https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81348
Michael Meissner <meissner at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |ASSIGNED Last reconfirmed| |2017-07-07 CC| |dje at gcc dot gnu.org Assignee|unassigned at gcc dot gnu.org |meissner at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #1 from Michael Meissner <meissner at gcc dot gnu.org> --- It is a typo in the power9 splitter for loading up a short value to an Altivec register, sign extending it in preparation for converting it to floating point. If you look at rs6000.md around line 943, the splitter uses operands[1] (which is a memory address), when it should use operands[0].