xe: jit: ir: fixup memory inefficiencies

Author: atkassen

src/gpu/intel/jit/ir/blocking.hpp

@@ -1,5 +1,5 @@
 /*******************************************************************************
-* Copyright 2023-2024 Intel Corporation
+* Copyright 2023-2025 Intel Corporation
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
@@ -50,7 +50,7 @@ class blocking_t {
         return loop_.is_empty() && thread_group_.is_empty() && iter_.is_empty();
     }
     bool is_spatial() const {
-        for (auto d : {pvars::iw, pvars::ow}) {
+        for (const auto &d : {pvars::iw, pvars::ow}) {
             if (iter_.has(d) && iter_[d] != 1) return true;
         }
         return false;

src/gpu/intel/jit/ir/config.hpp

@@ -130,11 +130,7 @@ class exec_cfg_param_t : public value_param_t<exec_config_t> {
     }
 
     std::vector<std::string> accepted_keys() const override {
-        std::vector<std::string> ret;
-        ret.push_back("regs");
-        ret.push_back("simd");
-        ret.push_back("vec");
-        return ret;
+        return {"regs", "simd", "vec"};
     }
 
     void set_from_str(

src/gpu/intel/jit/ir/gemm_schedule.hpp

@@ -1,5 +1,5 @@
 /*******************************************************************************
-* Copyright 2021-2024 Intel Corporation
+* Copyright 2021-2025 Intel Corporation
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
@@ -364,9 +364,9 @@ class loop_t {
         oss << " kind: " << kind_;
         if (unroll_factor_ != 1) oss << " unroll: " << unroll_factor_;
         std::vector<std::string> props;
-        if (is_root()) props.push_back("root");
-        if (is_fused_child()) props.push_back("fused");
-        if (is_split_parent()) props.push_back("split");
+        if (is_root()) props.emplace_back("root");
+        if (is_fused_child()) props.emplace_back("fused");
+        if (is_split_parent()) props.emplace_back("split");
         oss << "(" << make_seq_print_helper(props, ", ") << ")";
         return oss.str();
     }
@@ -623,8 +623,9 @@ class gemm_schedule_t {
         }
         auto &fused_loop = create_loop(fused_var, fused_bound);
         std::vector<std::reference_wrapper<loop_t>> loop_refs;
+        loop_refs.reserve(vars.size());
         for (auto &v : vars) {
-            loop_refs.push_back(find_loop(v));
+            loop_refs.emplace_back(find_loop(v));
         }
         fused_loop.set_fuse(loop_refs);
         set_bmnk_kind(fused_var, bmnk_kind(vars));

src/gpu/intel/jit/ir/ir.hpp

@@ -326,6 +326,7 @@ struct expr_cast_helper_t {
 
     static std::vector<T> call(const std::vector<expr_t> &exprs) {
         std::vector<T> ret;
+        ret.reserve(exprs.size());
         for (auto &e : exprs)
             ret.push_back(to_cpp<T>(e));
         return ret;
@@ -345,6 +346,7 @@ struct expr_cast_helper_t<expr_t> {
             = typename std::enable_if<std::is_arithmetic<U>::value>::type>
     static std::vector<expr_t> call(const std::vector<U> &vec) {
         std::vector<expr_t> ret;
+        ret.reserve(vec.size());
         for (auto &v : vec)
             ret.push_back(to_expr(v));
         return ret;

src/gpu/intel/jit/ir/tensor.hpp

@@ -553,7 +553,8 @@ class layout_t {
                     b_str.append(1, '*');
                 }
             }
-            ret = b_str + ret;
+            b_str += ret;
+            std::swap(ret, b_str);
             dense_stride = b.stride * b.block;
             seen[b.dim_idx] = true;
         }