In future tests it will be neccesary to create repositories with a set
number of tables. To make this easier, introduce the
`GIT_TEST_REFTABLE_AUTOCOMPACTION` environment variable that, when set
to false, disables autocompaction of reftables.
Signed-off-by: Justin Tobler <jltobler@gmail.com>
Acked-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ps/pack-refs-auto:
builtin/gc: pack refs when using `git maintenance run --auto`
builtin/gc: forward git-gc(1)'s `--auto` flag when packing refs
t6500: extract objects with "17" prefix
builtin/gc: move `struct maintenance_run_opts`
builtin/pack-refs: introduce new "--auto" flag
builtin/pack-refs: release allocated memory
refs/reftable: expose auto compaction via new flag
refs: remove `PACK_REFS_ALL` flag
refs: move `struct pack_refs_opts` to where it's used
t/helper: drop pack-refs wrapper
refs/reftable: print errors on compaction failure
reftable/stack: gracefully handle failed auto-compaction due to locks
reftable/stack: use error codes when locking fails during compaction
reftable/error: discern locked/outdated errors
reftable/stack: fix error handling in `reftable_stack_init_addition()`
Under normal circumstances, the "reftable" backend will automatically
perform compaction after appending to the stack. It is thus not
necessary and may even be considered wasteful to run git-pack-refs(1) in
"reftable"-backed repositories as it will cause the backend to compact
all tables into a single one. We do exactly that though when running
`git maintenance run --auto` or `git gc --auto`, which gets spawned by
Git after running some specific commands.
The `--auto` mode is typically only executing optimizations as needed.
To do so, we already use several heuristics for the various different
data structures in Git to determine whether to optimize them or not.
We do not use any heuristics for refs though and instead always optimize
them.
Introduce a new `PACK_REFS_AUTO` flag that can be passed to the backend.
When not handled by the backend we will continue to behave the exact
same as we do right now, that is we optimize refs unconditionally. This
is done for the "files" backend for now to retain current behaviour,
even though we may eventually also want to introduce heuristics here.
For the "reftable" backend though we already do have auto-compaction, so
we can easily reuse that logic to implement the new auto-packing flag.
Note that under normal circumstances, this should always end up being a
no-op. After all, we already invoke the code for every single addition
to the stack. But there are special cases where it can still be helpful
to execute the auto-compaction code explicitly:
- Concurrent writers may cause compaction to not run due to locks.
- Callers may decide to disable compaction altogether and then pack
refs at a later point due to various reasons.
- Other implementations of the reftable format may do compaction
differently or even not at all.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When git-pack-refs(1) fails in the reftable backend we end up printing
no error message at all, leaving the caller puzzled as to why compaction
has failed. Fix this.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The code to iterate over reflogs in the reftable has been optimized
to reduce memory allocation and deallocation.
Reviewed-by: Josh Steadmon <steadmon@google.com>
cf. <Ze9eX-aaWoVaqsPP@google.com>
* ps/reftable-reflog-iteration-perf:
refs/reftable: track last log record name via strbuf
reftable/record: use scratch buffer when decoding records
reftable/record: reuse message when decoding log records
reftable/record: reuse refnames when decoding log records
reftable/record: avoid copying author info
reftable/record: convert old and new object IDs to arrays
refs/reftable: reload correct stack when creating reflog iter
The code to iterate over refs with the reftable backend has seen
some optimization.
* ps/reftable-iteration-perf-part2:
refs/reftable: precompute prefix length
reftable: allow inlining of a few functions
reftable/record: decode keys in place
reftable/record: reuse refname when copying
reftable/record: reuse refname when decoding
reftable/merged: avoid duplicate pqueue emptiness check
reftable/merged: circumvent pqueue with single subiter
reftable/merged: handle subiter cleanup on close only
reftable/merged: remove unnecessary null check for subiters
reftable/merged: make subiters own their records
reftable/merged: advance subiter on subsequent iteration
reftable/merged: make `merged_iter` structure private
reftable/pq: use `size_t` to track iterator index
Clear the fallout from a fix for 2.44 regression.
* ps/reftable-repo-init-fix:
t0610: remove unused variable assignment
refs/reftable: don't fail empty transactions in repo without HEAD
"git for-each-ref" learned "--include-root-refs" option to show
even the stuff outside the 'refs/' hierarchy.
* kn/for-all-refs:
for-each-ref: add new option to include root refs
ref-filter: rename 'FILTER_REFS_ALL' to 'FILTER_REFS_REGULAR'
refs: introduce `refs_for_each_include_root_refs()`
refs: extract out `loose_fill_ref_dir_regular_file()`
refs: introduce `is_pseudoref()` and `is_headref()`
The reflog iterator enumerates all reflogs known to a ref backend. In
the "reftable" backend there is no way to list all existing reflogs
directly. Instead, we have to iterate through all reflog entries and
discard all those redundant entries for which we have already returned a
reflog entry.
This logic is implemented by tracking the last reflog name that we have
emitted to the iterator's user. If the next log record has the same name
we simply skip it until we find another record with a different refname.
This last reflog name is stored in a simple C string, which requires us
to free and reallocate it whenever we need to update the reflog name.
Convert it to use a `struct strbuf` instead, which reduces the number of
allocations. Before:
HEAP SUMMARY:
in use at exit: 13,473 bytes in 122 blocks
total heap usage: 1,068,485 allocs, 1,068,363 frees, 281,122,886 bytes allocated
After:
HEAP SUMMARY:
in use at exit: 13,473 bytes in 122 blocks
total heap usage: 68,485 allocs, 68,363 frees, 256,234,072 bytes allocated
Note that even after this change we still allocate quite a lot of data,
even though the number of allocations does not scale with the number of
log records anymore. This remainder comes mostly from decompressing the
log blocks, where we decompress each block into newly allocated memory.
This will be addressed at a later point in time.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 7af607c58d (reftable/record: store "val1" hashes as static arrays,
2024-01-03) and b31e3cc620 (reftable/record: store "val2" hashes as
static arrays, 2024-01-03) we have converted ref records to store their
object IDs in a static array. Convert log records to do the same so that
their old and new object IDs are arrays, too.
This change results in two allocations less per log record that we're
iterating over. Before:
HEAP SUMMARY:
in use at exit: 13,473 bytes in 122 blocks
total heap usage: 8,068,495 allocs, 8,068,373 frees, 401,011,862 bytes allocated
After:
HEAP SUMMARY:
in use at exit: 13,473 bytes in 122 blocks
total heap usage: 6,068,489 allocs, 6,068,367 frees, 361,011,822 bytes allocated
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When creating a new reflog iterator, we first have to reload the stack
that the iterator is being created. This is done so that any concurrent
writes to the stack are reflected. But `reflog_iterator_for_stack()`
always reloads the main stack, which is wrong.
Fix this and reload the correct stack.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ps/reftable-iteration-perf-part2:
refs/reftable: precompute prefix length
reftable: allow inlining of a few functions
reftable/record: decode keys in place
reftable/record: reuse refname when copying
reftable/record: reuse refname when decoding
reftable/merged: avoid duplicate pqueue emptiness check
reftable/merged: circumvent pqueue with single subiter
reftable/merged: handle subiter cleanup on close only
reftable/merged: remove unnecessary null check for subiters
reftable/merged: make subiters own their records
reftable/merged: advance subiter on subsequent iteration
reftable/merged: make `merged_iter` structure private
reftable/pq: use `size_t` to track iterator index
We're recomputing the prefix length on every iteration of the ref
iterator. Precompute it for another speedup when iterating over 1
million refs:
Benchmark 1: show-ref: single matching ref (revision = HEAD~)
Time (mean ± σ): 100.3 ms ± 3.7 ms [User: 97.3 ms, System: 2.8 ms]
Range (min … max): 97.5 ms … 139.7 ms 1000 runs
Benchmark 2: show-ref: single matching ref (revision = HEAD)
Time (mean ± σ): 95.8 ms ± 3.4 ms [User: 92.9 ms, System: 2.8 ms]
Range (min … max): 93.0 ms … 121.9 ms 1000 runs
Summary
show-ref: single matching ref (revision = HEAD) ran
1.05 ± 0.05 times faster than show-ref: single matching ref (revision = HEAD~)
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Under normal circumstances, it shouldn't ever happen that a repository
has no HEAD reference. In fact, git-update-ref(1) would fail any request
to delete the HEAD reference, and a newly initialized repository always
pre-creates it, too.
We have however changed git-clone(1) to partially initialize the
refdb just up to the point where remote helpers can find the
repository. With that change, we are going to run into a situation
where repositories have no refs at all.
Now there is a very particular edge case in this situation: when
preparing an empty ref transacton, we end up returning whatever value
`read_ref_without_reload()` returned to the caller. Under normal
conditions this would be fine: "HEAD" should usually exist, and thus the
function would return `0`. But if "HEAD" doesn't exist, the function
returns a positive value which we end up returning to the caller.
Fix this bug by resetting the return code to `0` and add a test.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The git-for-each-ref(1) command doesn't provide a way to print root refs
i.e pseudorefs and HEAD with the regular "refs/" prefixed refs.
This commit adds a new option "--include-root-refs" to
git-for-each-ref(1). When used this would also print pseudorefs and HEAD
for the current worktree.
Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The reflog iterator tries to resolve the corresponding ref for every
reflog that it is about to yield. Historically, this was done due to
multiple reasons:
- It ensures that the refname is safe because we end up calling
`check_refname_format()`. Also, non-conformant refnames are skipped
altogether.
- The iterator used to yield the resolved object ID as well as its
flags to the callback. This info was never used though, and the
corresponding parameters were dropped in the preceding commit.
- When a ref is corrupt then the reflog is not emitted at all.
We're about to introduce a new `git reflog list` subcommand that will
print all reflogs that the refdb knows about. Skipping over reflogs
whose refs are corrupted would be quite counterproductive in this case
as the user would have no way to learn about reflogs which may still
exist in their repository to help and rescue such a corrupted ref. Thus,
the only remaining reason for why we'd want to resolve the ref is to
verify its refname.
Refactor the code to call `check_refname_format()` directly instead of
trying to resolve the ref. This is significantly more efficient given
that we don't have to hit the object database anymore to list reflogs.
And second, it ensures that we end up showing reflogs of broken refs,
which will help to make the reflog more useful.
Note that this really only impacts the case where the corresponding ref
is corrupt. Reflogs for nonexistent refs would have been returned to the
caller beforehand already as we did not pass `RESOLVE_REF_READING` to
the function, and thus `refs_resolve_ref_unsafe()` would have returned
successfully in that case.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The ref and reflog iterators share much of the same underlying code to
iterate over the corresponding entries. This results in some weird code
because the reflog iterator also exposes an object ID as well as a flag
to the callback function. Neither of these fields do refer to the reflog
though -- they refer to the corresponding ref with the same name. This
is quite misleading. In practice at least the object ID cannot really be
implemented in any other way as a reflog does not have a specific object
ID in the first place. This is further stressed by the fact that none of
the callbacks except for our test helper make use of these fields.
Split up the infrastucture so that ref and reflog iterators use separate
callback signatures. This allows us to drop the nonsensical fields from
the reflog iterator.
Note that internally, the backends still use the same shared infra to
iterate over both types. As the backends should never end up being
called directly anyway, this is not much of a problem and thus kept
as-is for simplicity's sake.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the preceding commit we have converted the reflog iterator of the
"files" backend to be ordered, which was the only remaining ref iterator
that wasn't ordered. Refactor the ref iterator infrastructure so that we
always assume iterators to be ordered, thus simplifying the code.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When iterating through reflogs in a worktree we create a merged iterator
that merges reflogs from both refdbs. The resulting refs are ordered so
that instead we first return all worktree reflogs before we return all
common refs.
This is the only remaining case where a ref iterator returns entries in
a non-lexicographic order. The result would look something like the
following (listed with a command we introduce in a subsequent commit):
```
$ git reflog list
HEAD
refs/worktree/per-worktree
refs/heads/main
refs/heads/wt
```
So we first print the per-worktree reflogs in lexicographic order, then
the common reflogs in lexicographic order. This is confusing and not
consistent with how we print per-worktree refs, which are exclusively
sorted lexicographically.
Sort reflogs lexicographically in the same way as we sort normal refs.
As this is already implemented properly by the "reftable" backend via a
separate selection function, we simply pull out that logic and reuse it
for the "files" backend. As logs are properly sorted now, mark the
merged reflog iterator as sorted.
Tests will be added in a subsequent commit.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When copying a ref with the reftable backend we also copy the
corresponding log records. When seeking the first log record that we're
about to copy fails though we directly return from `write_copy_table()`
without doing any cleanup, leaking several allocated data structures.
Fix this by exiting via our common cleanup logic instead.
Reported-by: Jeff King <peff@peff.net> via Coverity
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Due to scalability issues, Shawn Pearce has originally proposed a new
"reftable" format more than six years ago [1]. Initially, this new
format was implemented in JGit with promising results. Around two years
ago, we have then added the "reftable" library to the Git codebase via
a4bbd13be3 (Merge branch 'hn/reftable', 2021-12-15). With this we have
landed all the low-level code to read and write reftables. Notably
missing though was the integration of this low-level code into the Git
code base in the form of a new ref backend that ties all of this
together.
This gap is now finally closed by introducing a new "reftable" backend
into the Git codebase. This new backend promises to bring some notable
improvements to Git repositories:
- It becomes possible to do truly atomic writes where either all refs
are committed to disk or none are. This was not possible with the
"files" backend because ref updates were split across multiple loose
files.
- The disk space required to store many refs is reduced, both compared
to loose refs and packed-refs. This is enabled both by the reftable
format being a binary format, which is more compact, and by prefix
compression.
- We can ignore filesystem-specific behaviour as ref names are not
encoded via paths anymore. This means there is no need to handle
case sensitivity on Windows systems or Unicode precomposition on
macOS.
- There is no need to rewrite the complete refdb anymore every time a
ref is being deleted like it was the case for packed-refs. This
means that ref deletions are now constant time instead of scaling
linearly with the number of refs.
- We can ignore file/directory conflicts so that it becomes possible
to store both "refs/heads/foo" and "refs/heads/foo/bar".
- Due to this property we can retain reflogs for deleted refs. We have
previously been deleting reflogs together with their refs to avoid
file/directory conflicts, which is not necessary anymore.
- We can properly enumerate all refs. With the "files" backend it is
not easily possible to distinguish between refs and non-refs because
they may live side by side in the gitdir.
Not all of these improvements are realized with the current "reftable"
backend implementation. At this point, the new backend is supposed to be
a drop-in replacement for the "files" backend that is used by basically
all Git repositories nowadays. It strives for 1:1 compatibility, which
means that a user can expect the same behaviour regardless of whether
they use the "reftable" backend or the "files" backend for most of the
part.
Most notably, this means we artificially limit the capabilities of the
"reftable" backend to match the limits of the "files" backend. It is not
possible to create refs that would end up with file/directory conflicts,
we do not retain reflogs, we perform stricter-than-necessary checks.
This is done intentionally due to two main reasons:
- It makes it significantly easier to land the "reftable" backend as
tests behave the same. It would be tough to argue for each and every
single test that doesn't pass with the "reftable" backend.
- It ensures compatibility between repositories that use the "files"
backend and repositories that use the "reftable" backend. Like this,
hosters can migrate their repositories to use the "reftable" backend
without causing issues for clients that use the "files" backend in
their clones.
It is expected that these artificial limitations may eventually go away
in the long term.
Performance-wise things very much depend on the actual workload. The
following benchmarks compare the "files" and "reftable" backends in the
current version:
- Creating N refs in separate transactions shows that the "files"
backend is ~50% faster. This is not surprising given that creating a
ref only requires us to create a single loose ref. The "reftable"
backend will also perform auto compaction on updates. In real-world
workloads we would likely also want to perform pack loose refs,
which would likely change the picture.
Benchmark 1: update-ref: create refs sequentially (refformat = files, refcount = 1)
Time (mean ± σ): 2.1 ms ± 0.3 ms [User: 0.6 ms, System: 1.7 ms]
Range (min … max): 1.8 ms … 4.3 ms 133 runs
Benchmark 2: update-ref: create refs sequentially (refformat = reftable, refcount = 1)
Time (mean ± σ): 2.7 ms ± 0.1 ms [User: 0.6 ms, System: 2.2 ms]
Range (min … max): 2.4 ms … 2.9 ms 132 runs
Benchmark 3: update-ref: create refs sequentially (refformat = files, refcount = 1000)
Time (mean ± σ): 1.975 s ± 0.006 s [User: 0.437 s, System: 1.535 s]
Range (min … max): 1.969 s … 1.980 s 3 runs
Benchmark 4: update-ref: create refs sequentially (refformat = reftable, refcount = 1000)
Time (mean ± σ): 2.611 s ± 0.013 s [User: 0.782 s, System: 1.825 s]
Range (min … max): 2.597 s … 2.622 s 3 runs
Benchmark 5: update-ref: create refs sequentially (refformat = files, refcount = 100000)
Time (mean ± σ): 198.442 s ± 0.241 s [User: 43.051 s, System: 155.250 s]
Range (min … max): 198.189 s … 198.670 s 3 runs
Benchmark 6: update-ref: create refs sequentially (refformat = reftable, refcount = 100000)
Time (mean ± σ): 294.509 s ± 4.269 s [User: 104.046 s, System: 190.326 s]
Range (min … max): 290.223 s … 298.761 s 3 runs
- Creating N refs in a single transaction shows that the "files"
backend is significantly slower once we start to write many refs.
The "reftable" backend only needs to update two files, whereas the
"files" backend needs to write one file per ref.
Benchmark 1: update-ref: create many refs (refformat = files, refcount = 1)
Time (mean ± σ): 1.9 ms ± 0.1 ms [User: 0.4 ms, System: 1.4 ms]
Range (min … max): 1.8 ms … 2.6 ms 151 runs
Benchmark 2: update-ref: create many refs (refformat = reftable, refcount = 1)
Time (mean ± σ): 2.5 ms ± 0.1 ms [User: 0.7 ms, System: 1.7 ms]
Range (min … max): 2.4 ms … 3.4 ms 148 runs
Benchmark 3: update-ref: create many refs (refformat = files, refcount = 1000)
Time (mean ± σ): 152.5 ms ± 5.2 ms [User: 19.1 ms, System: 133.1 ms]
Range (min … max): 148.5 ms … 167.8 ms 15 runs
Benchmark 4: update-ref: create many refs (refformat = reftable, refcount = 1000)
Time (mean ± σ): 58.0 ms ± 2.5 ms [User: 28.4 ms, System: 29.4 ms]
Range (min … max): 56.3 ms … 72.9 ms 40 runs
Benchmark 5: update-ref: create many refs (refformat = files, refcount = 1000000)
Time (mean ± σ): 152.752 s ± 0.710 s [User: 20.315 s, System: 131.310 s]
Range (min … max): 152.165 s … 153.542 s 3 runs
Benchmark 6: update-ref: create many refs (refformat = reftable, refcount = 1000000)
Time (mean ± σ): 51.912 s ± 0.127 s [User: 26.483 s, System: 25.424 s]
Range (min … max): 51.769 s … 52.012 s 3 runs
- Deleting a ref in a fully-packed repository shows that the "files"
backend scales with the number of refs. The "reftable" backend has
constant-time deletions.
Benchmark 1: update-ref: delete ref (refformat = files, refcount = 1)
Time (mean ± σ): 1.7 ms ± 0.1 ms [User: 0.4 ms, System: 1.2 ms]
Range (min … max): 1.6 ms … 2.1 ms 316 runs
Benchmark 2: update-ref: delete ref (refformat = reftable, refcount = 1)
Time (mean ± σ): 1.8 ms ± 0.1 ms [User: 0.4 ms, System: 1.3 ms]
Range (min … max): 1.7 ms … 2.1 ms 294 runs
Benchmark 3: update-ref: delete ref (refformat = files, refcount = 1000)
Time (mean ± σ): 2.0 ms ± 0.1 ms [User: 0.5 ms, System: 1.4 ms]
Range (min … max): 1.9 ms … 2.5 ms 287 runs
Benchmark 4: update-ref: delete ref (refformat = reftable, refcount = 1000)
Time (mean ± σ): 1.9 ms ± 0.1 ms [User: 0.5 ms, System: 1.3 ms]
Range (min … max): 1.8 ms … 2.1 ms 217 runs
Benchmark 5: update-ref: delete ref (refformat = files, refcount = 1000000)
Time (mean ± σ): 229.8 ms ± 7.9 ms [User: 182.6 ms, System: 46.8 ms]
Range (min … max): 224.6 ms … 245.2 ms 6 runs
Benchmark 6: update-ref: delete ref (refformat = reftable, refcount = 1000000)
Time (mean ± σ): 2.0 ms ± 0.0 ms [User: 0.6 ms, System: 1.3 ms]
Range (min … max): 2.0 ms … 2.1 ms 3 runs
- Listing all refs shows no significant advantage for either of the
backends. The "files" backend is a bit faster, but not by a
significant margin. When repositories are not packed the "reftable"
backend outperforms the "files" backend because the "reftable"
backend performs auto-compaction.
Benchmark 1: show-ref: print all refs (refformat = files, refcount = 1, packed = true)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.5 ms … 2.0 ms 1729 runs
Benchmark 2: show-ref: print all refs (refformat = reftable, refcount = 1, packed = true)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.5 ms … 1.8 ms 1816 runs
Benchmark 3: show-ref: print all refs (refformat = files, refcount = 1000, packed = true)
Time (mean ± σ): 4.3 ms ± 0.1 ms [User: 0.9 ms, System: 3.3 ms]
Range (min … max): 4.1 ms … 4.6 ms 645 runs
Benchmark 4: show-ref: print all refs (refformat = reftable, refcount = 1000, packed = true)
Time (mean ± σ): 4.5 ms ± 0.2 ms [User: 1.0 ms, System: 3.3 ms]
Range (min … max): 4.2 ms … 5.9 ms 643 runs
Benchmark 5: show-ref: print all refs (refformat = files, refcount = 1000000, packed = true)
Time (mean ± σ): 2.537 s ± 0.034 s [User: 0.488 s, System: 2.048 s]
Range (min … max): 2.511 s … 2.627 s 10 runs
Benchmark 6: show-ref: print all refs (refformat = reftable, refcount = 1000000, packed = true)
Time (mean ± σ): 2.712 s ± 0.017 s [User: 0.653 s, System: 2.059 s]
Range (min … max): 2.692 s … 2.752 s 10 runs
Benchmark 7: show-ref: print all refs (refformat = files, refcount = 1, packed = false)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.5 ms … 1.9 ms 1834 runs
Benchmark 8: show-ref: print all refs (refformat = reftable, refcount = 1, packed = false)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.4 ms … 2.0 ms 1840 runs
Benchmark 9: show-ref: print all refs (refformat = files, refcount = 1000, packed = false)
Time (mean ± σ): 13.8 ms ± 0.2 ms [User: 2.8 ms, System: 10.8 ms]
Range (min … max): 13.3 ms … 14.5 ms 208 runs
Benchmark 10: show-ref: print all refs (refformat = reftable, refcount = 1000, packed = false)
Time (mean ± σ): 4.5 ms ± 0.2 ms [User: 1.2 ms, System: 3.3 ms]
Range (min … max): 4.3 ms … 6.2 ms 624 runs
Benchmark 11: show-ref: print all refs (refformat = files, refcount = 1000000, packed = false)
Time (mean ± σ): 12.127 s ± 0.129 s [User: 2.675 s, System: 9.451 s]
Range (min … max): 11.965 s … 12.370 s 10 runs
Benchmark 12: show-ref: print all refs (refformat = reftable, refcount = 1000000, packed = false)
Time (mean ± σ): 2.799 s ± 0.022 s [User: 0.735 s, System: 2.063 s]
Range (min … max): 2.769 s … 2.836 s 10 runs
- Printing a single ref shows no real difference between the "files"
and "reftable" backends.
Benchmark 1: show-ref: print single ref (refformat = files, refcount = 1)
Time (mean ± σ): 1.5 ms ± 0.1 ms [User: 0.4 ms, System: 1.0 ms]
Range (min … max): 1.4 ms … 1.8 ms 1779 runs
Benchmark 2: show-ref: print single ref (refformat = reftable, refcount = 1)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.4 ms … 2.5 ms 1753 runs
Benchmark 3: show-ref: print single ref (refformat = files, refcount = 1000)
Time (mean ± σ): 1.5 ms ± 0.1 ms [User: 0.3 ms, System: 1.1 ms]
Range (min … max): 1.4 ms … 1.9 ms 1840 runs
Benchmark 4: show-ref: print single ref (refformat = reftable, refcount = 1000)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.5 ms … 2.0 ms 1831 runs
Benchmark 5: show-ref: print single ref (refformat = files, refcount = 1000000)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.5 ms … 2.1 ms 1848 runs
Benchmark 6: show-ref: print single ref (refformat = reftable, refcount = 1000000)
Time (mean ± σ): 1.6 ms ± 0.1 ms [User: 0.4 ms, System: 1.1 ms]
Range (min … max): 1.5 ms … 2.1 ms 1762 runs
So overall, performance depends on the usecases. Except for many
sequential writes the "reftable" backend is roughly on par or
significantly faster than the "files" backend though. Given that the
"files" backend has received 18 years of optimizations by now this can
be seen as a win. Furthermore, we can expect that the "reftable" backend
will grow faster over time when attention turns more towards
optimizations.
The complete test suite passes, except for those tests explicitly marked
to require the REFFILES prerequisite. Some tests in t0610 are marked as
failing because they depend on still-in-flight bug fixes. Tests can be
run with the new backend by setting the GIT_TEST_DEFAULT_REF_FORMAT
environment variable to "reftable".
There is a single known conceptual incompatibility with the dumb HTTP
transport. As "info/refs" SHOULD NOT contain the HEAD reference, and
because the "HEAD" file is not valid anymore, it is impossible for the
remote client to figure out the default branch without changing the
protocol. This shortcoming needs to be handled in a subsequent patch
series.
As the reftable library has already been introduced a while ago, this
commit message will not go into the details of how exactly the on-disk
format works. Please refer to our preexisting technical documentation at
Documentation/technical/reftable for this.
[1]: https://public-inbox.org/git/CAJo=hJtyof=HRy=2sLP0ng0uZ4=S-DpZ5dR1aF+VHVETKG20OQ@mail.gmail.com/
Original-idea-by: Shawn Pearce <spearce@spearce.org>
Based-on-patch-by: Han-Wen Nienhuys <hanwen@google.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
