GNU Scientific Library - Bugs: bug #59913, gsl 2.3.0 problem in...
You are not allowed to post comments on this tracker with your current authentication level.
bug #59913: gsl 2.3.0 problem in gsl_integration_cquad
Submitter: | Patrick Alken <psa> | ||
Submitted: | Wed 20 Jan 2021 02:41:50 AM UTC | ||
Category: | None | Severity: | 3 - Normal |
Operating System: | Status: | None | |
Assigned to: | None | Open/Closed: | Open |
Release: |
Attached Files
Depends on the following items: None found
Items that depend on this one: None found
Carbon-Copy List
Follows 1 latest change.
Date | Changed by | Updated Field | Previous Value | => | Replaced by |
---|---|---|---|---|---|
2021-01-20 | psa | Attached File | - | ![]() |
Added main.cpp, #50749 |
from buchholz =at= tugraz =dot= at
I found a problem in gsl (assuming the compiler/linker is not broken), which I would consider a bug.
A minimal working example is attached.
I compiled using g++ 6.3.0 and gsl 2.3.0. The compile command is given in the file.
The code uses an extension to cause the code to stop on floating point exceptions (found this via stack exchange https://linux.die.net/man/3/feenableexcept).
I would expect the code to just run through with output something like
result: 4.39016e-164
abserr: 0
nevals: 33
Instead i get the message
Gleitkomma-Ausnahme (german for floating point exception)
Running the program with gdb I get
Program received signal SIGFPE, Arithmetic exception.
0x00007ffff7a1b42f in gsl_integration_cquad () from /usr/lib/x86_64-linux-gnu/libgsl.so.19
(gdb) backtrace
#0 0x00007ffff7a1b42f in gsl_integration_cquad () from /usr/lib/x86_64-linux-gnu/libgsl.so.19
#1 0x0000555555554cee in main () at main.cpp:42
Changing the lower bound to 19.2 will cause the code to work, i.e. no floating point exception.
If I do not trap the exception the code works as expected, but if there is no real problem, then no exception should be raised.