Merge bitcoin#26252: refactor: Make 64-bit shift explicit

5c5b85d refactor: Make 64-bit shift explicit (Hennadii Stepanov)

Pull request description:

  [`std::array::at()`](https://en.cppreference.com/w/cpp/container/array/at) expects an argument of the `size_t` type. This PR avoids implicit type conversion (for both 64-bit and 32-bit systems).

  Also it enables MSVC warning [C4334](https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-3-c4334) for all codebase.

ACKs for top commit:
  MarcoFalke:
    ACK 5c5b85d 🚎
  jonatack:
    Code review ACK 5c5b85d

Tree-SHA512: fda850a42068f2ada9f877fac9ff8af1e22b5dcb3e708f5b95c316e77c52c72d33cd9ec6507a7f5d1731d1afdf5af6dc65025d388cc480f82c46f4d88ef2d306

Author: MacroFake

build_msvc/common.init.vcxproj.in

@@ -88,7 +88,7 @@
       <WarningLevel>Level3</WarningLevel>
       <PrecompiledHeader>NotUsing</PrecompiledHeader>
       <AdditionalOptions>/utf-8 /Zc:__cplusplus /std:c++20 %(AdditionalOptions)</AdditionalOptions>
-      <DisableSpecificWarnings>4018;4244;4267;4334;4715;4805</DisableSpecificWarnings>
+      <DisableSpecificWarnings>4018;4244;4267;4715;4805</DisableSpecificWarnings>
       <TreatWarningAsError>true</TreatWarningAsError>
       <PreprocessorDefinitions>_SILENCE_CXX17_CODECVT_HEADER_DEPRECATION_WARNING;ZMQ_STATIC;NOMINMAX;WIN32;HAVE_CONFIG_H;_CRT_SECURE_NO_WARNINGS;_SCL_SECURE_NO_WARNINGS;_CONSOLE;_WIN32_WINNT=0x0601;_WIN32_IE=0x0501;WIN32_LEAN_AND_MEAN;%(PreprocessorDefinitions)</PreprocessorDefinitions>
       <AdditionalIncludeDirectories>..\..\src;..\..\src\minisketch\include;..\..\src\univalue\include;..\..\src\secp256k1\include;..\..\src\leveldb\include;..\..\src\leveldb\helpers\memenv;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>

src/bech32.cpp

@@ -241,7 +241,7 @@ constexpr std::array<uint32_t, 25> GenerateSyndromeConstants() {
     std::array<uint32_t, 25> SYNDROME_CONSTS{};
     for (int k = 1; k < 6; ++k) {
         for (int shift = 0; shift < 5; ++shift) {
-            int16_t b = GF1024_LOG.at(1 << shift);
+            int16_t b = GF1024_LOG.at(size_t{1} << shift);
             int16_t c0 = GF1024_EXP.at((997*k + b) % 1023);
             int16_t c1 = GF1024_EXP.at((998*k + b) % 1023);
             int16_t c2 = GF1024_EXP.at((999*k + b) % 1023);