Many call sites use strbuf_init(&foo, 0) to initialize local
strbuf variable "foo" which has not been accessed since its
declaration. These can be replaced with a static initialization
using the STRBUF_INIT macro which is just as readable, saves a
function call, and takes up fewer lines.
Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If verification of path failed, it is always better to print an
error message saying this than relying on the caller function to
print a meaningful error message (especially when the callee already
prints error message for another situation).
Because the callers of add_index_entry_with_check() did not print
any error message, it resulted that the user would not notice the
problem when checkout of an invalid path failed.
Signed-off-by: Dmitry Potapov <dpotapov@gmail.com>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
On ARM I have the following compilation errors:
CC fast-import.o
In file included from cache.h:8,
from builtin.h:6,
from fast-import.c:142:
arm/sha1.h:14: error: conflicting types for 'SHA_CTX'
/usr/include/openssl/sha.h:105: error: previous declaration of 'SHA_CTX' was here
arm/sha1.h:16: error: conflicting types for 'SHA1_Init'
/usr/include/openssl/sha.h:115: error: previous declaration of 'SHA1_Init' was here
arm/sha1.h:17: error: conflicting types for 'SHA1_Update'
/usr/include/openssl/sha.h:116: error: previous declaration of 'SHA1_Update' was here
arm/sha1.h:18: error: conflicting types for 'SHA1_Final'
/usr/include/openssl/sha.h:117: error: previous declaration of 'SHA1_Final' was here
make: *** [fast-import.o] Error 1
This is because openssl header files are always included in
git-compat-util.h since commit 684ec6c63c whenever NO_OPENSSL is not
set, which somehow brings in <openssl/sha1.h> clashing with the custom
ARM version. Compilation of git is probably broken on PPC too for the
same reason.
Turns out that the only file requiring openssl/ssl.h and openssl/err.h
is imap-send.c. But only moving those problematic includes there
doesn't solve the issue as it also includes cache.h which brings in the
conflicting local SHA1 header file.
As suggested by Jeff King, the best solution is to rename our references
to SHA1 functions and structure to something git specific, and define those
according to the implementation used.
Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
This adds "--intent-to-add" option to "git add". This is to let the
system know that you will tell it the final contents to be staged later,
iow, just be aware of the presense of the path with the type of the blob
for now. It is implemented by staging an empty blob as the content.
With this sequence:
$ git reset --hard
$ edit newfile
$ git add -N newfile
$ edit newfile oldfile
$ git diff
the diff will show all changes relative to the current commit. Then you
can do:
$ git commit -a ;# commit everything
or
$ git commit oldfile ;# only oldfile, newfile not yet added
to pretend you are working with an index-free system like CVS.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* maint:
unpack_trees(): protect the handcrafted in-core index from read_cache()
git-p4: Fix one-liner in p4_write_pipe function.
Completion: add missing '=' for 'diff --diff-filter'
Fix 'git help help'
unpack_trees() rebuilds the in-core index from scratch by allocating a new
structure and finishing it off by copying the built one to the final
index.
The resulting in-core index is Ok for most use, but read_cache() does not
recognize it as such. The function is meant to be no-op if you already
have loaded the index, until you call discard_cache().
This change the way read_cache() detects an already initialized in-core
index, by introducing an extra bit, and marks the handcrafted in-core
index as initialized, to avoid this problem.
A better fix in the longer term would be to change the read_cache() API so
that it will always discard and re-read from the on-disk index to avoid
confusion. But there are higher level API that have relied on the current
semantics, and they and their users all need to get converted, which is
outside the scope of 'maint' track.
An example of such a higher level API is write_cache_as_tree(), which is
used by git-write-tree as well as later Porcelains like git-merge, revert
and cherry-pick. In the longer term, we should remove read_cache() from
there and add one to cmd_write_tree(); other callers expect that the
in-core index they prepared is what gets written as a tree so no other
change is necessary for this particular codepath.
The original version of this patch marked the index by pointing an
otherwise wasted malloc'ed memory with o->result.alloc, but this version
uses Linus's idea to use a new "initialized" bit, which is conceptually
much cleaner.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We do not have any more bits in the on-disk index flags word, but we would
need to have more in the future. Use the last remaining bits as a signal
to tell us that the index entry we are looking at is an extended one.
Since we do not understand the extended format yet, we will just error out
when we see it.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The ie_modified() function is the workhorse for refresh_cache_entry(),
i.e. checking if an index entry that is stat-dirty actually has changes.
After running quicker check to compare cached stat information with
results from the latest lstat(2) to answer "has modification" early, the
code goes on to check if there really is a change by comparing the staged
data with what is on the filesystem by asking ce_modified_check_fs().
However, this function always said "no change" for any gitlinks that has a
directory at the corresponding path. This made ie_modified() to miss
actual changes in the subproject.
The patch fixes this first by modifying an existing short-circuit logic
before calling the ce_modified_check_fs() function. It knows that for any
filesystem entity to which ie_match_stat() says its data has changed, if
its cached size is nonzero then the contents cannot match, which is a
correct optimization only for blob objects. We teach gitlink objects to
this special case, as we already know that any gitlink that
ie_match_stat() says is modified is indeed modified at this point in the
codepath.
With the above change, we could leave ce_modified_check_fs() broken, but
it also futureproofs the code by teaching it to use ce_compare_gitlink(),
instead of assuming (incorrectly) that any directory is unchanged.
Originally noticed by Alex Riesen on Cygwin.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A new configuration variable 'core.trustctime' is introduced to
allow ignoring st_ctime information when checking if paths
in the working tree has changed, because there are situations where
it produces too much false positives. Like when file system crawlers
keep changing it when scanning and using the ctime for marking scanned
files.
The default is to notice ctime changes.
Signed-off-by: Alex Riesen <raa.lkml@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The rewrite of git-mv from a shell script to a builtin was perhaps
a little too straightforward: the git add and git rm queues were
emulated directly, which resulted in a rather complicated code and
caused an inconsistent behaviour when moving dirty index entries;
git mv would update the entry based on working tree state,
except in case of overwrites, where the new entry would still have
sha1 of the old file.
This patch introduces rename_index_entry_at() into the index toolkit,
which will rename an entry while removing any entries the new entry
might render duplicate. This is then used in git mv instead
of all the file queues, resulting in a major simplification
of the code and an inevitable change in git mv -n output format.
Also the code used to refuse renaming overwriting symlink with a regular
file and vice versa; there is no need for that.
A few new tests have been added to the testsuite to reflect this change.
Signed-off-by: Petr Baudis <pasky@suse.cz>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A private function add_files_to_cache() in builtin-add.c was borrowed by
checkout and commit re-implementors without getting properly refactored to
more library-ish place. This does the refactoring.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git update-index --refresh", "git reset" and "git add --refresh" have
reported paths that have local modifications as "needs update" since the
beginning of git.
Although this is logically correct in that you need to update the index at
that path before you can commit that change, it is now becoming more and
more clear, especially with the continuous push for user friendliness
since 1.5.0 series, that the message is suboptimal. After all, the change
may be something the user might want to get rid of, and "updating" would
be absolutely a wrong thing to do if that is the case.
I prepared two alternatives to solve this. Both aim to reword the message
to more neutral "locally modified".
This patch is a more intrusive variant that changes the message for only
Porcelain commands ("add" and "reset") while keeping the plumbing
"update-index" intact.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
builtin-read-tree has a read_cache_unmerged() which is useful for other
builtins, for example builtin-merge uses it as well. Move it to
read-cache.c to avoid code duplication.
Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We use size=0 as the magic token to say the entry is known to be racily
clean, but a sequence that does:
- update the path with a non-empty blob and write the index;
- update an unrelated path and write the index -- this smudges
the above entry;
- truncate the path to size zero.
would make both the size field for the path in the index and the size on
the filesystem zero. We should not mistake it as a clean index entry.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When a cache entry has been marked as CE_VALID, the user has
promised us that any change in the work tree does not matter.
Just mark the entry as up-to-date, and continue.
Signed-off-by: Marius Storm-Olsen <marius@trolltech.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* jc/add-n-u:
Make git add -n and git -u -n output consistent
"git-add -n -u" should not add but just report
Conflicts:
builtin-add.c
builtin-mv.c
cache.h
read-cache.c
* js/ignore-submodule:
Ignore dirty submodule states during rebase and stash
Teach update-index about --ignore-submodules
diff options: Introduce --ignore-submodules
Like with the diff machinery, update-index should sometimes just
ignore submodules (e.g. to determine a clean state before a rebase).
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Update the programs which used the function (as add_file_to_cache).
Signed-off-by: Alex Riesen <raa.lkml@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The commit sequence used to do
if (file_exists(p->path))
add_file_to_cache(p->path, 0);
where both "file_exists()" and "add_file_to_cache()" needed to do a
lstat() on the path to do their work.
This cuts down 'lstat()' calls for the partial commit case by two
for each path we know about (because we do this twice per path).
Just move the lstat() to the caller instead (that's all that
"file_exists()" really does), and pass the stat information down to the
add_to_cache() function.
This essentially makes 'add_to_index()' the core function that adds a path
to the index, getting the index pointer, the pathname and the stat
information as arguments. There are then shorthand helper functions that
use this core function:
- 'add_to_cache()' is just 'add_to_index()' with the default index
- 'add_file_to_cache/index()' is the same, but does the lstat() call
itself, so you can pass just the pathname if you don't already have the
stat information available.
So old users of the 'add_file_to_xyzzy()' are essentially left unchanged,
and this just exposes the more generic helper function that can take
existing stat information into account.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* py/diff-submodule:
is_racy_timestamp(): do not check timestamp for gitlinks
diff-lib.c: rename check_work_tree_entity()
diff: a submodule not checked out is not modified
Add t7506 to test submodule related functions for git-status
t4027: test diff for submodule with empty directory
* lt/case-insensitive:
Make git-add behave more sensibly in a case-insensitive environment
When adding files to the index, add support for case-independent matches
Make unpack-tree update removed files before any updated files
Make branch merging aware of underlying case-insensitive filsystems
Add 'core.ignorecase' option
Make hash_name_lookup able to do case-independent lookups
Make "index_name_exists()" return the cache_entry it found
Move name hashing functions into a file of its own
Make unpack_trees_options bit flags actual bitfields
Because we do not even check the timestamp to determie if a gitlink
is up to date or not, triggering the racy-timestamp check for gitlinks
does not make sense.
This fixes the recently added test in t7506.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When writing the index out, we need to check the work tree again to see if
an entry whose timestamp indicates that it could be "racily clean", in
order to smudge it if it is stat-clean but with modified contents.
However, we can skip this step for entries marked with CE_UPTODATE,
which are known to be the really clean (i.e. the one we already have
checked when we prepared the index). This will reduce lstat(2) calls
necessary in git-status.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This expands on the previous patch, and allows "git add" to sanely handle
a filename that has changed case, keeping the case in the index constant,
and avoiding aliases.
In particular, if you have an index entry called "File", but the
checked-out tree is case-corrupted and has an entry called "file"
instead, doing a
git add .
(or naming "file" explicitly) will automatically notice that we have an
alias, and will replace the name "file" with the existing index
capitalization (ie "File").
However, if we actually have *both* a file called "File" and one called
"file", and they don't have the same lstat() information (ie we're on a
case-sensitive filesystem but have the "core.ignorecase" flag set), we
will error out if we try to add them both.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This simplifies the matching case of "I already have this file and it is
up-to-date" and makes it do the right thing in the face of
case-insensitive aliases.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It's really totally separate functionality, and if we want to start
doing case-insensitive hash lookups, I'd rather do it when it's
separated out.
It also renames "remove_index_entry()" to "remove_name_hash()", because
that really describes the thing better. It doesn't actually remove the
index entry, that's done by "remove_index_entry_at()", which is something
very different, despite the similarity in names.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This is in an effort to make the source index of 'unpack_trees()' as
being const, and thus making the compiler help us verify that we only
access it for reading.
The constification also extended to some of the hashing helpers that get
called indirectly.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This new helper is identical to base_name_compare(), except it compares
conflicting directory/file entries as equal in order to help handling DF
conflicts (thus the name).
Note that while a directory name compares as equal to a regular file
with the new helper, they then individually compare _differently_ to a
filename that has a dot after the basename (because '\0' < '.' < '/').
So a directory called "foo/" will compare equal to a file "foo", even
though "foo.c" will compare after "foo" and before "foo/"
This will be used by routines that want to traverse the git namespace
but then handle conflicting entries together when possible.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* db/checkout: (21 commits)
checkout: error out when index is unmerged even with -m
checkout: show progress when checkout takes long time while switching branches
Add merge-subtree back
checkout: updates to tracking report
builtin-checkout.c: Remove unused prefix arguments in switch_branches path
checkout: work from a subdirectory
checkout: tone down the "forked status" diagnostic messages
Clean up reporting differences on branch switch
builtin-checkout.c: fix possible usage segfault
checkout: notice when the switched branch is behind or forked
Build in checkout
Move code to clean up after a branch change to branch.c
Library function to check for unmerged index entries
Use diff -u instead of diff in t7201
Move create_branch into a library file
Build-in merge-recursive
Add "skip_unmerged" option to unpack_trees.
Discard "deleted" cache entries after using them to update the working tree
Send unpack-trees debugging output to stderr
Add flag to make unpack_trees() not print errors.
...
Conflicts:
Makefile
This makes the name hash removal function (which really just sets the
bit that disables lookups of it) available to external routines, and
makes read_cache_unmerged() use it when it drops an unmerged entry from
the index.
It's renamed to remove_index_entry(), and we drop the (unused) 'istate'
argument.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We handled the case of removing and re-inserting cache entries badly,
which is something that merging commonly needs to do (removing the
different stages, and then re-inserting one of them as the merged
state).
We even had a rather ugly special case for this failure case, where
replace_index_entry() basically turned itself into a no-op if the new
and the old entries were the same, exactly because the hash routines
didn't handle it on their own.
So what this patch does is to not just have the UNHASHED bit, but a
HASHED bit too, and when you insert an entry into the name hash, that
involves:
- clear the UNHASHED bit, because now it's valid again for lookup
(which is really all that UNHASHED meant)
- if we're being lazy, we're done here (but we still want to clear the
UNHASHED bit regardless of lazy mode, since we can become unlazy
later, and so we need the UNHASHED bit to always be set correctly,
even if we never actually insert the entry into the hash list)
- if it was already hashed, we just leave it on the list
- otherwise mark it HASHED and insert it into the list
this all means that unhashing and rehashing a name all just works
automatically. Obviously, you cannot change the name of an entry (that
would be a serious bug), but nothing can validly do that anyway (you'd
have to allocate a new struct cache_entry anyway since the name length
could change), so that's not a new limitation.
The code actually gets simpler in many ways, although the lazy hashing
does mean that there are a few odd cases (ie something can be marked
unhashed even though it was never on the hash in the first place, and
isn't actually marked hashed!).
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This creates a hash index of every single file added to the index.
Right now that hash index isn't actually used for much: I implemented a
"cache_name_exists()" function that uses it to efficiently look up a
filename in the index without having to do the O(logn) binary search,
but quite frankly, that's not why this patch is interesting.
No, the whole and only reason to create the hash of the filenames in the
index is that by modifying the hash function, you can fairly easily do
things like making it always hash equivalent names into the same bucket.
That, in turn, means that suddenly questions like "does this name exist
in the index under an _equivalent_ name?" becomes much much cheaper.
Guiding principles behind this patch:
- it shouldn't be too costly. In fact, my primary goal here was to
actually speed up "git commit" with a fully populated kernel tree, by
being faster at checking whether a file already existed in the index. I
did succeed, but only barely:
Best before:
[torvalds@woody linux]$ time git commit > /dev/null
real 0m0.255s
user 0m0.168s
sys 0m0.088s
Best after:
[torvalds@woody linux]$ time ~/git/git commit > /dev/null
real 0m0.233s
user 0m0.144s
sys 0m0.088s
so some things are actually faster (~8%).
Caveat: that's really the best case. Other things are invariably going
to be slightly slower, since we populate that index cache, and quite
frankly, few things really use it to look things up.
That said, the cost is really quite small. The worst case is probably
doing a "git ls-files", which will do very little except puopulate the
index, and never actually looks anything up in it, just lists it.
Before:
[torvalds@woody linux]$ time git ls-files > /dev/null
real 0m0.016s
user 0m0.016s
sys 0m0.000s
After:
[torvalds@woody linux]$ time ~/git/git ls-files > /dev/null
real 0m0.021s
user 0m0.012s
sys 0m0.008s
and while the thing has really gotten relatively much slower, we're
still talking about something almost unmeasurable (eg 5ms). And that
really should be pretty much the worst case.
So we lose 5ms on one "benchmark", but win 22ms on another. Pick your
poison - this patch has the advantage that it will _likely_ speed up
the cases that are complex and expensive more than it slows down the
cases that are already so fast that nobody cares. But if you look at
relative speedups/slowdowns, it doesn't look so good.
- It should be simple and clean
The code may be a bit subtle (the reasons I do hash removal the way I
do etc), but it re-uses the existing hash.c files, so it really is
fairly small and straightforward apart from a few odd details.
Now, this patch on its own doesn't really do much, but I think it's worth
looking at, if only because if done correctly, the name hashing really can
make an improvement to the whole issue of "do we have a filename that
looks like this in the index already". And at least it gets real testing
by being used even by default (ie there is a real use-case for it even
without any insane filesystems).
NOTE NOTE NOTE! The current hash is a joke. I'm ashamed of it, I'm just
not ashamed of it enough to really care. I took all the numbers out of my
nether regions - I'm sure it's good enough that it works in practice, but
the whole point was that you can make a really much fancier hash that
hashes characters not directly, but by their upper-case value or something
like that, and thus you get a case-insensitive hash, while still keeping
the name and the index itself totally case sensitive.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This moves a common boolean expression into a helper function,
and makes the comparison between filesystem timestamp and index
timestamp done in the function in line with the other places.
st.st_mtime should be casted to (unsigned int) when compared to
an index timestamp ce_mtime.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It is a D/F conflict if you want to add "foo/bar" to the index
when "foo" already exists. Also it is a conflict if you want to
add a file "foo" when "foo/bar" exists.
An exception is when the existing entry is there only to mark "I
used to be here but I am being removed". This is needed for
operations such as "git read-tree -m -u" that update the index
and then reflect the result to the work tree --- we need to
remember what to remove somewhere, and we use the index for
that. In such a case, an existing file "foo" is being removed
and we can create "foo/" directory and hang "bar" underneath it
without any conflict.
We used to use (ce->ce_mode == 0) to mark an entry that is being
removed, but (CE_REMOVE & ce->ce_flags) is used for that purpose
these days. An earlier commit forgot to convert the logic in
the code that checks D/F conflict condition.
The old code knew that "to be removed" entries cannot be at
higher stage and actively checked that condition, but it was an
unnecessary check. This patch removes the extra check as well.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Aside from the lstat(2) done for work tree files, there are
quite many lstat(2) calls in refname dwimming codepath. This
patch is not about reducing them.
* It adds a new ce_flag, CE_UPTODATE, that is meant to mark the
cache entries that record a regular file blob that is up to
date in the work tree. If somebody later walks the index and
wants to see if the work tree has changes, they do not have
to be checked with lstat(2) again.
* fill_stat_cache_info() marks the cache entry it just added
with CE_UPTODATE. This has the effect of marking the paths
we write out of the index and lstat(2) immediately as "no
need to lstat -- we know it is up-to-date", from quite a lot
fo callers:
- git-apply --index
- git-update-index
- git-checkout-index
- git-add (uses add_file_to_index())
- git-commit (ditto)
- git-mv (ditto)
* refresh_cache_ent() also marks the cache entry that are clean
with CE_UPTODATE.
* write_index is changed not to write CE_UPTODATE out to the
index file, because CE_UPTODATE is meant to be transient only
in core. For the same reason, CE_UPDATE is not written to
prevent an accident from happening.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We currently use lower 12-bit (masked with CE_NAMEMASK) in the
ce_flags field to store the length of the name in cache_entry,
without checking the length parameter given to
create_ce_flags(). This can make us store incorrect length.
Currently we are mostly protected by the fact that many
codepaths first copy the path in a variable of size PATH_MAX,
which typically is 4096 that happens to match the limit, but
that feels like a bug waiting to happen. Besides, that would
not allow us to shorten the width of CE_NAMEMASK to use the bits
for new flags.
This redefines the meaning of the name length stored in the
cache_entry. A name that does not fit is represented by storing
CE_NAMEMASK in the field, and the actual length needs to be
computed by actually counting the bytes in the name[] field.
This way, only the unusually long paths need to suffer.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This converts the index explicitly on read and write to its on-disk
format, allowing the in-core format to contain more flags, and be
simpler.
In particular, the in-core format is now host-endian (as opposed to the
on-disk one that is network endian in order to be able to be shared
across machines) and as a result we can dispense with all the
htonl/ntohl on accesses to the cache_entry fields.
This will make it easier to make use of various temporary flags that do
not exist in the on-disk format.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* jc/maint-add-sync-stat:
t2200: test more cases of "add -u"
git-add: make the entry stat-clean after re-adding the same contents
ce_match_stat, run_diff_files: use symbolic constants for readability
Conflicts:
builtin-add.c
Earlier in commit 0781b8a9b2
(add_file_to_index: skip rehashing if the cached stat already
matches), add_file_to_index() were taught not to re-add the path
if it already matches the index.
The change meant well, but was not executed quite right. It
used ie_modified() to see if the file on the work tree is really
different from the index, and skipped adding the contents if the
function says "not modified".
This was wrong. There are three possible comparison results
between the index and the file in the work tree:
- with lstat(2) we _know_ they are different. E.g. if the
length or the owner in the cached stat information is
different from the length we just obtained from lstat(2), we
can tell the file is modified without looking at the actual
contents.
- with lstat(2) we _know_ they are the same. The same length,
the same owner, the same everything (but this has a twist, as
described below).
- we cannot tell from lstat(2) information alone and need to go
to the filesystem to actually compare.
The last case arises from what we call 'racy git' situation,
that can be caused with this sequence:
$ echo hello >file
$ git add file
$ echo aeiou >file ;# the same length
If the second "echo" is done within the same filesystem
timestamp granularity as the first "echo", then the timestamp
recorded by "git add" and the timestamp we get from lstat(2)
will be the same, and we can mistakenly say the file is not
modified. The path is called 'racily clean'. We need to
reliably detect racily clean paths are in fact modified.
To solve this problem, when we write out the index, we mark the
index entry that has the same timestamp as the index file itself
(that is the time from the point of view of the filesystem) to
tell any later code that does the lstat(2) comparison not to
trust the cached stat info, and ie_modified() then actually goes
to the filesystem to compare the contents for such a path.
That's all good, but it should not be used for this "git add"
optimization, as the goal of "git add" is to actually update the
path in the index and make it stat-clean. With the false
optimization, we did _not_ cause any data loss (after all, what
we failed to do was only to update the cached stat information),
but it made the following sequence leave the file stat dirty:
$ echo hello >file
$ git add file
$ echo hello >file ;# the same contents
$ git add file
The solution is not to use ie_modified() which goes to the
filesystem to see if it is really clean, but instead use
ie_match_stat() with "assume racily clean paths are dirty"
option, to force re-adding of such a path.
There was another problem with "git add -u". The codepath
shares the same issue when adding the paths that are found to be
modified, but in addition, it asked "git diff-files" machinery
run_diff_files() function (which is "git diff-files") to list
the paths that are modified. But "git diff-files" machinery
uses the same ie_modified() call so that it does not report
racily clean _and_ actually clean paths as modified, which is
not what we want.
The patch allows the callers of run_diff_files() to pass the
same "assume racily clean paths are dirty" option, and makes
"git-add -u" codepath to use that option, to discover and re-add
racily clean _and_ actually clean paths.
We could further optimize on top of this patch to differentiate
the case where the path really needs re-adding (i.e. the content
of the racily clean entry was indeed different) and the case
where only the cached stat information needs to be refreshed
(i.e. the racily clean entry was actually clean), but I do not
think it is worth it.
This patch applies to maint and all the way up.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
ce_match_stat() can be told:
(1) to ignore CE_VALID bit (used under "assume unchanged" mode)
and perform the stat comparison anyway;
(2) not to perform the contents comparison for racily clean
entries and report mismatch of cached stat information;
using its "option" parameter. Give them symbolic constants.
Similarly, run_diff_files() can be told not to report anything
on removed paths. Also give it a symbolic constant for that.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
