[llvm-bugs] [Bug 17686] static_casting to a large unsigned long long from a double causes an FPE

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=17686

Vlad Tsyrklevich <vlad at tsyrklevich.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |vlad at tsyrklevich.net
         Resolution|FIXED                       |---
             Status|RESOLVED                    |REOPENED

--- Comment #11 from Vlad Tsyrklevich <vlad at tsyrklevich.net> ---
The reproducer in https://bugs.llvm.org/show_bug.cgi?id=30885 still continues
to reproduce. We're generating code that can cause an FPU exception in Jay's
example and his code is totally legal. It seems like Simon's patch fixed the
behavior when it was *profitable* but when it's not profitable it's still not
correct. e.g. Jay's example works correctly with -march=haswell but not
without. However, I'm not familiar enough with the codegen to say if
X86TargetLowering::shouldUseStrictFP_TO_INT() should always return true or what
the correct fix is.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20190107/bc85cd15/attachment.html>