summaryrefslogtreecommitdiff
path: root/t
AgeCommit message (Collapse)AuthorFilesLines
2019-07-30test-dir-iterator: do not assume errno valuesLibravatar Junio C Hamano2-3/+12
A few tests printed 'errno' as an integer and compared with hardcoded integers; this is obviously not portable. A two things to note are: - the string obtained by strerror() is not portable, and cannot be used for the purpose of these tests. - there unfortunately isn't a portable way to map error numbers to error names. As we only care about a few selected errors, just map the error number to the name before emitting for comparison. Reported-by: Randall S. Becker <rsbecker@nexbridge.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-11clone: copy hidden paths at local cloneLibravatar Matheus Tavares1-0/+9
Make the copy_or_link_directory function no longer skip hidden directories. This function, used to copy .git/objects, currently skips all hidden directories but not hidden files, which is an odd behaviour. The reason for that could be unintentional: probably the intention was to skip '.' and '..' only but it ended up accidentally skipping all directories starting with '.'. Besides being more natural, the new behaviour is more permissive to the user. Also adjust tests to reflect this behaviour change. Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Co-authored-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-11dir-iterator: add flags parameter to dir_iterator_beginLibravatar Matheus Tavares2-7/+107
Add the possibility of giving flags to dir_iterator_begin to initialize a dir-iterator with special options. Currently possible flags are: - DIR_ITERATOR_PEDANTIC, which makes dir_iterator_advance abort immediately in the case of an error, instead of keep looking for the next valid entry; - DIR_ITERATOR_FOLLOW_SYMLINKS, which makes the iterator follow symlinks and include linked directories' contents in the iteration. These new flags will be used in a subsequent patch. Also add tests for the flags' usage and adjust refs/files-backend.c to the new dir_iterator_begin signature. Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-11dir-iterator: refactor state machine modelLibravatar Matheus Tavares2-0/+18
dir_iterator_advance() is a large function with two nested loops. Let's improve its readability factoring out three functions and simplifying its mechanics. The refactored model will no longer depend on level.initialized and level.dir_state to keep track of the iteration state and will perform on a single loop. Also, dir_iterator_begin() currently does not check if the given string represents a valid directory path. Since the refactored model will have to stat() the given path at initialization, let's also check for this kind of error and make dir_iterator_begin() return NULL, on failures, with errno appropriately set. And add tests for this new behavior. Improve documentation at dir-iteration.h and code comments at dir-iterator.c to reflect the changes and eliminate possible ambiguities. Finally, adjust refs/files-backend.c to check for now possible dir_iterator_begin() failures. Original-patch-by: Daniel Ferreira <bnmvco@gmail.com> Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-11dir-iterator: add tests for dir-iterator APILibravatar Daniel Ferreira4-0/+90
Create t/helper/test-dir-iterator.c, which prints relevant information about a directory tree iterated over with dir-iterator. Create t/t0066-dir-iterator.sh, which tests that dir-iterator does iterate through a whole directory tree as expected. Signed-off-by: Daniel Ferreira <bnmvco@gmail.com> [matheus.bernardino: update to use test-tool and some minor aesthetics] Helped-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-11clone: better handle symlinked files at .git/objects/Libravatar Matheus Tavares1-7/+20
There is currently an odd behaviour when locally cloning a repository with symlinks at .git/objects: using --no-hardlinks all symlinks are dereferenced but without it, Git will try to hardlink the files with the link() function, which has an OS-specific behaviour on symlinks. On OSX and NetBSD, it creates a hardlink to the file pointed by the symlink whilst on GNU/Linux, it creates a hardlink to the symlink itself. On Manjaro GNU/Linux: $ touch a $ ln -s a b $ link b c $ ls -li a b c 155 [...] a 156 [...] b -> a 156 [...] c -> a But on NetBSD: $ ls -li a b c 2609160 [...] a 2609164 [...] b -> a 2609160 [...] c It's not good to have the result of a local clone to be OS-dependent and besides that, the current behaviour on GNU/Linux may result in broken symlinks. So let's standardize this by making the hardlinks always point to dereferenced paths, instead of the symlinks themselves. Also, add tests for symlinked files at .git/objects/. Note: Git won't create symlinks at .git/objects itself, but it's better to handle this case and be friendly with users who manually create them. Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Co-authored-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-11clone: test for our behavior on odd objects/* contentLibravatar Ævar Arnfjörð Bjarmason1-0/+111
Add tests for what happens when we perform a local clone on a repo containing odd files at .git/object directory, such as symlinks to other dirs, or unknown files. I'm bending over backwards here to avoid a SHA-1 dependency. See [1] for an earlier and simpler version that hardcoded SHA-1s. This behavior has been the same for a *long* time, but hasn't been tested for. There's a good post-hoc argument to be made for copying over unknown things, e.g. I'd like a git version that doesn't know about the commit-graph to copy it under "clone --local" so a newer git version can make use of it. In follow-up commits we'll look at changing some of this behavior, but for now, let's just assert it as-is so we'll notice what we'll change later. 1. https://public-inbox.org/git/20190226002625.13022-5-avarab@gmail.com/ Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> [matheus.bernardino: improved and split tests in more than one patch] Helped-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Matheus Tavares <matheus.bernardino@usp.br> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-13Merge branch 'sb/format-patch-base-patch-id-fix'Libravatar Junio C Hamano1-3/+29
The "--base" option of "format-patch" computed the patch-ids for prerequisite patches in an unstable way, which has been updated to compute in a way that is compatible with "git patch-id --stable". * sb/format-patch-base-patch-id-fix: format-patch: make --base patch-id output stable format-patch: inform user that patch-id generation is unstable
2019-06-13Merge branch 'nd/init-relative-template-fix'Libravatar Junio C Hamano2-4/+4
A relative pathname given to "git init --template=<path> <repo>" ought to be relative to the directory "git init" gets invoked in, but it instead was made relative to the repository, which has been corrected. * nd/init-relative-template-fix: init: make --template path relative to $CWD
2019-06-13Merge branch 'ab/send-email-transferencoding-fix'Libravatar Junio C Hamano1-20/+82
Since "git send-email" learned to take 'auto' as the value for the transfer-encoding, it by mistake stopped honoring the values given to the configuration variables sendemail.transferencoding and/or sendemail.<ident>.transferencoding. This has been corrected to (finally) redoing the order of setting the default, reading the configuration and command line options. * ab/send-email-transferencoding-fix: send-email: fix regression in sendemail.identity parsing send-email: document --no-[to|cc|bcc] send-email: fix broken transferEncoding tests send-email: remove cargo-culted multi-patch pattern in tests send-email: do defaults -> config -> getopt in that order send-email: rename the @bcclist variable for consistency send-email: move the read_config() function above getopts
2019-06-06Merge branch 'en/merge-directory-renames-fix'Libravatar Junio C Hamano1-0/+116
Recent code restructuring of merge-recursive engine introduced a regression dealing with rename/add conflict. * en/merge-directory-renames-fix: merge-recursive: restore accidentally dropped setting of path
2019-06-05merge-recursive: restore accidentally dropped setting of pathLibravatar Elijah Newren1-0/+116
In commit 8daec1df03de ("merge-recursive: switch from (oid,mode) pairs to a diff_filespec", 2019-04-05), we actually switched from (oid,mode,path) triplets to a diff_filespec -- but most callsites in the patch only needed to worry about oid and mode so the commit message focused on that. The oversight in the commit message apparently spilled over to the code as well; one of the dozen or so callsites accidentally dropped the setting of the path in the conversion. Restore the path setting in that location. Also, this pointed out that our testsuite was lacking a good rename/add test, at least one that involved the need for merge content with the rename. Add such a test, and since rename/add vs. add/rename could possibly be important, redo the merge the opposite direction to make sure we don't have issues with the direction of the merge. These testcases failed before restoring the setting of path, but with the paths appropriately set the testcases both pass. Reported-by: Ben Humphreys <behumphreys@atlassian.com> Based-on-patch-by: SZEDER Gábor <szeder.dev@gmail.com> Tested-by: Ben Humphreys <behumphreys@atlassian.com> Signed-off-by: Elijah Newren <newren@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-03Merge branch 'cc/list-objects-filter-wo-sparse-path'Libravatar Junio C Hamano2-84/+26
Disable "--filter=sparse:path=<path>" that would allow reading from paths on the filesystem. * cc/list-objects-filter-wo-sparse-path: list-objects-filter: disable 'sparse:path' filters
2019-05-30Merge branch 'js/rebase-deprecate-preserve-merges'Libravatar Junio C Hamano2-7/+13
A bit more leftover clean-up to deprepcate "rebase -p". * js/rebase-deprecate-preserve-merges: rebase docs: recommend `-r` over `-p` docs: say that `--rebase=preserve` is deprecated tests: mark a couple more test cases as requiring `rebase -p`
2019-05-30Merge branch 'sg/trace2-rename'Libravatar Junio C Hamano5-32/+32
Rename environment variables that are used to control the "trace2" mechanism to a more readable name. * sg/trace2-rename: trace2: document the supported values of GIT_TRACE2* env variables trace2: rename environment variables to GIT_TRACE2*
2019-05-30Merge branch 'nd/diff-parseopt'Libravatar Junio C Hamano4-0/+94
A brown-paper-bag bugfix to a change already in 'master'. * nd/diff-parseopt: parse-options: check empty value in OPT_INTEGER and OPT_ABBREV diff-parseopt: restore -U (no argument) behavior diff-parseopt: correct variable types that are used by parseopt
2019-05-29list-objects-filter: disable 'sparse:path' filtersLibravatar Christian Couder2-84/+26
If someone wants to use as a filter a sparse file that is in the repository, something like "--filter=sparse:oid=<ref>:<path>" already works. So 'sparse:path' is only interesting if the sparse file is not in the repository. In this case though the current implementation has a big security issue, as it makes it possible to ask the server to read any file, like for example /etc/password, and to explore the filesystem, as well as individual lines of files. If someone is interested in using a sparse file that is not in the repository as a filter, then at the minimum a config option, such as "uploadpack.sparsePathFilter", should be implemented first to restrict the directory from which the files specified by 'sparse:path' can be read. For now though, let's just disable 'sparse:path' filters. Helped-by: Matthew DeVore <matvore@google.com> Helped-by: Jeff Hostetler <git@jeffhostetler.com> Signed-off-by: Christian Couder <chriscool@tuxfamily.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-29diff-parseopt: restore -U (no argument) behaviorLibravatar Nguyễn Thái Ngọc Duy4-0/+94
Before d473e2e0e8 (diff.c: convert -U|--unified, 2019-01-27), -U and --unified are implemented with a custom parser opt_arg() in diff.c. I didn't check this code carefully and not realize that it's the equivalent of PARSE_OPT_NONEG | PARSE_OPT_OPTARG. In other words, if -U is specified without any argument, the option should be accepted, and the default value should be used. Without PARSE_OPT_OPTARG, parse_options() will reject this case and cause a regression. Reported-by: Bryan Turner <bturner@atlassian.com> Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-29send-email: fix regression in sendemail.identity parsingLibravatar Ævar Arnfjörð Bjarmason1-0/+64
Fix a regression in my recent 3494dfd3ee ("send-email: do defaults -> config -> getopt in that order", 2019-05-09). I missed that the $identity variable needs to be extracted from the command-line before we do the config reading, as it determines which config variable we should read first. See [1] for the report. The sendemail.identity feature was added back in 34cc60ce2b ("send-email: Add support for SSL and SMTP-AUTH", 2007-09-03), there were no tests to assert that it worked properly. So let's fix both the regression, and add some tests to assert that this is being parsed properly. While I'm at it I'm adding a --no-identity option to go with --[to|cc|bcc] variable, since the semantics are similar. It's like to/cc/bcc except that unlike those we don't support multiple identities, but we could now easily add it support for it if anyone cares. In just fixing the --identity command-line parsing bug I discovered that a narrow fix to that wouldn't do. In read_config() we had a state machine that would only set config values if they weren't set already, and thus by proxy we wouldn't e.g. set "to" based on sendemail.to if we'd seen sendemail.gmail.to before, with --identity=gmail. I'd modified some of the relevant code in 3494dfd3ee, but just reverting to that wouldn't do, since it would bring back the regression fixed in that commit. Refactor read_config() do what we actually mean here. We don't want to set a given sendemail.VAR if a sendemail.$identity.VAR previously set it. The old code was conflating this desire with the hardcoded defaults for these variables, and as discussed in 3494dfd3ee that was never going to work. Instead pass along the state of whether an identity config set something before, as distinguished from the state of the default just being false, or the default being a non-bool or true (e.g. --transferencoding). I'm still not happy with the test coverage here, e.g. there's nothing testing sendemail.smtpEncryption, but I only have so much time to fix this code. 1. https://public-inbox.org/git/5cddeb61.1c69fb81.47ed4.e648@mx.google.com/ Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-28tests: mark a couple more test cases as requiring `rebase -p`Libravatar Johannes Schindelin2-7/+13
The `--preserve-merges` option has been deprecated, and as a consequence we started to mark test cases that require that option to be supported, in preparation for removing that support eventually. Since we marked those test cases, a couple more crept into the test suite, and with this patch, we mark them, too. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-28trace2: rename environment variables to GIT_TRACE2*Libravatar SZEDER Gábor5-32/+32
For an environment variable that is supposed to be set by users, the GIT_TR2* env vars are just too unclear, inconsistent, and ugly. Most of the established GIT_* environment variables don't use abbreviations, and in case of the few that do (GIT_DIR, GIT_COMMON_DIR, GIT_DIFF_OPTS) it's quite obvious what the abbreviations (DIR and OPTS) stand for. But what does TR stand for? Track, traditional, trailer, transaction, transfer, transformation, transition, translation, transplant, transport, traversal, tree, trigger, truncate, trust, or ...?! The trace2 facility, as the '2' suffix in its name suggests, is supposed to eventually supercede Git's original trace facility. It's reasonable to expect that the corresponding environment variables follow suit, and after the original GIT_TRACE variables they are called GIT_TRACE2; there is no such thing is 'GIT_TR'. All trace2-specific config variables are, very sensibly, in the 'trace2' section, not in 'tr2'. OTOH, we don't gain anything at all by omitting the last three characters of "trace" from the names of these environment variables. So let's rename all GIT_TR2* environment variables to GIT_TRACE2*, before they make their way into a stable release. Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-19Merge branch 'jk/get-oid-indexed-object-name'Libravatar Junio C Hamano1-6/+2
The codepath to parse :<path> that obtains the object name for an indexed object has been made more robust. * jk/get-oid-indexed-object-name: get_oid: handle NULL repo->index
2019-05-19Merge branch 'tz/test-lib-check-working-jgit'Libravatar Junio C Hamano1-1/+1
A prerequiste check in the test suite to see if a working jgit is available was made more robust. * tz/test-lib-check-working-jgit: test-lib: try harder to ensure a working jgit
2019-05-19Merge branch 'es/check-non-portable-pre-5.10'Libravatar Junio C Hamano1-4/+5
Developer support update. * es/check-non-portable-pre-5.10: check-non-portable-shell: support Perl versions older than 5.10
2019-05-19Merge branch 'js/fsmonitor-refresh-after-discarding-index'Libravatar Junio C Hamano2-1/+31
The fsmonitor interface got out of sync after the in-core index file gets discarded, which has been corrected. * js/fsmonitor-refresh-after-discarding-index: fsmonitor: force a refresh after the index was discarded fsmonitor: demonstrate that it is not refreshed after discard_index()
2019-05-19Merge branch 'js/t5580-unc-alternate-test'Libravatar Junio C Hamano1-0/+12
An additional test for MinGW * js/t5580-unc-alternate-test: t5580: verify that alternates can be UNC paths
2019-05-19Merge branch 'bl/t4253-exit-code-from-format-patch'Libravatar Junio C Hamano1-2/+4
Avoid patterns to pipe output from a git command to feed another command in tests. * bl/t4253-exit-code-from-format-patch: t4253-am-keep-cr-dos: avoid using pipes
2019-05-19Merge branch 'dl/difftool-mergetool'Libravatar Junio C Hamano2-61/+126
Update "git difftool" and "git mergetool" so that the combinations of {diff,merge}.{tool,guitool} configuration variables serve as fallback settings of each other in a sensible order. * dl/difftool-mergetool: difftool: fallback on merge.guitool difftool: make --gui, --tool and --extcmd mutually exclusive mergetool: fallback to tool when guitool unavailable mergetool--lib: create gui_mode function mergetool: use get_merge_tool function t7610: add mergetool --gui tests t7610: unsuppress output
2019-05-19Merge branch 'js/t6500-use-windows-pid-on-mingw'Libravatar Junio C Hamano1-1/+9
Future-proof a test against an update to MSYS2 runtime v3.x series. * js/t6500-use-windows-pid-on-mingw: t6500(mingw): use the Windows PID of the shell
2019-05-19Merge branch 'jk/apache-lsan'Libravatar Junio C Hamano1-0/+1
Allow tests that involve httpd to be run under leak sanitizer, just like we can already do so under address sanitizer. * jk/apache-lsan: t/lib-httpd: pass LSAN_OPTIONS through apache
2019-05-19Merge branch 'nd/parse-options-aliases'Libravatar Junio C Hamano2-0/+20
Attempt to use an abbreviated option in "git clone --recurs" is responded by a request to disambiguate between --recursive and --recurse-submodules, which is bad because these two are synonyms. The parse-options API has been extended to define such synonyms more easily and not produce an unnecessary failure. * nd/parse-options-aliases: parse-options: don't emit "ambiguous option" for aliases
2019-05-19Merge branch 'dl/branch-from-3dot-merge-base'Libravatar Junio C Hamano2-30/+40
"git branch new A...B" and "git checkout -b new A...B" have been taught that in their contexts, the notation A...B means "the merge base between these two commits", just like "git checkout A...B" detaches HEAD at that commit. * dl/branch-from-3dot-merge-base: branch: make create_branch accept a merge base rev t2018: cleanup in current test
2019-05-19Merge branch 'ab/perf-installed-fix'Libravatar Junio C Hamano4-27/+53
Performance test framework has been broken and measured the version of Git that happens to be on $PATH, not the specified one to measure, for a while, which has been corrected. * ab/perf-installed-fix: perf-lib.sh: forbid the use of GIT_TEST_INSTALLED perf tests: add "bindir" prefix to git tree test results perf-lib.sh: remove GIT_TEST_INSTALLED from perf-lib.sh perf-lib.sh: make "./run <revisions>" use the correct gits perf aggregate: remove GIT_TEST_INSTALLED from --codespeed perf README: correct docs for 3c8f12c96c regression
2019-05-19Merge branch 'dl/warn-tagging-a-tag'Libravatar Junio C Hamano1-1/+1
Typofix. * dl/warn-tagging-a-tag: tag: fix typo in nested tagging hint
2019-05-19send-email: fix broken transferEncoding testsLibravatar Ævar Arnfjörð Bjarmason1-24/+11
I fixed a bug that had broken the reading of sendmail.transferEncoding in 3494dfd3ee ("send-email: do defaults -> config -> getopt in that order", 2019-05-09), but the test I added in that commit did nothing to assert the bug had been fixed. That issue originates in 8d81408435 ("git-send-email: add --transfer-encoding option", 2014-11-25) which first added the "sendemail.transferencoding=8bit". That test has never done anything meaningful. It tested that the "--transfer-encoding=8bit" option would turn on the 8bit Transfer-Encoding, but that was the default at the time (and now). As checking out 8d81408435 and editing the test to remove that option will reveal, supplying it never did anything. So when I copied it thinking it would work in 3494dfd3ee I copied a previously broken test, although I was making sure it did the right thing via da-hoc debugger inspection, so the bug was fixed. So fix the test I added in 3494dfd3ee, as well as the long-standing test added in 8d81408435. To test if we're actually setting the Transfer-Encoding let's set it to 7bit, not 8bit, as 7bit will error out on "email-using-8bit". This means that we can remove the "sendemail.transferencoding=7bit fails on 8bit data" test, since it was redundant, we now have other tests that assert that that'll fail. While I'm at it convert "git config <key> <value>" in the test setup to just "-c <key>=<value>" on the command-line. Then we don't need to cleanup after these tests, and there's no sense in asserting where config values come from in these tests, we can take that as a given. Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-19send-email: remove cargo-culted multi-patch pattern in testsLibravatar Ævar Arnfjörð Bjarmason1-7/+7
Change test code added in f434c083a0 ("send-email: add --no-cc, --no-to, and --no-bcc", 2010-03-07) which blindly copied a pattern from an earlier test added in 32ae83194b ("add a test for git-send-email for non-threaded mails", 2009-06-12) where the "$patches" variable was supplied more than once. As it turns out we didn't need more than one "$patches" for the test added in 32ae83194b either. The only tests that actually needed this sort of invocation were the tests added in 54aae5e1a0 ("t9001: send-email interation with --in-reply-to and --chain-reply-to", 2010-10-19). Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-15test-lib: try harder to ensure a working jgitLibravatar Todd Zullinger1-1/+1
The JGIT prereq uses `type jgit` to determine whether jgit is present. While this is usually sufficient, it won't help if the jgit found is badly broken. This wastes time running tests which fail due to no fault of our own. Use `jgit --version` instead, to guard against cases where jgit is present on the system, but will fail to run, e.g. because of some JRE issue, or missing Java dependencies. Checking that it gets far enough to process the '--version' argument isn't perfect, but seems to be good enough in practice. It's also consistent with how we detect some other dependencies, see e.g. the CURL and UNZIP prerequisites. Signed-off-by: Todd Zullinger <tmz@pobox.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-15get_oid: handle NULL repo->indexLibravatar Jeff King1-6/+2
When get_oid() and its helpers see an index name like ":.gitmodules", they try to load the index on demand, like: if (repo->index->cache) repo_read_index(repo); However, that misses the case when "repo->index" itself is NULL; we'll segfault in the conditional. This never happens with the_repository; there we always point its index field to &the_index. But a submodule repository may have a NULL index field until somebody calls repo_read_index(). This bug is triggered by t7411, but it was hard to notice because it's in an expect_failure block. That test was added by 2b1257e463 (t/helper: add test-submodule-nested-repo-config, 2018-10-25). Back then we had no easy way to access the .gitmodules blob of a submodule repo, so we expected (and got) an error message to that effect. Later, d9b8b8f896 (submodule-config.c: use repo_get_oid for reading .gitmodules, 2019-04-16) started looking in the correct repo, which is when we started triggering the segfault. With this fix, the test starts passing (once we clean it up as its comment instructs). Note that as far as I know, this bug could not be triggered outside of the test suite. It requires resolving an index name in a submodule, and all of the code paths (aside from test-tool) which do that either load the index themselves, or always pass the_repository. Ultimately it comes from 3a7a698e93 (sha1-name.c: remove implicit dependency on the_index, 2019-01-12), which replaced a check of "the_index.cache" with "repo->index->cache". So even if there is another way to trigger it, it wouldn't affect any versions before then. Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13Merge branch 'pw/clean-sequencer-state-upon-final-commit'Libravatar Junio C Hamano2-0/+75
"git chery-pick" (and "revert" that shares the same runtime engine) that deals with multiple commits got confused when the final step gets stopped with a conflict and the user concluded the sequence with "git commit". Attempt to fix it by cleaning up the state files used by these commands in such a situation. * pw/clean-sequencer-state-upon-final-commit: fix cherry-pick/revert status after commit commit/reset: try to clean up sequencer state
2019-05-13Merge branch 'jk/perf-aggregate-wo-libjson'Libravatar Junio C Hamano1-2/+2
The script to aggregate perf result unconditionally depended on libjson-perl even though it did not have to, which has been corrected. * jk/perf-aggregate-wo-libjson: t/perf: depend on perl JSON only when using --codespeed
2019-05-13Merge branch 'jk/p5302-avoid-collision-check-cost'Libravatar Junio C Hamano1-13/+18
Fix index-pack perf test so that the repeated invocations always run in an empty repository, which emulates the initial clone situation better. * jk/p5302-avoid-collision-check-cost: p5302: create the repo in each index-pack test
2019-05-13Merge branch 'ew/repack-with-bitmaps-by-default'Libravatar Junio C Hamano4-7/+21
The connectivity bitmaps are created by default in bare repositories now; also the pathname hash-cache is created by default to avoid making crappy deltas when repacking. * ew/repack-with-bitmaps-by-default: pack-objects: default to writing bitmap hash-cache t5310: correctly remove bitmaps for jgit test repack: enable bitmaps by default on bare repos
2019-05-13Merge branch 'js/partial-clone-connectivity-check'Libravatar Junio C Hamano1-0/+26
During an initial "git clone --depth=..." partial clone, it is pointless to spend cycles for a large portion of the connectivity check that enumerates and skips promisor objects (which by definition is all objects fetched from the other side). This has been optimized out. * js/partial-clone-connectivity-check: t/perf: add perf script for partial clones clone: do faster object check for partial clones
2019-05-13Merge branch 'jh/trace2-sid-fix'Libravatar Junio C Hamano3-21/+113
Polishing of the new trace2 facility continues. The system-level configuration can specify site-wide trace2 settings, which can be overridden with per-user configuration and environment variables. * jh/trace2-sid-fix: trace2: fixup access problem on /etc/gitconfig in read_very_early_config trace2: update docs to describe system/global config settings trace2: make SIDs more unique trace2: clarify UTC datetime formatting trace2: report peak memory usage of the process trace2: use system/global config for default trace2 settings config: add read_very_early_config() trace2: find exec-dir before trace2 initialization trace2: add absolute elapsed time to start event trace2: refactor setting process starting time config: initialize opts structure in repo_read_config()
2019-05-13difftool: fallback on merge.guitoolLibravatar Denton Liu1-0/+16
In git-difftool.txt, it says 'git difftool' falls back to 'git mergetool' config variables when the difftool equivalents have not been defined. However, when `diff.guitool` is missing, it doesn't fallback to anything. Make git-difftool fallback to `merge.guitool` when `diff.guitool` is missing. Signed-off-by: Denton Liu <liu.denton@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13difftool: make --gui, --tool and --extcmd mutually exclusiveLibravatar Denton Liu1-0/+8
In git-difftool, these options specify which tool to ultimately run. As a result, they are logically conflicting. Explicitly disallow these options from being used together. Signed-off-by: Denton Liu <liu.denton@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13mergetool: fallback to tool when guitool unavailableLibravatar Denton Liu1-0/+19
In git-difftool, if the tool is called with --gui but `diff.guitool` is not set, it falls back to `diff.tool`. Make git-mergetool also fallback from `merge.guitool` to `merge.tool` if the former is undefined. If git-difftool, when called with `--gui`, were to use `get_configured_mergetool` in a future patch, it would also get the fallback behavior in the following precedence: 1. diff.guitool 2. merge.guitool 3. diff.tool 4. merge.tool The behavior for when difftool or mergetool are called without `--gui` should be identical with or without this patch. Note that the search loop could be written as sections="merge" keys="tool" if diff_mode then sections="diff $sections" fi if gui_mode then keys="guitool $keys" fi merge_tool=$( IFS=' ' for key in $keys do for section in $sections do selected=$(git config $section.$key) if test -n "$selected" then echo "$selected" return fi done done) which would make adding a mode in the future much easier. However, adding a new mode will likely never happen as it is highly discouraged so, as a result, it is written in its current form so that it is more readable for future readers. Signed-off-by: Denton Liu <liu.denton@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13send-email: do defaults -> config -> getopt in that orderLibravatar Ævar Arnfjörð Bjarmason1-1/+12
Change the git-send-email command-line argument parsing and config reading code to parse those two in the right order. I.e. first we set our hardcoded defaults, then we read our config, and finally we read the command-line, with later sets overriding earlier sets. This fixes a bug introduced in e67a228cd8 ("send-email: automatically determine transfer-encoding", 2018-07-08). That change broke the reading of sendmail.transferencoding because it wasn't careful to update the code to parse them in the previous "defaults -> getopt -> config" order. But as we can see from the history for this file doing it this way was never what we actually wanted, it's just something we grew organically as of 5483c71d7a ("git-send-email: make options easier to configure.", 2007-06-27) and have been dealing with the fallout since, e.g. in 463b0ea22b ("send-email: Fix %config_path_settings handling", 2011-10-14). As can be seen in this change the only place where we actually want to do something clever is with the to/cc/bcc variables, where setting them on the command-line (or using --no-{to,cc,bcc}) should clear out values we grab from the config. All the rest are things where the command-line should simply override the config values, and by reading the config first the config code doesn't need all this "let's not set it, if it was on the command-line" special-casing, as [1] shows we'd otherwise need to care about the difference between whether something was a default or present in config to fix the bug in e67a228cd8. 1. https://public-inbox.org/git/20190508105607.178244-2-gitster@pobox.com/ Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13init: make --template path relative to $CWDLibravatar Nguyễn Thái Ngọc Duy2-4/+4
During git-init we chdir() to the target directory, but --template is not adjusted. So it's relative to the target directory instead of current directory. It would be ok if it's documented, but --template in git-init.txt mentions nothing about this behavior. Change it to be relative to $CWD, which is much more intuitive. The changes in the test suite show that this relative-to-target behavior is actually used. I just hope that it's only used in the test suite and it's safe to change. Otherwise, the other option is just document it (i.e. relative to target dir) and move on. Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13check-non-portable-shell: support Perl versions older than 5.10Libravatar Eric Sunshine1-4/+5
For thoroughness when checking for one-shot environment variable assignments at shell function call sites, check-non-portable-shell stitches together incomplete lines (those ending with backslash). This allows it to correctly flag such undesirable usage even when the variable assignment and function call are split across lines, for example: FOO=bar \ func where 'func' is a shell function. The stitching is accomplished like this: while (<>) { chomp; # stitch together incomplete lines (those ending with "\") while (s/\\$//) { $_ .= readline; chomp; } # detect unportable/undesirable shell constructs ... } Although this implementation is well supported in reasonably modern Perl versions (5.10 and later), it fails with older versions (such as Perl 5.8 shipped with ancient Mac OS 10.5). In particular, in older Perl versions, 'readline' is not connected to the file handle associated with the "magic" while (<>) {...} construct, so 'readline' throws a "readline() on unopened filehandle" error. Work around this problem by dropping readline() and instead incorporating the stitching of incomplete lines directly into the existing while (<>) {...} loop. Helped-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Eric Sunshine <sunshine@sunshineco.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>