summaryrefslogtreecommitdiffstats
path: root/llvm/lib
diff options
context:
space:
mode:
authorSander de Smalen <sander.desmalen@arm.com>2018-07-18 09:37:51 +0000
committerSander de Smalen <sander.desmalen@arm.com>2018-07-18 09:37:51 +0000
commitccdc7ebc1d99a3ccc34fe4a875b916bbb23b7298 (patch)
tree3f73a10d47100884ab090f976c4b163d5bf877fa /llvm/lib
parentc1090da8529d85e85ab8dbf141ebecd8926b00ab (diff)
downloadbcm5719-llvm-ccdc7ebc1d99a3ccc34fe4a875b916bbb23b7298.tar.gz
bcm5719-llvm-ccdc7ebc1d99a3ccc34fe4a875b916bbb23b7298.zip
[AArch64][SVE] Asm: Support for UDOT/SDOT instructions.
The signed/unsigned DOT instructions perform a dot-product on quadtuplets from two source vectors and accumulate the result in the destination register. The instructions come in two forms: Vector form, e.g. sdot z0.s, z1.b, z2.b - signed dot product on four 8-bit quad-tuplets, accumulating results in 32-bit elements. udot z0.d, z1.h, z2.h - unsigned dot product on four 16-bit quad-tuplets, accumulating results in 64-bit elements. Indexed form, e.g. sdot z0.s, z1.b, z2.b[3] - signed dot product on four 8-bit quad-tuplets with specified quadtuplet from second source vector, accumulating results in 32-bit elements. udot z0.d, z1.h, z2.h[1] - dot product on four 16-bit quad-tuplets with specified quadtuplet from second source vector, accumulating results in 64-bit elements. llvm-svn: 337372
Diffstat (limited to 'llvm/lib')
-rw-r--r--llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td6
-rw-r--r--llvm/lib/Target/AArch64/SVEInstrFormats.td66
2 files changed, 72 insertions, 0 deletions
diff --git a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
index b37c6967727..94f0be01b3c 100644
--- a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
+++ b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
@@ -72,6 +72,12 @@ let Predicates = [HasSVE] in {
defm SDIVR_ZPmZ : sve_int_bin_pred_arit_2_div<0b110, "sdivr">;
defm UDIVR_ZPmZ : sve_int_bin_pred_arit_2_div<0b111, "udivr">;
+ defm SDOT_ZZZ : sve_intx_dot<0b0, "sdot">;
+ defm UDOT_ZZZ : sve_intx_dot<0b1, "udot">;
+
+ defm SDOT_ZZZI : sve_intx_dot_by_indexed_elem<0b0, "sdot">;
+ defm UDOT_ZZZI : sve_intx_dot_by_indexed_elem<0b1, "udot">;
+
defm SXTB_ZPmZ : sve_int_un_pred_arit_0_h<0b000, "sxtb">;
defm UXTB_ZPmZ : sve_int_un_pred_arit_0_h<0b001, "uxtb">;
defm SXTH_ZPmZ : sve_int_un_pred_arit_0_w<0b010, "sxth">;
diff --git a/llvm/lib/Target/AArch64/SVEInstrFormats.td b/llvm/lib/Target/AArch64/SVEInstrFormats.td
index c759950daa7..4bcde3edbcf 100644
--- a/llvm/lib/Target/AArch64/SVEInstrFormats.td
+++ b/llvm/lib/Target/AArch64/SVEInstrFormats.td
@@ -1458,6 +1458,72 @@ multiclass sve_int_mlas_vvv_pred<bits<1> opc, string asm> {
}
//===----------------------------------------------------------------------===//
+// SVE Integer Dot Product Group
+//===----------------------------------------------------------------------===//
+
+class sve_intx_dot<bit sz, bit U, string asm, ZPRRegOp zprty1,
+ ZPRRegOp zprty2>
+: I<(outs zprty1:$Zda), (ins zprty1:$_Zda, zprty2:$Zn, zprty2:$Zm), asm,
+ "\t$Zda, $Zn, $Zm", "", []>, Sched<[]> {
+ bits<5> Zda;
+ bits<5> Zn;
+ bits<5> Zm;
+ let Inst{31-23} = 0b010001001;
+ let Inst{22} = sz;
+ let Inst{21} = 0;
+ let Inst{20-16} = Zm;
+ let Inst{15-11} = 0;
+ let Inst{10} = U;
+ let Inst{9-5} = Zn;
+ let Inst{4-0} = Zda;
+
+ let Constraints = "$Zda = $_Zda";
+}
+
+multiclass sve_intx_dot<bit opc, string asm> {
+ def _S : sve_intx_dot<0b0, opc, asm, ZPR32, ZPR8>;
+ def _D : sve_intx_dot<0b1, opc, asm, ZPR64, ZPR16>;
+}
+
+//===----------------------------------------------------------------------===//
+// SVE Integer Dot Product Group - Indexed Group
+//===----------------------------------------------------------------------===//
+
+class sve_intx_dot_by_indexed_elem<bit sz, bit U, string asm,
+ ZPRRegOp zprty1, ZPRRegOp zprty2,
+ ZPRRegOp zprty3, Operand itype>
+: I<(outs zprty1:$Zda), (ins zprty1:$_Zda, zprty2:$Zn, zprty3:$Zm, itype:$iop),
+ asm, "\t$Zda, $Zn, $Zm$iop",
+ "", []>, Sched<[]> {
+ bits<5> Zda;
+ bits<5> Zn;
+ let Inst{31-23} = 0b010001001;
+ let Inst{22} = sz;
+ let Inst{21} = 0b1;
+ let Inst{15-11} = 0;
+ let Inst{10} = U;
+ let Inst{9-5} = Zn;
+ let Inst{4-0} = Zda;
+
+ let Constraints = "$Zda = $_Zda";
+}
+
+multiclass sve_intx_dot_by_indexed_elem<bit opc, string asm> {
+ def _S : sve_intx_dot_by_indexed_elem<0b0, opc, asm, ZPR32, ZPR8, ZPR3b8, VectorIndexS> {
+ bits<2> iop;
+ bits<3> Zm;
+ let Inst{20-19} = iop;
+ let Inst{18-16} = Zm;
+ }
+ def _D : sve_intx_dot_by_indexed_elem<0b1, opc, asm, ZPR64, ZPR16, ZPR4b16, VectorIndexD> {
+ bits<1> iop;
+ bits<4> Zm;
+ let Inst{20} = iop;
+ let Inst{19-16} = Zm;
+ }
+}
+
+//===----------------------------------------------------------------------===//
// SVE Integer Arithmetic - Unary Predicated Group
//===----------------------------------------------------------------------===//
OpenPOWER on IntegriCloud