When "git merge-recursive" works on history with many criss-cross
merges in "verbose" mode, the names the command assigns to the
virtual merge bases could have overwritten each other by unintended
reuse of the same piece of memory.
* rs/pull-signed-tag:
commit: use FLEX_ARRAY in struct merge_remote_desc
merge-recursive: fix verbose output for multiple base trees
commit: factor out set_merge_remote_desc()
commit: use xstrdup() in get_merge_parent()
One of the indirect callers of make_virtual_commit() passes the result of
oid_to_hex() as the name, i.e. a pointer to a static buffer. Since the
function uses that string pointer directly in building a struct
merge_remote_desc, multiple entries can end up sharing the same name
inadvertently.
Fix that by calling set_merge_remote_desc(), which creates a copy of the
string, instead of building the struct by hand.
Signed-off-by: Rene Scharfe <l.s.r@web.de>
Reviewed-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Initialize a string_list right when it's defined. That's shorter, saves
a function call and makes it more obvious that we're using the NODUP
variant here.
Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Ever since 66a155b (Enable output buffering in merge-recursive.,
2007-01-14), we had a problem: When the merge failed in a fatal way, all
regular output was swallowed because we called die() and did not get a
chance to drain the output buffers.
To fix this, several modifications were necessary:
- we needed to stop die()ing, to give callers a chance to do something
when an error occurred (in this case, flush the output buffers),
- we needed to delay printing the error message so that the caller can
print the buffered output before that, and
- we needed to make sure that the output buffers are flushed even when
the return value indicates an error.
The first two changes were introduced through earlier commits in this
patch series, and this commit addresses the third one.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The recursive merge machinery accumulates its output in an output
buffer, to be flushed at the end of merge_recursive(). At this point,
we forgot to release the output buffer.
When calling merge_trees() (i.e. the non-recursive part of the recursive
merge) directly, the output buffer is never flushed because the caller
may be merge_recursive() which wants to flush the output itself.
For the same reason, merge_trees() cannot release the output buffer: it
may still be needed.
Forgetting to release the output buffer did not matter much when running
git-checkout, or git-merge-recursive, because we exited after the
operation anyway. Ever since cherry-pick learned to pick a commit range,
however, this memory leak had the potential of becoming a problem.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since 66a155b (Enable output buffering in merge-recursive., 2007-01-14),
we already accumulate the output in a buffer. The idea was to avoid
interfering with the progress output that goes to stderr, which is
unbuffered, when we write to stdout, which is buffered.
We extend that buffering to allow the caller to handle the output
(possibly suppressing it). This will help us when extending the
sequencer to do rebase -i's brunt work: it does not want the picks to
print anything by default but instead determine itself whether to print
the output or not.
Note that we also redirect the error messages into the output buffer
when the caller asked not to flush the output buffer, for two reasons:
1) to retain the correct output order, and 2) to allow the caller to
suppress *all* output.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 66a155b (Enable output buffering in merge-recursive., 2007-01-14), we
changed the code such that it prints the output in one go, to avoid
interfering with the progress output.
Let's make sure that the same holds true when outputting the commit
title: previously, we used several printf() statements to stdout and
assumed that stdout's buffer is large enough to hold the entire
commit title.
Apart from making that speculation unnecessary, we change the code to
add the message to the output buffer before flushing for another reason:
the next commit will introduce a new level of output buffering, where
the caller can request the output not to be flushed, but to be retained
for further processing.
This latter feature will be needed when teaching the sequencer to do
rebase -i's brunt work: it wants to control the output of the
cherry-picks (i.e. recursive merges).
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The data structure passed to the recursive merge machinery has a feature
where the caller can ask for the output to be buffered into a strbuf, by
setting the field 'buffer_output'.
Previously, we died without flushing, losing accumulated output. With
this patch, we show the output first, and only then print the error
message.
Currently, the only user of that buffering is merge_recursive() itself,
to avoid the progress output to interfere.
In the next patches, we will introduce a new buffer_output mode that
forces merge_recursive() to retain the output buffer for further
processing by the caller. If the caller asked for that, we will then
also write the error messages into the output buffer. This is necessary
to give the caller more control not only how to react in case of errors
but also control how/if to display the error messages.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The recursive merge machinery is supposed to be a library function, i.e.
it should return an error when it fails. Originally the functions were
part of the builtin "merge-recursive", though, where it was simpler to
call die() and be done with error handling.
The existing callers were already prepared to detect negative return
values to indicate errors and to behave as previously: exit with code 128
(which is the same thing that die() does, after printing the message).
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We are about to libify the recursive merge machinery, where we only
die() in case of a bug or memory contention. To that end, we must heed
negative return values as indicating errors.
This requires our functions to be careful to pass through error
conditions in call chains, and for quite a few functions this means
that they have to return values to begin with.
The next step will be to convert the places where we currently die() to
return negative values (read: -1) instead.
Note that we ignore errors reported by make_room_for_path(), consistent
with the previous behavior (update_file_flags() used the return value of
make_room_for_path() only to indicate an early return, but not a fatal
error): if the error is really a fatal error, we will notice later; If
not, it was not that serious a problem to begin with. (Witnesses in
favor of this reasoning are t4151-am-abort and t7610-mergetool, which
would start failing if we stopped on errors reported by
make_room_for_path()).
Also note: while this patch makes the code slightly less readable in
update_file_flags() (we introduce a new "goto free_buf;" instead of
an explicit "free(buf); return;"), it is a preparatory change for
the next patch where we will convert all of the die() calls in the same
function to go through the free_buf return path instead.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It is possible that a tree cannot be written (think: disk full). We
will want to give the caller a chance to clean up instead of letting
the program die() in such a case.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It is technically allowed, as per C89, for functions' return type to
be complete structs (i.e. *not* just pointers to structs).
However, it was just an oversight of this developer when converting
Python code to C code in 6d297f8 (Status update on merge-recursive in
C, 2006-07-08) which introduced such a return type.
Besides, by converting this construct to pass in the struct, we can now
start returning a value that can indicate errors in future patches. This
will help the current effort to libify merge-recursive.c.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There are a couple of places where return values never indicated errors
before, as we simply died instead of returning.
But now negative return values mean that there was an error and we have to
abort the operation. Let's do exactly that.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It can be puzzling to see that was_tracked() asks to get an index entry
by name, but does not take a negative return value for an answer.
The reason we have to do this is that cache_name_pos() only looks for
entries in stage 0, even if nobody asked for any stage in particular.
Let's rewrite the logic a little bit, to handle the easy case early: if
cache_name_pos() returned a non-negative position, we know it is a match,
and we do not even have to compare the name again (cache_name_pos() did
that for us already). We can say right away: yes, this file was tracked.
Only if there was no exact match do we need to look harder for any
matching entry in stage 2.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
While working on the patch series that avoids die()ing in recursive
merges, the issue came up that bug reports (i.e. die("BUG: ...")
constructs) should never be translated, as the target audience is the
Git developer community, not necessarily the current user, and hence
a translated message would make it *harder* to address the problem.
So let's stop translating the obvious ones. As it is really, really
outside the purview of this patch series to see whether there are more
die() statements that report bugs and are currently translated, that
task is left for another day and patch.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The vast majority of error messages in Git's source code which report a
bug use the convention to prefix the message with "BUG:".
As part of cleaning up merge-recursive to stop die()ing except in case of
detected bugs, let's just make the remainder of the bug reports consistent
with the de facto rule.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git merge" with renormalization did not work well with
merge-recursive, due to "safer crlf" conversion kicking in when it
shouldn't.
* jc/renormalize-merge-kill-safer-crlf:
merge: avoid "safer crlf" during recording of merge results
convert: unify the "auto" handling of CRLF
When merge_recursive() decides what the correct blob object merge
result for a path should be, it uses update_file_flags() helper
function to write it out to a working tree file and then calls
add_cacheinfo(). The add_cacheinfo() function in turn calls
make_cache_entry() to create a new cache entry to replace the
higher-stage entries for the path that represents the conflict.
The make_cache_entry() function calls refresh_cache_entry() to fill
in the cached stat information. To mark a cache entry as
up-to-date, the data is re-read from the file in the working tree,
and goes through convert_to_git() conversion to be compared with the
blob object name the new cache entry records.
It is important to note that this happens while the higher-stage
entries, which are going to be replaced with the new entry, are
still in the index. Unfortunately, the convert_to_git() conversion
has a misguided "safer crlf" mechanism baked in, and looks at the
existing cache entry for the path to decide how to convert the
contents in the working tree file. If our side (i.e. stage#2)
records a text blob with CRLF in it, even when the system is
configured to record LF in blobs and convert them to CRLF upon
checkout (and back to LF upon checkin), the "safer crlf" mechanism
stops us doing so.
This especially poses a problem during a renormalizing merge, where
the merge result for the path is computed by first "normalizing" the
blobs involved in the merge by using convert_to_working_tree()
followed by convert_to_git() with "safer crlf" disabled. The merge
result that is computed correctly and fed to add_cacheinfo() via
update_file_flags() does _not_ match what refresh_cache_entry() sees
by converting the working tree file via convert_to_git().
We can work this around by not refreshing the new cache entry in
make_cache_entry() called by add_cacheinfo(). After add_cacheinfo()
adds the new entry, we can call refresh_cache_entry() on that,
knowing that addition of this new cache entry would have removed the
stale cache entries that had CRLF in stage #2 that were carried over
before the renormalizing merge started and will not interfere with
the correct recording of the result.
The test update was taken from a series by Torsten Bögershausen
that attempted to fix this with a different approach.
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Reviewed-by: Torsten Bögershausen <tboegi@web.de>
Convert this function and the git merge-recursive subcommand to use
struct object_id.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert all but two of the static functions in this file to use struct
object_id.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert struct merge_file_info to use struct object_id. The following
Coccinelle semantic patch was used to implement this, followed by the
transformations in object_id.cocci:
@@
struct merge_file_info o;
@@
- o.sha
+ o.oid.hash
@@
struct merge_file_info *p;
@@
- p->sha
+ p->oid.hash
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert the anonymous struct within struct stage_data to use struct
object_id. The following Coccinelle semantic patch was used to
implement this, followed by the transformations in object_id.cocci:
@@
struct stage_data o;
expression E1;
@@
- o.stages[E1].sha
+ o.stages[E1].oid.hash
@@
struct stage_data *p;
expression E1;
@@
- p->stages[E1].sha
+ p->stages[E1].oid.hash
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert struct diff_filespec's sha1 member to use a struct object_id
called "oid" instead. The following Coccinelle semantic patch was used
to implement this, followed by the transformations in object_id.cocci:
@@
struct diff_filespec o;
@@
- o.sha1
+ o.oid.hash
@@
struct diff_filespec *p;
@@
- p->sha1
+ p->oid.hash
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move from unsigned char[20] to struct object_id continues.
* bc/object-id:
match-trees: convert several leaf functions to use struct object_id
tree-walk: convert tree_entry_extract() to use struct object_id
struct name_entry: use struct object_id instead of unsigned char sha1[20]
match-trees: convert shift_tree() and shift_tree_by() to use object_id
test-match-trees: convert to use struct object_id
sha1-name: introduce a get_oid() function
There were a few cases in merge-recursive that could result in a
check for the presence of files in the working copy while trying to
create a virtual merge base. These were rare and innocuous, but
somewhat illogical. The two cases were:
* When there was naming conflicts (e.g. a D/F conflict) and we had to
pick a new unique name for a file. Since the new name is somewhat
arbitrary, it didn't matter that we consulted the working copy to
avoid picking a filename it has, but since the virtual merge base is
never checked out, it's a waste of time and slightly odd to do so.
* When two different files get renamed to the same name (on opposite
sides of the merge), we needed to delete the original filenames from
the cache and possibly also the working directory. The caller's check
for determining whether to delete from the working directory was a
call to would_lose_untracked(). It turns out this didn't matter
because remove_file() had logic to avoid modifying the working
directory when creating a virtual merge base, but there is no reason
for the caller to check the working directory in such circumstances.
It's a waste of time, if not also a bit weird.
Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In commit 51931bf (merge-recursive: Improve handling of rename
target vs. directory addition, 2011-08-11), I apparently added two
lines of code that were immediately duplicated a few lines later.
No idea why, other than it seems pretty clear this was a mistake:
there is no need to remove the same file twice; removing it once is
sufficient...especially since the intervening line was working with
a different file entirely.
Signed-off-by: Elijah Newren <newren@gmail.com>
Reviewed-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Make the find-renames option follow the behaviour in git-diff, where it
resets the threshold when none is given. So, for instance,
"--find-renames=25 --find-renames" should result in the default
threshold (50%) instead of 25%.
Signed-off-by: Felipe Gonçalves Assis <felipegassis@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add strategy option find-renames, following git-diff interface. This
makes the option rename-threshold redundant.
Signed-off-by: Felipe Gonçalves Assis <felipegassis@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The recursive strategy turns on rename detection by default. Add a
strategy option to disable rename detection even for exact renames.
Signed-off-by: Felipe Gonçalves Assis <felipegassis@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert all instances of get_object_hash to use an appropriate reference
to the hash member of the oid member of struct object. This provides no
functional change, as it is essentially a macro substitution.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Jeff King <peff@peff.net>
struct object is one of the major data structures dealing with object
IDs. Convert it to use struct object_id instead of an unsigned char
array. Convert get_object_hash to refer to the new member as well.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Jeff King <peff@peff.net>
Convert most instances where the sha1 member of struct object is
dereferenced to use get_object_hash. Most instances that are passed to
functions that have versions taking struct object_id, such as
get_sha1_hex/get_oid_hex, or instances that can be trivially converted
to use struct object_id instead, are not converted.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Jeff King <peff@peff.net>
Merging a branch that removes a path and another that changes the
mode bits on the same path should have conflicted at the path, but
it didn't and silently favoured the removal.
* jk/delete-modechange-conflict:
merge: detect delete/modechange conflict
t6031: generalize for recursive and resolve strategies
t6031: move triple-rename test to t3030
If one side deletes a file and the other changes its
content, we notice and report a conflict. However, if
instead of changing the content, we change only the mode,
the merge does not notice (and the mode change is silently
dropped).
The trivial index merge notices the problem and correctly
leaves the conflict in the index, but both merge-recursive
and merge-one-file will silently resolve this in favor of
the deletion. In many cases that is a sane resolution, but
we should be punting to the user whenever there is any
question. So let's detect and treat this as a conflict (in
both strategies).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This would be a fairly routine use of xstrfmt, except that
we need to remember the length of the result to pass to
cache_name_pos. So just use a strbuf, which makes this
simple.
As a bonus, this gets rid of confusing references to
"pathlen+1". The "1" is for the trailing slash we added, but
that is automatically accounted for in the strbuf's len
parameter.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Call file_exists() instead of open-coding it. That's shorter, simpler
and the intent becomes clearer.
Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
These string_list instances were allocated by get_renames() and
get_unmerged for the sole use of this caller, and the function is
responsible for freeing them, not just their contents.
Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The get_merge_bases*() API was easy to misuse by careless
copy&paste coders, leaving object flags tainted in the commits that
needed to be traversed.
* jc/merge-bases:
get_merge_bases(): always clean-up object flags
bisect: clean flags after checking merge bases
This allows the callback to use 'base' as a temporary buffer to
quickly assemble full path "without" extra allocation. The callback
has to restore it afterwards of course.
Helped-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The callers of get_merge_bases() can choose to leave object flags
used during the merge-base traversal by passing cleanup=0 as a
parameter, but in practice a very few callers can afford to do so
(namely, "git merge-base"), as they need to compute merge base in
preparation for other processing of their own and they need to see
the object without contaminate flags.
Change the function signature of get_merge_bases_many() and
get_merge_bases() to drop the cleanup parameter, so that the
majority of the callers do not have to say ", 1" at the end.
Give a new get_merge_bases_many_dirty() API to support only a few
callers that know they do not need to spend cycles cleaning up the
object flags.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The lockfile API and its users have been cleaned up.
* mh/lockfile: (38 commits)
lockfile.h: extract new header file for the functions in lockfile.c
hold_locked_index(): move from lockfile.c to read-cache.c
hold_lock_file_for_append(): restore errno before returning
get_locked_file_path(): new function
lockfile.c: rename static functions
lockfile: rename LOCK_NODEREF to LOCK_NO_DEREF
commit_lock_file_to(): refactor a helper out of commit_lock_file()
trim_last_path_component(): replace last_path_elm()
resolve_symlink(): take a strbuf parameter
resolve_symlink(): use a strbuf for internal scratch space
lockfile: change lock_file::filename into a strbuf
commit_lock_file(): use a strbuf to manage temporary space
try_merge_strategy(): use a statically-allocated lock_file object
try_merge_strategy(): remove redundant lock_file allocation
struct lock_file: declare some fields volatile
lockfile: avoid transitory invalid states
git_config_set_multivar_in_file(): avoid call to rollback_lock_file()
dump_marks(): remove a redundant call to rollback_lock_file()
api-lockfile: document edge cases
commit_lock_file(): rollback lock file on failure to rename
...
Move the interface declaration for the functions in lockfile.c from
cache.h to a new file, lockfile.h. Add #includes where necessary (and
remove some redundant includes of cache.h by files that already
include builtin.h).
Move the documentation of the lock_file state diagram from lockfile.c
to the new header file.
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git merge-recursive" had a small bug that could have made it
mishandle "one side deleted, the other side did not touch it" in a
rare corner case, where the other side actually did touch to cause
the blob object names to be different but both blobs before and
after the change normalize to the same (e.g. correcting mistake to
check in a blob with CRLF line endings by replacing it with another
blob that records the same contents with LF line endings).
* sb/merge-recursive-copy-paste-fix:
merge-recursive: remove stale commented debugging code
merge-recursive: fix copy-paste mistake
The following issue was found by scan.coverity.com (ID: 1049510),
and claimed to be likely a copy-paste mistake.
Introduced in 331a1838b (2010-07-02, Try normalizing files
to avoid delete/modify conflicts when merging), which is
quite a long time ago, so I'm rather unsure if it's of any impact
or just went unnoticed.
The line after the changed line has a comparison of 'o.len' to 'a.len',
so we should assume the lengths may be different.
I'd be happy to have a test for this bug(?) attached to
t6031-merge-recursive.sh, but I did not manage to
come up with a test in a reasonable amount of time.
Signed-off-by: Stefan Beller <stefanbeller@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
CodingGuidelines states that the first #include in C files should be
git-compat-util.h or another header file that includes it, such as
cache.h or builtin.h.
Signed-off-by: David Aguilar <davvid@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
