[libc-commits] [libc] [libc] Cache old slabs when allocating GPU memory (PR #151866)

[Joseph Huber via libc-commits]

================
@@ -39,6 +40,9 @@ constexpr static uint32_t MIN_ALIGNMENT = MIN_SIZE - 1;
 // The number of times to attempt claiming an in-progress slab allocation.
 constexpr static uint32_t MAX_TRIES = 1024;
 
+// The number of previously allocated slabs we will keep in memory.
+constexpr static uint32_t CACHED_SLABS = 32;
----------------
jhuber6 wrote:

Right now this will use at most 64 MiB that becomes 'reserved' for the `malloc` implementation. It's a little high, but it'll only get that high if the user is *really* using `malloc`, so I think it's warranted. 

https://github.com/llvm/llvm-project/pull/151866