I originally assumed this is the same bug mentioned here
however after some regression testing it turned out that this issue is new to 2010 and cannot be reproduced in 2005 or 2008, while 737176 problem can be traced back at least to 2005, so this seems to be a different, even though possible related issue.
We've recently transitioned to 2010 from 2005 and this problem quickly became a blocker for us: with over 1M codeline base we cannot practically go through all our loop code and disable optimization there with #pragma optimize(). On the other hand, disabling optimization across the product has quite severe performance repercussions, with up to 100% performance hit, since many operations in our product are highly computational intensive. Would you have any other suggestions as to how to deal with this problem?
Thanks in advance!
Ilya Varaev & Michael Yakubovsky