This updates Boost to 1.88, which is needed because Clio wants to move to 1.88 as that fixes several ASAN false positives around coroutine usage. In order for Clio to move to newer boost, libXRPL needs to move too. Hence the changes in this PR. A lot has changed between 1.83 and 1.88 so there are lots of changes in the diff, especially in regards to Boost.Asio and coroutines in particular.
* XRPLF/develop:
fix: Modify jobs to use '>>' instead of 'tee' for GITHUB_OUTPUT (#5699)
refactor: Revamp CI workflows (#5661)
refactor: Decouple net from xrpld and move rpc-related classes to the rpc folder (#5477)
Set version to 2.6.0-rc2
docs: Updates list of maintainers and reviewers (#5687)
fix: Change log to debug level for AMM offer retrieval and IOU payment check (#5686)
fix: Add -Wno-deprecated-declarations for Clang only (#5680)
Update .git-blame-ignore-revs for #5657 (#5675)
Fix BUILD.md instruction (#5676)
Set version to 2.6.0-rc1
fix: Improve logging of the reason to refuse a peer connection (#5664)
fix: Make test suite names match the directory name (#5597)
chore: Run prettier on all files (#5657)
chore: Set CONAN_REMOTE_URL also for forks (#5662)
chore: Cleanup bin/ directory (#5660)
perf: Optimize hash performance by avoiding allocating hash state object (#5469)
We're currently calling `XXH3_createState` and `XXH3_freeState` when hashing an object. However, it may be slow because they call `malloc` and `free`, which may affect the performance. This change avoids the use of the streaming API as much as possible by using an internal buffer.
The current implementation of rngfill is prone to false warnings from GCC about array bounds violations. Looking at the code, the implementation naively manipulates both the bytes count and the buffer pointer directly to ensure the trailing memcpy doesn't overrun the buffer. As expressed, there is a data dependency on both fields between loop iterations.
Now, ideally, an optimizing compiler would realize that these dependencies were unnecessary and end up restructuring its intermediate representation into a functionally equivalent form with them absent. However, the point at which this occurs may be disjoint from when warning analyses are performed, potentially rendering them more difficult to
determine precisely.
In addition, it may also consume a portion of the budget the optimizer has allocated to attempting to improve a translation unit's performance. Given this is a function template which requires context-sensitive instantiation, this code would be more prone than most to being inlined, with a decrease in optimization budget corresponding to the effort the optimizer has already expended, having already optimized one or more calling functions. Thus, the scope for impacting the the ultimate quality of the code generated is elevated.
For this change, we rearrange things so that the location and contents of each memcpy can be computed independently, relying on a simple loop iteration counter as the only changing input between iterations.
* upstream/develop: (56 commits)
Remove `include(default)` from libxrpl profile (#5587)
refactor: Change boost::shared_mutex to std::shared_mutex (#5576)
Fix macos runner (#5585)
Remove the type filter from "ledger" RPC command (#4934)
refactor: Update date, libarchive, nudb, openssl, sqlite3, xxhash packages (#5567)
test: Run unit tests regardless of 'Supported' amendment status (#5537)
Retire Flow Cross amendment (#5562)
chore: Update CI to use Conan 2 (#5556)
fixAMMClawbackRounding: adjust last holder's LPToken balance (#5513)
chore: Add gcc-12 workaround (#5554)
Add MPT related txns into issuer's account history (#5530)
chore: Remove unused headers (#5526)
fix: add allowTrustLineLocking flag for account_info (#5525)
Downgrade required CMake version for Antithesis SDK (#5548)
fix: Link with boost libraries explicitly (#5546)
chore: Fix compilation error with clang-20 and cleanup (#5543)
test: Remove circular jtx.h dependencies (#5544)
Decouple CredentialHelpers from xrpld/app/tx (#5487)
fix: crash when trace-logging in tests (#5529)
test: switch some unit tests to doctest (#5383)
...
The codebase is filled with includes that are unused, and which thus can be removed. At the same time, the files often do not include all headers that contain the definitions used in those files. This change uses clang-format and clang-tidy to clean up the includes, with minor manual intervention to ensure the code compiles on all platforms.
Combine multiple related debug log data points into a single
message. Allows quick correlation of events that
previously were either not logged or, if logged, strewn
across multiple lines, making correlation difficult.
The Heartbeat Timer and consensus ledger accept processing
each have this capability.
Also guarantees that log entries will be written if the
node is a validator, regardless of log severity level.
Otherwise, the level of these messages is at INFO severity.
* Rename ASSERT to XRPL_ASSERT
* Upgrade to Anthithesis SDK 0.4.4, and use new 0.4.4 features
* automatic cast to bool, like assert
* Add instrumentation workflow to verify build with instrumentation enabled
* Copy Antithesis SDK version 0.4.0 to directory external/
* Add build option `voidstar` to enable instrumentation with Antithesis SDK
* Define instrumentation macros ASSERT and UNREACHABLE in terms of regular C assert
* Replace asserts with named ASSERT or UNREACHABLE
* Add UNREACHABLE to LogicError
* Document instrumentation macros in CONTRIBUTING.md
