So I've tried running a C++ code with with the cachegrind tool and it's showing way more refs than when I ran last time which was 2 years ago. Of course I ran it multiple times on both occasions with similar results each time.I'm using the same hardware environment as before.
Now the weird thing is it's showing approximately the same amount of cache misses, but with the big change in refs the overall percentage of misses is completely different. This is not desirable for me because the purpose of the runs are to show the inefficiency of linked lists with regards to the cache. On the old run the results were, as expected, a high percentage of cache misses.
Of course I tried it with multiple versions of valgrind including the one I used back then. The results were, as they should be, the same.
Anyone have an explanation for this? Thank you in advance.