Handling of "promisor packs" that allows certain objects to be
missing and lazily retrievable has been optimized (a bit).
* jk/promisor-optim:
revision: avoid parsing with --exclude-promisor-objects
lookup_unknown_object(): take a repository argument
is_promisor_object(): free tree buffer after parsing
When `git repack -A -d` is run in a partial clone, `pack-objects`
is invoked twice: once to repack all promisor objects, and once to
repack all non-promisor objects. The latter `pack-objects` invocation
is with --exclude-promisor-objects and --unpack-unreachable, which
loosens all objects unused during this invocation. Unfortunately,
this includes promisor objects.
Because the -d argument to `git repack` subsequently deletes all loose
objects also in packs, these just-loosened promisor objects will be
immediately deleted. However, this extra disk churn is unnecessary in
the first place. For example, in a newly-cloned partial repo that
filters all blob objects (e.g. `--filter=blob:none`), `repack` ends up
unpacking all trees and commits into the filesystem because every
object, in this particular case, is a promisor object. Depending on
the repo size, this increases the disk usage considerably: In my copy
of the linux.git, the object directory peaked 26GB of more disk usage.
In order to avoid this extra disk churn, pass the names of the promisor
packfiles as --keep-pack arguments to the second invocation of
`pack-objects`. This informs `pack-objects` that the promisor objects
are already in a safe packfile and, therefore, do not need to be
loosened.
For testing, we need to validate whether any object was loosened.
However, the "evidence" (loosened objects) is deleted during the
process which prevents us from inspecting the object directory.
Instead, let's teach `pack-objects` to count loosened objects and
emit via trace2 thus allowing inspecting the debug events after the
process is finished. This new event is used on the added regression
test.
Lastly, add a new perf test to evaluate the performance impact
made by this changes (tested on git.git):
Test HEAD^ HEAD
----------------------------------------------------------
5600.3: gc 134.38(41.93+90.95) 7.80(6.72+1.35) -94.2%
For a bigger repository, such as linux.git, the improvement is
even bigger:
Test HEAD^ HEAD
-------------------------------------------------------------------
5600.3: gc 6833.00(918.07+3162.74) 268.79(227.02+39.18) -96.1%
These improvements are particular big because every object in the
newly-cloned partial repository is a promisor object.
Reported-by: SZEDER Gábor <szeder.dev@gmail.com>
Helped-by: Jeff King <peff@peff.net>
Helped-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Rafael Silva <rafaeloliveira.cs@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We use struct object_id for the names of objects. It isn't intended to
be used for other hash values that don't name objects such as the pack
hash.
Because struct object_id will soon need to have its algorithm member
set, using it in this code path would mean that we didn't set that
member, only the hash member, which would result in a crash. For both
of these reasons, switch to using an unsigned char array of size
GIT_MAX_RAWSZ.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When "git pack-objects" makes a literal copy of a part of existing
packfile using the reachability bitmaps, its update to the progress
meter was broken.
* jk/pack-objects-bitmap-progress-fix:
pack-objects: update "nr_seen" progress based on pack-reused count
When providing an object filter, it is currently impossible to also
filter provided items. E.g. when executing `git rev-list HEAD` , the
commit this reference points to will be treated as user-provided and is
thus excluded from the filtering mechanism. This makes it harder than
necessary to properly use the new `--filter=object:type` filter given
that even if the user wants to only see blobs, he'll still see commits
of provided references.
Improve this by introducing a new `--filter-provided-objects` option
to the git-rev-parse(1) command. If given, then all user-provided
references will be subject to filtering.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
All of the other lookup_foo() functions take a repository argument, but
lookup_unknown_object() was never converted, and it uses the_repository
internally. Let's fix that.
We could leave a wrapper that uses the_repository, but there aren't that
many calls, so we'll just convert them all. I looked briefly at each
site to see if we had a repository struct (besides the_repository) we
could pass, but none of them do (so this conversion to pass
the_repository is a pure noop in each case, though it does take us one
step closer to eventually getting rid of the_repository).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When serving a clone or fetch with bitmaps, after deciding which objects
need to be sent our "pack reuse" mechanism kicks in: we try to send
more-or-less verbatim a bunch of objects from the beginning of the
bitmapped packfile without even adding them to the to_pack.objects
array.
After deciding which objects will be in the "reused" portion, we update
nr_result to account for those, and then trigger display_progress() to
show the user (who is undoubtedly dazzled that we managed to enumerate
so many objects so quickly).
But then something confusing happens: the "Enumerating objects" progress
meter jumps _backwards_, counting up from zero the number of objects we
actually add into to_pack.objects.
This worked correctly once upon a time, but was broken in 5af050437a
(pack-objects: show some progress when counting kept objects,
2018-04-15), when the latter half of that progress meter switched to
using a separate nr_seen counter, rather than nr_result. Nobody noticed
for two reasons:
- prior to the pack-reuse fixes from a14aebeac3 (Merge branch
'jk/packfile-reuse-cleanup', 2020-02-14), the reuse code almost
never kicked in anyway
- the output looks _kind of_ correct. The "backwards" moment is hard
to catch, because we overwrite the old progress number with the new
one, and the larger number is displayed only for a second. So unless
you look at that exact second, you just see the much smaller value,
counting up to the number of non-reused objects (though of course if
you catch it in stderr, or look at GIT_TRACE_PACKET from a server
with bitmaps, you can see both values).
This smaller output isn't wrong per se, but isn't counting what we ever
intended to. We should give the user the whole number of objects we
considered (which, as per 5af050437a's original purpose, is already
_not_ a count of what goes into to_pack.objects). The follow-on
"Counting objects" meter shows the actual number of objects we feed into
that array.
We can easily fix this by bumping (and showing) nr_seen for the
pack-reused objects. When the included test is run without this patch,
the second pack-objects invocation produces "Enumerating objects: 1" to
show the one loose object, even though the resulting pack has hundreds
of objects in it. With it, we jump to "Enumerating objects: 674" after
deciding on reuse, and then "675" when we add in the loose object.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When writing a new pack with a bitmap, it is sometimes convenient to
indicate some reference prefixes which should receive priority when
selecting which commits to receive bitmaps.
A truly motivated caller could accomplish this by setting
'pack.islandCore', (since all commits in the core island are similarly
marked as preferred) but this requires callers to opt into using delta
islands, which they may or may not want to do.
Introduce a new multi-valued configuration, 'pack.preferBitmapTips' to
allow callers to specify a list of reference prefixes. All references
which have a prefix contained in 'pack.preferBitmapTips' will mark their
tips as "preferred" in the same way as commits are marked as preferred
for selection by 'pack.islandCore'.
The choice of the verb "prefer" is intentional: marking the NEEDS_BITMAP
flag on an object does *not* guarantee that that object will receive a
bitmap. It merely guarantees that that commit will receive a bitmap over
any *other* commit in the same window by bitmap_writer_select_commits().
The test this patch adds reflects this quirk, too. It only tests that
a commit (which didn't receive bitmaps by default) is selected for
bitmaps after changing the value of 'pack.preferBitmapTips' to include
it. Other commits may lose their bitmaps as a byproduct of how the
selection process works (bitmap_writer_select_commits() ignores the
remainder of a window after seeing a commit with the NEEDS_BITMAP flag).
This configuration will aide in selecting important references for
multi-pack bitmaps, since they do not respect the same pack.islandCore
configuration. (They could, but doing so may be confusing, since it is
packs--not bitmaps--which are influenced by the delta-islands
configuration).
In a fork network repository (one which lists all forks of a given
repository as remotes), for example, it is useful to set
pack.preferBitmapTips to 'refs/remotes/<root>/heads' and
'refs/remotes/<root>/tags', where '<root>' is an opaque identifier
referring to the repository which is at the base of the fork chain.
Suggested-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git repack" so far has been only capable of repacking everything
under the sun into a single pack (or split by size). A cleverer
strategy to reduce the cost of repacking a repository has been
introduced.
* tb/geometric-repack:
builtin/pack-objects.c: ignore missing links with --stdin-packs
builtin/repack.c: reword comment around pack-objects flags
builtin/repack.c: be more conservative with unsigned overflows
builtin/repack.c: assign pack split later
t7703: test --geometric repack with loose objects
builtin/repack.c: do not repack single packs with --geometric
builtin/repack.c: add '--geometric' option
packfile: add kept-pack cache for find_kept_pack_entry()
builtin/pack-objects.c: rewrite honor-pack-keep logic
p5303: measure time to repack with keep
p5303: add missing &&-chains
builtin/pack-objects.c: add '--stdin-packs' option
revision: learn '--no-kept-objects'
packfile: introduce 'find_kept_pack_entry()'
As record_reused_object(offset, offset - hashfile_total(out)) said,
reused_chunk.difference should be the offset of original packfile minus
the offset of the generated packfile. But the comment presented an opposite way.
Signed-off-by: Han Xin <hanxin.hx@alibaba-inc.com>
Acked-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When 'git pack-objects --stdin-packs' encounters a commit in a pack, it
marks it as a starting point of a best-effort reachability traversal
that is used to populate the name-hash of the objects listed in the
given packs.
The traversal expects that it should be able to walk the ancestors of
all commits in a pack without issue. Ordinarily this is the case, but it
is possible to having missing parents from an unreachable part of the
repository. In that case, we'd consider any missing objects in the
unreachable portion of the graph to be junk.
This should be handled gracefully: since the traversal is best-effort
(i.e., we don't strictly need to fill in all of the name-hash fields),
we should simply ignore any missing links.
This patch does that (by setting the 'ignore_missing_links' bit on the
rev_info struct), and ensures we don't regress in the future by adding a
test which demonstrates this case.
It is a little over-eager, since it will also ignore missing links in
reachable parts of the packs (which would indicate a corrupted
repository), but '--stdin-packs' is explicitly *not* about reachability.
So this step isn't making anything worse for a repository which contains
packs missing reachable objects (since we never drop objects with
'--stdin-packs').
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add and apply a semantic patch for converting code that open-codes
CALLOC_ARRAY to use it instead. It shortens the code and infers the
element size automatically.
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Now that we have find_kept_pack_entry(), we don't have to manually keep
hunting through every pack to find a possible "kept" duplicate of the
object. This should be faster, assuming only a portion of your total
packs are actually kept.
Note that we have to re-order the logic a bit here; we can deal with the
disqualifying situations first (e.g., finding the object in a non-local
pack with --local), then "kept" situation(s), and then just fall back to
other "--local" conditions.
Here are the results from p5303 (measurements again taken on the
kernel):
Test HEAD^ HEAD
-----------------------------------------------------------------------------------------------
5303.5: repack (1) 57.26(54.59+10.84) 57.34(54.66+10.88) +0.1%
5303.6: repack with kept (1) 57.33(54.80+10.51) 57.38(54.83+10.49) +0.1%
5303.11: repack (50) 71.54(88.57+4.84) 71.70(88.99+4.74) +0.2%
5303.12: repack with kept (50) 85.12(102.05+4.94) 72.58(89.61+4.78) -14.7%
5303.17: repack (1000) 216.87(490.79+14.57) 217.19(491.72+14.25) +0.1%
5303.18: repack with kept (1000) 665.63(938.87+15.76) 246.12(520.07+14.93) -63.0%
and the --stdin-packs timings:
5303.7: repack with --stdin-packs (1) 0.01(0.01+0.00) 0.00(0.00+0.00) -100.0%
5303.13: repack with --stdin-packs (50) 3.53(12.07+0.24) 3.43(11.75+0.24) -2.8%
5303.19: repack with --stdin-packs (1000) 195.83(371.82+8.10) 130.50(307.15+7.66) -33.4%
So our repack with an empty .keep pack is roughly as fast as one without
a .keep pack up to 50 packs. But the --stdin-packs case scales a little
better, too.
Notably, it is faster than a repack of the same size and a kept pack. It
looks at fewer objects, of course, but the penalty for looking at many
packs isn't as costly.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In an upcoming commit, 'git repack' will want to create a pack comprised
of all of the objects in some packs (the included packs) excluding any
objects in some other packs (the excluded packs).
This caller could iterate those packs themselves and feed the objects it
finds to 'git pack-objects' directly over stdin, but this approach has a
few downsides:
- It requires every caller that wants to drive 'git pack-objects' in
this way to implement pack iteration themselves. This forces the
caller to think about details like what order objects are fed to
pack-objects, which callers would likely rather not do.
- If the set of objects in included packs is large, it requires
sending a lot of data over a pipe, which is inefficient.
- The caller is forced to keep track of the excluded objects, too, and
make sure that it doesn't send any objects that appear in both
included and excluded packs.
But the biggest downside is the lack of a reachability traversal.
Because the caller passes in a list of objects directly, those objects
don't get a namehash assigned to them, which can have a negative impact
on the delta selection process, causing 'git pack-objects' to fail to
find good deltas even when they exist.
The caller could formulate a reachability traversal themselves, but the
only way to drive 'git pack-objects' in this way is to do a full
traversal, and then remove objects in the excluded packs after the
traversal is complete. This can be detrimental to callers who care
about performance, especially in repositories with many objects.
Introduce 'git pack-objects --stdin-packs' which remedies these four
concerns.
'git pack-objects --stdin-packs' expects a list of pack names on stdin,
where 'pack-xyz.pack' denotes that pack as included, and
'^pack-xyz.pack' denotes it as excluded. The resulting pack includes all
objects that are present in at least one included pack, and aren't
present in any excluded pack.
To address the delta selection problem, 'git pack-objects --stdin-packs'
works as follows. First, it assembles a list of objects that it is going
to pack, as above. Then, a reachability traversal is started, whose tips
are any commits mentioned in included packs. Upon visiting an object, we
find its corresponding object_entry in the to_pack list, and set its
namehash parameter appropriately.
To avoid the traversal visiting more objects than it needs to, the
traversal is halted upon encountering an object which can be found in an
excluded pack (by marking the excluded packs as kept in-core, and
passing --no-kept-objects=in-core to the revision machinery).
This can cause the traversal to halt early, for example if an object in
an included pack is an ancestor of ones in excluded packs. But stopping
early is OK, since filling in the namehash fields of objects in the
to_pack list is only additive (i.e., having it helps the delta selection
process, but leaving it blank doesn't impact the correctness of the
resulting pack).
Even still, it is unlikely that this hurts us much in practice, since
the 'git repack --geometric' caller (which is introduced in a later
commit) marks small packs as included, and large ones as excluded.
During ordinary use, the small packs usually represent pushes after a
large repack, and so are unlikely to be ancestors of objects that
already exist in the repository.
(I found it convenient while developing this patch to have 'git
pack-objects' report the number of objects which were visited and got
their namehash fields filled in during traversal. This is also included
in the below patch via trace2 data lines).
Suggested-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Reviewed-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Introduce an on-disk file to record revindex for packdata, which
traditionally was always created on the fly and only in-core.
* tb/pack-revindex-on-disk:
t5325: check both on-disk and in-memory reverse index
pack-revindex: ensure that on-disk reverse indexes are given precedence
t: support GIT_TEST_WRITE_REV_INDEX
t: prepare for GIT_TEST_WRITE_REV_INDEX
Documentation/config/pack.txt: advertise 'pack.writeReverseIndex'
builtin/pack-objects.c: respect 'pack.writeReverseIndex'
builtin/index-pack.c: write reverse indexes
builtin/index-pack.c: allow stripping arbitrary extensions
pack-write.c: prepare to write 'pack-*.rev' files
packfile: prepare for the existence of '*.rev' files
The "pack-objects" command needs to iterate over all the tags when
automatic tag following is enabled, but it actually iterated over
all refs and then discarded everything outside "refs/tags/"
hierarchy, which was quite wasteful.
* jv/pack-objects-narrower-ref-iteration:
builtin/pack-objects.c: avoid iterating all refs
Add a new option that unconditionally enables the pack.writeReverseIndex
setting in order to run the whole test suite in a mode that generates
on-disk reverse indexes. Additionally, enable this mode in the second
run of tests under linux-gcc in 'ci/run-build-and-tests.sh'.
Once on-disk reverse indexes are proven out over several releases, we
can change the default value of that configuration to 'true', and drop
this patch.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Now that we have an implementation that can write the new reverse index
format, enable writing a .rev file in 'git pack-objects' by consulting
the pack.writeReverseIndex configuration variable.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In git-pack-objects, we iterate over all the tags if the --include-tag
option is passed on the command line. For some reason this uses
for_each_ref which is expensive if the repo has many refs. We should
use for_each_tag_ref instead.
Because the add_ref_tag callback will now only visit tags we
simplified it a bit.
The motivation for this change is that we observed performance issues
with a repository on gitlab.com that has 500,000 refs but only 2,000
tags. The fetch traffic on that repo is dominated by CI, and when we
changed CI to fetch with 'git fetch --no-tags' we saw a dramatic
change in the CPU profile of git-pack-objects. This lead us to this
particular ref walk. More details in:
https://gitlab.com/gitlab-com/gl-infra/scalability/-/issues/746#note_483546598
Signed-off-by: Jacob Vosmaer <jacob@gitlab.com>
Reviewed-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The peel_ref() interface is confusing and error-prone:
- it's typically used by ref iteration callbacks that have both a
refname and oid. But since they pass only the refname, we may load
the ref value from the filesystem again. This is inefficient, but
also means we are open to a race if somebody simultaneously updates
the ref. E.g., this:
int some_ref_cb(const char *refname, const struct object_id *oid, ...)
{
if (!peel_ref(refname, &peeled))
printf("%s peels to %s",
oid_to_hex(oid), oid_to_hex(&peeled);
}
could print nonsense. It is correct to say "refname peels to..."
(you may see the "before" value or the "after" value, either of
which is consistent), but mentioning both oids may be mixing
before/after values.
Worse, whether this is possible depends on whether the optimization
to read from the current iterator value kicks in. So it is actually
not possible with:
for_each_ref(some_ref_cb);
but it _is_ possible with:
head_ref(some_ref_cb);
which does not use the iterator mechanism (though in practice, HEAD
should never peel to anything, so this may not be triggerable).
- it must take a fully-qualified refname for the read_ref_full() code
path to work. Yet we routinely pass it partial refnames from
callbacks to for_each_tag_ref(), etc. This happens to work when
iterating because there we do not call read_ref_full() at all, and
only use the passed refname to check if it is the same as the
iterator. But the requirements for the function parameters are quite
unclear.
Instead of taking a refname, let's instead take an oid. That fixes both
problems. It's a little funny for a "ref" function not to involve refs
at all. The key thing is that it's optimizing under the hood based on
having access to the ref iterator. So let's change the name to make it
clear why you'd want this function versus just peel_object().
There are two other directions I considered but rejected:
- we could pass the peel information into the each_ref_fn callback.
However, we don't know if the caller actually wants it or not. For
packed-refs, providing it is essentially free. But for loose refs,
we actually have to peel the object, which would be wasteful in most
cases. We could likewise pass in a flag to the callback indicating
whether the peeled information is known, but that complicates those
callbacks, as they then have to decide whether to manually peel
themselves. Plus it requires changing the interface of every
callback, whether they care about peeling or not, and there are many
of them.
- we could make a function to return the peeled value of the current
iterated ref (computing it if necessary), and BUG() otherwise. I.e.:
int peel_current_iterated_ref(struct object_id *out);
Each of the current callers is an each_ref_fn callback, so they'd
mostly be happy. But:
- we use those callbacks with functions like head_ref(), which do
not use the iteration code. So we'd need to handle the fallback
case there, anyway.
- it's possible that a caller would want to call into generic code
that sometimes is used during iteration and sometimes not. This
encapsulates the logic to do the fast thing when possible, and
fallback when necessary.
The implementation is mostly obvious, but I want to call out a few
things in the patch:
- the test-tool coverage for peel_ref() is now meaningless, as it all
collapses to a single peel_object() call (arguably they were pretty
uninteresting before; the tricky part of that function is the
fast-path we see during iteration, but these calls didn't trigger
that). I've just dropped it entirely, though note that some other
tests relied on the tags we created; I've moved that creation to the
tests where it matters.
- we no longer need to take a ref_store parameter, since we'd never
look up a ref now. We do still rely on a global "current iterator"
variable which _could_ be kept per-ref-store. But in practice this
is only useful if there are multiple recursive iterations, at which
point the more appropriate solution is probably a stack of
iterators. No caller used the actual ref-store parameter anyway
(they all call the wrapper that passes the_repository).
- the original only kicked in the optimization when the "refname"
pointer matched (i.e., not string comparison). We do likewise with
the "oid" parameter here, but fall back to doing an actual oideq()
call. This in theory lets us kick in the optimization more often,
though in practice no current caller cares. It should never be
wrong, though (peeling is a property of an object, so two refs
pointing to the same object would peel identically).
- the original took care not to touch the peeled out-parameter unless
we found something to put in it. But no caller cares about this, and
anyway, it is enforced by peel_object() itself (and even in the
optimized iterator case, that's where we eventually end up). We can
shorten the code and avoid an extra copy by just passing the
out-parameter through the stack.
Signed-off-by: Jeff King <peff@peff.net>
Reviewed-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Replace direct accesses to the revindex with calls to
'offset_to_pack_pos()' and 'pack_pos_to_index()'.
Since this caller already had some error checking (it can jump to the
'give_up' label if it encounters an error), we can easily check whether
or not the provided offset points to an object in the given pack. This
error checking existed prior to this patch, too, since the caller checks
whether the return value from 'find_pack_revindex()' was NULL or not.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Replace a direct access to the revindex array with
'pack_pos_to_offset()'.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Replace direct revindex accesses with calls to 'pack_pos_to_offset()'
and 'pack_pos_to_index()'.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
First replace 'find_pack_revindex()' with its replacement
'offset_to_pack_pos()'. This prevents any bogus OFS_DELTA that may make
its way through until 'write_reuse_object()' from causing a bad memory
read (if 'revidx' is 'NULL')
Next, replace a direct access of '->nr' with the wrapper function
'pack_pos_to_index()'.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The on-disk bitmap format has a flag to mark a bitmap to be "reused".
This is a rather curious feature, and works like this:
- a run of pack-objects would decide to mark the last 80% of the
bitmaps it generates with the reuse flag
- the next time we generate bitmaps, we'd see those reuse flags from
the last run, and mark those commits as special:
- we'd be more likely to select those commits to get bitmaps in
the new output
- when generating the bitmap for a selected commit, we'd reuse the
old bitmap as-is (rearranging the bits to match the new pack, of
course)
However, neither of these behaviors particularly makes sense.
Just because a commit happened to be bitmapped last time does not make
it a good candidate for having a bitmap this time. In particular, we may
choose bitmaps based on how recent they are in history, or whether a ref
tip points to them, and those things will change. We're better off
re-considering fresh which commits are good candidates.
Reusing the existing bitmap _is_ a reasonable thing to do to save
computation. But only reusing exact bitmaps is a weak form of this. If
we have an old bitmap for A and now want a new bitmap for its child, we
should be able to compute that only by looking at trees and that are new
to the child. But this code would consider only exact reuse (which is
perhaps why it was eager to select those commits in the first place).
Furthermore, the recent switch to the reverse-edge algorithm for
generating bitmaps dropped this optimization entirely (and yet still
performs better).
So let's do a few cleanups:
- drop the whole "reusing bitmaps" phase of generating bitmaps. It's
not helping anything, and is mostly unused code (or worse, code that
is using CPU but not doing anything useful)
- drop the use of the on-disk reuse flag to select commits to bitmap
- stop setting the on-disk reuse flag in bitmaps we generate (since
nothing respects it anymore)
We will keep a few innards of the reuse code, which will help us
implement a more capable version of the "reuse" optimization:
- simplify rebuild_existing_bitmaps() into a function that only builds
the mapping of bits between the old and new orders, but doesn't
actually convert any bitmaps
- make rebuild_bitmap() public; we'll call it lazily to convert bitmaps
as we traverse (using the mapping created above)
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A few end-user facing messages have been updated to be
hash-algorithm agnostic.
* jc/object-names-are-not-sha-1:
messages: avoid SHA-1 in end-user facing messages
There are still a handful mentions of SHA-1 when we meant the
(hexadecimal) object names in end-user facing messages. Rewrite
them.
I was hoping that this can mostly be s/SHA-1/object name/, but
a few messages needed rephrasing to keep the result readable.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A new helper function has_object() has been introduced to make it
easier to mark object existence checks that do and don't want to
trigger lazy fetches, and a few such checks are converted using it.
* jt/has_object:
fsck: do not lazy fetch known non-promisor object
pack-objects: no fetch when allow-{any,promisor}
apply: do not lazy fetch when applying binary
sha1-file: introduce no-lazy-fetch has_object()
The argv_array API is useful for not just managing argv but any
"vector" (NULL-terminated array) of strings, and has seen adoption
to a certain degree. It has been renamed to "strvec" to reduce the
barrier to adoption.
* jk/strvec:
strvec: rename struct fields
strvec: drop argv_array compatibility layer
strvec: update documention to avoid argv_array
strvec: fix indentation in renamed calls
strvec: convert remaining callers away from argv_array name
strvec: convert more callers away from argv_array name
strvec: convert builtin/ callers away from argv_array name
quote: rename sq_dequote_to_argv_array to mention strvec
strvec: rename files from argv-array to strvec
argv-array: rename to strvec
argv-array: use size_t for count and alloc
The options --missing=allow-{any,promisor} were introduced in caf3827e2f
("rev-list: add list-objects filtering support", 2017-11-22) with the
following note in the commit message:
This patch introduces handling of missing objects to help
debugging and development of the "partial clone" mechanism,
and once the mechanism is implemented, for a power user to
perform operations that are missing-object aware without
incurring the cost of checking if a missing link is expected.
The idea that these options are missing-object aware (and thus do not
need to lazily fetch objects, unlike unaware commands that assume that
all objects are present) are assumed in later commits such as 07ef3c6604
("fetch test: use more robust test for filtered objects", 2020-01-15).
However, the current implementations of these options use
has_object_file(), which indeed lazily fetches missing objects. Teach
these implementations not to do so. Also, update the documentation of
these options to be clearer.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "argc" and "argv" names made sense when the struct was argv_array,
but now they're just confusing. Let's rename them to "nr" (which we use
for counts elsewhere) and "v" (which is rather terse, but reads well
when combined with typical variable names like "args.v").
Note that we have to update all of the callers immediately. Playing
tricks with the preprocessor is hard here, because we wouldn't want to
rewrite unrelated tokens.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We eventually want to drop the argv_array name and just use strvec
consistently. There's no particular reason we have to do it all at once,
or care about interactions between converted and unconverted bits.
Because of our preprocessor compat layer, the names are interchangeable
to the compiler (so even a definition and declaration using different
names is OK).
This patch converts all of the files in builtin/ to keep the diff to a
manageable size.
The conversion was done purely mechanically with:
git ls-files '*.c' '*.h' |
xargs perl -i -pe '
s/ARGV_ARRAY/STRVEC/g;
s/argv_array/strvec/g;
'
and then selectively staging files with "git add builtin/". We'll deal
with any indentation/style fallouts separately.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This requires updating #include lines across the code-base, but that's
all fairly mechanical, and was done with:
git ls-files '*.c' '*.h' |
xargs perl -i -pe 's/argv-array.h/strvec.h/'
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When an object to be packed is noticed to be missing, prefetch all
to-be-packed objects in one batch.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use oid_object_info_extended() instead of oid_object_info() because a
subsequent commit needs to specify an additional flag here.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Teach upload-pack to send part of its packfile response as URIs.
An administrator may configure a repository with one or more
"uploadpack.blobpackfileuri" lines, each line containing an OID, a pack
hash, and a URI. A client may configure fetch.uriprotocols to be a
comma-separated list of protocols that it is willing to use to fetch
additional packfiles - this list will be sent to the server. Whenever an
object with one of those OIDs would appear in the packfile transmitted
by upload-pack, the server may exclude that object, and instead send the
URI. The client will then download the packs referred to by those URIs
before performing the connectivity check.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There are many functions in commit.h that are more related to shallow
repositories than they are to any sort of generic commit machinery.
Likely this began when there were only a few shallow-related functions,
and commit.h seemed a reasonable enough place to put them.
But, now there are a good number of shallow-related functions, and
placing them all in 'commit.h' doesn't make sense.
This patch extracts a 'shallow.h', which takes all of the declarations
from 'commit.h' for functions which already exist in 'shallow.c'. We
will bring the remaining shallow-related functions defined in 'commit.c'
in a subsequent patch.
For now, move only the ones that already are implemented in 'shallow.c',
and update the necessary includes.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the codebase, there are many options which use OPTION_CALLBACK in a
plain ol' struct definition. However, we have the OPT_CALLBACK and
OPT_CALLBACK_F macros which are meant to abstract these plain struct
definitions away. These macros are useful as they semantically signal to
developers that these are just normal callback option with nothing fancy
happening.
Replace plain struct definitions of OPTION_CALLBACK with OPT_CALLBACK or
OPT_CALLBACK_F where applicable. The heavy lifting was done using the
following (disgusting) shell script:
#!/bin/sh
do_replacement () {
tr '\n' '\r' |
sed -e 's/{\s*OPTION_CALLBACK,\s*\([^,]*\),\([^,]*\),\([^,]*\),\([^,]*\),\([^,]*\),\s*0,\(\s*[^[:space:]}]*\)\s*}/OPT_CALLBACK(\1,\2,\3,\4,\5,\6)/g' |
sed -e 's/{\s*OPTION_CALLBACK,\s*\([^,]*\),\([^,]*\),\([^,]*\),\([^,]*\),\([^,]*\),\([^,]*\),\(\s*[^[:space:]}]*\)\s*}/OPT_CALLBACK_F(\1,\2,\3,\4,\5,\6,\7)/g' |
tr '\r' '\n'
}
for f in $(git ls-files \*.c)
do
do_replacement <"$f" >"$f.tmp"
mv "$f.tmp" "$f"
done
The result was manually inspected and then reformatted to match the
style of the surrounding code. Finally, using
`git grep OPTION_CALLBACK \*.c`, leftover results which were not handled
by the script were manually transformed.
Signed-off-by: Denton Liu <liu.denton@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code cleanup.
* jk/oid-array-cleanups:
oidset: stop referring to sha1-array
ref-filter: stop referring to "sha1 array"
bisect: stop referring to sha1_array
test-tool: rename sha1-array to oid-array
oid_array: rename source file from sha1-array
oid_array: use size_t for iteration
oid_array: use size_t for count and allocation
We renamed the actual data structure in 910650d2f8 (Rename sha1_array to
oid_array, 2017-03-31), but the file is still called sha1-array. Besides
being slightly confusing, it makes it more annoying to grep for leftover
occurrences of "sha1" in various files, because the header is included
in so many places.
Let's complete the transition by renaming the source and header files
(and fixing up a few comment references).
I kept the "-" in the name, as that seems to be our style; cf.
fc1395f4a4 (sha1_file.c: rename to use dash in file name, 2018-04-10).
We also have oidmap.h and oidset.h without any punctuation, but those
are "struct oidmap" and "struct oidset" in the code. We _could_ make
this "oidarray" to match, but somehow it looks uglier to me because of
the length of "array" (plus it would be a very invasive patch for little
gain).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The 'pack.useSparse' configuration variable now defaults to 'true',
enabling an optimization that has been experimental since Git 2.21.
* ds/default-pack-use-sparse-to-true:
pack-objects: flip the use of GIT_TEST_PACK_SPARSE
config: set pack.useSparse=true by default
SHA-256 transition continues.
* bc/sha-256-part-1-of-4: (22 commits)
fast-import: add options for rewriting submodules
fast-import: add a generic function to iterate over marks
fast-import: make find_marks work on any mark set
fast-import: add helper function for inserting mark object entries
fast-import: permit reading multiple marks files
commit: use expected signature header for SHA-256
worktree: allow repository version 1
init-db: move writing repo version into a function
builtin/init-db: add environment variable for new repo hash
builtin/init-db: allow specifying hash algorithm on command line
setup: allow check_repository_format to read repository format
t/helper: make repository tests hash independent
t/helper: initialize repository if necessary
t/helper/test-dump-split-index: initialize git repository
t6300: make hash algorithm independent
t6300: abstract away SHA-1-specific constants
t: use hash-specific lookup tables to define test constants
repository: require a build flag to use SHA-256
hex: add functions to parse hex object IDs in any algorithm
hex: introduce parsing variants taking hash algorithms
...
The environment variable GIT_TEST_PACK_SPARSE was previously used
to allow testing the --sparse option for "git pack-objects" in
the test suite. This allowed interesting cases of "git push" to
also test this algorithm.
Since pack.useSparse is now true by default, we do not need this
variable to _enable_ the --sparse option, but instead to _disable_
it. This flips how we work with the variable a bit.
When checking for the variable, default to a value of -1 for
"unset". If unset, then take the default from the repo settings,
which is currently 1. Then, the --[no-]sparse command-line option
will override either of these settings.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Code cleanup to use "struct object_id" more by replacing use of
"char *sha1"
* jk/nth-packed-object-id:
packfile: drop nth_packed_object_sha1()
packed_object_info(): use object_id internally for delta base
packed_object_info(): use object_id for returning delta base
pack-check: push oid lookup into loop
pack-check: convert "internal error" die to a BUG()
pack-bitmap: use object_id when loading on-disk bitmaps
pack-objects: use object_id struct in pack-reuse code
pack-objects: convert oe_set_delta_ext() to use object_id
pack-objects: read delta base oid into object_id struct
nth_packed_object_oid(): use customary integer return
The object reachability bitmap machinery and the partial cloning
machinery were not prepared to work well together, because some
object-filtering criteria that partial clones use inherently rely
on object traversal, but the bitmap machinery is an optimization
to bypass that object traversal. There however are some cases
where they can work together, and they were taught about them.
* jk/object-filter-with-bitmap:
rev-list --count: comment on the use of count_right++
pack-objects: support filters with bitmaps
pack-bitmap: implement BLOB_LIMIT filtering
pack-bitmap: implement BLOB_NONE filtering
bitmap: add bitmap_unset() function
rev-list: use bitmap filters for traversal
pack-bitmap: basic noop bitmap filter infrastructure
rev-list: allow commit-only bitmap traversals
t5310: factor out bitmap traversal comparison
rev-list: allow bitmaps when counting objects
rev-list: make --count work with --objects
rev-list: factor out bitmap-optimized routines
pack-bitmap: refuse to do a bitmap traversal with pathspecs
rev-list: fallback to non-bitmap traversal when filtering
pack-bitmap: fix leak of haves/wants object lists
pack-bitmap: factor out type iterator initialization
When the pack-reuse code is dumping an OFS_DELTA entry to a client that
doesn't support it, we re-write it as a REF_DELTA. To do so, we use
nth_packed_object_sha1() to get the oid, but that function is soon going
away in favor of the more type-safe nth_packed_object_id(). Let's switch
now in preparation.
Note that this does incur an extra hash copy (from the pack idx mmap to
the object_id and then to the output, rather than straight from mmap to
the output). But this is not worth worrying about. It's probably not
measurable even when it triggers, and this is fallback code that we
expect to trigger very rarely (since everybody supports OFS_DELTA these
days anyway).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We already store an object_id internally, and now our sole caller also
has one. Let's stop passing around the internal hash array, which adds a
bit of type safety.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we're considering reusing an on-disk delta, we get the oid of the
base as a pointer to unsigned char bytes of the hash, either into the
packfile itself (for REF_DELTA) or into the pack idx (using the revindex
to convert the offset into an index entry).
Instead, we'd prefer to use a more type-safe object_id as much as
possible. We can get the pack idx using nth_packed_object_id() instead.
For the packfile bytes, we can copy them out using oidread().
This doesn't even incur an extra copy overall, since the next thing we'd
always do with that pointer is pass it to can_reuse_delta(), which needs
an object_id anyway (and called oidread() itself). So this patch also
converts that function to take the object_id directly.
Note that we did previously use NULL as a sentinel value when the object
isn't a delta. We could probably get away with using the null oid for
this, but instead we'll use an explicit boolean flag, which should make
things more obvious for people reading the code later.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Our nth_packed_object_sha1() function returns NULL for error. So when we
wrapped it with nth_packed_object_oid(), we kept the same semantics. But
it's a bit funny, because the caller actually passes in an out
parameter, and the pointer we return is just that same struct they
passed to us (or NULL).
It's not too terrible, but it does make the interface a little
non-idiomatic. Let's switch to our usual "0 for success, negative for
error" return value. Most callers either don't check it, or are
trivially converted. The one that requires the biggest change is
actually improved, as we can ditch an extra aliased pointer variable.
Since we are changing the interface in a subtle way that the compiler
wouldn't catch, let's also change the name to catch any topics in
flight. We can drop the 'o' and make it nth_packed_object_id(). That's
slightly shorter, but also less redundant since the 'o' stands for
"object" already.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Avoid hard-coding a hash size, instead preferring to use the_hash_algo.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some codepaths were given a repository instance as a parameter to
work in the repository, but passed the_repository instance to its
callees, which has been cleaned up (somewhat).
* mt/use-passed-repo-more-in-funcs:
sha1-file: allow check_object_signature() to handle any repo
sha1-file: pass git_hash_algo to hash_object_file()
sha1-file: pass git_hash_algo to write_object_file_prepare()
streaming: allow open_istream() to handle any repo
pack-check: use given repo's hash_algo at verify_packfile()
cache-tree: use given repo's hash_algo at verify_one()
diff: make diff_populate_filespec() honor its repo argument
Just as rev-list recently learned to combine filters and bitmaps, let's
do the same for pack-objects. The infrastructure is all there; we just
need to pass along our filter options, and the pack-bitmap code will
decide to use bitmaps or not.
This unsurprisingly makes things faster for partial clones of large
repositories (here we're cloning linux.git):
Test HEAD^ HEAD
------------------------------------------------------------------------------
5310.11: simulated partial clone 38.94(37.28+5.87) 11.06(11.27+4.07) -71.6%
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently you can't use object filters with bitmaps, but we plan to
support at least some filters with bitmaps. Let's introduce some
infrastructure that will help us do that:
- prepare_bitmap_walk() now accepts a list_objects_filter_options
parameter (which can be NULL for no filtering; all the current
callers pass this)
- we'll bail early if the filter is incompatible with bitmaps (just as
we would if there were no bitmaps at all). Currently all filters are
incompatible.
- we'll filter the resulting bitmap; since there are no supported
filters yet, this is always a noop.
There should be no behavior change yet, but we'll support some actual
filters in a future patch.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Ever since we added reachability bitmap support, we've been able to use
it with rev-list to get the full list of objects, like:
git rev-list --objects --use-bitmap-index --all
But you can't do so without --objects, since we weren't ready to just
show the commits. However, the internals of the bitmap code are mostly
ready for this: they avoid opening up trees when walking to fill in the
bitmaps. We just need to actually pass in the rev_info to
traverse_bitmap_commit_list() so it knows which types to bother
triggering our callback for.
For completeness, the perf test now covers both the existing --objects
case, as well as the new commits-only behavior (the objects one got way
faster when we introduced bitmaps, but obviously isn't improved now).
Here are numbers for linux.git:
Test HEAD^ HEAD
------------------------------------------------------------------------
5310.7: rev-list (commits) 8.29(8.10+0.19) 1.76(1.72+0.04) -78.8%
5310.8: rev-list (objects) 8.06(7.94+0.12) 8.14(7.94+0.13) +1.0%
That run was cheating a little, as I didn't have any commit-graph in the
repository, and we'd built it by default these days when running git-gc.
Here are numbers with a commit-graph:
Test HEAD^ HEAD
------------------------------------------------------------------------
5310.7: rev-list (commits) 0.70(0.58+0.12) 0.51(0.46+0.04) -27.1%
5310.8: rev-list (objects) 6.20(6.09+0.10) 6.27(6.16+0.11) +1.1%
Still an improvement, but a lot less impressive.
We could have the perf script remove any commit-graph to show the
out-sized effect, but it probably makes sense to leave it in what would
be a more typical setup.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some callers of open_istream() at archive-tar.c and archive-zip.c are
capable of working on arbitrary repositories but the repo struct is not
passed down to open_istream(), which uses the_repository internally. For
now, that's not a problem since the said callers are only being called
with the_repository. But to be consistent and avoid future problems,
let's allow open_istream() to receive a struct repository and use that
instead of the_repository. This parameter addition will also be used in
a future patch to make sha1-file.c:check_object_signature() be able to
work on arbitrary repos.
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Additional checks are added in have_duplicate_entry() and
obj_is_packed() to avoid duplicate objects in the reuse
bitmap. It was probably buggy to not have such a check
before.
Git as a client would never both asks for a tag by sha1 and
specify "include-tag", but libgit2 will, so a libgit2 client
cloning from a Git server would trigger the bug.
If a client both asks for a tag by sha1 and specifies
"include-tag", we may end up including the tag in the reuse
bitmap (due to the first thing), and then later adding it to
the packlist (due to the second). This results in duplicate
objects in the pack, which git chokes on. We should notice
that we are already including it when doing the include-tag
portion, and avoid adding it to the packlist.
The simplest place to fix this is right in add_ref_tag(),
where we could avoid peeling the tag at all if we know that
we are already including it. However, this pushes the check
instead into have_duplicate_entry(). This fixes not only
this case, but also means that we cannot have any similar
problems lurking in other code.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The old code to reuse deltas from an existing packfile
just tried to dump a whole segment of the pack verbatim.
That's faster than the traditional way of actually adding
objects to the packing list, but it didn't kick in very
often. This new code is really going for a middle ground:
do _some_ per-object work, but way less than we'd
traditionally do.
The general strategy of the new code is to make a bitmap
of objects from the packfile we'll include, and then
iterate over it, writing out each object exactly as it is
in our on-disk pack, but _not_ adding it to our packlist
(which costs memory, and increases the search space for
deltas).
One complication is that if we're omitting some objects,
we can't set a delta against a base that we're not
sending. So we have to check each object in
try_partial_reuse() to make sure we have its delta.
About performance, in the worst case we might have
interleaved objects that we are sending or not sending,
and we'd have as many chunks as objects. But in practice
we send big chunks.
For instance, packing torvalds/linux on GitHub servers
now reused 6.5M objects, but only needed ~50k chunks.
Helped-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Let's refactor the way we check if an object is packed by
introducing obj_is_packed(). This function is now a simple
wrapper around packlist_find(), but it will evolve in a
following commit.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Let's make it possible to configure if we want pack reuse or not.
The main reason it might not be wanted is probably debugging and
performance testing, though pack reuse _might_ cause larger packs,
because we wouldn't consider the reused objects as bases for
finding new deltas.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Various fixes to codepaths gcc 9 had trouble following dataflow.
* jk/misc-uninitialized-fixes:
pack-objects: drop packlist index_pos optimization
test-read-cache: drop namelen variable
diff-delta: set size out-parameter to 0 for NULL delta
bulk-checkin: zero-initialize hashfile_checkpoint
pack-objects: use object_id in packlist_alloc()
git-am: handle missing "author" when parsing commit
xmalloc() used to have a mechanism to ditch memory and address
space resources as the last resort upon seeing an allocation
failure from the underlying malloc(), which made the code complex
and thread-unsafe with dubious benefit, as major memory resource
users already do limit their uses with various other mechanisms.
It has been simplified away.
* jk/drop-release-pack-memory:
packfile: drop release_pack_memory()
To see when packfile reuse kicks in or not, it is useful to
show reused packfile objects statistics in the output of
upload-pack.
Helped-by: James Ramsay <james@jramsay.com.au>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A mechanism to affect the default setting for a (related) group of
configuration variables is introduced.
* ds/feature-macros:
repo-settings: create feature.experimental setting
repo-settings: create feature.manyFiles setting
repo-settings: parse core.untrackedCache
commit-graph: turn on commit-graph by default
t6501: use 'git gc' in quiet mode
repo-settings: consolidate some config settings
Once upon a time, the code to add an object to our packing list in
pack-objects all lived in a single function. It computed the position
within the hash table once, then used it to check if the object was
already present, and if not, to add it.
Later, in 2834bc27c1 (pack-objects: refactor the packing list,
2013-10-24), this was split into two functions: packlist_find() and
packlist_alloc(). We ended up with an "index_pos" variable that gets
passed through several functions to make it from one to the other.
The resulting code is rather confusing to follow. The "index_pos"
variable is sometimes undefined, if we don't yet have a hash table. This
works out in practice because in that case packlist_alloc() won't use it
at all, since it will have to create/grow the hash table. But it's hard
to verify that, and it does cause gcc 9.2.1's -Wmaybe-uninitialized to
complain when compiled with "-flto -O3" (rightfully, since we do pass
the uninitialized value as a function parameter, even if nobody ends up
using it).
All of this is to save computing the hash index again when we're
inserting into the hash table, which I found doesn't make a measurable
difference in the program runtime (which is not surprising, since we're
doing all kinds of other heavyweight things for each object).
Let's just drop this index_pos variable entirely, simplifying the code
(and pleasing the compiler).
We might be better still refactoring this custom hash table to use one
of our existing implementations (an oidmap, or a kh_oid_map). I stopped
short of that here, but this would be the likely first step towards that
anyway.
Reported-by: Stephan Beyer <s-beyer@gmx.net>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The only caller of packlist_alloc() already has a "struct object_id",
and we immediately copy the hash they pass us into our own object_id.
Let's avoid the unnecessary round-trip to a raw sha1 pointer.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There are a few important config settings that are not loaded
during git_default_config. These are instead loaded on-demand.
Centralize these config options to a single scan, and store
all of the values in a repo_settings struct. The values for
each setting are initialized as negative to indicate "unset".
This centralization will be particularly important in a later
change to introduce "meta" config settings that change the
defaults for these config settings.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Long ago, in 97bfeb34df (Release pack windows before reporting out of
memory., 2006-12-24), we taught xmalloc() and friends to try unmapping
pack windows when malloc() failed. It's unlikely that his helps a lot in
practice, and it has some downsides. First, the downsides:
1. It makes xmalloc() not thread-safe. We've worked around this in
pack-objects.c, which installs its own locking version of the
try_to_free_routine(). But other threaded code doesn't.
2. It makes the system as a whole harder to reason about. Functions
which allocate heap memory under the hood may have farther-reaching
effects than expected.
That might be worth the tradeoff if there's a benefit. But in practice,
it seems unlikely. We're generally dealing with mmap'd files, so the OS
is going to do a much better job at responding to memory pressure by
dropping individual pages (the exception is systems with NO_MMAP, but
even there the OS can probably respond just as well with swapping).
So the only thing we're really freeing is address space. On 64-bit
systems, we have plenty of that to go around. On 32-bit systems, it
could possibly help. But around the same time we made two other changes:
77ccc5bbd1 (Introduce new config option for mmap limit., 2006-12-23) and
60bb8b1453 (Fully activate the sliding window pack access., 2006-12-23).
Together that means that a 32-bit system should have no more than 256MB
total of packed-git mmaps at one time, split between a few 32MB windows.
It's unlikely we have any address space problems since then, but we
don't have any data since the features were all added at the same time.
Likewise, xmmap() will try to free memory. At first glance, it seems
like we'd need this (when we try to mmap a new window, we might need to
close an old one to save address space on a 32-bit system). But we're
saved again by core.packedGitLimit: if we're going to exceed our 256MB
limit, we'll close an existing window before we even call mmap().
So it seems unlikely that this feature is actually doing anything
useful. And while we don't have reports of it harming anything (probably
because it rarely if ever kicks in), it would be nice to simplify the
system overall. This patch drops the whole try_to_free system from
xmalloc(), as well as the manual pack memory release in xmmap().
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Depending on various config options, a full repack may not be able to
build a reachability bitmap index (e.g., if pack.packSizeLimit forces us
to write multiple packs). In these cases pack-objects may write a
warning to stderr.
Since 36eba0323d (repack: enable bitmaps by default on bare repos,
2019-03-14), we may generate these warnings even when the user did not
explicitly ask for bitmaps. This has two downsides:
- it can be confusing, if they don't know what bitmaps are
- a daemonized auto-gc will write this to its log file, and the
presence of the warning may suppress further auto-gc (until
gc.logExpiry has elapsed)
Let's have repack communicate to pack-objects that the choice to turn on
bitmaps was not made explicitly by the user, which in turn allows
pack-objects to suppress these warnings.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The tree-walk API learned to pass an in-core repository
instance throughout more codepaths.
* nd/tree-walk-with-repo:
t7814: do not generate same commits in different repos
Use the right 'struct repository' instead of the_repository
match-trees.c: remove the_repo from shift_tree*()
tree-walk.c: remove the_repo from get_tree_entry_follow_symlinks()
tree-walk.c: remove the_repo from get_tree_entry()
tree-walk.c: remove the_repo from fill_tree_descriptor()
sha1-file.c: remove the_repo from read_object_with_reference()
Code clean-up to remove hardcoded SHA-1 hash from many places.
* jk/oidhash:
hashmap: convert sha1hash() to oidhash()
hash.h: move object_id definition from cache.h
khash: rename oid helper functions
khash: drop sha1-specific map types
pack-bitmap: convert khash_sha1 maps into kh_oid_map
delta-islands: convert island_marks khash to use oids
khash: rename kh_oid_t to kh_oid_set
khash: drop broken oid_map typedef
object: convert create_object() to use object_id
object: convert internal hash_obj() to object_id
object: convert lookup_object() to use object_id
object: convert lookup_unknown_object() to use object_id
pack-objects: convert locate_object_entry_hash() to object_id
pack-objects: convert packlist_find() to use object_id
pack-bitmap-write: convert some helpers to use object_id
upload-pack: rename a "sha1" variable to "oid"
describe: fix accidental oid/hash type-punning
The codepath to compute delta islands used to spew progress output
without giving the callers any way to squelch it, which has been
fixed.
* jk/delta-islands-progress-fix:
delta-islands: respect progress flag
The delta island code always prints "Marked %d islands", even if
progress has been suppressed with --no-progress or by sending stderr to
a non-tty.
Let's pass a progress boolean to load_delta_islands(). We already do
the same thing for the progress meter in resolve_tree_islands().
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There are no callers left of lookup_unknown_object() that aren't just
passing us the "hash" member of a "struct object_id". Let's take the
whole struct, which gets us closer to removing all raw sha1 variables.
It also matches the existing conversions of lookup_blob(), etc.
The conversions of callers were done by hand, but they're all mechanical
one-liners.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We take a raw hash pointer, but most of our callers have a "struct
object_id" already. Let's switch to taking the full struct, which will
let us continue removing uses of raw sha1 buffers.
There are two callers that do need special attention:
- in rebuild_existing_bitmaps(), we need to switch to
nth_packed_object_oid(). This incurs an extra hash copy over
pointing straight to the mmap'd sha1, but it shouldn't be measurable
compared to the rest of the operation.
- in can_reuse_delta() we already spent the effort to copy the sha1
into a "struct object_id", but now we just have to do so a little
earlier in the function (we can't easily convert that function's
callers because they may be pointing at mmap'd REF_DELTA blocks).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jk/unused-params-final-batch:
verify-commit: simplify parameters to run_gpg_verify()
show-branch: drop unused parameter from show_independent()
rev-list: drop unused void pointer from finish_commit()
remove_all_fetch_refspecs(): drop unused "remote" parameter
receive-pack: drop unused "commands" from prepare_shallow_update()
pack-objects: drop unused rev_info parameters
name-rev: drop unused parameters from is_better_name()
mktree: drop unused length parameter
wt-status: drop unused status parameter
read-cache: drop unused parameter from threaded load
clone: drop dest parameter from copy_alternates()
submodule: drop unused prefix parameter from some functions
builtin: consistently pass cmd_* prefix to parse_options
cmd_{read,write}_tree: rename "unused" variable that is used
The code to generate the multi-pack idx file was not prepared to
see too many packfiles and ran out of open file descriptor, which
has been corrected.
* ds/midx-too-many-packs:
midx: add packs to packed_git linked list
midx: pass a repository pointer
The connectivity bitmaps are created by default in bare
repositories now; also the pathname hash-cache is created by
default to avoid making crappy deltas when repacking.
* ew/repack-with-bitmaps-by-default:
pack-objects: default to writing bitmap hash-cache
t5310: correctly remove bitmaps for jgit test
repack: enable bitmaps by default on bare repos
When collecting the list of objects to pack in get_object_list(), we
pass our rev_info struct around to some functions that don't need it.
This is due to 03a9683d22 (Simplify is_kept_pack(), 2009-02-28), where
the kept-pack handling was moved out of the revision machinery.
Let's drop these unused parameters.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "git pack-objects" command learned to report the number of
objects it packed via the trace2 mechanism.
* jk/pack-objects-reports-num-objects-to-trace2:
pack-objects: write objects packed to trace2
Much of the multi-pack-index code focuses on the multi_pack_index
struct, and so we only pass a pointer to the current one. However,
we will insert a dependency on the packed_git linked list in a
future change, so we will need a repository reference. Inserting
these parameters is a significant enough change to split out.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is useful when investigating performance of pushes, and other times
when no progress information is written (because the pack is written to
stdout).
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Instead of storing unsigned char pointers in the hash tables, switch to
storing instances of struct object_id. Update several internal functions
and one external function to take pointers to struct object_id.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Enabling pack.writebitmaphashcache should always be a performance win.
It costs only 4 bytes per object on disk, and the timings in ae4f07fbcc
(pack-bitmap: implement optional name_hash cache, 2013-12-21) show it
improving fetch and partial-bitmap clone times by 40-50%.
The only reason we didn't enable it by default at the time is that early
versions of JGit's bitmap reader complained about the presence of
optional header bits it didn't understand. But that was changed in
JGit's d2fa3987a (Use bitcheck to check for presence of OPT_FULL option,
2013-10-30), which made it into JGit v3.5.0 in late 2014.
So let's turn this option on by default. It's backwards-compatible with
all versions of Git, and if you are also using JGit on the same
repository, you'd only run into problems using a version that's almost 5
years old.
We'll drop the manual setting from all of our test scripts, including
perf tests. This isn't strictly necessary, but it has two advantages:
1. If the hash-cache ever stops being enabled by default, our perf
regression tests will notice.
2. We can use the modified perf tests to show off the behavior of an
otherwise unconfigured repo, as shown below.
These are the results of a few of a perf tests against linux.git that
showed interesting results. You can see the expected speedup in 5310.4,
which was noted in ae4f07fbcc. Curiously, 5310.8 did not improve (and
actually got slower), despite seeing the opposite in ae4f07fbcc.
I don't have an explanation for that.
The tests from p5311 did not exist back then, but do show improvements
(a smaller pack due to better deltas, which we found in less time).
Test HEAD^ HEAD
-------------------------------------------------------------------------------------
5310.4: simulated fetch 7.39(22.70+0.25) 5.64(11.43+0.22) -23.7%
5310.8: clone (partial bitmap) 18.45(24.83+1.19) 19.94(28.40+1.36) +8.1%
5311.31: server (128 days) 0.41(1.13+0.05) 0.34(0.72+0.02) -17.1%
5311.32: size (128 days) 7.4M 7.0M -4.8%
5311.33: client (128 days) 1.33(1.49+0.06) 1.29(1.37+0.12) -3.0%
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When studying the performance of 'git push' we would like to know
how much time is spent at various parts of the command. One area
that could cause performance trouble is 'git pack-objects'.
Add trace2 regions around the three main actions taken in this
command:
1. Enumerate objects.
2. Prepare pack.
3. Write pack-file.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git pack-objects" learned another algorithm to compute the set of
objects to send, that trades the resulting packfile off to save
traversal cost to favor small pushes.
* ds/push-sparse-tree-walk:
pack-objects: create GIT_TEST_PACK_SPARSE
pack-objects: create pack.useSparse setting
revision: implement sparse algorithm
list-objects: consume sparse tree walk
revision: add mark_tree_uninteresting_sparse
The assumption to work on the single "in-core index" instance has
been reduced from the library-ish part of the codebase.
* nd/the-index-final:
cache.h: flip NO_THE_INDEX_COMPATIBILITY_MACROS switch
read-cache.c: remove the_* from index_has_changes()
merge-recursive.c: remove implicit dependency on the_repository
merge-recursive.c: remove implicit dependency on the_index
sha1-name.c: remove implicit dependency on the_index
read-cache.c: replace update_index_if_able with repo_&
read-cache.c: kill read_index()
checkout: avoid the_index when possible
repository.c: replace hold_locked_index() with repo_hold_locked_index()
notes-utils.c: remove the_repository references
grep: use grep_opt->repo instead of explict repo argument
"git pack-objects" incorrectly used uninitialized mutex, which has
been corrected.
* ph/pack-objects-mutex-fix:
pack-objects: merge read_lock and lock in packing_data struct
pack-objects: move read mutex to packing_data struct
The code to walk tree objects has been taught that we may be
working with object names that are not computed with SHA-1.
* bc/tree-walk-oid:
cache: make oidcpy always copy GIT_MAX_RAWSZ bytes
tree-walk: store object_id in a separate member
match-trees: use hashcpy to splice trees
match-trees: compute buffer offset correctly when splicing
tree-walk: copy object ID before use
Rename the packing_data lock to obd_lock and upgrade it to a recursive
mutex to make it suitable for current read_lock usages. Additionally
remove the superfluous #ifndef NO_PTHREADS guard around mutex
initialization in prepare_packing_data as the mutex functions
themselves are already protected.
Signed-off-by: Patrick Hogg <phogg@novamoon.net>
Helped-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
ac77d0c37 ("pack-objects: shrink size field in struct object_entry",
2018-04-14) added an extra usage of read_lock/read_unlock in the newly
introduced oe_get_size_slow for thread safety in parallel calls to
try_delta(). Unfortunately oe_get_size_slow is also used in serial
code, some of which is called before the first invocation of
ll_find_deltas. As such the read mutex is not guaranteed to be
initialized.
Resolve this by moving the read mutex to packing_data and initializing
it in prepare_packing_data which is initialized in cmd_pack_objects.
Signed-off-by: Patrick Hogg <phogg@novamoon.net>
Reviewed-by: Duy Nguyen <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
By default, index compat macros are off from now on, because they
could hide the_index dependency.
Only those in builtin can use it.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
