Andrew Troschinetz wrote: > On Jul 14, 2009, at 4:44 PM, Andrew Troschinetz wrote: > >> Testing with GCC 3.5.x to GCC 4.x has shown that this issue doesn't >> surface unless you attempt some mathematical operation on the float >> containing a byteswaped value. In other words: > > That was me talking about results on my x86_64 machine, I went to my > i686 machine instead of my x86_64 machine and tried this code: > http://codepad.org/OZhEz0az The x86_64 has a very different floating-point architecture from the x86. > If I use GCC 2.95, or GCC 3.4.6 to compile the code and run it, the > problem presents itself. But if I use GCC 4.1.2 the code compiles and > runs just fine. That's a mystery to me. Any attempt to load an sNaN into a floating- point register should generate this behaviour. gcc 4.4.0 "fails" in the same way. > Unfortunately we're in a position where we need to support at least GCC > 3.4.6 on i686 hardware, so this problem affects us. My co-worker has, > predictably, expressed dismay that returning a float as a byte-sequence, > accommodating unsigned int type, or a user defined swapped type is > unnecessarily complicated. And that if the test fails it is because of > some kind of misconfiguration or compiler bug. He actually suggested a > comment in the header file that says "this may not work on floating > point types" as a good resolution to this issue. D: Can't he find a task more suited to his skills? Farming, perhaps. > On Jul 15, 2009, at 12:20 PM, Andrew Haley wrote: > >> Well, it looks like our lovely (?) hardware is indeed silently >> converting a signaling NaN to a quiet NaN, as I guessed. Thanks for >> the test case; it's always nice to know the real reason for things. > > Doesn't this new information suggest that the problem is in the > compiler, and not the hardware? No. > And might there a way perhaps to force > GCC to generate code that will not cause this problem to arise? Maybe > something like --no-nan-quietification? No. An attempt to load a signaling NaN from memory into the FPU generates an invalid-operation exception. If that exception is masked, the FPU returns a quiet NaN. [1] Andrew. [1] 8.5.1.2. INVALID ARITHMETIC OPERAND EXCEPTION (#IA), IA-32 Intel Architecture Software Developer’s Manual, Basic Architecture, Order Number 245470, 2001.
