Re: Test on Ubuntu 12.04 64bits

[David Chisnall]

On 27 May 2012, at 18:31, Fred Kiefer wrote:

> That was a very nice explanation, but in this specific case, I don't see how 
> floating point pointers come into play. Did you look at the actual code in 
> question? This is just an integer swapping macro that operates on the fields 
> in a time zone structure.

Ooops, I misread the code - I thought it was byte-swapping floats using a 
function that expected ints.  In this case, I think it's a spurious error, and 
doesn't appear with clang 3.1, so presumably a fixed spurious error...

David

-- Sent from my brain