From 63889b02e7cccf4293f4a93cf71b516965cd01dc Mon Sep 17 00:00:00 2001
From: Eyck-Alexander Jentzsch <alex@minres.com>
Date: Wed, 19 Feb 2025 15:02:30 +0100
Subject: [PATCH] adds widening reductions

---
 gen_input/templates/interp/CORENAME.cpp.gtl | 13 +++++++++++
 src/vm/vector_functions.hpp                 | 26 ++++++++++++++-------
 2 files changed, 30 insertions(+), 9 deletions(-)

diff --git a/gen_input/templates/interp/CORENAME.cpp.gtl b/gen_input/templates/interp/CORENAME.cpp.gtl
index 50d663f..4b3aed8 100644
--- a/gen_input/templates/interp/CORENAME.cpp.gtl
+++ b/gen_input/templates/interp/CORENAME.cpp.gtl
@@ -481,6 +481,19 @@ if(vector != null) {%>
                 throw new std::runtime_error("Unsupported sew bit value");
         }
     }
+    void vector_red_wv(uint8_t* V, uint8_t funct6, uint8_t funct3, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t vd, uint8_t vs2, uint8_t vs1, uint8_t sew_val){
+        switch(sew_val){
+            case 0b000:
+                return softvector::vector_red_op<${vlen}, uint16_t, uint8_t>(V, funct6, funct3, vl, vstart, vtype, vm, vd, vs2, vs1);
+            case 0b001:
+                return softvector::vector_red_op<${vlen}, uint32_t, uint16_t>(V, funct6, funct3, vl, vstart, vtype, vm, vd, vs2, vs1);
+            case 0b010:
+                return softvector::vector_red_op<${vlen}, uint64_t, uint32_t>(V, funct6, funct3, vl, vstart, vtype, vm, vd, vs2, vs1);
+            case 0b011: // would require 128 bits vs2 value
+            default:
+                throw new std::runtime_error("Unsupported sew bit value");
+        }
+    }
 <%}%>
     uint64_t fetch_count{0};
     uint64_t tval{0};
diff --git a/src/vm/vector_functions.hpp b/src/vm/vector_functions.hpp
index 668d3bf..c3ece59 100644
--- a/src/vm/vector_functions.hpp
+++ b/src/vm/vector_functions.hpp
@@ -798,8 +798,14 @@ template <typename dest_elem_t, typename src_elem_t>
 std::function<void(dest_elem_t&, src_elem_t)> get_red_funct(unsigned funct6, unsigned funct3) {
     if(funct3 == OPIVV || funct3 == OPIVX || funct3 == OPIVI)
         switch(funct6) {
-        // case 0b110000: // VWREDSUMU
-        // case 0b110001: // VWREDSUM
+        case 0b110000: // VWREDSUMU
+            return [](dest_elem_t& running_total, src_elem_t vs2) { return running_total += static_cast<dest_elem_t>(vs2); };
+        case 0b110001: // VWREDSUM
+            return [](dest_elem_t& running_total, src_elem_t vs2) {
+                // cast the signed vs2 elem to unsigned to enable wraparound on overflow
+                return running_total += static_cast<dest_elem_t>(
+                           static_cast<std::make_signed_t<dest_elem_t>>(static_cast<std::make_signed_t<src_elem_t>>(vs2)));
+            };
         default:
             throw new std::runtime_error("Unknown funct6 in get_funct");
         }
@@ -814,18 +820,20 @@ std::function<void(dest_elem_t&, src_elem_t)> get_red_funct(unsigned funct6, uns
         case 0b000011: // VREDXOR
             return [](dest_elem_t& running_total, src_elem_t vs2) { running_total ^= vs2; };
         case 0b000100: // VREDMINU
-            return [](dest_elem_t& running_total, src_elem_t vs2) { running_total = std::min(running_total, vs2); };
+            return
+                [](dest_elem_t& running_total, src_elem_t vs2) { running_total = std::min(running_total, static_cast<dest_elem_t>(vs2)); };
         case 0b000101: // VREDMIN
             return [](dest_elem_t& running_total, src_elem_t vs2) {
-                running_total =
-                    std::min(static_cast<std::make_signed_t<dest_elem_t>>(running_total), static_cast<std::make_signed_t<src_elem_t>>(vs2));
+                running_total = std::min(static_cast<std::make_signed_t<dest_elem_t>>(running_total),
+                                         static_cast<std::make_signed_t<dest_elem_t>>(static_cast<std::make_signed_t<src_elem_t>>(vs2)));
             };
         case 0b000110: // VREDMAXU
-            return [](dest_elem_t& running_total, src_elem_t vs2) { running_total = std::max(running_total, vs2); };
+            return
+                [](dest_elem_t& running_total, src_elem_t vs2) { running_total = std::max(running_total, static_cast<dest_elem_t>(vs2)); };
         case 0b000111: // VREDMAX
             return [](dest_elem_t& running_total, src_elem_t vs2) {
-                running_total =
-                    std::max(static_cast<std::make_signed_t<dest_elem_t>>(running_total), static_cast<std::make_signed_t<src_elem_t>>(vs2));
+                running_total = std::max(static_cast<std::make_signed_t<dest_elem_t>>(running_total),
+                                         static_cast<std::make_signed_t<dest_elem_t>>(static_cast<std::make_signed_t<src_elem_t>>(vs2)));
             };
         default:
             throw new std::runtime_error("Unknown funct6 in get_funct");
@@ -840,7 +848,7 @@ void vector_red_op(uint8_t* V, unsigned funct6, unsigned funct3, uint64_t vl, ui
         return;
     uint64_t elem_count = VLEN * vtype.lmul() / vtype.sew();
     vmask_view mask_reg = read_vmask<VLEN>(V, elem_count);
-    auto vs1_elem = get_vreg<VLEN, src_elem_t>(V, vs1, elem_count)[0];
+    auto vs1_elem = get_vreg<VLEN, dest_elem_t>(V, vs1, elem_count)[0];
     auto vs2_view = get_vreg<VLEN, src_elem_t>(V, vs2, elem_count);
     auto vd_view = get_vreg<VLEN, dest_elem_t>(V, vd, elem_count);
     auto fn = get_red_funct<dest_elem_t, src_elem_t>(funct6, funct3);