-
Notifications
You must be signed in to change notification settings - Fork 511
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Snap solvespace RGBA pixel error #689
Comments
What GPU do you have? |
Intel hd4000 integrated with i7. Lenovo T430s
Téléchargez Outlook pour iOS<https://aka.ms/o0ukef>
…________________________________
De : whitequark <notifications@github.com>
Envoyé : Tuesday, September 1, 2020 2:50:30 AM
À : solvespace/solvespace <solvespace@noreply.github.com>
Cc : joellapointe1717 <joellapointe1717@hotmail.com>; Author <author@noreply.github.com>
Objet : Re: [solvespace/solvespace] Snap solvespace RGBA pixel error (#689)
What GPU do you have?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#689 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AN726H64QC672DR4W73VVVLSDSKTNANCNFSM4QQ5DQAA>.
|
Can you post the output of |
Sure, I’ll do it tomorrow morning because it is 3:12 AM un Montreal, Canada. Is it the complete command fromage terminal? (I’m not a programmer.) |
Yes, it is the complete command. |
Here it is :
|
@ppd can you take a look? seems like a Snap-related bug since the GL version is high enough |
Well, the snap/platform ships its own Mesa based on the Ubuntu 18.04 distribution, so I'm not sure how conclusive that glxinfo is in relation to the snap. However, this seems really strange. I suppose it's possible that there could be an incompatibility of older Mesa with newer DRM drivers, but I'have never experienced such a constellation. It could be wise to ship mesa-utils in the snap to aid debugging such problems more easily. I'll grab my Intel notebook later and investigate. |
I was worried about Intel HD4000 being too old to support OpenGL 3.2 (which IIRC is the version we require). But no, that seems fine, though it does seem that it used to be a problem in the past. |
Hi, |
What host system do you have? Do you have a dual GPU setup? |
no, i have a notebook with a Core i7-8565U and the built-in GPU. No other one built in |
And your OS is also Ubuntu 20.04? |
Ubuntu Budgie 18.04 |
Wayland or X11? @joellapointe1717 The same question: is it an X11 session (I believe that's the default), or is it a Wayland session? I can't reproduce atm, neither on my HD4000 nor on my 620. |
It is an X11 session (default, like you said) |
can I somehow create a debug log? |
when i run the snap from the command line i get this output: /snap/bin$ ./solvespace Missing (absent) translation for '&Line Styles...' (solvespace:11902): glibmm-CRITICAL **: 15:20:00.162: (solvespace:11902): glibmm-CRITICAL **: 15:20:00.168: |
Maybe try: |
I have the same issue, also on Ubuntu 18.04. What I did: $ sudo snap install --edge solvespace
$ solvespace
[...]
(solvespace:27195): glibmm-CRITICAL **: 17:53:02.604:
unhandled exception (type Glib::Error) in signal handler:
domain: gdk-gl-error-quark
code : 1
what : No available configurations for the given RGBA pixel format
(solvespace:27195): glibmm-CRITICAL **: 17:53:02.611:
unhandled exception (type Glib::Error) in signal handler:
domain: gdk-gl-error-quark
code : 1
what : No available configurations for the given RGBA pixel format |
solvespace works fine on the same system after I followed the "Build from source" instructions (built at 094eff7). It looks like only the snap package shows the error. |
There's something funky going on. I'll publish a snap with mesa-utils included for debugging. We must be doing something wrong. |
solvespace_3.0~79457971_amd64.snap.zip That's current master with glxinfo added on top (see https://github.com/ppd/solvespace/tree/glxinfo). Glxinfo is run exactly with the environment that the main solvespace app uses. Install like so wget https://github.com/solvespace/solvespace/files/5244486/solvespace_3.0.79457971_amd64.snap.zip
unzip solvespace_3.0.79457971_amd64.snap.zip
snap install --dangerous ./solvespace_3.0.79457971_amd64.snap Run solvespace.glxinfo Beware: you will need to reverse this manual installation: snap refresh --amend solvespace |
um, if you get only the output of glxinfo with solvesapce.glxinfo, here is the log |
Hi, |
I'm still investigating, but at this time, I'm a bit swamped with other stuff. Your glxinfo looks rather fine though. Maybe there's some strange corner case with GDK/Gtk not picking up the correct config when initializing. Could you do |
It's failing here: https://gitlab.gnome.org/GNOME/gtk/-/blob/master/gdk/x11/gdkglcontext-x11.c#L1159 I'm looking for a simple reproducer to check whether we're doing something wrong in solvespace's snap. Therefore, I've adapted Emmanuele Bassi's glarea-example to run as a snap: https://github.com/ppd/glarea-example/tree/snap
Install like so: Running |
Another option: Run solvespace with GtkInspector: And lastly, please check for AppArmor denials (https://snapcraft.io/docs/debugging-building-snaps): # install snappy-debug
snap install snappy-debug
# listen for denials
snappy-debug.security scanlog Then launch solvespace and check if snappy-debug spits out any interesting stuff. |
@gtxi8f4 Thank you very much. Could you do the |
and finally from /var/log/syslog with snappy-debug.security scanlog :
sudo journalctl --output=short --follow --all | sudo snappy-debug |
I have started it in the terminal: No further output. What I sent was on the syslog. but i have started the "original" solvespace from the snap. Not the one provided by you. But that doesn't make any difference, does it? And thanks for the hint with the code :-) |
It makes no difference. So you have no denials, and GTK is rather happy with your GL environment. It just doesn't seem to be able to find a valid pixel format. The GTK build in snapcraft's gnome-3-34 extension is not debug enabled, and I can't figure out how to run an X11 snap program with GDB atm. I'm not really sure how to debug this without it turning into a large effort. |
what is the difference between the last running snap version and this one? Apparently there were some major changes? |
Do you have a last running revision? That'd be interesting. I feel that the last relevant snap change wrt GTK would be this one: f5415b3#diff-c33a969de50ed245e7d91f6b05968401 |
well, not anymore :-) |
Well, here's 13b461d, the last commit before gnome-3-34: and here's f5415b3, the first commit with gnome-3-34: |
Unfortunately neither of them work. But I know that I had a working version 3 of solvespace via snap |
Are you not running your desktop in 32bit color? Try typing the following:
|
Here it is :
Thank you, I see that people are very motivated to rule out the problem... |
xwininfo -root|grep Depth |
@joellapointe1717 @gtxi8f4 I'm quite amazed to discover that the platform ships a GTK build with debug enabled. So you could try to see if GDK has some nicer debug messages for us:
|
What did you do? I just installed solvespace again via snap to call it with gdk-debug. But I don't need to run the debug anymore, because it runs again without problems :-) |
Nothing. Literally nothing. I wonder whether I could be related to state, i.e. something that's stored in If it has just started magically working for all users, it could also point to an (implicit) fix in a recent update of |
I had deleted solvespace with purge. But that did not help so far. Maybe it was because of a Linux-Update? There was something coming in several times during the days. If no explanation helps anymore, magic is always a possibility ;-) |
Thank you for your support and your time |
@joellapointe1717 I'd appreciate it very much if you could try the following once more and give us feedback:
|
My Lenovo T430s decided to not power up anymore (silent key as they say in ham radio jargon). So I’ll get access to a computer in the next few days. Sorry for the delay. |
@joellapointe1717 can you still reproduce this problem with recent builds? On the T430s or another computer? |
I found this issue while trying to get the solvespace build from nixpkgs running on arch with intel graphics (do not recommend). The most helpful debug output i got was through running with This suggests that the driver needs to be included in the snap; it looks like it currently isn't, but that doesn't really explain why it works for anyone: https://snapcraft.io/docs/gpu-support Apologies if that isn't useful or relevant info... |
@tomjnixon That's great info; thank you for contributing. The snap itself does not include the drivers, that's true. However, the connected platform snap does (in the Intel case), or it gets mounted from the host into the snap namespace (in the Nvidia case). I suspect this bug has been fixed by our recent move to the core20 base in 34efb6d. |
I'm not directly affected by this bug, but I did encounter this EXACT same error on my own custom made application.... using gstreamer library. I just thought I would help people look in the right direction. Hint: Somewhere along gstreamer and its dependencies by my understanding. I'll share my findings if I manage to resolve my problem. |
I receive this error off and on. I am running Linux Mint, using the current stable flatpak version of Solvespace. Unfortunately, I haven't figured out the cause or the definitive fix. I had just attempted to run Solvespace which wouldn't run. I had previously upgraded my OS (Linux Mint 21.0 > 21.1) and installed what I thought were all the available updates, and tried to run solvespace again which still didn't work. I noticed that Mint was notifying me of new updates. I installed them and then tried running solvespace again (not expecting it to work) but it actually worked and loaded fine. This photo shows the updates that were installed. When I previously had this problem, I may have fixed it in a similar way. Something to note - the nvidia driver I am running is 515.86.01, and the update history shows that I just installed nvidia 515.86.01 "flatpak runtime". I'm not very familiar with the details of how snap/flatpak works, but in my case, I am leaning towards a problem with graphics drivers being mismatched between the OS and the flatpak "runtime". Its possible that some have updated their graphics drivers, but a flatpak runtime isn't available yet for the new driver version. I also happened to take and save a glxinfo before and after the update (system glxinfo) but it doesn't show any differences from before and after. It's available if wanted. I hope this helps. |
In order to prevent this prehistoric laptop with a GeForce GT 330M from burning down while watching YT vids in the summer heat I thought it would be a good idea to install the Nvidia driver to get the HW acceleration/decoder running. Now getting the latest driver would be easy... if Nvidia hadn't dropped the support for the relic. Guess now what greeted me today when I started up solvespace to design a small adapter plate. Is there a possible fall back to software rendering? |
System information
SolveSpace version: 3.0~d3da2c8b
Operating system: Ubuntu 20.04
Expected behavior
Solvespace should just open and work!Actual behavior
Solvespace has an error : No available configuration for the given RGBA pixel format. (see picture)Additional information
The text was updated successfully, but these errors were encountered: