Valgrind

[Vasily Postnicov]

Let me correct my answer a bit. You calculate 'which' as a sum of floating
point comparisons with some "weight" from 1 to 8, but this is ok, as a
result of comparison is 0 or 1. But, there might be a problem. The greatest
possible 'which' is the sum of "weights" 1+2+3+4+8+12=30. (Assuming all
conditionals return 1). Your ctrltab array has only 16 elements. So if one
floating point conditional wrongly returns 1 (due to lack of precision) you
will get out-of-bounds error. I think this is the most probable scenario
2 февр. 2016 г. 22:34 пользователь "Pierre Abbat" <
phma at leaf.dragonflybsd.org> написал:

> On 02/ 2/16 06:22 PM, Vasily Postnicov wrote:
>
>> Do you mean an uninitialized pointer problem? Do you have a source code
>> of that software that you think causes a trouble? If you think your
>> memory is corrupted because of uninitialized pointer, I suggest to
>> compile problematic code with clang with AddressSanitizer. Just don't
>> forget to compile llvm with compiler-rt option. I hope this helps.
>> Didn't try this by myself, though
>>
>
> It's on Github. Follow the links from http://bezitopo.org/. It's not an
> uninitialized pointer, it's a double variable, which is a control point of
> a Bézier triangle. The function is returning something like 1e-300 instead
> of 280. I inserted a debugging statement to output "ctrlpt garbage" if it
> ever returned such a number. On Linux, inserting the statement made the bug
> go away, but it reappeared on DragonFly.
>
> I ran bezitest under Valgrind a few weeks ago and fixed some uninitialized
> variables and allocation mixups.
>
> Pierre
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dragonflybsd.org/pipermail/users/attachments/20160202/b49fd5c2/attachment-0007.html>