Граф коммитов

171 Коммитов

Автор SHA1 Сообщение Дата
Junio C Hamano 21fa5bb972 Merge branch 'rb/nonstop-config-mak-uname-update'
Build update.

* rb/nonstop-config-mak-uname-update:
  config.mak.uname: remove old NonStop compatibility settings
2020-12-18 15:15:18 -08:00
Randall S. Becker 731d578b4f config.mak.uname: remove old NonStop compatibility settings
The MKDIR_WO_TRAILING_SLASH and NO_SETITIMER options are no longer
needed on the NonStop platforms as both are now supported by the
oldest supported operating system revision.

Signed-off-by: Randall S. Becker <rsbecker@nexbridge.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2020-12-16 22:06:53 -08:00
Ævar Arnfjörð Bjarmason b990f02fd8 config.mak.uname: remove unused NEEDS_SSL_WITH_CURL flag
The NEEDS_SSL_WITH_CURL flag was still being set in one case, but
hasn't existed since 23c4bbe28e ("build: link with curl-defined
linker flags", 2018-11-03). Remove it, and a comment which referred to
it. See 6c109904bc ("Port to HP NonStop", 2012-09-19) for the initial
addition of the comment.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2020-11-11 12:46:08 -08:00
Ævar Arnfjörð Bjarmason a9c6123b64 config.mak.uname: remove unused the NO_R_TO_GCC_LINKER flag
The NO_R_TO_GCC_LINKER flag was still being on some platforms. It
hasn't been used since my 0f50c8e32c ("Makefile: remove the
NO_R_TO_GCC_LINKER flag", 2019-05-17).

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2020-11-11 12:46:06 -08:00
Junio C Hamano 7b304ab16c Merge branch 'cb/no-more-gmtime'
Code clean-up by removing a compatibility implementation of a
function we no longer use.

* cb/no-more-gmtime:
  compat: remove gmtime
2020-05-20 08:33:27 -07:00
Carlo Marcelo Arenas Belón 84b0115f0d compat: remove gmtime
ccd469450a (date.c: switch to reentrant {gm,local}time_r, 2019-11-28)
removes the only gmtime() call we had and moves to gmtime_r() which
doesn't have the same portability problems.

Remove the compat gmtime code since it is no longer needed, and confirm
by successfull running t4212 in FreeBSD 9.3 amd64 (the oldest I could
get a hold off).

Further work might be needed to ensure 32bit time_t systems (like FreeBSD
i386) will handle correctly the overflows tested in t4212, but that is
orthogonal to this change, and it doesn't change the current behaviour
as neither gmtime() or gmtime_r() will ever return NULL on those systems
because time_t is unsigned.

Signed-off-by: Carlo Marcelo Arenas Belón <carenas@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2020-05-14 13:52:27 -07:00
Junio C Hamano dd094c2b75 Merge branch 'es/bugreport'
The "bugreport" tool.

* es/bugreport:
  bugreport: drop extraneous includes
  bugreport: add compiler info
  bugreport: add uname info
  bugreport: gather git version and build info
  bugreport: add tool to generate debugging info
  help: move list_config_help to builtin/help
2020-05-01 13:39:59 -07:00
Junio C Hamano aabf3ea144 Merge branch 'ds/build-homebrew-gettext-fix'
Recent update to Homebrew used by macOS folks breaks build by
moving gettext library and necessary headers.

* ds/build-homebrew-gettext-fix:
  macOS/brew: let the build find gettext headers/libraries/msgfmt
2020-05-01 13:39:57 -07:00
Johannes Schindelin a0b3108618 macOS/brew: let the build find gettext headers/libraries/msgfmt
Apparently a recent Homebrew update now installs `gettext` into the
subdirectory /usr/local/opt/gettext/[lib/include].

Sometimes the ci job succeeds:
 brew link --force gettext
 Linking /usr/local/Cellar/gettext/0.20.1... 179 symlinks created

And sometimes installing the package "gettext" with force-link fails:
 brew link --force gettext
 Warning: Refusing to link macOS provided/shadowed software: gettext
 If you need to have gettext first in your PATH run:
  echo 'export PATH="/usr/local/opt/gettext/bin:$PATH"' >> ~/.bash_profile

(And the is not the final word either, since macOS itself says:
 The default interactive shell is now zsh.)

Anyway, The latter requires CFLAGS to include /usr/local/opt/gettext/include
and LDFLAGS to include /usr/local/opt/gettext/lib.

Likewise, the `msgfmt` tool is no longer in the `PATH`.

While it is unclear which change is responsible for this breakage (that
most notably only occurs on CI build agents that updated very recently),
https://github.com/Homebrew/homebrew-core/pull/53489 has fixed it.

Nevertheless, let's work around this issue, as there are still quite a
few build agents out there that need some help in this regard: we
explicitly do not call `brew update` in our CI/PR builds anymore.

Helped-by: Carlo Marcelo Arenas Belón <carenas@gmail.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2020-04-27 11:40:26 -07:00
Jessica Clarke 274a1328fb config.mak.uname: Define FREAD_READS_DIRECTORIES for GNU/Hurd
GNU/Hurd is another platform that behaves like this. Set it to
UnfortunatelyYes so that config directory files are correctly processed.
This fixes the corresponding 'proper error on directory "files"' test in
t1308-config-set.sh.

Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Jessica Clarke <jrtc27@jrtc27.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2020-04-22 10:19:28 -07:00
Emily Shaffer 709df95b78 help: move list_config_help to builtin/help
Starting in 3ac68a93fd, help.o began to depend on builtin/branch.o,
builtin/clean.o, and builtin/config.o. This meant that help.o was
unusable outside of the context of the main Git executable.

To make help.o usable by other commands again, move list_config_help()
into builtin/help.c (where it makes sense to assume other builtin libraries
are present).

When command-list.h is included but a member is not used, we start to
hear a compiler warning. Since the config list is generated in a fairly
different way than the command list, and since commands and config
options are semantically different, move the config list into its own
header and move the generator into its own script and build rule.

For reasons explained in 976aaedc (msvc: add a Makefile target to
pre-generate the Visual Studio solution, 2019-07-29), some build
artifacts we consider non-source files cannot be generated in the
Visual Studio environment, and we already have some Makefile tweaks
to help Visual Studio to use generated command-list.h header file.
Do the same to a new generated file, config-list.h, introduced by
this change.

Helped-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Emily Shaffer <emilyshaffer@google.com>
2020-04-16 15:22:16 -07:00
Johannes Schindelin 67af91c47a Sync with 2.23.1
* maint-2.23: (44 commits)
  Git 2.23.1
  Git 2.22.2
  Git 2.21.1
  mingw: sh arguments need quoting in more circumstances
  mingw: fix quoting of empty arguments for `sh`
  mingw: use MSYS2 quoting even when spawning shell scripts
  mingw: detect when MSYS2's sh is to be spawned more robustly
  t7415: drop v2.20.x-specific work-around
  Git 2.20.2
  t7415: adjust test for dubiously-nested submodule gitdirs for v2.20.x
  Git 2.19.3
  Git 2.18.2
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  ...
2019-12-06 16:31:39 +01:00
Johannes Schindelin 7fd9fd94fb Sync with 2.22.2
* maint-2.22: (43 commits)
  Git 2.22.2
  Git 2.21.1
  mingw: sh arguments need quoting in more circumstances
  mingw: fix quoting of empty arguments for `sh`
  mingw: use MSYS2 quoting even when spawning shell scripts
  mingw: detect when MSYS2's sh is to be spawned more robustly
  t7415: drop v2.20.x-specific work-around
  Git 2.20.2
  t7415: adjust test for dubiously-nested submodule gitdirs for v2.20.x
  Git 2.19.3
  Git 2.18.2
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  ...
2019-12-06 16:31:30 +01:00
Johannes Schindelin 5421ddd8d0 Sync with 2.21.1
* maint-2.21: (42 commits)
  Git 2.21.1
  mingw: sh arguments need quoting in more circumstances
  mingw: fix quoting of empty arguments for `sh`
  mingw: use MSYS2 quoting even when spawning shell scripts
  mingw: detect when MSYS2's sh is to be spawned more robustly
  t7415: drop v2.20.x-specific work-around
  Git 2.20.2
  t7415: adjust test for dubiously-nested submodule gitdirs for v2.20.x
  Git 2.19.3
  Git 2.18.2
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  ...
2019-12-06 16:31:23 +01:00
Johannes Schindelin fc346cb292 Sync with 2.20.2
* maint-2.20: (36 commits)
  Git 2.20.2
  t7415: adjust test for dubiously-nested submodule gitdirs for v2.20.x
  Git 2.19.3
  Git 2.18.2
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  t6130/t9350: prepare for stringent Win32 path validation
  quote-stress-test: allow skipping some trials
  quote-stress-test: accept arguments to test via the command-line
  tests: add a helper to stress test argument quoting
  mingw: fix quoting of arguments
  Disallow dubiously-nested submodule git directories
  ...
2019-12-06 16:31:12 +01:00
Johannes Schindelin d851d94151 Sync with 2.19.3
* maint-2.19: (34 commits)
  Git 2.19.3
  Git 2.18.2
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  t6130/t9350: prepare for stringent Win32 path validation
  quote-stress-test: allow skipping some trials
  quote-stress-test: accept arguments to test via the command-line
  tests: add a helper to stress test argument quoting
  mingw: fix quoting of arguments
  Disallow dubiously-nested submodule git directories
  protect_ntfs: turn on NTFS protection by default
  path: also guard `.gitmodules` against NTFS Alternate Data Streams
  ...
2019-12-06 16:30:49 +01:00
Johannes Schindelin 7c9fbda6e2 Sync with 2.18.2
* maint-2.18: (33 commits)
  Git 2.18.2
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  t6130/t9350: prepare for stringent Win32 path validation
  quote-stress-test: allow skipping some trials
  quote-stress-test: accept arguments to test via the command-line
  tests: add a helper to stress test argument quoting
  mingw: fix quoting of arguments
  Disallow dubiously-nested submodule git directories
  protect_ntfs: turn on NTFS protection by default
  path: also guard `.gitmodules` against NTFS Alternate Data Streams
  is_ntfs_dotgit(): speed it up
  ...
2019-12-06 16:30:38 +01:00
Johannes Schindelin 14af7ed5a9 Sync with 2.17.3
* maint-2.17: (32 commits)
  Git 2.17.3
  Git 2.16.6
  test-drop-caches: use `has_dos_drive_prefix()`
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  t6130/t9350: prepare for stringent Win32 path validation
  quote-stress-test: allow skipping some trials
  quote-stress-test: accept arguments to test via the command-line
  tests: add a helper to stress test argument quoting
  mingw: fix quoting of arguments
  Disallow dubiously-nested submodule git directories
  protect_ntfs: turn on NTFS protection by default
  path: also guard `.gitmodules` against NTFS Alternate Data Streams
  is_ntfs_dotgit(): speed it up
  mingw: disallow backslash characters in tree objects' file names
  ...
2019-12-06 16:29:15 +01:00
Johannes Schindelin 9ac92fed5b Sync with 2.15.4
* maint-2.15: (29 commits)
  Git 2.15.4
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  t6130/t9350: prepare for stringent Win32 path validation
  quote-stress-test: allow skipping some trials
  quote-stress-test: accept arguments to test via the command-line
  tests: add a helper to stress test argument quoting
  mingw: fix quoting of arguments
  Disallow dubiously-nested submodule git directories
  protect_ntfs: turn on NTFS protection by default
  path: also guard `.gitmodules` against NTFS Alternate Data Streams
  is_ntfs_dotgit(): speed it up
  mingw: disallow backslash characters in tree objects' file names
  path: safeguard `.git` against NTFS Alternate Streams Accesses
  clone --recurse-submodules: prevent name squatting on Windows
  is_ntfs_dotgit(): only verify the leading segment
  ...
2019-12-06 16:27:18 +01:00
Johannes Schindelin d3ac8c3f27 Sync with 2.14.6
* maint-2.14: (28 commits)
  Git 2.14.6
  mingw: handle `subst`-ed "DOS drives"
  mingw: refuse to access paths with trailing spaces or periods
  mingw: refuse to access paths with illegal characters
  unpack-trees: let merged_entry() pass through do_add_entry()'s errors
  quote-stress-test: offer to test quoting arguments for MSYS2 sh
  t6130/t9350: prepare for stringent Win32 path validation
  quote-stress-test: allow skipping some trials
  quote-stress-test: accept arguments to test via the command-line
  tests: add a helper to stress test argument quoting
  mingw: fix quoting of arguments
  Disallow dubiously-nested submodule git directories
  protect_ntfs: turn on NTFS protection by default
  path: also guard `.gitmodules` against NTFS Alternate Data Streams
  is_ntfs_dotgit(): speed it up
  mingw: disallow backslash characters in tree objects' file names
  path: safeguard `.git` against NTFS Alternate Streams Accesses
  clone --recurse-submodules: prevent name squatting on Windows
  is_ntfs_dotgit(): only verify the leading segment
  test-path-utils: offer to run a protectNTFS/protectHFS benchmark
  ...
2019-12-06 16:26:55 +01:00
Johannes Schindelin 9102f958ee protect_ntfs: turn on NTFS protection by default
Back in the DOS days, in the FAT file system, file names always
consisted of a base name of length 8 plus a file extension of length 3.
Shorter file names were simply padded with spaces to the full 8.3
format.

Later, the FAT file system was taught to support _also_ longer names,
with an 8.3 "short name" as primary file name. While at it, the same
facility allowed formerly illegal file names, such as `.git` (empty base
names were not allowed), which would have the "short name" `git~1`
associated with it.

For backwards-compatibility, NTFS supports alternative 8.3 short
filenames, too, even if starting with Windows Vista, they are only
generated on the system drive by default.

We addressed the problem that the `.git/` directory can _also_ be
accessed via `git~1/` (when short names are enabled) in 2b4c6efc82
(read-cache: optionally disallow NTFS .git variants, 2014-12-16), i.e.
since Git v1.9.5, by introducing the config setting `core.protectNTFS`
and enabling it by default on Windows.

In the meantime, Windows 10 introduced the "Windows Subsystem for Linux"
(short: WSL), i.e. a way to run Linux applications/distributions in a
thinly-isolated subsystem on Windows (giving rise to many a "2016 is the
Year of Linux on the Desktop" jokes). WSL is getting increasingly
popular, also due to the painless way Linux application can operate
directly ("natively") on files on Windows' file system: the Windows
drives are mounted automatically (e.g. `C:` as `/mnt/c/`).

Taken together, this means that we now have to enable the safe-guards of
Git v1.9.5 also in WSL: it is possible to access a `.git` directory
inside `/mnt/c/` via the 8.3 name `git~1` (unless short name generation
was disabled manually). Since regular Linux distributions run in WSL,
this means we have to enable `core.protectNTFS` at least on Linux, too.

To enable Services for Macintosh in Windows NT to store so-called
resource forks, NTFS introduced "Alternate Data Streams". Essentially,
these constitute additional metadata that are connected to (and copied
with) their associated files, and they are accessed via pseudo file
names of the form `filename:<stream-name>:<stream-type>`.

In a recent patch, we extended `core.protectNTFS` to also protect
against accesses via NTFS Alternate Data Streams, e.g. to prevent
contents of the `.git/` directory to be "tracked" via yet another
alternative file name.

While it is not possible (at least by default) to access files via NTFS
Alternate Data Streams from within WSL, the defaults on macOS when
mounting network shares via SMB _do_ allow accessing files and
directories in that way. Therefore, we need to enable `core.protectNTFS`
on macOS by default, too, and really, on any Operating System that can
mount network shares via SMB/CIFS.

A couple of approaches were considered for fixing this:

1. We could perform a dynamic NTFS check similar to the `core.symlinks`
   check in `init`/`clone`: instead of trying to create a symbolic link
   in the `.git/` directory, we could create a test file and try to
   access `.git/config` via 8.3 name and/or Alternate Data Stream.

2. We could simply "flip the switch" on `core.protectNTFS`, to make it
   "on by default".

The obvious downside of 1. is that it won't protect worktrees that were
clone with a vulnerable Git version already. We considered patching code
paths that check out files to check whether we're running on an NTFS
system dynamically and persist the result in the repository-local config
setting `core.protectNTFS`, but in the end decided that this solution
would be too fragile, and too involved.

The obvious downside of 2. is that everybody will have to "suffer" the
performance penalty incurred from calling `is_ntfs_dotgit()` on every
path, even in setups where.

After the recent work to accelerate `is_ntfs_dotgit()` in most cases,
it looks as if the time spent on validating ten million random
file names increases only negligibly (less than 20ms, well within the
standard deviation of ~50ms). Therefore the benefits outweigh the cost.

Another downside of this is that paths that might have been acceptable
previously now will be forbidden. Realistically, though, this is an
improvement because public Git hosters already would reject any `git
push` that contains such file names.

Note: There might be a similar problem mounting HFS+ on Linux. However,
this scenario has been considered unlikely and in light of the cost (in
the aforementioned benchmark, `core.protectHFS = true` increased the
time from ~440ms to ~610ms), it was decided _not_ to touch the default
of `core.protectHFS`.

This change addresses CVE-2019-1353.

Reported-by: Nicolas Joly <Nicolas.Joly@microsoft.com>
Helped-by: Garima Singh <garima.singh@microsoft.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
2019-12-05 15:36:51 +01:00
Johannes Schindelin 5d65ad17a9 vcxproj: include more generated files
In the CI builds, we bundle all generated files into a so-called
artifacts `.tar` file, so that the test phase can fan out into multiple
parallel builds.

This patch makes sure that all files are included in the `vcxproj`
target which are needed for that artifacts `.tar` file.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-10-06 09:07:44 +09:00
Johannes Schindelin 030a628b81 vcxproj: only copy `git-remote-http.exe` once it was built
In b18ae14a8f (vcxproj: also link-or-copy builtins, 2019-07-29), we
started to copy or hard-link the built-ins as a post-build step of the
`git` project.

At the same time, we tried to copy or hard-link `git-remote-http.exe`,
but it is quite possible that it was not built at that time.

Let's move that latter task into a post-install step of the
`git-remote-http` project instead.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-10-06 09:07:44 +09:00
Johannes Schindelin b18ae14a8f vcxproj: also link-or-copy builtins
The default location for `.exe` files linked by Visual Studio depends on
the mode (debug vs release) and the architecture. Meaning: after a full
build, there is a `git.exe` in the top-level directory, but none of the
built-ins are linked..

When running a test script in Git Bash, it therefore would pick up the
wrong, say, `git-receive-pack.exe`: the one installed at the same time
as the Git Bash.

Absolutely not what we want. We want to have confidence that our test
covers the MSVC-built Git executables, and not some random stuff.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-29 14:51:43 -07:00
Johannes Schindelin 976aaedca0 msvc: add a Makefile target to pre-generate the Visual Studio solution
The entire idea of generating the VS solution makes only sense if we
generate it via Continuous Integration; otherwise potential users would
still have to download the entire Git for Windows SDK.

If we pre-generate the Visual Studio solution, Git can be built entirely
within Visual Studio, and the test scripts can be run in a regular Git
for Windows (e.g. the Portable Git flavor, which does not include a full
GCC toolchain and therefore weighs only about a tenth of Git for
Windows' SDK).

So let's just add a target in the Makefile that can be used to generate
said solution; The generated files will then be committed so that they
can be pushed to a branch ready to check out by Visual Studio users.

To make things even more useful, we also generate and commit other files
that are required to run the test suite, such as templates and
bin-wrappers: with this, developers can run the test suite in a regular
Git Bash after building the solution in Visual Studio.

Note: for this build target, we do not actually need to initialize the
`vcpkg` system, so we don't.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-07-29 14:51:43 -07:00
Junio C Hamano 32749c35a4 Merge branch 'js/mingw-gcc-stack-protect'
Windows update.

* js/mingw-gcc-stack-protect:
  mingw: enable stack smashing protector
2019-07-11 15:16:47 -07:00
Johannes Schindelin a1c5e906c8 mingw: enable stack smashing protector
To reduce Git for Windows' attack surface, we started using the Address
Space Layout Randomization and Data Execution Prevention features in
ce6a158561 (mingw: enable DEP and ASLR, 2019-05-08).

To remove yet another attack vector, let's make use of gcc's stack
smashing protector that helps detect stack buffer overruns early.

Rather than using -fstack-protector, we use -fstack-protector-strong
because on Windows: The latter appears to strike a better balance
between the performance impact and the provided safety.

In a non-scientific test (time git log --grep=is -p), best of 5 timings
went from 23.009s to 22.997s, i.e. the performance impact was *well*
lost in the noise.

This fixes https://github.com/git-for-windows/git/issues/501

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-27 12:53:45 -07:00
Jeff Hostetler 556702f86c msvc: add a compile-time flag to allow detailed heap debugging
MS Visual C comes with a few neat features we can use to analyze the
heap consumption (i.e. leaks, max memory, etc).

With this patch, we introduce support via the build-time flag
`USE_MSVC_CRTDBG`.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-25 10:46:57 -07:00
Jeff Hostetler dce7d29551 msvc: support building Git using MS Visual C++
With this patch, Git can be built using the Microsoft toolchain, via:

	make MSVC=1 [DEBUG=1]

Third party libraries are built from source using the open source
"vcpkg" tool set. See https://github.com/Microsoft/vcpkg

On a first build, the vcpkg tools and the third party libraries are
automatically downloaded and built. DLLs for the third party libraries
are copied to the top-level (and t/helper) directory to facilitate
debugging. See compat/vcbuild/README.

A series of .bat files are invoked by the Makefile to find the location
of the installed version of Visual Studio and the associated compiler
tools (essentially replicating the environment setup performed by a
"Developer Command Prompt"). This should find the most recent VS2015 or
VS2017 installation. Output from these scripts are used by the Makefile
to define compiler and linker pathnames and -I and -L arguments.

The build produces .pdb files for both debug and release builds.

Note: This commit was squashed from an organic series of commits
developed between 2016 and 2018 in Git for Windows' `master` branch.
This combined commit eliminates the obsolete commits related to fetching
NuGet packages for third party libraries. It is difficult to use NuGet
packages for C/C++ sources because they may be built by earlier versions
of the MSVC compiler and have CRT version and linking issues.

Additionally, the C/C++ NuGet packages that we were using tended to not
be updated concurrently with the sources.  And in the case of cURL and
OpenSSL, this could expose us to security issues.

Helped-by: Yue Lin Ho <b8732003@student.nsysu.edu.tw>
Helped-by: Philip Oakley <philipoakley@iee.org>
Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-25 10:46:57 -07:00
Johannes Schindelin 963e1543ed msvc: fix dependencies of compat/msvc.c
The file compat/msvc.c includes compat/mingw.c, which means that we have
to recompile compat/msvc.o if compat/mingw.c changes.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-20 14:03:05 -07:00
Johannes Schindelin 396ff7547d mingw: replace mingw_startup() hack
Git for Windows has special code to retrieve the command-line parameters
(and even the environment) in UTF-16 encoding, so that they can be
converted to UTF-8. This is necessary because Git for Windows wants to
use UTF-8 encoded strings throughout its code, and the main() function
does not get the parameters in that encoding.

To do that, we used the __wgetmainargs() function, which is not even a
Win32 API function, but provided by the MINGW "runtime" instead.

Obviously, this method would not work with any compiler other than GCC,
and in preparation for compiling with Visual C++, we would like to avoid
precisely that.

Lucky us, there is a much more elegant way: we can simply implement the
UTF-16 variant of `main()`: `wmain()`.

To make that work, we need to link with -municode. The command-line
parameters are passed to `wmain()` encoded in UTF-16, as desired, and
this method also works with GCC, and also with Visual C++ after
adjusting the MSVC linker flags to force it to use `wmain()`.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-20 14:03:05 -07:00
Johannes Schindelin 6a1ce2ed3a mingw: fix a typo in the msysGit-specific section
The msysGit project (i.e. Git for Windows 1.x' SDK) is safely dead for
*years* already. This is probably the reason why nobody caught this typo
until Carlo Arenas spotted a copy-edited version of it nearby.

It is probably about time to rip out the remainders of msysGit/MSys1
support, but that can safely wait a bit longer, and we can at least fix
the typo for now.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-06-20 14:03:05 -07:00
Junio C Hamano f42bee7d51 Merge branch 'id/windows-dep-aslr'
Allow DEP and ASLR for Windows build to for security hardening.

* id/windows-dep-aslr:
  mingw: enable DEP and ASLR
  mingw: do not let ld strip relocations
2019-05-19 16:45:27 +09:00
Junio C Hamano 40bef4992e Merge branch 'cc/access-on-aix-workaround'
Workaround for standard-compliant but less-than-useful behaviour of
access(2) for the root user.

* cc/access-on-aix-workaround:
  git-compat-util: work around for access(X_OK) under root
2019-05-13 23:50:35 +09:00
İsmail Dönmez ce6a158561 mingw: enable DEP and ASLR
Enable DEP (Data Execution Prevention) and ASLR (Address Space Layout
Randomization) support. This applies to both 32bit and 64bit builds
and makes it substantially harder to exploit security holes in Git by
offering a much more unpredictable attack surface.

ASLR interferes with GDB's ability to set breakpoints. A similar issue
holds true when compiling with -O2 (in which case single-stepping is
messed up because GDB cannot map the code back to the original source
code properly). Therefore we simply enable ASLR only when an
optimization flag is present in the CFLAGS, using it as an indicator
that the developer does not want to debug in GDB anyway.

Signed-off-by: İsmail Dönmez <ismail@i10z.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13 10:55:20 +09:00
İsmail Dönmez 598b6c3a92 mingw: do not let ld strip relocations
This is the first step for enabling ASLR (Address Space Layout
Randomization) support. We want to enable ASLR for better protection
against exploiting security holes in Git: it makes it harder to attack
software by making code addresses unpredictable.

The problem fixed by this commit is that `ld.exe` seems to be stripping
relocations which in turn will break ASLR support. We just make sure
it's not stripping the main executable entry.

Signed-off-by: İsmail Dönmez <ismail@i10z.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-13 10:55:18 +09:00
Junio C Hamano 1ff440fb41 Merge branch 'cc/aix-has-fileno-as-a-macro'
AIX shared the same build issues with other BSDs around fileno(fp),
which has been corrected.

* cc/aix-has-fileno-as-a-macro:
  Makefile: use fileno macro work around on AIX
2019-05-09 00:37:28 +09:00
Junio C Hamano 70542df566 Merge branch 'ss/msvc-path-utils-fix'
An earlier update for MinGW and Cygwin accidentally broke MSVC build,
which has been fixed.

* ss/msvc-path-utils-fix:
  MSVC: include compat/win32/path-utils.h for MSVC, too, for real_path()
2019-05-09 00:37:26 +09:00
Clément Chigot 400caafb2b git-compat-util: work around for access(X_OK) under root
On AIX, access(X_OK) may succeed when run as root even if the
execution isn't possible. This behavior is allowed by POSIX
which says:

  ... for a process with appropriate privileges, an implementation
  may indicate success for X_OK even if execute permission is not
  granted to any user.

It can lead hook programs to have their execution refused:

   git commit -m content
   fatal: cannot exec '.git/hooks/pre-commit': Permission denied

Add NEED_ACCESS_ROOT_HANDLER in order to use an access helper function.
It checks with stat if any executable flags is set when the current user
is root.

Signed-off-by: Clément Chigot <clement.chigot@atos.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-04-25 17:49:44 +09:00
Clément Chigot ee662bf5c6 Makefile: use fileno macro work around on AIX
Declare FILENO_IS_A_MACRO on AIX

On AIX, fileno(fp) is a macro and need to use the work around already made for BSD's.

Signed-off-by: Clément Chigot <clement.chigot@atos.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-04-25 17:49:20 +09:00
Junio C Hamano aa8c8d914e Merge branch 'js/macos-gettext-build'
Build with gettext breaks on recent macOS w/ Homebrew when
/usr/local/bin is not on PATH, which has been corrected.

* js/macos-gettext-build:
  macOS: make sure that gettext is found
2019-04-25 16:41:25 +09:00
Johannes Schindelin 92a1377a2a macOS: make sure that gettext is found
Due to reasons (some XCode versions seem to include gettext, some
don't?), Homebrew does not expose the libraries and headers in
/usr/local/ by default anymore.

Let's help find them again.

Note: for some reason, this is a change of behavior caused by the
upgrade to Mojave, identified in our Azure Pipeline; it seems that
Homebrew used to add the /usr/local/ directories to the include and link
search path before, but now it no longer does.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-04-15 12:03:55 +09:00
Sven Strickroth 22c3634c0f MSVC: include compat/win32/path-utils.h for MSVC, too, for real_path()
A path such as 'c:/somepath/submodule/../.git/modules/submodule' wasn't
resolved correctly any more, because the *nix variant of offset_1st_component
is used instead of the Win32 specific version.

Regression was introduced in commit 1cadad6f6 when mingw_offset_1st_component
was moved from mingw.c which is included by msvc.c to a separate file. Then,
the new file "compat/win32/path-utils.h" was only included for the __CYGWIN__
and __MINGW32__ cases in git-compat-util.h, the case for _MSC_VER was missing.

Signed-off-by: Sven Strickroth <email@cs-ware.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-04-09 20:40:07 +09:00
Junio C Hamano be9f1c0224 Sync with maint
* maint:
  mingw: allow building with an MSYS2 runtime v3.x
2019-03-11 16:16:34 +09:00
Johannes Schindelin aeb582a983 mingw: allow building with an MSYS2 runtime v3.x
Recently the Git for Windows project started the upgrade process to
a MSYS2 runtime version based on Cygwin v3.x.

This has the very notable consequence that `$(uname -r)` no longer
reports a version starting with "2", but a version with "3".

That breaks our build, as df5218b4c3 (config.mak.uname: support MSys2,
2016-01-13) simply did not expect the version reported by `uname -r` to
depend on the underlying Cygwin version: it expected the reported
version to match the "2" in "MSYS2".

So let's invert that test case to test for *anything else* than a
version starting with "1" (for MSys). That should safeguard us for the
future, even if Cygwin ends up releasing versionsl like 314.272.65536.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-03-11 11:02:16 +09:00
Johannes Schindelin 6053c04b88 mingw: drop MakeMaker reference
In 20d2a30f8f (Makefile: replace perl/Makefile.PL with simple make
rules, 2017-12-10), Git stopped using MakeMaker. Therefore, that
definition in the MINGW-specific section became useless.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-03-08 10:42:10 +09:00
Jeff Hostetler 353d3d77f4 trace2: collect Windows-specific process information
Add platform-specific interface to log information about the current
process.

On Windows, this interface is used to indicate whether the git process
is running under a debugger and list names of the process ancestors.

Information for other platforms is left for a future effort.

Signed-off-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-02-22 15:27:59 -08:00
Junio C Hamano f26493b787 Merge branch 'rs/bash-is-in-coreutils-on-nonstop'
* rs/bash-is-in-coreutils-on-nonstop:
  config.mak.uname: move location of bash on NonStop to CoreUtils
2019-02-13 18:18:43 -08:00
Junio C Hamano 8593e8a618 Merge branch 'js/mingw-host-cpu'
Windows update.

* js/mingw-host-cpu:
  mingw: use a more canonical method to fix the CPU reporting
2019-02-13 18:18:43 -08:00
Junio C Hamano 1db999ce8d Merge branch 'nd/fileno-may-be-macro'
* nd/fileno-may-be-macro:
  git-compat-util: work around fileno(fp) that is a macro
2019-02-13 18:18:41 -08:00