Strange Error. Have you seen this before?

Quote:

---

Originally Posted by Philip Nicoletti

one problem:

in class eulerian ... you set the zInMeters vector to all zeros, but
in shear() you take the log of each element in the vector.

---

That is just an artifact of my trials to figure out the problem. You are right that it will cause a problem if I leave it to all zeros. I have changed that now.
Thanks.

Quote:

---

Originally Posted by Martin O

I could be mistaken, but shouldn't you make your %.o be dependent on %.cpp and %.h? Its been a while since I played with makefiles so I could be wrong.

---

Even I am not very familiar with make files. I just took a tutorial online to build it.
Anyways, thanks for pointing that out. I will take a look at it.
Thanks

I copied your code and compiled it with microsoft visual studio 6.0 and it ran
ok. (last output statement was time: 9999). Apart from rearranging the declaration of some integers (VS6.0 is non-standard), i had to comment out the include file (omp.h) in advectpar.cpp. It does seem to not need this file.
Maybe there is something in there that is going wrong.

Quote:

---

Originally Posted by innovaltec

I copied your code and compiled it with microsoft visual studio 6.0 and it ran
ok. (last output statement was time: 9999). Apart from rearranging the declaration of some integers (VS6.0 is non-standard), i had to comment out the include file (omp.h) in advectpar.cpp. It does seem to not need this file.
Maybe there is something in there that is going wrong.

---

Thanks for taking time to run my code. I think it runs fine when all the files are freshly compiled. For example (as one of the poster suggested) , if I remove all the object files and recompile, I get no errors at all. If I just change in one file (for example add a new variable in a header file), I get the error (sometimes).

OMP.H file is for OpenMP parallel programming. I was just playing with that.

Thanks again.

Can anybody suggest anything else as there seems to be a problem in my code which i m unable to detect.

Thanks guys.

Quote:

---

Originally Posted by pipa

Can anybody suggest anything else as there seems to be a problem in my code which i m unable to detect.

Thanks guys.

---

Maybe your make utility is messed up, and the problem has nothing to do with your code.

You stated that if you recompile everything, the problem goes away. So is your make utility recompiling all of the modules that use the header?

Regards,

Paul McKenzie

Quote:

---

Originally Posted by pipa

I think it runs fine when all the files are freshly compiled. For example (as one of the poster suggested) , if I remove all the object files and recompile, I get no errors at all. If I just change in one file (for example add a new variable in a header file), I get the error (sometimes).

---

Your makefile doesn't appear to be using header files as dependencies. So, if you change a header and just re-run make, it isn't going to detect any changes and nothing is going to be re-built.

make only checks dependencies for changes, it doesn't know to check the header files that are included inside the cpp / c files.

Quote:

---

Originally Posted by Martin O

I could be mistaken, but shouldn't you make your %.o be dependent on %.cpp and %.h? Its been a while since I played with makefiles so I could be wrong.

---

+1.

Quote:

---

Originally Posted by pm_kirkham

+1.

---

Can you please explain that in little detail? I didn't understand ur post.
Thanks