Add some tests to demonstrate the current behavior around adding files
outside of the sparse-checkout cone. Currently, untracked files are
handled differently from tracked files. A future change will make these
cases be handled the same way.
Further expand checking that a failed 'git add' does not stage changes
to the index.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit b681b191 introduced the support of systemd timers for git
maintenance.
A test is leveraging the `systemd-analyze verify` utility to verify the
correctness of the systemd unit files generated by git.
But on some systems, although the `systemd-analyze` tool is installed
and supports the `verify` subcommand, it fails with some permission
errors.
So, instead of only checking if the `verify` subcommand exists, a more
reliable way of detecting whether `systemd-analyze verify` can be used
is to try to use it.
The SYSTEMD_ANALYZE prerequisite is now trying to run `systemd-analyze
verify` on a systemd unit file which is shipped by systemd itself.
We can reasonably think that, on systemd hosts, this file is present and
valid.
Signed-off-by: Lénaïc Huard <lenaic@lhuard.fr> Signed-off-by: Junio C Hamano <gitster@pobox.com>
builtin/blame.c: refactor commit_info_init() to COMMIT_INFO_INIT macro
Remove the commit_info_init() function addded in ea02ffa3857 (mailmap:
simplify map_user() interface, 2013-01-05) and instead initialize the
"struct commit_info" with a macro.
This is the more idiomatic pattern in the codebase, and doesn't leave
us wondering when we see the *_init() function if this struct needs
more complex initialization than a macro can provide.
The get_commit_info() function is only called by the three callers
being changed here immediately after initializing the struct with the
macros, so by moving the initialization to the callers we don't need
to do it in get_commit_info() anymore.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
daemon.c: refactor hostinfo_init() to HOSTINFO_INIT macro
Remove the hostinfo_init() function added in 01cec54e135 (daemon:
deglobalize hostname information, 2015-03-07) and instead initialize
the "struct hostinfo" with a macro.
This is the more idiomatic pattern in the codebase, and doesn't leave
us wondering when we see the *_init() function if this struct needs
more complex initialization than a macro can provide.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
cbtree.h: define cb_init() in terms of CBTREE_INIT
Use the same pattern for cb_init() as the one established in the
recent refactoring of other such patterns in 5726a6b4012 (*.c *_init(): define in terms of corresponding *_INIT
macro, 2021-07-01).
It has been pointed out[1] that we could perhaps use this C99
replacement of using a compound literal for all of these:
*t = (struct cb_tree){ 0 };
But let's just stick to the existing pattern established in 5726a6b4012 for now, we can leave another weather balloon for some
other time.
Move various *_INIT macros to use designated initializers. This helps
readability. I've only picked those leftover macros that were not
touched by another in-flight series of mine which changed others, but
also how initialization was done.
In the case of SUBMODULE_ALTERNATE_SETUP_INIT I've left an explicit
initialization of "error_mode", even though
SUBMODULE_ALTERNATE_ERROR_IGNORE itself is defined as "0". Let's not
peek under the hood and assume that enum fields we know the value of
will stay at "0".
The change to "TESTSUITE_INIT" in "t/helper/test-run-command.c" was
part of an earlier on-list version[1] of c90be786da9 (test-tool
run-command: fix flip-flop init pattern, 2021-09-11).
Change the initialization of "struct strbuf" changed in cbc0f81d96f (strbuf: use designated initializers in STRBUF_INIT,
2017-07-10) to omit specifying "alloc" and "len", as we do with other
"alloc" and "len" (or "nr") in similar structs.
Let's likewise omit the explicit initialization of all fields in the
"struct ipc_client_connect_option" struct added in 59c7b88198a (simple-ipc: add win32 implementation, 2021-03-15).
Do the same for a few other initializers, e.g. STRVEC_INIT and
CACHE_DEF_INIT.
Finally, start incrementally changing the same pattern in
"t/helper/test-run-command.c". This change was part of an earlier
on-list version[1] of c90be786da9 (test-tool run-command: fix
flip-flop init pattern, 2021-09-11).
*.[ch] *_INIT macros: use { 0 } for a "zero out" idiom
In C it isn't required to specify that all members of a struct are
zero'd out to 0, NULL or '\0', just providing a "{ 0 }" will
accomplish that.
Let's also change code that provided N zero'd fields to just
provide one, and change e.g. "{ NULL }" to "{ 0 }" for
consistency. I.e. even if the first member is a pointer let's use "0"
instead of "NULL". The point of using "0" consistently is to pick one,
and to not have the reader wonder why we're not using the same pattern
everywhere.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
This macro was added and used in c68f8375760 (implement fetching of
moved submodules, 2017-10-16) but its last user went away in be76c212823 (fetch: ensure submodule objects fetched, 2018-12-06).
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Makefile: pass -Wno-pendantic under GENERATE_COMPILATION_DATABASE=yes
The same bug fixed in the "COMPUTE_HEADER_DEPENDENCIES=auto" mode in
the preceding commit was also present with
"GENERATE_COMPILATION_DATABASE=yes". Let's fix it so it works again
with "DEVOPTS=1".
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Comment important codepaths regarding nuking untracked files/dirs
In the last few commits we focused on code in unpack-trees.c that
mistakenly removed untracked files or directories. There may be more of
those, but in this commit we change our focus: callers of toplevel
commands that are expected to remove untracked files or directories.
As noted previously, we have toplevel commands that are expected to
delete untracked files such as 'read-tree --reset', 'reset --hard', and
'checkout --force'. However, that does not mean that other highlevel
commands that happen to call these other commands thought about or
conveyed to users the possibility that untracked files could be removed.
Audit the code for such callsites, and add comments near existing
callsites to mention whether these are safe or not.
My auditing is somewhat incomplete, though; it skipped several cases:
* git-rebase--preserve-merges.sh: is in the process of being
deprecated/removed, so I won't leave a note that there are
likely more bugs in that script.
* contrib/git-new-workdir: why is the -f flag being used in a new
empty directory?? It shouldn't hurt, but it seems useless.
* git-p4.py: Don't see why -f is needed for a new dir (maybe it's
not and is just superfluous), but I'm not at all familiar with
the p4 stuff
* git-archimport.perl: Don't care; arch is long since dead
* git-cvs*.perl: Don't care; cvs is long since dead
Also, the reset --hard in builtin/worktree.c looks safe, due to only
running in an empty directory.
Signed-off-by: Elijah Newren <newren@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.
Remove ignored files by default when they are in the way
Change several commands to remove ignored files by default when they are
in the way. Since some commands (checkout, merge) take a
--no-overwrite-ignore option to allow the user to configure this, and it
may make sense to add that option to more commands (and in the case of
merge, actually plumb that configuration option through to more of the
backends than just the fast-forwarding special case), add little
comments about where such flags would be used.
Incidentally, this fixes a test failure in t7112.
Signed-off-by: Elijah Newren <newren@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
unpack-trees: introduce preserve_ignored to unpack_trees_options
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>
read-tree, merge-recursive: overwrite ignored files by default
This fixes a long-standing patchwork of ignored files handling in
read-tree and merge-recursive, called out and suggested by Junio long
ago. Quoting from commit dcf0c16ef1 ("core.excludesfile clean-up"
2007-11-16):
git-read-tree takes --exclude-per-directory=<gitignore>,
not because the flexibility was needed. Again, this was
because the option predates the standardization of the ignore
files.
...
On the other hand, I think it makes perfect sense to fix
git-read-tree, git-merge-recursive and git-clean to follow the
same rule as other commands. I do not think of a valid use case
to give an exclude-per-directory that is nonstandard to
read-tree command, outside a "negative" test in the t1004 test
script.
This patch is the first step to untangle this mess.
The next step would be to teach read-tree, merge-recursive and
clean (in C) to use setup_standard_excludes().
History shows each of these were partially or fully fixed:
* clean was taught the new trick in 1617adc7a0 ("Teach git clean to
use setup_standard_excludes()", 2007-11-14).
* read-tree was primarily used by checkout & merge scripts. checkout
and merge later became builtins and were both fixed to use the new
setup_standard_excludes() handling in fc001b526c ("checkout,merge:
loosen overwriting untracked file check based on info/exclude",
2011-11-27). So the primary users were fixed, though read-tree
itself was not.
* merge-recursive has now been replaced as the default merge backend
by merge-ort. merge-ort fixed this by using
setup_standard_excludes() starting early in its implementation; see
commit 6681ce5cf6 ("merge-ort: add implementation of checkout()",
2020-12-13), largely due to its design depending on checkout() and
thus being influenced by the checkout code. However,
merge-recursive itself was not fixed here, in part because its
design meant it had difficulty differentiating between untracked
files, ignored files, leftover tracked files that haven't been
removed yet due to order of processing files, and files written by
itself due to collisions).
Make the conversion more complete by now handling read-tree and
handling at least the unpack_trees() portion of merge-recursive. While
merge-recursive is on its way out, fixing the unpack_trees() portion is
easy and facilitates some of the later changes in this series. Note
that fixing read-tree makes the --exclude-per-directory option to
read-tree useless, so we remove it from the documentation (though we
continue to accept it if passed).
The read-tree changes happen to fix a bug in t1013.
Signed-off-by: Elijah Newren <newren@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
lazyload.h: use an even more generic function pointer than FARPROC
gcc will helpfully raise a -Wcast-function-type warning when casting
between functions that might have incompatible return types
(ex: GetUserNameExW returns bool which is only half the size of the
return type from FARPROC which is long long), so create a new type that
could be used as a completely generic function pointer and cast through
it instead.
Additionaly remove the -Wno-incompatible-pointer-types temporary
flag added in 27e0c3c (win32: allow building with pedantic mode
enabled, 2021-09-03), as it will be no longer needed.
Signed-off-by: Carlo Marcelo Arenas Belón <carenas@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:48:48 +0000 (14:48 -0400)]
refs: drop "broken" flag from for_each_fullref_in()
No callers pass in anything but "0" here. Likewise to our sibling
functions. Note that some of them ferry along the flag, but none of
their callers pass anything but "0" either.
Nor is anybody likely to change that. Callers which really want to see
all of the raw refs use for_each_rawref(). And anybody interested in
iterating a subset of the refs will likely be happy to use the
now-default behavior of showing broken refs, but omitting dangling
symlinks.
So we can get rid of this whole feature.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:48:20 +0000 (14:48 -0400)]
ref-filter: drop broken-ref code entirely
Now that none of our callers passes the INCLUDE_BROKEN flag, we can drop
it entirely, along with the code to plumb it through to the
for_each_fullref_in() functions.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Of the ref-filter callers, for-each-ref and git-branch both set the
INCLUDE_BROKEN flag (but git-tag does not, which is a weird
inconsistency). But now that GIT_REF_PARANOIA is on by default, that
produces almost the same outcome for all three.
The one exception is that GIT_REF_PARANOIA will omit dangling symrefs.
That's a better behavior for these tools, as they would never include
such a symref in the main output anyway (they can't, as it doesn't point
to an object). Instead they issue a warning to stderr. But that warning
is somewhat useless; a dangling symref is a perfectly reasonable thing
to have in your repository, and is not a sign of corruption. It's much
friendlier to just quietly ignore it.
And in terms of robustness, the warning gains us little. It does not
impact the exit code of either tool. So while the warning _might_ clue
in a user that they have an unexpected broken symref, it would not help
any kind of scripted use.
This patch converts for-each-ref and git-branch to stop using the
INCLUDE_BROKEN flag. That gives them more reasonable behavior, and
harmonizes them with git-tag.
We have to change one test to adapt to the situation. t1430 tries to
trigger all of the REF_ISBROKEN behaviors from the underlying ref code.
It uses for-each-ref to do so (because there isn't any other mechanism).
That will no longer issue a warning about the symref which points to an
invalid name, as it's considered dangling (and we can instead be sure
that it's _not_ mentioned on stderr). Note that we do still complain
about the illegally named "broken..symref"; its problem is not that it's
dangling, but the name of the symref itself is illegal.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:46:37 +0000 (14:46 -0400)]
repack, prune: drop GIT_REF_PARANOIA settings
Now that GIT_REF_PARANOIA is the default, we don't need to selectively
enable it for destructive operations. In fact, it's harmful to do so,
because it overrides any GIT_REF_PARANOIA=0 setting that the user may
have provided (because they're trying to work around some corruption).
With these uses gone, we can further clean up the ref_paranoia global,
and make it a static variable inside the refs code.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:46:13 +0000 (14:46 -0400)]
refs: turn on GIT_REF_PARANOIA by default
The original point of the GIT_REF_PARANOIA flag was to include broken
refs in iterations, so that possibly-destructive operations would not
silently ignore them (and would generally instead try to operate on the
oids and fail when the objects could not be accessed).
We already turned this on by default for some dangerous operations, like
"repack -ad" (where missing a reachability tip would mean dropping the
associated history). But it was not on for general use, even though it
could easily result in the spreading of corruption (e.g., imagine
cloning a repository which simply omits some of its refs because
their objects are missing; the result quietly succeeds even though you
did not clone everything!).
This patch turns on GIT_REF_PARANOIA by default. So a clone as mentioned
above would actually fail (upload-pack tells us about the broken ref,
and when we ask for the objects, pack-objects fails to deliver them).
This may be inconvenient when working with a corrupted repository, but:
- we are better off to err on the side of complaining about
corruption, and then provide mechanisms for explicitly loosening
safety.
- this is only one type of corruption anyway. If we are missing any
other objects in the history that _aren't_ ref tips, then we'd
behave similarly (happily show the ref, but then barf when we
started traversing).
We retain the GIT_REF_PARANOIA variable, but simply default it to "1"
instead of "0". That gives the user an escape hatch for loosening this
when working with a corrupt repository. It won't work across a remote
connection to upload-pack (because we can't necessarily set environment
variables on the remote), but there the client has other options (e.g.,
choosing which refs to fetch).
As a bonus, this also makes ref iteration faster in general (because we
don't have to call has_object_file() for each ref), though probably not
noticeably so in the general case. In a repo with a million refs, it
shaved a few hundred milliseconds off of upload-pack's advertisement;
that's noticeable, but most repos are not nearly that large.
The possible downside here is that any operation which iterates refs but
doesn't ever open their objects may now quietly claim to have X when the
object is corrupted (e.g., "git rev-list new-branch --not --all" will
treat a broken ref as uninteresting). But again, that's not really any
different than corruption below the ref level. We might have
refs/heads/old-branch as non-corrupt, but we are not actively checking
that we have the entire reachable history. Or the pointed-to object
could even be corrupted on-disk (but our "do we have it" check would
still succeed). In that sense, this is merely bringing ref-corruption in
line with general object corruption.
One alternative implementation would be to actually check for broken
refs, and then _immediately die_ if we see any. That would cause the
"rev-list --not --all" case above to abort immediately. But in many ways
that's the worst of all worlds:
- it still spends time looking up the objects an extra time
- it still doesn't catch corruption below the ref level
- it's even more inconvenient; with the current implementation of
GIT_REF_PARANOIA for something like upload-pack, we can make
the advertisement and let the client choose a non-broken piece of
history. If we bail as soon as we see a broken ref, they cannot even
see the advertisement.
The test changes here show some of the fallout. A non-destructive "git
repack -adk" now fails by default (but we can override it). Deleting a
broken ref now actually tells the hooks the correct "before" state,
rather than a confusing null oid.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:42:38 +0000 (14:42 -0400)]
refs: omit dangling symrefs when using GIT_REF_PARANOIA
Dangling symrefs aren't actually a corruption problem. It's perfectly
fine for refs/remotes/origin/HEAD to point to an unborn branch. And in
particular, if you are trying to establish reachability, a symref that
points nowhere doesn't matter either way. Any ref it could point to will
be examined during the rest of the traversal.
It's possible that a symref pointing nowhere _could_ be a sign that the
ref it was meant to point to was deleted accidentally (e.g., via
corruption). But there is no particular reason to think that is true for
any given case, and in the meantime, GIT_REF_PARANOIA kicking in
automatically for some operations means they'll fail unnecessarily.
So let's loosen it just a bit. The new test in t5312 shows off an
example that is safe, but currently fails (and no longer does after this
patch).
Note that we don't do anything if the caller explicitly asked for
DO_FOR_EACH_INCLUDE_BROKEN. In that case they may be looking for
dangling symrefs themselves, and setting GIT_REF_PARANOIA should not
_loosen_ things from what the caller asked for.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:41:32 +0000 (14:41 -0400)]
refs: add DO_FOR_EACH_OMIT_DANGLING_SYMREFS flag
When the DO_FOR_EACH_INCLUDE_BROKEN flag is used, we include both actual
corrupt refs (illegal names, missing objects), but also symrefs that
point to nothing. This latter is not really a corruption, but just
something that may happen normally. For example, the symref at
refs/remotes/origin/HEAD may point to a tracking branch which is later
deleted. (The local HEAD may also be unborn, of course, but we do not
access it through ref iteration).
Most callers of for_each_ref() etc, do not care. They don't pass
INCLUDE_BROKEN, so don't see it at all. But for those which do pass it,
this somewhat-normal state causes extra warnings (e.g., from
for-each-ref) or even aborts operations (destructive repacks with
GIT_REF_PARANOIA set).
This patch just introduces the flag and the mechanism; there are no
callers yet (and hence no tests). Two things to note on the
implementation:
- we actually skip any symref that does not resolve to a ref. This
includes ones which point to an invalidly-named ref. You could argue
this is a more serious breakage than simple dangling. But the
overall effect is the same (we could not follow the symref), as well
as the impact on things like REF_PARANOIA (either way, a symref we
can't follow won't impact reachability, because we'll see the ref
itself during iteration). The underlying resolution function doesn't
distinguish these two cases (they both get REF_ISBROKEN).
- we change the iterator in refs/files-backend.c where we check
INCLUDE_BROKEN. There's a matching spot in refs/packed-backend.c,
but we don't know need to do anything there. The packed backend does
not support symrefs at all.
The resulting set of flags might be a bit easier to follow if we broke
this down into "INCLUDE_CORRUPT_REFS" and "INCLUDE_DANGLING_SYMREFS".
But there are a few reasons not do so:
- adding a new OMIT_DANGLING_SYMREFS flag lets us leave existing
callers intact, without changing their behavior (and some of them
really do want to see the dangling symrefs; e.g., t5505 has a test
which expects us to report when a symref becomes dangling)
- they're not actually independent. You cannot say "include dangling
symrefs" without also including refs whose objects are not
reachable, because dangling symrefs by definition do not have an
object. We could tweak the implementation to distinguish this, but
in practice nobody wants to ask for that. Adding the OMIT flag keeps
the implementation simple and makes sure we don't regress the
current behavior.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:39:44 +0000 (14:39 -0400)]
refs-internal.h: reorganize DO_FOR_EACH_* flag documentation
The documentation for the DO_FOR_EACH_* flags is sprinkled over the
refs-internal.h file. We define the two flags in one spot, and then
describe them in more detail far away from there, in the definitions of
refs_ref_iterator_begin() and ref_iterator_advance_fn().
Let's try to organize this a bit better:
- convert the #defines to an enum. This makes it clear that they are
related, and that the enum shows the complete set of flags.
- combine all descriptions for each flag in a single spot, next to the
flag's definition
- use the enum rather than a bare int for functions which take the
flags. This helps readers realize which flags can be used.
- clarify the mention of flags for ref_iterator_advance_fn(). It does
not take flags itself, but is meant to depend on ones set up
earlier.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:37:58 +0000 (14:37 -0400)]
refs-internal.h: move DO_FOR_EACH_* flags next to each other
There are currently two DO_FOR_EACH_* flags, which must not have their
bits overlap. Yet they're defined hundreds of lines apart. Let's move
them next to each other to make it clear that they are related and are a
complete set (which matters if you are adding a new flag and would like
to know what the next available bit is).
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:37:11 +0000 (14:37 -0400)]
t5312: be more assertive about command failure
When repacking or pruning in a corrupted repository, our tests in t5312
argue that it is OK to complete the operation or bail, as long as we
don't actually delete the objects pointed to by the corruption.
This isn't a wrong line of reasoning, but the tests are a bit permissive
by using test_might_fail. The fact is that we _do_ bail currently, and
if we ever stopped doing so, that would be worthy of a human
investigating. So let's switch these to test_must_fail.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:36:42 +0000 (14:36 -0400)]
t5312: test non-destructive repack
In t5312, we create a state with a broken ref, and then make sure that
destructive repacks don't silently ignore the breakage (where a
destructive repack is one that might drop objects). But we don't check
the behavior of non-destructive repacks at all (i.e., ones where we'd
keep unreachable objects).
So let's add a test to confirm the current behavior, which is that
they are allowed (i.e., ignoring the breakage and considering any
objects it points to as unreachable). This may change in the future, but
we'd like for the test suite to alert us to that fact.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:36:18 +0000 (14:36 -0400)]
t5312: create bogus ref as necessary
Some tests in t5312 create an illegally-named ref, and then see how
various operations handle it. But between those operations, we also do
some more setup (e.g., repacking), and we are subtly depending on how
those setup steps react to the illegal ref.
To future-proof us against those behaviors changing, let's instead
create and clean up our bogus ref on demand in the tests that need it.
This has two small extra advantages:
- the tests are more stand-alone; we do not need an extra test to clean
up the ref before moving on to other parts of the script
- the creation and cleanup is together in one helper function. Because
these depend on touching the refs in the filesystem directly, they
may need to be tweaked for a world with alternate backends (they have
not been noticed so far in the reftable work because with a non-file
backend the tests don't fail; they simply become uninteresting noops
because the broken ref isn't read at all).
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:35:29 +0000 (14:35 -0400)]
t5312: drop "verbose" helper
t5312 has several uses of the "verbose" helper, as described in 8ad1652418 (t5304: use helper to report failure of "test foo = bar",
2014-10-10). Back then the "-x" trace option for tests was new, and was
not as pleasant to use (e.g., some tests failed under "-x", we did not
support BASH_XTRACEFD, etc).
These days it is clear that "-x" is the preferred way to get extra
output, and we don't need to mark up individual tests. Let's get rid of
the uses of "verbose" here, as one step toward eradicating it totally.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:34:04 +0000 (14:34 -0400)]
t5600: provide detached HEAD for corruption failures
When checking how git-clone behaves when it fails, we stimulate some
failures by trying to do a clone from a local repository whose objects
have been removed. Because these clones use local optimizations, there's
a subtle dependency in how the corruption is handled on the sending
side.
If upload-pack does not show us the broken refs (which it does not
currently), then we see only HEAD (which is itself broken), and clone
that as a detached HEAD. When we try to write the ref, we notice that we
never got the object and bail.
But if upload-pack _does_ show us the broken refs (which it may in a
future patch), then we'll realize that HEAD is a symref and just write
that. You'd think we'd fail when writing out the refs themselves, but we
don't; we do a bulk write and skip the connectivity check because of our
--local optimizations. For the non-bare case, we do notice the problem
when we try to checkout. But for a bare repository, we unexpectedly
complete the clone successfully!
At first glance this may seem like a bug. But the whole point of those
local optimizations is to give up some safety for speed. If you want to
be careful, you should be using "--no-local", which would notice that
the pack did not transfer sufficient objects. We could do that in these
tests, but part of the point is for them to fail at specific moments
(and indeed, we have a later test that checks for transport failure).
However, we can make this less subtle and future-proof it against
changes on the upload-pack side by just having an explicit detached
HEAD in the corrupted repo. Now we'll fail as expected during the ref
write if any ref _or_ HEAD is corrupt, whether we're --bare or not.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:33:00 +0000 (14:33 -0400)]
t5516: don't use HEAD ref for invalid ref-deletion tests
A few tests in t5516 want to assert that we can delete a corrupted ref
whose pointed-to object is missing. They do so by using the "main"
branch, which is also pointed to by HEAD.
This does work, but only because of a subtle assumption about the
implementation. We do not block the deletion because of the invalid ref,
but we _also_ do not notice that the deleted branch is pointed to by
HEAD. And so the safety rule of "do not allow HEAD to be deleted in a
non-bare repository" does not kick in, and the test passes.
Let's instead use a non-HEAD branch. That still tests what we care about
here (deleting a corrupt ref), but without implicitly depending on our
failure to notice that we're deleting HEAD. That will future proof the
test against that behavior changing.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Fri, 24 Sep 2021 18:32:17 +0000 (14:32 -0400)]
t7900: clean up some more broken refs
The "incremental-repack task" test replaces the object directory with a
known state. As a result, some of our refs point to objects that are not
included in that state.
Commit 3cf5f221be (t7900: clean up some broken refs, 2021-01-19) cleaned
up some of those (that were causing warnings to stderr from the
maintenance process). But there are a few more that were missed. These
aren't hurting anything for now, but it's certainly an unexpected state
to leave the test repository in, and it will become a problem if repack
ever gets more picky about broken refs.
Let's clean up those additional refs (which are all in refs/remotes,
with nothing there that isn't broken), and add an extra "for-each-ref"
call to assert that we've got everything.
Signed-off-by: Jeff King <peff@peff.net> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
http: check CURLE_SSL_PINNEDPUBKEYNOTMATCH when emitting errors
Change the error shown when a http.pinnedPubKey doesn't match to point
the http.pinnedPubKey variable added in aeff8a61216 (http: implement
public key pinning, 2016-02-15), e.g.:
git -c http.pinnedPubKey=sha256/someNonMatchingKey ls-remote https://github.com/git/git.git
fatal: unable to access 'https://github.com/git/git.git/' with http.pinnedPubkey configuration: SSL: public key does not match pinned public key!
Before this we'd emit the exact same thing without the " with
http.pinnedPubkey configuration". The advantage of doing this is that
we're going to get a translated message (everything after the ":" is
hardcoded in English in libcurl), and we've got a reference to the
git-specific configuration variable that's causing the error.
Unfortunately we can't test this easily, as there are no tests that
require https:// in the test suite, and t/lib-httpd.sh doesn't know
how to set up such tests. See [1] for the start of a discussion about
what it would take to have divergent "t/lib-httpd/apache.conf" test
setups. #leftoverbits
Andrzej Hunt [Sun, 26 Sep 2021 15:58:33 +0000 (15:58 +0000)]
connect: also update offset for features without values
parse_feature_value() takes an offset, and uses it to seek past the
point in features_list that we've already seen. However if the feature
being searched for does not specify a value, the offset is not
updated. Therefore if we call parse_feature_value() in a loop on a
value-less feature, we'll keep on parsing the same feature over and over
again. This usually isn't an issue: there's no point in using
next_server_feature_value() to search for repeated instances of the same
capability unless that capability typically specifies a value - but a
broken server could send a response that omits the value for a feature
even when we are expecting a value.
Therefore we add an offset update calculation for the no-value case,
which helps ensure that loops using next_server_feature_value() will
always terminate.
next_server_feature_value(), and the offset calculation, were first
added in 2.28 in 2c6a403d96 (connect: add function to parse multiple
v1 capability values, 2020-05-25).
Thanks to Peff for authoring the test.
Co-authored-by: Jeff King <peff@peff.net> Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Andrzej Hunt <andrzej@ahunt.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
hook-list.h: add a generated list of hooks, like config-list.h
Make githooks(5) the source of truth for what hooks git supports, and
punt out early on hooks we don't know about in find_hook(). This
ensures that the documentation and the C code's idea about existing
hooks doesn't diverge.
We still have Perl and Python code running its own hooks, but that'll
be addressed by Emily Shaffer's upcoming "git hook run" command.
This resolves a long-standing TODO item in bugreport.c of there being
no centralized listing of hooks, and fixes a bug with the bugreport
listing only knowing about 1/4 of the p4 hooks. It didn't know about
the recent "reference-transaction" hook either.
We could make the find_hook() function die() or BUG() out if the new
known_hook() returned 0, but let's make it return NULL just as it does
when it can't find a hook of a known type. Making it die() is overly
anal, and unlikely to be what we need in catching stupid typos in the
name of some new hook hardcoded in git.git's sources. By making this
be tolerant of unknown hook names, changes in a later series to make
"git hook run" run arbitrary user-configured hook names will be easier
to implement.
I have not been able to directly test the CMake change being made
here. Since 4c2c38e800 (ci: modification of main.yml to use cmake for
vs-build job, 2020-06-26) some of the Windows CI has a hard dependency
on CMake, this change works there, and is to my eyes an obviously
correct use of a pattern established in previous CMake changes,
namely:
- 061c2240b1 (Introduce CMake support for configuring Git,
2020-06-12)
- 709df95b78 (help: move list_config_help to builtin/help,
2020-04-16)
- 976aaedca0 (msvc: add a Makefile target to pre-generate the Visual
Studio solution, 2019-07-29)
The LC_ALL=C is needed because at least in my locale the dash ("-") is
ignored for the purposes of sorting, which results in a different
order. I'm not aware of anything in git that has a hard dependency on
the order, but e.g. the bugreport output would end up using whatever
locale was in effect when git was compiled.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Helped-by: René Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
hook.c users: use "hook_exists()" instead of "find_hook()"
Use the new hook_exists() function instead of find_hook() where the
latter was called in boolean contexts. This make subsequent changes in
a series where we further refactor the hook API clearer, as we won't
conflate wanting to get the path of the hook with checking for its
existence.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
hook.c: add a hook_exists() wrapper and use it in bugreport.c
Add a boolean version of the find_hook() function for those callers
who are only interested in checking whether the hook exists, not what
the path to it is.
Signed-off-by: Emily Shaffer <emilyshaffer@google.com> Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
hook.[ch]: move find_hook() from run-command.c to hook.c
Move the find_hook() function from run-command.c to a new hook.c
library. This change establishes a stub library that's pretty
pointless right now, but will see much wider use with Emily Shaffer's
upcoming "configuration-based hooks" series.
Eventually all the hook related code will live in hook.[ch]. Let's
start that process by moving the simple find_hook() function over
as-is.
Signed-off-by: Emily Shaffer <emilyshaffer@google.com> Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Johannes Sixt [Sun, 26 Sep 2021 10:05:11 +0000 (03:05 -0700)]
lazyload.h: fix warnings about mismatching function pointer types
Here, GCC warns about every use of the INIT_PROC_ADDR macro, for example:
In file included from compat/mingw.c:8:
compat/mingw.c: In function 'mingw_strftime':
compat/win32/lazyload.h:38:12: warning: assignment to
'size_t (*)(char *, size_t, const char *, const struct tm *)'
{aka 'long long unsigned int (*)(char *, long long unsigned int,
const char *, const struct tm *)'} from incompatible pointer type
'FARPROC' {aka 'long long int (*)()'} [-Wincompatible-pointer-types]
38 | (function = get_proc_addr(&proc_addr_##function))
| ^
compat/mingw.c:1014:6: note: in expansion of macro 'INIT_PROC_ADDR'
1014 | if (INIT_PROC_ADDR(strftime))
| ^~~~~~~~~~~~~~
(message wrapped for convenience). Insert a cast to keep the compiler
happy. A cast is fine in these cases because they are generic function
pointer values that have been looked up in a DLL.
Helped-by: Carlo Marcelo Arenas Belón <carenas@gmail.com> Signed-off-by: Johannes Sixt <j6t@kdbg.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The tests in t3705-add-sparse-checkout.sh check to see how 'git add'
behaves with paths outside the sparse-checkout definition. These
currently check to see if a given warning is present but not that the
index is not updated with the sparse entries. Add a new
'test_sparse_entry_unstaged' helper to be sure 'git add' is behaving
correctly.
We need to modify setup_sparse_entry to actually commit the sparse_entry
file so it exists at HEAD and as an entry in the index, but its exact
contents are not staged in the index.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t2500: add various tests for nuking untracked files
Noting that unpack_trees treats reset=1 & update=1 as license to nuke
untracked files, I looked for code paths that use this combination and
tried to generate testcases which demonstrated unintentional loss of
untracked files and directories. I found several.
I also include testcases for `git reset --{hard,merge,keep}`. A hard
reset is perhaps the most direct test of unpack_tree's reset=1 behavior,
but we cannot make `git reset --hard` preserve untracked files without
some migration work.
Also, the two commands `checkout --force` (because of the --force) and
`read-tree --reset` (because it's plumbing and we need to keep it
backward compatible) were left out as we expect those to continue
removing untracked files and directories.
Signed-off-by: Elijah Newren <newren@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
54fd3243da ("rebase -i: reread the todo list if `exec` touched it",
2017-04-26) sought to reread the todo list after running an exec
command only if it had been changed. To accomplish this it checks the
stat data of the todo list after running an exec command to see if it
has changed. Unfortunately there are two problems, firstly the
implementation is buggy we actually reread the list after each exec
which is quadratic in the number of commit lookups and secondly the
design is predicated on using nanosecond time stamps which are not the
default.
The implementation bug stems from the fact that we write a new todo
list to disk before running each command but do not update the stat
data to reflect this[1].
The design problem is that it is possible for the user to edit the
todo list without changing its size or inode which means we have to
rely on the mtime to tell us if it has changed. Unfortunately unless
git is built with USE_NSEC it is possible for the original and edited
list to share the same mtime.
Ideally "git rebase --edit-todo" would set a flag that we would then
check in sequencer.c. Unfortunately this is approach will not work as
there are scripts in the wild that write to the todo list directly
without running "git rebase --edit-todo". Instead of relying on stat
data this patch simply reads the possibly edited todo list and
compares it to the original with memcmp(). This is much faster than
reparsing the todo list each time. This patch reduces the time to run
git rebase -r -xtrue v2.32.0~100 v2.32.0
which runs 419 exec commands by 6.6%. For comparison fixing the
implementation bug in stat based approach reduces the time by a
further 1.4% and is indistinguishable from never rereading the todo
list.
Reported-by: SZEDER Gábor <szeder.dev@gmail.com> Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk> Signed-off-by: Junio C Hamano <gitster@pobox.com>
This code is heavily indented and obscures the high level logic within
the loop. Let's move it to its own function before modifying it in the
next commit. Note that there is a subtle change in behavior if the
todo list cannot be reread. Previously todo_list->current was
incremented before returning, now it returns immediately.
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk> Signed-off-by: Junio C Hamano <gitster@pobox.com>
This comment added in dfea575017 (Makefile: lazily compute header
dependencies, 2010-01-26) has been out of date since 92b88eba9f (Makefile: use `git ls-files` to list header files, if
possible, 2019-03-04), when we did exactly what it tells us not to do
and added $(GENERATED_H) to $(OBJECTS) dependencies.
The rest of it was also somewhere between inaccurate and outdated,
since as of b8ba629264 (Makefile: fold MISC_H into LIB_H, 2012-06-20)
it's not followed by a list of header files, that got moved earlier in
the file into LIB_H in 60d24dd255 (Makefile: fold XDIFF_H and VCSSVN_H
into LIB_H, 2012-07-06).
Let's just remove it entirely, to the extent that we have anything
useful to say here the comment on the
"USE_COMPUTED_HEADER_DEPENDENCIES" variable a few lines above this
change does the job for us.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Makefile: don't perform "mv $@+ $@" dance for $(GENERATED_H)
Change the "cmd.sh > $@+ && mv $@+ $@" pattern used for generating the
config-list.h and command-list.h to just "cmd.sh >$@". This was needed
as a guard to ensure that we don't have an empty file if the script
failed, but since 7b76d6bf221 (Makefile: add and use the
".DELETE_ON_ERROR" flag, 2021-06-29) GNU make ensures that doesn't
happen.
There's still a lot of other places in the Makefile where we
needlessly use this pattern, but I'm just changing these because I'm
about to add a new $(GENERATED_H) target, let's have them all look and
act the same way.
Even with ".DELETE_ON_ERROR" there is still a point to using the "mv
$@+ $@" pattern in some cases, e.g. to ensure that you have a working
binary during recompilation (see [1] for the start of a long
discussion about that), but that doesn't apply here. Nothing external
uses $(GENERATED_H) directly, it's only ever used in the context of
the Makefile's own dependency (re-)generation.
Change various places that hardcode the names of these two files to
refer to either $(GENERATED_H), or to a new generated-hdrs
target. That target is consistent with the *-objs targets I recently
added in 029bac01a8 (Makefile: add {program,xdiff,test,git,fuzz}-objs
& objects targets, 2021-02-23).
A subsequent commit will add a new generated hook-list.h. By doing
this refactoring we'll only need to add the new file to the
GENERATED_H variable, not EXCEPT_HDRS, the vcbuild/README etc.
Hardcoding command-list.h there seems to have been a case of
copy/paste programming in 976aaedca0 (msvc: add a Makefile target to
pre-generate the Visual Studio solution, 2019-07-29). The
config-list.h was added later in 709df95b78 (help: move
list_config_help to builtin/help, 2020-04-16).
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Fix a bug in 44c9e8594e (Fix up header file dependencies and add
sparse checking rules, 2005-07-03), we never marked the phony "check"
target as such.
Perhaps we should just remove it, since as of a combination of 912f9980d2 (Makefile: help people who run 'make check' by mistake,
2008-11-11) 0bcd9ae85d (sparse: Fix errors due to missing
target-specific variables, 2011-04-21) we've been suggesting the user
run "make sparse" directly.
But under that mode it still does something, as well as directing the
user to run "make test" under non-sparse. So let's punt that and
narrowly fix the PHONY bug.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
bundle: remove ignored & undocumented "--verbose" flag
In 73c3253d75e (bundle: framework for options before bundle file,
2019-11-10) the "git bundle" command was refactored to use
parse_options(). In that refactoring it started understanding the
"--verbose" flag before the subcommand, e.g.:
git bundle --verbose verify --quiet
However, nothing ever did anything with this "verbose" variable, and
the change wasn't documented. It appears to have been something that
escaped the lab, and wasn't flagged by reviewers at the time. Let's
just remove it.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Reviewed-by: Taylor Blau <me@ttaylorr.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Thu, 23 Sep 2021 20:44:48 +0000 (13:44 -0700)]
Merge branch 'cb/unix-sockets-with-windows'
Adjust credential-cache helper to Windows.
* cb/unix-sockets-with-windows:
git-compat-util: include declaration for unix sockets in windows
credential-cache: check for windows specific errors
t0301: fixes for windows compatibility
Junio C Hamano [Thu, 23 Sep 2021 20:44:48 +0000 (13:44 -0700)]
Merge branch 'ab/retire-option-argument'
An oddball OPTION_ARGUMENT feature has been removed from the
parse-options API.
* ab/retire-option-argument:
parse-options API: remove OPTION_ARGUMENT feature
difftool: use run_command() API in run_file_diff()
difftool: prepare "diff" cmdline in cmd_difftool()
difftool: prepare "struct child_process" in cmd_difftool()
Junio C Hamano [Thu, 23 Sep 2021 20:44:48 +0000 (13:44 -0700)]
Merge branch 'mr/bisect-in-c-4'
Rewrite of "git bisect" in C continues.
* mr/bisect-in-c-4:
bisect--helper: retire `--bisect-next-check` subcommand
bisect--helper: reimplement `bisect_run` shell function in C
bisect--helper: reimplement `bisect_visualize()` shell function in C
run-command: make `exists_in_PATH()` non-static
t6030-bisect-porcelain: add test for bisect visualize
t6030-bisect-porcelain: add tests to control bisect run exit cases
Junio C Hamano [Thu, 23 Sep 2021 20:44:47 +0000 (13:44 -0700)]
Merge branch 'ab/http-drop-old-curl-plus'
Conditional compilation around versions of libcURL has been
straightened out.
* ab/http-drop-old-curl-plus:
http: don't hardcode the value of CURL_SOCKOPT_OK
http: centralize the accounting of libcurl dependencies
http: correct curl version check for CURLOPT_PINNEDPUBLICKEY
http: correct version check for CURL_HTTP_VERSION_2
http: drop support for curl < 7.18.0 (again)
Makefile: drop support for curl < 7.9.8 (again)
INSTALL: mention that we need libcurl 7.19.4 or newer to build
INSTALL: reword and copy-edit the "libcurl" section
INSTALL: don't mention the "curl" executable at all
Replace a handcrafted data structure used to keep track of bad
objects in the packfile API by an oidset.
* rs/packfile-bad-object-list-in-oidset:
packfile: use oidset for bad objects
packfile: convert has_packed_and_bad() to object_id
packfile: convert mark_bad_packed_object() to object_id
midx: inline nth_midxed_pack_entry()
oidset: make oidset_size() an inline function
tests: add a test mode for SANITIZE=leak, run it in CI
While git can be compiled with SANITIZE=leak, we have not run
regression tests under that mode. Memory leaks have only been fixed as
one-offs without structured regression testing.
This change adds CI testing for it. We'll now build and small set of
whitelisted t00*.sh tests under Linux with a new job called
"linux-leaks".
The CI target uses a new GIT_TEST_PASSING_SANITIZE_LEAK=true test
mode. When running in that mode, we'll assert that we were compiled
with SANITIZE=leak. We'll then skip all tests, except those that we've
opted-in by setting "TEST_PASSES_SANITIZE_LEAK=true".
A test setting "TEST_PASSES_SANITIZE_LEAK=true" setting can in turn
make use of the "SANITIZE_LEAK" prerequisite, should they wish to
selectively skip tests even under
"GIT_TEST_PASSING_SANITIZE_LEAK=true". In the preceding commit we
started doing this in "t0004-unwritable.sh" under SANITIZE=leak, now
it'll combine nicely with "GIT_TEST_PASSING_SANITIZE_LEAK=true".
This is how tests that don't set "TEST_PASSES_SANITIZE_LEAK=true" will
be skipped under GIT_TEST_PASSING_SANITIZE_LEAK=true:
$ GIT_TEST_PASSING_SANITIZE_LEAK=true ./t0001-init.sh
1..0 # SKIP skip all tests in t0001 under SANITIZE=leak, TEST_PASSES_SANITIZE_LEAK not set
The intent is to add more TEST_PASSES_SANITIZE_LEAK=true annotations
as follow-up change, but let's start small to begin with.
In ci/run-build-and-tests.sh we make use of the default "*" case to
run "make test" without any GIT_TEST_* modes. SANITIZE=leak is known
to fail in combination with GIT_TEST_SPLIT_INDEX=true in
t0016-oidmap.sh, and we're likely to have other such failures in
various GIT_TEST_* modes. Let's focus on getting the base tests
passing, we can expand coverage to GIT_TEST_* modes later.
It would also be possible to implement a more lightweight version of
this by only relying on setting "LSAN_OPTIONS". See
<YS9OT/pn5rRK9cGB@coredump.intra.peff.net>[1] and
<YS9ZIDpANfsh7N+S@coredump.intra.peff.net>[2] for a discussion of
that. I've opted for this approach of adding a GIT_TEST_* mode instead
because it's consistent with how we handle other special test modes.
Being able to add a "!SANITIZE_LEAK" prerequisite and calling
"test_done" early if it isn't satisfied also means that we can more
incrementally add regression tests without being forced to fix
widespread and hard-to-fix leaks at the same time.
We have tests that do simple checking of some tool we're interested
in, but later on in the script might be stressing trace2, or common
sources of leaks like "git log" in combination with the tool (e.g. the
commit-graph tests). To be clear having a prerequisite could also be
accomplished by using "LSAN_OPTIONS" directly.
On the topic of "LSAN_OPTIONS": It would be nice to have a mode to
aggregate all failures in our various scripts, see [2] for a start at
doing that which sets "log_path" in "LSAN_OPTIONS". I've punted on
that for now, it can be added later.
As of writing this we've got major regressions between master..seen,
i.e. the t000*.sh tests and more fixed since 31f9acf9ce2 (Merge branch
'ah/plugleaks', 2021-08-04) have regressed recently.
See the discussion at <87czsv2idy.fsf@evledraar.gmail.com>[3] about
the lack of this sort of test mode, and 0e5bba53af (add UNLEAK
annotation for reducing leak false positives, 2017-09-08) for the
initial addition of SANITIZE=leak.
See also 09595ab381 (Merge branch 'jk/leak-checkers', 2017-09-19), 7782066f67 (Merge branch 'jk/apache-lsan', 2019-05-19) and the recent 936e58851a (Merge branch 'ah/plugleaks', 2021-05-07) for some of the
past history of "one-off" SANITIZE=leak (and more) fixes.
As noted in [5] we can't support this on OSX yet until Clang 14 is
released, at that point we'll probably want to resurrect that
"osx-leaks" job.
Makefile: add SANITIZE=leak flag to GIT-BUILD-OPTIONS
When SANITIZE=leak is specified we'll now add a SANITIZE_LEAK flag to
GIT-BUILD-OPTIONS, this can then be picked up by the test-lib.sh,
which sets a SANITIZE_LEAK prerequisite.
We can then skip specific tests that are known to fail under
SANITIZE=leak, add one such annotation to t0004-unwritable.sh, which
now passes under SANITIZE=leak.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
David Aguilar [Thu, 23 Sep 2021 04:12:50 +0000 (21:12 -0700)]
difftool: fix symlink-file writing in dir-diff mode
The difftool dir-diff mode handles symlinks by replacing them with their
readlink(2) values. This allows diff tools to see changes to symlinks
as if they were regular text diffs with the old and new path values.
This is analogous to what "git diff" displays when symlinks change.
The temporary diff directories that are created initially contain
symlinks because they get checked-out using a temporary index that
retains the original symlinks as checked-in to the repository.
A bug was introduced when difftool was rewritten in C that made
difftool write the readlink(2) contents into the pointed-to file rather
than the symlink itself. The write was going through the symlink and
writing to its target rather than writing to the symlink path itself.
Replace symlinks with raw text files by unlinking the symlink path
before writing the readlink(2) content into them.
When 18ec800512 (difftool: handle modified symlinks in dir-diff mode,
2017-03-15) added handling for modified symlinks this bug got recorded
in the test suite. The tests included the pointed-to symlink target
paths. These paths were being reported because difftool was erroneously
writing to them, but they should have never been reported nor written.
Correct the modified-symlinks test cases by removing the target files
from the expected output.
Add a test to ensure that symlinks are written with the readlink(2)
values and that the target files contain their original content.
Reported-by: Alan Blotz <work@blotz.org> Helped-by: Đoàn Trần Công Danh <congdanhqx@gmail.com> Signed-off-by: David Aguilar <davvid@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
help: move column config discovery to help.c library
When a git_config() call was added in dbfae689690 (help: reuse
print_columns() for help -a, 2012-04-13) to read the column config
we'd always use the resulting "colopts" variable.
Then in 63eae83f8f3 (help: add "-a --verbose" to list all commands
with synopsis, 2018-05-20) we started only using the "colopts" config
under "--all" if "--no-verbose" was also given, but the "git_config()"
call was not moved inside the "verbose" branch of the code.
This change effectively does that, we'll only call list_commands()
under "--all --no-verbose", so let's have it look up the config it
needs. See 26c7d067832 (help -a: improve and make --verbose default, 2018-09-29) for another case in help.c where we look up config.
The get_colopts() function is named for consistency with the existing
get_alias() function added in 26c7d067832.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
help / completion: make "git help" do the hard work
The "help" builtin has been able to emit configuration variables since e17ca926371 (completion: drop the hard coded list of config vars,
2018-05-26), but it hasn't produced exactly the format the completion
script wanted. Let's do that.
We got partway there in 2675ea1cc0f (completion: use 'sort -u' to
deduplicate config variable names, 2019-08-13) and d9438873c4d (completion: deduplicate configuration sections,
2019-08-13), but after both we still needed some sorting,
de-duplicating and awk post-processing of the list.
We can instead simply do the relevant parsing ourselves (we were doing
most of it already), and call string_list_remove_duplicates() after
already sorting the list, so the caller doesn't need to invoke "sort
-u". The "--config-for-completion" output is the same as before after
being passed through "sort -u".
Then add a new "--config-sections-for-completion" option. Under that
output we'll emit config sections like "alias" (instead of "alias." in
the --config-for-completion output).
We need to be careful to leave the "--config-for-completion" option
compatible with users git, but are still running a shell with an older
git-completion.bash. If we e.g. changed the option name they'd see
messages about git-completion.bash being unable to find the
"--config-for-completion" option.
Such backwards compatibility isn't something we should bend over
backwards for, it's only helping users who:
* Upgrade git
* Are in an old shell
* The git-completion.bash in that shell hasn't cached the old
"--config-for-completion" output already.
But since it's easy in this case to retain compatibility, let's do it,
the older versions of git-completion.bash won't care that the input
they get doesn't change after a "sort -u".
While we're at it let's make "--config-for-completion" die if there's
anything left over in "argc", and do the same in the new
"--config-sections-for-completion" option.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
help tests: test --config-for-completion option & output
Add a regression test for the --config-for-completion option, this was
tested for indirectly with the test added in 7a09a8f093e (completion:
add tests for 'git config' completion, 2019-08-13), but let's do it
directly here as well.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
As preceding commits have incrementally established all of the --all,
--guides, --config and hidden --config-for-completion options are
mutually exclusive. So let's use OPT_CMDMODE() to parse the
command-line instead, and take advantage of its conflicting options
detection.
This is the first command with a hidden CMDMODE, so let's introduce a
OPT_CMDMODE_F() macro to go along with OPT_CMDMODE().
I think this makes the usage information that we emit slightly worse,
e.g. before we'd emit:
$ git help --all --config
fatal: --config and --all cannot be combined
usage: git help [-a|--all] [--[no-]verbose]]
[[-i|--info] [-m|--man] [-w|--web]] [<command>]
or: git help [-g|--guides]
or: git help [-c|--config]
[...]
$
And now:
$ git help --all --config
error: option `config' is incompatible with --all
$
But improving that is a general topic for parse-options.c improvement,
i.e. we should probably emit the full usage in that case.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
help: correct logic error in combining --all and --guides
The --all and --guides commands could be combined, which wouldn't have
any impact on the output except for:
git help --all --guides --no-verbose
Listing the guide alongside that output was clearly not intended, so
let's error out here. See 002b726a400 (builtin/help.c: add
list_common_guides_help() function, 2013-04-02) for the initial
implementation.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
help: correct logic error in combining --all and --config
Fix a bug in the --config option that's been there ever since its
introduction in 3ac68a93fd2 (help: add --config to list all available
config, 2018-05-26). Die when --all and --config are combined,
combining them doesn't make sense.
The code for the --config option when combined with an earlier
refactoring done to support the --guide option in 65f98358c0c (builtin/help.c: add --guide option, 2013-04-02) would
cause us to take the "--all" branch early and ignore the --config
option.
Let's instead list these as incompatible, both in the synopsis and
help output, and enforce it in the code itself.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a missing test for checking what the --config output added in ac68a93fd2 (help: add --config to list all available config,
2018-05-26) looks like. We should not be emitting anything except
config variables and the brief usage information at the end here.
The second test regexp here might not match three-level variables in
general, as their second level could contain ".", but in this case
we're always emitting what we extract from the documentation, so it's
all strings like:
foo.<name>.bar
If we did introduce something like variable example content here we'd
like this to break, since we'd then be likely to break the
git-completion.bash.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
help: correct usage & behavior of "git help --guides"
As noted in 65f98358c0c (builtin/help.c: add --guide option,
2013-04-02) and a133737b809 (doc: include --guide option description
for "git help", 2013-04-02) which introduced the --guide option, it
cannot be combined with e.g. <command>.
Change the command and the "SYNOPSIS" section to reflect that desired
behavior. Now that we assert this in code we don't need to
exhaustively describe the previous confusing behavior in the
documentation either, instead of silently ignoring the provided
argument we'll now error out.
The "We're done. Ignore any remaining args" comment added in 15f7d494380 (builtin/help.c: split "-a" processing into two,
2013-04-02) can now be removed, it's obvious that we're asserting the
behavior with the check of "argc".
The "--config" option is still missing from the synopsis, it will be
added in a subsequent commit where we'll fix bugs in its
implementation.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Thu, 23 Sep 2021 16:47:05 +0000 (09:47 -0700)]
Merge branch 'jk/grep-haystack-is-read-only' into hm/paint-hits-in-log-grep
* jk/grep-haystack-is-read-only:
grep: store grep_source buffer as const
grep: mark "haystack" buffers as const
grep: stop modifying buffer in grep_source_1()
grep: stop modifying buffer in show_line()
grep: stop modifying buffer in strip_timestamp
Makefile: make COMPUTE_HEADER_DEPENDENCIES=auto work with DEVOPTS=pedantic
The "COMPUTE_HEADER_DEPENDENCIES" feature added in [1] was extended to
use auto-detection in [2], that "auto" detection has always piped
STDERR to /dev/null, so any failures on compilers that didn't support
these GCC flags would silently fall back to
"COMPUTE_HEADER_DEPENDENCIES=no".
Later when -Wpedantic support was added to DEVOPTS in [3] we started
passing -Wpedantic in combination with -Werror to the compiler
here. Note (to the pedantic): [3] actually passed "-pedantic", but it
and "-Wpedantic" are synonyms.
Turning on -Wpedantic in [3] broke the auto-detection, since this
relies on compiling an empty program. GCC would loudly complain on
STDERR:
/dev/null:1: error: ISO C forbids an empty translation unit
[-Werror=pedantic]
cc1: note: unrecognized command-line option
‘-Wno-pedantic-ms-format’ may have been intended to silence
earlier diagnostics
cc1: all warnings being treated as errors
But as that ended up in the "$(dep_check)" variable due to the "2>&1"
in [2] we didn't see it.
Then when [4] made DEVOPTS=pedantic the default specifying
"DEVELOPER=1" would effectively set "COMPUTE_HEADER_DEPENDENCIES=no".
To fix these issues let's unconditionally pass -Wno-pedantic after
$(ALL_CFLAGS), we might get a -Wpedantic via config.mak.dev after, or
the builder might specify it via CFLAGS. In either case this will undo
current and future problems with -Wpedantic.
I think it would make sense to simply remove the "2>&1", it would mean
that anyone using a non-GCC-like compiler would get warnings under
COMPUTE_HEADER_DEPENDENCIES=auto, e.g on AIX's xlc would emit:
/opt/IBM/xlc/13.1.3/bin/.orig/xlc: 1501-208 (S) command option D is missing a subargument
Non-zero 40 exit with COMPUTE_HEADER_DEPENDENCIES=auto, set it to "yes" or "no" to quiet auto-detect
And on Solaris with SunCC:
cc: Warning: Option -x passed to ld, if ld is invoked, ignored otherwise
cc: refused to overwrite input file by output file: /dev/null
cc: Warning: Option -x passed to ld, if ld is invoked, ignored otherwise
cc: refused to overwrite input file by output file: /dev/null
Non-zero 1 exit with COMPUTE_HEADER_DEPENDENCIES=auto, set it to "yes" or "no" to quiet auto-detect
Both could be quieted by setting COMPUTE_HEADER_DEPENDENCIES=no
explicitly, as suggested, but let's see if this'll fix it without
emitting too much noise at those that aren't using "gcc" or "clang".
1. f2fabbf76e4 (Teach Makefile to check header dependencies,
2010-01-26)
2. 111ee18c31f (Makefile: Use computed header dependencies if the
compiler supports it, 2011-08-18)
3. 729b3925ed9 (Makefile: add a DEVOPTS flag to get pedantic
compilation, 2018-07-24)
4. 6a8cbc41bac (developer: enable pedantic by default, 2021-09-03)
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change the "sparse" target and its *.sp dependencies to be
non-.PHONY. Before this change "make sparse" would take ~5s to re-run
all the *.c files through "cgcc", after it it'll create an empty *.sp
file sitting alongside the *.c file, only if the *.c file or its
dependencies are newer than the *.sp is the *.sp re-made.
We ensure that the recursive dependencies are correct by depending on
the *.o file, which in turn will have correct dependencies by either
depending on all header files, or under
"COMPUTE_HEADER_DEPENDENCIES=yes" the headers it needs.
This means that a plain "make sparse" is much slower, as we'll now
need to make the *.o files just to create the *.sp files, but
incrementally creating the *.sp files is *much* faster and less
verbose, it thus becomes viable to run "sparse" along with "all" as
e.g. "git rebase --exec 'make all sparse'".
On my box with -j8 "make sparse" was fast before, or around 5 seconds,
now it only takes that long the first time, and the common case is
<100ms, or however long it takes GNU make to stat the *.sp file and
see that all the corresponding *.c file and its dependencies are
older.
See 0bcd9ae85d7 (sparse: Fix errors due to missing target-specific
variables, 2011-04-21) for the modern implementation of the sparse
target being changed here.
It is critical that we use -Wsparse-error here, otherwise the error
would only show up once, but we'd successfully create the empty *.sp
file, and running a second time wouldn't show the error. I'm therefore
not putting it into SPARSE_FLAGS or SP_EXTRA_FLAGS, it's not optional,
the Makefile logic won't behave properly without it.
Appending to $@ without a move is OK here because we're using the
.DELETE_ON_ERROR Makefile feature. See 7b76d6bf221 (Makefile: add and
use the ".DELETE_ON_ERROR" flag, 2021-06-29). GNU make ensures that on
error this file will be removed.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Wed, 22 Sep 2021 22:11:53 +0000 (18:11 -0400)]
http: match headers case-insensitively when redacting
When HTTP/2 is in use, we fail to correctly redact "Authorization" (and
other) headers in our GIT_TRACE_CURL output.
We get the headers in our CURLOPT_DEBUGFUNCTION callback, curl_trace().
It passes them along to curl_dump_header(), which in turn checks
redact_sensitive_header(). We see the headers as a text buffer like:
Host: ...
Authorization: Basic ...
After breaking it into lines, we match each header using skip_prefix().
This is case-sensitive, even though HTTP headers are case-insensitive.
This has worked reliably in the past because these headers are generated
by curl itself, which is predictable in what it sends.
But when HTTP/2 is in use, instead we get a lower-case "authorization:"
header, and we fail to match it. The fix is simple: we should match with
skip_iprefix().
Testing is more complicated, though. We do have a test for the redacting
feature, but we don't hit the problem case because our test Apache setup
does not understand HTTP/2. You can reproduce the issue by applying this
on top of the test change in this patch:
- it's not necessarily portable. The http2 apache module might not be
available on all systems. Further, the http2 module isn't compatible
with the prefork mpm, so we have to switch to something else. But we
don't necessarily know what's available. It would be nice if we
could have conditional config, but IfModule only tells us if a
module is already loaded, not whether it is available at all.
This might be a non-issue. The http tests are already optional, and
modern-enough systems may just have both of these. But...
- if we do this, then we'd no longer be testing HTTP/1.1 at all. I'm
not sure how much that matters since it's all handled by curl under
the hood, but I'd worry that some detail leaks through. We'd
probably want two scripts running similar tests, one with HTTP/2 and
one with HTTP/1.1.
- speaking of which, a later test fails with the patch above! The
problem is that it is making sure we used a chunked
transfer-encoding by looking for that header in the trace. But
HTTP/2 doesn't support that, as it has its own streaming mechanisms
(the overall operation works fine; we just don't see the header in
the trace).
Furthermore, even with the changes above, this test still does not
detect the current failure, because we see _both_ HTTP/1.1 and HTTP/2
requests, which confuse it. Quoting only the interesting bits from the
resulting trace file, we first see:
So the client does properly redact there, because we're speaking
HTTP/1.1, and the server indicates it can do the upgrade. And then the
client will make further requests using HTTP/2:
And there we can see that the credential is _not_ redacted. This part of
the test is what gets confused:
# Ensure that there is no "Basic" followed by a base64 string, but that
# the auth details are redacted
! grep "Authorization: Basic [0-9a-zA-Z+/]" trace &&
grep "Authorization: Basic <redacted>" trace
The first grep does not match the un-redacted HTTP/2 header, because
it insists on an uppercase "A". And the second one does find the
HTTP/1.1 header. So as far as the test is concerned, everything is OK,
but it failed to notice the un-redacted lines.
We can make this test (and the other related ones) more robust by adding
"-i" to grep case-insensitively. This isn't really doing anything for
now, since we're not actually speaking HTTP/2, but it future-proofs the
tests for a day when we do (either we add explicit HTTP/2 test support,
or it's eventually enabled by default by our Apache+curl test setup).
And it doesn't hurt in the meantime for the tests to be more careful.
The change to use "grep -i", coupled with the changes to use HTTP/2
shown above, causes the test to fail with the current code, and pass
after this patch is applied.
And finally, there's one other way to demonstrate the issue (and how I
actually found it originally). Looking at GIT_TRACE_CURL output against
github.com, you'll see the unredacted output, even if you didn't set
http.version. That's because setting it is only necessary for curl to
send the extra headers in its HTTP/1.1 request that say "Hey, I speak
HTTP/2; upgrade if you do, too". But for a production site speaking
https, the server advertises via ALPN, a TLS extension, that it supports
HTTP/2, and the client can immediately start using it.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
test-lib tests: get rid of copy/pasted mock test code
Now that we've split up the write_sub_test_lib_test*() and
run_sub_test_lib_test*() functions let's fix those tests in
t0000-basic.sh that were verbosely copy/pasting earlier tests.
That we caught all of them was asserted with a follow-up change that's
not part of this series[1], we might add such a duplication check at
some later time, but for now let's just one-off remove the duplicate
boilerplate.
Improve the testing for test-lib.sh itself to assert that we have a
exit code of 1, not any non-zero. Improves code added in 0445e6f0a12 (test-lib: '--run' to run only specific tests,
2014-04-30).
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
test-lib tests: refactor common part of check_sub_test_lib_test*()
Refactor the two check_sub_test_lib_test*() functions to avoid
duplicating the same comparison they did of stdout. This duplication
was initially added when check_sub_test_lib_test_err() was added in 0445e6f0a12 (test-lib: '--run' to run only specific tests,
2014-04-30).
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
test-lib tests: avoid subshell for "test_cmp" for readability
The use of a sub-shell for running the test_cmp of stdout/stderr for
the test author was introduced in this form in 565b6fa87bb (tests:
refactor mechanics of testing in a sub test-lib, 2012-12-16), but from
looking at the history that seemed to have diligently copied my
original ad-hoc implementation in 7b905119703 (t/t0000-basic.sh: Run
the passing TODO test inside its own test-lib, 2010-08-19).
There's no reason to use a subshell here, we try to avoid it in
general. It also improves readability, if the test fails we print out
the relative path in the trash directory that needs to be looked
at.
Before that was mostly obscured, since the "write_sub_test_lib_test"
will pick the directory for you from the test name.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
projects / thirdparty / git.git / log
