Fixed
Status Update
No update yet.
Comments
uc...@google.com
you explicitly mention "brew", which is nothing to do with us, but imply that when you downloaded "platform tools" (presumably from developer.google.com), that worked. so i think you need to talk to the (home?)brew folks...
de...@googlemail.com
No by brew I meant that my platform tools recently updated. I ran brew through an update and as a part of the update it updates all the packages you've installed using brew, and android platfrom tools is one of them. Brew doesn't let you downgrade platform tools version, and neither does the Android website. But I found the previous version of platform tools on another website and uninstalled the 34.0 version and used the 33.0.3 and it was working. So no I wasn't implying that brew broke it, I shouldn't have even included brew in my explanation, just explaining that I recently updated the platform tools and that's when this issue came up.
lf...@google.com
In fact, I just installed the latest Android Platform Tools from https://developer.android.com/studio/releases/platform-tools and it has the same issue.
sa...@gmail.com
did you adb kill-server to ensure you're actually running the new one?
lf...@google.com
I forgot to mention I am on Mac.
ak...@gmail.com
Yes I did that.
lf...@google.com
Are you not gonna fix it even though it was working in the 33.0.3 release?
ak...@gmail.com
Are you not gonna fix it even though it was working in the 33.0.3 release?
at the moment, this is at best on the "maybe there's the a problem?" pile. if macOS was as broken as you suggest, we'd probably have heard more than one complaint by now. so in the meantime, there are lots of other more bugs to look at, that we're already convinced are real.
you can make this bug more convincing by trying to work out what the actual variable is --- is it some specific model of mac? is it only some specific macOS versions? android versions? is there anything interesting in logcat?
ak...@gmail.com
If you want to know specific Mac's and MacOS versions it's broken on, I don't have any time for that, nor do I have the equipment for that, I have one Macbook. I can let you know I am on MacOS version 12.6.3 and am on a MacBook Pro (16-inch, 2019). As far as it being broken on other platforms, I may have time to test on Windows and seem if the platform tools on Windows has the same issue in the latest version. And I don't know how logcat would show any issues if the adb shell command is working just fine over USB, it's just the install function that isn't working and is hanging. I haven't tested all the adb functions nor do I use them all, so no I wasn't suggesting all of ADB was broken on Mac either. I specifically mentioned the adb install command.
lf...@google.com
Also the fact that it was working in the previous platform tools release, and then you assuming that I would make a bug up just because is pretty far out there. It was working and I just noticed it because platform tools was recently updated. As far as specific Android versions, so far I have tested on Android 7 and Android 9 and it still works on Android 7 but does not work on Android 9. Those are the two main Android versions I work on so as far as other Android versions it working or not, I don't know.
lf...@google.com
Also the fact that it was working in the previous platform tools release, and then you assuming that I would make a bug up just because is pretty far out there.
not make it up, just be confused/mistaken. you'd be surprised how many of our bug reports are "user error" or "bad cable". is anyone else on your team having similar troubles? have you tried --no-streaming to confirm it's specifically the streaming?
ak...@gmail.com
Yeah I get that. I tried multiple USB cables too and multiple tablets on Android 9 and still got the same issue. And I tried on the Windows platfrom tools and adb there doesn't have the same issue. As far as anyone else on my team having the same issue, I'm not sure because they don't really update their platform tools. They are on older versions. I also just tried the no-streaming option and it was still having the same hang issue, just get's stuck on performing push install instead.
lf...@google.com
does export ADB_OSX_USB_CLEAR_ENDPOINTS=1 (
ak...@gmail.com
+1 to the comment from enh@ about ADB_OSX_USB_CLEAR_ENDPOINTS - so far I have been unable to reproduce this failure so cannot state conclusively whether this flag can help you out or not.
This particular change was in response to a possible edge-case, which you might have been encountering.
And, as enh@ mentioned we have many cases of confusion (often from the part of the end-users). Thanks for your patience while we work through this!
This particular change was in response to a possible edge-case, which you might have been encountering.
And, as enh@ mentioned we have many cases of confusion (often from the part of the end-users). Thanks for your patience while we work through this!
lf...@google.com
And, it's interesting that it does not regress on Android 7/Mac OS 12.6.3/platform tools 34.0.0.
lf...@google.com
Yes I just tested and that flag fixes the issue.
ap...@google.com
lf...@google.com
lf...@google.com
@al...@ifit.com : Can you please upgrade your kernel version (from 12.6.3) and retest against 34.0.0? (because we may be forced to rollback the revert deployed in 34.0.1, since the revert caused disruption for an internal user who is a counter-example to your situation).
ur...@gmail.com
I don't have a way to upgrade my kernel version. I can test on a different tablet with different android version that may have newer kernel version, but don't have a way to just update the kernel version.
lf...@google.com
If it's a kernel issue with my tablet, I'll just have to live with it and leave that ADB_OSX_USB_CLEAR_ENDPOINTS=1 flag set. I haven't updated to the 34.0.1 version yet because I was on vacation.
ma...@gmail.com
I don't have a way to upgrade my kernel version
he actually meant "upgrade macOS" because of "I can let you know I am on MacOS version 12.6.3 and am on a MacBook Pro (16-inch, 2019)" from earlier. we added this patch because people on macOS 13 seem to need it (and that's all we can test on, because Google requires us to be on the current version for security reasons!). so anecdotally ("lack of evidence") we think that this is probably macOS 12 vs 13 rather than arm64 vs intel, and think that should solve your problem?
sa...@gmail.com
Oh gotcha. Yes I can update my MacOS tonight after work and test tomorrow.
lf...@google.com
thanks! (and apologies in advance in case we have more questions later as we try to work out what the active ingredient is here ... i'll keep my fingers crossed that it's just this :-) )
ma...@gmail.com
@al...@ifit.com: If you had a chance to do a host kernel upgrade (on your M1) and re-test, please do let us know the results (so we can plan appropriately) - thanks!
lf...@google.com
Actually I'm not on an apple chip on my Mac, I'm on an intel i9. And yes I have updated, just haven't had a chance to test yet. Am testing right now.
ma...@gmail.com
Ok so I tested it out and it's not working on 34.0.0 when I'm on Mac OS version 13.2.1. I tried both with the ADB_OSX_USB_CLEAR_ENDPOINTS=1 unset and set and neither worked.
lf...@google.com
Thank you for testing it out - this result appears to be a disparity from your earlier test results when 34.0.0 with ADB_OSX_USB_CLEAR_ENDPOINTS=1 bypassed the disconnect issue on a 12.6.3 host system kernel :(
Will update here when we decide on next steps.
Will update here when we decide on next steps.
ma...@gmail.com
Yes that is correct. The ADB_OSX_USB_CLEAR_ENDPOINTS=1 fixed the issue when I was on previous MacOS version. Do you want me to update to 34.0.1?
na...@gmail.com
Do you want me to update to 34.0.1?
that would be an interesting datapoint, but that should be equivalent to 34.0.0 with ADB_OSX_USB_CLEAR_ENDPOINTS=1... (you remembered the adb kill-server dance, and remembered to use export?)
yo...@gmail.com
Yes I remembered to do that. I've tried to do it multiple times. I'm setting in it my bash_profile so when I'm changing it in there, I remember to source my bash_profile before I run the commands too.
lf...@google.com
Also I just tested again on the latest MacOS with platform tools/adb version 33.0.3 and the install worked successfully both with the ADB_OSX_USB_CLEAR_ENDPOINTS=1 and the ADB_OSX_USB_CLEAR_ENDPOINTS=0 but I'm guessing that's because that flag wasn't added until version 34.0.0?
na...@gmail.com
>>I'm guessing that's because that flag wasn't added until version 34.0.0?
yes, that's correct.
yes, that's correct.
lf...@google.com
Ok any updates on what else I can try?
lf...@google.com
Please upload the adb host side logs when testing against both 33.0.3 and 34.0.0; here're the instructions:
- export ADB_TRACE=all (re-initialize host/server by running `adb kill-server`)
- Run your tests (with and without the ADB_OSX_USB_CLEARENDPOINTS variable set against 34.0.0)
- Attach logs (located under $TMPDIR) named appropriately
- export ADB_TRACE=all (re-initialize host/server by running `adb kill-server`)
- Run your tests (with and without the ADB_OSX_USB_CLEARENDPOINTS variable set against 34.0.0)
- Attach logs (located under $TMPDIR) named appropriately
lf...@google.com
Android Studio QA tested this combination (Intel Mac with 13.x kernel against Android 9) and observed no errors - @jainne Please correct me if my assumption is incorrect.
na...@gmail.com
I'm having trouble finding the logs under $TMPDIR.
na...@gmail.com
you're on a mac, right? i don't think that's usually set there, so adb should have defaulted to /tmp...
na...@gmail.com
Yeah I'm on Mac. I checked in the /tmp folder too and didn't see anything there.
lf...@google.com
$echo $TMPDIR
/var/folders/_k/jg980rm50yxgx0wh0kvy15qm00_ccw/T/
/var/folders/_k/jg980rm50yxgx0wh0kvy15qm00_ccw/T$ls -al adb*
-rw-r-----@ 1 <snip> 5001 27834843 Mar 29 09:28 adb.340380.log
This's what it should look like - also look under /tmp (as enh@ pointed out).
PS: I'd export TMPDIR, do a kill-server and run your test again.
/var/folders/_k/jg980rm50yxgx0wh0kvy15qm00_ccw/T/
/var/folders/_k/jg980rm50yxgx0wh0kvy15qm00_ccw/T$ls -al adb*
-rw-r-----@ 1 <snip> 5001 27834843 Mar 29 09:28 adb.340380.log
This's what it should look like - also look under /tmp (as enh@ pointed out).
PS: I'd export TMPDIR, do a kill-server and run your test again.
lf...@google.com
Ok I got them to show up. I did the one with the ADB_OSX_USB_CLEARENDPOINTS set to 0 first, and then 1 2nd, and didn't want to delete the file because I'm not sure I would find it again so then 2nd file has stuff from the 1st one too I think.
lf...@google.com
Any insights from the logs?
ap...@google.com
I just upgraded and have the same problem. If I use emulators for debugging everything is fine. When I connect my usb debugging device, Flutter Daemon stops working.
ap...@google.com
ps. I'm on MacOs Ventura 13.3
na...@gmail.com
UPDATE: I connected my phone with WiFi debugging and using adb connect <ip-address>:<port-number> command, it works. It would appear to be just a USB Debugging issue.
lf...@google.com
I have the same problem, my mac version is MacOs Ventura13.3. This phenomenon does not occur on all devices, only in certain.
lf...@google.com
The problem is still the same after upgrading to version 34.0.1. I had to rollback to 33.0.3 to be able to install APK on my devices.
ng...@gmail.com
Yes I have also noticed all these things. WiFi debugging works fine when trying to do an ADB install. And it works on 33.0.3.
ar...@gmail.com
Since we have a fair number of folks who're impacted, please assist in identifying the active ingredient here:
- Are you all on Intel Mac host laptops (vs Intel ARM/M1)? If so, please let us know your host kernel version (e.g. 13.3).
- Since it *appears* that the device-side is also a dependency, please provide the details about the Android device as well (both successful as well as unsuccessful), and any other relevant information (e.g workaround such as 33.0.3, enable the environment variable for 34.0.0 and later).
@do.topevolutions@gmail:
>>If I use emulators for debugging everything is fine. When I connect my usb debugging device, Flutter Daemon stops working.
What's the emulator Android API level (which succeeded)?
What's the physical device API level (that failed)?
Do *all* devices succeed against 33.0.3?
@zhoujianbo1205@gmail:
@nenti1980@gmail:
@christophe.dongieux@gmail:
Please provide additional information for the table below (Intel Mac vs ARM M1 Mac? Macbook kernel version? successfully connected Android version? unsuccessful Android version? Workaround if/any)
@alech@ifit; Host: Intel Mac/i9 formerly 12.6.3 kernel; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: 34.0.0, 34.0.1 worked with env var setting, 33.0.3 succeeded
@alech@ifit; Host: Intel Mac/i9 upgraded to 13.3 kernel; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: 34.0.0, 34.0.1 failed, 33.0.3 succeeded
@do.topevolutions@gmail;Host: <Intel Mac? 13.3 Ventura kernel>; Successful device:<emulator Android ver?>; Unsuccessful device:<physical device version: ?> Workaround: <33.0.3 works?>
@zhoujianbo1205@gmail; Host: <Intel Mac?> 13.3 Ventura kernel; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: <?>
@nenti1980@gmail; Host: <Intel Mac? kernel?>; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround:<?>
@christophe.dongieux@gmail; Host: <Intel Mac? kernel?>; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: 34.0.0, 34.0.1 failed, 33.0.3 succeeded
- Are you all on Intel Mac host laptops (vs Intel ARM/M1)? If so, please let us know your host kernel version (e.g. 13.3).
- Since it *appears* that the device-side is also a dependency, please provide the details about the Android device as well (both successful as well as unsuccessful), and any other relevant information (e.g workaround such as 33.0.3, enable the environment variable for 34.0.0 and later).
@do.topevolutions@gmail:
>>If I use emulators for debugging everything is fine. When I connect my usb debugging device, Flutter Daemon stops working.
What's the emulator Android API level (which succeeded)?
What's the physical device API level (that failed)?
Do *all* devices succeed against 33.0.3?
@zhoujianbo1205@gmail:
@nenti1980@gmail:
@christophe.dongieux@gmail:
Please provide additional information for the table below (Intel Mac vs ARM M1 Mac? Macbook kernel version? successfully connected Android version? unsuccessful Android version? Workaround if/any)
@alech@ifit; Host: Intel Mac/i9 formerly 12.6.3 kernel; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: 34.0.0, 34.0.1 worked with env var setting, 33.0.3 succeeded
@alech@ifit; Host: Intel Mac/i9 upgraded to 13.3 kernel; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: 34.0.0, 34.0.1 failed, 33.0.3 succeeded
@do.topevolutions@gmail;Host: <Intel Mac? 13.3 Ventura kernel>; Successful device:<emulator Android ver?>; Unsuccessful device:<physical device version: ?> Workaround: <33.0.3 works?>
@zhoujianbo1205@gmail; Host: <Intel Mac?> 13.3 Ventura kernel; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: <?>
@nenti1980@gmail; Host: <Intel Mac? kernel?>; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround:<?>
@christophe.dongieux@gmail; Host: <Intel Mac? kernel?>; Successful device:<Android 13?>; Unsuccessful device:<?> Workaround: 34.0.0, 34.0.1 failed, 33.0.3 succeeded
ap...@google.com
The device I have been testing on is Android 9 that has had the issues. In the past it worked on Android 7, though I haven't tested recently. Only workaround I have found for Android 9 after updating to MacOS version 13.3 is using previous platform tools version 33.0.3.
ap...@google.com
Android
fo...@gmail.com
Boa noite sou novo aqui mecânico de motos a 30 anos preciso evoluir rs
bs...@gmail.com
Boa noite como sou mecânico de motos preciso de velocidade 🚄
Description
AI-181.5540.7.32.5056338, JRE 1.8.0_152-release-1136-b06x64 JetBrains s.r.o, OS Mac OS X(x86_64) v10.14.1 unknown, screens 1440x900
Android Gradle Plugin: 3.2.1
Gradle: 4.6
NDK: from local.properties: 18.1.5063045; latest from SDK: 18.1.5063045;
LLDB: pinned revision 3.1 not found; latest from SDK: (package not found);
CMake: from local.properties: (not specified); latest from SDK: (not found); from PATH: (not found);
IMPORTANT: Please read