mirrors/dynarmic
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

decoder/arm: Improve performance of arm decoding by adding LUT

Browse source
This commit is contained in:
Merry 2024-01-13 15:04:33 +00:00
parent 1c97fd5ec5
commit 331b41bc93
1 changed files with 25 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

30
src/dynarmic/frontend/A32/decoder/arm.h
View file

@ -23,9 +23,18 @@ namespace Dynarmic::A32 {
template<typename Visitor> template<typename Visitor>
using ArmMatcher = Decoder::Matcher<Visitor, u32>; using ArmMatcher = Decoder::Matcher<Visitor, u32>;
template<typename Visitor>
using ArmDecodeTable = std::array<std::vector<ArmMatcher<Visitor>>, 0x1000>;
namespace detail {
inline size_t ToFastLookupIndexArm(u32 instruction) {
return ((instruction >> 4) & 0x00F) | ((instruction >> 16) & 0xFF0);
}
} // namespace detail
template<typename V> template<typename V>
std::vector<ArmMatcher<V>> GetArmDecodeTable() { ArmDecodeTable<V> GetArmDecodeTable() {
std::vector<ArmMatcher<V>> table = { std::vector<ArmMatcher<V>> list = {
#define INST(fn, name, bitstring) DYNARMIC_DECODER_GET_MATCHER(ArmMatcher, fn, name, Decoder::detail::StringToArray<32>(bitstring)), #define INST(fn, name, bitstring) DYNARMIC_DECODER_GET_MATCHER(ArmMatcher, fn, name, Decoder::detail::StringToArray<32>(bitstring)),
#include "./arm.inc" #include "./arm.inc"
@ -34,10 +43,20 @@ std::vector<ArmMatcher<V>> GetArmDecodeTable() {
}; };
// If a matcher has more bits in its mask it is more specific, so it should come first. // If a matcher has more bits in its mask it is more specific, so it should come first.
std::stable_sort(table.begin(), table.end(), [](const auto& matcher1, const auto& matcher2) { std::stable_sort(list.begin(), list.end(), [](const auto& matcher1, const auto& matcher2) {
return mcl::bit::count_ones(matcher1.GetMask()) > mcl::bit::count_ones(matcher2.GetMask()); return mcl::bit::count_ones(matcher1.GetMask()) > mcl::bit::count_ones(matcher2.GetMask());
}); });
ArmDecodeTable<V> table{};
for (size_t i = 0; i < table.size(); ++i) {
for (auto matcher : list) {
const auto expect = detail::ToFastLookupIndexArm(matcher.GetExpected());
const auto mask = detail::ToFastLookupIndexArm(matcher.GetMask());
if ((i & mask) == expect) {
table[i].push_back(matcher);
}
}
}
return table; return table;
} }
@ -47,8 +66,9 @@ std::optional<std::reference_wrapper<const ArmMatcher<V>>> DecodeArm(u32 instruc
const auto matches_instruction = [instruction](const auto& matcher) { return matcher.Matches(instruction); }; const auto matches_instruction = [instruction](const auto& matcher) { return matcher.Matches(instruction); };
auto iter = std::find_if(table.begin(), table.end(), matches_instruction); const auto& subtable = table[detail::ToFastLookupIndexArm(instruction)];
return iter != table.end() ? std::optional<std::reference_wrapper<const ArmMatcher<V>>>(*iter) : std::nullopt; auto iter = std::find_if(subtable.begin(), subtable.end(), matches_instruction);
return iter != subtable.end() ? std::optional<std::reference_wrapper<const ArmMatcher<V>>>(*iter) : std::nullopt;
} }
} // namespace Dynarmic::A32 } // namespace Dynarmic::A32