Status Update
Comments
en...@google.com <en...@google.com> #2
al...@ifit.com <al...@ifit.com> #3
Could you please try running with -Pandroid.useDexArchive=false? This disables the new dexing pipeline, and fallbacks to the old one.
al...@ifit.com <al...@ifit.com> #4
en...@google.com <en...@google.com> #5
al...@ifit.com <al...@ifit.com> #6
With -Pandroid.useDexArchive=false it looks like this
:app:transformClassesWithDesugarForNonpayItalyCompatDebugAndroidTest
:app:transformClassesWithPreDexForNonpayItalyCompatDebugAndroidTest
Dex: warning: Ignoring InnerClasses attribute for an anonymous inner class
(org.ccil.cowan.tagsoup.Parser$1) that doesn't come with an
associated EnclosingMethod attribute. This class was probably produced by a
compiler that did not target the modern .class file format. The recommended
solution is to recompile the class from source, using an up-to-date compiler
and without specifying any "-target" type options. The consequence of ignoring
this warning is that reflective operations on this class will incorrectly
indicate that it is *not* an inner class.
:app:transformDexWithDexForNonpayItalyCompatDebugAndroidTest FAILED
FAILURE: Build failed with an exception.
* What went wrong:
Execution failed for task ':app:transformDexWithDexForNonpayItalyCompatDebugAndroidTest'.
> com.android.build.api.transform.TransformException: com.android.ide.common.process.ProcessException: java.util.concurrent.ExecutionException: com.android.dex.DexIndexOverflowException: method ID not in [0, 0xffff]: 65536
al...@ifit.com <al...@ifit.com> #7
Just run
./gradlew spoonGermanyDebugAndroidTest
or
./gradlew spoonGermanyDebugAndroidTest -Pandroid.useDexArchive=false
and you will see the issue
al...@ifit.com <al...@ifit.com> #8
Please note that if you are using plugin's internal or Retrolamba support for Java 8 language features, for each lambda expression you will generate 3 additional references, so it might affect your reference count quite significantly.
en...@google.com <en...@google.com> #9
Could you please clarify?
Thank you for your time
al...@ifit.com <al...@ifit.com> #10
I suggest you take a look at Proguard rules that you can specify when running the tests -
al...@ifit.com <al...@ifit.com> #11
Thank you for the link and clarification.
Pardon my many questions I just want to understand the problem better.
We use proguard for the app and define proguard rules which classes to keep for the test runner that we need early on in the start of the test run.
Why did it work with older build tools?
Also why doesnt multidex work for the test app or proguard rules from the debug builds?
Thank you for the quick response
en...@google.com <en...@google.com> #12
The fact that it worked with 2.3.0 and not with 2.4.0-alpha4 is worrying. Because I am unable to reproduce with the test project you've provided, can you please compare the contents of build/intermediates/transforms/preDex/androidTest/ when building tests:
- from clean, with 2.3.0
- from clean, with 2.4.0-alpha4, with -Pandroid.useDexArchive=false
The content of these 2 directories should be the same. Please report any differences.
The support in the test runner is missing to be able to run test apk with multiple DEX files. I don't have much more info on it, sorry.
al...@ifit.com <al...@ifit.com> #13
sorry I am a bit restricted in the work environment so I cannot build the demo project from this location but I executed a diff on a real project where the issue happens for dex folders (predex is not available in 2.3.0) its disabled for our project.
this is for 2.3.0 and 2.4.0-alpha4 in location build/intermediates/transforms/dex/androidTest/nonPayMexicoBleeding for following command:
./gradlew clean connectedNonpayMexicoBleedingDebugAndroidTest
I used flag -Pandroid.useDexArchive=false for 2.4.0 only.
diff -rq nonpayMexicoBleeding\ 2.3.0/ nonpayMexicoBleeding\ 2.4.0-alpha4/
Result:
Only in nonpayMexicoBleeding 2.3.0/debug/folders/1000: 1
Only in nonpayMexicoBleeding 2.3.0/debug/folders/1000: 10
Only in nonpayMexicoBleeding 2.4.0-alpha4/debug/folders/1000: 1f
Does this help?
en...@google.com <en...@google.com> #14
./gradlew clean connectedNonpayMexicoBleedingDebugAndroidTest
Gradle 3.4.0-alpha7 breaks
Commit: ca9e7f7134e5418f9e27e3164efe25521a324a46
Gradle 3.1.2 works
Commit: cfa48cf0b25427d48b328ef6cf80d51fc4e5933b
Cheers
sh...@google.com <sh...@google.com> #15
sh...@google.com <sh...@google.com> #16
al...@ifit.com <al...@ifit.com> #17
en...@google.com <en...@google.com>
sh...@google.com <sh...@google.com>
go...@gmail.com <go...@gmail.com> #18
- If the tested app is mono-dex, test will be mono dex for < 21, and multidex for 21+.
- If the tested app is native multidex, test is as well.
- And if the tested app is legacy multidex, test is mono dex (because currently the test runner does not support pre 21 test multidex apks).
WDYT?
on...@gmail.com <on...@gmail.com> #19
Sounds like a solution that would make it work on gradle 3.4.0>= ?
What are the side effects from the architecture point of view, performance etc, depending on the effort there would it pay off to bring full multidex support to test runner and test apk both?
I always find it confusing what is the difference between:
-multidex test apk
-multidex test runner
I thought that multidex apk is already supported by MultidexInstrumentationTestRunner
so that should work, or are you suggesting to introduce the multidex support to the test runner also (what is less effort and more benefit from your side? )
Best regards & thanks for your time
sh...@google.com <sh...@google.com> #20
For api levels 21+, ART supports multidex for both the test apk and tested apk, so nothing needs to be done there, and there is no performance penalty. However, multidex for < 21 requires patching the class loader, which is done by the test runner for the tested APK. The test runner is unable to patch the class loader for the test apk though, and we this change is not going to change that. So if you are testing your application with min sdk below 21, you will still have to make sure that the test apk contains a single DEX file.
al...@ifit.com <al...@ifit.com> #21
thank you for the explanation. But I still dont understand why did it work before for api <21, what broke it?
Will that be fixed?
Thank you for your time.
al...@ifit.com <al...@ifit.com> #22
en...@google.com <en...@google.com> #23
If this is a decision from the Android team, could you at least give us instructions how to do this patching on our own. I am amazed that we did not see so much breaking problems up until now, but you are telling us that technically we had a bug/s that is now exposed with the change you guys did now.
Is there a way to add this manual patching by ourselves then, and how to do it?
Thx for the support.
al...@ifit.com <al...@ifit.com> #24
en...@google.com <en...@google.com> #25
Is there a way to connect this bug to that commit and somehow get an ETA, we rely heavily on our instrumentation for a safety net and this is a show stopper for us so I need to relay and prioritize this information accordingly in the company.
To be honest the biggest worry for me here is that some of the classes you have there are package protected e.g. MultidexExtractor (from the first glance) so I would have to find a way to find proper list of dex files created from test apk, instrumentation and the app and sum them together in the test runner, not sure can I duplicate this behaviour but I will take a look in it in more detail starting next work week.
sh...@google.com <sh...@google.com> #26
Could we link this bug to the commits and merge them in 2.4.0?
Thanks
al...@ifit.com <al...@ifit.com> #27
@Mario The change is part of the multidex support library, and that is released as part of the SDK (not plugin nor Android Studio). Once the change is in, I will work on adding the support in the plugin. Unfortunately, I cannot provide any ETA. The owner of that change is aware of this bug (it is referenced from an internal bug report that is assigned to him).
al...@ifit.com <al...@ifit.com> #28
1. using native multidex, the fix will be available in the next gradle plugin releases (after 2.4.0-alpha7)
2. using legacy (i.e. support) multidex, we'll have to wait until a new release of com.android.support:multidex:1.0.x library
Do I have that right?
sh...@google.com <sh...@google.com> #29
al...@ifit.com <al...@ifit.com> #30
This causes quite a bit of pain for my team. We currently have a test apk with more than 65k referenced methods (we're already running proguard on the test apk), and we're depending on it building properly. Our minSdkVersion is 18, but we only run our instrumentation tests on API 21+, so the old behavior allowed us to continue running our tests.
With the new behavior in AGP 3.0, our test apk fails to even build. It seems like my only options are:
1) Reduce the number of referenced methods
2) Create a new variant or flavor with minSdk 21+ and use that for running tests
3) Use minSdk 21 for all debug builds
The first option isn't really feasible, and the second is massively painful because it requires us to retrain 50+ devs on which gradle commands to use to run tests (since the commands change when adding a variant or flavor). The third option also isn't ideal because then we lose lint validation for our real minSdkVersion and lose the ability to easily run debug builds on older android versions manually.
Is there any way to keep the behavior from AGP 2.3.x for building test apks until a proper solution for legacy multidex is available?
en...@google.com <en...@google.com> #31
al...@ifit.com <al...@ifit.com> #32
al...@ifit.com <al...@ifit.com> #33
since the multidex 1.0.2 is live I would expect that multidex works fine on api <21 for test apks, but this is not the case.
Did I misunderstand something when you said this will be backported to legacy multidex so test apk can contain multiple dex files also or?
classpath 'com.android.tools.build:gradle:3.0.0-alpha7'
multidex : '1.0.2',
Android Studio 3.0 Canary 7
Build #AI-171.4182969, built on July 14, 2017
JRE: 1.8.0_152-release-884-b01 x86_64
JVM: OpenJDK 64-Bit Server VM by JetBrains s.r.o
Mac OS X 10.12.4
sh...@google.com <sh...@google.com> #34
al...@ifit.com <al...@ifit.com> #35
Thank you for the answer but I don't understand what is inside multidex 1.0.2 then.
in support lib Revision 25.4.0 it is stated:
Concurrent with this support library release, we are also releasing multidex version 1.0.2. This version includes the following important changes:
Allows multidexing of instrumentation APK.
Deprecates MultiDexTestRunner (AndroidJUnitRunner should be used instead).
We use multidex 1.0.2. Does this mean this fix will come out in 1.0.3, do you have any eta window on this topic?
Cheers
sh...@google.com <sh...@google.com> #36
sh...@google.com <sh...@google.com> #37
Looking forward to the integration let us know when you manage to merge it.
Have a nice weekend!
al...@ifit.com <al...@ifit.com> #38
en...@google.com <en...@google.com> #39
thank you
al...@ifit.com <al...@ifit.com> #40
sh...@google.com <sh...@google.com> #41
does this mean this will come out with beta4?
al...@ifit.com <al...@ifit.com> #42
al...@ifit.com <al...@ifit.com> #43
Simply create a "CustomerTestRunner" and override "onCreate" to call "MultiDex.installInstrumentation(context, targetContext)". See this example:
do...@gmail.com <do...@gmail.com> #44
do...@gmail.com <do...@gmail.com> #45
zh...@gmail.com <zh...@gmail.com> #47
Can anyone give us an update and an ETA for this ?
We use build tools 26.0.1, gap 3 beta 5, ASTL runner 1.0.1.
ch...@gmail.com <ch...@gmail.com> #48
al...@ifit.com <al...@ifit.com> #49
run new tests.
Jared Burrows
*jaredsburrows@gmail.com <jaredsburrows@gmail.com>*
LinkedIn:
On Thu, Sep 14, 2017 at 12:14 PM, <buganizer-system@google.com> wrote:
sh...@google.com <sh...@google.com> #50
al...@ifit.com <al...@ifit.com> #51
yo...@gmail.com <yo...@gmail.com> #52
ma...@gmail.com <ma...@gmail.com> #54
ma...@gmail.com <ma...@gmail.com> #55
zh...@gmail.com <zh...@gmail.com> #56
[Deleted User] <[Deleted User]> #57
al...@ifit.com <al...@ifit.com> #58
sh...@google.com <sh...@google.com> #59
al...@ifit.com <al...@ifit.com> #60
sh...@google.com <sh...@google.com>
al...@ifit.com <al...@ifit.com> #61
sh...@google.com <sh...@google.com> #62
al...@ifit.com <al...@ifit.com> #63
All the changes needed are submitted to the four components (build tools, MultidexTestRunner, AndroidJunitRunner and android gradle plugin) and we'll update this bug as soon as it is released, so it will be in Android Gradle Plugin 3.1.
Note that both the deprecated MultiDexTestRunner and the replacement AndroidJUnitRunner will support legacy multidex test APKs
ki...@gmail.com <ki...@gmail.com> #64
Doing it in 3.1 will require everyone who wants to have multidex for test APK working to start using unstable version of AS and AGP.
ra...@gmail.com <ra...@gmail.com> #65
> `Using multidex to create a test APK is not currently supported.`
jq...@gmail.com <jq...@gmail.com> #66
da...@gmail.com <da...@gmail.com> #67
de...@gmail.com <de...@gmail.com> #68
ol...@gmail.com <ol...@gmail.com> #69
th...@gmail.com <th...@gmail.com> #70
va...@hotmail.com <va...@hotmail.com>
sc...@gmail.com <sc...@gmail.com> #71
ja...@gmail.com <ja...@gmail.com> #72
de...@gmail.com <de...@gmail.com> #73
You need to use Android plugin for Gradle 3.1.0-alpha04+, any compatible version of Gradle (e.g. 4.4) and BuildTools (e.g. 27.0.3). Multidex library 1.0.2 will be added by the plugin as a dependency automatically. With this, your build will be successful.
To run tests successfully, the test runner needs to support legacy test APK. AndroidJUnitRunner supports that, so you can use that.
sh...@google.com <sh...@google.com> #74
For example following your suggestion of tl;dr version the test crashes because some files are not found in the main dex file during the app start:
01-09 10:31:06.965 4861-4861/? W/dalvikvm: Class resolved by unexpected DEX: Ltimber/log/Timber$Tree;(0xa75def58):0x99f02000 ref [Lcom/google/devtools/build/android/desugar/runtime/ThrowableExtension;] Lcom/google/devtools/build/android/desugar/runtime/ThrowableExtension;(0xa75def58):0x99caf000
01-09 10:31:06.965 4861-4861/? W/dalvikvm: (Ltimber/log/Timber$Tree; had used a different Lcom/google/devtools/build/android/desugar/runtime/ThrowableExtension; during pre-verification)
01-09 10:31:06.965 4861-4861/? D/AndroidRuntime: Shutting down VM
01-09 10:31:06.965 4861-4861/? W/dalvikvm: threadid=1: thread exiting with uncaught exception (group=0xa66de228)
01-09 10:31:06.975 4861-4861/? E/InstrumentationResultPrinter: Failed to mark test No Tests as finished after process crash
01-09 10:31:06.975 4861-4861/? E/MonitoringInstr: Exception encountered by: Thread[main,5,main]. Dumping thread state to outputs and pining for the fjords.
java.lang.IllegalAccessError: Class ref in pre-verified class resolved to unexpected implementation
at timber.log.Timber$Tree.getStackTraceString(Timber.java:569)
at timber.log.Timber$Tree.prepareLog(Timber.java:544)
at timber.log.Timber$Tree.i(Timber.java:452)
at timber.log.Timber$1.i(Timber.java:288)
at timber.log.Timber.i(Timber.java:63)
Do you have any idea where does this pops out from?
ms...@masslight.com <ms...@masslight.com> #75
This seems to be Java 8 specific bug, and I've filled
sh...@google.com <sh...@google.com> #76
ms...@masslight.com <ms...@masslight.com> #77
mo...@gmail.com <mo...@gmail.com> #78
sh...@google.com <sh...@google.com> #79
ms...@masslight.com <ms...@masslight.com> #80
Until recently we had a big module, with all our code + all our instrurmentation tests code. Then we could check and the plugin (3.1.alpha and 3.1) work well to enable multidex in tests apks.
But we recently dissociated this big module: one contains the production code, the other one contains the instrumentation code. And then, the new plugin breaks: we can still a multidex production app, but the instrumentation test apk is not generated. We hit the 65K limit and the build reports this as an issue. Also, the android test don't have a task: transformClassesWithMultidexlistFor<variant>AndroidTest
What can go wrong ?
ms...@masslight.com <ms...@masslight.com> #81
* a few classes in production code
* quite a lot of classes of instrumentation tests
The module_apk wires together all other modules that are all library modules.
When adding multidexEnabled to module_apk, the production apk is multidexed, but not the test apk. There is no transform / task for the AndroidTest variant related to multidex.
When adding multidexEnabled to all default configuration of all modules, all of them have both instrumentation and production multidex tasks & transformations, all except module_apk...
We have no idea what is going wrong.
mo...@gmail.com <mo...@gmail.com> #82
I've filled
sh...@google.com <sh...@google.com>
si...@gmail.com <si...@gmail.com> #83
Execution failed for task ':app:transformClassesWithMultidexlistForLive'.
> com.android.build.api.transform.TransformException: Error while generating the main dex list.
sh...@google.com <sh...@google.com>
po...@gmail.com <po...@gmail.com> #84
da...@gmail.com <da...@gmail.com>
rk...@hubspot.com <rk...@hubspot.com> #85
no...@gmail.com <no...@gmail.com> #86
sh...@google.com <sh...@google.com> #87
Yes, please file a new bug, and paste the link here. Issue in #78 was due to a custom plugin which would disable multidex. Thanks!
aa...@gmail.com <aa...@gmail.com> #89
var loop = function() { // do cool stuff }
ls...@gmail.com <ls...@gmail.com> #91
sh...@google.com <sh...@google.com> #92
Alternatively, set the env variable mentioned in
ew...@dropbox.com <ew...@dropbox.com> #93
ve...@whirlpool.com <ve...@whirlpool.com> #95
ve...@whirlpool.com <ve...@whirlpool.com> #96
dj...@gmail.com <dj...@gmail.com> #97
doc/api/n-api.md
sh...@google.com <sh...@google.com> #98
It should *not* hang. If it does, the logs need to be uploaded to an appropriate bug (spanning `adb kill-server` )
rk...@hubspot.com <rk...@hubspot.com> #99
I'm on 34.0.5 now for a few weeks now and the problem is still the same as with previous version - adb only works for a few minutes after killing it.
sh...@google.com <sh...@google.com> #100
rk...@hubspot.com <rk...@hubspot.com> #101
Hello, I'm mostly experiencing this while developing on the Android emulators, but it's happening on all emulators and on a physical device (specs below). In terms of connection it happens on the emulator, usb-C cable (tried 2 different ones, with 2 different original Apple hubs) and on Wi-Fi debugging connection.
In terms of reproducibility, it's reproducible 100%. But I can't trigger it. I stop ADB by running killall -9 adb
, Android studio reconnects it and it works for cca. 15-30 minutes. And then at some point, suddenly, stops working.
Symptoms:
- logcat in AS stops updating
- trying to install the app through AS doesn't work (says "waiting for devices" but never finishes)
- debugger doesn't connect
- adb commands mostly don't respond, except for --help or --version (I guess because they're static and don't rely on server)
About ADB_LUBUSB, I'm only running adb from Android Studio. How do I tell it to export that value to 1? I also don't know how to get logs on adb. Googling didn't help, I only found instructions on how to get logs off the Android device that's connected over adb, not for the adb itself. Any tips?
Computer
Apple M1 Max
macOS Ventura 13.6
Android Studio
Android Studio Giraffe | 2022.3.1 Patch 4
Build #AI-223.8836.35.2231.11090377, built on November 13, 2023
Runtime version: 17.0.6+0-17.0.6b829.9-10027231 aarch64
VM: OpenJDK 64-Bit Server VM by JetBrains s.r.o.
Devices
Emulator Pixel 4, API 33, Android 13.0 Google Play, arm64-v8a (used mostly)
Emulator Pixel 6a, API 34, Android 14.0 Google Play, arm64-v8a
OnePlus 8, Android 13
Java version
➜ ~ java --version
openjdk 17.0.6 2023-01-17
OpenJDK Runtime Environment Temurin-17.0.6+10 (build 17.0.6+10)
OpenJDK 64-Bit Server VM Temurin-17.0.6+10 (build 17.0.6+10, mixed mode)
ADB version
➜ ~ adb --version
Android Debug Bridge version 1.0.41
Version 34.0.5-10900879
Installed as /Users/.../Library/Android/sdk/platform-tools//adb
Running on Darwin 22.6.0 (arm64)
rk...@hubspot.com <rk...@hubspot.com> #102
I set the ADB_LIBUSB to 1 and it still happened:
➜ ~ killall -9 adb
➜ ~ export ADB_LIBUSB=1
➜ ~ echo $ADB_LIBUSB
1
➜ ~ adb devices
List of devices attached
emulator-5554 device
// 1 hour later ...
➜ ~ echo $ADB_LIBUSB
1
➜ ~ adb devices
^C <- I had to kill the command manually as it didn't return anything for 1 minute
➜ ~ killall -9 adb
➜ ~ adb devices
List of devices attached
emulator-5554 device
sh...@google.com <sh...@google.com>
sc...@gmail.com <sc...@gmail.com> #103
From: <buganizer-system@google.com>
Date: Mon, Oct 16, 2023, 6:42 PM
Subject: Re:
latest platform tools version 34.0.0
To: <b-system+404257552@google.com>
Cc: <schultrodney928@gmail.com>
Replying to this email means your email address will be shared with the
team that works on this product.
*Changed*
status: In Progress (Accepted) → Fixed
*sh...@google.com <sh...@google.com> added
<
_______________________________
*Reference Info: 270205252 ADB install not working on Android 9 with latest
platform tools version 34.0.0*
component: Android Public Tracker > App Development > SDK > platform tools
> adb <
status: Fixed
reporter: al...@ifit.com
assignee: sh...@google.com
cc: al...@ifit.com, ja...@google.com, sh...@google.com, and 3 more
type: Bug
access level: Default access
priority: P2
severity: S2
blocking: 269001731 <
duplicate issue: 288902771
<
<
retention: Component default
Generated by Google IssueTracker notification system
You're receiving this email because you are subscribed to updates on Google
IssueTracker
<
starred.
Unsubscribe from this issue.
<
sh...@google.com <sh...@google.com> #104
identify the approximate timestamp at which you suspect connectivity dropped.
Host debugging:
- set ADB_TRACE (dynamic setting)
- aosp-master$ export ADB_TRACE=all
- On MacOS, the logs can be found in the directory pointed by the TMPDIR environment variable.
-$ tail -f /tmp/adb*.log
adb I 02-08 09:30:25 139670 139670 main.cpp:62] Android Debug Bridge version 1.0.41
<snip>
Device-side (adbd) debugging:
Set trace_mask on
Restart the daemon and analyze the logs. Caveat is that the device-logs tend to be noisy.
$ adb shell getprop persist.adb.trace_mask
1
$ adb shell setprop persist.adb.trace_mask 0
$ adb shell pkill adbd
$ adb shell
# cd /data/adb
/data/adb # ls -al
total 23
drwx------ 2 root root 3488 2022-02-08 18:04 .
drwxrwx--x 49 system system 4096 2022-01-18 12:13 ..
-rw------- 1 root root 8521 2022-02-08 18:05 adb-2022-02-08-18-04-49-18527
Error(s) that you may run into, and resolution:
$ adb shell setprop persist.adb.trace_mask 0
Failed to set property 'persist.adb.trace_mask' to '0'.
See dmesg for error reason.
$ adb root
$ adb shell setprop persist.adb.trace_mask 0
>>happens on the emulator, usb-C cable (tried 2 different ones, with 2 different original Apple hubs) and on Wi-Fi debugging connection.
FYI, `WiFi Debugging` runs over a different transport therefore it might make sense to open a separate bug for any failure you
experience over that transport. In fact, libusb transport is also distinct from the legacy transport (as far as the possible failure point which is likely USB, is concerned).
One other question: do you experience this failure from adb alone? (if you eliminate Android Studio from this equation)
Studio interacts with the adb server (and is itself the equivalent of the adb client).
If not, the Studio logs are also relevant.
ga...@gmail.com <ga...@gmail.com> #107
El mar., 28 de noviembre de 2023 18:33, <buganizer-system@google.com>
escribió:
li...@gmail.com <li...@gmail.com> #108
sc...@gmail.com <sc...@gmail.com> #109
he...@gmail.com <he...@gmail.com> #110
ra...@gmail.com <ra...@gmail.com> #111
ra...@gmail.com <ra...@gmail.com> #112
ch...@gmail.com <ch...@gmail.com> #113
hs...@gmail.com <hs...@gmail.com> #114
فعلت تحديث كامل للفون software
rk...@hubspot.com <rk...@hubspot.com> #115
Hey sh... sorry for the late reply. The problem is still here. I tried the steps you wrote above and I can't activate the root user, I get this:
➜ ~ adb shell setprop persist.adb.trace_mask 0
Failed to set property 'persist.adb.trace_mask' to '0'.
See dmesg for error reason.
➜ ~ adb root
adbd cannot run as root in production builds
➜ ~
What does this mean? Any way around that?
ja...@gmail.com <ja...@gmail.com> #116
rk...@hubspot.com <rk...@hubspot.com> #117
I managed to find a solution to the problem adbd cannot run as root in production builds
, I had to use the emulator without Play Store included, I had only the Google APIs.
rk...@hubspot.com <rk...@hubspot.com> #118
Regarding the original issue, this weekend I updated my Android Studio to the stable Iguana 2023.2.1 and I also update SDK which included ADB 35.0.0-11411520 (1.0.41) and for the past 27 hours I didn't have to kill adb once. So it seems the problem is fixed, at least for me. Thanks!
ja...@gmail.com <ja...@gmail.com> #119
ia...@gmail.com <ia...@gmail.com> #120
How install
ka...@gmail.com <ka...@gmail.com> #121
al...@gmail.com <al...@gmail.com> #122
[Deleted User] <[Deleted User]> #123
Reinstall all
rk...@hubspot.com <rk...@hubspot.com> #124
I'm sad to inform you that the issue reappeared. My whole team and I are back to killing adb on an hourly basis. :(
rk...@hubspot.com <rk...@hubspot.com> #125
I'm on
Android Debug Bridge version 1.0.41
Version 35.0.0-11411520
Running on Darwin 23.5.0 (arm64)
And Android Studio is
Android Studio Jellyfish | 2023.3.1 Patch 1
Build #AI-233.14808.21.2331.11842104, built on May 15, 2024
Runtime version: 17.0.10+0-17.0.10b1087.21-11572160 aarch64
VM: OpenJDK 64-Bit Server VM by JetBrains s.r.o.
ab...@gmail.com <ab...@gmail.com> #126
#270205252
ab...@gmail.com <ab...@gmail.com> #128
Job
kr...@gmail.com <kr...@gmail.com> #131
This fix has made my custom AOSP based device not being recognized as an Android device (not there at all on the adb devices
list).
Only downgrading to 34.0.4
helps.
OSX Sonoma 14.5, MacBook Air M2.
Let me know if you'd like me to provide some debug logs to help with fixing that in the latest version.
am...@gmail.com <am...@gmail.com> #132
I need to Update my New Phone right now to android 15
em...@sumup.com <em...@sumup.com> #133
Hello, this issue is very present for us. We have a custom AOSP-based device (Android 10) that now often fails to be shown in the device list or hangs on the install command. Sometimes, downgrading to adb prior to 34.0.5 fixes the issue, sometimes upgrading above 34.0.4 fixes it. Sometimes it cannot be fixed. So far we were unable to find a correlation between OS&hardware and the step that fixes it.
Here are adb host logs from a device that attempted connecting with adb 34.0.4-10411341. It is an M1 Pro with Ventura 13.6.1.
One file has the logs where the device was not listed with adb devices
, the other file are the logs from where adb devices
showed the device, but then adb install
hung indefinitely.
en...@google.com <en...@google.com> #134
both of those traces seem to be using the non-libusb macOS usb backend, which is no longer supported. please repro with a 35.* adb using libusb.
em...@sumup.com <em...@sumup.com> #135
The device is a MacBook Pro M3 with Sonoma 14.5.
I've not been able to find someone to reproduce the indefinite install hang on 35+, yet, but I know we had this in the past (unfortunately without collecting host logs). I'll continue to ask around.
m2...@gmail.com <m2...@gmail.com>
am...@gmail.com <am...@gmail.com> #136
b....@gmail.com <b....@gmail.com> #137 Restricted
b....@gmail.com <b....@gmail.com> #138 Restricted
mw...@gmail.com <mw...@gmail.com> #139
#404
Description