[LLVMbugs] [Bug 22917] fatal error in backend for g++.dg/other/pr59492.C

[bugzilla-daemon at llvm.org]

https://llvm.org/bugs/show_bug.cgi?id=22917

Reid Kleckner <rnk at google.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
                 CC|                            |rnk at google.com
         Resolution|---                         |INVALID

--- Comment #3 from Reid Kleckner <rnk at google.com> ---
The __builtin_* intrinsics aren't really the public user interface for this
functionality. The intrinsics documented by Intel provided by the *mmintrin.h
headers are the public facing interface. If you include immintrin.h and use
_rdrand32_step() without -march=ivybridge or better, you'll get an undeclared
id error from the frontend. This is a bad diagnostic and we could do better,
but this is a much broader issue that affects all conditionally supported
intrinsics provided by immintrin.h.

Also, passing 100% of the GCC test suite is not a goal of the Clang project.
People have looked into passing it before, and the consensus is that the
remaining failures are not worth fixing because they don't help us compile more
real world code.

-- 
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/20150316/070c0612/attachment.html>