You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A strange behavior has been observed in specific OS / GPU vendor combinations where resizing an OpenGL framebuffer by resizing the container window will cause a hard crash during the call to SwapBuffers.
Although the cause is unknown, and seems to be down to specific issues with drivers and/or software configuration, a possible workaround is to scale the control along only one axis at a time, as documented elsewhere: ocornut/imgui#3321
Although not performance critical, and technically not a critical blocker for everyone, it will noticeably and dramatically affect systems who suffer from this specific configuration, so any mitigation measures which do not harm existing behavior will be appreciated.
The text was updated successfully, but these errors were encountered:
It appears that the issue was indeed caused by having one instance of Nahimic running in the background as discussed in ocornut/imgui#3321 and ocornut/imgui#5354.
The fix PR reduces the probability of occurrence but does not completely eliminate it if this service is present and running.
If anyone else hits upon this, it's highly recommended to thoroughly check background processes and services for Nahimic and other malware/bloat-ware mentioned in the above issues.
A strange behavior has been observed in specific OS / GPU vendor combinations where resizing an OpenGL framebuffer by resizing the container window will cause a hard crash during the call to
SwapBuffers
.Although the cause is unknown, and seems to be down to specific issues with drivers and/or software configuration, a possible workaround is to scale the control along only one axis at a time, as documented elsewhere: ocornut/imgui#3321
Although not performance critical, and technically not a critical blocker for everyone, it will noticeably and dramatically affect systems who suffer from this specific configuration, so any mitigation measures which do not harm existing behavior will be appreciated.
The text was updated successfully, but these errors were encountered: