Message ID | 9a63abfe3b812a32d69c7393004bea4f88971559.1725280479.git.ps@pks.im (mailing list archive) |
---|---|
State | New |
Headers | show |
Series | refs/files: use heuristic to decide whether to repack with `--auto` | expand |
Patrick Steinhardt <ps@pks.im> writes: > The `--auto` flag for git-pack-refs(1) allows the ref backend to decide > whether or not a repack is in order. This switch has been introduced > mostly with the "reftable" backend in mind, which already knows to > auto-compact its tables during normal operations. When the flag is set, > then it will use the same auto-compaction mechanism and thus end up > doing nothing in most cases. > > The "files" backend does not have any such heuristic yet and instead Nit: s/instead/will instead/ > packs any loose references unconditionally. So we rewrite the complete > "packed-refs" file even if there's only a single loose reference to be > packed. > > Even worse, starting with 9f6714ab3e (builtin/gc: pack refs when using > `git maintenance run --auto`, 2024-03-25), `git pack-refs --auto` is > unconditionally executed via our auto maintenance, so we end up repacking > references every single time auto maintenance kicks in. And while that > commit already mentioned that the "files" backend unconditionally packs > refs now, the author obviously didn't quite think about the consequences > thereof. So while the idea was sound, we really should have added a > heuristic to the "files" backend before implementing it. > > Introduce a heuristic that decides whether or not it is worth to pack > loose references. The important factors to decide here are the number of > loose references in comparison to the overall size of the "packed-refs" > file. The bigger the "packed-refs" file, the longer it takes to rewrite > it and thus we scale up the limit of allowed loose references before we > repack. > > As is the nature of heuristics, this mechansim isn't obviously > "correct", but should rather be seen as a tradeoff between how much > resources we spend packing refs and how inefficient the ref store > becomes. For all I can say, we have successfully been using the exact > same heuristic in Gitaly for several years by now. > > Signed-off-by: Patrick Steinhardt <ps@pks.im> > --- > refs/files-backend.c | 75 +++++++++++++++++++++++++++++++ > refs/packed-backend.c | 18 ++++++++ > refs/packed-backend.h | 7 +++ > t/t0601-reffiles-pack-refs.sh | 85 ++++++++++++++++++++++++++++++----- > 4 files changed, 175 insertions(+), 10 deletions(-) > > diff --git a/refs/files-backend.c b/refs/files-backend.c > index 1cff65f6ae5..261e2b8570f 100644 > --- a/refs/files-backend.c > +++ b/refs/files-backend.c > @@ -1311,6 +1311,78 @@ static int should_pack_ref(struct files_ref_store *refs, > return 0; > } > > +static size_t fastlog2(size_t sz) Nit: We already `reftable/stack_test.c:fastlog2` and `bisect.c:log2i`. I wonder if it would be nice to consolidate all of them. But I guess it's okay, since the reftable code anyways is isolated from the rest of the Git code. > +{ > + size_t l = 0; > + if (!sz) > + return 0; > + for (; sz; sz /= 2) > + l++; > + return l - 1; > +} [snip] The rest of the patch looks great!
On Tue, Sep 03, 2024 at 02:00:16AM -0700, karthik nayak wrote: > Patrick Steinhardt <ps@pks.im> writes: > > diff --git a/refs/files-backend.c b/refs/files-backend.c > > index 1cff65f6ae5..261e2b8570f 100644 > > --- a/refs/files-backend.c > > +++ b/refs/files-backend.c > > @@ -1311,6 +1311,78 @@ static int should_pack_ref(struct files_ref_store *refs, > > return 0; > > } > > > > +static size_t fastlog2(size_t sz) > > Nit: We already `reftable/stack_test.c:fastlog2` and `bisect.c:log2i`. I > wonder if it would be nice to consolidate all of them. But I guess it's > okay, since the reftable code anyways is isolated from the rest of the > Git code. Yeah, the version we have here is in fact an (almost exact) copy of the function in our reftable codebase. And as you note, we do not want to deduplicate these such that the reftable library is somewhat isolated from the rest of the codebase. I also noticed `log2i()`, but honestly the only reason why I didn't reuse it is that I had no clue where to put it. There isn't any header that would be a good fit for it, and creating a new "math.h" header for a single function felt overblown to me. So I decided to just not bother. I'm happy to adjust though if somebody has a suggestion for where to put it. Patrick
Patrick Steinhardt <ps@pks.im> writes: > I also noticed `log2i()`, but honestly the only reason why I didn't > reuse it is that I had no clue where to put it. There isn't any header > that would be a good fit for it, and creating a new "math.h" header for > a single function felt overblown to me. So I decided to just not bother. > I'm happy to adjust though if somebody has a suggestion for where to put > it. Given the existing contents of wrapper.h (near the end of the file), I think wrapper.c would be a good place to do so. Shouldn't this essentially be a call to ffs() with the argument tweaked, by the way?
On Tue, Sep 03, 2024 at 11:23:12AM -0700, Junio C Hamano wrote: > Patrick Steinhardt <ps@pks.im> writes: > > > I also noticed `log2i()`, but honestly the only reason why I didn't > > reuse it is that I had no clue where to put it. There isn't any header > > that would be a good fit for it, and creating a new "math.h" header for > > a single function felt overblown to me. So I decided to just not bother. > > I'm happy to adjust though if somebody has a suggestion for where to put > > it. > > Given the existing contents of wrapper.h (near the end of the file), > I think wrapper.c would be a good place to do so. Indeed. Quite a grab bag of functions, thanks for the hint! > Shouldn't this essentially be a call to ffs() with the argument > tweaked, by the way? We are looking for the last set bit, not for the first set bit. We can massage things a bit, but wouldn't that require us to be aware of the platforms endianess? In any case, GCC is clever enough to notice what we're doing: fastlog2(unsigned long): xor eax, eax test rdi, rdi je .L5 bsr rax, rdi .L5: ret The `log2i()` function looks a bit less efficient: log2i(unsigned long): test rdi, rdi je .L3 bsr rdi, rdi lea eax, [rdi+1] cdqe ret .L3: xor eax, eax ret Clang isn't yet clever enough with v18.1, but is with trunk: log2i(unsigned long): bsr rax, rdi inc rax test rdi, rdi cmove rax, rdi ret fastlog2(unsigned long): test rdi, rdi je .LBB1_1 shr rdi bsr rcx, rdi mov eax, 127 cmovne rax, rcx xor eax, -64 add eax, 65 ret .LBB1_1: mov eax, -1 ret So with the following definition we're optimizing both with GCC and Clang: size_t fastlog2(size_t sz) { size_t l = 0; if (!sz) return 0; for (; sz; sz >>= 1) l++; return l; } I'd thus say we can just pick that function instead of caring about platform endianess with `ffs()`. Patrick
On Tue, Sep 03, 2024 at 02:00:16AM -0700, karthik nayak wrote: > Patrick Steinhardt <ps@pks.im> writes: > > > The `--auto` flag for git-pack-refs(1) allows the ref backend to decide > > whether or not a repack is in order. This switch has been introduced > > mostly with the "reftable" backend in mind, which already knows to > > auto-compact its tables during normal operations. When the flag is set, > > then it will use the same auto-compaction mechanism and thus end up > > doing nothing in most cases. > > > > The "files" backend does not have any such heuristic yet and instead > > Nit: s/instead/will instead/ > > > packs any loose references unconditionally. So we rewrite the complete > > "packed-refs" file even if there's only a single loose reference to be > > packed. Revisiting this: isn't the current version of this sentence correct? Replacing it with "will instead" certainly makes it incorrect without also changing "packs" to "pack". Patrick
Patrick Steinhardt <ps@pks.im> writes: > In any case, GCC is clever enough to notice what we're doing: > > fastlog2(unsigned long): > xor eax, eax > test rdi, rdi > je .L5 > bsr rax, rdi > .L5: > ret Nice. Aiming to compile to "bsr" is very good. > So with the following definition we're optimizing both with GCC and > Clang: > > size_t fastlog2(size_t sz) > { > size_t l = 0; > if (!sz) > return 0; > for (; sz; sz >>= 1) > l++; > return l; > } > > I'd thus say we can just pick that function instead of caring about > platform endianess with `ffs()`. The above loop that compilers seem to know to reduce to "bsr" is good. FWIW, because the definition of "first bit" in ffs() is "least significant bit", you do not need to worry about endianness at all, but what we want is most significant, so it is not directly usable.
Patrick Steinhardt <ps@pks.im> writes: > On Tue, Sep 03, 2024 at 02:00:16AM -0700, karthik nayak wrote: >> Patrick Steinhardt <ps@pks.im> writes: >> >> > The `--auto` flag for git-pack-refs(1) allows the ref backend to decide >> > whether or not a repack is in order. This switch has been introduced >> > mostly with the "reftable" backend in mind, which already knows to >> > auto-compact its tables during normal operations. When the flag is set, >> > then it will use the same auto-compaction mechanism and thus end up >> > doing nothing in most cases. >> > >> > The "files" backend does not have any such heuristic yet and instead >> >> Nit: s/instead/will instead/ >> >> > packs any loose references unconditionally. So we rewrite the complete >> > "packed-refs" file even if there's only a single loose reference to be >> > packed. > > Revisiting this: isn't the current version of this sentence correct? > Replacing it with "will instead" certainly makes it incorrect without > also changing "packs" to "pack". > > Patrick You're right, I did missread.
diff --git a/refs/files-backend.c b/refs/files-backend.c index 1cff65f6ae5..261e2b8570f 100644 --- a/refs/files-backend.c +++ b/refs/files-backend.c @@ -1311,6 +1311,78 @@ static int should_pack_ref(struct files_ref_store *refs, return 0; } +static size_t fastlog2(size_t sz) +{ + size_t l = 0; + if (!sz) + return 0; + for (; sz; sz /= 2) + l++; + return l - 1; +} + +static int should_pack_refs(struct files_ref_store *refs, + struct pack_refs_opts *opts) +{ + struct ref_iterator *iter; + size_t packed_size; + size_t refcount = 0; + size_t limit; + int ret; + + if (!(opts->flags & PACK_REFS_AUTO)) + return 1; + + ret = packed_refs_size(refs->packed_ref_store, &packed_size); + if (ret < 0) + die("cannot determine packed-refs size"); + + /* + * Packing loose references into the packed-refs file scales with the + * number of references we're about to write. We thus decide whether we + * repack refs by weighing the current size of the packed-refs file + * against the number of loose references. This is done such that we do + * not repack too often on repositories with a huge number of + * references, where we can expect a lot of churn in the number of + * references. + * + * As a heuristic, we repack if the number of loose references in the + * repository exceeds `log2(nr_packed_refs) * 5`, where we estimate + * `nr_packed_refs = packed_size / 100`, which scales as following: + * + * - 1kB ~ 10 packed refs: 16 refs + * - 10kB ~ 100 packed refs: 33 refs + * - 100kB ~ 1k packed refs: 49 refs + * - 1MB ~ 10k packed refs: 66 refs + * - 10MB ~ 100k packed refs: 82 refs + * - 100MB ~ 1m packed refs: 99 refs + * + * We thus allow roughly 16 additional loose refs per factor of ten of + * packed refs. This heuristic may be tweaked in the future, but should + * serve as a sufficiently good first iteration. + */ + limit = fastlog2(packed_size / 100) * 5; + if (limit < 16) + limit = 16; + + iter = cache_ref_iterator_begin(get_loose_ref_cache(refs, 0), NULL, + refs->base.repo, 0); + while ((ret = ref_iterator_advance(iter)) == ITER_OK) { + if (should_pack_ref(refs, iter->refname, iter->oid, + iter->flags, opts)) + refcount++; + if (refcount >= limit) { + ref_iterator_abort(iter); + return 1; + } + } + + if (ret != ITER_DONE) + die("error while iterating over references"); + + return 0; +} + static int files_pack_refs(struct ref_store *ref_store, struct pack_refs_opts *opts) { @@ -1323,6 +1395,9 @@ static int files_pack_refs(struct ref_store *ref_store, struct strbuf err = STRBUF_INIT; struct ref_transaction *transaction; + if (!should_pack_refs(refs, opts)) + return 0; + transaction = ref_store_transaction_begin(refs->packed_ref_store, &err); if (!transaction) return -1; diff --git a/refs/packed-backend.c b/refs/packed-backend.c index 7a0a695ca2f..07c57fd541a 100644 --- a/refs/packed-backend.c +++ b/refs/packed-backend.c @@ -1250,6 +1250,24 @@ int packed_refs_is_locked(struct ref_store *ref_store) return is_lock_file_locked(&refs->lock); } +int packed_refs_size(struct ref_store *ref_store, + size_t *out) +{ + struct packed_ref_store *refs = packed_downcast(ref_store, REF_STORE_READ, + "packed_refs_size"); + struct stat st; + + if (stat(refs->path, &st) < 0) { + if (errno != ENOENT) + return -1; + *out = 0; + return 0; + } + + *out = st.st_size; + return 0; +} + /* * The packed-refs header line that we write out. Perhaps other traits * will be added later. diff --git a/refs/packed-backend.h b/refs/packed-backend.h index 09437ad13bd..9481d5e7c2e 100644 --- a/refs/packed-backend.h +++ b/refs/packed-backend.h @@ -27,6 +27,13 @@ int packed_refs_lock(struct ref_store *ref_store, int flags, struct strbuf *err) void packed_refs_unlock(struct ref_store *ref_store); int packed_refs_is_locked(struct ref_store *ref_store); +/* + * Obtain the size of the `packed-refs` file. Reports `0` as size in case there + * is no packed-refs file. Returns 0 on success, negative otherwise. + */ +int packed_refs_size(struct ref_store *ref_store, + size_t *out); + /* * Return true if `transaction` really needs to be carried out against * the specified packed_ref_store, or false if it can be skipped diff --git a/t/t0601-reffiles-pack-refs.sh b/t/t0601-reffiles-pack-refs.sh index ed9652bb829..d8cbd3f202b 100755 --- a/t/t0601-reffiles-pack-refs.sh +++ b/t/t0601-reffiles-pack-refs.sh @@ -362,21 +362,86 @@ test_expect_success 'pack-refs does not drop broken refs during deletion' ' for command in "git pack-refs --all --auto" "git maintenance run --task=pack-refs --auto" do - test_expect_success "$command unconditionally packs loose refs" ' + test_expect_success "$command does not repack below 16 refs without packed-refs" ' test_when_finished "rm -rf repo" && git init repo && ( cd repo && git config set maintenance.auto false && - test_commit initial && - - git update-ref refs/heads/something HEAD && - test_path_is_file .git/refs/heads/something && - git rev-parse refs/heads/something >expect && - $command && - test_path_is_missing .git/refs/heads/something && - git rev-parse refs/heads/something >actual && - test_cmp expect actual + git commit --allow-empty --message "initial" && + + # Create 14 additional references, which brings us to + # 15 together with the default branch. + printf "create refs/heads/loose-%d HEAD\n" $(test_seq 14) >stdin && + git update-ref --stdin <stdin && + test_path_is_missing .git/packed-refs && + git pack-refs --auto --all && + test_path_is_missing .git/packed-refs && + + # Create the 16th reference, which should cause us to repack. + git update-ref refs/heads/loose-15 HEAD && + git pack-refs --auto --all && + test_path_is_file .git/packed-refs + ) + ' + + test_expect_success "$command does not repack below 16 refs with small packed-refs" ' + test_when_finished "rm -rf repo" && + git init repo && + ( + cd repo && + git config set maintenance.auto false && + git commit --allow-empty --message "initial" && + + git pack-refs --all && + test_line_count = 2 .git/packed-refs && + + # Create 15 loose references. + printf "create refs/heads/loose-%d HEAD\n" $(test_seq 15) >stdin && + git update-ref --stdin <stdin && + git pack-refs --auto --all && + test_line_count = 2 .git/packed-refs && + + # Create the 16th loose reference, which should cause us to repack. + git update-ref refs/heads/loose-17 HEAD && + git pack-refs --auto --all && + test_line_count = 18 .git/packed-refs + ) + ' + + test_expect_success "$command scales with size of packed-refs" ' + test_when_finished "rm -rf repo" && + git init repo && + ( + cd repo && + git config set maintenance.auto false && + git commit --allow-empty --message "initial" && + + # Create 99 packed refs. This should cause the heuristic + # to require more than the minimum amount of loose refs. + test_seq 99 | + while read i + do + printf "create refs/heads/packed-%d HEAD\n" $i || return 1 + done >stdin && + git update-ref --stdin <stdin && + git pack-refs --all && + test_line_count = 101 .git/packed-refs && + + # Create 24 loose refs, which should not yet cause us to repack. + printf "create refs/heads/loose-%d HEAD\n" $(test_seq 24) >stdin && + git update-ref --stdin <stdin && + git pack-refs --auto --all && + test_line_count = 101 .git/packed-refs && + + # Create another handful of refs to cross the border. + # Note that we explicitly do not check for strict + # boundaries here, as this also depends on the size of + # the object hash. + printf "create refs/heads/addn-%d HEAD\n" $(test_seq 10) >stdin && + git update-ref --stdin <stdin && + git pack-refs --auto --all && + test_line_count = 135 .git/packed-refs ) ' done
The `--auto` flag for git-pack-refs(1) allows the ref backend to decide whether or not a repack is in order. This switch has been introduced mostly with the "reftable" backend in mind, which already knows to auto-compact its tables during normal operations. When the flag is set, then it will use the same auto-compaction mechanism and thus end up doing nothing in most cases. The "files" backend does not have any such heuristic yet and instead packs any loose references unconditionally. So we rewrite the complete "packed-refs" file even if there's only a single loose reference to be packed. Even worse, starting with 9f6714ab3e (builtin/gc: pack refs when using `git maintenance run --auto`, 2024-03-25), `git pack-refs --auto` is unconditionally executed via our auto maintenance, so we end up repacking references every single time auto maintenance kicks in. And while that commit already mentioned that the "files" backend unconditionally packs refs now, the author obviously didn't quite think about the consequences thereof. So while the idea was sound, we really should have added a heuristic to the "files" backend before implementing it. Introduce a heuristic that decides whether or not it is worth to pack loose references. The important factors to decide here are the number of loose references in comparison to the overall size of the "packed-refs" file. The bigger the "packed-refs" file, the longer it takes to rewrite it and thus we scale up the limit of allowed loose references before we repack. As is the nature of heuristics, this mechansim isn't obviously "correct", but should rather be seen as a tradeoff between how much resources we spend packing refs and how inefficient the ref store becomes. For all I can say, we have successfully been using the exact same heuristic in Gitaly for several years by now. Signed-off-by: Patrick Steinhardt <ps@pks.im> --- refs/files-backend.c | 75 +++++++++++++++++++++++++++++++ refs/packed-backend.c | 18 ++++++++ refs/packed-backend.h | 7 +++ t/t0601-reffiles-pack-refs.sh | 85 ++++++++++++++++++++++++++++++----- 4 files changed, 175 insertions(+), 10 deletions(-)