This lets us put more allocations on an "almost as fast" path after a flush.
This results in around a 4% reduction in malloc cycles in prod workloads
(corresponding to about a 0.1% reduction in overall cycles).
With this, we track all of the empty, full, and low water states together. This
simplifies a lot of the tracking logic, since we now don't need the
cache_bin_info_t for state queries (except for some debugging).
I.e. the tcache code just calls a cache-bin function to finish flush (and move
pointers around, etc.). It doesn't directly access the cache-bin's owned memory
any more.
Previously, we took an array of cache_bin_info_ts and an index, and dereferenced
ourselves. But infos for other cache_bins aren't relevant to any particular
cache bin, so that should be the caller's job.
`tcache_bin_info` is not accessed on malloc fast path but the
compiler reserves a register for it, as well as an additional
register for `tcache_bin_info[ind].stack_size`. The optimization
gets rid of the need for the two registers.
The -1 value of low_water indicates if the cache has been depleted and
refilled. Track the status explicitly in the tcache struct.
This allows the fast path to check if (cur_ptr > low_water), instead of >=,
which avoids reaching slow path when the last item is allocated.
With the cache bin metadata switched to pointers, ncached_max is usually
accessed and timed by sizeof(ptr). Store the results in tcache_bin_info for
direct access, and add a helper function for the ncached_max value.
Implement the pointer-based metadata for tcache bins --
- 3 pointers are maintained to represent each bin;
- 2 of the pointers are compressed on 64-bit;
- is_full / is_empty done through pointer comparison;
Comparing to the previous counter based design --
- fast-path speed up ~15% in benchmarks
- direct pointer comparison and de-reference
- no need to access tcache_bin_info in common case
This eliminates the need for the arena stats code to "know" about tcaches; all
that it needs is a cache_bin_array_descriptor_t to tell it where to find
cache_bins whose stats it should aggregate.
This is the first step towards breaking up the tcache and arena (since they
interact primarily at the bin level). It should also make a future arena
caching implementation more straightforward.
