Assigned
Status Update
No update yet.
Comments
we...@gmail.com
Thank you for reporting this issue. For us to further investigate this issue, please provide the following additional information:
Android build
Which Android build are you using? (e.g. OPP1.170223.012)
Steps to reproduce
What steps are needed to reproduce this issue?
Please provide sample project or apk to reproduce the issue. Also mention the steps to be followed for reproducing the issue with the given sample project or apk.
Frequency
How frequently does this issue occur? (e.g 100% of the time, 10% of the time)
Screen record of the issue, for clarity
Please capture screen record or video of the issue using following steps:
adb shell screenrecord /sdcard/video.mp4
Subsequently use following command to pull the recorded file:
adb pull /sdcard/video.mp4
Attach the file to this issue.
NOTE : To avoid leaking private information, please share screenshots and bugreports only in Google Drive. Share files with android-bugreport@google.com and include only Google drive links in your bug. Bug report attachments should not be included directly in issue reports.
Android build
Which Android build are you using? (e.g. OPP1.170223.012)
Steps to reproduce
What steps are needed to reproduce this issue?
Please provide sample project or apk to reproduce the issue. Also mention the steps to be followed for reproducing the issue with the given sample project or apk.
Frequency
How frequently does this issue occur? (e.g 100% of the time, 10% of the time)
Screen record of the issue, for clarity
Please capture screen record or video of the issue using following steps:
adb shell screenrecord /sdcard/video.mp4
Subsequently use following command to pull the recorded file:
adb pull /sdcard/video.mp4
Attach the file to this issue.
NOTE : To avoid leaking private information, please share screenshots and bugreports only in Google Drive. Share files with android-bugreport@google.com and include only Google drive links in your bug. Bug report attachments should not be included directly in issue reports.
we...@gmail.com
Android 9 on Nokia 7 Plus
Build number 00WW_3_54H_SP03
Android Studio 3.5.1 on macOS 10.13.6
compileSdkVersion 29
buildToolsVersion '29.0.2'
minSdkVersion 17
targetSdkVersion 28
implementation 'androidx.biometric:biometric:1.0.0-rc01'
Code taken from "Allow for fallback to non-biometric credentials" in
https://developer.android.com/training/sign-in/biometric-auth
Problem:
1. Show biometric prompt.
2. Tap "USE PASSWORD" instead of scanning fingerprint.
3. Unlock using pattern.
4. Activity stops responding.
See attached MP4 video.
This happens consistently.
Project attached in MyApplication.zip
Build number 00WW_3_54H_SP03
Android Studio 3.5.1 on macOS 10.13.6
compileSdkVersion 29
buildToolsVersion '29.0.2'
minSdkVersion 17
targetSdkVersion 28
implementation 'androidx.biometric:biometric:1.0.0-rc01'
Code taken from "Allow for fallback to non-biometric credentials" in
Problem:
1. Show biometric prompt.
2. Tap "USE PASSWORD" instead of scanning fingerprint.
3. Unlock using pattern.
4. Activity stops responding.
See attached MP4 video.
This happens consistently.
Project attached in MyApplication.zip
we...@gmail.com
Same issue with androidx.biometric:biometric:1.0.0-rc02
Android 9 on Nokia 7 Plus
Build number 00WW_3_54H_SP03
Android Studio 3.5.1 on macOS 10.13.6
compileSdkVersion 29
buildToolsVersion '29.0.2'
minSdkVersion 17
targetSdkVersion 28
implementation 'androidx.biometric:biometric:1.0.0-rc02'
Code taken from "Allow for fallback to non-biometric credentials" in
https://developer.android.com/training/sign-in/biometric-auth
Problem:
1. Show biometric prompt.
2. Tap "USE PASSWORD" instead of scanning fingerprint.
3. Unlock using pattern.
4. Activity stops responding.
See attached MP4 video.
This happens consistently.
Android 9 on Nokia 7 Plus
Build number 00WW_3_54H_SP03
Android Studio 3.5.1 on macOS 10.13.6
compileSdkVersion 29
buildToolsVersion '29.0.2'
minSdkVersion 17
targetSdkVersion 28
implementation 'androidx.biometric:biometric:1.0.0-rc02'
Code taken from "Allow for fallback to non-biometric credentials" in
Problem:
1. Show biometric prompt.
2. Tap "USE PASSWORD" instead of scanning fingerprint.
3. Unlock using pattern.
4. Activity stops responding.
See attached MP4 video.
This happens consistently.
go...@chromium.org
I am facing the exact same issue.
Android 9 on Moto Z2 Play (XT1710-07)
Build number PPS29.133-30
Android Studio 3.5.1 on Windows 10 Pro
compileSdkVersion 28
buildToolsVersion '28.0.3'
minSdkVersion 23
targetSdkVersion 28
implementation 'androidx.biometric:biometric:1.0.0-rc02'
Android 9 on Moto Z2 Play (XT1710-07)
Build number PPS29.133-30
Android Studio 3.5.1 on Windows 10 Pro
compileSdkVersion 28
buildToolsVersion '28.0.3'
minSdkVersion 23
targetSdkVersion 28
implementation 'androidx.biometric:biometric:1.0.0-rc02'
dg...@chromium.org
we...@gmail.com
[Deleted User] <[Deleted User]>
This error case appears to be triggered by creating BiometricPrompt immediately before calling authenticate(), rather than recreating the prompt in the Activity's onCreate() (or Fragment's onCreateView()) lifecycle method as intended. That's obviously an easy mistake to make, so I'll work on making sure the documentation is clearer here.
The fact that the activity appears to hang afterward is definitely a bug and will be fixed by aosp/1151594, likely in a 1.0.1 bugfix release. For now, I'm closing this bug as a duplicate of
The fact that the activity appears to hang afterward is definitely a bug and will be fixed by aosp/1151594, likely in a 1.0.1 bugfix release. For now, I'm closing this bug as a duplicate of
te...@gmail.com
Correction:
we...@gmail.com
Must it be in onCreate?
I'm doing this in onResume instead of onCreate so that the prompt shows up when the user comes back to my app.
I'm doing this in onResume instead of onCreate so that the prompt shows up when the user comes back to my app.
da...@chromium.org
Project: platform/frameworks/support
Branch: androidx-master-dev
commit df08869e2c6f2d066f1caf143cebdb965c1b5a2c
Author: Curtis Belmonte <curtislb@google.com>
Date: Mon Nov 18 15:17:00 2019
Fix BiometricPrompt + device credential error case
BiometricPrompt currently has a bug in the following usage scenario:
1. The prompt is not instantiated in onCreate() or onCreateView()
2. Device credential auth is enabled with setDeviceCredentialAllowed()
3. authenticate() is called twice in the same activity/fragment
When this happens, the handler bridge is incorrectly reset before the
handler activity can be launched, preventing the prompt from being shown
at all. While this isn't a recommended usage pattern for
BiometricPrompt at the moment, it still shouldn't cause the prompt to
get stuck in this state. This commit therefore ensures that the above
scenario no longer triggers the bug.
Test: Manually, using sample app from
Test: ./gradlew biometric:test
Test: ./gradlew biometric:connectedAndroidTest
Fixes: 143097321
Change-Id: Ie7e343d9323437d9765e512a09a0e25a6c86dd8e
M biometric/src/main/java/androidx/biometric/BiometricPrompt.java
M biometric/src/main/java/androidx/biometric/DeviceCredentialHandlerActivity.java
https://android-review.googlesource.com/1168851
Branch: androidx-master-dev
commit df08869e2c6f2d066f1caf143cebdb965c1b5a2c
Author: Curtis Belmonte <curtislb@google.com>
Date: Mon Nov 18 15:17:00 2019
Fix BiometricPrompt + device credential error case
BiometricPrompt currently has a bug in the following usage scenario:
1. The prompt is not instantiated in onCreate() or onCreateView()
2. Device credential auth is enabled with setDeviceCredentialAllowed()
3. authenticate() is called twice in the same activity/fragment
When this happens, the handler bridge is incorrectly reset before the
handler activity can be launched, preventing the prompt from being shown
at all. While this isn't a recommended usage pattern for
BiometricPrompt at the moment, it still shouldn't cause the prompt to
get stuck in this state. This commit therefore ensures that the above
scenario no longer triggers the bug.
Test: Manually, using sample app from
Test: ./gradlew biometric:test
Test: ./gradlew biometric:connectedAndroidTest
Fixes: 143097321
Change-Id: Ie7e343d9323437d9765e512a09a0e25a6c86dd8e
M biometric/src/main/java/androidx/biometric/BiometricPrompt.java
M biometric/src/main/java/androidx/biometric/DeviceCredentialHandlerActivity.java
da...@chromium.org
Yes, initializing the prompt in onResume() should be fine as well.
Unmarked as duplicate, since aosp/1168851 should be a more proper fix for this. Targeting it for inclusion in 1.0.1.
Unmarked as duplicate, since aosp/1168851 should be a more proper fix for this. Targeting it for inclusion in 1.0.1.
da...@chromium.org
ve...@chromium.org
I believe I am facing a related issue. I am using BiometricPrompt with setDeviceCredentialAllowed. This is fully working on an older device running Oreo (now that I recreate the activity each time after using BiometricPrompt). The user can choose to use their fingerprint to authenticate, or can select to use the pin. In both cases, once the user successfully authenticates, onAuthenticationSucceeded is called on the BiometricPrompt.AuthenticationCallback.
However when used in the emulator, running Android 10, as soon as the user selects the "Use PIN" option, onAuthenticationError is called with errorCode=5 and errString=Authentication canceled
However when used in the emulator, running Android 10, as soon as the user selects the "Use PIN" option, onAuthenticationError is called with errorCode=5 and errString=Authentication canceled
te...@gmail.com
Yes I am facing the same issue. Once I get the errCode = 5, the activity starting acting very strange. It then again gives an errCode = 10 before crashing again.
te...@gmail.com
Addition to my previous post "https://crbug.com/chromium/1480717#c14 " (I don’t know how to edit posts here):
Note for clearification: on the video I slightly move cursor over the tab, hovering close icon and leaving it over and over again
Note for clearification: on the video I slightly move cursor over the tab, hovering close icon and leaving it over and over again
en...@chromium.org
ping tangm@ and cc sunnpys@
https://crbug.com/chromium/1480717#c9 says DCompPresenter is the cause - should we consider disabling this feature?
su...@chromium.org
Does --disable-direct-composition-video-overlays also fix the problem?
https://crbug.com/chromium/1480717#c16 : DCompPresenter shipped a while back (M114/5?), so I'm not sure if disabling it is risk free now.
ra...@microsoft.com
[Empty comment from Monorail migration]
wi...@microsoft.com
Similar sounding issue reported for Edge + Media Foundation (uses DComp) presentation: https://www.reddit.com/r/MicrosoftEdge/comments/164emku/anyone_know_why_i_get_these_black_squares_when_i/
Sounded like there may have been a known NVidia driver issue (I'm trying to see if the Windows team had any further details here)
Sounded like there may have been a known NVidia driver issue (I'm trying to see if the Windows team had any further details here)
su...@chromium.org
[Empty comment from Monorail migration]
ma...@gmail.com
di...@gmail.com
This is the same bug as 1155285, and its been passed over to Microsoft as a possible OS issue.
we...@gmail.com
Bug #1155285 can be replicated in Chromium builds 1136035 and lower, while bug #1480717 cannot.
ky...@chromium.org
[Empty comment from Monorail migration]
ky...@chromium.org
[Empty comment from Monorail migration]
ky...@gmail.com
I'm able to reproduce with same approach from https://www.youtube.com/watch?v=EHAtNa7I8Vg on beta 120.0.6099.56.
di...@gmail.com
> The result of this bug is the same. As in, the kind of artifacts are the same. But the cause of this bug is different from bug #1155285
No, its the same bug, the manifestation has began to occur in more conditions due to software changes at both the browser and OS end.
No, its the same bug, the manifestation has began to occur in more conditions due to software changes at both the browser and OS end.
di...@gmail.com
This no longer reproduces under 26002 and also fixed in 22635 after .2841.
ky...@chromium.org
[Empty comment from Monorail migration]
ra...@microsoft.com
@kyle, @Dale, @wezelkrozum, and @teakiska2212, does this bug still repro for you? Does disabling DCompPresenter still fix the issue?
ra...@microsoft.com
@patrik.2199, when you say "Disabling DCompPresenter never worked for me at all", do you mean that running Chrome with the "--disable-features=DCompPresenter" still exhibits the bug?
hi...@chromium.org
[Empty comment from Monorail migration]
pa...@gmail.com
@rafae...@microsoft.com, exactly. I even tried with --disable-features=DCompPresenter --disable-direct-composition-video-overlays but as seen on my attached video, it still happens...
pe...@chromium.org
I still see this every day on my Windows nvidia on chrome. Perhaps a bisect is still possible.
di...@gmail.com
This checkerboarding issue seems to occur not just through the web browsers usage of the overlays, but in combination with CEF apps,
is your discord perhaps maximised behind the browser window? The composition issue is a result of a race condition in the compositor that is allowing apps using it to corrupt outside of their own borders, including on top of surfaces occluding them too
is your discord perhaps maximised behind the browser window? The composition issue is a result of a race condition in the compositor that is allowing apps using it to corrupt outside of their own borders, including on top of surfaces occluding them too
pe...@chromium.org
In my case I run vs-code so probably this qualifies. I can try running just chrome and see if this happens again.
pa...@gmail.com
I just tried with Discord closed, only Chrome is open, and it still happens... Installed 22631.3085 a few days ago, as well as the new 551.23 NVIDIA driver and new Chrome 121.0.6167.86, none of them fixed it...
zh...@bytedance.com
@sunnyps and @tangm. Any update with this bug? Should we disable MPO for NVIDIA drivers?
One of my collegue reproduce the issue (using Win10 + Chrome 121 + GTX1660S + 2K144Hz monitor), he ask me how to fix this, based on his test, ` --disable-direct-composition-video-overlays=1` temperately fix the issue.
One of my collegue reproduce the issue (using Win10 + Chrome 121 + GTX1660S + 2K144Hz monitor), he ask me how to fix this, based on his test, ` --disable-direct-composition-video-overlays=1` temperately fix the issue.
ra...@microsoft.com
Earlier in the this issue, a couple of command line flags were mistyped. Recent changes to the blocklist mean that in order to disable direct composition video overlays and software overlays you need to pass the browser --disable_direct_composition_video_overlays=1 and --disable_direct_composition_sw_video_overlays=1 respectively.
For those getting the bug, does using combinations of these flags make the bug disappear? When applying any command line flag, don't forget to fully exit all browser processes for the flag to take effect, otherwise it will be ignored.
jb...@gmail.com
Since I updated my NVIDIA drivers to 551.52, I'm seeing constant checkerboarding in Microsoft Edge, so I'll try these command line flags and see if they make a difference. I'm running Windows 11 23H2 22631.3155.
Supposedly this issue is fixed in Windows 11 22635.2841. The reports since then are for 22631.3007 and 22631.3085. Aren't these versions behind 22635.2841 and therefore don't have the fix yet? Also, has Microsoft confirmed that direct composition overlay fixes were made in 22635.2841 and later?
Supposedly this issue is fixed in Windows 11 22635.2841. The reports since then are for 22631.3007 and 22631.3085. Aren't these versions behind 22635.2841 and therefore don't have the fix yet? Also, has Microsoft confirmed that direct composition overlay fixes were made in 22635.2841 and later?
zh...@bytedance.com
@refael, I've confirm from my collegue that only passing `--disable_direct_composition_video_overlays=1` could fix the problem.
I don't have a high refresh rate monitor though, and I can't reproduce it using 60fps monitor. But I saw someone previously mentionedhttps://chromium.googlesource.com/chromium/src.git/+/02c3a71f121a8b7efc2600d5802ec23b1a5d95c3 is the root cause.
Can someone in this thread do a bisect (using:https://www.chromium.org/developers/bisect-builds-py/ ) to double check if this CL is the root cause?
I don't have a high refresh rate monitor though, and I can't reproduce it using 60fps monitor. But I saw someone previously mentioned
Can someone in this thread do a bisect (using:
ra...@microsoft.com
But I saw someone previously mentioned
https://chromium.googlesource.com/chromium/src.git/+/02c3a71f121a8b7efc2600d5802ec23b1a5d95c3 is the root cause.
In --disable_direct_composition_video_overlays=1 and --disable_direct_composition_sw_video_overlays=1
jb...@gmail.com
I added --disable_direct_composition_video_overlays=1 and --disable_direct_composition_sw_video_overlays=1 to my Microsoft Edge command line arguments and still experience random checkerboarding. I rebooted my PC after making this change to ensure no Edge processes were running in the background.
Here's the full command line:
"C:\Program Files (x86)\Microsoft\Edge\Application\msedge.exe" --profile-directory=Default --disable_direct_composition_video_overlays=1 --disable_direct_composition_sw_video_overlays=1
Also, I've been experiencing checkerboarding since at least November 2021 when I built my current PC. At first, I thought my graphics card was bad, but then I realized it only happened in Chromium apps and chalked it up to a bug. I've also since upgraded from a 3080FE to a 4090FE and it still happens, so it's not a hardware issue.
Here's the full command line:
"C:\Program Files (x86)\Microsoft\Edge\Application\msedge.exe" --profile-directory=Default --disable_direct_composition_video_overlays=1 --disable_direct_composition_sw_video_overlays=1
Also, I've been experiencing checkerboarding since at least November 2021 when I built my current PC. At first, I thought my graphics card was bad, but then I realized it only happened in Chromium apps and chalked it up to a bug. I've also since upgraded from a 3080FE to a 4090FE and it still happens, so it's not a hardware issue.
ra...@microsoft.com
@Jbelair, what do you mean when you say "I added XYZ to my Microsoft Edge command line arguments"? Do you see these command line arguments in the Task Manager details view when you add the "Command line" column?
With certain features enabled, Edge starts soon after login. So if you, for instance, modified a shortcut to have command line flags, they could still be getting ignored when you click on the shortcut. The best way is to fully terminate all Edge processes first before running with command line flags.
jb...@gmail.com
@Jbelair, what do you mean when you say "I added XYZ to my Microsoft Edge command line arguments"? Do you see these command line arguments in the Task Manager details view when you add the "Command line" column?
With certain features enabled, Edge starts soon after login. So if you, for instance, modified a shortcut to have command line flags, they could still be getting ignored when you click on the shortcut. The best way is to fully terminate all Edge processes first before running with command line flags.
@ra...@microsoft.com - I changed them for the Edge shortcut, but I see what you're saying about Edge auto-running on startup now. So, no, I wasn't see the command-line arguments in Task Manager, but after killing all Edge processes and clicking on the shortcut, now I am. Thank you for the clarification.
Now that I'm running Edge with direct composition overlays disabled, I'll try this out and report my findings. Sorry for the confusion.
jb...@gmail.com
@ra...@microsoft.com - Since adding --disable_direct_composition_video_overlays=1 and --disable_direct_composition_sw_video_overlays=1, I have not experienced any checkerboarding. I confirmed the running Edge Brower process has these command line arguments in Task Manager.
pa...@gmail.com
I can also confirm adding --disable_direct_composition_video_overlays=1 and --disable_direct_composition_sw_video_overlays=1 fixes this issue, without slowdown/flicker unlike the previous fixes.
zh...@bytedance.com
@rafael, I think I can stable reproduce this issue on my 4k + 60fps monitor now. I've tested several NVIDIA, AMD and Intel GPU . Here is my result:
1. Intel A380 + Latest Driver + NV12 overlay support: Scaling = not reproduable.
2. AMD RX6600 + Latest Driver + NV12 overlay support: Scaling = not reproduable.
3. RTX3050 + Latest Driver + NV12 overlay support: Direct = reproduable.
4. RTX4070 Super + Latest Driver + NV12 overlay support: Direct = reproduable.
So this seems to be a NVIDIA only issue from my test result.
And, about the command line switch:
1. Only adding `--disable_direct_composition_sw_video_overlays=1` = still reproduable.
1. Only adding ` --disable_direct_composition_video_overlays=1` -> not reproduable.
So this seems only related to ` --disable_direct_composition_video_overlays=1`.
1. Intel A380 + Latest Driver + NV12 overlay support: Scaling = not reproduable.
2. AMD RX6600 + Latest Driver + NV12 overlay support: Scaling = not reproduable.
3. RTX3050 + Latest Driver + NV12 overlay support: Direct = reproduable.
4. RTX4070 Super + Latest Driver + NV12 overlay support: Direct = reproduable.
So this seems to be a NVIDIA only issue from my test result.
And, about the command line switch:
1. Only adding `--disable_direct_composition_sw_video_overlays=1` = still reproduable.
1. Only adding ` --disable_direct_composition_video_overlays=1` -> not reproduable.
So this seems only related to ` --disable_direct_composition_video_overlays=1`.
le...@google.com
[GPU triage] friendly ping. @sunny, @rafael, @tangm do you have any update on this? Thanks. Not sure this is a driver issue or not and whether there is any action we should do.
pe...@chromium.org
This impacts my personal Windows computer every few minutes
ra...@microsoft.com
Thank you, everyone for your patience.
The root cause of the issue is a bug in Windows that manifests when drivers use a certain feature. The fix is in prerelease versions of Windows which you can test drive if you are a Windows Insider. Testing is in progress but I'm hoping we'll be able to service the fix to in-market Windows versions.
pa...@gmail.com
26xxx Windows builds have this fixed, but I find it extreme that it is not fixed since 2021 for stable versions... Thousands of people RMAd their GPUs thinking it's a hardware issue.
di...@gmail.com
nobody is rmaing their gpu thinking this is a hardware issue, don't dramatise the situation.
Tentative word is that kb5035942 fixes this, but i've never had a replication on my desktop pascal cards to confirm or refute.
Tentative word is that kb5035942 fixes this, but i've never had a replication on my desktop pascal cards to confirm or refute.
we...@gmail.com
@di...@gmail.com
KB5035942 was installed on my pc 2 days ago. I still experience 100+ artifacts daily.
"nobody is rmaing their gpu thinking this is a hardware issue":
https://www.reddit.com/r/chrome/comments/15uuyry/comment/k0r8qfy/ - "I sent it to RMA and they said the GPU was fine, so they send it back to me. :("
KB5035942 was installed on my pc 2 days ago. I still experience 100+ artifacts daily.
"nobody is rmaing their gpu thinking this is a hardware issue":
ya...@gmail.com
KB5035942 (os build 22621.3374 or 22631.3374) was not the one that worked for me ; Its the 24H2 or 26090.1 os build That did , If you sign up to window's insider program and pick the Dev channel and go check for updates it should automatically install, Im pretty sure you can also find the iso here as long as you are signed up to the insider program : https://www.microsoft.com/en-us/software-download/windowsinsiderpreviewiso just make sure that the os build that starts with 26xxx (Like the guy said couple comments ago yesterday).
This is, of course, still temporary, and being on a Dev build is not a solution. I just got my laptop a couple of days ago, and from what I can see, this issue has been going on for a pretty long time. But it's very disappointing that these massive corporations can't fix a bug as small as this. These cards and computers ain't cheap. Fix. your. stuff. !
This is, of course, still temporary, and being on a Dev build is not a solution. I just got my laptop a couple of days ago, and from what I can see, this issue has been going on for a pretty long time. But it's very disappointing that these massive corporations can't fix a bug as small as this. These cards and computers ain't cheap. Fix. your. stuff. !
[Deleted User] <[Deleted User]>
I have the same problem with anything involving chromium on Windows, including WebView2. It is a constant and annoying bug, is there any ETA on when the solution will be released for stable versions, as we are many users who use Windows 10 currently for its great stability. There are more than 55 people who have reported this issue on forums such as Reddit e.g.: https://www.reddit.com/r/nvidia/comments/1bvn19o/comment/ky0jalw/
[Deleted User] <[Deleted User]>
I request that this bug, which has been active for at least one year, be taken seriously.
es...@gmail.com
In Chrome, about:flags and switching ANGLE backend to OpenGL fixes this for now ...
I do hope the Windows fix is expedited to stable soon.
I do hope the Windows fix is expedited to stable soon.
ma...@gmail.com
Windows 11 update kb5035942 made it better for my system for whatever reason I dont see the random artefacts anymore when visiting twitter.com I was able to reproduce the issue by:
1. opening a twitch,tv video playing
2. openingtwitter.com feed in another tab
3. now move the mouse cursor over chrome addons icons on the top bar
whenever the mouse leaves/enters an icon and there is a short animation of background color change of the icons in the Chrome addon bar, there is a flicker on the render of Chrome or checkerboards
Same I can reproduce by:
1. opening atwitch.tv stream with chat running
2. opening OBS Studio
3. minimize OBS Studio to task bar
Now there are random flickering and jumping elements in Chrome. If you close OBS or maximize again the issue vanishes.
1. opening a twitch,tv video playing
2. opening
3. now move the mouse cursor over chrome addons icons on the top bar
whenever the mouse leaves/enters an icon and there is a short animation of background color change of the icons in the Chrome addon bar, there is a flicker on the render of Chrome or checkerboards
Same I can reproduce by:
1. opening a
2. opening OBS Studio
3. minimize OBS Studio to task bar
Now there are random flickering and jumping elements in Chrome. If you close OBS or maximize again the issue vanishes.
ra...@microsoft.com
Thank you everyone for your patience.
As I wrote previously, the root cause of the issue is a bug in Windows that manifests when drivers use a certain feature. The fix is in prerelease versions of Windows which you can test drive if you are a Windows Insider. The fix is (now) also making its way through the Windows Servicing branches for in-market versions of Windows. The fix for each Windows version is fairly unique so we're taking our time to make sure it is ready for general consumption and doesn't cause new issues.
In the meantime, the best way to work around the issue locally is to run the browser with the "--disable_direct_composition_video_overlays=1" command line switch.
jb...@gmail.com
Thank you for the update,
- Will the latest versions of Windows get the highest priority in terms of backporting the fix?
- Since the fix is unique to each version of Windows, will they be rolled out all at once, or as they are certified for each Windows servicing branch?
- Could you provide a rough estimate of how much time it might take for the fix to make its way into the in-market versions of Windows? (assuming the latest servicing [23H2] branch gets top priority)
Thank you!
es...@gmail.com
That is somewhat incorrect,- Opera 109.0.5097.38 on Windows 11 22631.3374
will still black screen with "--disable_direct_composition_video_overlays=1"
The absolute fix is switching about:flags ANGLE to OpenGL ...
will still black screen with "--disable_direct_composition_video_overlays=1"
The absolute fix is switching about:flags ANGLE to OpenGL ...
pa...@gmail.com
Setting it to OpenGL causes white flash while switching to YouTube tab… 26100 RTM works properly.
ra...@microsoft.com
That is somewhat incorrect,- Opera 109.0.5097.38 on Windows 11 22631.3374 will still black screen with "--disable_direct_composition_video_overlays=1"
Opera is a closed source browser so it's hard for me to comment on its behavior. Due to the bug, DirectComposition surfaces are unreliable so if Opera is using them to draw its UI or something else, then you'll still see the bug.
The absolute fix is switching about:flags ANGLE to OpenGL ...
If you use OpenGL drivers on Windows, things like DRM videos will not work super well. If you never view those, the workaround might be satisfactory to you.
Since the fix is unique to each version of Windows, will they be rolled out all at once, or as they are certified for each Windows servicing branch?
As I said previously, Windows Insider builds of Windows already have the fix. As we port the fix to earlier versions of Windows, we'll release those to users.
Could you provide a rough estimate of how much time it might take for the fix to make its way into the in-market versions of Windows? (assuming the latest servicing [23H2] branch gets top priority)
I do not have an ETA at the moment. I'll let you know more as I find out.
es...@gmail.com
Thanks for the comments,- it could very well be the case that Opera behaves differently from Chrome even with that flag ...
I am using regular RTX 3060 552.12 driver, but with the ANGLE setting everything seems to work, I might not have tried DRM though!
I am using regular RTX 3060 552.12 driver, but with the ANGLE setting everything seems to work, I might not have tried DRM though!
te...@gmail.com
ra...@microsoft.com
Thank you.
Thank you.
lu...@onet.pl
Hello. I have the same problem. Windows 11 22631.3447 . Switching to OpenGL doesn't help. Artifacts still appear when browsing Facebook. Edge and Chrome browser. For me, what Windows does is a joke. The problem has been going on for so long and they still haven't fixed anything. ASUS RTX 4070 ProArt
le...@gmail.com
Also experiencing black grid artifacts on Netflix. RTX 4080 Super. Driver 551.86 on Windows 11 (23H2).
So far I have only tried disabling hardware acceleration in the Chrome settings, which seems to work, but is obviously not ideal.
So far I have only tried disabling hardware acceleration in the Chrome settings, which seems to work, but is obviously not ideal.
un...@gmail.com
Comment has been deleted.
il...@gmail.com
Comment has been deleted.
ni...@gmail.com
I have the same issue, however I was unable to replicate it through the playlist method and it also doesn't appear to be quite the same.
Basically I also have a checkerboard pattern appearing from time to time but it does so when I reply to comments from the bell icon on the top right. I am unable to replicate the issue consistently, I will upload a video or photo whenever I stumble upon it.
I'm running Windows 11 Pro build 22631 on a Threadripper 7960X TRX50 build and an RTX 4090, however the issue was also present while I had a Ryzen 2700X and X470 chipset so I doubt it's hardware dependant.
Basically I also have a checkerboard pattern appearing from time to time but it does so when I reply to comments from the bell icon on the top right. I am unable to replicate the issue consistently, I will upload a video or photo whenever I stumble upon it.
I'm running Windows 11 Pro build 22631 on a Threadripper 7960X TRX50 build and an RTX 4090, however the issue was also present while I had a Ryzen 2700X and X470 chipset so I doubt it's hardware dependant.
fl...@gmail.com
Windows 11 pro 22631
cpu ryzen 9 5900x
nvidia rtx 3080
same issue appears in my browser from time to time... dont even need to replicate it... this was for example during a twitch stream and wasnt even open in that tab, but it appeared over my messenger chat
got on this page via nvidias post for their new update... fix coming via windows update or so they claim
cpu ryzen 9 5900x
nvidia rtx 3080
same issue appears in my browser from time to time... dont even need to replicate it... this was for example during a twitch stream and wasnt even open in that tab, but it appeared over my messenger chat
got on this page via nvidias post for their new update... fix coming via windows update or so they claim
sa...@gmail.com
Not sure if this is driver-related but it seemed to happen when I updated to 552.22, but when I stream on discord now, the video is a couple of seconds delayed while the audio stays consistent. This never happened before but I do remember the checkerboard artifacting at one point that got fixed. Would this issue I am having currently be at all correlated to the recent drivers? Thanks.
br...@gmail.com
Hello, I have the same problem, I occasionally experience black grid artifacts only on Electron based apps like Discord or Chromium web browsers. Other GPU-intensive applications work just fine.
My hardware/software specs:
CPU: Intel Core i9-13900K Processor
GPU: Gigabyte GeForce RTX 4080 SUPER AERO OC 16G
Motherboard: Gigabyte Z790 Aorus Master
RAM: Kingston Fury Renegade DDR5 32GB 6400MT/s (KF564C32RS-32)
OS: Windows 11 Pro 23H2, Build 22631.3527
Driver: Nvidia Driver GeForce Game Ready 552.22
My hardware/software specs:
CPU: Intel Core i9-13900K Processor
GPU: Gigabyte GeForce RTX 4080 SUPER AERO OC 16G
Motherboard: Gigabyte Z790 Aorus Master
RAM: Kingston Fury Renegade DDR5 32GB 6400MT/s (KF564C32RS-32)
OS: Windows 11 Pro 23H2, Build 22631.3527
Driver: Nvidia Driver GeForce Game Ready 552.22
do...@gmail.com
have the same issues nothing help with 4070ti super
di...@gmail.com
Please stop with the Me too's, its not an issue where pile-ons will get things moving faster and the replication of the issue was already managed over a year ago
The problem manifests in a Windows composition API due in part most likely to a case not covered where Tiled rendering (Or atleast Tiled Cache) capable GPU's can experience artifacts when Intermediate resources are released while presentation is happening.
The origin problem can manifest a different set of artifacts on devices that don't have this tiled cache (flicker, abnormal tearing) which have for the most part gone unnoticed.
This issue has gone back as far as 2021 at this point, with most reports wrongly pointed at nvidia on third party forums and reddit without actually submitting actional feedback to their feedback email.
The problem manifests in a Windows composition API due in part most likely to a case not covered where Tiled rendering (Or atleast Tiled Cache) capable GPU's can experience artifacts when Intermediate resources are released while presentation is happening.
The origin problem can manifest a different set of artifacts on devices that don't have this tiled cache (flicker, abnormal tearing) which have for the most part gone unnoticed.
This issue has gone back as far as 2021 at this point, with most reports wrongly pointed at nvidia on third party forums and reddit without actually submitting actional feedback to their feedback email.
ti...@gmail.com
Which part of the stack is making use of that supposedly broken composition API? The browser, the OS or the GPU driver?
ti...@gmail.com
On a side-note: the stress test software OCCT also suffered from checkerboard patterns in combination with NVidia drivers. The latest release notes say:
> The "Checkerboard/Desktop" Nvidia driver glitch also has been eradicated from this release with a surprising workaround on the tech side of things ! Took me two solid weeks to get it done.
Two weeks for a niche application with single developer vs. years of unsolved graphic glitches in the most widespread combination of OS + drivers + browser engine with large developer teams. Food for thoughts...
> The "Checkerboard/Desktop" Nvidia driver glitch also has been eradicated from this release with a surprising workaround on the tech side of things ! Took me two solid weeks to get it done.
Two weeks for a niche application with single developer vs. years of unsolved graphic glitches in the most widespread combination of OS + drivers + browser engine with large developer teams. Food for thoughts...
lu...@onet.pl
Changing the browser to chrome://flags/ helped me
ANGLE
D3D9
Now I don't have a chessboard. Works on Chrome and EDGE
ANGLE
D3D9
Now I don't have a chessboard. Works on Chrome and EDGE
ab...@gmail.com
Hi
I just managed to solve it.
For context:-
I had a lot of electron artifacts across different applications (especially when special animations or videos were involved).
These applications include VS Code, Github Desktop, Popcorn time, Google chrome on heavy graphics load, etc
I also develop electron applications, and one of my projects required a bunch of overlaying video content which also causes lag and stutters. (worked fined until electron v23.x)
I have since been trying to find and research solutions and narrowed it down to a bug related to MS, Chromium or Nvidia.
My system spec:-
AMD Ryzen 2700x
RTX 2080
Windows 10
Dual monitor setup (idk if this affects it, but just putting it across just in case)
The solution:-
This is what worked for me. Maybe it helps others too.
Download latest nvidia control panel.
In manage 3D settings, global settings, change the following two options:-
OpenGL GDI compatability - Prefer compatible (this will be set to auto by default)
Vertical sync - Off (this is by default set to On, and I had tuned it to fast mode before, now set it back to off instead)
I just managed to solve it.
For context:-
I had a lot of electron artifacts across different applications (especially when special animations or videos were involved).
These applications include VS Code, Github Desktop, Popcorn time, Google chrome on heavy graphics load, etc
I also develop electron applications, and one of my projects required a bunch of overlaying video content which also causes lag and stutters. (worked fined until electron v23.x)
I have since been trying to find and research solutions and narrowed it down to a bug related to MS, Chromium or Nvidia.
My system spec:-
AMD Ryzen 2700x
RTX 2080
Windows 10
Dual monitor setup (idk if this affects it, but just putting it across just in case)
The solution:-
This is what worked for me. Maybe it helps others too.
Download latest nvidia control panel.
In manage 3D settings, global settings, change the following two options:-
OpenGL GDI compatability - Prefer compatible (this will be set to auto by default)
Vertical sync - Off (this is by default set to On, and I had tuned it to fast mode before, now set it back to off instead)
ra...@microsoft.com
Update: The Win11 version of the fix is due to be released as part of an optional Windows update the last week of May. All updates are subject to a gradual rollout which then becomes mandatory during June.
The Win10 version of the fix is still under development.
As I said previously, the bug can be worked around by specifying the --disable_direct_composition_video_overlays=1 command line flag. For Edge, you will need to disable StartupBoost from edge://settings/system for command line flags to take effect.
jb...@gmail.com
Phenomenal news. Thank you for the update!
pe...@gmail.com
yo tengo windows 10 y tengo varios problemas en google chrome artefactos, parpadeos y puntos rojos en algunos videos al extender o minimizar el google chrome, en firefox no me pasa solo en google chrome
di...@gmail.com
KB5037853 contains the fix for 226x1.
pa...@gmail.com
After the fix, instead of checkerboard artifacts, sometimes another tab's partial content becomes visible on the current tab in chromium-based browsers... This has never happened before the fix.
ra...@microsoft.com
@patrik, when you notice another tab's partial content becomes visible, is the only difference on you machine the presence of KB5037853?
Does the problem go away when you run with the --disable_direct_composition_video_overlays=1 command line flag?
Please attach an about:gpu log to the bug.
pa...@gmail.com
To clarify, I'm now running 24H2 (26100.712), so I'm not sure if it's caused by the fix, or it's an entirely new bug to be introduced in 24H2. It occurs similarly like the checkerboard artifacts did, but it's less common and I can't trigger it (for example, I was on my Grafana dashboard, then suddenly a part of the third tab (which was Facebook) appeared for a few seconds.). I didn't have any video playing.
I'm trying with --disable_direct_composition_video_overlays=1 and will report back if it happens again.
GPU log attached.
I'm trying with --disable_direct_composition_video_overlays=1 and will report back if it happens again.
GPU log attached.
we...@gmail.com
Here is a link to my comment from then that contains the links to the 4 webpages (in case it could help you to recreate the issue more frequently):
PS: I first started to notice it in these comments (on 05-12-2020):
pa...@gmail.com
I have disabled direct composition video overlays 3 days ago and haven't seen the tab windows glitch since then, so it looks like it's another DComp bug... I guess we'll see more complaints as people get the fix with the June patch tuesday.
lu...@onet.pl
For me, the chessboard disappeared after the update.
pa...@gmail.com
"After the fix, instead of checkerboard artifacts, sometimes another tab's partial content becomes visible on the current tab in chromium-based browsers... This has never happened before the fix."
This has been happening for me since the checkerboard pattern appeared, it is also more frequent than the checkerboard patterns. I'm on Win10 by the way, so I still don't have the so called fix. I will attach the log Microsoft asked from you, if anything else is needed, let me know.
Edit: Also, I watch a lot of content with subtitles, and the subtitles are also behaving like this, so the previous line comes back for a moment, it's happening everywhere (YouTube, (HBO) Max, etc.)
This has been happening for me since the checkerboard pattern appeared, it is also more frequent than the checkerboard patterns. I'm on Win10 by the way, so I still don't have the so called fix. I will attach the log Microsoft asked from you, if anything else is needed, let me know.
Edit: Also, I watch a lot of content with subtitles, and the subtitles are also behaving like this, so the previous line comes back for a moment, it's happening everywhere (YouTube, (HBO) Max, etc.)
zm...@google.com
[gpu triage] This is a long standing bug, drop to P2
ne...@gmail.com
Comment has been deleted.
ne...@gmail.com
After fixing the checkerboard, there is a chance of catching the following rendering error ( 26100.1000 and 22631.3737 ) Microsoft, can you seriously not fix this??? Your capitalization is $3.321 Trillion
di...@gmail.com
I do not experience that issue and it looks like one fixed by nvidia already in R540 onwards.
Description
Chrome Version : 115.0.5736.0 (Official Build) (64-bit)
URLs (if applicable) :
Other browsers tested:
Add OK or FAIL after other browsers where you have tested this issue:
Safari: PASS
Firefox: PASS
Edge: FAIL (Because it's also based on chromium)
Any animation or edit in the html or css seems to be able to trigger these artifacts randomly every few seconds. Sometimes they are white squares and sometimes black squares.
But one scenario in which a lot of artifacts appear is when I'm on a personal YouTube playlist webpage that displays at least 20 playlist items and YouTube's miniplayer in the bottom right corner of the webpage. Then a lot of artifacts appear when I drag playlist items to other positions.
Requirements:
Reproduction steps:
Prerequisites: Meet the requirements above. (And don't use DevTools or a screenrecorder to record this bug, use a camera instead.)
(1) (Optional, if not already done earlier) Create a personal playlist on youtube.com with at least 20 videos
(2) Go to the playlist's webpage
(3) Click on a video in the playlist
(4) Pause the video
(5) Click on the playlist title to navigate back to the playlist page
(6) The playlist page should now also display a miniplayer at the bottom right of the webpage that contains the video you had just clicked
(7) Scroll down to the bottom of the playlist
(8) (Optional, if not already done at the start) Close any other open Chrome or Chromium windows.
(9) Drag up to 5 times a video for at least 5 seconds to the left, right, up and down. That should be a large enough timespan to trigger these checkerboard artifacts.
Triage steps:
I believe this is a regression, because I can reproduce this in Chromium version 115.0.5736.0 build 1136075 and later. But not in version 115.0.5736.0 build 1136035 and earlier. I also noticed that ANGLE was updated between these builds from version 2.1.20985 to 2.1.20986 when I compared the information in chrome://gpu. So, this could be a regression in ANGLE.
I've compared the following Chromium build snapshots:https://commondatastorage.googleapis.com/chromium-browser-snapshots/index.html?prefix=Win_x64/1136035/ https://commondatastorage.googleapis.com/chromium-browser-snapshots/index.html?prefix=Win_x64/1136075/
Build 1136035:
Build 1136075: