-
Notifications
You must be signed in to change notification settings - Fork 492
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
OpenCPN segfaults on startup #1527
Comments
Sorry, my fault: I ran the executable directly from the build dir, without |
Now, to be able to cope with this we need more information. As a starter: could you run the program in gdb, let it crash, make a stacktrace using the |
Starting program: /usr/local/src/OpenCPN/build/opencpn Thread 1 "opencpn" received signal SIGSEGV, Segmentation fault. #0 0x0000555555a371de in Piano::BuildGLTexture() (this=0x55555643f000) at /usr/local/src/OpenCPN/src/piano.cpp:324 |
OK, looks like some graphics stuff. I'm not in this business, and cannot help. Have you done a git bisect before? That would probably be extremely useful here. |
Sorry, I don't have much time to debug this right now. |
hm, again (I should do something else!) Line piano.cpp line 322-324 reads
What's wrong with this? Corrupted stack? |
@bdbcat: This is far above my paygrade... I can reproduce the same crash on Fedora 30. Git bisect result:
Bisect log:
|
This is pretty bad, blocks all testing on current master... |
Please let me know, I'm available to rebuild once presumably fixed. |
I pushed a possible fix, just simplifying the needlessly obtuse code. The failure must be related to compiler optimization? |
Rebuilt, still segfaults |
|
|
One more commit, fixing this time? |
Definitely I have no luck (how do I miss a deb, even more!).
|
pcav... |
Mixed repos==trouble :) |
The official debian package is accepted, but still at 4.8.8. I'm about to update to 5.0.0, will look into it later. |
I can now compile and run on Fedora 30 without the previous crash. Instead, here is an assert message box at startup:
Annoying, but not blocking. Using a more elaborated cmake invocation:
and it looks OK (that is, starts without crash or assert messages.) |
I missed this, thanks a lot! |
Sorry to hijack the thread: AFAICT there is no make uninstall, am I wrong? |
No, you are not. There should be.... |
Thanks. For the record, I opened ticket #1531 not to forget. |
Great. Does this mean that you can now run this without crash i. e., that this bug could be closed? |
Yes, thanks! |
A fresh install on a Debian Sid box apparently went fine, but the program refuses to start:
The text was updated successfully, but these errors were encountered: