From 800c3385348537b93d986053438e1789ceab10be Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Kamil=20=C5=9Aliwak?= <kamil.sliwak@codepoets.it>
Date: Sat, 23 Jan 2021 18:07:53 +0100
Subject: [PATCH] Scripts for generating and comparing bytecode reports for
 solc-bin binaries

---
 .../bytecode_reports_for_modified_binaries.sh | 133 ++++++++++++++++++
 scripts/solc-bin/compare_bytecode_reports.sh  |  64 +++++++++
 2 files changed, 197 insertions(+)
 create mode 100755 scripts/solc-bin/bytecode_reports_for_modified_binaries.sh
 create mode 100755 scripts/solc-bin/compare_bytecode_reports.sh

diff --git a/scripts/solc-bin/bytecode_reports_for_modified_binaries.sh b/scripts/solc-bin/bytecode_reports_for_modified_binaries.sh
new file mode 100755
index 000000000..111c19637
--- /dev/null
+++ b/scripts/solc-bin/bytecode_reports_for_modified_binaries.sh
@@ -0,0 +1,133 @@
+#!/usr/bin/env bash
+
+# ------------------------------------------------------------------------------
+# Determines versions of all release binaries from solc-bin repo modified in the
+# specified commit range, finds all release binaries from one, selected platform
+# that match these versions and uses them to produce bytecode reports.
+#
+# The script is meant to be backwards-compatible with older versions of the
+# compiler. Reports are generated via the CLI interface (rather than Standard
+# JSON) in case of native binaries because the JSON interface was not available
+# in very old versions.
+#
+# The script creates one report per binary and file names follow the pattern
+# 'report-<binary name>.txt'.
+#
+# Usage:
+#    <script name>.sh <PLATFORM> <BASE_REF> <TOP_REF> <SOLC_BIN_DIR> <SOLIDITY_DIR>
+#
+# PLATFORM: Platform name, corresponding the one of the top-level directories
+#     in solc-bin.
+# BASE_REF..TOP_REF: Commit range in the solc-bin repository to search for
+#     modified binaries.
+# SOLC_BIN_DIR: Directory containing a checkout of the ethereum/solc-bin
+#    repository with full history. Must be an absolute path.
+# SOLIDITY_DIR: Directory containing a checkout of the ethereum/solidity
+#    repository with full history. Bytecode report will be generated using
+#    scripts from the currently checked out revision. Must be an absolute path.
+#
+# Example:
+#    <script name>.sh linux-amd64 gh-pages pr-branch "$PWD/solc-bin" "$PWD/solidity"
+# ------------------------------------------------------------------------------
+# This file is part of solidity.
+#
+# solidity is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# solidity is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with solidity.  If not, see <http://www.gnu.org/licenses/>
+#
+# (c) 2020 solidity contributors.
+#------------------------------------------------------------------------------
+
+# FIXME: Can't use set -u because the old Bash on macOS treats empty arrays as unbound variables
+set -eo pipefail
+
+(( $# == 5 )) || { echo "ERROR: Not enough arguments."; exit 1; }
+
+platform="$1"
+base_ref="$2"
+top_ref="$3"
+solc_bin_dir="$4"
+solidity_dir="$5"
+
+report_dir="$PWD"
+tmp_dir=$(mktemp -d -t bytecode-reports-XXXXXX)
+solcjs_dir="$tmp_dir/solcjs"
+script_dir="$solidity_dir/scripts"
+
+cd "$tmp_dir"
+
+git clone https://github.com/ethereum/solc-js.git "$solcjs_dir"
+cd "$solcjs_dir"
+npm install
+rm soljson.js
+
+cd "${solc_bin_dir}/${platform}/"
+echo "Commit range: ${base_ref}..${top_ref}"
+
+modified_release_versions=$(
+    git diff --name-only "${base_ref}" |
+    sed -n -E 's/^[^\/]+\/(solc|soljson)-[0-9a-zA-Z-]+-v([0-9.]+)\+commit\.[0-9a-f]+(.[^.]+)?$/\2/p' |
+    sort -V |
+    uniq
+)
+echo "Release versions modified in the commit range:"
+echo "$modified_release_versions"
+
+# NOTE: We want perform the check when the soljson-* files in bin/ and wasm/ are modified too
+# because in that case the symlinks in emscripten-wasm32/ and emscripten-asmjs/ might remain
+# unchanged but were're assuming that these directories are never directly used as a platform name.
+[[ $platform != bin && $platform != wasm ]] || { echo "Invalid platform name."; exit 1; }
+
+platform_binaries="$(git ls-files "solc-${platform}-v*+commit.*" | sort -V)"
+
+for binary_name in $platform_binaries; do
+    solidity_version=$(echo "$binary_name" | sed -n -E 's/^solc-'"${platform}"'-v([0-9.]+)\+commit.+$/\1/p')
+
+    if echo "$modified_release_versions" | grep -x "$solidity_version"; then
+        echo "Binary ${binary_name} (version ${solidity_version}) matches one of the modified versions."
+
+        work_dir="${tmp_dir}/${binary_name}"
+        mkdir "$work_dir"
+        cd "$work_dir"
+
+        # While bytecode scripts come from the latest compiler, the test files should come from
+        # the Solidity version we're running them against to avoid errors due to breaking syntax changes.
+        git clone --branch "v${solidity_version}" "$solidity_dir" "${work_dir}/solidity/"
+        "${script_dir}/isolate_tests.py" "${work_dir}/solidity/test/"
+
+        if [[ $platform == emscripten-wasm32 ]] || [[ $platform == emscripten-asmjs ]]; then
+            ln -sf "${solc_bin_dir}/${platform}/${binary_name}" "${solcjs_dir}/soljson.js"
+            ln -s "${solcjs_dir}" solc-js
+            cp "${script_dir}/bytecodecompare/prepare_report.js" prepare_report.js
+
+            # shellcheck disable=SC2035
+            ./prepare_report.js --strip-smt-pragmas *.sol > "${report_dir}/report-${binary_name}.txt"
+        else
+            yul_optimizer_flags=()
+            if [[ $solidity_version == 0.6.0 ]] || [[ $solidity_version == 0.6.1 ]]; then
+                yul_optimizer_flags+=(--force-no-optimize-yul)
+            fi
+
+            "${script_dir}/bytecodecompare/prepare_report.py" "${solc_bin_dir}/${platform}/${binary_name}" \
+                --interface cli \
+                --smt-use strip-pragmas \
+                --report-file "${report_dir}/report-${binary_name}.txt" \
+                "${yul_optimizer_flags[@]}"
+        fi
+
+        rm -r "${work_dir}/solidity/"
+    else
+        echo "Binary ${binary_name} (version ${solidity_version}) does not match any modified version. Skipping."
+    fi
+done
+
+rm -r "$tmp_dir"
diff --git a/scripts/solc-bin/compare_bytecode_reports.sh b/scripts/solc-bin/compare_bytecode_reports.sh
new file mode 100755
index 000000000..7637a102c
--- /dev/null
+++ b/scripts/solc-bin/compare_bytecode_reports.sh
@@ -0,0 +1,64 @@
+#!/usr/bin/env bash
+
+# ------------------------------------------------------------------------------
+# Compares bytecode reports from multiple compiler versions and platforms to
+# determine if there are any differences.
+#
+# The script does not accept any arguments.
+#
+# The reports should be placed in subdirectories of the current working
+# directory and their names should follow one of the following patterns:
+#
+#     report-solc-<platform>-v<solidity version>+commit.<commit hash>.txt
+#     report-soljson-<platform>-v<solidity version>+commit.<commit hash>.js.txt
+#
+# Reports corresponding to the same version and commit hash are grouped together
+# and the script succeeds only if the files within each group are identical.
+# ------------------------------------------------------------------------------
+# This file is part of solidity.
+#
+# solidity is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# solidity is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with solidity.  If not, see <http://www.gnu.org/licenses/>
+#
+# (c) 2020 solidity contributors.
+#------------------------------------------------------------------------------
+
+set -euo pipefail
+
+report_files="$(find . -type f -name 'report-*.txt')"
+[[ $report_files != "" ]] || { echo "No reports found in the working directory."; exit 0; }
+
+echo "Available reports:"
+echo "$report_files"
+
+versions_in_report_names=$(
+    echo "$report_files" |
+    sed -n -E 's/^\.\/[^\/]+\/report-(solc|soljson)-[0-9a-zA-Z-]+-v([0-9.]+\+commit\.[0-9a-f]+)(.[^.]+)?\.txt$/\2/p' |
+    sort -V |
+    uniq
+)
+
+num_failed_comparisons=0
+for solidity_version_and_commit in $versions_in_report_names; do
+    echo "Comparing reports for Solidity ${solidity_version_and_commit}:"
+    mapfile -t report_files_for_version < <(
+        echo "$report_files" |
+        sed -n -E '/^\.\/[^\/]+\/report-(solc|soljson)-[0-9a-zA-Z-]+-v'"${solidity_version_and_commit//\+/\\+}"'+(.[^.]+)?\.txt$/p'
+    )
+
+    diff --report-identical-files --brief --from-file "${report_files_for_version[@]}" || ((++num_failed_comparisons))
+    echo
+done
+
+(( num_failed_comparisons == 0 )) || { echo "Found bytecode differences in ${num_failed_comparisons} versions"; exit 1; }
+echo "No bytecode differences found."