luke
/
yosys
mirror of https://github.com/YosysHQ/yosys.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Avoid racing accesses to shards[0] in ShardedHashtable parallel destruction

This commit is contained in:
Robert O'Callahan 2026-04-08 21:56:36 +00:00
parent a96cf8cc2b
commit 1f6559a5cf
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
kernel/threading.h
View File

@ -567,7 +567,7 @@ public:
// Call this for each shard to implement parallel destruction. For very large `ShardedHashtable`s, // Call this for each shard to implement parallel destruction. For very large `ShardedHashtable`s,
// deleting all elements of all shards on a single thread can be a performance bottleneck. // deleting all elements of all shards on a single thread can be a performance bottleneck.
void clear(const ThreadIndex &shard) { void clear(const ThreadIndex &shard) {
AccumulatedValueEquality equality = shards[0].key_eq(); AccumulatedValueEquality equality = shards[shard.thread_num].key_eq();
shards[shard.thread_num] = Shard(0, AccumulatedValueHashOp(), equality); shards[shard.thread_num] = Shard(0, AccumulatedValueHashOp(), equality);
} }
private: private: