On 2018-12-03, Stuart Henderson <s...@spacehopper.org> wrote: > # Defend against clang being used on x86-32 without SSE2 enabled. As current > # versions of clang do not understand -fexcess-precision=standard, the use of > # x87 floating point operations leads to problems like isinf possibly > returning > # false for a value that is infinite when converted from the 80bit register to > # the 8byte memory representation. > > I'm unsure which approach I like least, gcc or -msse2...
A third possibility would be to ignore this corner case and disable the check. I think from a practical point of view this would be fine. The number of people who do scientific computing in Postgres on i386 should be very limited. However, I imagine this would burn whatever goodwill we have with upstream. -- Christian "naddy" Weisgerber na...@mips.inka.de