Change output format of vget_lane, for better readability.

6 files changed, 40 insertions, 4 deletions

diff --git a/compute_ref.axf b/compute_ref.axf
index f195b9f..ccce337 100644
--- a/compute_ref.axf
+++ b/compute_ref.axf
diff --git a/compute_ref.gccarm b/compute_ref.gccarm
index 72d22a9..5384166 100755
--- a/compute_ref.gccarm
+++ b/compute_ref.gccarm
diff --git a/compute_ref.gccarm-rvct b/compute_ref.gccarm-rvct
index 5aa78f3..9500185 100644
--- a/compute_ref.gccarm-rvct
+++ b/compute_ref.gccarm-rvct
diff --git a/ref-rvct-neon.txt b/ref-rvct-neon.txt
index 8ef013c..93f1a76 100644
--- a/ref-rvct-neon.txt
+++ b/ref-rvct-neon.txt
@@ -748,7 +748,25 @@ VSET_LANE/VSET_LANEQ:16:result_uint64x2 [] = { fffffffffffffff0, 11,  }
 VSET_LANE/VSET_LANEQ:17:result_float32x4 [] = { c1800000 -0x1.0000000p+4 -16, c1700000 -0x1.e000000p+3 -15, c1600000 -0x1.c000000p+3 -14, 41333333 0x1.6666660p+3 11.2,  }
 
 VGET_LANE/VGETQ_LANE output:
-fffffff7, fffffff3, fffffff1, fffffffffffffff0, f6, fff2, fffffff1, fffffffffffffff0, c1700000, ffffffff, fffffff5, fffffff3, fffffffffffffff1, fe, fff6, fffffff2, fffffffffffffff1, c1500000, 
+vget_lane_s8: fffffff7
+vget_lane_s16: fffffff3
+vget_lane_s32: fffffff1
+vget_lane_s64: fffffffffffffff0
+vget_lane_u8: f6
+vget_lane_u16: fff2
+vget_lane_u32: fffffff1
+vget_lane_u64: fffffffffffffff0
+vget_lane_f32: c1700000
+vgetq_lane_s8: ffffffff
+vgetq_lane_s16: fffffff5
+vgetq_lane_s32: fffffff3
+vgetq_lane_s64: fffffffffffffff1
+vgetq_lane_u8: fe
+vgetq_lane_u16: fff6
+vgetq_lane_u32: fffffff2
+vgetq_lane_u64: fffffffffffffff1
+vgetq_lane_f32: c1500000
+
 
 VQSUB/VQSUBQ overflow output:
 VQSUB/VQSUBQ:0:vqsub_s8 Neon overflow 0
diff --git a/ref-rvct.txt b/ref-rvct.txt
index 6aafe88..4855998 100644
--- a/ref-rvct.txt
+++ b/ref-rvct.txt
@@ -748,7 +748,25 @@ VSET_LANE/VSET_LANEQ:16:result_uint64x2 [] = { fffffffffffffff0, 11,  }
 VSET_LANE/VSET_LANEQ:17:result_float32x4 [] = { c1800000 -0x1.0000000p+4 -16, c1700000 -0x1.e000000p+3 -15, c1600000 -0x1.c000000p+3 -14, 41333333 0x1.6666660p+3 11.2,  }
 
 VGET_LANE/VGETQ_LANE output:
-fffffff7, fffffff3, fffffff1, fffffffffffffff0, f6, fff2, fffffff1, fffffffffffffff0, c1700000, ffffffff, fffffff5, fffffff3, fffffffffffffff1, fe, fff6, fffffff2, fffffffffffffff1, c1500000, 
+vget_lane_s8: fffffff7
+vget_lane_s16: fffffff3
+vget_lane_s32: fffffff1
+vget_lane_s64: fffffffffffffff0
+vget_lane_u8: f6
+vget_lane_u16: fff2
+vget_lane_u32: fffffff1
+vget_lane_u64: fffffffffffffff0
+vget_lane_f32: c1700000
+vgetq_lane_s8: ffffffff
+vgetq_lane_s16: fffffff5
+vgetq_lane_s32: fffffff3
+vgetq_lane_s64: fffffffffffffff1
+vgetq_lane_u8: fe
+vgetq_lane_u16: fff6
+vgetq_lane_u32: fffffff2
+vgetq_lane_u64: fffffffffffffff1
+vgetq_lane_f32: c1500000
+
 
 VQSUB/VQSUBQ overflow output:
 VQSUB/VQSUBQ:0:vqsub_s8 Neon overflow 0
diff --git a/ref_vget_lane.c b/ref_vget_lane.c
index a1ab6a9..9743529 100644
--- a/ref_vget_lane.c
+++ b/ref_vget_lane.c
@@ -37,7 +37,7 @@ void exec_vget_lane (void)
 #define TEST_VGET_LANE(Q, T1, T2, W, N, L)				\
   VAR(var, T1, W) = vget##Q##_lane_##T2##W(VECT_VAR(vector, T1, W, N), L); \
   vst1##Q##_##T2##W(VECT_VAR(result, T1, W, N), VECT_VAR(vector, T1, W, N)); \
-  fprintf(ref_file, "%" PRIx##W ", ", VAR(var, T1, W))
+  fprintf(ref_file, "%s: %" PRIx##W "\n", "vget"STR(Q)"_lane_"STR(T2##W), VAR(var, T1, W))
 
   /* Special variant for floating-point */
   union {
@@ -49,7 +49,7 @@ void exec_vget_lane (void)
   VAR(var, T1, W) = vget##Q##_lane_##T2##W(VECT_VAR(vector, T1, W, N), L); \
   vst1##Q##_##T2##W(VECT_VAR(result, T1, W, N), VECT_VAR(vector, T1, W, N)); \
   var_int##W##_float##W.var_float##W = VAR(var, T1, W);		\
-  fprintf(ref_file, "%" PRIx##W ", ", var_int##W##_float##W.var_int##W)
+  fprintf(ref_file, "%s: %" PRIx##W "\n", "vget"STR(Q)"_lane_"STR(T2##W), var_int##W##_float##W.var_int##W)
 
   /* With ARM RVCT, we need to declare variables before any executable
      statement */