Skip to content

use proper cache-line size variable in work-stealing queue #53035

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jan 24, 2024
Merged

use proper cache-line size variable in work-stealing queue #53035

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 9 additions & 0 deletions src/julia_atomics.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,6 +56,15 @@ enum jl_memory_order {
jl_memory_order_seq_cst
};

/**
* Cache line size
*/
#if defined(_CPU_AARCH64_) && defined(_OS_DARWIN_) // Apple silicon has 128 cache lines
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

So this would be wrong for a Linux system running on Apple Silicon (e.g. Asahi Linux)?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes.

#define JL_CACHE_BYTE_ALIGNMENT 128
#else
#define JL_CACHE_BYTE_ALIGNMENT 64
#endif

/**
* Thread synchronization primitives:
*
Expand Down
5 changes: 0 additions & 5 deletions src/julia_internal.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -485,11 +485,6 @@ STATIC_INLINE uint8_t JL_CONST_FUNC jl_gc_szclass_align8(unsigned sz) JL_NOTSAFE
}

#define JL_SMALL_BYTE_ALIGNMENT 16
#if defined(_CPU_AARCH64_) && defined(_OS_DARWIN_) // Apple silicon has 128 cache lines
#define JL_CACHE_BYTE_ALIGNMENT 128
#else
#define JL_CACHE_BYTE_ALIGNMENT 64
#endif
// JL_HEAP_ALIGNMENT is the maximum alignment that the GC can provide
#define JL_HEAP_ALIGNMENT JL_SMALL_BYTE_ALIGNMENT
#define GC_MAX_SZCLASS (2032-sizeof(void*))
Expand Down
2 changes: 1 addition & 1 deletion src/work-stealing-queue.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,7 +36,7 @@ static inline ws_array_t *create_ws_array(size_t capacity, int32_t eltsz) JL_NOT

typedef struct {
_Atomic(int64_t) top;
char _padding[128 - sizeof(int64_t)];
char _padding[JL_CACHE_BYTE_ALIGNMENT - sizeof(_Atomic(int64_t))];
_Atomic(int64_t) bottom; // put on a separate cache line. conservatively estimate cache line size as 128 bytes
_Atomic(ws_array_t *) array;
} ws_queue_t;
Expand Down