Open Jacalz opened 4 years ago
Is this still present on 1.3.2?
Yes I can still replicate this on the latest tag in the develop branch.
Is it at all like this issue? https://askubuntu.com/questions/67567/how-to-fix-error-xlib-extension-nv-glx-missing-on-display
Nope. This is a desktop computer. I guess I could technically could force the intel iGPU to be turned off in the bios but I very much doubt that it would make a difference as I am using the Nvidia GPU and nothing else.
Sorry I wasn't suggesting tweaking the hardware, was just trying to figure some more out. Which graphics driver is X using?
I didn’t think you did. I just added my thoughts and I guess that they came out a bit wrong 🙃
It should be using the proprietary Nvidia driver 450 or possibly 380, but I will need to double check that when I have time 😅
I'm also getting Xlib: extension "NV-GLX" missing on display ":0"
on close. I'm also on Solus. Is this a Solus-specific thing? Any fix for it?
This is present on Fedora Workstation 33 running Gnome Shell 3.38.2 and nvidia 455 driver via dnf
xorg-x11-drv-nvidia.x86_64 3:455.45.01-1.fc33 @rpmfusion-nonfree-updates
Confirming this is still present on Ubuntu 20.04 LTS (linux/amd64) running Gnome 3.36.8 and nVidia driver 390. Go version 1.16.5 and Fyne version 2.0.3
Exact error text is: Xlib: extension "NV-GLX" missing on display ":1"
Everything looks fine in the app (graphics-wise). The warning on os.exit is just unsightly is all. Hope that helps.
I'm seeing this as well on Ubuntu 18.04
Also present for :0
on gentoo with nvidia-drivers-515.65.01
and x11-libs/libxcb-1.15-r1
.
Edit: I ran it for you with strace.
$ LANG=C strace ./main 2>&1 | grep -i glx
openat(AT_FDCWD, "/usr/lib64/libGLX.so.0", O_RDONLY|O_CLOEXEC) = 3
writev(7, [{iov_base="-\t\5\0\6\0\0\3\6\0a\33cursor\1\0^\0\10\0\7\0\0\3\6\0\0\3"..., iov_len=220}, {iov_base="GLX", iov_len=3}, {iov_base="\0", iov_len=1}], 3) = 224
writev(7, [{iov_base="b\t\3\0\3\0\0\3", iov_len=8}, {iov_base="GLX", iov_len=3}, {iov_base="\0", iov_len=1}], 3) = 12
openat(AT_FDCWD, "/usr/lib64/libGLX_nvidia.so.0", O_RDONLY|O_CLOEXEC) = 12
openat(AT_FDCWD, "/usr/lib64/libGLX.so.0.0.0", O_RDONLY) = 12
openat(AT_FDCWD, "/usr/lib64/libGLX_nvidia.so.515.65.01", O_RDONLY) = 12
openat(AT_FDCWD, "/usr/lib64/libGLX_nvidia.so.515.65.01", O_RDONLY) = 12
openat(AT_FDCWD, "/usr/lib64/libGLX_nvidia.so.515.65.01", O_RDONLY) = 13
writev(7, [{iov_base="b\23\3\0\3\0\0\0", iov_len=8}, {iov_base="GLX", iov_len=3}, {iov_base="\0", iov_len=1}], 3) = 12
writev(7, [{iov_base="b\7\3\0\3\0\0\0", iov_len=8}, {iov_base="GLX", iov_len=3}, {iov_base="\0", iov_len=1}], 3) = 12
writev(7, [{iov_base="b\7\4\0\6\0\0\0", iov_len=8}, {iov_base="NV-GLX", iov_len=6}, {iov_base="\0\0", iov_len=2}], 3) = 16
openat(AT_FDCWD, "/usr/lib64/libxcb-glx.so.0", O_RDONLY|O_CLOEXEC) = 12
writev(7, [{iov_base="b\0\4\0\6\0\0\0NV-GLX\0\0", iov_len=16}], 1) = 16
openat(AT_FDCWD, "/usr/lib64/libGLX.so.0.0.0", O_RDONLY) = 14
openat(AT_FDCWD, "/usr/lib64/libGLX_nvidia.so.515.65.01", O_RDONLY) = 14
openat(AT_FDCWD, "/usr/lib64/libxcb-glx.so.0.0.0", O_RDONLY) = 14
openat(AT_FDCWD, "/usr/lib64/libGLX.so.0.0.0", O_RDONLY) = 14
openat(AT_FDCWD, "/usr/lib64/libGLX_nvidia.so.515.65.01", O_RDONLY) = 14
openat(AT_FDCWD, "/usr/lib64/libxcb-glx.so.0.0.0", O_RDONLY) = 14
memfd_create("/.glXXXXXX", MFD_CLOEXEC) = 26
futex(0x7f03ce724e68, FUTEX_WAIT_PRIVATE, 2, NULLXlib: extension "NV-GLX" missing on display ":0".
It looks like this is the same problem for Manjaro (with X11 and Nvidia GPU).
Does the system also have an integrated GPU?
I can't speak for anyone else here but FYI: When I initially saw this, it was on a desktop computer without an integrated GPU :)
Describe the bug:
It looks like https://github.com/fyne-io/fyne/pull/1153 accidentally introduced a weird Xlib error on application close. I don't know why I didn't notice it when testing the PR though. Seems to just be a warning and not something that break anything as of now, however it would be good to get rid of it :)
To Reproduce:
Steps to reproduce the behaviour:
Screenshots:
Example code:
It works fine to test this using
fyne_demo
.Device (please complete the following information):