Add support for vector types and vector constants.

tests_lit/llvm2ice_tests/vector-pool.ll

+; This file tests support for vector constant pooling.
+
+; The order in which vector constants appear in the output is unspecified.
+; Hence, we check for the presence of only one constant per run line.
+;
+; RUN: %llvm2ice -O2 --verbose none %s | FileCheck %s --check-prefix=I16
+; RUN: %llvm2ice -Om1 --verbose none %s | FileCheck %s --check-prefix=I16
+; RUN: %llvm2ice -O2 --verbose none %s | FileCheck %s --check-prefix=FLOAT
+; RUN: %llvm2ice -Om1 --verbose none %s | FileCheck %s --check-prefix=FLOAT
+; RUN: %llvm2ice -O2 --verbose none %s | FileCheck %s --check-prefix=I1
+; RUN: %llvm2ice -Om1 --verbose none %s | FileCheck %s --check-prefix=I1
+;
+; RUN: %llvm2ice --verbose none %s | FileCheck --check-prefix=ERRORS %s
+; RUN: %llvm2iceinsts %s | %szdiff %s | FileCheck --check-prefix=DUMP %s
+
+define <8 x i16> @test_v8i16() {
+entry:
+  ret <8 x i16> <i16 0, i16 1, i16 2, i16 3, i16 4, i16 5, i16 6, i16 7>
+}
+
+; I16:      .long 0x10000
+; I16-NEXT: .long 0x30002
+; I16-NEXT: .long 0x50004
+; I16-NEXT: .long 0x70006
+
+define <4 x float> @test_v4f32() {
+entry:
+  ret <4 x float> <float 4.000000e+00, float 3.000000e+00, float 2.000000e+00, float 1.000000e+00>
+}
+
+; FLOAT:      .long 0x40800000
+; FLOAT-NEXT: .long 0x40400000
+; FLOAT-NEXT: .long 0x40000000
+; FLOAT-NEXT: .long 0x3f800000
+
+define <8 x i1> @test_v4i1() {
+entry:
+  ret <8 x i1> <i1 0, i1 1, i1 0, i1 1, i1 1, i1 1, i1 1, i1 1>
+}
+
+; I1:      .short 0x0
+; I1-NEXT: .short 0x1
+; I1-NEXT: .short 0x0
+; I1-NEXT: .short 0x1
+; I1-NEXT: .short 0x1
+; I1-NEXT: .short 0x1
+; I1-NEXT: .short 0x1
+; I1-NEXT: .short 0x1
+
+; ERRORS-NOT: ICE translation error
+; DUMP-NOT: SZ