ARCHIVE/rippled
1
0
Fork 0
mirror of https://github.com/XRPLF/rippled.git synced 2026-06-04 01:06:48 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'develop' into gregtatcam/mpt/audit-attackathon-fixes-1

Browse Source
This commit is contained in:
Gregory Tsipenyuk
2026-05-28 16:01:43 -04:00
parent b2a0a2b31a 2f3558c610
commit 949deb5dc8
37 changed files with 1203 additions and 483 deletions
Download Patch File Download Diff File Expand all files Collapse all files

137
include/xrpl/basics/Number.h
View File

@@ -7,7 +7,9 @@
#include <limits>
#include <optional>
#include <ostream>
#include <set>
#include <string>
#include <unordered_map>
namespace xrpl {
@@ -44,11 +46,11 @@ isPowerOfTen(T value)
* * min is a power of 10, and
* * max = min * 10 - 1.
*
* The mantissa_scale enum indicates whether the range is "small" or "large".
* This intentionally restricts the number of MantissaRanges that can be
* instantiated to two: one for each scale.
* The MantissaScale enum indicates properties of the range: size, and some behavioral
* options. This intentionally restricts the number of unique MantissaRanges that can
* be instantiated: one for each scale.
*
* The "small" scale is based on the behavior of STAmount for IOUs. It has a min
* The "Small" scale is based on the behavior of STAmount for IOUs. It has a min
* value of 10^15, and a max value of 10^16-1. This was sufficient for
* uses before Lending Protocol was implemented, mostly related to AMM.
*
@@ -59,29 +61,54 @@ isPowerOfTen(T value)
* STNumber field type, and for internal calculations. That necessitated the
* "large" scale.
*
* The "large" scale is intended to represent all values that can be represented
* The "Large" scales are intended to represent all values that can be represented
* by an STAmount - IOUs, XRP, and MPTs. It has a min value of 10^18, and a max
* value of 10^19-1.
* value of 10^19-1. "LargeLegacy" is like "Large", but preserves
* a rounding error when a computation results in a mantissa of
* Number::kMaxRep that needs to be rounded up, but rounds down
* instead. It will maintain consistent behavior until the fixCleanup3_2_0
* amendment is enabled.
*
* Note that if the mentioned amendments are eventually retired, this class
* should be left in place, but the "small" scale option should be removed. This
* should be left in place, but the "Small" scale option should be removed. This
* will allow for future expansion beyond 64-bits if it is ever needed.
*/
struct MantissaRange
struct MantissaRange final
{
using rep = std::uint64_t;
enum class MantissaScale { Small, Large };
enum class MantissaScale {
Small,
// LargeLegacy can be removed when fixCleanup3_2_0 is retired
LargeLegacy,
Large,
};
// This entire enum can be removed when fixCleanup3_2_0 is retired
enum class CuspRoundingFix : bool {
Disabled = false,
Enabled = true,
};
explicit constexpr MantissaRange(MantissaScale scale)
: min(getMin(scale)), log(logTen(min).value_or(-1)), scale(scale)
: min(getMin(scale))
, cuspRoundingFixEnabled(isCuspFixEnabled(scale))
, log(logTen(min).value_or(-1))
, scale(scale)
{
}
rep min;
rep max{(min * 10) - 1};
CuspRoundingFix cuspRoundingFixEnabled;
int log;
MantissaScale scale;
static MantissaRange const&
getMantissaRange(MantissaScale scale);
static std::set<MantissaScale> const&
getAllScales();
private:
static constexpr rep
getMin(MantissaScale scale)
@@ -90,15 +117,35 @@ private:
{
case MantissaScale::Small:
return 1'000'000'000'000'000ULL;
case MantissaScale::LargeLegacy:
case MantissaScale::Large:
return 1'000'000'000'000'000'000ULL;
default:
// Since this can never be called outside a non-constexpr
// context, this throw assures that the build fails if an
// If called in a constexpr context, this throw assures that the build fails if an
// invalid scale is used.
throw std::runtime_error("Unknown mantissa scale");
throw std::runtime_error("Unknown mantissa scale"); // LCOV_EXCL_LINE
}
}
static constexpr CuspRoundingFix
isCuspFixEnabled(MantissaScale scale)
{
switch (scale)
{
case MantissaScale::Small:
case MantissaScale::LargeLegacy:
return CuspRoundingFix::Disabled;
case MantissaScale::Large:
return CuspRoundingFix::Enabled;
default:
// If called in a constexpr context, this throw assures that the build fails if an
// invalid scale is used.
throw std::runtime_error("Unknown mantissa scale"); // LCOV_EXCL_LINE
}
}
static std::unordered_map<MantissaScale, MantissaRange> const&
getRanges();
};
// Like std::integral, but only 64-bit integral types.
@@ -203,7 +250,7 @@ concept Integral64 = std::is_same_v<T, std::int64_t> || std::is_same_v<T, std::u
* amendments are enabled to determine which result to expect.
*
*/
class Number
class Number final
{
using rep = std::int64_t;
using internalrep = MantissaRange::rep;
@@ -424,49 +471,29 @@ public:
return kRange.get().log;
}
/// oneSmall is needed because the ranges are private
static constexpr Number
oneSmall();
/// oneLarge is needed because the ranges are private
static constexpr Number
oneLarge();
// And one is needed because it needs to choose between oneSmall and
// oneLarge based on the current range
static Number
one();
template <Integral64 T>
template <
auto MinMantissa,
auto MaxMantissa,
Integral64 T = std::decay_t<decltype(MinMantissa)>,
Integral64 TMax = std::decay_t<decltype(MaxMantissa)>>
[[nodiscard]]
std::pair<T, int>
normalizeToRange(T minMantissa, T maxMantissa) const;
normalizeToRange() const;
private:
static thread_local RoundingMode mode;
// The available ranges for mantissa
static constexpr MantissaRange kSmallRange{MantissaRange::MantissaScale::Small};
static_assert(isPowerOfTen(kSmallRange.min));
static_assert(kSmallRange.min == 1'000'000'000'000'000LL);
static_assert(kSmallRange.max == 9'999'999'999'999'999LL);
static_assert(kSmallRange.log == 15);
static_assert(kSmallRange.min < kMaxRep);
static_assert(kSmallRange.max < kMaxRep);
static constexpr MantissaRange kLargeRange{MantissaRange::MantissaScale::Large};
static_assert(isPowerOfTen(kLargeRange.min));
static_assert(kLargeRange.min == 1'000'000'000'000'000'000ULL);
static_assert(kLargeRange.max == internalrep(9'999'999'999'999'999'999ULL));
static_assert(kLargeRange.log == 18);
static_assert(kLargeRange.min < kMaxRep);
static_assert(kLargeRange.max > kMaxRep);
// The range for the mantissa when normalized.
// Use reference_wrapper to avoid making copies, and prevent accidentally
// changing the values inside the range.
static thread_local std::reference_wrapper<MantissaRange const> kRange;
void
normalize();
normalize(MantissaRange const& range);
/** Normalize Number components to an arbitrary range.
*
@@ -481,7 +508,8 @@ private:
T& mantissa,
int& exponent,
internalrep const& minMantissa,
internalrep const& maxMantissa);
internalrep const& maxMantissa,
MantissaRange::CuspRoundingFix cuspRoundingFixEnabled);
template <class T>
friend void
@@ -490,7 +518,8 @@ private:
T& mantissa,
int& exponent,
MantissaRange::rep const& minMantissa,
MantissaRange::rep const& maxMantissa);
MantissaRange::rep const& maxMantissa,
MantissaRange::CuspRoundingFix cuspRoundingFixEnabled);
[[nodiscard]] bool
isnormal() const noexcept;
@@ -526,7 +555,7 @@ static constexpr Number kNumZero{};
inline Number::Number(bool negative, internalrep mantissa, int exponent, Normalized)
: Number(negative, mantissa, exponent, Unchecked{})
{
normalize();
normalize(kRange);
}
inline Number::Number(internalrep mantissa, int exponent, Normalized)
@@ -696,10 +725,19 @@ Number::isnormal() const noexcept
kMinExponent <= exponent_ && exponent_ <= kMaxExponent);
}
template <Integral64 T>
template <auto MinMantissa, auto MaxMantissa, Integral64 T, Integral64 TMax>
std::pair<T, int>
Number::normalizeToRange(T minMantissa, T maxMantissa) const
Number::normalizeToRange() const
{
static_assert(std::is_same_v<T, std::uint64_t> || std::is_same_v<T, std::int64_t>);
static_assert(std::is_same_v<T, TMax>);
auto constexpr kMIN = static_cast<T>(MinMantissa);
auto constexpr kMAX = static_cast<T>(MaxMantissa);
static_assert(kMIN > 0);
static_assert(kMIN % 10 == 0);
static_assert(kMAX % 10 == 9);
static_assert((kMAX + 1) / 10 == kMIN);
bool negative = negative_;
internalrep mantissa = mantissa_;
int exponent = exponent_;
@@ -711,7 +749,10 @@ Number::normalizeToRange(T minMantissa, T maxMantissa) const
"xrpl::Number::normalizeToRange",
"Number is non-negative for unsigned range.");
}
Number::normalize(negative, mantissa, exponent, minMantissa, maxMantissa);
// Don't need to worry about the cuspRounding fix because rounding up will never take the
// mantissa over maxMantissa with a ones digit value other than 0. 0 can safely be truncated.
Number::normalize(
negative, mantissa, exponent, kMIN, kMAX, MantissaRange::CuspRoundingFix::Disabled);
auto const sign = negative ? -1 : 1;
return std::make_pair(static_cast<T>(sign * mantissa), exponent);
@@ -763,6 +804,8 @@ to_string(MantissaRange::MantissaScale const& scale)
{
case MantissaRange::MantissaScale::Small:
return "small";
case MantissaRange::MantissaScale::LargeLegacy:
return "largeLegacy";
case MantissaRange::MantissaScale::Large:
return "large";
default:

13
include/xrpl/protocol/Rules.h
View File

@@ -122,4 +122,17 @@ private:
std::optional<Rules> saved_;
};
class NumberSO;
class NumberMantissaScaleGuard;
bool
useRulesGuards(Rules const& rules);
void
createGuards(
Rules const& rules,
std::optional<NumberSO>& stNumberSO,
std::optional<CurrentTransactionRulesGuard>& rulesGuard,
std::optional<NumberMantissaScaleGuard>& mantissaScaleGuard);
} // namespace xrpl

2
include/xrpl/protocol/STAmount.h
View File

@@ -559,7 +559,7 @@ STAmount::fromNumber(A const& a, Number const& number)
return STAmount{asset, intValue, 0, negative};
}
auto const [mantissa, exponent] = working.normalizeToRange(kMinValue, kMaxValue);
auto const [mantissa, exponent] = working.normalizeToRange<kMinValue, kMaxValue>();
return STAmount{asset, mantissa, exponent, negative};
}

2
include/xrpl/protocol/detail/features.macro
View File

@@ -15,7 +15,7 @@
// Add new amendments to the top of this list.
// Keep it sorted in reverse chronological order.
XRPL_FIX (Cleanup3_2_0, Supported::No, VoteBehavior::DefaultNo)
XRPL_FIX (Cleanup3_2_0, Supported::Yes, VoteBehavior::DefaultNo)
XRPL_FEATURE(MPTokensV2, Supported::No, VoteBehavior::DefaultNo)
XRPL_FIX (Cleanup3_1_3, Supported::Yes, VoteBehavior::DefaultYes)
XRPL_FIX (BatchInnerSigs, Supported::No, VoteBehavior::DefaultNo)

4
include/xrpl/protocol_autogen/README.md
View File

@@ -15,8 +15,8 @@ Generation requires a one-time setup step to create a virtual environment
and install Python dependencies, followed by running the generation target:
```bash
cmake --build . --target setup_code_gen # create venv and install dependencies (once)
cmake --build . --target code_gen # generate code
cmake --build . --target setup_code_gen # create venv and install dependencies (once)
cmake --build . --target code_gen # generate code
```
By default, `CODEGEN_VENV_DIR` points to `.venv` in the project root. The