Make Sharded an enum and specialize it for the single thread case
#114860
Conversation
|
r? @b-naber (rustbot has picked a reviewer for you, use r? to override) |
rustbot
added
S-waiting-on-review
|
@bors try @rust-timer queue |
This comment has been minimized.
This comment has been minimized.
|
⌛ Trying commit c737c62 with merge c2631968523852a0815358b7f3e17eb91a677173... |
rustbot
added
the
S-waiting-on-perf
|
☀️ Try build successful - checks-actions |
This comment has been minimized.
This comment has been minimized.
|
Finished benchmarking commit (c2631968523852a0815358b7f3e17eb91a677173): comparison URL. Overall result: no relevant changes - no action neededBenchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf. @bors rollup=never Instruction countThis benchmark run did not return any relevant results for this metric. Max RSS (memory usage)ResultsThis is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.
CyclesResultsThis is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.
Binary sizeThis benchmark run did not return any relevant results for this metric. Bootstrap: 632.472s -> 632.68s (0.03%) |
rustbot
removed
the
S-waiting-on-perf
| /// a single shard to be used for greater cache efficiency. | ||
| /// A single field is used when the compiler uses only one thread. | ||
| pub enum Sharded<T> { | ||
| Single(Lock<T>), |
There was a problem hiding this comment.
We can use RefCell directly if it brought visible performeance benefit. Then we need to add Sharded to marker.rs
There was a problem hiding this comment.
That doesn't work as we need to return a single type for both cases. With #111713 landed however we can add a lock operation to Sharded that combines the branches on is_dyn_thread_safe.
There was a problem hiding this comment.
Yea I mean Single(RefCell<T>)
There was a problem hiding this comment.
Oh fn lock_shards requires the return type Lock<T>. Just forget it :)
rustbot
added
the
A-query-system
This comment has been minimized.
This comment has been minimized.
|
Thanks! It looks good to me. cc @nnethercote @cjgillot |
| use std::borrow::Borrow; | ||
| use std::collections::hash_map::RawEntryMut; | ||
| use std::hash::{Hash, Hasher}; | ||
| use std::mem; | ||
|
|
||
| #[cfg(parallel_compiler)] | ||
| // 32 shards is sufficient to reduce contention on an 8-core Ryzen 7 1700, | ||
| // but this should be tested on higher core count CPUs. How the `Sharded` type gets used | ||
| // may also affect the ideal number of shards. | ||
| const SHARD_BITS: usize = 5; |
There was a problem hiding this comment.
This is still used by get_shard_hash. Should it keep both possible values? Or should get_shard_hash be refactored out?
There was a problem hiding this comment.
get_shard_hash isn't actually used for anything with cfg(not(parallel_compiler)).
There was a problem hiding this comment.
Then, should it be cfg(parallel_compiler) to avoid accidental use?
There was a problem hiding this comment.
The function is still used with cfg(not(parallel_compiler)), but it's value doesn't end up used with optimizations applied as get_shard_by_index ignores the hash then.
|
@bors r+ |
bors
added
S-waiting-on-bors
|
⌛ Testing commit 0823f0c with merge 9cd8d80c729f199c56bac370baa2684a7235ae8e... |
|
The job Click to see the possible cause of the failure (guessed by this bot) |
|
💔 Test failed - checks-actions |
bors
added
S-waiting-on-review
|
That test failure looks unrelated to this PR. It's also ignored on MSVC. I wonder why MINGW is left out? |
It uses different debuginfo and unwinding so often unaffected by many MSVC issues regarding unwinding. |
|
Let's do a retry to see if the test is flaky? |
|
@bors r+ |
|
💡 This pull request was already approved, no need to approve it again.
|
bors
added
S-waiting-on-bors
|
☀️ Test successful - checks-actions |
|
Finished benchmarking commit (840ed5d): comparison URL. Overall result: ❌✅ regressions and improvements - no action needed@rustbot label: -perf-regression Instruction countThis is a highly reliable metric that was used to determine the overall result at the top of this comment.
Max RSS (memory usage)ResultsThis is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.
CyclesResultsThis is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.
Binary sizeThis benchmark run did not return any relevant results for this metric. Bootstrap: 636.596s -> 632.475s (-0.65%) |
This changes
Shardedto use a single shard by an enum, reducing the size ofShardedfor greater cache efficiency.Performance improvement with 1 thread and
cfg(parallel_compiler):I did see an unexpected 0.23% change for the serial compiler, so this could use a perf run to see if that reproduces.
cc @SparrowLii