Many git commands that deal with working tree files try to remove a
directory that becomes empty (i.e. "git switch" from a branch that
has the directory to another branch that does not would attempt
remove all files in the directory and the directory itself). This
drops users into an unfamiliar situation if the command was run in
a subdirectory that becomes subject to removal due to the command.
The commands have been taught to keep an empty directory if it is
the directory they were started in to avoid surprising users.
* en/keep-cwd:
t2501: simplify the tests since we can now assume desired behavior
dir: new flag to remove_dir_recurse() to spare the original_cwd
dir: avoid incidentally removing the original_cwd in remove_path()
stash: do not attempt to remove startup_info->original_cwd
rebase: do not attempt to remove startup_info->original_cwd
clean: do not attempt to remove startup_info->original_cwd
symlinks: do not include startup_info->original_cwd in dir removal
unpack-trees: add special cwd handling
unpack-trees: refuse to remove startup_info->original_cwd
setup: introduce startup_info->original_cwd
t2501: add various tests for removing the current working directory
The sparse-index/sparse-checkout feature had a bug in its use of
the matching code to determine which path is in or outside the
sparse checkout patterns.
* ds/sparse-deep-pattern-checkout-fix:
unpack-trees: use traverse_path instead of name
t1092: add deeper changes during a checkout
Various operating modes of "git reset" have been made to work
better with the sparse index.
* vd/sparse-reset:
unpack-trees: improve performance of next_cache_entry
reset: make --mixed sparse-aware
reset: make sparse-aware (except --mixed)
reset: integrate with sparse index
reset: expand test coverage for sparse checkouts
sparse-index: update command for expand/collapse test
reset: preserve skip-worktree bit in mixed reset
reset: rename is_missing to !is_in_reset_tree
When running commands such as `git reset --hard` from a subdirectory, if
that subdirectory is in the way of adding needed files, bail with an
error message.
Note that this change looks kind of like it duplicates the new lines of
code from the previous commit in verify_clean_subdirectory(). However,
when we are preserving untracked files, we would rather any error
messages about untracked files being in the way take precedence over
error messages about a subdirectory that happens to be the_original_cwd
being in the way. But in the UNPACK_RESET_OVERWRITE_UNTRACKED case,
there is no untracked checking to be done, so we simply add a special
case near the top of verify_absent_1.
Acked-by: Derrick Stolee <stolee@gmail.com>
Acked-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the past, when a directory needs to be removed to make room for a
file, we have always errored out when that directory contains any
untracked (but not ignored) files. Add an extra condition on that: also
error out if the directory is the current working directory we inherited
from our parent process.
Acked-by: Derrick Stolee <stolee@gmail.com>
Acked-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The sparse_dir_matches_path() method compares a cache entry that is a
sparse directory entry against a 'struct traverse_info *info' and a
'struct name_entry *p' to see if the cache entry has exactly the right
name for those other inputs.
This method was introduced in 523506d (unpack-trees: unpack sparse
directory entries, 2021-07-14), but included a significant mistake. The
path comparisons used 'info->name' instead of 'info->traverse_path'.
Since 'info->name' only stores a single tree entry name while
'info->traverse_path' stores the full path from root, this method does
not work when 'info' is in a subdirectory of a directory. Replacing the
right strings and their corresponding lengths make the method work
properly.
The previous change included a failing test that exposes this issue.
That test now passes. The critical detail is that as we go deep into
unpack_trees(), the logic for merging a sparse directory entry with a
tree entry during 'git checkout' relies on this
sparse_dir_matches_path() in order to avoid calling
traverse_trees_recursive() during unpack_callback() in this hunk:
if (!is_sparse_directory_entry(src[0], names, info) &&
traverse_trees_recursive(n, dirmask, mask & ~dirmask,
names, info) < 0) {
return -1;
}
For deep paths, the short-circuit never occurred and
traverse_trees_recursive() was being called incorrectly and that was
causing other strange issues. Specifically, the error message from the
now-passing test previously included this:
error: Your local changes to the following files would be overwritten by checkout:
deep/deeper1/deepest2/a
deep/deeper1/deepest3/a
Please commit your changes or stash them before you switch branches.
Aborting
These messages occurred because the 'current' cache entry in
twoway_merge() was showing as NULL because the index did not contain
entries for the paths contained within the sparse directory entries. We
instead had 'oldtree' given as the entry at HEAD and 'newtree' as the
entry in the target tree. This led to reject_merge() listing these
paths.
Now that sparse_dir_matches_path() works the same for deep paths as it
does for shallow depths, the rest of the logic kicks in to properly
handle modifying the sparse directory entries as designed.
Reported-by: Gustave Granroth <gus.gran@gmail.com>
Reported-by: Mike Marcelais <michmarc@exchange.microsoft.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
To find the first non-unpacked cache entry, `next_cache_entry` iterates
through index, starting at `cache_bottom`. The performance of this in full
indexes is helped by `cache_bottom` advancing with each invocation of
`mark_ce_used` (called by `unpack_index_entry`). However, the presence of
sparse directories can prevent the `cache_bottom` from advancing in a sparse
index case, effectively forcing `next_cache_entry` to search from the
beginning of the index each time it is called.
The `cache_bottom` must be preserved for the sparse index (see 17a1bb570b
(unpack-trees: preserve cache_bottom, 2021-07-14)). Therefore, to retain the
benefit `cache_bottom` provides in non-sparse index cases, a separate `hint`
position indicates the first position `next_cache_entry` should search,
updated each execution with a new position.
Signed-off-by: Victoria Dye <vdye@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Leakfix.
* ab/unpack-trees-leakfix:
sequencer: fix a memory leak in do_reset()
sequencer: add a "goto cleanup" to do_reset()
unpack-trees: don't leak memory in verify_clean_subdirectory()
Various fixes in code paths that move untracked files away to make room.
* en/removing-untracked-fixes:
Documentation: call out commands that nuke untracked files/directories
Comment important codepaths regarding nuking untracked files/dirs
unpack-trees: avoid nuking untracked dir in way of locally deleted file
unpack-trees: avoid nuking untracked dir in way of unmerged file
Change unpack_trees' 'reset' flag into an enum
Remove ignored files by default when they are in the way
unpack-trees: make dir an internal-only struct
unpack-trees: introduce preserve_ignored to unpack_trees_options
read-tree, merge-recursive: overwrite ignored files by default
checkout, read-tree: fix leak of unpack_trees_options.dir
t2500: add various tests for nuking untracked files
Fix two different but related memory leaks in
verify_clean_subdirectory(). We leaked both the "pathbuf" if
read_directory() returned non-zero, and we never cleaned up our own
"struct dir_struct" either.
* "pathbuf": When the read_directory() call followed by the
free(pathbuf) was added in c81935348b (Fix switching to a branch
with D/F when current branch has file D., 2007-03-15) we didn't
bother to free() before we called die().
But when this code was later libified in 203a2fe117 (Allow callers
of unpack_trees() to handle failure, 2008-02-07) we started to leak
as we returned data to the caller. This fixes that memory leak,
which can be observed under SANITIZE=leak with e.g. the
"t1001-read-tree-m-2way.sh" test.
* "struct dir_struct": We've leaked the dir_struct ever since this
code was added back in c81935348b.
When that commit was written there wasn't an equivalent of
dir_clear(). Since it was added in 270be81604 (dir.c: provide
clear_directory() for reclaiming dir_struct memory, 2013-01-06)
we've omitted freeing the memory allocated here.
This memory leak could also be observed under SANITIZE=leak and the
"t1001-read-tree-m-2way.sh" test.
This makes all the test in "t1001-read-tree-m-2way.sh" pass under
"GIT_TEST_PASSING_SANITIZE_LEAK=true", we'd previously die in tests
25, 26 & 28.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Traditionally, unpack_trees_options->reset was used to signal that it
was okay to delete any untracked files in the way. This was used by
`git read-tree --reset`, but then started appearing in other places as
well. However, many of the other uses should not be deleting untracked
files in the way. Change this value to an enum so that a value of 1
(i.e. "true") can be split into two:
UNPACK_RESET_PROTECT_UNTRACKED,
UNPACK_RESET_OVERWRITE_UNTRACKED
In order to catch accidental misuses (i.e. where folks call it the way
they traditionally used to), define the special enum value of
UNPACK_RESET_INVALID = 1
which will trigger a BUG().
Modify existing callers so that
read-tree --reset
reset --hard
checkout --force
continue using the UNPACK_RESET_OVERWRITE_UNTRACKED logic, while other
callers, including
am
checkout without --force
stash (though currently dead code; reset always had a value of 0)
numerous callers from rebase/sequencer to reset_head()
will use the new UNPACK_RESET_PROTECT_UNTRACKED value.
Also, note that it has been reported that 'git checkout <treeish>
<pathspec>' currently also allows overwriting untracked files[1]. That
case should also be fixed, but it does not use unpack_trees() and thus
is outside the scope of the current changes.
[1] https://lore.kernel.org/git/15dad590-087e-5a48-9238-5d2826950506@gmail.com/
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Avoid accidental misuse or confusion over ownership by clearly making
unpack_trees_options.dir an internal-only variable.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently, every caller of unpack_trees() that wants to ensure ignored
files are overwritten by default needs to:
* allocate unpack_trees_options.dir
* flip the DIR_SHOW_IGNORED flag in unpack_trees_options.dir->flags
* call setup_standard_excludes
AND then after the call to unpack_trees() needs to
* call dir_clear()
* deallocate unpack_trees_options.dir
That's a fair amount of boilerplate, and every caller uses identical
code. Make this easier by instead introducing a new boolean value where
the default value (0) does what we want so that new callers of
unpack_trees() automatically get the appropriate behavior. And move all
the handling of unpack_trees_options.dir into unpack_trees() itself.
While preserve_ignored = 0 is the behavior we feel is the appropriate
default, we defer fixing commands to use the appropriate default until a
later commit. So, this commit introduces several locations where we
manually set preserve_ignored=1. This makes it clear where code paths
were previously preserving ignored files when they should not have been;
a future commit will flip these to instead use a value of 0 to get the
behavior we want.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In cone mode, the sparse-index code path learned to remove ignored
files (like build artifacts) outside the sparse cone, allowing the
entire directory outside the sparse cone to be removed, which is
especially useful when the sparse patterns change.
* ds/sparse-index-ignored-files:
sparse-checkout: clear tracked sparse dirs
sparse-index: add SPARSE_INDEX_MEMORY_ONLY flag
attr: be careful about sparse directories
sparse-checkout: create helper methods
sparse-index: use WRITE_TREE_MISSING_OK
sparse-index: silently return when cache tree fails
unpack-trees: fix nested sparse-dir search
sparse-index: silently return when not using cone-mode patterns
t7519: rewrite sparse index test
Code clean up to migrate callers from older advice_config[] based
API to newer advice_if_enabled() and advice_enabled() API.
* ab/retire-advice-config:
advice: move advice.graftFileDeprecated squashing to commit.[ch]
advice: remove use of global advice_add_embedded_repo
advice: remove read uses of most global `advice_` variables
advice: add enum variants for missing advice variables
The iterated search in find_cache_entry() was recently modified to
include a loop that searches backwards for a sparse directory entry that
matches the given traverse_info and name_entry. However, the string
comparison failed to actually concatenate those two strings, so this
failed to find a sparse directory when it was not a top-level directory.
This caused some errors in rare cases where a 'git checkout' spanned a
diff that modified files within the sparse directory entry, but we could
not correctly find the entry.
Helped-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Helped-by: René Scharfe <l.s.r@web.de>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Reviewed-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The 'Filtering contents...' progress report from delayed checkout is
displayed even when checkout and clone are invoked with --quiet or
--no-progress. Furthermore, it is displayed unconditionally, without
first checking whether stdout is a tty. Let's fix these issues and also
add some regression tests for the two code paths that currently use
delayed checkout: unpack_trees.c:check_updates() and
builtin/checkout.c:checkout_worktree().
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In c4a09cc9cc (Merge branch 'hw/advise-ng', 2020-03-25), a new API for
accessing advice variables was introduced and deprecated `advice_config`
in favor of a new array, `advice_setting`.
This patch ports all but two uses which read the status of the global
`advice_` variables over to the new `advice_enabled` API. We'll deal
with advice_add_embedded_repo and advice_graft_file_deprecated
separately.
Signed-off-by: Ben Boeckel <mathstuf@gmail.com>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git read-tree" had a codepath where blobs are fetched one-by-one
from the promisor remote, which has been corrected to fetch in bulk.
* jt/bulk-prefetch:
cache-tree: prefetch in partial clone read-tree
unpack-trees: refactor prefetching code
Refactor the prefetching code in unpack-trees.c into its own function,
because it will be used elsewhere in a subsequent commit.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When running unpack_trees() with a sparse index, we attempt to operate
on the index without expanding the sparse directory entries. Thus, we
operate by manipulating entire directories and passing them to the
unpack function. In the case of the 'git checkout' command, this is the
twoway_merge() function.
There are several cases in twoway_merge() that handle different
situations. One new one to add is the case of a directory/file conflict
where the directory is sparse. Before the sparse index, such a conflict
would appear as a list of file additions and deletions. Now,
twoway_merge() initializes 'current', 'oldtree', and 'newtree' from
src[0], src[1], and src[2], then sets 'oldtree' to NULL because it is
equal to the df_conflict_entry. The way to determine that we have a
directory/file conflict is to test that 'current' and 'newtree' disagree
on being sparse directory entries.
When we are in this case, we want to resolve the situation by calling
merged_entry(). This allows replacing the 'current' entry with the
'newtree' entry. This is important for cases where we want to run 'git
checkout' across the conflict and have the new HEAD represent the new
file type at that path. The first NEEDSWORK comment dropped in t1092
demonstrates this necessary behavior.
However, we still are in a confusing state when 'current' corresponds to
a staged change within a sparse directory that is not present at HEAD.
This should be atypical, because it requires adding a change outside of
the sparse-checkout cone, but it is possible. Since we are unable to
determine that this is a staged change within twoway_merge(), we cannot
add a case to reject the merge at this point. I believe this is due to
the use of df_conflict_entry in the place of 'oldtree' instead of using
the valud at HEAD, which would provide some perspective to this
decision. Any change that would allow this differentiation for staged
entries would need to involve information further up in unpack_trees().
That work should be done, sometime, because we are further confusing the
behavior of a directory/file conflict when staging a change in the
directory. The two cases 'checkout behaves oddly with df-conflict-?' in
t1092 demonstrate that even without a sparse-checkout, Git is not
consistent in its behavior. Neither of the two options seems correct,
either. This change makes the sparse-index behave differently than the
typcial sparse-checkout case, but it does match the full checkout
behavior in the df-conflict-2 case.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
During unpack_callback(), index entries are compared against tree
entries. These are matched according to names and types. One goal is to
decide if we should recurse into subtrees or simply operate on one index
entry.
In the case of a sparse-directory entry, we do not want to recurse into
that subtree and instead simply compare the trees. In some cases, we
might want to perform a merge operation on the entry, such as during
'git checkout <commit>' which wants to replace a sparse tree entry with
the tree for that path at the target commit. We extend the logic within
unpack_single_entry() to create a sparse-directory entry in this case,
and then that is sent to call_unpack_fn().
There are some subtleties in this process. For instance, we need to
update find_cache_entry() to allow finding a sparse-directory entry that
exactly matches a given path. Use the new helper method
sparse_dir_matches_path() for this. We also need to ignore conflict
markers in the case that the entries correspond to directories and we
already have a sparse directory entry.
Reviewed-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the next change, we will use this method to unpack a sparse directory
entry, so change the name to unpack_single_entry() so these entries
apply. The new name reflects that we will not recurse into trees in
order to resolve the conflicts.
Reviewed-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As we further integrate the sparse-index into unpack-trees, we need to
ensure that we compare sparse directory entries correctly with other
entries. This affects searching for an exact path as well as sorting
index entries.
Sparse directory entries contain the trailing directory separator. This
is important for the sorting, in particular. Thus, within
do_compare_entry() we stop using S_IFREG in all cases, since sparse
directories should use S_IFDIR to indicate that the comparison should
treat the entry name as a dirctory.
Within compare_entry(), it first calls do_compare_entry() to check the
leading portion of the name. When the input path is a directory name, we
could match exactly already. Thus, we should return 0 if we have an
exact string match on a sparse directory entry. The final check is a
length comparison between the strings.
Reviewed-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The cache_bottom member of 'struct unpack_trees_options' is used to
track the range of index entries corresponding to a node of the cache
tree. While recursing with traverse_by_cache_tree(), this value is
preserved on the call stack using a local and then restored as that
method returns.
The mark_ce_used() method normally modifies the cache_bottom member when
it refers to the marked cache entry. However, sparse directory entries
are stored as nodes in the cache-tree data structure as of 2de37c53
(cache-tree: integrate with sparse directory entries, 2021-03-30). Thus,
the cache_bottom will be modified as the cache-tree walk advances. Do
not update it as well within mark_ce_used().
Reviewed-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The final part of "parallel checkout".
* mt/parallel-checkout-part-3:
ci: run test round with parallel-checkout enabled
parallel-checkout: add tests related to .gitattributes
t0028: extract encoding helpers to lib-encoding.sh
parallel-checkout: add tests related to path collisions
parallel-checkout: add tests for basic operations
checkout-index: add parallel checkout support
builtin/checkout.c: complete parallel checkout support
make_transient_cache_entry(): optionally alloc from mem_pool
Allow make_transient_cache_entry() to optionally receive a mem_pool
struct in which it should allocate the entry. This will be used in the
following patch, to store some transient entries which should persist
until parallel checkout finishes.
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The checkout machinery has been taught to perform the actual
write-out of the files in parallel when able.
* mt/parallel-checkout-part-2:
parallel-checkout: add design documentation
parallel-checkout: support progress displaying
parallel-checkout: add configuration options
parallel-checkout: make it truly parallel
unpack-trees: add basic support for parallel checkout
Builds on top of the sparse-index infrastructure to mark operations
that are not ready to mark with the sparse index, causing them to
fall back on fully-populated index that they always have worked with.
* ds/sparse-index-protections: (47 commits)
name-hash: use expand_to_path()
sparse-index: expand_to_path()
name-hash: don't add directories to name_hash
revision: ensure full index
resolve-undo: ensure full index
read-cache: ensure full index
pathspec: ensure full index
merge-recursive: ensure full index
entry: ensure full index
dir: ensure full index
update-index: ensure full index
stash: ensure full index
rm: ensure full index
merge-index: ensure full index
ls-files: ensure full index
grep: ensure full index
fsck: ensure full index
difftool: ensure full index
commit: ensure full index
checkout: ensure full index
...
Make parallel checkout configurable by introducing two new settings:
checkout.workers and checkout.thresholdForParallelism. The first defines
the number of workers (where one means sequential checkout), and the
second defines the minimum number of entries to attempt parallel
checkout.
To decide the default value for checkout.workers, the parallel version
was benchmarked during three operations in the linux repo, with cold
cache: cloning v5.8, checking out v5.8 from v2.6.15 (checkout I) and
checking out v5.8 from v5.7 (checkout II). The four tables below show
the mean run times and standard deviations for 5 runs in: a local file
system on SSD, a local file system on HDD, a Linux NFS server, and
Amazon EFS (all on Linux). Each parallel checkout test was executed with
the number of workers that brings the best overall results in that
environment.
Local SSD:
Sequential 10 workers Speedup
Clone 8.805 s ± 0.043 s 3.564 s ± 0.041 s 2.47 ± 0.03
Checkout I 9.678 s ± 0.057 s 4.486 s ± 0.050 s 2.16 ± 0.03
Checkout II 5.034 s ± 0.072 s 3.021 s ± 0.038 s 1.67 ± 0.03
Local HDD:
Sequential 10 workers Speedup
Clone 32.288 s ± 0.580 s 30.724 s ± 0.522 s 1.05 ± 0.03
Checkout I 54.172 s ± 7.119 s 54.429 s ± 6.738 s 1.00 ± 0.18
Checkout II 40.465 s ± 2.402 s 38.682 s ± 1.365 s 1.05 ± 0.07
Linux NFS server (v4.1, on EBS, single availability zone):
Sequential 32 workers Speedup
Clone 240.368 s ± 6.347 s 57.349 s ± 0.870 s 4.19 ± 0.13
Checkout I 242.862 s ± 2.215 s 58.700 s ± 0.904 s 4.14 ± 0.07
Checkout II 65.751 s ± 1.577 s 23.820 s ± 0.407 s 2.76 ± 0.08
EFS (v4.1, replicated over multiple availability zones):
Sequential 32 workers Speedup
Clone 922.321 s ± 2.274 s 210.453 s ± 3.412 s 4.38 ± 0.07
Checkout I 1011.300 s ± 7.346 s 297.828 s ± 0.964 s 3.40 ± 0.03
Checkout II 294.104 s ± 1.836 s 126.017 s ± 1.190 s 2.33 ± 0.03
The above benchmarks show that parallel checkout is most effective on
repositories located on an SSD or over a distributed file system. For
local file systems on spinning disks, and/or older machines, the
parallelism does not always bring a good performance. For this reason,
the default value for checkout.workers is one, a.k.a. sequential
checkout.
To decide the default value for checkout.thresholdForParallelism,
another benchmark was executed in the "Local SSD" setup, where parallel
checkout showed to be beneficial. This time, we compared the runtime of
a `git checkout -f`, with and without parallelism, after randomly
removing an increasing number of files from the Linux working tree. The
"sequential fallback" column below corresponds to the executions where
checkout.workers was 10 but checkout.thresholdForParallelism was equal
to the number of to-be-updated files plus one (so that we end up writing
sequentially). Each test case was sampled 15 times, and each sample had
a randomly different set of files removed. Here are the results:
sequential fallback 10 workers speedup
10 files 772.3 ms ± 12.6 ms 769.0 ms ± 13.6 ms 1.00 ± 0.02
20 files 780.5 ms ± 15.8 ms 775.2 ms ± 9.2 ms 1.01 ± 0.02
50 files 806.2 ms ± 13.8 ms 767.4 ms ± 8.5 ms 1.05 ± 0.02
100 files 833.7 ms ± 21.4 ms 750.5 ms ± 16.8 ms 1.11 ± 0.04
200 files 897.6 ms ± 30.9 ms 730.5 ms ± 14.7 ms 1.23 ± 0.05
500 files 1035.4 ms ± 48.0 ms 677.1 ms ± 22.3 ms 1.53 ± 0.09
1000 files 1244.6 ms ± 35.6 ms 654.0 ms ± 38.3 ms 1.90 ± 0.12
2000 files 1488.8 ms ± 53.4 ms 658.8 ms ± 23.8 ms 2.26 ± 0.12
From the above numbers, 100 files seems to be a reasonable default value
for the threshold setting.
Note: Up to 1000 files, we observe a drop in the execution time of the
parallel code with an increase in the number of files. This is a rather
odd behavior, but it was observed in multiple repetitions. Above 1000
files, the execution time increases according to the number of files, as
one would expect.
About the test environments: Local SSD tests were executed on an
i7-7700HQ (4 cores with hyper-threading) running Manjaro Linux. Local
HDD tests were executed on an Intel(R) Xeon(R) E3-1230 (also 4 cores
with hyper-threading), HDD Seagate Barracuda 7200.14 SATA 3.1, running
Debian. NFS and EFS tests were executed on an Amazon EC2 c5n.xlarge
instance, with 4 vCPUs. The Linux NFS server was running on a m6g.large
instance with 2 vCPUSs and a 1 TB EBS GP2 volume. Before each timing,
the linux repository was removed (or checked out back to its previous
state), and `sync && sysctl vm.drop_caches=3` was executed.
Co-authored-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This new interface allows us to enqueue some of the entries being
checked out to later uncompress them, apply in-process filters, and
write out the files in parallel. For now, the parallel checkout
machinery is enabled by default and there is no user configuration, but
run_parallel_checkout() just writes the queued entries in sequence
(without spawning additional workers). The next patch will actually
implement the parallelism and, later, we will make it configurable.
Note that, to avoid potential data races, not all entries are eligible
for parallel checkout. Also, paths that collide on disk (e.g.
case-sensitive paths in case-insensitive file systems), are detected by
the parallel checkout code and skipped, so that they can be safely
sequentially handled later. The collision detection works like the
following:
- If the collision was at basename (e.g. 'a/b' and 'a/B'), the framework
detects it by looking for EEXIST and EISDIR errors after an
open(O_CREAT | O_EXCL) failure.
- If the collision was at dirname (e.g. 'a/b' and 'A'), it is detected
at the has_dirs_only_path() check, which is done for the leading path
of each item in the parallel checkout queue.
Both verifications rely on the fact that, before enqueueing an entry for
parallel checkout, checkout_entry() makes sure that there is no file at
the entry's path and that its leading components are all real
directories. So, any later change in these conditions indicates that
there was a collision (either between two parallel-eligible entries or
between an eligible and an ineligible one).
After all parallel-eligible entries have been processed, the collided
(and thus, skipped) entries are sequentially fed to checkout_entry()
again. This is similar to the way the current code deals with
collisions, overwriting the previously checked out entries with the
subsequent ones. The only difference is that, since we no longer create
the files in the same order that they appear on index, we are not able
to determine which of the colliding entries will survive on disk (for
the classic code, it is always the last entry).
Co-authored-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Co-authored-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Preparatory API changes for parallel checkout.
* mt/parallel-checkout-part-1:
entry: add checkout_entry_ca() taking preloaded conv_attrs
entry: move conv_attrs lookup up to checkout_entry()
entry: extract update_ce_after_write() from write_entry()
entry: make fstat_output() and read_blob_entry() public
entry: extract a header file for entry.c functions
convert: add classification for conv_attrs struct
convert: add get_stream_filter_ca() variant
convert: add [async_]convert_to_working_tree_ca() variants
convert: make convert_attrs() and convert structs public
When "git checkout" removes a path that does not exist in the
commit it is checking out, it wasn't careful enough not to follow
symbolic links, which has been corrected.
* mt/checkout-remove-nofollow:
checkout: don't follow symlinks when removing entries
symlinks: update comment on threaded_check_leading_path()
The index_pos_by_traverse_info() currently throws a BUG() when a
directory entry exists exactly in the index. We need to consider that it
is possible to have a directory in a sparse index as long as that entry
is itself marked with the skip-worktree bit.
The 'pos' variable is assigned a negative value if an exact match is not
found. Since a directory name can be an exact match, it is no longer an
error to have a nonnegative 'pos' value.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The next change will translate full indexes into sparse indexes at write
time. The existing logic provides a way for every sparse index to be
expanded to a full index at read time. However, there are cases where an
index is written and then continues to be used in-memory to perform
further updates.
unpack_trees() is frequently called after such a write. In particular,
commands like 'git reset' do this double-update of the index.
Ensure that we have a full index when entering unpack_trees(), but only
when command_requires_full_index is true. This is always true at the
moment, but we will later relax that after unpack_trees() is updated to
handle sparse directory entries.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The declarations of entry.c's public functions and structures currently
reside in cache.h. Although not many, they contribute to the size of
cache.h and, when changed, cause the unnecessary recompilation of
modules that don't really use these functions. So let's move them to a
new entry.h header. While at it let's also move a comment related to
checkout_entry() from entry.c to entry.h as it's more useful to describe
the function there.
Original-patch-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git stash show" learned to optionally show untracked part of the
stash.
* dl/stash-show-untracked:
stash show: learn stash.showIncludeUntracked
stash show: teach --include-untracked and --only-untracked
The data structure used by fsmonitor interface was not properly
duplicated during an in-core merge, leading to use-after-free etc.
* js/fsmonitor-unpack-fix:
fsmonitor: do not forget to release the token in `discard_index()`
fsmonitor: fix memory corruption in some corner cases
At 1d718a5108 ("do not overwrite untracked symlinks", 2011-02-20),
symlink.c:check_leading_path() started returning different codes for
FL_ENOENT and FL_SYMLINK. But one of its callers, unlink_entry(), was
not adjusted for this change, so it started to follow symlinks on the
leading path of to-be-removed entries. Fix that and add a regression
test.
Note that since 1d718a5108 check_leading_path() no longer differentiates
the case where it found a symlink in the path's leading components from
the cases where it found a regular file or failed to lstat() the
component. So, a side effect of this current patch is that
unlink_entry() now returns early in all of these three cases. And
because we no longer try to unlink such paths, we also don't get the
warning from remove_or_warn().
For the regular file and symlink cases, it's questionable whether the
warning was useful in the first place: unlink_entry() removes tracked
paths that should no longer be present in the state we are checking out
to. If the path had its leading dir replaced by another file, it means
that the basename already doesn't exist, so there is no need for a
warning. Sure, we are leaving a regular file or symlink behind at the
path's dirname, but this file is either untracked now (so again, no
need to warn), or it will be replaced by a tracked file during the next
phase of this checkout operation.
As for failing to lstat() one of the leading components, the basename
might still exist only we cannot unlink it (e.g. due to the lack of the
required permissions). Since the user expect it to be removed
(especially with checkout's --no-overlay option), add back the warning
in this more relevant case.
Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 56c6910028 (fsmonitor: change last update timestamp on the
index_state to opaque token, 2020-01-07), we forgot to adjust the part
of `unpack_trees()` that copies the FSMonitor "last-update" information
that we copy from the source index to the result index since 679f2f9fdd
(unpack-trees: skip stat on fsmonitor-valid files, 2019-11-20).
Since the "last-update" information is no longer a 64-bit number, but a
free-form string that has been allocated, we need to duplicate it rather
than just copying it.
This is important because there _are_ cases when `unpack_trees()` will
perform a oneway merge that implicitly calls `refresh_fsmonitor()`
(which will allocate that "last-update" token). This happens _after_
that token was copied into the result index. However, we _then_ call
`check_updates()` on that index, which will _also_ call
`refresh_fsmonitor()`, accessing the "last-update" string, which by now
would be released already.
In the instance that lead to this patch, this caused a segmentation
fault during a lengthy, complicated rebase involving the todo command
`reset` that (crucially) had to updated many files. Unfortunately, it
seems very hard to trigger that crash, therefore this patch is not
accompanied by a regression test.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stash entries can be made with untracked files via
`git stash push --include-untracked`. However, because the untracked
files are stored in the third parent of the stash entry and not the
stash entry itself, running `git stash show` does not include the
untracked files as part of the diff.
With --include-untracked, untracked paths, which are recorded in the
third-parent if it exists, are shown in addition to the paths that have
modifications between the stash base and the working tree in the stash.
It is possible to manually craft a malformed stash entry where duplicate
untracked files in the stash entry will mask tracked files. We detect
and error out in that case via a custom unpack_trees() callback:
stash_worktree_untracked_merge().
Also, teach stash the --only-untracked option which only shows the
untracked files of a stash entry. This is similar to `git show stash^3`
but it is nice to provide a convenient abstraction for it so that users
do not have to think about the underlying implementation.
Signed-off-by: Denton Liu <liu.denton@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
