feat(version): merge release_v4.8.2 into master#6875
Merged
Conversation
Signed-off-by: InventiveCoder <liucongcong@outlook.com>
Signed-off-by: xiaochangbai <704566072@qq.com>
opt(framework): optimization Help Instruction
Signed-off-by: imalasong <imalasong@qq.com>
# Conflicts: # framework/src/main/java/org/tron/program/FullNode.java
fix(doc): fix some typos in comment
…develop Merge master into develop
* fix readme typo and format
Removed 4 config files from framework/src/main/resources that were not referenced anywhere in the codebase: - config-backup.conf: no references - config-beta.conf: no references - config-test-net.conf: no references - config-localtest.conf: only used in tests, shadowed by src/test/resources/config-localtest.conf on the test classpath
…6586) * test: add dual DB engine (LevelDB + RocksDB) test coverage
* feat(vm): check CREATE2 max depth under Osaka * feat(vm): clear residual logs on exception escape paths under Osaka Mirror the in-try failure path: failed txs must not leak event logs into TransactionInfo/Bloom. Gated on Osaka to preserve replay.
… triggers (#6833) On a chain reorg, BlockLogTrigger/TransactionLogTrigger subscribers previously never learned that an orphaned block's events were rolled back. Add the same "removed" semantics already used by ContractTrigger. - add a `removed` field to BlockLogTrigger/TransactionLogTrigger and the matching setRemoved() on their capsules - Manager (event version 0): refactor postBlockTrigger to emit a single real-time block (with removed), move the solidified-mode batch into postSolidityTrigger, add reOrgBlockTrigger to re-emit erased blocks with removed=true, rename reApplyLogsFilter to reApplyBlockEvents and re-emit the re-applied fork branch's block/tx triggers (forward), thread removed through processTransactionTrigger/postTransactionTrigger - RealtimeEventService (event version 1): post block/tx triggers synchronously to the plugin with the removed flag instead of the async triggerCapsuleQueue, avoiding overwrite of the shared cached capsule; drop the unused manager field - tests: capsule removed passthrough, version-0 emit (testReOrgBlockTriggerRemoved), rewritten RealtimeEventServiceTest, updated ManagerTest stub
Improve the hand-rolled JsonFormat parser used by HTTP wallet APIs so field-heavy and deeply nested JSON no longer overflows the request thread stack. mergeField now consumes one field per call and leaves comma iteration to the caller. JsonFormat also enforces Constant.MAX_NESTING_DEPTH for object/array descent and returns ParseException instead of allowing StackOverflowError to escape. Compatibility notes: direct JsonFormat.merge callers now reject nesting beyond 100 levels; trailing commas on known-field parse paths (top-level fields, known message fields, and known repeated fields) may be accepted, while trailing commas inside unknown nested object/array fields remain rejected; malicious deep-nesting requests now surface as parse errors instead of container-level HTTP 500s.
* fix(jsonrpc): enforce maxBlockRange on eth_getFilterLogs * style(jsonrpc): fix styles and comments
…#6851) Validate the inventory type at the inbound entry points and reject any value other than TRX or BLOCK with P2pException(BAD_MESSAGE), before the type is used for cache insertion or outbound fetch construction: - P2pEventHandlerImpl.checkInvRateLimit: add else branch for unknown type - InventoryMsgHandler.check: add type allowlist check - FetchInvDataMsgHandler.check: validate raw getInventoryType() instead of getInvMessageType() (which mapped non-BLOCK types to TRX) Use getTypeValue() when building the error message to avoid calling getNumber() on an UNRECOGNIZED enum value. Add regression tests for all three entry points. Co-authored-by: wb <wb@wbdeMacBook-Air.local>
#6857) * fix(vm): self-dispatch vote-witness cost by proposal flag getVoteWitnessCost3 falls back to cost2 when Osaka is off, and cost2 to the legacy cost when energy adjustment is off. The energy then stays correct even if a stale cost function lingers in the shared jump table after a reorg, or is read by a constant call whose view has the proposal off. * fix(vm): isolate constant-call config view to prevent global pollution Constant calls bound to a lagging solidity/PBFT snapshot loaded their flags into the process-global VMConfig, racing block processing during a proposal's activation-to-solidification window and risking a fork. Route a constant call's config into a thread-local snapshot; the block/broadcast path keeps writing (and reading) the global.
1. Treat known protobuf fields with literal null as absent so HTTP APIs recover the 4.8.1 fastjson-compatible behavior for null object fields. 2. This relaxes JSON-RPC buildTransaction ABI handling because that path now accepts null ABI fields instead of returning invalid abi. 3. TRON's JsonFormat now aligns more closely with protobuf Java JsonFormat: field-level null is treated as absent, while null elements inside repeated arrays remain invalid.
The switchFork new-branch apply loop only re-validated each block's witness signature; the transactions inside were applied through the isVerified cache. That flag caches a signature-verification result computed against a specific account-permission state, but a fork switch re-applies blocks on a rewound, diverged chain state where those permissions may have changed. Trusting the stale flag lets a transaction skip verification and be accepted with a signature no longer valid under the fork-chain state. - Manager.switchFork: clear isVerified on every transaction of each block on the branch being switched to, before applyBlock, forcing full signature re-validation against the fork-chain state. The switch-back path (original main branch) is left untouched: it reproduces the exact original state, so its cached verifications stay valid and resetting them would only cost perf. - ManagerTest: add switchForkShouldResetTransactionSignVerifiedOnNewBranch, which drives a real reorg and asserts the new branch's transaction gets its cached verification flag cleared (fails without the fix).
chore(release): record master merge into release_v4.8.2
lvs0075
approved these changes
Jul 15, 2026
CodeNinjaEvan
approved these changes
Jul 15, 2026
kuny0707
approved these changes
Jul 15, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
19 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR merges
release_v4.8.2intomaster.Main Changes
For more details, see the GreatVoyage-v4.8.2 milestone.