<div dir="ltr"><div class="gmail_extra">The problem is we "guess" that the host architecture is i686, and fail when we find out that we can't target it (__i686__ is not defined).</div><div class="gmail_extra">This guess originates from get_host_triple() call in cmake/modules/GetHostTriple.cmake.</div><div class="gmail_extra">Probably, it makes sense to fix that to return i568 on openSUSE, and then test for i586 in the same way we already test for i386/i686.</div><div class="gmail_extra"><br></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Jul 17, 2015 at 11:23 AM, Hans Wennborg <span dir="ltr"><<a href="mailto:hans@chromium.org" target="_blank">hans@chromium.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Seems on OpenSUSE x86, it's called i586, not i686 :-(<br>
<br>
+Alexey: do you think we can handle this in the compiler-rt cmake<br>
files somehow? Maybe try targeting both i686 and i586 unless that<br>
would break something else?<br>
<span class="im"><br>
On Fri, Jul 17, 2015 at 1:31 AM, Nikola Smiljanic <<a href="mailto:popizdeh@gmail.com">popizdeh@gmail.com</a>> wrote:<br>
> CMake Error at projects/compiler-rt/cmake/config-ix.cmake:125 (message):<br>
> Cannot compile for i686<br>
><br>
> CMakeError.log attached, seems like #include checks are failing? This is x86<br>
> openSUSE.<br>
><br>
</span><span class="im">> On Fri, Jul 17, 2015 at 9:14 AM, Hans Wennborg <<a href="mailto:hans@chromium.org">hans@chromium.org</a>> wrote:<br>
>><br>
>> Hi Jack,<br>
>><br>
>> On Thu, Jul 16, 2015 at 4:03 PM, Jack Howarth<br>
</span><div class=""><div class="h5">>> <<a href="mailto:howarth.mailing.lists@gmail.com">howarth.mailing.lists@gmail.com</a>> wrote:<br>
>> > Hans,<br>
>> > Do we intend to leave -fopenmp defaulted to the no-op libgomp<br>
>> > support for 3.7.0 or do the sensible thing by applying...<br>
>> ><br>
>> > Index: CMakeLists.txt<br>
>> > ===================================================================<br>
>> > --- CMakeLists.txt (revision 242425)<br>
>> > +++ CMakeLists.txt (working copy)<br>
>> > @@ -182,7 +182,7 @@ set(GCC_INSTALL_PREFIX "" CACHE PATH "Di<br>
>> > set(DEFAULT_SYSROOT "" CACHE PATH<br>
>> > "Default <path> to all compiler invocations for --sysroot=<path>." )<br>
>> ><br>
>> > -set(CLANG_DEFAULT_OPENMP_RUNTIME "libgomp" CACHE STRING<br>
>> > +set(CLANG_DEFAULT_OPENMP_RUNTIME "libomp" CACHE STRING<br>
>> > "Default OpenMP runtime used by -fopenmp.")<br>
>> ><br>
>> > set(CLANG_VENDOR "" CACHE STRING<br>
>> ><br>
>> > so that the new llvm openmp library (which passes a three stage<br>
>> > bootstrap as part of an<br>
>> > in-tree build of llvm/cfe/compiler-rt/polly/libc++/openmp on<br>
>> > x86_64-apple-darwin).<br>
>><br>
>> I'm not fully aware of the implications of this, but if we do want to<br>
>> change it, it needs to be done on trunk first.<br>
>><br>
>> If you get it through review and committed to trunk, I'm open to merging<br>
>> it.<br>
>><br>
>> I assume utils/release/test-release.sh would also need an update so<br>
>> the library gets built?<br>
>><br>
>> Thanks,<br>
>> Hans<br>
><br>
><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature"><div dir="ltr">Alexey Samsonov<br><a href="mailto:vonosmas@gmail.com" target="_blank">vonosmas@gmail.com</a></div></div>
</div></div>