Fix regression described in:
https://github.com/git-for-windows/git/issues/1392
which was introduced in:
b2353379bb
Problem Symptoms
================
When the user has a .gitignore file that is a symlink, the fscache
optimization introduced above caused the stat-data from the symlink,
rather that of the target file, to be returned. Later when the ignore
file was read, the buffer length did not match the stat.st_size field
and we called die("cannot use <path> as an exclude file")
Optimization Rationale
======================
The above optimization calls lstat() before open() primarily to ask
fscache if the file exists. It gets the current stat-data as a side
effect essentially for free (since we already have it in memory).
If the file does not exist, it does not need to call open(). And
since very few directories have .gitignore files, we can greatly
reduce time spent in the filesystem.
Discussion of Fix
=================
The above optimization calls lstat() rather than stat() because the
fscache only intercepts lstat() calls. Calls to stat() stay directed
to the mingw_stat() completly bypassing fscache. Furthermore, calls
to mingw_stat() always call {open, fstat, close} so that symlinks are
properly dereferenced, which adds *additional* open/close calls on top
of what the original code in dir.c is doing.
Since the problem only manifests for symlinks, we add code to overwrite
the stat-data when the path is a symlink. This preserves the effect of
the performance gains provided by the fscache in the normal case.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Make fscache_enabled() function public rather than static.
Remove unneeded fscache_is_enabled() function.
Change is_fscache_enabled() macro to call fscache_enabled().
is_fscache_enabled() now takes a pathname so that the answer
is more precise and mean "is fscache enabled for this pathname",
since fscache only stores repo-relative paths and not absolute
paths, we can avoid attempting lookups for absolute paths.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Teach read_directory_recursive() and add_excludes() to
be aware of optional fscache and avoid trying to open()
and fstat() non-existant ".gitignore" files in every
directory in the worktree.
The current code in add_excludes() calls open() and then
fstat() for a ".gitignore" file in each directory present
in the worktree. Change that when fscache is enabled to
call lstat() first and if present, call open().
This seems backwards because both lstat needs to do more
work than fstat. But when fscache is enabled, fscache will
already know if the .gitignore file exists and can completely
avoid the IO calls. This works because of the lstat diversion
to mingw_lstat when fscache is enabled.
This reduced status times on a 350K file enlistment of the
Windows repo on a NVMe SSD by 0.25 seconds.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Teach "add" to use preload-index and fscache features
to improve performance on very large repositories.
During an "add", a call is made to run_diff_files()
which calls check_remove() for each index-entry. This
calls lstat(). On Windows, the fscache code intercepts
the lstat() calls and builds a private cache using the
FindFirst/FindNext routines, which are much faster.
Somewhat independent of this, is the preload-index code
which distributes some of the start-up costs across
multiple threads.
We need to keep the call to read_cache() before parsing the
pathspecs (and hence cannot use the pathspecs to limit any preload)
because parse_pathspec() is using the index to determine whether a
pathspec is, in fact, in a submodule. If we would not read the index
first, parse_pathspec() would not error out on a path that is inside
a submodule, and t7400-submodule-basic.sh would fail with
not ok 47 - do not add files from a submodule
We still want the nice preload performance boost, though, so we simply
call read_cache_preload(&pathspecs) after parsing the pathspecs.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Teach FSCACHE to remember "not found" directories.
This is a performance optimization.
FSCACHE is a performance optimization available for Windows. It
intercepts Posix-style lstat() calls into an in-memory directory
using FindFirst/FindNext. It improves performance on Windows by
catching the first lstat() call in a directory, using FindFirst/
FindNext to read the list of files (and attribute data) for the
entire directory into the cache, and short-cut subsequent lstat()
calls in the same directory. This gives a major performance
boost on Windows.
However, it does not remember "not found" directories. When STATUS
runs and there are missing directories, the lstat() interception
fails to find the parent directory and simply return ENOENT for the
file -- it does not remember that the FindFirst on the directory
failed. Thus subsequent lstat() calls in the same directory, each
re-attempt the FindFirst. This completely defeats any performance
gains.
This can be seen by doing a sparse-checkout on a large repo and
then doing a read-tree to reset the skip-worktree bits and then
running status.
This change reduced status times for my very large repo by 60%.
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
If multiple threads access a directory that is not yet in the cache, the
directory will be loaded by each thread. Only one of the results is added
to the cache, all others are leaked. This wastes performance and memory.
On cache miss, add a future object to the cache to indicate that the
directory is currently being loaded. Subsequent threads register themselves
with the future object and wait. When the first thread has loaded the
directory, it replaces the future object with the result and notifies
waiting threads.
Signed-off-by: Karsten Blees <blees@dcon.de>
Checking the work tree status is quite slow on Windows, due to slow
`lstat()` emulation (git calls `lstat()` once for each file in the
index). Windows operating system APIs seem to be much better at scanning
the status of entire directories than checking single files.
Add an `lstat()` implementation that uses a cache for lstat data. Cache
misses read the entire parent directory and add it to the cache.
Subsequent `lstat()` calls for the same directory are served directly
from the cache.
Also implement `opendir()`/`readdir()`/`closedir()` so that they create
and use directory listings in the cache.
The cache doesn't track file system changes and doesn't plug into any
modifying file APIs, so it has to be explicitly enabled for git functions
that don't modify the working copy.
Note: in an earlier version of this patch, the cache was always active and
tracked file system changes via ReadDirectoryChangesW. However, this was
much more complex and had negative impact on the performance of modifying
git commands such as 'git checkout'.
Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Add a macro to mark code sections that only read from the file system,
along with a config option and documentation.
This facilitates implementation of relatively simple file system level
caches without the need to synchronize with the file system.
Enable read-only sections for 'git status' and preload_index.
Signed-off-by: Karsten Blees <blees@dcon.de>
Emulating the POSIX lstat API on Windows via GetFileAttributes[Ex] is quite
slow. Windows operating system APIs seem to be much better at scanning the
status of entire directories than checking single files. A caching
implementation may improve performance by bulk-reading entire directories
or reusing data obtained via opendir / readdir.
Make the lstat implementation pluggable so that it can be switched at
runtime, e.g. based on a config option.
Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Emulating the POSIX `dirent` API on Windows via
`FindFirstFile()`/`FindNextFile()` is pretty staightforward, however,
most of the information provided in the `WIN32_FIND_DATA` structure is
thrown away in the process. A more sophisticated implementation may
cache this data, e.g. for later reuse in calls to `lstat()`.
Make the `dirent` implementation pluggable so that it can be switched at
runtime, e.g. based on a config option.
Define a base DIR structure with pointers to `readdir()`/`closedir()`
that match the `opendir()` implementation (similar to vtable pointers in
Object-Oriented Programming). Define `readdir()`/`closedir()` so that
they call the function pointers in the `DIR` structure. This allows to
choose the `opendir()` implementation on a call-by-call basis.
Make the fixed-size `dirent.d_name` buffer a flex array, as `d_name` may
be implementation specific (e.g. a caching implementation may allocate a
`struct dirent` with _just_ the size needed to hold the `d_name` in
question).
Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
We will use them in the upcoming "FSCache" patches (to accelerate
sequential lstat() calls).
Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
In MSYS2, we have two Python interpreters at our disposal, so we can
include the Python stuff in the build.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Make `git config --system` work like you think it should on Windows: it
should edit the file that is located in `<Git>\etc\gitconfig`.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
It would appear that least the Isilon network filesystem (and possibly
other network filesystems, too), report non-standard error values when
trying to access a non-existing directory.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This topic branch teaches `git clean` to respect NTFS junctions and Unix
bind mounts: it will now stop at those boundaries.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
In Git for Windows, there is an option to mark the .git directory as
hidden. Our test cases verify this by using the system utility
`attrib.exe`.
This file name is unfortunately quite generic, and overlaps with a
Unix-y utility that might be hiding the system one in the `PATH`.
Let's specify explicitly which `attrib` to use.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
After writing to `stdout` and before reading from `stdin`, it is a good
idea to flush the former.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
It is a good idea in general to avoid pipes in test cases, as it makes
things more debuggable to have files to inspect (instead of ephemereal
piped data that is long gone).
This also seemed to work around a problem where MSYS2' Perl would
segfault which may, or may not, still be present today.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This topic branch avoids spawning `gzip` when asking `git archive` to
create `.tar.gz` files.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
When compiling Git with a runtime prefix (so that it can be installed
into any location, finding its libexec/ directory relative to the
location of the `git` executable), it is convenient to provide
"absolute" Unix-y paths e.g. for http.sslCAInfo, and have those absolute
paths be resolved relative to the runtime prefix.
This patch makes it so for Windows. It is up for discussion whether we
want this for other platforms, too, as long as building with
RUNTIME_PREFIX.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This topic branch allows us to specify absolute paths without the drive
prefix e.g. when cloning.
Example:
C:\Users\me> git clone https://github.com/git/git \upstream-git
This will clone into a new directory C:\upstream-git, in line with how
Windows interprets absolute paths.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
These fixes were necessary for Sverre Rabbelier's remote-hg to work,
but for some magic reason they are not necessary for the current
remote-hg. Makes you wonder how that one gets away with it.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
After importing anything with fast-import, we should always let the
garbage collector do its job, since the objects are written to disk
inefficiently.
This brings down an initial import of http://selenic.com/hg from about
230 megabytes to about 14.
In the future, we may want to make this configurable on a per-remote
basis, or maybe teach fast-import about it in the first place.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This topic branch fixes a corner case that is amazingly common in this
developer's workflow: in a `git stash -p`, splitting a hunk and stashing
only part of it runs into a (known) bug where the partial hunk cannot be
applied in reverse.
It is one of those "good enough" fixes, not a full fix, though, as the
full fix would require a 3-way merge between `stash^` and the *worktree*
(not `HEAD`), with `stash` as merge base (i.e. a `git revert`, but on
top of the current worktree).
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This is the final leg of the journey to a fully built-in `git add`: the
`git add -i` and `git add -p` modes were re-implemented in C, but they
lacked support for a couple of config settings.
The one that sticks out most is the `interactive.singleKey` setting: it
was not only particularly hard to get to work, especially on Windows. It
is also the setting that seems to be incomplete already in the Perl
version: while the name suggests that it applies to the main loop of
`git add --interactive`, or to the file selections in that command, it
does not. Only the `git add --patch` mode respects that setting.
As it is outside the purpose of the conversion of
`git-add--interactive.perl` to C, we will leave that loose end for some
future date.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
When trying to stash part of the worktree changes by splitting a hunk
and then only partially accepting the split bits and pieces, the user
is presented with a rather cryptic error:
error: patch failed: <file>:<line>
error: test: patch does not apply
Cannot remove worktree changes
and the command would fail to stash the desired parts of the worktree
changes (even if the `stash` ref was actually updated correctly).
We even have a test case demonstrating that failure, carrying it for
four years already.
The explanation: when splitting a hunk, the changed lines are no longer
separated by more than 3 lines (which is the amount of context lines
Git's diffs use by default), but less than that. So when staging only
part of the diff hunk for stashing, the resulting diff that we want to
apply to the worktree in reverse will contain those changes to be
dropped surrounded by three context lines, but since the diff is
relative to HEAD rather than to the worktree, these context lines will
not match.
Example time. Let's assume that the file README contains these lines:
We
the
people
and the worktree added some lines so that it contains these lines
instead:
We
are
the
kind
people
and the user tries to stash the line containing "are", then the command
will internally stage this line to a temporary index file and try to
revert the diff between HEAD and that index file. The diff hunk that
`git stash` tries to revert will look somewhat like this:
@@ -1776,3 +1776,4
We
+are
the
people
It is obvious, now, that the trailing context lines overlap with the
part of the original diff hunk that the user did *not* want to stash.
Keeping in mind that context lines in diffs serve the primary purpose of
finding the exact location when the diff does not apply precisely (but
when the exact line number in the file to be patched differs from the
line number indicated in the diff), we work around this by reducing the
amount of context lines: the diff was just generated.
Note: this is not a *full* fix for the issue. Just as demonstrated in
t3701's 'add -p works with pathological context lines' test case, there
are ambiguities in the diff format. It is very rare in practice, of
course, to encounter such repeated lines.
The full solution for such cases would be to replace the approach of
generating a diff from the stash and then applying it in reverse by
emulating `git revert` (i.e. doing a 3-way merge). However, in `git
stash -p` it would not apply to `HEAD` but instead to the worktree,
which makes this non-trivial to implement as long as we also maintain a
scripted version of `add -i`.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
In 7e9e048661 (stash -p: demonstrate failure of split with mixed y/n,
2015-04-16), a regression test for a known breakage that was added to
the test script `t3904-stash-patch.sh` that demonstrated that splitting
a hunk and trying to stash only part of that split hunk fails (but
shouldn't).
As expected, it still fails, but for the wrong reason: once the bug is
fixed, we would expect stderr to show nothing, yet the regression test
expects stderr to show something.
Let's fix that by telling that regression test case to expect nothing to
be printed to stderr.
While at it, also drop the obvious left-over from debugging where the
regression test did not mind `git stash -p` to return a non-zero exit
status.
Of course, the regression test still fails, but this time for the
correct reason.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This job runs the test suite twice, once in regular mode, and once with
a whole slew of `GIT_TEST_*` variables set.
Now that the built-in version of `git add --interactive` is
feature-complete, let's also throw `GIT_TEST_MULTI_PACK_INDEX` into that
fray.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
When `interactive.singlekey = true`, we react immediately to keystrokes,
even to Escape sequences (e.g. when pressing a cursor key).
The problem with Escape sequences is that we do not really know when
they are done, and as a heuristic we poll standard input for half a
second to make sure that we got all of it.
While waiting half a second is not asking for a whole lot, it can become
quite annoying over time, therefore with this patch, we read the
terminal capabilities (if available) and extract known Escape sequences
from there, then stop polling immediately when we detected that the user
pressed a key that generated such a known sequence.
This recapitulates the remaining part of b5cc003253 (add -i: ignore
terminal escape sequences, 2011-05-17).
Note: We do *not* query the terminal capabilities directly. That would
either require a lot of platform-specific code, or it would require
linking to a library such as ncurses.
Linking to a library in the built-ins is something we try very hard to
avoid (we even kicked the libcurl dependency to a non-built-in remote
helper, just to shave off a tiny fraction of a second from Git's startup
time). And the platform-specific code would be a maintenance nightmare.
Even worse: in Git for Windows' case, we would need to query MSYS2
pseudo terminals, which `git.exe` simply cannot do (because it is
intentionally *not* an MSYS2 program).
To address this, we simply spawn `infocmp -L -1` and parse its output
(which works even in Git for Windows, because that helper is included in
the end-user facing installations).
This is done only once, as in the Perl version, but it is done only when
the first Escape sequence is encountered, not upon startup of `git add
-i`; This saves on startup time, yet makes reacting to the first Escape
sequence slightly more sluggish. But it allows us to keep the
terminal-related code encapsulated in the `compat/terminal.c` file.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
This recapitulates part of b5cc003253 (add -i: ignore terminal escape
sequences, 2011-05-17):
add -i: ignore terminal escape sequences
On the author's terminal, the up-arrow input sequence is ^[[A, and
thus fat-fingering an up-arrow into 'git checkout -p' is quite
dangerous: git-add--interactive.perl will ignore the ^[ and [
characters and happily treat A as "discard everything".
As a band-aid fix, use Term::Cap to get all terminal capabilities.
Then use the heuristic that any capability value that starts with ^[
(i.e., \e in perl) must be a key input sequence. Finally, given an
input that starts with ^[, read more characters until we have read a
full escape sequence, then return that to the caller. We use a
timeout of 0.5 seconds on the subsequent reads to avoid getting stuck
if the user actually input a lone ^[.
Since none of the currently recognized keys start with ^[, the net
result is that the sequence as a whole will be ignored and the help
displayed.
Note that we leave part for later which uses "Term::Cap to get all
terminal capabilities", for several reasons:
1. it is actually not really necessary, as the timeout of 0.5 seconds
should be plenty sufficient to catch Escape sequences,
2. it is cleaner to keep the change to special-case Escape sequences
separate from the change that reads all terminal capabilities to
speed things up, and
3. in practice, relying on the terminal capabilities is a bit overrated,
as the information could be incomplete, or plain wrong. For example,
in this developer's tmux sessions, the terminal capabilities claim
that the "cursor up" sequence is ^[M, but the actual sequence
produced by the "cursor up" key is ^[[A.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
The Perl version of `git add -p` supports this config setting to allow
users to input commands via single characters (as opposed to having to
press the <Enter> key afterwards).
This is an opt-in feature because it requires Perl packages
(Term::ReadKey and Term::Cap, where it tries to handle an absence of the
latter package gracefully) to work. Note that at least on Ubuntu, that
Perl package is not installed by default (it needs to be installed via
`sudo apt-get install libterm-readkey-perl`), so this feature is
probably not used a whole lot.
In C, we obviously do not have these packages available, but we just
introduced `read_single_keystroke()` that is similar to what
Term::ReadKey provides, and we use that here.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Typically, input on the command-line is line-based. It is actually not
really easy to get single characters (or better put: keystrokes).
We provide two implementations here:
- One that handles `/dev/tty` based systems as well as native Windows.
The former uses the `tcsetattr()` function to put the terminal into
"raw mode", which allows us to read individual keystrokes, one by one.
The latter uses `stty.exe` to do the same, falling back to direct
Win32 Console access.
Thanks to the refactoring leading up to this commit, this is a single
function, with the platform-specific details hidden away in
conditionally-compiled code blocks.
- A fall-back which simply punts and reads back an entire line.
Note that the function writes the keystroke into an `strbuf` rather than
a `char`, in preparation for reading Escape sequences (e.g. when the
user hit an arrow key). This is also required for UTF-8 sequences in
case the keystroke corresponds to a non-ASCII letter.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
