llvm-project/llvm/test/CodeGen/PowerPC/2006-01-11-darwin-fp-argume...

; RUN: llc < %s | not grep {, f1}

target datalayout = "E-p:32:32"
target triple = "powerpc-apple-darwin8.2.0"

; Dead argument should reserve an FP register.
define double @bar(double %DEAD, double %X, double %Y) {
        %tmp.2 = fadd double %X, %Y              ; <double> [#uses=1]
        ret double %tmp.2
}
Eliminate more uses of llvm-as and llvm-dis. llvm-svn: 81293 2009-09-09 08:09:15 +08:00			`; RUN: llc < %s \| not grep {, f1}`
Testcase the PPC backend is miscompiling. llvm-svn: 25216 2006-01-12 02:15:37 +08:00
Byebye llvm-upgrade! llvm-svn: 48762 2008-03-25 12:26:08 +08:00			`target datalayout = "E-p:32:32"`
Testcase the PPC backend is miscompiling. llvm-svn: 25216 2006-01-12 02:15:37 +08:00			`target triple = "powerpc-apple-darwin8.2.0"`

			`; Dead argument should reserve an FP register.`
Byebye llvm-upgrade! llvm-svn: 48762 2008-03-25 12:26:08 +08:00			`define double @bar(double %DEAD, double %X, double %Y) {`
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897 2009-06-05 06:49:04 +08:00			`%tmp.2 = fadd double %X, %Y ; <double> [#uses=1]`
Testcase the PPC backend is miscompiling. llvm-svn: 25216 2006-01-12 02:15:37 +08:00			`ret double %tmp.2`
			`}`