llvm/test/CodeGen/PowerPC/vrspill.ll
Bill Schmidt 11729224bf [PowerPC] Disable fast-isel for existing -O0 tests for PowerPC.
This is a preliminary patch for fast instruction selection on
PowerPC.  Code generation can differ between DAG isel and fast isel.
Existing tests that specify -O0 were written to expect DAG isel.  Make
this explicit by adding -fast-isel=false to the tests.

In some cases specifying -fast-isel=false produces different code even
when there isn't a fast instruction selector specified.  This is
because TM.Options.EnableFastISel = 1 at -O0 whether or not a FastISel
object exists.  Thus disabling fast isel can actually produce less
conservative code.  Because of this, some of the expected code
generation in the -O0 tests needs to be adjusted.

In particular, handling of function arguments is less conservative
with -fast-isel=false (see isOnlyUsedInEntryBlock() in
SelectionDAGBuilder.cpp).  This results in fewer stack accesses and,
in some cases, reduced stack size as uselessly loaded values are no
longer stored back to spill locations in the stack.

No functional change with this patch; test case adjustments only.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183939 91177308-0d34-0410-b5e6-96231b3b80d8
2013-06-13 20:23:34 +00:00

19 lines
631 B
LLVM

; RUN: llc -O0 -mtriple=powerpc-unknown-linux-gnu -mattr=+altivec -verify-machineinstrs < %s | FileCheck %s
; RUN: llc -O0 -mtriple=powerpc64-unknown-linux-gnu -mattr=+altivec -verify-machineinstrs -fast-isel=false < %s | FileCheck %s
; This verifies that we generate correct spill/reload code for vector regs.
define void @addrtaken(i32 %i, <4 x float> %w) nounwind {
entry:
%i.addr = alloca i32, align 4
%w.addr = alloca <4 x float>, align 16
store i32 %i, i32* %i.addr, align 4
store <4 x float> %w, <4 x float>* %w.addr, align 16
call void @foo(i32* %i.addr)
ret void
}
; CHECK: stvx 2,
declare void @foo(i32*)