server-skynet-source-3rd-je.../include/jemalloc/internal/hpdata.h
David Goldblatt 30b9e8162b HPA: Generalize purging.
Previously, we would purge a hugepage only when it's completely empty.  With
this change, we can purge even when only partially empty.  Although the
heuristic here is still fairly primitive, this infrastructure can scale to
become more advanced.
2021-02-04 20:58:31 -08:00

275 lines
8.0 KiB
C

#ifndef JEMALLOC_INTERNAL_HPDATA_H
#define JEMALLOC_INTERNAL_HPDATA_H
#include "jemalloc/internal/flat_bitmap.h"
#include "jemalloc/internal/ph.h"
#include "jemalloc/internal/ql.h"
#include "jemalloc/internal/typed_list.h"
/*
* The metadata representation we use for extents in hugepages. While the PAC
* uses the edata_t to represent both active and inactive extents, the HP only
* uses the edata_t for active ones; instead, inactive extent state is tracked
* within hpdata associated with the enclosing hugepage-sized, hugepage-aligned
* region of virtual address space.
*
* An hpdata need not be "truly" backed by a hugepage (which is not necessarily
* an observable property of any given region of address space). It's just
* hugepage-sized and hugepage-aligned; it's *potentially* huge.
*/
typedef struct hpdata_s hpdata_t;
struct hpdata_s {
/*
* We likewise follow the edata convention of mangling names and forcing
* the use of accessors -- this lets us add some consistency checks on
* access.
*/
/*
* The address of the hugepage in question. This can't be named h_addr,
* since that conflicts with a macro defined in Windows headers.
*/
void *h_address;
/* Its age (measured in psset operations). */
uint64_t h_age;
/* Whether or not we think the hugepage is mapped that way by the OS. */
bool h_huge;
/*
* Whether or not some thread is purging this hpdata (i.e. has called
* hpdata_purge_begin but not yet called hpdata_purge_end), or
* hugifying it. Only one thread at a time is allowed to change a
* hugepage's state.
*/
bool h_mid_purge;
bool h_mid_hugify;
/* Whether or not the hpdata is a the psset. */
bool h_in_psset;
union {
/* When nonempty, used by the psset bins. */
phn(hpdata_t) ph_link;
/*
* When empty (or not corresponding to any hugepage), list
* linkage.
*/
ql_elm(hpdata_t) ql_link;
};
/* The length of the largest contiguous sequence of inactive pages. */
size_t h_longest_free_range;
/* Number of active pages. */
size_t h_nactive;
/* A bitmap with bits set in the active pages. */
fb_group_t active_pages[FB_NGROUPS(HUGEPAGE_PAGES)];
/*
* Number of dirty pages, and a bitmap tracking them. This really means
* "dirty" from the OS's point of view; it includes both active and
* inactive pages that have been touched by the user.
*/
size_t h_ndirty;
/* The dirty pages (using the same definition as above). */
fb_group_t dirty_pages[FB_NGROUPS(HUGEPAGE_PAGES)];
};
static inline void *
hpdata_addr_get(const hpdata_t *hpdata) {
return hpdata->h_address;
}
static inline void
hpdata_addr_set(hpdata_t *hpdata, void *addr) {
assert(HUGEPAGE_ADDR2BASE(addr) == addr);
hpdata->h_address = addr;
}
static inline uint64_t
hpdata_age_get(const hpdata_t *hpdata) {
return hpdata->h_age;
}
static inline void
hpdata_age_set(hpdata_t *hpdata, uint64_t age) {
hpdata->h_age = age;
}
static inline bool
hpdata_huge_get(const hpdata_t *hpdata) {
return hpdata->h_huge;
}
static inline bool
hpdata_changing_state_get(const hpdata_t *hpdata) {
return hpdata->h_mid_purge || hpdata->h_mid_hugify;
}
static inline bool
hpdata_mid_purge_get(const hpdata_t *hpdata) {
return hpdata->h_mid_purge;
}
static inline bool
hpdata_mid_hugify_get(const hpdata_t *hpdata) {
return hpdata->h_mid_hugify;
}
static inline bool
hpdata_in_psset_get(const hpdata_t *hpdata) {
return hpdata->h_in_psset;
}
static inline void
hpdata_in_psset_set(hpdata_t *hpdata, bool in_psset) {
hpdata->h_in_psset = in_psset;
}
static inline size_t
hpdata_longest_free_range_get(const hpdata_t *hpdata) {
return hpdata->h_longest_free_range;
}
static inline void
hpdata_longest_free_range_set(hpdata_t *hpdata, size_t longest_free_range) {
assert(longest_free_range <= HUGEPAGE_PAGES);
hpdata->h_longest_free_range = longest_free_range;
}
static inline size_t
hpdata_nactive_get(hpdata_t *hpdata) {
return hpdata->h_nactive;
}
static inline size_t
hpdata_ndirty_get(hpdata_t *hpdata) {
return hpdata->h_ndirty;
}
static inline void
hpdata_assert_empty(hpdata_t *hpdata) {
assert(fb_empty(hpdata->active_pages, HUGEPAGE_PAGES));
assert(hpdata->h_nactive == 0);
}
/*
* Only used in tests, and in hpdata_assert_consistent, below. Verifies some
* consistency properties of the hpdata (e.g. that cached counts of page stats
* match computed ones).
*/
static inline bool
hpdata_consistent(hpdata_t *hpdata) {
if(fb_urange_longest(hpdata->active_pages, HUGEPAGE_PAGES)
!= hpdata_longest_free_range_get(hpdata)) {
return false;
}
if (fb_scount(hpdata->active_pages, HUGEPAGE_PAGES, 0, HUGEPAGE_PAGES)
!= hpdata->h_nactive) {
return false;
}
if (fb_scount(hpdata->dirty_pages, HUGEPAGE_PAGES, 0, HUGEPAGE_PAGES)
!= hpdata->h_ndirty) {
return false;
}
if (hpdata->h_ndirty < hpdata->h_nactive) {
return false;
}
if (hpdata->h_huge && hpdata->h_ndirty != HUGEPAGE_PAGES) {
return false;
}
return true;
}
static inline void
hpdata_assert_consistent(hpdata_t *hpdata) {
assert(hpdata_consistent(hpdata));
}
TYPED_LIST(hpdata_list, hpdata_t, ql_link)
typedef ph(hpdata_t) hpdata_age_heap_t;
ph_proto(, hpdata_age_heap_, hpdata_age_heap_t, hpdata_t);
static inline bool
hpdata_empty(hpdata_t *hpdata) {
return hpdata->h_nactive == 0;
}
void hpdata_init(hpdata_t *hpdata, void *addr, uint64_t age);
/*
* Given an hpdata which can serve an allocation request, pick and reserve an
* offset within that allocation.
*/
void *hpdata_reserve_alloc(hpdata_t *hpdata, size_t sz);
void hpdata_unreserve(hpdata_t *hpdata, void *begin, size_t sz);
/*
* The hpdata_purge_prepare_t allows grabbing the metadata required to purge
* subranges of a hugepage while holding a lock, drop the lock during the actual
* purging of them, and reacquire it to update the metadata again.
*/
typedef struct hpdata_purge_state_s hpdata_purge_state_t;
struct hpdata_purge_state_s {
size_t npurged;
fb_group_t to_purge[FB_NGROUPS(HUGEPAGE_PAGES)];
size_t next_purge_search_begin;
};
/*
* Initializes purge state. The access to hpdata must be externally
* synchronized with other hpdata_* calls.
*
* You can tell whether or not a thread is purging or hugifying a given hpdata
* via hpdata_changing_state_get(hpdata). Racing hugification or purging
* operations aren't allowed.
*
* Once you begin purging, you have to follow through and call hpdata_purge_next
* until you're done, and then end. Allocating out of an hpdata undergoing
* purging is not allowed.
*/
void hpdata_purge_begin(hpdata_t *hpdata, hpdata_purge_state_t *purge_state);
/*
* If there are more extents to purge, sets *r_purge_addr and *r_purge_size to
* true, and returns true. Otherwise, returns false to indicate that we're
* done.
*
* This requires exclusive access to the purge state, but *not* to the hpdata.
* In particular, unreserve calls are allowed while purging (i.e. you can dalloc
* into one part of the hpdata while purging a different part).
*/
bool hpdata_purge_next(hpdata_t *hpdata, hpdata_purge_state_t *purge_state,
void **r_purge_addr, size_t *r_purge_size);
/*
* Updates the hpdata metadata after all purging is done. Needs external
* synchronization.
*/
void hpdata_purge_end(hpdata_t *hpdata, hpdata_purge_state_t *purge_state);
/*
* Similarly, when hugifying , callers can do the metadata modifications while
* holding a lock (thereby setting the change_state field), but actually do the
* operation without blocking other threads.
*
* Unlike most metadata operations, hugification ending should happen while an
* hpdata is in the psset (or upcoming hugepage collections). This is because
* while purge/use races are unsafe, purge/hugepageify races are perfectly
* reasonable.
*/
void hpdata_hugify_begin(hpdata_t *hpdata);
void hpdata_hugify_end(hpdata_t *hpdata);
/*
* Tell the hpdata that it's no longer a hugepage (all its pages are still
* counted as dirty, though; an explicit purge call is required to change that).
*
* This should only be done after starting to purge, and before actually purging
* any contents.
*/
void hpdata_dehugify(hpdata_t *hpdata);
#endif /* JEMALLOC_INTERNAL_HPDATA_H */