[PATCH] D137946: [X86] Allow no X87 on 32-bit

[LuoYuanke via Phabricator via llvm-commits]

LuoYuanke added inline comments.


================
Comment at: llvm/lib/Target/X86/X86InstrFormats.td:479
 
 // FpI_ - Floating Point Pseudo Instruction template. Not Predicated.
 class FpI_<dag outs, dag ins, FPFormat fp, list<dag> pattern>
----------------
Remove "Not Predicated" of the comments?


================
Comment at: llvm/test/CodeGen/X86/no-ret-in-x87-reg.ll:140
+; NOX87-NEXT:    addl $24, %esp
+; NOX87-NEXT:    movzwl %cx, %ecx
 ; NOX87-NEXT:    retl
----------------
Is it because the last 16 bit value is stored in `cx`?


================
Comment at: llvm/test/CodeGen/X86/no-ret-in-x87-reg.ll:195
+entry:
+  %0 = sitofp i64 %a to x86_fp80
+  ret x86_fp80 %0
----------------
Add a case to convert from `x86_fp80` to `int`?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D137946/new/

https://reviews.llvm.org/D137946