[PATCH 2/2] devtools/dump-cpu-flags: add tool to update CPU flags table

[Andre Muezerie <andremue@linux.microsoft.com>]

This patchset allows users to specify the CPU for which the generated
code should be optimized for by passing the CPU name.

MSVC does not provide this functionality natively, so logic was
added. This additional logic relies on a table which stores instruction
set availability (like AXV512F) for different CPUs.
To make it easier to update this table a new devtool is introduced
with this patch. The new tool generates the table entries for all CPUs
listed in an input file using a recent version of the compiler, which
has all the information needed. This reduces enormously the amount
of work needed to update the table in msvc/meson.build and makes the
process much less error prone.

Signed-off-by: Andre Muezerie <andremue@linux.microsoft.com>
---
 devtools/dump-cpu-flags/README.md          |  25 +++++
 devtools/dump-cpu-flags/cpu-names.txt      | 120 +++++++++++++++++++++
 devtools/dump-cpu-flags/dump-cpu-flags.cpp | 119 ++++++++++++++++++++
 devtools/dump-cpu-flags/dump-cpu-flags.py  |  41 +++++++
 4 files changed, 305 insertions(+)
 create mode 100644 devtools/dump-cpu-flags/README.md
 create mode 100644 devtools/dump-cpu-flags/cpu-names.txt
 create mode 100644 devtools/dump-cpu-flags/dump-cpu-flags.cpp
 create mode 100644 devtools/dump-cpu-flags/dump-cpu-flags.py

diff --git a/devtools/dump-cpu-flags/README.md b/devtools/dump-cpu-flags/README.md
new file mode 100644
index 0000000000..3db69f9f8f
--- /dev/null
+++ b/devtools/dump-cpu-flags/README.md
@@ -0,0 +1,25 @@
+# Generating updated CPU flags
+
+File `config\x86\msvc\meson.build` has a table with flags indicating instruction set support for a variety of CPU types.
+
+Script `dump-cpu-flags.py` can be used to generate updated entries for this table.
+
+The CPU names are stored in file `cpu-names.txt`, which is consumed by `dump-cpu-flags.py`. The formatting used in that file is described at the top of the file itself.
+
+The script relies on the information embedded in the g++ compiler. This means that an updated table can automatically be generated by switching to a newer version of the compiler. This avoids the need to manually edit the entries, which is error prone. With the script the table entries can just copied and pasted into `meson.build`. The only thing that might need to be done is adding new CPU names to cpu-names.txt, when new CPUs are released.
+
+**NOTE**: CPUs not known to the compiler will result in errors, which can be ignored (`dump-cpu-flags.py` will ignore these errors and continue). For best results use the latest g++ compiler available.
+
+Below is a sample output, where an error was logged because the compiler did not know about a CPU named ‘raptorlake’.
+
+```sh
+$ ./dump-cpu-flags.py
+       'x86-64-v2': [],
+       'x86-64-v3': ['AVX', 'AVX2'],
+       'x86-64-v4': ['AVX', 'AVX2', 'AVX512F', 'AVX512VL', 'AVX512BW', 'AVX512DQ', 'AVX512CD'],
+       'alderlake': ['AVX', 'PCLMUL', 'RDRND', 'AVX2', 'RDSEED', 'AES', 'VPCLMULQDQ', 'GFNI'],
+cc1plus: error: bad value (‘raptorlake’) for ‘-march=’ switch
+cc1plus: note: valid arguments to ‘-march=’ switch are: nocona core2 nehalem corei7 westmere sandybridge...
+      'silvermont': ['PCLMUL', 'RDRND'],
+             'slm': ['PCLMUL', 'RDRND'],
+```
\ No newline at end of file
diff --git a/devtools/dump-cpu-flags/cpu-names.txt b/devtools/dump-cpu-flags/cpu-names.txt
new file mode 100644
index 0000000000..5ceaf05c0d
--- /dev/null
+++ b/devtools/dump-cpu-flags/cpu-names.txt
@@ -0,0 +1,120 @@
+# This file is consumed by dump-cpu-flags.py. It should contain CPU names,
+# one per line. When the given CPU has a 32 bit architecture, it must be
+# indicated so by appending ", 32" to the line.
+# Always use the latest compiler available, otherwise it might not know
+# about some CPUs listed here.
+# The latest CPU names can be obtained from:
+# https://gcc.gnu.org/onlinedocs/gcc/x86-Options.html
+#
+
+x86-64
+x86-64-v2
+x86-64-v3
+x86-64-v4
+i386, 32
+i486, 32
+i586, 32
+pentium, 32
+lakemont, 32
+pentium-mmx, 32
+pentiumpro, 32
+i686, 32
+pentium2, 32
+pentium3, 32
+pentium3m, 32
+pentium-m, 32
+pentium4, 32
+pentium4m, 32
+prescott, 32
+nocona
+core2
+nehalem
+corei7
+westmere
+sandybridge
+corei7-avx
+ivybridge
+core-avx-i
+haswell
+core-avx2
+broadwell
+skylake
+skylake-avx512
+cascadelake
+cannonlake
+cooperlake
+icelake-client
+icelake-server
+tigerlake
+rocketlake
+alderlake
+raptorlake,
+meteorlake,
+gracemont
+arrowlake
+arrowlake-s
+lunarlake
+pantherlake
+sapphirerapids
+emeraldrapids
+graniterapids
+graniterapids-d
+diamondrapids
+bonnell
+atom
+silvermont
+slm
+goldmont
+goldmont-plus
+tremont
+sierraforest
+grandridge
+clearwaterforest
+k6, 32
+k6-2, 32
+k6-3, 32
+athlon, 32
+athlon-tbird, 32
+athlon-4, 32
+athlon-xp, 32
+athlon-mp, 32
+k8
+opteron
+athlon64
+athlon-fx
+k8-sse3
+opteron-sse3
+athlon64-sse3
+amdfam10
+barcelona
+bdver1
+bdver2
+bdver3
+bdver4
+znver1
+znver2
+znver3
+znver4
+znver5
+btver1
+btver2
+winchip-c6, 32
+winchip2, 32
+c3, 32
+c3-2, 32
+c7, 32
+samuel-2, 32
+nehemiah, 32
+esther, 32
+eden-x2
+eden-x4
+nano
+nano-1000
+nano-2000
+nano-3000
+nano-x2
+nano-x4
+lujiazui
+yongfeng
+shijidadao
+geode, 32
diff --git a/devtools/dump-cpu-flags/dump-cpu-flags.cpp b/devtools/dump-cpu-flags/dump-cpu-flags.cpp
new file mode 100644
index 0000000000..3bd89c29e0
--- /dev/null
+++ b/devtools/dump-cpu-flags/dump-cpu-flags.cpp
@@ -0,0 +1,119 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2025 Microsoft Corporation
+ */
+
+#include <iostream>
+#include <iomanip>
+#include <string>
+#include <vector>
+
+enum option {
+    FILTER_OMIT_SSE_SETS = 1,
+};
+
+std::vector<std::string> get_cpu_flags(option options)
+{
+    std::vector<std::string> cpu_flags;
+
+    if (!(options & FILTER_OMIT_SSE_SETS)) {
+#ifdef __SSE__
+        cpu_flags.push_back("SSE");
+#endif
+#ifdef __SSE2__
+        cpu_flags.push_back("SSE2");
+#endif
+#ifdef __SSE3__
+        cpu_flags.push_back("SSE3");
+#endif
+#ifdef __SSSE3__
+        cpu_flags.push_back("SSEE3");
+#endif
+#ifdef __SSE4_1__
+        cpu_flags.push_back("SSE4_1");
+#endif
+#ifdef __SSE4_2__
+        cpu_flags.push_back("SSE4_2");
+#endif
+    }
+
+#ifdef __AVX__
+    cpu_flags.push_back("AVX");
+#endif
+#ifdef __PCLMUL__
+    cpu_flags.push_back("PCLMUL");
+#endif
+#ifdef __RDRND__
+    cpu_flags.push_back("RDRND");
+#endif
+#ifdef __AVX2__
+    cpu_flags.push_back("AVX2");
+#endif
+#ifdef __RDSEED__
+    cpu_flags.push_back("RDSEED");
+#endif
+#ifdef __AES__
+    cpu_flags.push_back("AES");
+#endif
+#ifdef __VPCLMULQDQ__
+    cpu_flags.push_back("VPCLMULQDQ");
+#endif
+#ifdef __AVX512F__
+    cpu_flags.push_back("AVX512F");
+#endif
+#ifdef __AVX512VL__
+    cpu_flags.push_back("AVX512VL");
+#endif
+#ifdef __AVX512BW__
+    cpu_flags.push_back("AVX512BW");
+#endif
+#ifdef __AVX512DQ__
+    cpu_flags.push_back("AVX512DQ");
+#endif
+#ifdef __AVX512CD__
+    cpu_flags.push_back("AVX512CD");
+#endif
+#ifdef __AVX512IFMA__
+    cpu_flags.push_back("AVX512IFMA");
+#endif
+#ifdef __GFNI__
+    cpu_flags.push_back("GFNI");
+#endif
+    return cpu_flags;
+}
+
+void dump_cpu_flags(const std::string &cpu_name, const std::vector<std::string> &cpu_flags)
+{
+    std::string cpu_name_quoted = std::string("'") + cpu_name + "'";
+    std::cout << std::setw(18) << cpu_name_quoted << ": [";
+    for (size_t i = 0; i < cpu_flags.size(); ++i) {
+        if (i > 0)
+            std::cout << ", ";
+
+        std::cout << "'" << cpu_flags[i] << "'";
+    }
+    std::cout << "],\n";
+}
+
+bool does_cpu_meet_dpdk_requirements()
+{
+#ifdef __SSE4_2__
+    return true;
+#endif
+
+    return false;
+}
+
+int main(int argc, char *argv[])
+{
+    if (argc < 2) {
+        std::cout << "Usage: " << argv[0] << " <cpu_name>\n";
+        return -1;
+    }
+
+    if (does_cpu_meet_dpdk_requirements()) {
+        std::vector<std::string> cpu_flags = get_cpu_flags(FILTER_OMIT_SSE_SETS);
+        dump_cpu_flags(argv[1], cpu_flags);
+    }
+
+    return 0;
+}
diff --git a/devtools/dump-cpu-flags/dump-cpu-flags.py b/devtools/dump-cpu-flags/dump-cpu-flags.py
new file mode 100644
index 0000000000..660a4a6699
--- /dev/null
+++ b/devtools/dump-cpu-flags/dump-cpu-flags.py
@@ -0,0 +1,41 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2025 Microsoft Corporation
+
+"""
+This script generates a table which lists the flags indicating which instruction sets are
+supported for each CPU type.
+The CPU names are stored in file cpu-names.txt, which is consumed by this script.
+The script relies on the information embedded in the g++ compiler. This means that an updated
+table can automatically be generated by switching to a newer version of the compiler.
+The only thing that might need be done is adding new CPU names to cpu-names.txt, when new
+CPUs are released in the market.
+
+NOTE: CPUs not known to the compiler will result in errors, which can be ignored (this script
+will ignore these errors and continue). For best results use the latest g++ compiler available.
+"""
+
+import subprocess
+
+with open("cpu-names.txt", "r") as file:
+    for line in file:
+        line = line.strip()
+        if line.startswith("#") or line == "":
+            continue
+
+        words = line.split(",")
+        cpu_name = words[0].strip()
+        if len(words) > 1:
+            nbits = words[1].strip()
+        else:
+            nbits = ""
+
+        if nbits == "32":
+            result = subprocess.run(["g++", "dump-cpu-flags.cpp", "-o",
+                                     "dump-cpu-flags", f"-march={cpu_name}", "-m32"])
+        else:
+            result = subprocess.run(["g++", "dump-cpu-flags.cpp", "-o",
+                                     "dump-cpu-flags", f"-march={cpu_name}"])
+
+        if result.returncode == 0:
+            subprocess.run(["./dump-cpu-flags", cpu_name])
-- 
2.48.1.vfs.0.0