a65ccc1b9f
On processors supporting vector registers and SIMD instructions, enable i128 as legal type in VRs. This allows many operations to be implemented via native instructions directly in VRs (including add, subtract, logical operations and shifts). For a few other operations (e.g. multiply and divide, as well as atomic operations), we need to move the i128 value back to a GPR pair to use the corresponding instruction there. Overall, this is still beneficial. The patch includes the following LLVM changes: - Enable i128 as legal type - Set up legal operations (in SystemZInstrVector.td) - Custom expansion for i128 add/subtract with carry - Custom expansion for i128 comparisons and selects - Support for moving i128 to/from GPR pairs when required - Handle 128-bit integer constant values everywhere - Use i128 as intrinsic operand type where appropriate - Updated and new test cases In addition, clang builtins are updated to reflect the intrinsic operand type changes (which also improves compatibility with GCC).
225 lines
6.8 KiB
LLVM
225 lines
6.8 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
|
|
; Test high-part i64->i128 multiplications on z13.
|
|
;
|
|
; RUN: llc < %s -mtriple=s390x-linux-gnu -mcpu=z13 | FileCheck %s
|
|
|
|
; Check zero-extended multiplication in which only the high part is used.
|
|
define i64 @f1(i64 %dummy, i64 %a, i64 %b) {
|
|
; CHECK-LABEL: f1:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlgr %r2, %r4
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check sign-extended multiplication in which only the high part is used.
|
|
; This needs a rather convoluted sequence.
|
|
define i64 @f2(i64 %dummy, i64 %a, i64 %b) {
|
|
; CHECK-LABEL: f2:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: srag %r1, %r4, 63
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: srag %r0, %r3, 63
|
|
; CHECK-NEXT: ngr %r1, %r3
|
|
; CHECK-NEXT: mlgr %r2, %r4
|
|
; CHECK-NEXT: ngr %r0, %r4
|
|
; CHECK-NEXT: agr %r0, %r1
|
|
; CHECK-NEXT: sgr %r2, %r0
|
|
; CHECK-NEXT: br %r14
|
|
%ax = sext i64 %a to i128
|
|
%bx = sext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check zero-extended multiplication in which only part of the high half
|
|
; is used.
|
|
define i64 @f3(i64 %dummy, i64 %a, i64 %b) {
|
|
; CHECK-LABEL: f3:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlgr %r2, %r4
|
|
; CHECK-NEXT: srlg %r2, %r2, 3
|
|
; CHECK-NEXT: br %r14
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 67
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check zero-extended multiplication in which the result is split into
|
|
; high and low halves.
|
|
define i64 @f4(i64 %dummy, i64 %a, i64 %b) {
|
|
; CHECK-LABEL: f4:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlgr %r2, %r4
|
|
; CHECK-NEXT: ogr %r2, %r3
|
|
; CHECK-NEXT: br %r14
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
%low = trunc i128 %mulx to i64
|
|
%or = or i64 %high, %low
|
|
ret i64 %or
|
|
}
|
|
|
|
; Check division by a constant, which should use multiplication instead.
|
|
define i64 @f5(i64 %dummy, i64 %a) {
|
|
; CHECK-LABEL: f5:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: llihf %r0, 1782028570
|
|
; CHECK-NEXT: oilf %r0, 598650223
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlgr %r2, %r0
|
|
; CHECK-NEXT: srlg %r2, %r2, 9
|
|
; CHECK-NEXT: br %r14
|
|
%res = udiv i64 %a, 1234
|
|
ret i64 %res
|
|
}
|
|
|
|
; Check MLG with no displacement.
|
|
define i64 @f6(i64 %dummy, i64 %a, ptr %src) {
|
|
; CHECK-LABEL: f6:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, 0(%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%b = load i64, ptr %src
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check the high end of the aligned MLG range.
|
|
define i64 @f7(i64 %dummy, i64 %a, ptr %src) {
|
|
; CHECK-LABEL: f7:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, 524280(%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%ptr = getelementptr i64, ptr %src, i64 65535
|
|
%b = load i64, ptr %ptr
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check the next doubleword up, which requires separate address logic.
|
|
; Other sequences besides this one would be OK.
|
|
define i64 @f8(i64 %dummy, i64 %a, ptr %src) {
|
|
; CHECK-LABEL: f8:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: agfi %r4, 524288
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, 0(%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%ptr = getelementptr i64, ptr %src, i64 65536
|
|
%b = load i64, ptr %ptr
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check the high end of the negative aligned MLG range.
|
|
define i64 @f9(i64 %dummy, i64 %a, ptr %src) {
|
|
; CHECK-LABEL: f9:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, -8(%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%ptr = getelementptr i64, ptr %src, i64 -1
|
|
%b = load i64, ptr %ptr
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check the low end of the MLG range.
|
|
define i64 @f10(i64 %dummy, i64 %a, ptr %src) {
|
|
; CHECK-LABEL: f10:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, -524288(%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%ptr = getelementptr i64, ptr %src, i64 -65536
|
|
%b = load i64, ptr %ptr
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check the next doubleword down, which needs separate address logic.
|
|
; Other sequences besides this one would be OK.
|
|
define i64 @f11(ptr %dest, i64 %a, ptr %src) {
|
|
; CHECK-LABEL: f11:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: agfi %r4, -524296
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, 0(%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%ptr = getelementptr i64, ptr %src, i64 -65537
|
|
%b = load i64, ptr %ptr
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|
|
; Check that MLG allows an index.
|
|
define i64 @f12(ptr %dest, i64 %a, i64 %src, i64 %index) {
|
|
; CHECK-LABEL: f12:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: # kill: def $r3d killed $r3d def $r2q
|
|
; CHECK-NEXT: mlg %r2, 524287(%r5,%r4)
|
|
; CHECK-NEXT: # kill: def $r2d killed $r2d killed $r2q
|
|
; CHECK-NEXT: br %r14
|
|
%add1 = add i64 %src, %index
|
|
%add2 = add i64 %add1, 524287
|
|
%ptr = inttoptr i64 %add2 to ptr
|
|
%b = load i64, ptr %ptr
|
|
%ax = zext i64 %a to i128
|
|
%bx = zext i64 %b to i128
|
|
%mulx = mul i128 %ax, %bx
|
|
%highx = lshr i128 %mulx, 64
|
|
%high = trunc i128 %highx to i64
|
|
ret i64 %high
|
|
}
|
|
|