Commit 9ac3f0e5 authored by Duy Nguyen's avatar Duy Nguyen Committed by Junio C Hamano

pack-objects: fix performance issues on packing large deltas

Let's start with some background about oe_delta_size() and
oe_set_delta_size(). If you already know, skip the next paragraph.

These two are added in 0aca34e8 (pack-objects: shrink delta_size
field in struct object_entry - 2018-04-14) to help reduce 'struct
object_entry' size. The delta size field in this struct is reduced to
only contain max 1MB. So if any new delta is produced and larger than
1MB, it's dropped because we can't really save such a large size
anywhere. Fallback is provided in case existing packfiles already have
large deltas, then we can retrieve it from the pack.

While this should help small machines repacking large repos without
large deltas (i.e. less memory pressure), dropping large deltas during
the delta selection process could end up with worse pack files. And if
existing packfiles already have >1MB delta and pack-objects is
instructed to not reuse deltas, all of them will be dropped on the
floor, and the resulting pack would be definitely bigger.

There is also a regression in terms of CPU/IO if we have large on-disk
deltas because fallback code needs to parse the pack every time the
delta size is needed and just access to the mmap'd pack data is enough
for extra page faults when memory is under pressure.

Both of these issues were reported on the mailing list. Here's some
numbers for comparison.

    Version  Pack (MB)  MaxRSS(kB)  Time (s)
    -------  ---------  ----------  --------
     2.17.0     5498     43513628    2494.85
     2.18.0    10531     40449596    4168.94

This patch provides a better fallback that is

- cheaper in terms of cpu and io because we won't have to read
  existing pack files as much

- better in terms of pack size because the pack heuristics is back to
  2.17.0 time, we do not drop large deltas at all

If we encounter any delta (on-disk or created during try_delta phase)
that is larger than the 1MB limit, we stop using delta_size_ field for
this because it can't contain such size anyway. A new array of delta
size is dynamically allocated and can hold all the deltas that 2.17.0
can. This array only contains delta sizes that delta_size_ can't
contain.

With this, we do not have to drop deltas in try_delta() anymore. Of
course the downside is we use slightly more memory, even compared to
2.17.0. But since this is considered an uncommon case, a bit more
memory consumption should not be a problem.

Delta size limit is also raised from 1MB to 16MB to better cover
common case and avoid that extra memory consumption (99.999% deltas in
this reported repo are under 12MB; Jeff noted binary artifacts topped
out at about 3MB in some other private repos). Other fields are
shuffled around to keep this struct packed tight. We don't use more
memory in common case even with this limit update.

A note about thread synchronization. Since this code can be run in
parallel during delta searching phase, we need a mutex. The realloc
part in packlist_alloc() is not protected because it only happens
during the object counting phase, which is always single-threaded.

Access to e->delta_size_ (and by extension
pack->delta_size[e - pack->objects]) is unprotected as before, the
thread scheduler in pack-objects must make sure "e" is never updated
by two different threads.

The area under the new lock is as small as possible, avoiding locking
at all in common case, since lock contention with high thread count
could be expensive (most blobs are small enough that delta compute
time is short and we end up taking the lock very often). The previous
attempt to always hold a lock in oe_delta_size() and
oe_set_delta_size() increases execution time by 33% when repacking
linux.git with with 40 threads.
Reported-by: Elijah Newren's avatarElijah Newren <newren@gmail.com>
Helped-by: Elijah Newren's avatarElijah Newren <newren@gmail.com>
Helped-by: default avatarJeff King <peff@peff.net>
Signed-off-by: Duy Nguyen's avatarNguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: default avatarJunio C Hamano <gitster@pobox.com>
parent f6a5576d
...@@ -2008,10 +2008,6 @@ static int try_delta(struct unpacked *trg, struct unpacked *src, ...@@ -2008,10 +2008,6 @@ static int try_delta(struct unpacked *trg, struct unpacked *src,
delta_buf = create_delta(src->index, trg->data, trg_size, &delta_size, max_size); delta_buf = create_delta(src->index, trg->data, trg_size, &delta_size, max_size);
if (!delta_buf) if (!delta_buf)
return 0; return 0;
if (delta_size >= (1U << OE_DELTA_SIZE_BITS)) {
free(delta_buf);
return 0;
}
if (DELTA(trg_entry)) { if (DELTA(trg_entry)) {
/* Prefer only shallower same-sized deltas. */ /* Prefer only shallower same-sized deltas. */
...@@ -2263,6 +2259,7 @@ static void init_threaded_search(void) ...@@ -2263,6 +2259,7 @@ static void init_threaded_search(void)
pthread_mutex_init(&cache_mutex, NULL); pthread_mutex_init(&cache_mutex, NULL);
pthread_mutex_init(&progress_mutex, NULL); pthread_mutex_init(&progress_mutex, NULL);
pthread_cond_init(&progress_cond, NULL); pthread_cond_init(&progress_cond, NULL);
pthread_mutex_init(&to_pack.lock, NULL);
old_try_to_free_routine = set_try_to_free_routine(try_to_free_from_threads); old_try_to_free_routine = set_try_to_free_routine(try_to_free_from_threads);
} }
......
...@@ -14,6 +14,7 @@ then ...@@ -14,6 +14,7 @@ then
export GIT_TEST_SPLIT_INDEX=yes export GIT_TEST_SPLIT_INDEX=yes
export GIT_TEST_FULL_IN_PACK_ARRAY=true export GIT_TEST_FULL_IN_PACK_ARRAY=true
export GIT_TEST_OE_SIZE=10 export GIT_TEST_OE_SIZE=10
export GIT_TEST_OE_DELTA_SIZE=5
make --quiet test make --quiet test
fi fi
......
...@@ -146,6 +146,8 @@ void prepare_packing_data(struct packing_data *pdata) ...@@ -146,6 +146,8 @@ void prepare_packing_data(struct packing_data *pdata)
pdata->oe_size_limit = git_env_ulong("GIT_TEST_OE_SIZE", pdata->oe_size_limit = git_env_ulong("GIT_TEST_OE_SIZE",
1U << OE_SIZE_BITS); 1U << OE_SIZE_BITS);
pdata->oe_delta_size_limit = git_env_ulong("GIT_TEST_OE_DELTA_SIZE",
1UL << OE_DELTA_SIZE_BITS);
} }
struct object_entry *packlist_alloc(struct packing_data *pdata, struct object_entry *packlist_alloc(struct packing_data *pdata,
...@@ -160,6 +162,8 @@ struct object_entry *packlist_alloc(struct packing_data *pdata, ...@@ -160,6 +162,8 @@ struct object_entry *packlist_alloc(struct packing_data *pdata,
if (!pdata->in_pack_by_idx) if (!pdata->in_pack_by_idx)
REALLOC_ARRAY(pdata->in_pack, pdata->nr_alloc); REALLOC_ARRAY(pdata->in_pack, pdata->nr_alloc);
if (pdata->delta_size)
REALLOC_ARRAY(pdata->delta_size, pdata->nr_alloc);
} }
new_entry = pdata->objects + pdata->nr_objects++; new_entry = pdata->objects + pdata->nr_objects++;
......
...@@ -2,6 +2,7 @@ ...@@ -2,6 +2,7 @@
#define PACK_OBJECTS_H #define PACK_OBJECTS_H
#include "object-store.h" #include "object-store.h"
#include "thread-utils.h"
#define OE_DFS_STATE_BITS 2 #define OE_DFS_STATE_BITS 2
#define OE_DEPTH_BITS 12 #define OE_DEPTH_BITS 12
...@@ -12,7 +13,7 @@ ...@@ -12,7 +13,7 @@
* above this limit. Don't lower it too much. * above this limit. Don't lower it too much.
*/ */
#define OE_SIZE_BITS 31 #define OE_SIZE_BITS 31
#define OE_DELTA_SIZE_BITS 20 #define OE_DELTA_SIZE_BITS 23
/* /*
* State flags for depth-first search used for analyzing delta cycles. * State flags for depth-first search used for analyzing delta cycles.
...@@ -92,11 +93,12 @@ struct object_entry { ...@@ -92,11 +93,12 @@ struct object_entry {
*/ */
unsigned delta_size_:OE_DELTA_SIZE_BITS; /* delta data size (uncompressed) */ unsigned delta_size_:OE_DELTA_SIZE_BITS; /* delta data size (uncompressed) */
unsigned delta_size_valid:1; unsigned delta_size_valid:1;
unsigned char in_pack_header_size;
unsigned in_pack_idx:OE_IN_PACK_BITS; /* already in pack */ unsigned in_pack_idx:OE_IN_PACK_BITS; /* already in pack */
unsigned z_delta_size:OE_Z_DELTA_BITS; unsigned z_delta_size:OE_Z_DELTA_BITS;
unsigned type_valid:1; unsigned type_valid:1;
unsigned type_:TYPE_BITS;
unsigned no_try_delta:1; unsigned no_try_delta:1;
unsigned type_:TYPE_BITS;
unsigned in_pack_type:TYPE_BITS; /* could be delta */ unsigned in_pack_type:TYPE_BITS; /* could be delta */
unsigned preferred_base:1; /* unsigned preferred_base:1; /*
* we do not pack this, but is available * we do not pack this, but is available
...@@ -106,17 +108,16 @@ struct object_entry { ...@@ -106,17 +108,16 @@ struct object_entry {
unsigned tagged:1; /* near the very tip of refs */ unsigned tagged:1; /* near the very tip of refs */
unsigned filled:1; /* assigned write-order */ unsigned filled:1; /* assigned write-order */
unsigned dfs_state:OE_DFS_STATE_BITS; unsigned dfs_state:OE_DFS_STATE_BITS;
unsigned char in_pack_header_size;
unsigned depth:OE_DEPTH_BITS; unsigned depth:OE_DEPTH_BITS;
/* /*
* pahole results on 64-bit linux (gcc and clang) * pahole results on 64-bit linux (gcc and clang)
* *
* size: 80, bit_padding: 20 bits, holes: 8 bits * size: 80, bit_padding: 9 bits
* *
* and on 32-bit (gcc) * and on 32-bit (gcc)
* *
* size: 76, bit_padding: 20 bits, holes: 8 bits * size: 76, bit_padding: 9 bits
*/ */
}; };
...@@ -128,6 +129,7 @@ struct packing_data { ...@@ -128,6 +129,7 @@ struct packing_data {
uint32_t index_size; uint32_t index_size;
unsigned int *in_pack_pos; unsigned int *in_pack_pos;
unsigned long *delta_size;
/* /*
* Only one of these can be non-NULL and they have different * Only one of these can be non-NULL and they have different
...@@ -138,10 +140,29 @@ struct packing_data { ...@@ -138,10 +140,29 @@ struct packing_data {
struct packed_git **in_pack_by_idx; struct packed_git **in_pack_by_idx;
struct packed_git **in_pack; struct packed_git **in_pack;
#ifndef NO_PTHREADS
pthread_mutex_t lock;
#endif
uintmax_t oe_size_limit; uintmax_t oe_size_limit;
uintmax_t oe_delta_size_limit;
}; };
void prepare_packing_data(struct packing_data *pdata); void prepare_packing_data(struct packing_data *pdata);
static inline void packing_data_lock(struct packing_data *pdata)
{
#ifndef NO_PTHREADS
pthread_mutex_lock(&pdata->lock);
#endif
}
static inline void packing_data_unlock(struct packing_data *pdata)
{
#ifndef NO_PTHREADS
pthread_mutex_unlock(&pdata->lock);
#endif
}
struct object_entry *packlist_alloc(struct packing_data *pdata, struct object_entry *packlist_alloc(struct packing_data *pdata,
const unsigned char *sha1, const unsigned char *sha1,
uint32_t index_pos); uint32_t index_pos);
...@@ -330,18 +351,34 @@ static inline unsigned long oe_delta_size(struct packing_data *pack, ...@@ -330,18 +351,34 @@ static inline unsigned long oe_delta_size(struct packing_data *pack,
{ {
if (e->delta_size_valid) if (e->delta_size_valid)
return e->delta_size_; return e->delta_size_;
return oe_size(pack, e);
/*
* pack->detla_size[] can't be NULL because oe_set_delta_size()
* must have been called when a new delta is saved with
* oe_set_delta().
* If oe_delta() returns NULL (i.e. default state, which means
* delta_size_valid is also false), then the caller must never
* call oe_delta_size().
*/
return pack->delta_size[e - pack->objects];
} }
static inline void oe_set_delta_size(struct packing_data *pack, static inline void oe_set_delta_size(struct packing_data *pack,
struct object_entry *e, struct object_entry *e,
unsigned long size) unsigned long size)
{ {
if (size < pack->oe_delta_size_limit) {
e->delta_size_ = size; e->delta_size_ = size;
e->delta_size_valid = e->delta_size_ == size; e->delta_size_valid = 1;
if (!e->delta_size_valid && size != oe_size(pack, e)) } else {
BUG("this can only happen in check_object() " packing_data_lock(pack);
"where delta size is the same as entry size"); if (!pack->delta_size)
ALLOC_ARRAY(pack->delta_size, pack->nr_alloc);
packing_data_unlock(pack);
pack->delta_size[e - pack->objects] = size;
e->delta_size_valid = 0;
}
} }
#endif #endif
...@@ -315,6 +315,10 @@ packs on demand. This normally only happens when the object size is ...@@ -315,6 +315,10 @@ packs on demand. This normally only happens when the object size is
over 2GB. This variable forces the code path on any object larger than over 2GB. This variable forces the code path on any object larger than
<n> bytes. <n> bytes.
GIT_TEST_OE_DELTA_SIZE=<n> exercises the uncomon pack-objects code
path where deltas larger than this limit require extra memory
allocation for bookkeeping.
Naming Tests Naming Tests
------------ ------------
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment