[libc] Increase the maximum RPC port size for future hardware (#188756)
Summary: We store the locks in local device memory for performance and simplicity. The number here needs to correspond to the maximum occupancy so that we never have a situation where a GPU thread is blocking another GPU thread. The number now is sufficient for most hardware, but modern compute chips like the MI300x are already pushing ~12000 resident waves. This has ABI impliciations so I'd like to bump it up sooner rather than later. The ABI change is within what OpenMP expects, LLVM major versions, and it will be caught statically so there's no risk of silent corruption (size doesn't match).
This commit is contained in:
parent
ffd6a13b5f
commit
3a56470a0e
@ -69,7 +69,9 @@ struct Header {
|
||||
};
|
||||
|
||||
/// The maximum number of parallel ports that the RPC interface can support.
|
||||
constexpr static uint64_t MAX_PORT_COUNT = 4096;
|
||||
/// This should be greater than the expected hardware's occupancy to ensure the
|
||||
/// interface is non-blocking.
|
||||
constexpr static uint64_t MAX_PORT_COUNT = 16384;
|
||||
|
||||
/// A common process used to synchronize communication between a client and a
|
||||
/// server. The process contains a read-only inbox and a write-only outbox used
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user