-
-
Notifications
You must be signed in to change notification settings - Fork 283
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
Examples all end with SIGSEGV #512
Comments
Also (hopefully related) I'm seeing lots of failures in
Should I be expecting these tests to pass? |
Maybe try a debug build ( |
Results:
I also ran with
|
This ran valgrind on bash only. You need |
Ah OK, new output here: https://pastebin.com/nwwf7Dnd |
points to this line in
Somehow Indices in BLAS are 1-based and |
I'm trying to compile with MKL blas, but I don't think it is working. I've downloaded the MKL libs and can see the
as per the ipopt instructions. But in the configure output I see:
Does this mean it didn't find MKL and is just using the netlib/openblas versions? (Tests still segfault as before) |
There seems to be something wrong in configure that it didn't pick up your flags. I'll fix. But in the meanwhile, change Figuring out what the issue could be with your |
with I get: and also: but now all the tests fail with:
but it is present...
|
But it is not in the library search path. Add MKLROOT to you LD_LIBRARY_PATH. |
I see! OK that gets me back to some tests failing with segfaults still. Full vagrind debug output here: https://pastebin.com/zvtJ7g4D |
And you have also rebuild HSL to use MKL? |
That fixed it! Happy days :)
So I guess there is a problem with openblas from the fedora 34 repos. I saw a few tickets on it (eg, OpenMathLib/OpenBLAS#2839) though I can't tell if they are related. But at least the issues seem fixed with MKL. Thanks for all your help and patience! |
Interesting development. I recompiled without Recompiling with the |
Did you install before building the examples? |
Yes, I think it fails otherwise..? |
Or the examples are build against some older Ipopt installation. Could also be that there are now both (broken) shared and (working) static libs installed, but the examples link against the shared libs. |
Hi,
I'm trying to test my installation and check the different solvers are compiled correctly, but every one exits with
Segmentation fault (core dumped)
.This is an example log from
journalctl
:Some examples show the licence and the selected solver. In
ScalableProblems
I see theNumber of nonzeros in....
lines and larger problems seem to take longer at least, but I can't tell if they are completing and just not exiting properly or if they are actually failing.The text was updated successfully, but these errors were encountered: