it's not clear how.

[vanitasvitae]

@polyzen I changed some permissions, but I changed them back to check, if this causes the crashes. Still crashes. I found that there is a subcategory in the permissions concerning email stuff (reading attachements and so on) and K-9 is disabled in all of these categories. When I switch the on and go a level up and click the permission again, K-9 is disabled again. May that cause K9 to crash?

PS: Also I realized that photos that I take are not saved to memory. This affects Camera as well as Focal. Permissions aren't touched and seem to be right here. Should we open a new Issue for this or is this related?

[vanitasvitae]

@polyzen I changed some permissions, but I changed them back to check, if this causes the crashes. Still crashes. I found that there is a subcategory in the permissions concerning email stuff (reading attachements and so on) and K-9 is disabled in all of these categories. When I switch the on and go a level up and click the permission again, K-9 is disabled again. May that cause K9 to crash?

PS: Also I realized that photos that I take are not saved to memory. This affects Camera as well as Focal. Permissions aren't touched and seem to be right here. Should we open a new Issue for this or is this related?

[thestinger]

There's a new release but it's still problematic. I don't need new bug reports just yet. I need to fix the existing problems which is taking quite some time, and those may be the causes of other problems.

Applications crashing due to handling the Android 6.0 dynamic permissions poorly isn't a bug though. It's not a CopperheadOS feature.

[thestinger]

There's a new release but it's still problematic. I don't need new bug reports just yet. I need to fix the existing problems which is taking quite some time, and those may be the causes of other problems.

Applications crashing due to handling the Android 6.0 dynamic permissions poorly isn't a bug though. It's not a CopperheadOS feature.

[thestinger]

One of the Firefox compatibility issues has been fixed so it now works in userdebug builds. It doesn't yet work in user builds so it won't work with the published releases. It seems like it must be something like a missing SELinux rule.

[thestinger]

One of the Firefox compatibility issues has been fixed so it now works in userdebug builds. It doesn't yet work in user builds so it won't work with the published releases. It seems like it must be something like a missing SELinux rule.

[vanitasvitae]

Okay. I appreciate your hard work on this project. Btw, Android IMSI Catcher Detector also crashes with and without PaX-softmode enabled. Didn't want to open a new issue since you said you don't need new bug reports just yet :)

[vanitasvitae]

Okay. I appreciate your hard work on this project. Btw, Android IMSI Catcher Detector also crashes with and without PaX-softmode enabled. Didn't want to open a new issue since you said you don't need new bug reports just yet :)

[thestinger]

I think the user build SELinux rules are too strict and it could be the same issue across most of these applications. I've found a few issues with user builds so far but it's important to get it working. It seems that AOSP is only really used for eng and userdebug builds in practice. There are some rough edges to fix.

[thestinger]

I think the user build SELinux rules are too strict and it could be the same issue across most of these applications. I've found a few issues with user builds so far but it's important to get it working. It seems that AOSP is only really used for eng and userdebug builds in practice. There are some rough edges to fix.

[vanitasvitae]

Btw: Are there any plans to root the ROM by default (like in the CM-builds?) Tried to do root my device from within, but then I end up in a bootloop :/

[vanitasvitae]

Btw: Are there any plans to root the ROM by default (like in the CM-builds?) Tried to do root my device from within, but then I end up in a bootloop :/

[thestinger]

The main release will be unrooted to avoid the attack surface of a root implementation. CyanogenMod's implementation was found to be quite insecure and while the obvious issues are now fixed it is still a liability. There can be an alternative channel with root and more debugging support down the road when the project has more resources.

[thestinger]

The main release will be unrooted to avoid the attack surface of a root implementation. CyanogenMod's implementation was found to be quite insecure and while the obvious issues are now fixed it is still a liability. There can be an alternative channel with root and more debugging support down the road when the project has more resources.

[polyzen]

@vanitasvitae, this[] worked for me. Used the update.zip format
[] http://forum.xda-developers.com/android/software-hacking/wip-selinux-capable-superuser-t3216394

On Tue, Dec 22, 2015 at 3:24 PM, vanitasvitae notifications@github.com
wrote:

Btw: Are there any plans to root the ROM by default (like in the
CM-builds?) Tried to do root my device from within, but then I end up in a
bootloop :/

—
Reply to this email directly or view it on GitHub
#105 (comment)
.

[polyzen]

@vanitasvitae, this[] worked for me. Used the update.zip format
[] http://forum.xda-developers.com/android/software-hacking/wip-selinux-capable-superuser-t3216394

On Tue, Dec 22, 2015 at 3:24 PM, vanitasvitae notifications@github.com
wrote:

Btw: Are there any plans to root the ROM by default (like in the
CM-builds?) Tried to do root my device from within, but then I end up in a
bootloop :/

—
Reply to this email directly or view it on GitHub
#105 (comment)
.

[vanitasvitae]

@polyzen Nice, it worked for me too. Thanks 👍

[vanitasvitae]

@polyzen Nice, it worked for me too. Thanks 👍

[thestinger]

Finally narrowed down the Firefox issue and have a solution that should work. Firefox breaks with the combination of PR_SET_DUMPABLE == 0 (which is the case in user builds) and hidepid=2, but not either alone. It crashes because it doesn't have proper error handling and that makes the problem seem a lot worse than it really turned out to be.

Part A (from earlier): https://github.com/CopperheadOS/kernel_lge_hammerhead/commit/18b4a32dcaf788cceb482b96161045b553e1ae3b (was missing in PaX upstream, going to be fixed there too)
Part B (the remaining issue): https://github.com/CopperheadOS/kernel_lge_hammerhead/commit/5275290bc83b37b33a09080b4fd54f6a8fc6cab7 (a backport to both the 3.4 and 3.10 Android kernels)

[thestinger]

Finally narrowed down the Firefox issue and have a solution that should work. Firefox breaks with the combination of PR_SET_DUMPABLE == 0 (which is the case in user builds) and hidepid=2, but not either alone. It crashes because it doesn't have proper error handling and that makes the problem seem a lot worse than it really turned out to be.

Part A (from earlier): https://github.com/CopperheadOS/kernel_lge_hammerhead/commit/18b4a32dcaf788cceb482b96161045b553e1ae3b (was missing in PaX upstream, going to be fixed there too)
Part B (the remaining issue): https://github.com/CopperheadOS/kernel_lge_hammerhead/commit/5275290bc83b37b33a09080b4fd54f6a8fc6cab7 (a backport to both the 3.4 and 3.10 Android kernels)

[thestinger]

Going to be fixed in the next release later today.

[thestinger]

Going to be fixed in the next release later today.

[polyzen]

\o/

[polyzen]

\o/

[polyzen]

@vanitasvitae, sorry, did not respond to your earlier message.

I found that there is a subcategory in the permissions concerning email stuff (reading attachements and so on) and K-9 is disabled in all of these categories. When I switch the on and go a level up and click the permission again, K-9 is disabled again. May that cause K9 to crash?

The permissions were the same here; haven't had to change them. It's been functional; can't recall if it's crashed..

PS: Also I realized that photos that I take are not saved to memory. This affects Camera as well as Focal. Permissions aren't touched and seem to be right here. Should we open a new Issue for this or is this related?

Haven't had trouble here, either. I'm on hammerhead-ota_update-2015.12.18.13.08.17.zip

[polyzen]

@vanitasvitae, sorry, did not respond to your earlier message.

I found that there is a subcategory in the permissions concerning email stuff (reading attachements and so on) and K-9 is disabled in all of these categories. When I switch the on and go a level up and click the permission again, K-9 is disabled again. May that cause K9 to crash?

The permissions were the same here; haven't had to change them. It's been functional; can't recall if it's crashed..

PS: Also I realized that photos that I take are not saved to memory. This affects Camera as well as Focal. Permissions aren't touched and seem to be right here. Should we open a new Issue for this or is this related?

Haven't had trouble here, either. I'm on hammerhead-ota_update-2015.12.18.13.08.17.zip

[thestinger]

The new release is uploaded for the Nexus 9 and will be uploaded for the Nexus 5 soon.

[thestinger]

The new release is uploaded for the Nexus 9 and will be uploaded for the Nexus 5 soon.

[thestinger]

The Nexus 5 release is out now.

[thestinger]

The Nexus 5 release is out now.

[polyzen]

👍

Note: Updating doesn't keep gapps and su installations intact. You will have to reinstall them directly after updating (before booting). Install opengapps last due to this: opengapps/opengapps#201. Edit: Looks like superuser.zip flashes fine after flashing opengapps.

[polyzen]

👍

Note: Updating doesn't keep gapps and su installations intact. You will have to reinstall them directly after updating (before booting). Install opengapps last due to this: opengapps/opengapps#201. Edit: Looks like superuser.zip flashes fine after flashing opengapps.

[vanitasvitae]

Unfortunately the update 2015.12.23 does not fix this issue for me.
I still cannot take pictures and any app that shows html content crashes :/
I will try 2015.12.24, to see, if the crashes disappear.

Edit: The strange thing is that it worked for me twice right after installing the updates. I don't know what I did then, but a few hours later it crashes again.

[vanitasvitae]

Unfortunately the update 2015.12.23 does not fix this issue for me.
I still cannot take pictures and any app that shows html content crashes :/
I will try 2015.12.24, to see, if the crashes disappear.

Edit: The strange thing is that it worked for me twice right after installing the updates. I don't know what I did then, but a few hours later it crashes again.

[thestinger]

This bug tracks the Firefox two issues, not the other issues. I haven't done anything about the other issues yet.

[thestinger]

This bug tracks the Firefox two issues, not the other issues. I haven't done anything about the other issues yet.

[thestinger]

#112 is the next issue that's going to be worked on and it may be what you're running into.

[thestinger]

#112 is the next issue that's going to be worked on and it may be what you're running into.

[vanitasvitae]

Ah okay, sorry. Orfox also still crashes on my device (2015-12-23).

[vanitasvitae]

Ah okay, sorry. Orfox also still crashes on my device (2015-12-23).

[thestinger]

I'm aware that there are general stability issues but that's not related to the issues here. If it ever works for you, then you aren't running into this (and I know you're not because it's fixed).

[thestinger]

I'm aware that there are general stability issues but that's not related to the issues here. If it ever works for you, then you aren't running into this (and I know you're not because it's fixed).

[vanitasvitae]

Okay thank you 👍 Let me know if I can help with logs or something. I'd really like to help in some way.

[vanitasvitae]

Okay thank you 👍 Let me know if I can help with logs or something. I'd really like to help in some way.

[thestinger]

#112 tracks whatever is causing the inconsistent access errors. I think it's one bug across all the places where it's happening. It's probably an issue with one of the Bionic changes. There were some minor tweaks as part of the port of AOSP and one of those might have been wrong. It needs to be narrowed down by reverting Bionic changes, doing a clean build and then thoroughly testing to see if it fixes the problem. It's very time consuming so that's why it takes a while to fix these issues despite days of work.

[thestinger]

#112 tracks whatever is causing the inconsistent access errors. I think it's one bug across all the places where it's happening. It's probably an issue with one of the Bionic changes. There were some minor tweaks as part of the port of AOSP and one of those might have been wrong. It needs to be narrowed down by reverting Bionic changes, doing a clean build and then thoroughly testing to see if it fixes the problem. It's very time consuming so that's why it takes a while to fix these issues despite days of work.