This commit introduces the infrastructure for the core.fsync
configuration knob. The repository components we want to sync
are identified by flags so that we can turn on or off syncing
for specific components.
If core.fsyncObjectFiles is set and the core.fsync configuration
also includes FSYNC_COMPONENT_LOOSE_OBJECT, we will fsync any
loose objects. This picks the strictest data integrity behavior
if core.fsync and core.fsyncObjectFiles are set to conflicting values.
This change introduces the currently unused fsync_component
helper, which will be used by a later patch that adds fsyncing to
the refs backend.
Actual configuration and documentation of the fsync components
list are in other patches in the series to separate review of
the underlying mechanism from the policy of how it's configured.
Helped-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Neeraj Singh <neerajsi@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As a small courtesy to users, report what limit was breached. This
is especially useful when a push exceeds a server-defined limit, since
the user is unlikely to have configured the limit (their host did).
Also demonstrate the human-readable message in a test.
Helped-by: Taylor Blau <me@ttaylorr.com>
Helped-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Matt Cooper <vtbassmatt@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
They are all replaced by "the option '%s' requires '%s'", which is a
new string but replaces 17 previous unique strings.
Signed-off-by: Jean-Noël Avila <jn.avila@free.fr>
Reviewed-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Emir and Jean-Noël reported typos in some i18n messages when preparing
l10n for git 2.34.0.
* Fix unstable spelling of config variable "gpg.ssh.defaultKeyCommand"
which was introduced in commit fd9e226776 (ssh signing: retrieve a
default key from ssh-agent, 2021-09-10).
* Add missing space between "with" and "--python" which was introduced
in commit bd0708c7eb (ref-filter: add %(raw) atom, 2021-07-26).
* Fix unmatched single quote in 'builtin/index-pack.c' which was
introduced in commit 8737dab346 (index-pack: refactor renaming in
final(), 2021-09-09)
[1] https://github.com/git-l10n/git-po/pull/567
Reported-by: Emir Sarı <bitigchi@me.com>
Reported-by: Jean-Noël Avila <jn.avila@free.fr>
Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git fsck" has been taught to report mismatch between expected and
actual types of an object better.
* ab/fsck-unexpected-type:
fsck: report invalid object type-path combinations
fsck: don't hard die on invalid object types
object-file.c: stop dying in parse_loose_header()
object-file.c: return ULHR_TOO_LONG on "header too long"
object-file.c: use "enum" return type for unpack_loose_header()
object-file.c: simplify unpack_loose_short_header()
object-file.c: make parse_loose_header_extended() public
object-file.c: return -1, not "status" from unpack_loose_header()
object-file.c: don't set "typep" when returning non-zero
cat-file tests: test for current --allow-unknown-type behavior
cat-file tests: add corrupt loose object test
cat-file tests: test for missing/bogus object with -t, -s and -p
cat-file tests: move bogus_* variable declarations earlier
fsck tests: test for garbage appended to a loose object
fsck tests: test current hash/type mismatch behavior
fsck tests: refactor one test to use a sub-repo
fsck tests: add test for fsck-ing an unknown type
Improve the error that's emitted in cases where we find a loose object
we parse, but which isn't at the location we expect it to be.
Before this change we'd prefix the error with a not-a-OID derived from
the path at which the object was found, due to an emergent behavior in
how we'd end up with an "OID" in these codepaths.
Now we'll instead say what object we hashed, and what path it was
found at. Before this patch series e.g.:
$ git hash-object --stdin -w -t blob </dev/null
e69de29bb2
$ mv objects/e6/ objects/e7
Would emit ("[...]" used to abbreviate the OIDs):
git fsck
error: hash mismatch for ./objects/e7/9d[...] (expected e79d[...])
error: e79d[...]: object corrupt or missing: ./objects/e7/9d[...]
Now we'll instead emit:
error: e69d[...]: hash-path mismatch, found at: ./objects/e7/9d[...]
Furthermore, we'll do the right thing when the object type and its
location are bad. I.e. this case:
$ git hash-object --stdin -w -t garbage --literally </dev/null
8315a83d2acc4c174aed59430f9a9c4ed926440f
$ mv objects/83 objects/84
As noted in an earlier commits we'd simply die early in those cases,
until preceding commits fixed the hard die on invalid object type:
$ git fsck
fatal: invalid object type
Now we'll instead emit sensible error messages:
$ git fsck
error: 8315[...]: hash-path mismatch, found at: ./objects/84/15[...]
error: 8315[...]: object is of unknown type 'garbage': ./objects/84/15[...]
In both fsck.c and object-file.c we're using null_oid as a sentinel
value for checking whether we got far enough to be certain that the
issue was indeed this OID mismatch.
We need to add the "object corrupt or missing" special-case to deal
with cases where read_loose_object() will return an error before
completing check_object_signature(), e.g. if we have an error in
unpack_loose_rest() because we find garbage after the valid gzip
content:
$ git hash-object --stdin -w -t blob </dev/null
e69de29bb2
$ chmod 755 objects/e6/9de29bb2d1d6434b8b29ae775ad8c2e48c5391
$ echo garbage >>objects/e6/9de29bb2d1d6434b8b29ae775ad8c2e48c5391
$ git fsck
error: garbage at end of loose object 'e69d[...]'
error: unable to unpack contents of ./objects/e6/9d[...]
error: e69d[...]: object corrupt or missing: ./objects/e6/9d[...]
There is currently some weird messaging in the edge case when the two
are combined, i.e. because we're not explicitly passing along an error
state about this specific scenario from check_stream_oid() via
read_loose_object() we'll end up printing the null OID if an object is
of an unknown type *and* it can't be unpacked by zlib, e.g.:
$ git hash-object --stdin -w -t garbage --literally </dev/null
8315a83d2acc4c174aed59430f9a9c4ed926440f
$ chmod 755 objects/83/15a83d2acc4c174aed59430f9a9c4ed926440f
$ echo garbage >>objects/83/15a83d2acc4c174aed59430f9a9c4ed926440f
$ /usr/bin/git fsck
fatal: invalid object type
$ ~/g/git/git fsck
error: garbage at end of loose object '8315a83d2acc4c174aed59430f9a9c4ed926440f'
error: unable to unpack contents of ./objects/83/15a83d2acc4c174aed59430f9a9c4ed926440f
error: 8315a83d2acc4c174aed59430f9a9c4ed926440f: object corrupt or missing: ./objects/83/15a83d2acc4c174aed59430f9a9c4ed926440f
error: 0000000000000000000000000000000000000000: object is of unknown type 'garbage': ./objects/83/15a83d2acc4c174aed59430f9a9c4ed926440f
[...]
I think it's OK to leave that for future improvements, which would
involve enum-ifying more error state as we've done with "enum
unpack_loose_header_result" in preceding commits. In these
increasingly more obscure cases the worst that can happen is that
we'll get slightly nonsensical or inapplicable error messages.
There's other such potential edge cases, all of which might produce
some confusing messaging, but still be handled correctly as far as
passing along errors goes. E.g. if check_object_signature() returns
and oideq(real_oid, null_oid()) is true, which could happen if it
returns -1 due to the read_istream() call having failed.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The order in which various files that make up a single (conceptual)
packfile has been reevaluated and straightened up. This matters in
correctness, as an incomplete set of files must not be shown to a
running Git.
* tb/pack-finalize-ordering:
pack-objects: rename .idx files into place after .bitmap files
pack-write: split up finish_tmp_packfile() function
builtin/index-pack.c: move `.idx` files into place last
index-pack: refactor renaming in final()
builtin/repack.c: move `.idx` files into place last
pack-write.c: rename `.idx` files after `*.rev`
pack-write: refactor renaming in finish_tmp_packfile()
bulk-checkin.c: store checksum directly
pack.h: line-wrap the definition of finish_tmp_packfile()
Support an arbitrary file descriptor expression in the semantic patch
for replacing open+die_errno with xopen, not just an identifier, and
apply it. This makes the error message at the single affected place
more consistent and reduces code duplication.
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In a similar spirit as preceding patches to `git repack` and `git
pack-objects`, fix the identical problem in `git index-pack`.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Refactor the renaming in final() into a helper function, this is
similar in spirit to a preceding refactoring of finish_tmp_packfile()
in pack-write.c.
Before e37d0b8730 (builtin/index-pack.c: write reverse indexes,
2021-01-25) it probably wasn't worth it to have this sort of helper,
due to the differing "else if" case for "pack" files v.s. "idx" files.
But since we've got "rev" as well now, let's do the renaming via a
helper, this is both a net decrease in lines, and improves the
readability, since we can easily see at a glance that the logic for
writing these three types of files is exactly the same, aside from the
obviously differing cases of "*final_name" being NULL, and
"make_read_only_if_same" being different.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a --progress-title option to index-pack, when data is piped into
index-pack its progress is a proxy for whatever's feeding it data.
This option will allow us to set a more relevant progress bar title in
"git bundle unbundle", and is also used in my "bundle-uri" RFC
patches[1] by a new caller in fetch-pack.c.
The code change in cmd_index_pack() won't handle
"--progress-title=xyz", only "--progress-title xyz", and the "(i+1)"
style (as opposed to "i + 1") is a bit odd.
Not using the "--long-option=value" style is inconsistent with
existing long options handled by cmd_index_pack(), but makes the code
that needs to call it better (two strvec_push(), instead of needing a
strvec_pushf()). Since the option is internal-only the inconsistency
shouldn't matter.
I'm copying the pattern to handle it as-is from the handling of the
existing "-o" option in the same function, see 9cf6d3357a (Add
git-index-pack utility, 2005-10-12) for its addition. That's a short
option, but the code to implement the two is the same in functionality
and style. Eventually we'd like to migrate all of this this to
parse_options(), which would make these differences in behavior go
away.
1. https://lore.kernel.org/git/RFC-cover-00.13-0000000000-20210805T150534Z-avarab@gmail.com/
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add and apply a semantic patch for using xopen() instead of calling
open(2) and die() or die_errno() explicitly. This makes the error
messages more consistent and shortens the code.
Signed-off-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
9cf6d3357a (Add git-index-pack utility, 2005-10-12) and
466dbc42f5 (receive-pack: Send internal errors over side-band #2,
2010-02-10) we added these static functions and forward-declared their
__attribute__((printf)).
I think this may have been to work around some compiler limitation at
the time, but in any case we have a lot of code that uses the briefer
way of declaring these that I'm using here, so if we had any such
issues with compilers we'd have seen them already.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we're hashing a value which is going to be an object ID, we want to
zero-pad that value if necessary. To do so, use the final_oid_fn
instead of the final_fn anytime we're going to create an object ID to
ensure we perform this operation.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the future, we'll want oidread to automatically set the hash
algorithm member for an object ID we read into it, so ensure we use
oidread instead of hashcpy everywhere we're copying a hash value into a
struct object_id.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Fsck API clean-up.
* ab/fsck-api-cleanup:
fetch-pack: use new fsck API to printing dangling submodules
fetch-pack: use file-scope static struct for fsck_options
fetch-pack: don't needlessly copy fsck_options
fsck.c: move gitmodules_{found,done} into fsck_options
fsck.c: add an fsck_set_msg_type() API that takes enums
fsck.c: pass along the fsck_msg_id in the fsck_error callback
fsck.[ch]: move FOREACH_FSCK_MSG_ID & fsck_msg_id from *.c to *.h
fsck.c: give "FOREACH_MSG_ID" a more specific name
fsck.c: undefine temporary STR macro after use
fsck.c: call parse_msg_type() early in fsck_set_msg_type()
fsck.h: re-order and re-assign "enum fsck_msg_type"
fsck.h: move FSCK_{FATAL,INFO,ERROR,WARN,IGNORE} into an enum
fsck.c: refactor fsck_msg_type() to limit scope of "int msg_type"
fsck.c: rename remaining fsck_msg_id "id" to "msg_id"
fsck.c: remove (mostly) redundant append_msg_id() function
fsck.c: rename variables in fsck_set_msg_type() for less confusion
fsck.h: use "enum object_type" instead of "int"
fsck.h: use designed initializers for FSCK_OPTIONS_{DEFAULT,STRICT}
fsck.c: refactor and rename common config callback
Refactor the check added in 5476e1efde (fetch-pack: print and use
dangling .gitmodules, 2021-02-22) to make use of us now passing the
"msg_id" to the user defined "error_func". We can now compare against
the FSCK_MSG_GITMODULES_MISSING instead of parsing the generated
message.
Let's also replace register_found_gitmodules() with directly
manipulating the "gitmodules_found" member. A recent commit moved it
into "fsck_options" so we could do this here.
I'm sticking this callback in fsck.c. Perhaps in the future we'd like
to accumulate such callbacks into another file (maybe fsck-cb.c,
similar to parse-options-cb.c?), but while we've got just the one
let's just put it into fsck.c.
A better alternative in this case would be some library some more
obvious library shared by fetch-pack.c ad builtin/index-pack.c, but
there isn't such a thing.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change the behavior of the .gitmodules validation added in
5476e1efde (fetch-pack: print and use dangling .gitmodules,
2021-02-22) so we're using one "fsck_options".
I found that code confusing to read. One might think that not setting
up the error_func earlier means that we're relying on the "error_func"
not being set in some code in between the two hunks being modified
here.
But we're not, all we're doing in the rest of "cmd_index_pack()" is
further setup by calling fsck_set_msg_types(), and assigning to
do_fsck_object.
So there was no reason in 5476e1efde to make a shallow copy of the
fsck_options struct before setting error_func. Let's just do this
setup at the top of the function, along with the "walk" assignment.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change the fsck_error callback to also pass along the
fsck_msg_id. Before this change the only way to get the message id was
to parse it back out of the "message".
Let's pass it down explicitly for the benefit of callers that might
want to use it, as discussed in [1].
Passing the msg_type is now redundant, as you can always get it back
from the msg_id, but I'm not changing that convention. It's really
common to need the msg_type, and the report() function itself (which
calls "fsck_error") needs to call fsck_msg_type() to discover
it. Let's not needlessly re-do that work in the user callback.
1. https://lore.kernel.org/git/87blcja2ha.fsf@evledraar.gmail.com/
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the FSCK_{FATAL,INFO,ERROR,WARN,IGNORE} defines into a new
fsck_msg_type enum.
These defines were originally introduced in:
- ba002f3b28 (builtin-fsck: move common object checking code to
fsck.c, 2008-02-25)
- f50c440730 (fsck: disallow demoting grave fsck errors to warnings,
2015-06-22)
- efaba7cc77 (fsck: optionally ignore specific fsck issues
completely, 2015-06-22)
- f27d05b170 (fsck: allow upgrading fsck warnings to errors,
2015-06-22)
The reason these were defined in two different places is because we
use FSCK_{IGNORE,INFO,FATAL} only in fsck.c, but FSCK_{ERROR,WARN} are
used by external callbacks.
Untangling that would take some more work, since we expose the new
"enum fsck_msg_type" to both. Similar to "enum object_type" it's not
worth structuring the API in such a way that only those who need
FSCK_{ERROR,WARN} pass around a different type.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change the fsck_walk_func to use an "enum object_type" instead of an
"int" type. The types are compatible, and ever since this was added in
355885d531 (add generic, type aware object chain walker, 2008-02-25)
we've used entries from object_type (OBJ_BLOB etc.).
So this doesn't really change anything as far as the generated code is
concerned, it just gives the compiler more information and makes this
easier to read.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.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>
The approach to "fsck" the incoming objects in "index-pack" is
attractive for performance reasons (we have them already in core,
inflated and ready to be inspected), but fundamentally cannot be
applied fully when we receive more than one pack stream, as a tree
object in one pack may refer to a blob object in another pack as
".gitmodules", when we want to inspect blobs that are used as
".gitmodules" file, for example. Teach "index-pack" to emit
objects that must be inspected later and check them in the calling
"fetch-pack" process.
* jt/transfer-fsck-across-packs:
fetch-pack: print and use dangling .gitmodules
fetch-pack: with packfile URIs, use index-pack arg
http-fetch: allow custom index-pack args
http: allow custom index-pack args
Teach index-pack to print dangling .gitmodules links after its "keep" or
"pack" line instead of declaring an error, and teach fetch-pack to check
such lines printed.
This allows the tree side of the .gitmodules link to be in one packfile
and the blob side to be in another without failing the fsck check,
because it is now fetch-pack which checks such objects after all
packfiles have been downloaded and indexed (and not index-pack on an
individual packfile, as it is before this commit).
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
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>
Teach 'git index-pack' to optionally write and verify reverse index with
'--[no-]rev-index', as well as respecting the 'pack.writeReverseIndex'
configuration option.
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
To derive the filename for a .idx file, 'git index-pack' uses
derive_filename() to strip the '.pack' suffix and add the new suffix.
Prepare for stripping off suffixes other than '.pack' by making the
suffix to strip a parameter of derive_filename(). In order to make this
consistent with the "suffix" parameter which does not begin with a ".",
an additional check in derive_filename.
Suggested-by: Jeff King <peff@peff.net>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Drop the last remnant of "sha1" in this file and rename it to reflect
that we're not just able to handle SHA-1 these days.
Signed-off-by: Martin Ågren <martin.agren@gmail.com>
Reviewed-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A pack and its matching .idx file are limited to 2^32 objects, because
the pack format contains a 32-bit field to store the number of objects.
Hence we use uint32_t in the code.
But the byte count of even a .idx file can be much larger than that,
because it stores at least a hash and an offset for each object. So
using SHA-1, a v2 .idx file will cross the 4GB boundary at 153,391,650
objects. This confuses load_idx(), which computes the minimum size like
this:
unsigned long min_size = 8 + 4*256 + nr*(hashsz + 4 + 4) + hashsz + hashsz;
Even though min_size will be big enough on most 64-bit platforms, the
actual arithmetic is done as a uint32_t, resulting in a truncation. We
actually exceed that min_size, but then we do:
unsigned long max_size = min_size;
if (nr)
max_size += (nr - 1)*8;
to account for the variable-sized table. That computation doesn't
overflow quite so low, but with the truncation for min_size, we end up
with a max_size that is much smaller than our actual size. So we
complain that the idx is invalid, and can't find any of its objects.
We can fix this case by casting "nr" to a size_t, which will do the
multiplication in 64-bits (assuming you're on a 64-bit platform; this
will never work on a 32-bit system since we couldn't map the whole .idx
anyway). Likewise, we don't have to worry about further additions,
because adding a smaller number to a size_t will convert the other side
to a size_t.
A few notes:
- obviously we could just declare "nr" as a size_t in the first place
(and likewise, packed_git.num_objects). But it's conceptually a
uint32_t because of the on-disk format, and we correctly treat it
that way in other contexts that don't need to compute byte offsets
(e.g., iterating over the set of objects should and generally does
use a uint32_t). Switching to size_t would make all of those other
cases look wrong.
- it could be argued that the proper type is off_t to represent the
file offset. But in practice the .idx file must fit within memory,
because we mmap the whole thing. And the rest of the code (including
the idx_size variable we're comparing against) uses size_t.
- we'll add the same cast to the max_size arithmetic line. Even though
we're adding to a larger type, which will convert our result, the
multiplication is still done as a 32-bit value and can itself
overflow. I didn't check this with my test case, since it would need
an even larger pack (~530M objects), but looking at compiler output
shows that it works this way. The standard should agree, but I
couldn't find anything explicit in 6.3.1.8 ("usual arithmetic
conversions").
The case in load_idx() was the most immediate one that I was able to
trigger. After fixing it, looking up actual objects (including the very
last one in sha1 order) works in a test repo with 153,725,110 objects.
That's because bsearch_hash() works with uint32_t entry indices, and the
actual byte access:
int cmp = hashcmp(table + mi * stride, sha1);
is done with "stride" as a size_t, causing the uint32_t "mi" to be
promoted to a size_t. This is the way most code will access the index
data.
However, I audited all of the other byte-wise accesses of
packed_git.index_data, and many of the others are suspect (they are
similar to the max_size one, where we are adding to a properly sized
offset or directly to a pointer, but the multiplication in the
sub-expression can overflow). I didn't trigger any of these in practice,
but I believe they're potential problems, and certainly adding in the
cast is not going to hurt anything here.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Hotfix and clean-up for the jt/threaded-index-pack topic that has
graduated to v2.29-rc0.
* jk/index-pack-hotfixes:
index-pack: make get_base_data() comment clearer
index-pack: drop type_cas mutex
index-pack: restore "resolving deltas" progress meter
A comment mentions that we may free cached delta bases via
find_unresolved_deltas(), but that function went away in f08cbf60fe
(index-pack: make quantum of work smaller, 2020-09-08). Since we need to
rewrite that comment anyway, make the entire comment clearer.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The type_cas lock lost all of its callers in f08cbf60fe (index-pack:
make quantum of work smaller, 2020-09-08), so we can safely delete it.
The compiler didn't alert us that the variable became unused, because we
still call pthread_mutex_init() and pthread_mutex_destroy() on it.
It's worth considering also whether that commit was in error to remove
the use of the lock. Why don't we need it now, if we did before, as
described in ab791dd138 (index-pack: fix race condition with duplicate
bases, 2014-08-29)? I think the answer is that we now look at and assign
the child_obj->real_type field in the main thread while holding the
work_lock(). So we don't have to worry about racing with the worker
threads.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit f08cbf60fe (index-pack: make quantum of work smaller, 2020-09-08)
refactored the main loop in threaded_second_pass(), but also deleted the
call to display_progress() at the top of the loop. This means that users
typically see no progress at all during the delta resolution phase (and
for large repositories, Git appears to hang).
This looks like an accident that was unrelated to the intended change of
that commit, since we continue to update nr_resolved_deltas in
resolve_delta(). Let's restore the call to get that progress back.
We'll also add a test that confirms we generate the expected progress.
This isn't perfect, as it wouldn't catch a bug where progress was
delayed to the end. That was probably possible to trigger when receiving
a thin pack, because we'd eventually call display_progress() from
fix_unresolved_deltas(), but only once after doing all the work.
However, since our test case generates a complete pack, it reliably
demonstrates this particular bug and its fix. And we can't do better
without making the test racy.
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git index-pack" learned to resolve deltified objects with greater
parallelism.
* jt/threaded-index-pack:
index-pack: make quantum of work smaller
index-pack: make resolve_delta() assume base data
index-pack: calculate {ref,ofs}_{first,last} early
index-pack: remove redundant child field
index-pack: unify threaded and unthreaded code
index-pack: remove redundant parameter
Documentation: deltaBaseCacheLimit is per-thread
Currently, when index-pack resolves deltas, it does not split up delta
trees into threads: each delta base root (an object that is not a
REF_DELTA or OFS_DELTA) can go into its own thread, but all deltas on
that root (direct or indirect) are processed in the same thread.
This is a problem when a repository contains a large text file (thus,
delta-able) that is modified many times - delta resolution time during
fetching is dominated by processing the deltas corresponding to that
text file.
This patch contains a solution to that. When cloning using
git -c core.deltabasecachelimit=1g clone \
https://fuchsia.googlesource.com/third_party/vulkan-cts
on my laptop, clone time improved from 3m2s to 2m5s (using 3 threads,
which is the default).
The solution is to have a global work stack. This stack contains delta
bases (objects, whether appearing directly in the packfile or generated
by delta resolution, that themselves have delta children) that need to
be processed; whenever a thread needs work, it peeks at the top of the
stack and processes its next unprocessed child. If a thread finds the
stack empty, it will look for more delta base roots to push on the stack
instead.
The main weakness of having a global work stack is that more time is
spent in the mutex, but profiling has shown that most time is spent in
the resolution of the deltas themselves, so this shouldn't be an issue
in practice. In any case, experimentation (as described in the clone
command above) shows that this patch is a net improvement.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A subsequent commit will make the quantum of work smaller, necessitating
more locking. This commit allows resolve_delta() to be called outside
the lock.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is refactoring 2 of 2 to simplify struct base_data.
Whenever we make a struct base_data, immediately calculate its delta
children. This eliminates confusion as to when the
{ref,ofs}_{first,last} fields are initialized.
Before this patch, the delta children were calculated at the last
possible moment. This allowed the members of struct base_data to be
populated in any order, superficially useful when we have the object
contents before the struct object_entry. But this makes reasoning about
the state of struct base_data more complicated, hence this patch.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is refactoring 1 of 2 to simplify struct base_data.
In index-pack, each thread maintains a doubly-linked list of the delta
chain that it is currently processing (the "base" and "child" pointers
in struct base_data). When a thread exceeds the delta base cache limit
and needs to reclaim memory, it uses the "child" pointers to traverse
the lineage, reclaiming the memory of the eldest delta bases first.
A subsequent patch will perform memory reclaiming in a different way and
will thus no longer need the "child" pointer. Because the "child"
pointer is redundant even now, remove it so that the aforementioned
subsequent patch will be clearer. In the meantime, reclaim memory in the
reverse order of the "base" pointers.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
find_{ref,ofs}_delta_{,children} take an enum object_type parameter, but
the object type is already present in the name of the function. Remove
that parameter from these functions.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit b8a2486f15 (index-pack: support multithreaded delta resolving,
2012-05-06) describes an experiment that shows that setting the number
of threads for index-pack higher than 3 does not help.
I repeated that experiment using a more modern version of Git and a more
modern CPU and got different results.
Here are timings for p5302 against linux.git run on my laptop, a Core
i9-9880H with 8 cores plus hyperthreading (so online-cpus returns 16):
5302.3: index-pack 0 threads 256.28(253.41+2.79)
5302.4: index-pack 1 threads 257.03(254.03+2.91)
5302.5: index-pack 2 threads 149.39(268.34+3.06)
5302.6: index-pack 4 threads 94.96(294.10+3.23)
5302.7: index-pack 8 threads 68.12(339.26+3.89)
5302.8: index-pack 16 threads 70.90(655.03+7.21)
5302.9: index-pack default number of threads 116.91(290.05+3.21)
You can see that wall-clock times continue to improve dramatically up to
the number of cores, but bumping beyond that (into hyperthreading
territory) does not help (and in fact hurts a little).
Here's the same experiment on a machine with dual Xeon 6230's, totaling
40 cores (80 with hyperthreading):
5302.3: index-pack 0 threads 310.04(302.73+6.90)
5302.4: index-pack 1 threads 310.55(302.68+7.40)
5302.5: index-pack 2 threads 178.17(304.89+8.20)
5302.6: index-pack 5 threads 99.53(315.54+9.56)
5302.7: index-pack 10 threads 72.80(327.37+12.79)
5302.8: index-pack 20 threads 60.68(357.74+21.66)
5302.9: index-pack 40 threads 58.07(454.44+67.96)
5302.10: index-pack 80 threads 59.81(720.45+334.52)
5302.11: index-pack default number of threads 134.18(309.32+7.98)
The results are similar; things stop improving at 40 threads. Curiously,
going from 20 to 40 really doesn't help much, either (and increases CPU
time considerably). So that may represent an actual barrier to
parallelism, where we lose out due to context-switching and loss of
cache locality, but don't reap the wall-clock benefits due to contention
of our coarse-grained locks.
So what's a good default value? It's clear that the current cap of 3 is
too low; our default values are 42% and 57% slower than the best times
on each machine. The results on the 40-core machine imply that 20
threads is an actual barrier regardless of the number of cores, so we'll
take that as a maximum. We get the best results on these machines at
half of the online-cpus value. That's presumably a result of the
hyperthreading. That's common on multi-core Intel processors, but not
necessarily elsewhere. But if we take it as an assumption, we can
perform optimally on hyperthreaded machines and still do much better
than the status quo on other machines, as long as we never half below
the current value of 3.
So that's what this patch does.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
git index-pack is usually run in a repository, but need not be. Since
packs don't contains information on the algorithm in use, instead
relying on context, add an option to index-pack to tell it which one
we're using in case someone runs it outside of a repository. Since
using --stdin necessarily implies a repository, don't allow specifying
an object format if it's provided to prevent users from passing an
option that won't work. Add documentation for this option.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Both v2 pack index files and the v3 format specified as part of the
NewHash work have similar data starting at the CRC table. Much of the
existing code wants to read either this table or the offset entries
following it, and in doing so computes the offset each time.
In order to share as much code between v2 and v3, compute the offset of
the CRC table and store it when the pack is opened. Use this value to
compute offsets to not only the CRC table, but to the offset entries
beyond it.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There are 3 callers to promisor_remote_get_direct() that first check if
the number of objects to be fetched is equal to 0. Fold that check into
promisor_remote_get_direct(), and in doing so, be explicit as to what
promisor_remote_get_direct() does if oid_nr is 0 (it returns 0, success,
immediately).
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The index-pack code now diagnoses a bad input packstream that
records the same object twice when it is used as delta base; the
code used to declare a software bug when encountering such an
input, but it is an input error.
* jk/index-pack-dupfix:
index-pack: downgrade twice-resolved REF_DELTA to die()
When we're resolving a REF_DELTA, we compare-and-swap its type from
REF_DELTA to whatever real type the base object has, as discussed in
ab791dd138 (index-pack: fix race condition with duplicate bases,
2014-08-29). If the old type wasn't a REF_DELTA, we consider that a
BUG(). But as discussed in that commit, we might see this case whenever
we try to resolve an object twice, which may happen because we have
multiple copies of the base object.
So this isn't a bug at all, but rather a sign that the input pack is
broken. And indeed, this case is triggered already in t5309.5 and
t5309.6, which create packs with delta cycles and duplicate bases. But
we never noticed because those tests are marked expect_failure.
Those tests were added by b2ef3d9ebb (test index-pack on packs with
recoverable delta cycles, 2013-08-23), which was leaving the door open
for cases that we theoretically _could_ handle. And when we see an
already-resolved object like this, in theory we could keep going after
confirming that the previously resolved child->real_type matches
base->obj->real_type. But:
- enforcing the "only resolve once" rule here saves us from an
infinite loop in other parts of the code. If we keep going, then the
delta cycle in t5309.5 causes us to loop infinitely, as
find_ref_delta_children() doesn't realize which objects have already
been resolved. So there would be more changes needed to make this
case work, and in the meantime we'd be worse off.
- any pack that triggers this is broken anyway. It either has a
duplicate base object, or it has a cycle which causes us to bring in
a duplicate via --fix-thin. In either case, we'd end up rejecting
the pack in write_idx_file(), which also detects duplicates.
So the tests have little value in documenting what we _could_ be doing
(and have been neglected for 6+ years). Let's switch them to confirming
that we handle this case cleanly (and switch out the BUG() for a more
informative die() so that we do so).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some callers of check_object_signature() can work on arbitrary
repositories, but the repo does not get passed to this function.
Instead, the_repository is always used internally. To fix possible
inconsistencies, allow the function to receive a struct repository and
make those callers pass on the repo being handled.
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Neeraj Singh
Matt Cooper
Jean-Noël Avila
Jiang Xin
Junio C Hamano
Ævar Arnfjörð Bjarmason
René Scharfe
Taylor Blau
brian m. carlson
Jonathan Tan
Martin Ågren
Jeff King
Matheus Tavares