nihilist/tor
1
0
Fork 0
mirror of https://gitlab.torproject.org/tpo/core/tor.git synced 2024-11-30 15:43:32 +01:00
Code Issues Packages Projects Releases Wiki Activity
825aa705b1
tor/configure.ac

2839 lines
92 KiB
Plaintext
Raw Normal View History

Happy new year! svn:r5949
2006-02-09 06:46:49 +01:00
dnl Copyright (c) 2001-2004, Roger Dingledine
Update some copyright notices: it is now 2008. svn:r13412
2008-02-07 06:31:47 +01:00
dnl Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson
Bump copyright date to 2019
2019-01-16 18:33:22 +01:00
dnl Copyright (c) 2007-2019, The Tor Project, Inc.
update copyright notices. svn:r3982
2005-04-01 22:15:56 +02:00
dnl See LICENSE for licensing information
added automake/autoconf support. When in doubt, "aclocal && autoconf && autoheader && automake" from the top dir. svn:r10
2002-06-29 01:26:42 +02:00
Document required autotools versions CentOS 6 is roughly the oldest thing we care about developers still using, and it has autoconf 2.63 / automake 1.11. These are both older than openssl 1.0.0, so anybody who can't upgrade past those probably can't upgrade to a modern openssl either. And since only people building from git or editing configure.ac/Makefile.am need to use autotools, I'm not totally enthused about keeping support for old ones anyway. Closes ticket 17732.
2016-02-23 17:08:04 +01:00
AC_PREREQ([2.63])
version: Bump version to 0.4.8.8-dev
2023-11-03 14:50:01 +01:00
AC_INIT([tor],[0.4.8.8-dev])
Fix everything that previously referred to src/or
2018-07-05 23:05:54 +02:00
AC_CONFIG_SRCDIR([src/app/main/tor_main.c])
Use a nicely written autoconf macro to determine the sign of a type This beats our old implementation, which wouldn't work when cross-compiling
2013-02-07 22:23:48 +01:00
AC_CONFIG_MACRO_DIR([m4])
Move Automake options to configure.ac This will centralize the version configuration in one file to simplify future changes. Also fixes some typos in the Automake options comments.
2016-02-26 13:53:28 +01:00
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
configure_flags="$*"
Add the compiler name, version and libs used to compile This changes the behaviour of `tor --version` in such a way. ```console src/app/tor --version Tor version 0.4.5.1-alpha-dev (git-46ccde66a97d7985). Tor is running on Linux with Libevent 2.1.12-stable, OpenSSL 1.1.1h, Zlib 1.2.11, Liblzma 5.2.4, Libzstd 1.4.5 and Glibc 2.31 as libc. Tor compiled with GCC version 10.2.0 ``` Fixes #32102
2020-11-17 10:43:00 +01:00
AC_DEFINE_UNQUOTED([CONFIG_FLAGS], ["$configure_flags"], [Flags passed to configure])
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
Rewrite updateVersions script in Python, add datestamp functionality. This updateVersions.pl script was one of the only essential perl scripts left in out maint system, and was the only one that used autoconf to fill in the script. This script adds a feature to define an APPROX_RELEASE_DATE macro that is updated when the version changes. We'll use this to implement prop297, so that we have an accurate view of when a release date happens.
2018-11-25 02:03:48 +01:00
# DO NOT EDIT THIS DEFINITION BY HAND UNLESS YOU KNOW WHAT YOU'RE DOING.
#
# The update_versions.py script updates this definition when the
# version number changes. Tor uses it to make sure that it
# only shuts down for missing "required protocols" when those protocols
# are listed as required by a consensus after this date.
version: Bump version to 0.4.8.8-dev
2023-11-03 14:50:01 +01:00
AC_DEFINE(APPROX_RELEASE_DATE, ["2023-11-03"], # for 0.4.8.8-dev
Rewrite updateVersions script in Python, add datestamp functionality. This updateVersions.pl script was one of the only essential perl scripts left in out maint system, and was the only one that used autoconf to fill in the script. This script adds a feature to define an APPROX_RELEASE_DATE macro that is updated when the version changes. We'll use this to implement prop297, so that we have an accurate view of when a release date happens.
2018-11-25 02:03:48 +01:00
[Approximate date when this software was released. (Updated when the version changes.)])
Move Automake options to configure.ac This will centralize the version configuration in one file to simplify future changes. Also fixes some typos in the Automake options comments.
2016-02-26 13:53:28 +01:00
# "foreign" means we don't follow GNU package layout standards
# "1.11" means we require automake version 1.11 or newer
# "subdir-objects" means put .o files in the same directory as the .c files
Report errors when updating configuration files
2016-12-15 18:02:30 +01:00
AM_INIT_AUTOMAKE([foreign 1.11 subdir-objects -Wall -Werror])
Move Automake options to configure.ac This will centralize the version configuration in one file to simplify future changes. Also fixes some typos in the Automake options comments.
2016-02-26 13:53:28 +01:00
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_ac_n_warnings=0
tor_incr_n_warnings() {
tor_ac_n_warnings=`expr $tor_ac_n_warnings + 1`
}
Enable the automake silent rules. $ make V=1 # will temporarily disable them otherwise you see: CC foo.c rather than the giant long bulid line. This makes it significantly easier to spot compiler warnings etc. Additionally, make them conditional, so we won't error on automake < 1.11 (commits squashed by nickm.)
2012-08-02 16:01:25 +02:00
m4_ifdef([AM_SILENT_RULES], [AM_SILENT_RULES([yes])])
s/AM_CONFIG_HEADER/AC_CONFIG_HEADERS/ Apparently the former will stop working with newer auto* (commit message by nickm)
2012-08-23 18:35:21 +02:00
AC_CONFIG_HEADERS([orconfig.h])
added automake/autoconf support. When in doubt, "aclocal && autoconf && autoheader && automake" from the top dir. svn:r10
2002-06-29 01:26:42 +02:00
Use the Autoconf macro AC_USE_SYSTEM_EXTENSIONS The Autoconf macro AC_USE_SYSTEM_EXTENSIONS defines preprocessor macros which turn on extensions to C and POSIX. The macro also makes it easier for developers to use the extensions without needing (or forgetting) to define them manually. The macro can be safely used because it was introduced in Autoconf 2.60 and Tor requires Autoconf 2.63 and above.
2016-06-17 15:43:25 +02:00
AC_USE_SYSTEM_EXTENSIONS
Override unaligned-access-ok check when $host_cpu is ia64. Apparently, ia64-linux reacts to unaligned access by making the kernel gripe. Re-run autogen.sh if you need config.guess/sub. svn:r1882
2004-05-17 22:53:04 +02:00
AC_CANONICAL_HOST
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
PKG_PROG_PKG_CONFIG
configure: Give a more useful message when pkg-config fails When pkg-config is not installed, or a library that depends on pkg-config is not found, tell the user what to do to fix the problem. Fixes bug 31922; bugfix on 0.3.1.1-alpha.
2019-10-03 08:31:20 +02:00
if test "x$PKG_CONFIG" = "x" ; then
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
pkg_config_user_action="install pkg-config, and check the PKG_CONFIG_PATH environment variable"
configure: Give a more useful message when pkg-config fails When pkg-config is not installed, or a library that depends on pkg-config is not found, tell the user what to do to fix the problem. Fixes bug 31922; bugfix on 0.3.1.1-alpha.
2019-10-03 08:31:20 +02:00
AC_MSG_NOTICE([Some libraries need pkg-config, including systemd, nss, lzma, zstd, and custom mallocs.])
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
AC_MSG_NOTICE([To use those libraries, $pkg_config_user_action.])
configure: Give a more useful message when pkg-config fails When pkg-config is not installed, or a library that depends on pkg-config is not found, tell the user what to do to fix the problem. Fixes bug 31922; bugfix on 0.3.1.1-alpha.
2019-10-03 08:31:20 +02:00
else
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
pkg_config_user_action="check the PKG_CONFIG_PATH environment variable"
configure: Give a more useful message when pkg-config fails When pkg-config is not installed, or a library that depends on pkg-config is not found, tell the user what to do to fix the problem. Fixes bug 31922; bugfix on 0.3.1.1-alpha.
2019-10-03 08:31:20 +02:00
fi
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
force pkg-config to only use --prefix when cross-compiling The current pkg-config setup has no sense of whether it is cross-compiling, so it will detect things on the build system that are not present or are wrong for the host system. This forces the cross-compiling build to only look for pkg-config .pc files in --prefix. A version of this has been the setup for many years with the Android builds. Fixes #32191 Signed-off-by: Hans-Christoph Steiner <hans@eds.org>
2019-10-29 21:13:56 +01:00
if test "x$PKG_CONFIG_PATH" = "x" && test "x$prefix" != "xNONE" && test "$host" != "$build"; then
export PKG_CONFIG_PATH=$prefix/lib/pkgconfig
AC_MSG_NOTICE([set PKG_CONFIG_PATH=$PKG_CONFIG_PATH to support cross-compiling])
fi
configure: Add --enable-gpl option This change on its own doesn't use the option for anything, but it includes support for configure and a message in 'tor --version' Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-08 18:08:23 +01:00
# License options
AC_ARG_ENABLE(gpl,
AS_HELP_STRING(--enable-gpl, [allow the inclusion of GPL-licensed code, building a version of tor and libtor covered by the GPL rather than its usual 3-clause BSD license]))
license_option=BSD
AS_IF([test "x$enable_gpl" = xyes],
[
AC_DEFINE(ENABLE_GPL, 1, [Defined if tor is building in GPL-licensed mode.])
license_option=GPL
])
# Optional features
r14238@tombo: nickm | 2008-02-17 14:03:44 -0500 Add openbsd memory allocator discussed in bug 468, to make it easier for linux users to get non-awful allocation patterns. Use --enable-openbsd-malloc to turn it on. Needs more testing. svn:r13544
2008-02-17 20:03:49 +01:00
AC_ARG_ENABLE(openbsd-malloc,
Mark openbsd-malloc as deprecated
2018-08-03 13:38:47 +02:00
AS_HELP_STRING(--enable-openbsd-malloc, [use malloc code from OpenBSD. Linux only. Deprecated: see --with-malloc]))
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
AC_ARG_ENABLE(static-openssl,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-static-openssl, [link against a static openssl library. Requires --with-openssl-dir]))
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
AC_ARG_ENABLE(static-libevent,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-static-libevent, [link against a static libevent library. Requires --with-libevent-dir]))
Add --enable-static-zlib option Works like the --enable-static-openssl/libevent options. Requires --with-zlib-dir to be set. Note that other dependencies might still pull in a dynamicly linked zlib, if you don't link them in statically too.
2010-04-14 16:48:05 +02:00
AC_ARG_ENABLE(static-zlib,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-static-zlib, [link against a static zlib library. Requires --with-zlib-dir]))
add --enable-static-tor to our configure script This implements the feature request in bug #2702
2011-03-10 10:22:32 +01:00
AC_ARG_ENABLE(static-tor,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-static-tor, [create an entirely static Tor binary. Requires --with-openssl-dir and --with-libevent-dir and --with-zlib-dir]))
Start work on fancy compiler tricks to expose extra stuff to our tests This is mainly a matter of automake trickery: we build each static library in two versions now: one with the TOR_UNIT_TESTS macro defined, and one without. When TOR_UNIT_TESTS is defined, we can enable mocking and expose more functions. When it's not defined, we can lock the binary down more. The alternatives would be to have alternate build modes: a "testing configuration" for building the libraries with test support, and a "production configuration" for building them without. I don't favor that approach, since I think it would mean more people runnning binaries build for testing, or more people not running unit tests.
2013-06-06 20:56:05 +02:00
AC_ARG_ENABLE(unittests,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-unittests, [don't build unit tests for Tor. Risky!]))
Coverage support: build with --enable-coverage to have tests run with gcov If you pass the --enable-coverage flag on the command line, we build our testing binaries with appropriate options eo enable coverage testing. We also build a "tor-cov" binary that has coverage enabled, for integration tests. On recent OSX versions, test coverage only works with clang, not gcc. So we warn about that. Also add a contrib/coverage script to actually run gcov with the appropriate options to generate useful .gcov files. (Thanks to automake, the .o files will not have the names that gcov expects to find.) Also, remove generated gcda and gcno files on clean.
2013-07-09 18:37:11 +02:00
AC_ARG_ENABLE(coverage,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-coverage, [enable coverage support in the unit-test build]))
Make the no-assertions-during-coverage check into a configure option Closes ticket 18242. The rationale here is that I like having coverage on by default in my own working directory, but I always want assertions turned on unless I'm doing branch coverage specifically.
2016-02-04 18:51:52 +01:00
AC_ARG_ENABLE(asserts-in-tests,
AS_HELP_STRING(--disable-asserts-in-tests, [disable tor_assert() calls in the unit tests, for branch coverage]))
New option "--disable-system-torrc" to not read torrc from etc Implements 13037.
2015-01-06 23:07:01 +01:00
AC_ARG_ENABLE(system-torrc,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-system-torrc, [don't look for a system-wide torrc file]))
Add libfuzzer support.
2016-12-19 21:17:42 +01:00
AC_ARG_ENABLE(libfuzzer,
libfuzzer tweaks per recommendations
2017-01-09 16:44:55 +01:00
AS_HELP_STRING(--enable-libfuzzer, [build extra fuzzers based on 'libfuzzer']))
AC_ARG_ENABLE(oss-fuzz,
AS_HELP_STRING(--enable-oss-fuzz, [build extra fuzzers based on 'oss-fuzz' environment]))
Code to disable memory sentinels for fuzzing This feature makes it possible to turn off memory sentinels (like those used for safety in buffers.c and memarea.c) when fuzzing, so that we can catch bugs that they would otherwise prevent.
2017-02-13 15:10:11 +01:00
AC_ARG_ENABLE(memory-sentinels,
AS_HELP_STRING(--disable-memory-sentinels, [disable code that tries to prevent some kinds of memory access bugs. For fuzzing only.]))
Add a mostly disabled feature to debug restarting in-process For 23847, we want Tor to be able to shut down and then restart in the same process. Here's a patch to make the Tor binary do that. To test it, you need to build with --enable-restart-debugging, and then you need to set the environment variable TOR_DEBUG_RESTART. With this option, Tor will then run for 5 seconds, then restart itself in-process without exiting. This only happens once. You can change the 5-second interval using TOR_DEBUG_RESTART_AFTER_SECONDS. Implements ticket 24583.
2017-12-11 16:57:00 +01:00
AC_ARG_ENABLE(restart-debugging,
AS_HELP_STRING(--enable-restart-debugging, [Build Tor with support for debugging in-process restart. Developers only.]))
Add configure option to control ZSTD_STATIC_LINKING_ONLY
2018-02-06 17:58:05 +01:00
AC_ARG_ENABLE(zstd-advanced-apis,
AS_HELP_STRING(--disable-zstd-advanced-apis, [Build without support for zstd's "static-only" APIs.]))
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
AC_ARG_ENABLE(nss,
AS_HELP_STRING(--enable-nss, [Use Mozilla's NSS TLS library. (EXPERIMENTAL)]))
Add an --enable-pic option for people who hack Tor into libraries Closes ticket 23846
2018-09-12 16:10:29 +02:00
AC_ARG_ENABLE(pic,
AS_HELP_STRING(--enable-pic, [Build Tor's binaries as position-independent code, suitable to link as a library.]))
Start work on fancy compiler tricks to expose extra stuff to our tests This is mainly a matter of automake trickery: we build each static library in two versions now: one with the TOR_UNIT_TESTS macro defined, and one without. When TOR_UNIT_TESTS is defined, we can enable mocking and expose more functions. When it's not defined, we can lock the binary down more. The alternatives would be to have alternate build modes: a "testing configuration" for building the libraries with test support, and a "production configuration" for building them without. I don't favor that approach, since I think it would mean more people runnning binaries build for testing, or more people not running unit tests.
2013-06-06 20:56:05 +02:00
Re-think our doxygen warnings wrt missing docs and fatalness. Up till now, we have warned about all missing documentation, which meant that we could never make doxygen warnings fatal. This has led to our doxygen output getting full of errors several times in the past. This commit changes our approach to doxygen warnings: missing documentation warnings are only on when the user asks for them with --enable-missing-doc-warnings. When that option is not present, doxygen respects the --enable-fatal-warnings flag. Closes ticket 32385.
2019-11-05 18:24:02 +01:00
AC_ARG_ENABLE(missing-doc-warnings,
Clarify that "missing-doc-warnings" makes doxygen warnings nonfatal
2019-11-06 22:59:41 +01:00
AS_HELP_STRING(--enable-missing-doc-warnings, [Tell doxygen to warn about missing documentation. Makes doxygen warnings nonfatal.]))
Re-think our doxygen warnings wrt missing docs and fatalness. Up till now, we have warned about all missing documentation, which meant that we could never make doxygen warnings fatal. This has led to our doxygen output getting full of errors several times in the past. This commit changes our approach to doxygen warnings: missing documentation warnings are only on when the user asks for them with --enable-missing-doc-warnings. When that option is not present, doxygen respects the --enable-fatal-warnings flag. Closes ticket 32385.
2019-11-05 18:24:02 +01:00
if test "$enable_missing_doc_warnings" = "yes"; then
DOXYGEN_FATAL_WARNINGS=NO
DOXYGEN_WARN_ON_MISSING=YES
elif test "$enable_fatal_warnings" = "yes"; then
# Fatal warnings from doxygen are nice, but not if we're warning about
# missing documentation.
DOXYGEN_FATAL_WARNINGS=YES
DOXYGEN_WARN_ON_MISSING=NO
else
DOXYGEN_FATAL_WARNINGS=NO
DOXYGEN_WARN_ON_MISSING=NO
fi
AC_SUBST(DOXYGEN_FATAL_WARNINGS)
AC_SUBST(DOXYGEN_WARN_ON_MISSING)
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$enable_coverage" != "xyes" -a "x$enable_asserts_in_tests" = "xno" ; then
Fix the --disable-asserts-in-tests configure option
2016-02-05 14:40:07 +01:00
AC_MSG_ERROR([Can't disable assertions outside of coverage build])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
AM_CONDITIONAL(UNITTESTS_ENABLED, test "x$enable_unittests" != "xno")
AM_CONDITIONAL(COVERAGE_ENABLED, test "x$enable_coverage" = "xyes")
AM_CONDITIONAL(DISABLE_ASSERTS_IN_UNIT_TESTS, test "x$enable_asserts_in_tests" = "xno")
Add libfuzzer support.
2016-12-19 21:17:42 +01:00
AM_CONDITIONAL(LIBFUZZER_ENABLED, test "x$enable_libfuzzer" = "xyes")
libfuzzer tweaks per recommendations
2017-01-09 16:44:55 +01:00
AM_CONDITIONAL(OSS_FUZZ_ENABLED, test "x$enable_oss_fuzz" = "xyes")
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
AM_CONDITIONAL(USE_NSS, test "x$enable_nss" = "xyes")
When enabling NSS, disable OpenSSL. We used to link both libraries at once, but now that I'm working on TLS, there's nothing left to keep OpenSSL around for when NSS is enabled. Note that this patch causes a couple of places that still assumed OpenSSL to be disabled when NSS is enabled - tor-gencert - pbkdf2
2018-08-12 23:18:41 +02:00
AM_CONDITIONAL(USE_OPENSSL, test "x$enable_nss" != "xyes")
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
configure.ac: Define ENABLE_COVERAGE macro Allow conditional compilation based on whether the "--enable-coverage" configure option was specified.
2021-09-23 15:21:13 +02:00
if test "x$enable_coverage" = "xyes"; then
AC_DEFINE(ENABLE_COVERAGE, 1,
[Defined if coverage support is enabled for the unit tests])
fi
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
if test "x$enable_nss" = "xyes"; then
AC_DEFINE(ENABLE_NSS, 1,
Fix documentation of ENABLE_NSS
2018-09-04 15:51:39 +02:00
[Defined if we're building with NSS.])
When enabling NSS, disable OpenSSL. We used to link both libraries at once, but now that I'm working on TLS, there's nothing left to keep OpenSSL around for when NSS is enabled. Note that this patch causes a couple of places that still assumed OpenSSL to be disabled when NSS is enabled - tor-gencert - pbkdf2
2018-08-12 23:18:41 +02:00
else
AC_DEFINE(ENABLE_OPENSSL, 1,
[Defined if we're building with OpenSSL or LibreSSL])
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
fi
Coverage support: build with --enable-coverage to have tests run with gcov If you pass the --enable-coverage flag on the command line, we build our testing binaries with appropriate options eo enable coverage testing. We also build a "tor-cov" binary that has coverage enabled, for integration tests. On recent OSX versions, test coverage only works with clang, not gcc. So we warn about that. Also add a contrib/coverage script to actually run gcov with the appropriate options to generate useful .gcov files. (Thanks to automake, the .o files will not have the names that gcov expects to find.) Also, remove generated gcda and gcno files on clean.
2013-07-09 18:37:11 +02:00
add --enable-static-tor to our configure script This implements the feature request in bug #2702
2011-03-10 10:22:32 +01:00
if test "$enable_static_tor" = "yes"; then
enable_static_libevent="yes";
enable_static_openssl="yes";
enable_static_zlib="yes";
configure: Fix the --enable-static-tor switch The "-static" compile flag was set globally which means that all autoconf test were attempting to be built statically and lead to failures of detecting OpenSSL libraries and others. This commit adds this flag only to the "tor" binary build. There is also a fix on where to find libevent.a since it is using libtool, it is in .libs/. At this commit, there are still warnings being emitted that informs the user that the built binary must still be linked dynamically with glibc. Fixes #40111 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-11-18 17:24:38 +01:00
TOR_STATIC_LDFLAGS="-static"
add --enable-static-tor to our configure script This implements the feature request in bug #2702
2011-03-10 10:22:32 +01:00
fi
configure: Fix the --enable-static-tor switch The "-static" compile flag was set globally which means that all autoconf test were attempting to be built statically and lead to failures of detecting OpenSSL libraries and others. This commit adds this flag only to the "tor" binary build. There is also a fix on where to find libevent.a since it is using libtool, it is in .libs/. At this commit, there are still warnings being emitted that informs the user that the built binary must still be linked dynamically with glibc. Fixes #40111 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-11-18 17:24:38 +01:00
AC_SUBST(TOR_STATIC_LDFLAGS)
r14238@tombo: nickm | 2008-02-17 14:03:44 -0500 Add openbsd memory allocator discussed in bug 468, to make it easier for linux users to get non-awful allocation patterns. Use --enable-openbsd-malloc to turn it on. Needs more testing. svn:r13544
2008-02-17 20:03:49 +01:00
New option "--disable-system-torrc" to not read torrc from etc Implements 13037.
2015-01-06 23:07:01 +01:00
if test "$enable_system_torrc" = "no"; then
AC_DEFINE(DISABLE_SYSTEM_TORRC, 1,
[Defined if we're not going to look for a torrc in SYSCONF])
fi
Code to disable memory sentinels for fuzzing This feature makes it possible to turn off memory sentinels (like those used for safety in buffers.c and memarea.c) when fuzzing, so that we can catch bugs that they would otherwise prevent.
2017-02-13 15:10:11 +01:00
if test "$enable_memory_sentinels" = "no"; then
AC_DEFINE(DISABLE_MEMORY_SENTINELS, 1,
[Defined if we're turning off memory safety code to look for bugs])
fi
Allow excluding documentation from the build
2019-06-16 19:18:30 +02:00
AC_ARG_ENABLE(manpage,
AS_HELP_STRING(--disable-manpage, [Disable manpage generation.]))
AC_ARG_ENABLE(html-manual,
AS_HELP_STRING(--disable-html-manual, [Disable HTML documentation.]))
Add configure switch to disable use of asciidoc Also break the build if that switch isn't used and asciidoc isn't available.
2010-03-01 03:17:48 +01:00
AC_ARG_ENABLE(asciidoc,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-asciidoc, [don't use asciidoc (disables building of manpages)]),
Add configure switch to disable use of asciidoc Also break the build if that switch isn't used and asciidoc isn't available.
2010-03-01 03:17:48 +01:00
[case "${enableval}" in
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
"yes") asciidoc=true ;;
"no") asciidoc=false ;;
Add configure switch to disable use of asciidoc Also break the build if that switch isn't used and asciidoc isn't available.
2010-03-01 03:17:48 +01:00
*) AC_MSG_ERROR(bad value for --disable-asciidoc) ;;
esac], [asciidoc=true])
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
# systemd notify support
AC_ARG_ENABLE(systemd,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-systemd, [enable systemd notification support]),
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
[case "${enableval}" in
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
"yes") systemd=true ;;
"no") systemd=false ;;
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
* ) AC_MSG_ERROR(bad value for --enable-systemd) ;;
esac], [systemd=auto])
Add a mostly disabled feature to debug restarting in-process For 23847, we want Tor to be able to shut down and then restart in the same process. Here's a patch to make the Tor binary do that. To test it, you need to build with --enable-restart-debugging, and then you need to set the environment variable TOR_DEBUG_RESTART. With this option, Tor will then run for 5 seconds, then restart itself in-process without exiting. This only happens once. You can change the 5-second interval using TOR_DEBUG_RESTART_AFTER_SECONDS. Implements ticket 24583.
2017-12-11 16:57:00 +01:00
if test "$enable_restart_debugging" = "yes"; then
AC_DEFINE(ENABLE_RESTART_DEBUGGING, 1,
[Defined if we're building with support for in-process restart debugging.])
fi
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
Add configure option to control ZSTD_STATIC_LINKING_ONLY
2018-02-06 17:58:05 +01:00
if test "$enable_zstd_advanced_apis" != "no"; then
AC_DEFINE(ENABLE_ZSTD_ADVANCED_APIS, 1,
[Defined if we're going to try to use zstd's "static-only" APIs.])
fi
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
# systemd support
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$enable_systemd" = "xno"; then
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
have_systemd=no;
else
PKG_CHECK_MODULES(SYSTEMD,
[libsystemd-daemon],
have_systemd=yes,
have_systemd=no)
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$have_systemd" = "xno"; then
If libsystemd-daemon.pc is not found, look for libsystemd (Thanks to toralf for noting that the build was broken, and to Sebastian for telling me that my fix still wasn't right.)
2015-01-12 15:49:52 +01:00
AC_MSG_NOTICE([Okay, checking for systemd a different way...])
PKG_CHECK_MODULES(SYSTEMD,
[libsystemd],
have_systemd=yes,
have_systemd=no)
fi
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$have_systemd" = "xyes"; then
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
AC_DEFINE(HAVE_SYSTEMD,1,[Have systemd])
Better workaround for CFLAGS issues from #14072; fixes #14162 When I applied patch fcc78e5f8a3249eadfea31db, I somehow broke stack trace symbols on Linux. I'll leave it to others to figure out why that happens. This should be better. Really. Fixes bug 14162; bug not in any released version of Tor.
2015-01-10 23:21:11 +01:00
TOR_SYSTEMD_CFLAGS="${SYSTEMD_CFLAGS}"
Use package-config output for -lsystemd correctly In systemd 209, they deprecated -lsystemd-daemon in favor of -lsystemd. So we'd better actually look at the pkg-config output, or we'll get warnings on newer distributions. For some as-yet-unknown-to-me reason, setting CFLAGS so early makes it so -O2 -g doesn't get added to it later. So, adding it myself later. Perhaps a better fix here can be found. Fixes 14072; bugfix on 0.2.6.2-alpha. Based on a patch from h.venev
2015-01-06 20:05:35 +01:00
TOR_SYSTEMD_LIBS="${SYSTEMD_LIBS}"
Fix --enable-systemd builds on systems with libsystemd but not systemd Fixes bug 16164; bugfix on 0.2.6.3-alpha. Patch from Peter Palfrader.
2015-05-26 15:39:53 +02:00
PKG_CHECK_MODULES(LIBSYSTEMD209, [libsystemd >= 209],
configure: Fix a typo in the systemd version message
2019-10-03 08:32:30 +02:00
[AC_DEFINE(HAVE_SYSTEMD_209,1,[Have systemd v209 or greater])], [])
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
fi
Better workaround for CFLAGS issues from #14072; fixes #14162 When I applied patch fcc78e5f8a3249eadfea31db, I somehow broke stack trace symbols on Linux. I'll leave it to others to figure out why that happens. This should be better. Really. Fixes bug 14162; bug not in any released version of Tor.
2015-01-10 23:21:11 +01:00
AC_SUBST(TOR_SYSTEMD_CFLAGS)
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
AC_SUBST(TOR_SYSTEMD_LIBS)
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$enable_systemd" = "xyes" -a "x$have_systemd" != "xyes" ; then
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
AC_MSG_ERROR([Explicitly requested systemd support, but systemd not found, $pkg_config_user_action, or set SYSTEMD_CFLAGS and SYSTEMD_LIBS.])
add support for systemd notification protocol This permit for now to signal readiness in a cleaner way to systemd.
2014-02-21 00:24:25 +01:00
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
case "$host" in
Stop solaris from warning about localtime_r and friends. AFAICT, no other platforms want this. svn:r4827
2005-08-24 04:11:33 +02:00
*-*-solaris* )
AC_DEFINE(_REENTRANT, 1, [Define on some platforms to activate x_r() functions in time.h])
;;
esac
r8948@totoro: nickm | 2006-10-08 17:04:53 -0400 Add an --enable-gcc-warnings flag to configure.in svn:r8655
2006-10-08 23:05:11 +02:00
AC_ARG_ENABLE(gcc-warnings,
Make advisory-warnings on by default. Add --enable-fatal-warnings to control -Werror. Closes ticket 19044.
2016-05-23 20:39:56 +02:00
AS_HELP_STRING(--enable-gcc-warnings, [deprecated alias for enable-fatal-warnings]))
AC_ARG_ENABLE(fatal-warnings,
AS_HELP_STRING(--enable-fatal-warnings, [tell the compiler to treat all warnings as errors.]))
New configure option: --enable-gcc-warnings-advisory the new configure option --enable-gcc-warnings-advisory enables all the gcc flags that --enable-gcc-warnings does with the exception of -Werror.
2010-01-24 19:32:41 +01:00
AC_ARG_ENABLE(gcc-warnings-advisory,
Make advisory-warnings on by default. Add --enable-fatal-warnings to control -Werror. Closes ticket 19044.
2016-05-23 20:39:56 +02:00
AS_HELP_STRING(--disable-gcc-warnings-advisory, [disable the regular verbose warnings]))
r8948@totoro: nickm | 2006-10-08 17:04:53 -0400 Add an --enable-gcc-warnings flag to configure.in svn:r8655
2006-10-08 23:05:11 +02:00
Add support for gcc compiler/linker hardening flags. This patch adds support for two new configure options: '--enable-gcc-hardening' This sets CFLAGS to include: "-D_FORTIFY_SOURCE=2 -fstack-protector-all" "-fwrapv -fPIE -Wstack-protector -Wformat -Wformat-security" "-Wpointer-sign" It sets LDFLAGS to include: "-pie" '--enable-linker-hardening' This sets LDFLAGS to include: " -z relro -z now"
2009-11-07 00:45:27 +01:00
dnl Others suggest '/gs /safeseh /nxcompat /dynamicbase' for non-gcc on Windows
AC_ARG_ENABLE(gcc-hardening,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-gcc-hardening, [disable compiler security checks]))
Add support for gcc compiler/linker hardening flags. This patch adds support for two new configure options: '--enable-gcc-hardening' This sets CFLAGS to include: "-D_FORTIFY_SOURCE=2 -fstack-protector-all" "-fwrapv -fPIE -Wstack-protector -Wformat -Wformat-security" "-Wpointer-sign" It sets LDFLAGS to include: "-pie" '--enable-linker-hardening' This sets LDFLAGS to include: " -z relro -z now"
2009-11-07 00:45:27 +01:00
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
dnl Deprecated --enable-expensive-hardening but keep it for now for backward compat.
New --enable-expensive-hardening option It turns on -fsanitize=address and -fsanitize=ubsan if they work. Most relays won't want this. Some clients may. Ticket 11477.
2014-04-14 23:10:34 +02:00
AC_ARG_ENABLE(expensive-hardening,
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
AS_HELP_STRING(--enable-expensive-hardening, [enable more fragile and expensive compiler hardening; makes Tor slower]))
AC_ARG_ENABLE(fragile-hardening,
AS_HELP_STRING(--enable-fragile-hardening, [enable more fragile and expensive compiler hardening; makes Tor slower]))
if test "x$enable_expensive_hardening" = "xyes" || test "x$enable_fragile_hardening" = "xyes"; then
fragile_hardening="yes"
configure.ac: Define ENABLE_FRAGILE_HARDENING macro Allow conditional compilation based on whether the "--enable-fragile-hardening" configure option was specified.
2021-09-22 20:42:45 +02:00
AC_DEFINE(ENABLE_FRAGILE_HARDENING, 1, [Defined if we're building with additional, fragile and expensive compiler hardening])
Enable DEBUG_SMARTLIST for entire codebase when expensive hardening is on
2018-06-03 13:48:20 +02:00
AC_DEFINE(DEBUG_SMARTLIST, 1, [Enable smartlist debugging])
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
fi
New --enable-expensive-hardening option It turns on -fsanitize=address and -fsanitize=ubsan if they work. Most relays won't want this. Some clients may. Ticket 11477.
2014-04-14 23:10:34 +02:00
Add --enable-all-bugs-are-fatal option to configure This option enables the ALL_BUGS_ARE_FATAL macro. And use --enable-all-bugs-are-fatal in the Travis configuration.
2020-03-13 19:24:54 +01:00
AC_ARG_ENABLE(all-bugs-are-fatal,
AS_HELP_STRING(--enable-all-bugs-are-fatal, [force all soft asserts in Tor codebase (tor_assert_nonfatal(), BUG(), etc.) to act as hard asserts (tor_assert() and equivalents); makes Tor fragile; only recommended for dev builds]))
if test "x$enable_all_bugs_are_fatal" = "xyes"; then
AC_DEFINE(ALL_BUGS_ARE_FATAL, 1, [All assert failures are fatal])
fi
Add support for gcc compiler/linker hardening flags. This patch adds support for two new configure options: '--enable-gcc-hardening' This sets CFLAGS to include: "-D_FORTIFY_SOURCE=2 -fstack-protector-all" "-fwrapv -fPIE -Wstack-protector -Wformat -Wformat-security" "-Wpointer-sign" It sets LDFLAGS to include: "-pie" '--enable-linker-hardening' This sets LDFLAGS to include: " -z relro -z now"
2009-11-07 00:45:27 +01:00
dnl Linker hardening options
dnl Currently these options are ELF specific - you can't use this with MacOSX
AC_ARG_ENABLE(linker-hardening,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-linker-hardening, [disable linker security fixups]))
Add support for gcc compiler/linker hardening flags. This patch adds support for two new configure options: '--enable-gcc-hardening' This sets CFLAGS to include: "-D_FORTIFY_SOURCE=2 -fstack-protector-all" "-fwrapv -fPIE -Wstack-protector -Wformat -Wformat-security" "-Wpointer-sign" It sets LDFLAGS to include: "-pie" '--enable-linker-hardening' This sets LDFLAGS to include: " -z relro -z now"
2009-11-07 00:45:27 +01:00
This patch changes the default location where config and data files are stored when the --enable-local-appdata option is configured. This changes the Windows path from %APPDATA% to a host local %USERPROFILE%\Local Settings\Application Data\ path (aka, LOCAL_APPDATA). Patch from coderman. svn:r18122
2009-01-16 00:07:11 +01:00
AC_ARG_ENABLE(local-appdata,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--enable-local-appdata, [default to host local application data paths on Windows]))
This patch changes the default location where config and data files are stored when the --enable-local-appdata option is configured. This changes the Windows path from %APPDATA% to a host local %USERPROFILE%\Local Settings\Application Data\ path (aka, LOCAL_APPDATA). Patch from coderman. svn:r18122
2009-01-16 00:07:11 +01:00
if test "$enable_local_appdata" = "yes"; then
AC_DEFINE(ENABLE_LOCAL_APPDATA, 1,
[Defined if we default to host local appdata paths on Windows])
fi
Better detection for certain broken cross-compiling situations. Fixes bug 9869; patch from Benedikt Gollatz.
2013-10-21 19:30:38 +02:00
AC_ARG_ENABLE(tool-name-check,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-tool-name-check, [check for sanely named toolchain when cross-compiling]))
Better detection for certain broken cross-compiling situations. Fixes bug 9869; patch from Benedikt Gollatz.
2013-10-21 19:30:38 +02:00
New --disable-seccomp option to turn off support for seccomp. Fixes 11628.
2014-04-28 17:11:11 +02:00
AC_ARG_ENABLE(seccomp,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-seccomp, [do not attempt to use libseccomp]))
New --disable-seccomp option to turn off support for seccomp. Fixes 11628.
2014-04-28 17:11:11 +02:00
More generic passphrase hashing code, including scrypt support Uses libscrypt when found; otherwise, we don't have scrypt and we only support openpgp rfc2440 s2k hashing, or pbkdf2. Includes documentation and unit tests; coverage around 95%. Remaining uncovered code is sanity-checks that shouldn't be reachable fwict.
2014-08-28 23:38:22 +02:00
AC_ARG_ENABLE(libscrypt,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--disable-libscrypt, [do not attempt to use libscrypt]))
More generic passphrase hashing code, including scrypt support Uses libscrypt when found; otherwise, we don't have scrypt and we only support openpgp rfc2440 s2k hashing, or pbkdf2. Includes documentation and unit tests; coverage around 95%. Remaining uncovered code is sanity-checks that shouldn't be reachable fwict.
2014-08-28 23:38:22 +02:00
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
dnl --- Tracing Options. ---
trace: Add a basic event-tracing infrastructure. This commit adds the src/trace directory containing the basics for our tracing subsystem. It is not used in the code base. The "src/trace/debug.h" file contains an example on how we can map our tor trace events to log_debug(). The tracing subsystem can only be enabled by tracing framework at compile time. This commit introduces the "--enable-tracing-debug" option that will make all "tor_trace()" function be maped to "log_debug()". Closes #13802 Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-01-23 20:28:10 +01:00
trace: Add LTTng-UST interface support No probes at this point. They are per subsystem and thus in later commits. Part of #32910
2020-02-11 17:26:04 +01:00
TOR_TRACE_LIBS=
dnl LTTng instrumentation option.
AC_ARG_ENABLE(tracing-instrumentation-lttng,
AS_HELP_STRING([--enable-tracing-instrumentation-lttng],
[build with LTTng-UST instrumentation]))
AM_CONDITIONAL([USE_TRACING_INSTRUMENTATION_LTTNG],
[test "x$enable_tracing_instrumentation_lttng" = "xyes"])
if test "x$enable_tracing_instrumentation_lttng" = "xyes"; then
AC_CHECK_HEADERS([lttng/tracepoint.h], [],
[AC_MSG_ERROR([LTTng instrumentation headers not found.
On Debian, apt install liblttng-ust-dev"])], [])
AC_DEFINE([USE_TRACING_INSTRUMENTATION_LTTNG], [1], [Using LTTng instrumentation])
TOR_TRACE_LIBS="-llttng-ust -ldl"
trace: Emit a warning if tracing is built in Built in tracing should _not_ be run if it was not set on purpose. Warn as loud as we can in order to inform the user that they are running a version with tracing capabilities built in. This commit also adds a subsys stub because utlimately the logging will happen in the init phase but because the default log file is not set in the sys_logging init function, the stub is not useful for now. Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-03-11 17:12:28 +01:00
have_tracing=1
trace: Add LTTng-UST interface support No probes at this point. They are per subsystem and thus in later commits. Part of #32910
2020-02-11 17:26:04 +01:00
fi
trace: Add USDT probes generation support This commit adds both configure options and probe generation for tracepoints. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-14 20:58:09 +01:00
dnl USDT instrumentation option.
AC_ARG_ENABLE(tracing-instrumentation-usdt,
AS_HELP_STRING([--enable-tracing-instrumentation-usdt],
[build with tracing USDT instrumentation]))
AM_CONDITIONAL([USE_TRACING_INSTRUMENTATION_USDT],
[test "x$enable_tracing_instrumentation_usdt" = "xyes"])
if test "x$enable_tracing_instrumentation_usdt" = "xyes"; then
AC_CHECK_HEADERS([sys/sdt.h], [],
[AC_MSG_ERROR([USDT instrumentation requires sys/sdt.h header.
On Debian, apt install systemtap-sdt-dev])], [])
configure: Check STAP_PROBEV macro when building with USDT tracing It turns out that STAP_PROBEV() is not available on FreeBSD thus having sdt/sdt.h is not enough. Look for it now at configure time. Closes #40174 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-11-03 14:59:29 +01:00
AC_MSG_CHECKING([STAP_PROBEV()])
AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
#define SDT_USE_VARIADIC
#include <sys/sdt.h>
void test(void)
{
STAP_PROBEV(p, n, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12);
}
]])], [
AC_MSG_RESULT([yes])
dnl LTTng generates USDT probes if the UST library was built with
dnl --with-sdt. There is unfortunately no way to check that so we always
dnl build the USDT probes even though LTTng instrumentation was requested.
AC_DEFINE([USE_TRACING_INSTRUMENTATION_USDT], [1], [Using USDT instrumentation])
have_tracing=1
], [
AC_MSG_RESULT([no])
AC_MSG_ERROR([USDT tracing support requires STAP_PROBEV()])
])
trace: Add USDT probes generation support This commit adds both configure options and probe generation for tracepoints. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-14 20:58:09 +01:00
fi
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
dnl Tracepoints event to debug logs.
AC_ARG_ENABLE(tracing-instrumentation-log-debug,
AS_HELP_STRING([--enable-tracing-instrumentation-log-debug],
[build with tracing event to debug log]),
AC_DEFINE([USE_TRACING_INSTRUMENTATION_LOG_DEBUG], [1],
[Tracepoints to log debug]), [])
AM_CONDITIONAL([USE_TRACING_INSTRUMENTATION_LOG_DEBUG],
[test "x$enable_tracing_instrumentation_log_debug" = "xyes"])
trace: Emit a warning if tracing is built in Built in tracing should _not_ be run if it was not set on purpose. Warn as loud as we can in order to inform the user that they are running a version with tracing capabilities built in. This commit also adds a subsys stub because utlimately the logging will happen in the init phase but because the default log file is not set in the sys_logging init function, the stub is not useful for now. Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-03-11 17:12:28 +01:00
if test "x$enable_tracing_instrumentation_log_debug" = "xyes"; then
have_tracing=1
fi
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
trace: Add USDT probes generation support This commit adds both configure options and probe generation for tracepoints. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-14 20:58:09 +01:00
dnl Define that tracing is supported if any instrumentation is used.
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
AM_COND_IF([USE_TRACING_INSTRUMENTATION_LOG_DEBUG],
AC_DEFINE([HAVE_TRACING], [1], [Compiled with tracing support]))
trace: Add USDT probes generation support This commit adds both configure options and probe generation for tracepoints. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-14 20:58:09 +01:00
AM_COND_IF([USE_TRACING_INSTRUMENTATION_USDT],
AC_DEFINE([HAVE_TRACING], [1], [Compiled with tracing support]))
trace: Add LTTng-UST interface support No probes at this point. They are per subsystem and thus in later commits. Part of #32910
2020-02-11 17:26:04 +01:00
AM_COND_IF([USE_TRACING_INSTRUMENTATION_LTTNG],
AC_DEFINE([HAVE_TRACING], [1], [Compiled with tracing support]))
trace: Emit a warning if tracing is built in Built in tracing should _not_ be run if it was not set on purpose. Warn as loud as we can in order to inform the user that they are running a version with tracing capabilities built in. This commit also adds a subsys stub because utlimately the logging will happen in the init phase but because the default log file is not set in the sys_logging init function, the stub is not useful for now. Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-03-11 17:12:28 +01:00
AM_CONDITIONAL([USE_TRACING], [test "x$have_tracing" = x1 ])
trace: Add LTTng-UST interface support No probes at this point. They are per subsystem and thus in later commits. Part of #32910
2020-02-11 17:26:04 +01:00
dnl Finally, define the trace libs.
AC_SUBST([TOR_TRACE_LIBS])
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
dnl -- End Tracing Options. --
trace: Add a basic event-tracing infrastructure. This commit adds the src/trace directory containing the basics for our tracing subsystem. It is not used in the code base. The "src/trace/debug.h" file contains an example on how we can map our tor trace events to log_debug(). The tracing subsystem can only be enabled by tracing framework at compile time. This commit introduces the "--enable-tracing-debug" option that will make all "tor_trace()" function be maped to "log_debug()". Closes #13802 Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-01-23 20:28:10 +01:00
Add support for Android's logging subsystem. This patch adds support for Android's logging subsystem in Tor. When debugging Android applications it is useful to be able to collect information about the application running on the platform via the various system services that is available on the platform. This patch allows you to add "Log notice android" to your torrc and have Tor send everything above and including the notice severity to Android's ring buffer which can be inspected using the 'adb logcat' program. See: https://bugs.torproject.org/24362
2017-12-11 14:12:25 +01:00
dnl Enable Android only features.
AC_ARG_ENABLE(android,
AS_HELP_STRING(--enable-android, [build with Android features enabled]))
AM_CONDITIONAL([USE_ANDROID], [test "x$enable_android" = "xyes"])
if test "x$enable_android" = "xyes"; then
AC_DEFINE([USE_ANDROID], [1], [Compile with Android specific features enabled])
fi
mod: Build system changes for dirauth module Make our build system support a disable dirauth module option. It can only be disabled explicitly with: $ ./configure --disable-module-dirauth If *not* specified that is enabled, an automake conditional variable is set to true and a defined value for the C code: AM_CONDITIONAL: BUILD_MODULE_DIRAUTH AC_DEFINE: HAVE_MODULE_DIRAUTH=1 This introduces the dirauth/ module directory in src/or/ for which .c files are only compiled if the BUILD_MODULE_DIRAUTH is set. All the header files are compiled in regardless of the support so we can use the alternative entry point functions of the dirauth subsystem. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-03 21:06:34 +02:00
dnl ---
dnl Tor modules options. These options are namespaced with --disable-module-XXX
dnl ---
build: Always compile module support for tests The --disable-module-* configure option removes code from the final binary but we still build the unit tests with the disable module(s) so we can actually test that code path all the time and not forget about it. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-27 17:33:22 +02:00
dnl All our modules.
hs_pow: Make proof-of-work support optional in configure This adds a new "pow" module for the user-visible proof of work support in ./configure, and this disables src/feature/hs/hs_pow at compile-time. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-09 00:44:55 +01:00
m4_define([MODULES], [relay dirauth dircache pow])
configure: Add a relay module, with --disable-module-relay The Makefile and C defines don't actually do anything yet. Part of 32123.
2019-10-17 08:07:52 +02:00
configure: Don't print disable option for module that can't be disabled This is currently for the dircache module that can not be disabled by itself, it is only disabled from the relay module. Thus, we should not print in the configure summary the disable option. Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-01-14 15:25:46 +01:00
# Some modules are only disabled through another option. For those, we don't
# want to print the help in the summary at the end of the configure. Any entry
# in the following set will not print the "--disable-module-NAME" command in
# the summary.
m4_set_add_all([MODULES_WITH_NO_OPTIONS], [dircache])
configure: Add a relay module, with --disable-module-relay The Makefile and C defines don't actually do anything yet. Part of 32123.
2019-10-17 08:07:52 +02:00
dnl Relay module.
hs_pow: Make proof-of-work support optional in configure This adds a new "pow" module for the user-visible proof of work support in ./configure, and this disables src/feature/hs/hs_pow at compile-time. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-09 00:44:55 +01:00
m4_define([module_option_hints(relay)],
[AS_IF([test "x$value" = x1], [HINT_OPT([--disable-module-relay])],
[HINT_NONE])])
configure: Add a relay module, with --disable-module-relay The Makefile and C defines don't actually do anything yet. Part of 32123.
2019-10-17 08:07:52 +02:00
AC_ARG_ENABLE([module-relay],
AS_HELP_STRING([--disable-module-relay],
configure: Make --disable-module-relay imply --disable-module-dirauth Part of 32123.
2019-10-17 08:10:45 +02:00
[Build tor without the Relay modules: tor can not run as a relay, bridge, or authority. Implies --disable-module-dirauth]))
configure: Add a relay module, with --disable-module-relay The Makefile and C defines don't actually do anything yet. Part of 32123.
2019-10-17 08:07:52 +02:00
AM_CONDITIONAL(BUILD_MODULE_RELAY, [test "x$enable_module_relay" != "xno"])
AM_COND_IF(BUILD_MODULE_RELAY,
AC_DEFINE([HAVE_MODULE_RELAY], [1],
[Compile with Relay feature support]))
build: Always compile module support for tests The --disable-module-* configure option removes code from the final binary but we still build the unit tests with the disable module(s) so we can actually test that code path all the time and not forget about it. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-27 17:33:22 +02:00
Define a "dircache" module. For now, this module is enabled whenever the relay module is enabled, and disabled whenever the relay module is disabled. Though they are logically separate, the use cases for running one without the other are rare enough that we don't really want to support compiling them independently.
2020-01-09 03:03:20 +01:00
dnl Dircache module. (This cannot be enabled or disabled independently of
Document why dircache is not included in --list-modules
2020-01-15 18:46:10 +01:00
dnl the relay module. It is not listed by --list-modules for this reason.)
Define a "dircache" module. For now, this module is enabled whenever the relay module is enabled, and disabled whenever the relay module is disabled. Though they are logically separate, the use cases for running one without the other are rare enough that we don't really want to support compiling them independently.
2020-01-09 03:03:20 +01:00
AM_CONDITIONAL(BUILD_MODULE_DIRCACHE,
[test "x$enable_module_relay" != "xno"])
AM_COND_IF(BUILD_MODULE_DIRCACHE,
AC_DEFINE([HAVE_MODULE_DIRCACHE], [1],
[Compile with directory cache support]))
mod: Build system changes for dirauth module Make our build system support a disable dirauth module option. It can only be disabled explicitly with: $ ./configure --disable-module-dirauth If *not* specified that is enabled, an automake conditional variable is set to true and a defined value for the C code: AM_CONDITIONAL: BUILD_MODULE_DIRAUTH AC_DEFINE: HAVE_MODULE_DIRAUTH=1 This introduces the dirauth/ module directory in src/or/ for which .c files are only compiled if the BUILD_MODULE_DIRAUTH is set. All the header files are compiled in regardless of the support so we can use the alternative entry point functions of the dirauth subsystem. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-03 21:06:34 +02:00
dnl Directory Authority module.
hs_pow: Make proof-of-work support optional in configure This adds a new "pow" module for the user-visible proof of work support in ./configure, and this disables src/feature/hs/hs_pow at compile-time. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-09 00:44:55 +01:00
m4_define([module_option_hints(dirauth)],
[AS_IF([test "x$value" = x1], [HINT_OPT([--disable-module-dirauth])],
[HINT_NONE])])
mod: Build system changes for dirauth module Make our build system support a disable dirauth module option. It can only be disabled explicitly with: $ ./configure --disable-module-dirauth If *not* specified that is enabled, an automake conditional variable is set to true and a defined value for the C code: AM_CONDITIONAL: BUILD_MODULE_DIRAUTH AC_DEFINE: HAVE_MODULE_DIRAUTH=1 This introduces the dirauth/ module directory in src/or/ for which .c files are only compiled if the BUILD_MODULE_DIRAUTH is set. All the header files are compiled in regardless of the support so we can use the alternative entry point functions of the dirauth subsystem. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-03 21:06:34 +02:00
AC_ARG_ENABLE([module-dirauth],
AS_HELP_STRING([--disable-module-dirauth],
configure: Interpret --disable-module-dirauth=no correctly Fixes bug 32134; bugfix on 0.3.4.1-alpha.
2019-10-17 08:04:39 +02:00
[Build tor without the Directory Authority module: tor can not run as a directory authority or bridge authority]))
configure: Make --disable-module-relay imply --disable-module-dirauth Part of 32123.
2019-10-17 08:10:45 +02:00
AM_CONDITIONAL(BUILD_MODULE_DIRAUTH,[test "x$enable_module_dirauth" != "xno" && test "x$enable_module_relay" != "xno"])
configure: Interpret --disable-module-dirauth=no correctly Fixes bug 32134; bugfix on 0.3.4.1-alpha.
2019-10-17 08:04:39 +02:00
AM_COND_IF(BUILD_MODULE_DIRAUTH,
AC_DEFINE([HAVE_MODULE_DIRAUTH], [1],
[Compile with Directory Authority feature support]))
mod: Build system changes for dirauth module Make our build system support a disable dirauth module option. It can only be disabled explicitly with: $ ./configure --disable-module-dirauth If *not* specified that is enabled, an automake conditional variable is set to true and a defined value for the C code: AM_CONDITIONAL: BUILD_MODULE_DIRAUTH AC_DEFINE: HAVE_MODULE_DIRAUTH=1 This introduces the dirauth/ module directory in src/or/ for which .c files are only compiled if the BUILD_MODULE_DIRAUTH is set. All the header files are compiled in regardless of the support so we can use the alternative entry point functions of the dirauth subsystem. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-03 21:06:34 +02:00
hs_pow: Make proof-of-work support optional in configure This adds a new "pow" module for the user-visible proof of work support in ./configure, and this disables src/feature/hs/hs_pow at compile-time. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-09 00:44:55 +01:00
dnl Hidden Service Proof-of-Work module.
m4_define([module_option_hints(pow)],
[AS_IF([test "x$value" = x1], [HINT_OPT([--disable-module-pow])],
[AS_IF([test "x$license_option" != "xGPL"], [HINT_OPT([requires --enable-gpl])],
[HINT_NONE])])])
AC_ARG_ENABLE([module-pow],
AS_HELP_STRING([--disable-module-pow],
[Build tor without proof-of-work denial of service mitigation, normally available when building with --enable-gpl]))
AM_CONDITIONAL(BUILD_MODULE_POW,
[test "x$license_option" = "xGPL" && test "x$enable_module_pow" != "xno"])
AM_COND_IF(BUILD_MODULE_POW,
AC_DEFINE([HAVE_MODULE_POW], [1], [Compile with proof-of-work support]))
build: Always compile module support for tests The --disable-module-* configure option removes code from the final binary but we still build the unit tests with the disable module(s) so we can actually test that code path all the time and not forget about it. Signed-off-by: David Goulet <dgoulet@torproject.org>
2018-04-27 17:33:22 +02:00
dnl Helper variables.
TOR_MODULES_ALL_ENABLED=
AC_DEFUN([ADD_MODULE], [
MODULE=m4_toupper($1)
TOR_MODULES_ALL_ENABLED="${TOR_MODULES_ALL_ENABLED} -DHAVE_MODULE_${MODULE}=1"
])
m4_foreach_w([module], MODULES, [ADD_MODULE([module])])
AC_SUBST(TOR_MODULES_ALL_ENABLED)
Use the old replacement when AM_PROG_AR doesn't exist.
2016-12-23 16:44:34 +01:00
dnl check for the correct "ar" when cross-compiling.
dnl (AM_PROG_AR was new in automake 1.11.2, which we do not yet require,
dnl so kludge up a replacement for the case where it isn't there yet.)
m4_ifdef([AM_PROG_AR],
[AM_PROG_AR],
[AN_MAKEVAR([AR], [AC_PROG_AR])
AN_PROGRAM([ar], [AC_PROG_AR])
AC_DEFUN([AC_PROG_AR], [AC_CHECK_TOOL([AR], [ar], [:])])
AC_PROG_AR])
Look for correct "ar" to cross-compile For some reason, autoconf doesn't by default have an "AC_PROG_AR" for this -- possibly it's assumed that any "ar" you have will work everyplace. Fixes bug 3909; found by sid77. This fixes a build issue first present in fdbdb4dc1500, but the bug (of not using a correct ar) has been in every Tor version ever: it just didn't matter until then.
2011-09-02 17:51:49 +02:00
Better detection for certain broken cross-compiling situations. Fixes bug 9869; patch from Benedikt Gollatz.
2013-10-21 19:30:38 +02:00
dnl Check whether the above macro has settled for a simply named tool even
dnl though we're cross compiling. We must do this before running AC_PROG_CC,
dnl because that will find any cc on the system, not only the cross-compiler,
dnl and then verify that a binary built with this compiler runs on the
dnl build system. It will then come to the false conclusion that we're not
dnl cross-compiling.
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$enable_tool_name_check" != "xno"; then
if test "x$ac_tool_warned" = "xyes"; then
Better detection for certain broken cross-compiling situations. Fixes bug 9869; patch from Benedikt Gollatz.
2013-10-21 19:30:38 +02:00
AC_MSG_ERROR([We are cross compiling but could not find a properly named toolchain. Do you have your cross-compiling toolchain in PATH? (You can --disable-tool-name-check to ignore this.)])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
elif test "x$ac_ct_AR" != "x" -a "x$cross_compiling" = "xmaybe"; then
Better detection for certain broken cross-compiling situations. Fixes bug 9869; patch from Benedikt Gollatz.
2013-10-21 19:30:38 +02:00
AC_MSG_ERROR([We think we are cross compiling but could not find a properly named toolchain. Do you have your cross-compiling toolchain in PATH? (You can --disable-tool-name-check to ignore this.)])
fi
fi
added automake/autoconf support. When in doubt, "aclocal && autoconf && autoheader && automake" from the top dir. svn:r10
2002-06-29 01:26:42 +02:00
AC_PROG_CC
r17090@catbus: nickm | 2007-12-11 18:05:24 -0500 Fix some autoconf bugs, and prevent more from occurring. Partial backport candidate. svn:r12767
2007-12-12 00:06:48 +01:00
AC_PROG_CPP
added automake/autoconf support. When in doubt, "aclocal && autoconf && autoheader && automake" from the top dir. svn:r10
2002-06-29 01:26:42 +02:00
AC_PROG_MAKE_SET
AC_PROG_RANLIB
Use the Autoconf macro for finding sed The AC_PROG_SED macro was added in Autoconf 2.59b and the minimum supported Autoconf version is now 2.63 so we can safely assume it is available.
2016-12-15 16:33:19 +01:00
AC_PROG_SED
autoconf 2.59 appears not to support AC_PROG_SED
2009-08-21 05:02:57 +02:00
Add the compiler name, version and libs used to compile This changes the behaviour of `tor --version` in such a way. ```console src/app/tor --version Tor version 0.4.5.1-alpha-dev (git-46ccde66a97d7985). Tor is running on Linux with Libevent 2.1.12-stable, OpenSSL 1.1.1h, Zlib 1.2.11, Liblzma 5.2.4, Libzstd 1.4.5 and Glibc 2.31 as libc. Tor compiled with GCC version 10.2.0 ``` Fixes #32102
2020-11-17 10:43:00 +01:00
AX_COMPILER_VERSION
AX_COMPILER_VENDOR
AC_DEFINE_UNQUOTED([COMPILER_VERSION],["$ax_cv_c_compiler_version"], [Compiler version])
AC_DEFINE_UNQUOTED([COMPILER_VENDOR],["$ax_cv_c_compiler_vendor"], [Compiler vendor])
AS_CASE($ax_cv_c_compiler_vendor,
[gnu], AC_DEFINE([COMPILER],["GCC"],[Compiler name]),
AC_DEFINE([COMPILER],[],[Compiler name])
)
Run check-spaces only when Perl is available Also permit users to override the Perl variable with relative paths.
2016-12-15 13:02:42 +01:00
AC_ARG_VAR([PERL], [path to Perl binary])
AC_CHECK_PROGS([PERL], [perl])
AM_CONDITIONAL(USE_PERL, [test "x$ac_cv_prog_PERL" != "x"])
Make `check-docs` work from out-of-tree builds.
2015-03-11 10:47:13 +01:00
Allow generating documentation from asciidoc in the Makefile
2010-01-18 22:15:38 +01:00
dnl check for asciidoc and a2x
AC_PATH_PROG([ASCIIDOC], [asciidoc], none)
whoops; mistaek in a496010642c2. so many comma
2014-02-18 02:15:59 +01:00
AC_PATH_PROGS([A2X], [a2x a2x.py], none)
Add configure switch to disable use of asciidoc Also break the build if that switch isn't used and asciidoc isn't available.
2010-03-01 03:17:48 +01:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
AM_CONDITIONAL(USE_ASCIIDOC, test "x$asciidoc" = "xtrue")
Allow excluding documentation from the build
2019-06-16 19:18:30 +02:00
AM_CONDITIONAL(BUILD_MANPAGE, [test "x$enable_manpage" != "xno"])
AM_CONDITIONAL(BUILD_HTML_DOCS, [test "x$enable_html_manual" != "xno"])
Allow generating documentation from asciidoc in the Makefile
2010-01-18 22:15:38 +01:00
Changes to tor-fw-helper, some based on Nick's review * MINIUPNPC rather than the generic UPNP * Nick suggested a better abstraction model for tor-fw-helper * Fix autoconf to build with either natpmp or miniupnpc * Add AM_PROG_CC_C_O to fix automake complaint * update spec to address nickm's concern * refactor nat-pmp to match upnp state * we prefer tor_snprintf to snprintf * link properlty for tor_snprintf * rename test_commandline_options to log_commandline_options * cast this uint as an int * detect possible FD_SETSIZE errors * make note about future enhancements for natpmp * add upnp enhancement note * ChangeLog entry * doxygen and check-spaces cleanup * create tor-fw-helper.1.txt
2010-09-24 21:52:07 +02:00
AM_PROG_CC_C_O
Try making our configure.ac script build with AC 2.70. In versions <=2.69, according to the autoconf docs, AC_PROG_CC_C99 is needed with some compilers, if they require extra arguments to build C99 programs. In versions >=2.70, AC_PROG_CC checks for these compilers automatically, and so the AC_PROG_CC_C99 macro is obsolete. So, what can you do if you want your script to work right with both autoconf versions? IIUC, neither including AC_PROG_CC_C99 macro nor leaving it out will give you the right behavior with both versions. It looks like you need to look at the autoconf version explicitly. (Now, the autoconf manual implies that it's "against autoconf philosophy" to look at the autoconf version rather than trying the behavior to see if it works, but they don't actually tell you how to detect recoverably at autoconf-time whether a macro is obsolete or not, and I can't find a way to do that.) So, is it safe to use m4_version_prereq, like I do here? It isn't listed in the autoconf 2.63 manual (which is the oldest version we support). But a mailing list message [1] (which added the documentation back in 2008) implies that m4_version_prereq has been there since "at least back to autoconf 2.59". https://lists.gnu.org/archive/html/autoconf-patches/2008-12/msg00025.html So I think this will work. I am basing this patch against Tor 0.3.5 since, if autoconf 2.70 becomes widespread before 0.3.5 is unsupported, we might need this patch to continue 0.3.5 development. But I don't think we should backport farther than 0.4.5 until/unless that actually happens. This is part of a fix for #40355.
2021-04-07 16:18:44 +02:00
dnl Before autoconf 2.70, AC_PROG_CC_C99 is supposedly necessary for some
Fix typos
2021-11-12 13:23:08 +01:00
dnl compilers if you want C99 support. Starting with 2.70, it is obsolete and
Try making our configure.ac script build with AC 2.70. In versions <=2.69, according to the autoconf docs, AC_PROG_CC_C99 is needed with some compilers, if they require extra arguments to build C99 programs. In versions >=2.70, AC_PROG_CC checks for these compilers automatically, and so the AC_PROG_CC_C99 macro is obsolete. So, what can you do if you want your script to work right with both autoconf versions? IIUC, neither including AC_PROG_CC_C99 macro nor leaving it out will give you the right behavior with both versions. It looks like you need to look at the autoconf version explicitly. (Now, the autoconf manual implies that it's "against autoconf philosophy" to look at the autoconf version rather than trying the behavior to see if it works, but they don't actually tell you how to detect recoverably at autoconf-time whether a macro is obsolete or not, and I can't find a way to do that.) So, is it safe to use m4_version_prereq, like I do here? It isn't listed in the autoconf 2.63 manual (which is the oldest version we support). But a mailing list message [1] (which added the documentation back in 2008) implies that m4_version_prereq has been there since "at least back to autoconf 2.59". https://lists.gnu.org/archive/html/autoconf-patches/2008-12/msg00025.html So I think this will work. I am basing this patch against Tor 0.3.5 since, if autoconf 2.70 becomes widespread before 0.3.5 is unsupported, we might need this patch to continue 0.3.5 development. But I don't think we should backport farther than 0.4.5 until/unless that actually happens. This is part of a fix for #40355.
2021-04-07 16:18:44 +02:00
dnl forbidden.
m4_version_prereq([2.70], [:], [AC_PROG_CC_C99])
First implementation of tor-fw-helper. tor-fw-helper is a command-line tool to wrap and abstract various firewall port-forwarding tools. This commit matches the state of Jacob's tor-fw-helper branch as of 23 September 2010. (commit msg by Nick)
2010-04-17 02:45:12 +02:00
configure: reject Python 2. Python 2 will be end-of-life as of 1 Jan 2020, so we can finally stop supporting it. As a first step, we should make our configure script stop accepting python 2 as something acceptable to run our tests with. Closes ticket 32608.
2019-11-25 19:01:04 +01:00
AC_CACHE_CHECK([for Python 3], [tor_cv_PYTHON],
[AC_PATH_PROGS_FEATURE_CHECK([PYTHON], [ \
Prefer recent Python 3 in autoconf. Closes ticket 26372.
2018-06-21 15:28:50 +02:00
python3 \
python3.8 python3.7 python3.6 python3.5 python3.4 \
configure: reject Python 2. Python 2 will be end-of-life as of 1 Jan 2020, so we can finally stop supporting it. As a first step, we should make our configure script stop accepting python 2 as something acceptable to run our tests with. Closes ticket 32608.
2019-11-25 19:01:04 +01:00
python ],
[["$ac_path_PYTHON" -c 'import sys; sys.exit(sys.version_info[0]<3)' && tor_cv_PYTHON="$ac_path_PYTHON" ac_path_PYTHON_found=:]] )])
AC_SUBST([PYTHON], [$tor_cv_PYTHON])
PYTHON="$tor_cv_PYTHON"
When python is available, run the commandline unit tests from "make check"
2013-10-11 19:05:58 +02:00
if test "x$PYTHON" = "x"; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
configure: reject Python 2. Python 2 will be end-of-life as of 1 Jan 2020, so we can finally stop supporting it. As a first step, we should make our configure script stop accepting python 2 as something acceptable to run our tests with. Closes ticket 32608.
2019-11-25 19:01:04 +01:00
AC_MSG_WARN([Python 3 unavailable; some tests will not be run.])
When python is available, run the commandline unit tests from "make check"
2013-10-11 19:05:58 +02:00
fi
configure: reject Python 2. Python 2 will be end-of-life as of 1 Jan 2020, so we can finally stop supporting it. As a first step, we should make our configure script stop accepting python 2 as something acceptable to run our tests with. Closes ticket 32608.
2019-11-25 19:01:04 +01:00
When python is available, run the commandline unit tests from "make check"
2013-10-11 19:05:58 +02:00
AM_CONDITIONAL(USEPYTHON, [test "x$PYTHON" != "x"])
Use autoconf's FLEXIBLE_ARRAY_MEMBER for unspecified-length arrays C99 allows a syntax for structures whose last element is of unspecified length: struct s { int elt1; ... char last_element[]; }; Recent (last-5-years) autoconf versions provide an AC_C_FLEXIBLE_ARRAY_MEMBER test that defines FLEXIBLE_ARRAY_MEMBER to either no tokens (if you have c99 flexible array support) or to 1 (if you don't). At that point you just use offsetof [STRUCT_OFFSET() for us] to see where last_element begins, and allocate your structures like: struct s { int elt1; ... char last_element[FLEXIBLE_ARRAY_MEMBER]; }; tor_malloc(STRUCT_OFFSET(struct s, last_element) + n_elements*sizeof(char)); The advantages are: 1) It's easier to see which structures and elements are of unspecified length. 2) The compiler and related checking tools can also see which structures and elements are of unspecified length, in case they wants to try weird bounds-checking tricks or something. 3) The compiler can warn us if we do something dumb, like try to stack-allocate a flexible-length structure.
2011-01-06 21:59:05 +01:00
ifdef([AC_C_FLEXIBLE_ARRAY_MEMBER], [
AC_C_FLEXIBLE_ARRAY_MEMBER
], [
dnl Maybe we've got an old autoconf...
AC_CACHE_CHECK([for flexible array members],
tor_cv_c_flexarray,
[AC_COMPILE_IFELSE(
AC_LANG_PROGRAM([
struct abc { int a; char b[]; };
], [
struct abc *def = malloc(sizeof(struct abc)+sizeof(char));
def->b[0] = 33;
]),
[tor_cv_c_flexarray=yes],
[tor_cv_c_flexarray=no])])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_flexarray" = "yes"; then
Provide missing description for our FLEXIBLE_ARRAY_MEMBER defines Apparently autoheader throws a tantrum if you say 'AC_DEFINE([a], [b])'. Instead you must say 'AC_DEFINE([a], [b], [description of a])'. We were running into this in our replacement definitions for FLEXIBLE_ARRAY_MEMBER, which were only getting built on autoconf versions before 2.61 -- and this made us stop working with those autoconf versinos. Fixes bug 2430; bugfix on 0.2.3.1-alpha.
2011-09-14 01:51:37 +02:00
AC_DEFINE([FLEXIBLE_ARRAY_MEMBER], [], [Define to nothing if C supports flexible array members, and to 1 if it does not.])
Use autoconf's FLEXIBLE_ARRAY_MEMBER for unspecified-length arrays C99 allows a syntax for structures whose last element is of unspecified length: struct s { int elt1; ... char last_element[]; }; Recent (last-5-years) autoconf versions provide an AC_C_FLEXIBLE_ARRAY_MEMBER test that defines FLEXIBLE_ARRAY_MEMBER to either no tokens (if you have c99 flexible array support) or to 1 (if you don't). At that point you just use offsetof [STRUCT_OFFSET() for us] to see where last_element begins, and allocate your structures like: struct s { int elt1; ... char last_element[FLEXIBLE_ARRAY_MEMBER]; }; tor_malloc(STRUCT_OFFSET(struct s, last_element) + n_elements*sizeof(char)); The advantages are: 1) It's easier to see which structures and elements are of unspecified length. 2) The compiler and related checking tools can also see which structures and elements are of unspecified length, in case they wants to try weird bounds-checking tricks or something. 3) The compiler can warn us if we do something dumb, like try to stack-allocate a flexible-length structure.
2011-01-06 21:59:05 +01:00
else
Provide missing description for our FLEXIBLE_ARRAY_MEMBER defines Apparently autoheader throws a tantrum if you say 'AC_DEFINE([a], [b])'. Instead you must say 'AC_DEFINE([a], [b], [description of a])'. We were running into this in our replacement definitions for FLEXIBLE_ARRAY_MEMBER, which were only getting built on autoconf versions before 2.61 -- and this made us stop working with those autoconf versinos. Fixes bug 2430; bugfix on 0.2.3.1-alpha.
2011-09-14 01:51:37 +02:00
AC_DEFINE([FLEXIBLE_ARRAY_MEMBER], [1], [Define to nothing if C supports flexible array members, and to 1 if it does not.])
Use autoconf's FLEXIBLE_ARRAY_MEMBER for unspecified-length arrays C99 allows a syntax for structures whose last element is of unspecified length: struct s { int elt1; ... char last_element[]; }; Recent (last-5-years) autoconf versions provide an AC_C_FLEXIBLE_ARRAY_MEMBER test that defines FLEXIBLE_ARRAY_MEMBER to either no tokens (if you have c99 flexible array support) or to 1 (if you don't). At that point you just use offsetof [STRUCT_OFFSET() for us] to see where last_element begins, and allocate your structures like: struct s { int elt1; ... char last_element[FLEXIBLE_ARRAY_MEMBER]; }; tor_malloc(STRUCT_OFFSET(struct s, last_element) + n_elements*sizeof(char)); The advantages are: 1) It's easier to see which structures and elements are of unspecified length. 2) The compiler and related checking tools can also see which structures and elements are of unspecified length, in case they wants to try weird bounds-checking tricks or something. 3) The compiler can warn us if we do something dumb, like try to stack-allocate a flexible-length structure.
2011-01-06 21:59:05 +01:00
fi
])
Require two c99 features (midblock decls, designated initializers) c99 lets us do neat stuff like: { int j, k; foo(&j, &k); int z = j + k; } and also struct point { int x; int y; }; struct point pt = { .x=5, .y=5 }; This commit makes the configure scripts check to make sure your compiler implements them. It also disables our longstanding warning about midblock declarations. Closes ticket 13233.
2014-09-25 17:20:04 +02:00
AC_CACHE_CHECK([for working C99 mid-block declaration syntax],
tor_cv_c_c99_decl,
[AC_COMPILE_IFELSE(
[AC_LANG_PROGRAM([], [int x; x = 3; int y; y = 4 + x;])],
[tor_cv_c_c99_decl=yes],
[tor_cv_c_c99_decl=no] )])
if test "$tor_cv_c_c99_decl" != "yes"; then
AC_MSG_ERROR([Your compiler doesn't support c99 mid-block declarations. This is required as of Tor 0.2.6.x])
fi
AC_CACHE_CHECK([for working C99 designated initializers],
tor_cv_c_c99_designated_init,
[AC_COMPILE_IFELSE(
[AC_LANG_PROGRAM([struct s { int a; int b; };],
[[ struct s ss = { .b = 5, .a = 6 }; ]])],
[tor_cv_c_c99_designated_init=yes],
[tor_cv_c_c99_designated_init=no] )])
if test "$tor_cv_c_c99_designated_init" != "yes"; then
AC_MSG_ERROR([Your compiler doesn't support c99 designated initializers. This is required as of Tor 0.2.6.x])
fi
Add a fallthrough macro. This macro defers to __attribute__((fallthrough)) on GCC (and clang). Previously we had been using GCC's magic /* fallthrough */ comments, but clang very sensibly doesn't accept those. Since not all compiler recognize it, we only define it when our configure script detects that it works. Part of a fix for 34078.
2020-05-06 16:24:21 +02:00
saved_CFLAGS="$CFLAGS"
CFLAGS="$CFLAGS -Werror"
AC_CACHE_CHECK([for __attribute__((fallthrough))],
tor_cv_c_attr_fallthrough,
[AC_COMPILE_IFELSE(
[AC_LANG_PROGRAM([extern int x; void fn(void) ;],
[[ switch (x) { case 1: fn(); __attribute__((fallthrough));
case 2: fn(); break; } ]])],
[tor_cv_c_attr_fallthrough=yes],
[tor_cv_c_attr_fallthrough=no] )])
CFLAGS="$saved_CFLAGS"
Fix use of non-portable == in configure.ac. Fixes bug 34233. (This has bug has been backported to 0.3.5, but only released in 0.4.3, so it only needs a changes file there.)
2020-05-15 15:58:49 +02:00
if test "$tor_cv_c_attr_fallthrough" = "yes"; then
Add a fallthrough macro. This macro defers to __attribute__((fallthrough)) on GCC (and clang). Previously we had been using GCC's magic /* fallthrough */ comments, but clang very sensibly doesn't accept those. Since not all compiler recognize it, we only define it when our configure script detects that it works. Part of a fix for 34078.
2020-05-06 16:24:21 +02:00
AC_DEFINE(HAVE_ATTR_FALLTHROUGH, [1], [defined if we have the fallthrough attribute.])
fi
Apply RPM-related cleanups from John Bashinski svn:r5639
2005-12-24 05:03:39 +01:00
TORUSER=_tor
AC_ARG_WITH(tor-user,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--with-tor-user=NAME, [specify username for tor daemon]),
Apply RPM-related cleanups from John Bashinski svn:r5639
2005-12-24 05:03:39 +01:00
[
TORUSER=$withval
]
)
AC_SUBST(TORUSER)
TORGROUP=_tor
AC_ARG_WITH(tor-group,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--with-tor-group=NAME, [specify group name for tor daemon]),
Apply RPM-related cleanups from John Bashinski svn:r5639
2005-12-24 05:03:39 +01:00
[
TORGROUP=$withval
]
)
AC_SUBST(TORGROUP)
r8920@totoro: nickm | 2006-10-07 11:40:06 -0400 Fix IRIX builds so that IRIX no longer self-detects as win32. For once, this is not the fault of Windows: Apparently, some helpful person decided that the the MIPSpro C Compiler should give a warning on #error rather than actually failing. How... "innovative." (Bug found by pnx; thanks!) svn:r8635
2006-10-07 17:40:33 +02:00
Change WIN32 to _WIN32 in configure.in, too.
2012-01-31 18:05:03 +01:00
dnl If _WIN32 is defined and non-zero, we are building for win32
r8920@totoro: nickm | 2006-10-07 11:40:06 -0400 Fix IRIX builds so that IRIX no longer self-detects as win32. For once, this is not the fault of Windows: Apparently, some helpful person decided that the the MIPSpro C Compiler should give a warning on #error rather than actually failing. How... "innovative." (Bug found by pnx; thanks!) svn:r8635
2006-10-07 17:40:33 +02:00
AC_MSG_CHECKING([for win32])
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
AC_RUN_IFELSE([AC_LANG_SOURCE([
r9081@totoro: nickm | 2006-10-19 11:00:51 -0400 Try to detect windows properly when cross-compiling svn:r8755
2006-10-19 17:02:06 +02:00
int main(int c, char **v) {
Change WIN32 to _WIN32 in configure.in, too.
2012-01-31 18:05:03 +01:00
#ifdef _WIN32
#if _WIN32
r8920@totoro: nickm | 2006-10-07 11:40:06 -0400 Fix IRIX builds so that IRIX no longer self-detects as win32. For once, this is not the fault of Windows: Apparently, some helpful person decided that the the MIPSpro C Compiler should give a warning on #error rather than actually failing. How... "innovative." (Bug found by pnx; thanks!) svn:r8635
2006-10-07 17:40:33 +02:00
return 0;
#else
return 1;
#endif
#else
return 2;
#endif
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
}])],
r8920@totoro: nickm | 2006-10-07 11:40:06 -0400 Fix IRIX builds so that IRIX no longer self-detects as win32. For once, this is not the fault of Windows: Apparently, some helpful person decided that the the MIPSpro C Compiler should give a warning on #error rather than actually failing. How... "innovative." (Bug found by pnx; thanks!) svn:r8635
2006-10-07 17:40:33 +02:00
bwin32=true; AC_MSG_RESULT([yes]),
r9081@totoro: nickm | 2006-10-19 11:00:51 -0400 Try to detect windows properly when cross-compiling svn:r8755
2006-10-19 17:02:06 +02:00
bwin32=false; AC_MSG_RESULT([no]),
bwin32=cross; AC_MSG_RESULT([cross])
r8920@totoro: nickm | 2006-10-07 11:40:06 -0400 Fix IRIX builds so that IRIX no longer self-detects as win32. For once, this is not the fault of Windows: Apparently, some helpful person decided that the the MIPSpro C Compiler should give a warning on #error rather than actually failing. How... "innovative." (Bug found by pnx; thanks!) svn:r8635
2006-10-07 17:40:33 +02:00
)
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$bwin32" = "cross"; then
r9081@totoro: nickm | 2006-10-19 11:00:51 -0400 Try to detect windows properly when cross-compiling svn:r8755
2006-10-19 17:02:06 +02:00
AC_MSG_CHECKING([for win32 (cross)])
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
AC_COMPILE_IFELSE([AC_LANG_SOURCE([
Change WIN32 to _WIN32 in configure.in, too.
2012-01-31 18:05:03 +01:00
#ifdef _WIN32
r9081@totoro: nickm | 2006-10-19 11:00:51 -0400 Try to detect windows properly when cross-compiling svn:r8755
2006-10-19 17:02:06 +02:00
int main(int c, char **v) {return 0;}
#else
#error
int main(int c, char **v) {return x(y);}
#endif
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
])],
r9081@totoro: nickm | 2006-10-19 11:00:51 -0400 Try to detect windows properly when cross-compiling svn:r8755
2006-10-19 17:02:06 +02:00
bwin32=true; AC_MSG_RESULT([yes]),
bwin32=false; AC_MSG_RESULT([no]))
fi
More windows header stuff. Will it work this time?
2015-07-09 20:58:16 +02:00
AH_BOTTOM([
#ifdef _WIN32
Implement WIN32 tor_cond_wait using condition variables #30187 Fix bug where running a relay on Windows would use 100% CPU after some time. Makes Windows >= Vista the required Windows version to build and run tor.
2020-10-31 19:33:33 +01:00
/* Defined to access windows functions and definitions for >=WinVista */
More windows header stuff. Will it work this time?
2015-07-09 20:58:16 +02:00
# ifndef WINVER
Implement WIN32 tor_cond_wait using condition variables #30187 Fix bug where running a relay on Windows would use 100% CPU after some time. Makes Windows >= Vista the required Windows version to build and run tor.
2020-10-31 19:33:33 +01:00
# define WINVER 0x0600
More windows header stuff. Will it work this time?
2015-07-09 20:58:16 +02:00
# endif
Implement WIN32 tor_cond_wait using condition variables #30187 Fix bug where running a relay on Windows would use 100% CPU after some time. Makes Windows >= Vista the required Windows version to build and run tor.
2020-10-31 19:33:33 +01:00
/* Defined to access _other_ windows functions and definitions for >=WinVista */
More windows header stuff. Will it work this time?
2015-07-09 20:58:16 +02:00
# ifndef _WIN32_WINNT
Implement WIN32 tor_cond_wait using condition variables #30187 Fix bug where running a relay on Windows would use 100% CPU after some time. Makes Windows >= Vista the required Windows version to build and run tor.
2020-10-31 19:33:33 +01:00
# define _WIN32_WINNT 0x0600
More windows header stuff. Will it work this time?
2015-07-09 20:58:16 +02:00
# endif
/* Defined to avoid including some windows headers as part of Windows.h */
# ifndef WIN32_LEAN_AND_MEAN
# define WIN32_LEAN_AND_MEAN 1
# endif
#endif
])
fixup! Move format_win32_error into lib/log/
2018-06-27 18:35:48 +02:00
AM_CONDITIONAL(WIN32, test "x$bwin32" = "xtrue")
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
AM_CONDITIONAL(BUILD_NT_SERVICES, test "x$bwin32" = "xtrue")
Add a small library to emulate tor_run_main() with exec()
2017-10-19 01:34:19 +02:00
AM_CONDITIONAL(BUILD_LIBTORRUNNER, test "x$bwin32" != "xtrue")
r8920@totoro: nickm | 2006-10-07 11:40:06 -0400 Fix IRIX builds so that IRIX no longer self-detects as win32. For once, this is not the fault of Windows: Apparently, some helpful person decided that the the MIPSpro C Compiler should give a warning on #error rather than actually failing. How... "innovative." (Bug found by pnx; thanks!) svn:r8635
2006-10-07 17:40:33 +02:00
r9326@totoro: nickm | 2006-11-15 16:17:35 -0500 Try to detect mipspro compiler. Add -c99 to cflags if we find it. svn:r8955
2006-11-15 22:17:48 +01:00
dnl Enable C99 when compiling with MIPSpro
AC_MSG_CHECKING([for MIPSpro compiler])
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM(, [
r9326@totoro: nickm | 2006-11-15 16:17:35 -0500 Try to detect mipspro compiler. Add -c99 to cflags if we find it. svn:r8955
2006-11-15 22:17:48 +01:00
#if (defined(__sgi) && defined(_COMPILER_VERSION))
#error
return x(y);
#endif
Fix MIPSpro and time_t signedness detection 3d6e2830876 silenced the autogen.sh warnings as it was supposed to, but introduced two bugs. Fix them.
2010-10-11 19:00:43 +02:00
])],
r12061@catbus: nickm | 2007-03-04 14:41:10 -0500 Run configure.in through autoupdate svn:r9729
2007-03-04 20:49:04 +01:00
bmipspro=false; AC_MSG_RESULT(no),
Fix MIPSpro and time_t signedness detection 3d6e2830876 silenced the autogen.sh warnings as it was supposed to, but introduced two bugs. Fix them.
2010-10-11 19:00:43 +02:00
bmipspro=true; AC_MSG_RESULT(yes))
r9326@totoro: nickm | 2006-11-15 16:17:35 -0500 Try to detect mipspro compiler. Add -c99 to cflags if we find it. svn:r8955
2006-11-15 22:17:48 +01:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$bmipspro" = "true"; then
r9326@totoro: nickm | 2006-11-15 16:17:35 -0500 Try to detect mipspro compiler. Add -c99 to cflags if we find it. svn:r8955
2006-11-15 22:17:48 +01:00
CFLAGS="$CFLAGS -c99"
fi
r15210@catbus: nickm | 2007-09-20 13:04:05 -0400 Re-optimize counter-mode: save about 15% on my core2 by (1) not regenerating the entire counter buffer every time we encrypt a block of keystream (2) using the platform-optimized htonl to convert to big-endian (It's a single instruction on 486 and later ) and (3) not even keeping a separate "counter" and "buffer" when the platform is big-endian. The third still needs testing. svn:r11536
2007-09-20 19:07:45 +02:00
AC_C_BIGENDIAN
-lm should not be hardcoded. On some platforms (Haiku/BeOS) libm lives in libcore. Also added 'network' to the list of libraries to search for connect().
2011-12-17 14:30:41 +01:00
AC_SEARCH_LIBS(socket, [socket network])
oops, and make make work too on solaris svn:r1316
2004-03-20 21:08:19 +01:00
AC_SEARCH_LIBS(gethostbyname, [nsl])
link with -ldl if you need it. svn:r5450
2005-11-23 21:15:39 +01:00
AC_SEARCH_LIBS(dlopen, [dl])
r9103@Kushana: nickm | 2006-10-05 23:27:35 -0400 Check whether inet_aton requires libresolv. Apparently this is another solaris thing. svn:r8602
2006-10-06 05:27:40 +02:00
AC_SEARCH_LIBS(inet_aton, [resolv])
FreeBSD needs -lexecinfo to get backtrace()
2015-09-25 02:59:59 +02:00
AC_SEARCH_LIBS(backtrace, [execinfo])
Fix --enable-static-tor on OpenBSD Previously we'd been using "we have clock_gettime()" as a proxy for "we need -lrt to link a static libevent". But that's not really accurate: we should only add -lrt if searching for clock_gettime function adds -lrt to our libraries.
2012-02-13 01:54:13 +01:00
saved_LIBS="$LIBS"
AC_SEARCH_LIBS([clock_gettime], [rt])
if test "$LIBS" != "$saved_LIBS"; then
# Looks like we need -lrt for clock_gettime().
have_rt=yes
fi
Fix bug 161: when --disable-threads is set, do not search for or link against pthreads libraries svn:r4545
2005-07-13 19:31:52 +02:00
Don't search for -lpthread on Windows If we're building for Windows, we want to use windows threads no matter what, and we don't want to link a pthread library even if it is present. Fixes bug 27081; bugfix on 1790dc67607799a in 0.1.0.1-rc.
2018-08-08 15:56:51 +02:00
if test "$bwin32" = "false"; then
AC_SEARCH_LIBS(pthread_create, [pthread])
AC_SEARCH_LIBS(pthread_detach, [pthread])
fi
Very experimental code to try to make libevent work from bad /usr/local/libs svn:r4238
2005-05-17 03:51:23 +02:00
Re-remove the --disable-threads configure option
2015-02-09 09:09:00 +01:00
AM_CONDITIONAL(THREADS_WIN32, test "$bwin32" = "true")
AM_CONDITIONAL(THREADS_PTHREADS, test "$bwin32" = "false")
Split threading-related code out of compat.c Also, re-enable the #if'd out condition-variable code. Work queues are going to make us hack on all of this stuff a bit more closely, so it might not be a terrible idea to make it easier to hack.
2013-09-23 03:30:46 +02:00
Split long lines in configure.in and Makefile.am files Having very long single lines with lots and lots of things in them tends to make files hard to diff and hard to merge. Since our tools are one-line-at-a-time, we should try to construct lists that way too, within reason. This incidentally turned up a few headers in configure.in that we were for some reason searching for twice.
2010-11-11 20:22:48 +01:00
AC_CHECK_FUNCS(
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
_NSGetEnviron \
Try a little harder to only use SecureZeroMemory when it's present We could be using AC_CHECK_FUNC_DECL too, but it shouldn't be needed.
2016-01-11 15:02:42 +01:00
RtlSecureZeroMemory \
SecureZeroMemory \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
accept4 \
backtrace \
backtrace_symbols_fd \
Isolate the "socketpair or a pipe" logic for alerting main thread This way we can use the linux eventfd extension where available. Using EVFILT_USER on the BSDs will be a teeny bit trickier, and will require libevent hacking.
2013-09-25 02:43:48 +02:00
eventfd \
Use memset_s or explicit_bzero when available.
2016-01-07 21:53:24 +01:00
explicit_bzero \
Use timingsafe_memcmp() where available. See ticket 17944; patch from "logan".
2015-12-29 15:43:01 +01:00
timingsafe_memcmp \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
flock \
err/log: Stop closing stderr and stdout during shutdown Closing these file descriptors can hide sanitiser logs. Instead, flush the logs before tor exits, using fsync(). Some Windows environments don't have fsync(), so we check for it at compile time. Fixes bug 33087; bugfix on 0.4.1.6.
2020-02-12 03:47:15 +01:00
fsync \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
ftime \
get_current_dir_name \
getaddrinfo \
Integrate getdelim() and getline() support into Tor.
2018-07-10 16:23:29 +02:00
getdelim \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
getifaddrs \
Integrate getdelim() and getline() support into Tor.
2018-07-10 16:23:29 +02:00
getline \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
getrlimit \
gettimeofday \
gmtime_r \
Check the libc version to decide whether to allow openat.
2017-11-16 20:06:38 +01:00
gnu_get_libc_version \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
inet_aton \
ioctl \
issetugid \
llround \
localtime_r \
lround \
Code for anonymous mappings via mmap() or CreateFileMapping(). Using an anonymous mmap() is a good way to get pages that we can set kernel-level flags on, like minherit() or madvise() or mlock(). We're going to use that so that we can make uninheritable locked pages to store PRNG data.
2019-02-05 22:58:56 +01:00
madvise \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
memmem \
memset_s \
Code for anonymous mappings via mmap() or CreateFileMapping(). Using an anonymous mmap() is a good way to get pages that we can set kernel-level flags on, like minherit() or madvise() or mlock(). We're going to use that so that we can make uninheritable locked pages to store PRNG data.
2019-02-05 22:58:56 +01:00
minherit \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
mmap \
Isolate the "socketpair or a pipe" logic for alerting main thread This way we can use the linux eventfd extension where available. Using EVFILT_USER on the BSDs will be a teeny bit trickier, and will require libevent hacking.
2013-09-25 02:43:48 +02:00
pipe \
pipe2 \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
prctl \
Add a tor_getpass to read passphrases. Needs better backend.
2015-03-01 15:35:36 +01:00
readpassphrase \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
rint \
sigaction \
Prefer use of __MINGW_PRINTF/SCANF_FORMAT if available. Mingw headers sometimes like to define alternative scanf/printf format attributes depending on whether they're using clang, UCRT, MINGW_ANSI_STDIO, or the microsoft version of printf/scanf. This change attempts to use the right one on the given platform. This is an attempt to fix part of #40355.
2021-11-05 17:08:48 +01:00
snprintf \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
socketpair \
Add a compat function to check how much disk space is free. Closes ticket 16734.
2015-08-05 20:01:49 +02:00
statvfs \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
strncasecmp \
strcasecmp \
strlcat \
strlcpy \
Define a strnlen replacement on platforms (win32) that lack it Right now this is only needed for test_util_format_time_interval, so define it as a static function. We can move it into compat later if we need to.
2014-10-13 20:59:17 +02:00
strnlen \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
strptime \
strtok_r \
strtoull \
sysconf \
get_total_system_memory(): see how much RAM we have
2014-04-03 17:46:01 +02:00
sysctl \
Use native timegm when available. Continue having a tor_gmtime_impl() unit test so that we can detect any problems in our replacement function; add a new test function to make sure that gmtime<->timegm are a round-trip on now-ish times. This is a fix for bug #40383, wherein we ran into trouble because tor_timegm() does not believe that time_t should include a count of leap seconds, but FreeBSD's gmtime believes that it should. This disagreement meant that for a certain amount of time each day, instead of calculating the most recent midnight, our voting-schedule functions would calculate the second-most-recent midnight, and lead to an assertion failure. I am calling this a bugfix on 0.2.0.3-alpha when we first started calculating our voting schedule in this way.
2021-06-01 22:18:23 +02:00
timegm \
Windows lacks truncate(3). Fix the new crypto tests, which used truncate(3).
2016-05-16 15:25:19 +02:00
truncate \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
uname \
Basic unit test for condition variables.
2013-09-28 05:15:53 +02:00
usleep \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
vasprintf \
Prefer use of __MINGW_PRINTF/SCANF_FORMAT if available. Mingw headers sometimes like to define alternative scanf/printf format attributes depending on whether they're using clang, UCRT, MINGW_ANSI_STDIO, or the microsoft version of printf/scanf. This change attempts to use the right one on the given platform. This is an attempt to fix part of #40355.
2021-11-05 17:08:48 +01:00
_vscprintf \
vsnprintf
Split long lines in configure.in and Makefile.am files Having very long single lines with lots and lots of things in them tends to make files hard to diff and hard to merge. Since our tools are one-line-at-a-time, we should try to construct lists that way too, within reason. This incidentally turned up a few headers in configure.in that we were for some reason searching for twice.
2010-11-11 20:22:48 +01:00
)
r18793@catbus: nickm | 2008-03-13 14:09:19 -0400 Add a malloc_good_size() implementation to OpenBSD_malloc_Linux.c. Also, make configure.in not use support functions for the platform malloc when we are not using the platform mallocs. svn:r14010
2008-03-13 19:11:33 +01:00
Ignore mach_approximate_time when targeting macOS 10.9 and earlier Work around two different bugs in the OS X 10.10 and later SDKs that would prevent us from successfully targeting earlier versions of OS X. Fixes bug 26876; bugfix on 0.3.3.1-alpha.
2018-07-20 04:26:33 +02:00
# Apple messed up when they added some functions: they
Try to work around breakage in the OSX 10.12 SDK. Apple is supposed to decorate their function declarations with macros to indicate which OSX version introduced them, so that you can tell the compiler that you want to build against certain versions of OSX. But they forgot to do that for clock_gettime() and getentropy(), both of which they introduced in 10.12. This patch adds a kludge to the configure.ac script where, if we detect that we are targeting OSX 10.11 or earlier, we don't even probe to see if the two offending functions are present. Closes ticket 20235.
2016-11-23 01:24:13 +01:00
# forgot to decorate them with appropriate AVAILABLE_MAC_OS_VERSION
Ignore mach_approximate_time when targeting macOS 10.9 and earlier Work around two different bugs in the OS X 10.10 and later SDKs that would prevent us from successfully targeting earlier versions of OS X. Fixes bug 26876; bugfix on 0.3.3.1-alpha.
2018-07-20 04:26:33 +02:00
# checks.
# We should only probe for these functions if we are sure that we
# are not targeting OS X 10.9 or earlier.
AC_MSG_CHECKING([for a pre-Yosemite OS X build target])
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
#ifdef __APPLE__
# include <AvailabilityMacros.h>
# ifndef MAC_OS_X_VERSION_10_10
# define MAC_OS_X_VERSION_10_10 101000
# endif
# if defined(MAC_OS_X_VERSION_MIN_REQUIRED)
# if MAC_OS_X_VERSION_MIN_REQUIRED < MAC_OS_X_VERSION_10_10
# error "Running on Mac OS X 10.9 or earlier"
# endif
# endif
#endif
]], [[]])],
[on_macos_pre_10_10=no ; AC_MSG_RESULT([no])],
[on_macos_pre_10_10=yes; AC_MSG_RESULT([yes])])
if test "$on_macos_pre_10_10" = "no"; then
AC_CHECK_FUNCS(
mach_approximate_time \
)
fi
# We should only probe for these functions if we are sure that we
Fix spelling mistakes corresponding to ticket #23650
2018-01-24 09:55:15 +01:00
# are not targeting OSX 10.11 or earlier.
Try to work around breakage in the OSX 10.12 SDK. Apple is supposed to decorate their function declarations with macros to indicate which OSX version introduced them, so that you can tell the compiler that you want to build against certain versions of OSX. But they forgot to do that for clock_gettime() and getentropy(), both of which they introduced in 10.12. This patch adds a kludge to the configure.ac script where, if we detect that we are targeting OSX 10.11 or earlier, we don't even probe to see if the two offending functions are present. Closes ticket 20235.
2016-11-23 01:24:13 +01:00
AC_MSG_CHECKING([for a pre-Sierra OSX build target])
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
Try to work around breakage in the OSX 10.12 SDK. Apple is supposed to decorate their function declarations with macros to indicate which OSX version introduced them, so that you can tell the compiler that you want to build against certain versions of OSX. But they forgot to do that for clock_gettime() and getentropy(), both of which they introduced in 10.12. This patch adds a kludge to the configure.ac script where, if we detect that we are targeting OSX 10.11 or earlier, we don't even probe to see if the two offending functions are present. Closes ticket 20235.
2016-11-23 01:24:13 +01:00
#ifdef __APPLE__
# include <AvailabilityMacros.h>
Spell MAC_OS_X_VERSION_10_12 correctly. Fixes 20935.
2016-12-12 04:17:14 +01:00
# ifndef MAC_OS_X_VERSION_10_12
# define MAC_OS_X_VERSION_10_12 101200
Try to work around breakage in the OSX 10.12 SDK. Apple is supposed to decorate their function declarations with macros to indicate which OSX version introduced them, so that you can tell the compiler that you want to build against certain versions of OSX. But they forgot to do that for clock_gettime() and getentropy(), both of which they introduced in 10.12. This patch adds a kludge to the configure.ac script where, if we detect that we are targeting OSX 10.11 or earlier, we don't even probe to see if the two offending functions are present. Closes ticket 20235.
2016-11-23 01:24:13 +01:00
# endif
# if defined(MAC_OS_X_VERSION_MIN_REQUIRED)
Spell MAC_OS_X_VERSION_10_12 correctly. Fixes 20935.
2016-12-12 04:17:14 +01:00
# if MAC_OS_X_VERSION_MIN_REQUIRED < MAC_OS_X_VERSION_10_12
Try to work around breakage in the OSX 10.12 SDK. Apple is supposed to decorate their function declarations with macros to indicate which OSX version introduced them, so that you can tell the compiler that you want to build against certain versions of OSX. But they forgot to do that for clock_gettime() and getentropy(), both of which they introduced in 10.12. This patch adds a kludge to the configure.ac script where, if we detect that we are targeting OSX 10.11 or earlier, we don't even probe to see if the two offending functions are present. Closes ticket 20235.
2016-11-23 01:24:13 +01:00
# error "Running on Mac OSX 10.11 or earlier"
# endif
# endif
#endif
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]], [[]])],
Try to work around breakage in the OSX 10.12 SDK. Apple is supposed to decorate their function declarations with macros to indicate which OSX version introduced them, so that you can tell the compiler that you want to build against certain versions of OSX. But they forgot to do that for clock_gettime() and getentropy(), both of which they introduced in 10.12. This patch adds a kludge to the configure.ac script where, if we detect that we are targeting OSX 10.11 or earlier, we don't even probe to see if the two offending functions are present. Closes ticket 20235.
2016-11-23 01:24:13 +01:00
[on_macos_pre_10_12=no ; AC_MSG_RESULT([no])],
[on_macos_pre_10_12=yes; AC_MSG_RESULT([yes])])
if test "$on_macos_pre_10_12" = "no"; then
AC_CHECK_FUNCS(
clock_gettime \
getentropy \
)
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$bwin32" != "true"; then
r9384@Kushana: nickm | 2006-10-24 22:51:42 -0400 Libevent-1.2 exports, but does not define in its headers, strlcpy. Try to fix this in configure.in by checking for most functions before we check for libevent. svn:r8826
2006-10-25 04:51:45 +02:00
AC_CHECK_HEADERS(pthread.h)
AC_CHECK_FUNCS(pthread_create)
Don't call pthread_condattr_setclock() unless it exists Fixes bug 17819; bugfix on 0.2.6.3-alpha (specifically, d684dbb0).
2015-12-16 15:23:44 +01:00
AC_CHECK_FUNCS(pthread_condattr_setclock)
r9384@Kushana: nickm | 2006-10-24 22:51:42 -0400 Libevent-1.2 exports, but does not define in its headers, strlcpy. Try to fix this in configure.in by checking for most functions before we check for libevent. svn:r8826
2006-10-25 04:51:45 +02:00
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$bwin32" = "true"; then
Nth time is maybe the charm for fixing windows readpassword build errors
2015-07-12 20:34:11 +02:00
AC_CHECK_DECLS([SecureZeroMemory, _getwch], , , [
#include <windows.h>
#include <conio.h>
#include <wchar.h>
])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
AM_CONDITIONAL(BUILD_READPASSPHRASE_C,
test "x$ac_cv_func_readpassphrase" = "xno" && test "$bwin32" = "false")
Add the openssh 6.8p1 readpassphrase implementation This way glibc users don't have to fall back to getpass. Windows users are still out of luck
2015-06-17 16:41:22 +02:00
Check if glob() is available at build-time. This patch disables the glob() support in the path library if glob() is unavailable at build-time. This currently happens with the Android NDK used for Tor Browser. See: https://bugs.torproject.org/tpo/core/tor/40114
2020-09-10 00:08:54 +02:00
AC_CHECK_FUNCS(glob)
Remove workaround code for systems where free(NULL) is busted. Add an autoconf test to make sure we won't regret it. Closes ticket 24484.
2018-01-25 23:04:12 +01:00
AC_MSG_CHECKING([whether free(NULL) works])
AC_RUN_IFELSE([AC_LANG_PROGRAM([
#include <stdlib.h>
], [
char *p = NULL;
free(p);
])],
[free_null_ok=true; AC_MSG_RESULT(yes)],
[free_null_ok=false; AC_MSG_RESULT(no)],
[free_null_ok=cross; AC_MSG_RESULT(cross)])
if test "$free_null_ok" = "false"; then
AC_MSG_ERROR([Your libc implementation doesn't allow free(NULL), as required by C99.])
fi
Very experimental code to try to make libevent work from bad /usr/local/libs svn:r4238
2005-05-17 03:51:23 +02:00
dnl ------------------------------------------------------
More comments for configure.in, and also make configure.in not croak on old libevents svn:r4287
2005-05-23 08:24:00 +02:00
dnl Where do you live, libevent? And how do we call you?
Very experimental code to try to make libevent work from bad /usr/local/libs svn:r4238
2005-05-17 03:51:23 +02:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$bwin32" = "true"; then
r12611@catbus: nickm | 2007-04-30 22:49:00 -0400 For reasons which make sense to somebody, I'm sure, mingw gcc wants the libraries to appear at the end of the command line. This is done by specifying them with LDADD in Makefile.am, not LDFLAGS. If anybody can explain to me why mingw thinks "gcc -o foo foo.o -lbar" is fine, whereas "gcc -lbar -o foo foo.o" is Doubleplusbad UnMingwThink, I'd quite appreciate it. Until then, I'll just do what seems to work, and hope we don't blunder across any other great slumbering cthonian deities of arbitrary syntax. svn:r10082
2007-05-01 04:53:32 +02:00
TOR_LIB_WS32=-lws2_32
Correct reference to libiphlpapi from libiphlapi
2011-08-22 18:38:22 +02:00
TOR_LIB_IPHLPAPI=-liphlpapi
Add support for patterns on %include #25140 Also adds generic tor_glob function to expand globs.
2020-06-03 23:09:42 +02:00
TOR_LIB_SHLWAPI=-lshlwapi
r12607@catbus: nickm | 2007-04-30 21:36:28 -0400 More attempt to fix win32 building. This time, with extra linking. svn:r10080
2007-05-01 03:36:32 +02:00
TOR_LIB_GDI=-lgdi32
22839: Build tor with rust enabled on win - make tor_util static library name configurable - fix Rust libary dependency order for Windows
2017-08-10 10:08:41 +02:00
TOR_LIB_USERENV=-luserenv
build: updated configure.ac to allow mingw builds with static openssl and libevent
2020-09-14 19:01:43 +02:00
TOR_LIB_BCRYPT=-lbcrypt
TOR_LIB_CRYPT32=-lcrypt32
svn:r9728
2007-03-04 20:49:02 +01:00
else
r12607@catbus: nickm | 2007-04-30 21:36:28 -0400 More attempt to fix win32 building. This time, with extra linking. svn:r10080
2007-05-01 03:36:32 +02:00
TOR_LIB_WS32=
TOR_LIB_GDI=
22839: Build tor with rust enabled on win - make tor_util static library name configurable - fix Rust libary dependency order for Windows
2017-08-10 10:08:41 +02:00
TOR_LIB_USERENV=
svn:r9728
2007-03-04 20:49:02 +01:00
fi
r12607@catbus: nickm | 2007-04-30 21:36:28 -0400 More attempt to fix win32 building. This time, with extra linking. svn:r10080
2007-05-01 03:36:32 +02:00
AC_SUBST(TOR_LIB_WS32)
AC_SUBST(TOR_LIB_GDI)
Correct reference to libiphlpapi from libiphlapi
2011-08-22 18:38:22 +02:00
AC_SUBST(TOR_LIB_IPHLPAPI)
build: updated configure.ac to allow mingw builds with static openssl and libevent
2020-09-14 19:01:43 +02:00
AC_SUBST(TOR_LIB_BCRYPT)
AC_SUBST(TOR_LIB_CRYPT32)
Add support for patterns on %include #25140 Also adds generic tor_glob function to expand globs.
2020-06-03 23:09:42 +02:00
AC_SUBST(TOR_LIB_SHLWAPI)
22839: Build tor with rust enabled on win - make tor_util static library name configurable - fix Rust libary dependency order for Windows
2017-08-10 10:08:41 +02:00
AC_SUBST(TOR_LIB_USERENV)
svn:r9728
2007-03-04 20:49:02 +01:00
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
tor_libevent_pkg_redhat="libevent"
We need the -dev package in both of these configure cases svn:r10365
2007-05-27 17:20:38 +02:00
tor_libevent_pkg_debian="libevent-dev"
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
tor_libevent_devpkg_redhat="libevent-devel"
tor_libevent_devpkg_debian="libevent-dev"
Fix libevent autoconf bug #2698
2011-03-17 02:35:31 +01:00
dnl On Gnu/Linux or any place we require it, we'll add librt to the Libevent
dnl linking for static builds.
STATIC_LIBEVENT_FLAGS=""
if test "$enable_static_libevent" = "yes"; then
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$have_rt" = "yes"; then
Fix libevent autoconf bug #2698
2011-03-17 02:35:31 +01:00
STATIC_LIBEVENT_FLAGS=" -lrt "
fi
fi
build: updated configure.ac to allow mingw builds with static openssl and libevent
2020-09-14 19:01:43 +02:00
TOR_SEARCH_LIBRARY(libevent, $trylibeventdir, [-levent $STATIC_LIBEVENT_FLAGS $TOR_LIB_IPHLPAPI $TOR_LIB_BCRYPT $TOR_LIB_WS32], [
Change WIN32 to _WIN32 in configure.in, too.
2012-01-31 18:05:03 +01:00
#ifdef _WIN32
r19797@catbus: nickm | 2008-05-16 15:42:53 -0400 Try to build correctly on win32 with libevent versions 1.4.x or greater. 1.4.5 should remove the need to do this, but hey. Backport candidate. svn:r14640
2008-05-16 21:44:57 +02:00
#include <winsock2.h>
#endif
r11596@catbus: nickm | 2007-01-30 01:43:40 -0500 Macro-ize our "search for a library and find out how to link against it" code. This will help make our handling of libevent and openssl more similar, and with any luck better. Not pushing to main branch until after 0.1.2. svn:r9726
2007-03-04 20:47:34 +01:00
#include <sys/time.h>
resume building on irix64 svn:r6198
2006-03-19 21:36:26 +01:00
#include <sys/types.h>
Raise libevent dependency to 2.0.10-stable or newer Only some very ancient distributions don't ship with Libevent 2 anymore, even the oldest supported Ubuntu LTS version has it. This allows us to get rid of a lot of compat code.
2016-07-03 19:42:36 +02:00
#include <event2/event.h>], [
Change WIN32 to _WIN32 in configure.in, too.
2012-01-31 18:05:03 +01:00
#ifdef _WIN32
More muddling around to get configure.in to recognize recent libevents. Really, libevent should be fixed too so that the next poor sucker doesn't have to do this. svn:r14641
2008-05-16 22:13:10 +02:00
#include <winsock2.h>
#endif
Use event_base_new(), not event_init(), to detect libevent 2. (event_init() is obsoleted in libevent 2.) Fixes bug 21051; bugfix on 0.2.9.1-alpha when we dropped libevent 1 support.
2016-12-23 14:18:31 +01:00
struct event_base;
Add a missing prototype to our libevent configure stanza. Fixes bug 25474; bugfix on 0.3.2.5-alpha.
2018-03-13 13:37:26 +01:00
struct event_base *event_base_new(void);
void event_base_free(struct event_base *);],
r19797@catbus: nickm | 2008-05-16 15:42:53 -0400 Try to build correctly on win32 with libevent versions 1.4.x or greater. 1.4.5 should remove the need to do this, but hey. Backport candidate. svn:r14640
2008-05-16 21:44:57 +02:00
[
Change WIN32 to _WIN32 in configure.in, too.
2012-01-31 18:05:03 +01:00
#ifdef _WIN32
More muddling around to get configure.in to recognize recent libevents. Really, libevent should be fixed too so that the next poor sucker doesn't have to do this. svn:r14641
2008-05-16 22:13:10 +02:00
{WSADATA d; WSAStartup(0x101,&d); }
r19797@catbus: nickm | 2008-05-16 15:42:53 -0400 Try to build correctly on win32 with libevent versions 1.4.x or greater. 1.4.5 should remove the need to do this, but hey. Backport candidate. svn:r14640
2008-05-16 21:44:57 +02:00
#endif
Fix configure libevent memory leak Breaks build with externally-specified hardening flags.
2017-10-29 14:14:58 +01:00
event_base_free(event_base_new());
r19797@catbus: nickm | 2008-05-16 15:42:53 -0400 Try to build correctly on win32 with libevent versions 1.4.x or greater. 1.4.5 should remove the need to do this, but hey. Backport candidate. svn:r14640
2008-05-16 21:44:57 +02:00
], [--with-libevent-dir], [/opt/libevent])
Very experimental code to try to make libevent work from bad /usr/local/libs svn:r4238
2005-05-17 03:51:23 +02:00
Only use -levent when checking functions if we will use it to link. Fixes 19904; bugfix on b62abf9f21499ab; patch from Rubiate.
2016-10-06 15:16:21 +02:00
dnl Determine the incantation needed to link libevent.
r12594@catbus: nickm | 2007-04-30 18:31:48 -0400 Oops: detect event_get_version and friends correctly again. svn:r10073
2007-05-01 00:42:47 +02:00
save_LIBS="$LIBS"
save_LDFLAGS="$LDFLAGS"
r12984@catbus: nickm | 2007-05-25 19:59:41 -0400 Do not alter CPPFLAGS until we are actually done with the rest of configure.in: this way, we notice if there are library paths that we are missing. svn:r10345
2007-05-26 01:59:47 +02:00
save_CPPFLAGS="$CPPFLAGS"
Only use -levent when checking functions if we will use it to link. Fixes 19904; bugfix on b62abf9f21499ab; patch from Rubiate.
2016-10-06 15:16:21 +02:00
LIBS="$STATIC_LIBEVENT_FLAGS $TOR_LIB_WS32 $save_LIBS"
Fix a variable handling mistake when testing for libevent functions in configure.in. Found by Riastradh. svn:r16793
2008-09-08 03:57:02 +02:00
LDFLAGS="$TOR_LDFLAGS_libevent $LDFLAGS"
r12984@catbus: nickm | 2007-05-25 19:59:41 -0400 Do not alter CPPFLAGS until we are actually done with the rest of configure.in: this way, we notice if there are library paths that we are missing. svn:r10345
2007-05-26 01:59:47 +02:00
CPPFLAGS="$TOR_CPPFLAGS_libevent $CPPFLAGS"
Update Tor to use Libevent 2.0 APIs when available. This patch adds a new compat_libevent.[ch] set of files, and moves our Libevent compatibility and utilitity functions there. We build them into a separate .a so that nothing else in src/commmon depends on Libevent (partially fixing bug 507). Also, do not use our own built-in evdns copy when we have Libevent 2.0, whose evdns is finally good enough (thus fixing Bug 920).
2009-06-04 07:05:23 +02:00
Fix libevent linking on openbsd. Closes ticket 19902; bugfix on 0.2.9.1-alpha; patch from rubiate
2016-09-08 16:08:29 +02:00
AC_CHECK_HEADERS(event2/event.h event2/dns.h event2/bufferevent_ssl.h)
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
if test "$enable_static_libevent" = "yes"; then
if test "$tor_cv_library_libevent_dir" = "(system)"; then
AC_MSG_ERROR("You must specify an explicit --with-libevent-dir=x option when using --enable-static-libevent")
else
configure: Revert back the static libevent path From c618c4f27986825df2026fb22b8ee222edd1afa1, we changed the static libevent.a path to be able to use the git repository or tarball directly but that broke the "make install" setup that Tor Browser is using with Android. In other words, the git repository and tarball put the "libevent.a" in ".libs/" where "make install" puts it in "lib/". Using the --with-libevent-dir=..., which is mandatory for static libevent, autoconf will take the path and use it for the includes (-I) and library (-L) for which if it finds a "include/" and a "lib/" in the root, it will use those. However, with the git repo or tarball, the "lib/" doesn't exists thus autoconf sets the library search path to be at the root and thus fails to find the libevent.a in ".libs/". This is a whole lot more work to make both cases work in our configure.ac thus I'm reverting the change here to the Tor Browser case works again and the work around for others is to either symlink the libevent.a at the root or use a temporary make install directory. One long term fix here would likely be to ask libevent to symblink the .a into the root along the .la files and likely do the same for .so. Or, use the "lib/" structure to contain the .a + .so files. Would be better than doing ninji-tsu in our configure.ac Fixes #40225 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-12-11 15:52:23 +01:00
TOR_LIBEVENT_LIBS="$TOR_LIBDIR_libevent/libevent.a $STATIC_LIBEVENT_FLAGS"
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
fi
else
Quote variables in case they contain spaces
2015-11-26 13:18:40 +01:00
if test "x$ac_cv_header_event2_event_h" = "xyes"; then
Try to fix windows build patch from rubiate on #16651
2015-12-09 14:59:01 +01:00
AC_SEARCH_LIBS(event_new, [event event_core], , AC_MSG_ERROR("libevent2 is installed but linking it failed while searching for event_new"))
AC_SEARCH_LIBS(evdns_base_new, [event event_extra], , AC_MSG_ERROR("libevent2 is installed but linking it failed while searching for evdns_base_new"))
Attempt to make openbsd compilation happier with libevent2 installed Fix for bug 16651; patch from "rubiate".
2015-11-25 15:43:12 +01:00
Quote variables in case they contain spaces
2015-11-26 13:18:40 +01:00
if test "$ac_cv_search_event_new" != "none required"; then
Attempt to make openbsd compilation happier with libevent2 installed Fix for bug 16651; patch from "rubiate".
2015-11-25 15:43:12 +01:00
TOR_LIBEVENT_LIBS="$ac_cv_search_event_new"
fi
Quote variables in case they contain spaces
2015-11-26 13:18:40 +01:00
if test "$ac_cv_search_evdns_base_new" != "none required"; then
Attempt to make openbsd compilation happier with libevent2 installed Fix for bug 16651; patch from "rubiate".
2015-11-25 15:43:12 +01:00
TOR_LIBEVENT_LIBS="$ac_cv_search_evdns_base_new $TOR_LIBEVENT_LIBS"
fi
else
Fix libevent linking on openbsd. Closes ticket 19902; bugfix on 0.2.9.1-alpha; patch from rubiate
2016-09-08 16:08:29 +02:00
AC_MSG_ERROR("libevent2 is required but the headers could not be found")
Attempt to make openbsd compilation happier with libevent2 installed Fix for bug 16651; patch from "rubiate".
2015-11-25 15:43:12 +01:00
fi
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
fi
Only use -levent when checking functions if we will use it to link. Fixes 19904; bugfix on b62abf9f21499ab; patch from Rubiate.
2016-10-06 15:16:21 +02:00
dnl Now check for particular libevent functions.
AC_CHECK_FUNCS([evutil_secure_rng_set_urandom_device_file \
evutil_secure_rng_add_bytes \
Check if libevent comes with evdns_base_get_nameserver_addr()
2018-10-20 19:18:54 +02:00
evdns_base_get_nameserver_addr \
Only use -levent when checking functions if we will use it to link. Fixes 19904; bugfix on b62abf9f21499ab; patch from Rubiate.
2016-10-06 15:16:21 +02:00
])
Attempt to make openbsd compilation happier with libevent2 installed Fix for bug 16651; patch from "rubiate".
2015-11-25 15:43:12 +01:00
LIBS="$save_LIBS"
LDFLAGS="$save_LDFLAGS"
CPPFLAGS="$save_CPPFLAGS"
Raise libevent dependency to 2.0.10-stable or newer Only some very ancient distributions don't ship with Libevent 2 anymore, even the oldest supported Ubuntu LTS version has it. This allows us to get rid of a lot of compat code.
2016-07-03 19:42:36 +02:00
dnl Check that libevent is at least at version 2.0.10, the first stable
dnl release of its series
CPPFLAGS="$CPPFLAGS $TOR_CPPFLAGS_libevent"
AC_MSG_CHECKING([whether Libevent is new enough])
AC_COMPILE_IFELSE([AC_LANG_SOURCE([
Detect Libevent version at configure time when using bufferevents
2010-09-27 19:24:19 +02:00
#include <event2/event.h>
Raise libevent dependency to 2.0.10-stable or newer Only some very ancient distributions don't ship with Libevent 2 anymore, even the oldest supported Ubuntu LTS version has it. This allows us to get rid of a lot of compat code.
2016-07-03 19:42:36 +02:00
#if !defined(LIBEVENT_VERSION_NUMBER) || LIBEVENT_VERSION_NUMBER < 0x02000a00
Detect Libevent version at configure time when using bufferevents
2010-09-27 19:24:19 +02:00
#error
int x = y(zz);
#else
int x = 1;
#endif
Raise libevent dependency to 2.0.10-stable or newer Only some very ancient distributions don't ship with Libevent 2 anymore, even the oldest supported Ubuntu LTS version has it. This allows us to get rid of a lot of compat code.
2016-07-03 19:42:36 +02:00
])], [ AC_MSG_RESULT([yes]) ],
[ AC_MSG_RESULT([no])
AC_MSG_ERROR([Libevent is not new enough. We require 2.0.10-stable or later]) ] )
LIBS="$save_LIBS"
LDFLAGS="$save_LDFLAGS"
CPPFLAGS="$save_CPPFLAGS"
Link to libevent_openssl statically when requested When configure tor with --enable-bufferevents and --enable-static-libevent, libevent_openssl would still be linked dynamically. Fix this and refactor src/or/Makefile.am along the way.
2011-05-06 15:34:00 +02:00
AC_SUBST(TOR_LIBEVENT_LIBS)
Teach read_event/write_event manipulators about bufferevents. Add an --enable-bufferevents config switch.
2009-07-31 17:11:45 +02:00
-lm should not be hardcoded. On some platforms (Haiku/BeOS) libm lives in libcore. Also added 'network' to the list of libraries to search for connect().
2011-12-17 14:30:41 +01:00
dnl ------------------------------------------------------
dnl Where do you live, libm?
dnl On some platforms (Haiku/BeOS) the math library is
dnl part of libroot. In which case don't link against lm
TOR_LIB_MATH=""
save_LIBS="$LIBS"
Tweak the haiku-support patches
2011-12-19 17:26:50 +01:00
AC_SEARCH_LIBS(pow, [m], , AC_MSG_ERROR([Could not find pow in libm or libc.]))
if test "$ac_cv_search_pow" != "none required"; then
TOR_LIB_MATH="$ac_cv_search_pow"
-lm should not be hardcoded. On some platforms (Haiku/BeOS) libm lives in libcore. Also added 'network' to the list of libraries to search for connect().
2011-12-17 14:30:41 +01:00
fi
LIBS="$save_LIBS"
AC_SUBST(TOR_LIB_MATH)
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
dnl ------------------------------------------------------
dnl Hello, NSS. You're new around here.
if test "x$enable_nss" = "xyes"; then
PKG_CHECK_MODULES(NSS,
[nss],
[have_nss=yes],
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
[have_nss=no; AC_MSG_ERROR([You asked for NSS but I can't find it, $pkg_config_user_action, or set NSS_CFLAGS and NSS_LIBS.])])
Add a configure switch to build with NSS. When it is set, include the NSS headers and libraries as appropriate. Doesn't actually use them yet, though.
2018-07-11 18:57:30 +02:00
AC_SUBST(NSS_CFLAGS)
AC_SUBST(NSS_LIBS)
fi
Very experimental code to try to make libevent work from bad /usr/local/libs svn:r4238
2005-05-17 03:51:23 +02:00
dnl ------------------------------------------------------
Make OpenSSL configure.in logic match libevent configure.in logic. svn:r5873
2006-01-27 20:38:16 +01:00
dnl Where do you live, openssl? And how do we call you?
make configure work on solaris without mods svn:r1315
2004-03-20 21:00:58 +01:00
When enabling NSS, disable OpenSSL. We used to link both libraries at once, but now that I'm working on TLS, there's nothing left to keep OpenSSL around for when NSS is enabled. Note that this patch causes a couple of places that still assumed OpenSSL to be disabled when NSS is enabled - tor-gencert - pbkdf2
2018-08-12 23:18:41 +02:00
if test "x$enable_nss" != "xyes"; then
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
tor_openssl_pkg_redhat="openssl"
Updated openssl configuration advice for Debian
2013-09-02 20:54:50 +02:00
tor_openssl_pkg_debian="libssl-dev"
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
tor_openssl_devpkg_redhat="openssl-devel"
tor_openssl_devpkg_debian="libssl-dev"
r17853@tombo: nickm | 2008-08-22 12:55:00 -0400 Oops. We had been telling people to use --with-ssl-dir, which had not worked since 0.2.0.1-alpha: --with-openssl-dir was the one that worked. Make them both work, document --with-openssl-dir, and mark --with-ssl-dir as obsolete. Found by "Dave". svn:r16625
2008-08-22 18:55:28 +02:00
ALT_openssl_WITHVAL=""
AC_ARG_WITH(ssl-dir,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--with-ssl-dir=PATH, [obsolete alias for --with-openssl-dir]),
r17853@tombo: nickm | 2008-08-22 12:55:00 -0400 Oops. We had been telling people to use --with-ssl-dir, which had not worked since 0.2.0.1-alpha: --with-openssl-dir was the one that worked. Make them both work, document --with-openssl-dir, and mark --with-ssl-dir as obsolete. Found by "Dave". svn:r16625
2008-08-22 18:55:28 +02:00
[
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$withval" != "xno" && test "x$withval" != "x"; then
r17853@tombo: nickm | 2008-08-22 12:55:00 -0400 Oops. We had been telling people to use --with-ssl-dir, which had not worked since 0.2.0.1-alpha: --with-openssl-dir was the one that worked. Make them both work, document --with-openssl-dir, and mark --with-ssl-dir as obsolete. Found by "Dave". svn:r16625
2008-08-22 18:55:28 +02:00
ALT_openssl_WITHVAL="$withval"
fi
])
When detecting OpenSSL, look for 1.0.1, and in homebrew location Previously we would detect the system openssl on OSX, and then fail to use it, since we required Open 1.0.1 or later. That's silly! Instead of looking for RAND_add(), look for TLSv1_1_method(): it was introduced in 1.0.1, and is also present in LibreSSL. Also, add the hombebrew path to our search path here. Fixes bug 23602; bugfix on 0.2.7.2-alpha.
2017-09-20 21:54:36 +02:00
AC_MSG_NOTICE([Now, we'll look for OpenSSL >= 1.0.1])
build: updated configure.ac to allow mingw builds with static openssl and libevent
2020-09-14 19:01:43 +02:00
TOR_SEARCH_LIBRARY(openssl, $tryssldir, [-lssl -lcrypto $TOR_LIB_GDI $TOR_LIB_WS32 $TOR_LIB_CRYPT32],
Edit our openssl detection in autoconf to tolerate no-deprecated. When openssl is built with no-deprecated, the TLSv1_1_method() function isn't visible in the headers. That's sad, because that method is what we were looking at. Instead, we now look at SSL_CIPHER_get_id(), which is present in OpenSSL 1.0.1 and later, which is _not_ deprecated, and which is also present in LibreSSL. Fixes ticket 25353. Not a bugfix exactly -- we never really worked with this configuration.
2018-03-05 21:34:04 +01:00
[#include <openssl/ssl.h>
char *getenv(const char *);],
[struct ssl_cipher_st;
unsigned SSL_CIPHER_get_id(const struct ssl_cipher_st *);
char *getenv(const char *);],
dnl This funny-looking test program calls getenv, so that the compiler
dnl will neither make code that call SSL_CIPHER_get_id(NULL) [producing
dnl a crash], nor optimize out the call to SSL_CIPHER_get_id().
dnl We look for SSL_cipher_get_id() because it is present in
dnl OpenSSL >=1.0.1, because it is not deprecated, and because Tor
dnl depends on it.
[if (getenv("THIS_SHOULDNT_BE_SET_X201803")) SSL_CIPHER_get_id((void *)0);], [],
Remove /usr/athena from search path in configure.ac Fixes #24363 Signed-off-by: Fernando Fernandez Mancera <ffernandezmancera@gmail.com>
2017-11-21 16:53:37 +01:00
[/usr/local/opt/openssl /usr/local/openssl /usr/lib/openssl /usr/local/ssl /usr/lib/ssl /usr/local /opt/openssl])
Make OpenSSL configure.in logic match libevent configure.in logic. svn:r5873
2006-01-27 20:38:16 +01:00
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
if test "$enable_static_openssl" = "yes"; then
if test "$tor_cv_library_openssl_dir" = "(system)"; then
AC_MSG_ERROR("You must specify an explicit --with-openssl-dir=x option when using --enable-static-openssl")
else
build: updated configure.ac to allow mingw builds with static openssl and libevent
2020-09-14 19:01:43 +02:00
TOR_OPENSSL_LIBS="$TOR_LIBDIR_openssl/libssl.a $TOR_LIBDIR_openssl/libcrypto.a $TOR_LIB_WS32 $TOR_LIB_CRYPT32 $TOR_LIB_BCRYPT"
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
fi
else
Link libssl and libcrypto in the right order. For most linking setups, this doesn't matter. But for some setups, when statically linking openssl, it does matter, since you need to link things with dependencies before you link things they depend on. Fix for bug 1237.
2010-02-02 22:12:45 +01:00
TOR_OPENSSL_LIBS="-lssl -lcrypto"
Add --enable-static-(openssl|libevent) options These options only work when using --with-(openssl|libevent)-dir to explicitly pick a libevent or openssl location.
2010-01-24 20:34:47 +01:00
fi
AC_SUBST(TOR_OPENSSL_LIBS)
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
dnl Now validate openssl, and check for particular openssl functions.
Stop looking at session->ciphers when possible If the OpenSSL team accepts my patch to add an SSL_get_client_ciphers function, this patch will make Tor use it when available, thereby working better with openssl 1.1.
2015-05-26 17:05:36 +02:00
save_LIBS="$LIBS"
save_LDFLAGS="$LDFLAGS"
save_CPPFLAGS="$CPPFLAGS"
LIBS="$TOR_OPENSSL_LIBS $LIBS"
LDFLAGS="$TOR_LDFLAGS_openssl $LDFLAGS"
CPPFLAGS="$TOR_CPPFLAGS_openssl $CPPFLAGS"
Fail in configure when openssl is too old. #16901.
2015-09-01 15:02:12 +02:00
configure: disable OpenSSL deprecation warnings with OpenSSL >= 3 We can't do this in the C headers, since by the time we include `opensslv.h` in order to check the openssl version number, we will have included `openssl/macros.h`, which is the thing that checks whether we disabled deprecation warnings.
2020-10-28 15:30:24 +01:00
dnl Tor currently uses a number of APIs that are deprecated in OpenSSL 3.0.0
dnl and later. We want to migrate away from them, but that will be a lot of
dnl work. (See ticket tor#40166.) For now, we disable the deprecation
dnl warnings.
AC_MSG_CHECKING([for OpenSSL >= 3.0.0])
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
#include <openssl/opensslv.h>
Reverse the direction of the test for openssl 3.0.0 Previously the logic was reversed, and always gave the wrong answer. This has no other effect than to change whether we suppress deprecated API warnings. Fixes #40429; bugfix on 0.3.5.13.
2021-11-05 18:23:05 +01:00
#if !defined(LIBRESSL_VERSION_NUMBER) && OPENSSL_VERSION_NUMBER >= 0x30000000L
configure: disable OpenSSL deprecation warnings with OpenSSL >= 3 We can't do this in the C headers, since by the time we include `opensslv.h` in order to check the openssl version number, we will have included `openssl/macros.h`, which is the thing that checks whether we disabled deprecation warnings.
2020-10-28 15:30:24 +01:00
#error "you_have_version_3"
#endif
]], [[]])],
[ AC_MSG_RESULT([no]) ],
[ AC_MSG_RESULT([yes]);
AC_DEFINE(OPENSSL_SUPPRESS_DEPRECATED, 1, [disable openssl deprecated-function warnings]) ])
AC_MSG_CHECKING([for OpenSSL < 1.0.1])
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
Fail in configure when openssl is too old. #16901.
2015-09-01 15:02:12 +02:00
#include <openssl/opensslv.h>
LibreSSL 3.5 compatibility LibreSSL is now closer to OpenSSL 1.1 than OpenSSL 1.0. According to https://undeadly.org/cgi?action=article;sid=20220116121253, this is the intention of OpenBSD developers. According to #40630, many special cases are needed to compile Tor against LibreSSL 3.5 when using Tor's OpenSSL 1.0 compatibility mode, whereas only a small number of #defines are required when using OpenSSL 1.1 compatibility mode. One additional workaround is required for LibreSSL 3.4 compatibility. Compiles and passes unit tests with LibreSSL 3.4.3 and 3.5.1.
2022-07-05 17:37:30 +02:00
#if OPENSSL_VERSION_NUMBER < 0x1000100fL
Fail in configure when openssl is too old. #16901.
2015-09-01 15:02:12 +02:00
#error "too old"
#endif
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]], [[]])],
configure: disable OpenSSL deprecation warnings with OpenSSL >= 3 We can't do this in the C headers, since by the time we include `opensslv.h` in order to check the openssl version number, we will have included `openssl/macros.h`, which is the thing that checks whether we disabled deprecation warnings.
2020-10-28 15:30:24 +01:00
[ AC_MSG_RESULT([no]) ],
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
[ AC_MSG_ERROR([OpenSSL is too old. We require 1.0.1 or later. You can specify a path to a newer one with --with-openssl-dir.]) ])
Fail in configure when openssl is too old. #16901.
2015-09-01 15:02:12 +02:00
Give an error message if LibreSSL's TLSv1.3 APIs aren't what we need From LibreSSL versions 3.2.1 through 3.4.0, our configure script would conclude that TLSv1.3 as supported, but it actually wasn't. This led to annoying breakage like #40128 and #40445. Now we give an error message if we try to build with one of those versions. Closes #40511.
2021-11-06 16:04:08 +01:00
AC_MSG_CHECKING([whether LibreSSL TLS 1.3 APIs are busted])
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
#include <openssl/opensslv.h>
#if defined(LIBRESSL_VERSION_NUMBER) && \
LIBRESSL_VERSION_NUMBER >= 0x3020100fL && \
LIBRESSL_VERSION_NUMBER < 0x3040100fL
#error "oh no"
#endif
]], [[]])],
[ AC_MSG_RESULT([no]) ],
[ AC_MSG_ERROR([This version of LibreSSL won't work with Tor. Please upgrade to LibreSSL 3.4.1 or later. (Or downgrade to 3.2.0 if you really must.)]) ])
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
Check that openssl has ECC support during configure This allows builds on machines with a crippled openssl to fail early during configure. Bugfix on 0.2.7.1-alpha, which introduced the requirement for ECC support. Fixes bug 17109.
2015-09-19 16:36:55 +02:00
#include <openssl/opensslv.h>
#include <openssl/evp.h>
#if defined(OPENSSL_NO_EC) || defined(OPENSSL_NO_ECDH) || defined(OPENSSL_NO_ECDSA)
#error "no ECC"
#endif
#if !defined(NID_X9_62_prime256v1) || !defined(NID_secp224r1)
#error "curves unavailable"
#endif
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]], [[]])],
Check that openssl has ECC support during configure This allows builds on machines with a crippled openssl to fail early during configure. Bugfix on 0.2.7.1-alpha, which introduced the requirement for ECC support. Fixes bug 17109.
2015-09-19 16:36:55 +02:00
[ : ],
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
[ AC_MSG_ERROR([OpenSSL is built without full ECC support, including curves P256 and P224. You can specify a path to one with ECC support with --with-openssl-dir.]) ])
Check that openssl has ECC support during configure This allows builds on machines with a crippled openssl to fail early during configure. Bugfix on 0.2.7.1-alpha, which introduced the requirement for ECC support. Fixes bug 17109.
2015-09-19 16:36:55 +02:00
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
dnl Let's see if we have a version mismatch between includes and libs.
AC_MSG_CHECKING([for significant mismatch between openssl headers and libraries])
ac_retval=foo
Use AC_LANG_SOURCE
2020-09-23 16:53:09 +02:00
AC_RUN_IFELSE([AC_LANG_SOURCE([AC_LANG_PROGRAM([[
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
#include <openssl/opensslv.h>
#include <openssl/crypto.h>
]], [[
/* Include major, minor, and fix, but not patch or status. */
unsigned long mask = 0xfffff000;
unsigned long linking = OpenSSL_version_num() & mask;
unsigned long running = OPENSSL_VERSION_NUMBER & mask;
return !(linking==running);
Use AC_LANG_SOURCE
2020-09-23 16:53:09 +02:00
]])])], [openssl_ver_mismatch=no], [
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
# This is a kludge to figure out whether compilation failed, or whether
# running the program failed.
Make our configure.ac script use the portable `=` comparison Back in 0.4.5.1-alpha we added an `==` comparison, which isn't portable. Fixes bug #40298; bugfix on 0.4.5.1-alpha.
2021-02-17 18:36:23 +01:00
if test "$ac_retval" = "1"; then
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
openssl_ver_mismatch=inconclusive
else
openssl_ver_mismatch=yes
fi], [openssl_ver_mismatch=cross])
AC_MSG_RESULT([$openssl_ver_mismatch])
Run correctly on OpenBSD systems without SSL_METHOD.get_cipher_by_char Also, make sure we will compile correctly on systems where they finally rip it out. Fixes issue #13325. Caused by this openbsd commit: ​http://marc.info/?l=openbsd-cvs&m=140768179627976&w=2 Reported by Fredzupy.
2014-10-03 18:15:09 +02:00
AC_CHECK_MEMBERS([struct ssl_method_st.get_cipher_by_char], , ,
[#include <openssl/ssl.h>
])
Use openssl's version of sha3 when available. Part of 28837.
2018-12-13 19:25:30 +01:00
dnl OpenSSL functions which we might not have. In theory, we could just
dnl check the openssl version number, but in practice that gets pretty
dnl confusing with LibreSSL, OpenSSL, and various distributions' patches
dnl to them.
Stop looking at session->ciphers when possible If the OpenSSL team accepts my patch to add an SSL_get_client_ciphers function, this patch will make Tor use it when available, thereby working better with openssl 1.1.
2015-05-26 17:05:36 +02:00
AC_CHECK_FUNCS([ \
Only use OpenSSL kdf support if it is present. We have to check for ERR_load_KDF_strings() here, since that's the only one that's actually a function rather than a macro. Fixes compilation with LibreSSL. Fixes bug 26712; bug not in any released Tor.
2018-07-11 16:10:42 +02:00
ERR_load_KDF_strings \
Use openssl's version of sha3 when available. Part of 28837.
2018-12-13 19:25:30 +01:00
EVP_PBE_scrypt \
SSL_CIPHER_find \
SSL_CTX_set1_groups_list \
SSL_CTX_set_security_level \
Stop looking at session->ciphers when possible If the OpenSSL team accepts my patch to add an SSL_get_client_ciphers function, this patch will make Tor use it when available, thereby working better with openssl 1.1.
2015-05-26 17:05:36 +02:00
SSL_SESSION_get_master_key \
Use openssl's version of sha3 when available. Part of 28837.
2018-12-13 19:25:30 +01:00
SSL_get_client_ciphers \
SSL_get_client_random \
Stop looking at session->ciphers when possible If the OpenSSL team accepts my patch to add an SSL_get_client_ciphers function, this patch will make Tor use it when available, thereby working better with openssl 1.1.
2015-05-26 17:05:36 +02:00
SSL_get_server_random \
Use openssl's version of sha3 when available. Part of 28837.
2018-12-13 19:25:30 +01:00
TLS_method \
Stop looking at session->ciphers when possible If the OpenSSL team accepts my patch to add an SSL_get_client_ciphers function, this patch will make Tor use it when available, thereby working better with openssl 1.1.
2015-05-26 17:05:36 +02:00
])
Check if OpenSSL includes scrypt.
2015-07-06 20:31:01 +02:00
Support LibreSSL with opaque structures Determining if OpenSSL structures are opaque now uses an autoconf check instead of comparing the version number. Some definitions have been moved to their own check as assumptions which were true for OpenSSL with opaque structures did not hold for LibreSSL. Closes ticket 21359.
2017-02-01 12:10:46 +01:00
dnl Check if OpenSSL structures are opaque
AC_CHECK_MEMBERS([SSL.state], , ,
[#include <openssl/ssl.h>
])
In relay_digest_matches(), use stack instead of heap. We'd been using crypto_digest_dup() and crypto_digest_assign() here, but they aren't necessary. Instead we can just use the stack to store the previous state of the SHA_CTX and avoid a malloc/free pair. Closes ticket 24914.
2018-01-24 18:33:13 +01:00
AC_CHECK_SIZEOF(SHA_CTX, , [AC_INCLUDES_DEFAULT()
#include <openssl/sha.h>
])
When enabling NSS, disable OpenSSL. We used to link both libraries at once, but now that I'm working on TLS, there's nothing left to keep OpenSSL around for when NSS is enabled. Note that this patch causes a couple of places that still assumed OpenSSL to be disabled when NSS is enabled - tor-gencert - pbkdf2
2018-08-12 23:18:41 +02:00
fi # enable_nss
Use openssl's version of sha3 when available. Part of 28837.
2018-12-13 19:25:30 +01:00
dnl We will someday make KECCAK_TINY optional, but for now we still need
dnl it for SHAKE, since OpenSSL's SHAKE can't be squeezed more than
dnl once. See comment in the definition of crypto_xof_t.
dnl AM_CONDITIONAL(BUILD_KECCAK_TINY,
dnl test "x$ac_cv_func_EVP_sha3_256" != "xyes")
AM_CONDITIONAL(BUILD_KECCAK_TINY, true)
When enabling NSS, disable OpenSSL. We used to link both libraries at once, but now that I'm working on TLS, there's nothing left to keep OpenSSL around for when NSS is enabled. Note that this patch causes a couple of places that still assumed OpenSSL to be disabled when NSS is enabled - tor-gencert - pbkdf2
2018-08-12 23:18:41 +02:00
dnl ======================================================================
dnl Can we use KIST?
sched: Detect KIST support at compile time Add a detection for the KIST scheduler in our build system and set HAVE_KIST_SUPPORT if available. Adapt the should use kist function with this new compile option. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-08-31 15:56:55 +02:00
dnl Define the set of checks for KIST scheduler support.
AC_DEFUN([CHECK_KIST_SUPPORT],[
dnl KIST needs struct tcp_info and for certain members to exist.
AC_CHECK_MEMBERS(
[struct tcp_info.tcpi_unacked, struct tcp_info.tcpi_snd_mss],
, ,[[#include <netinet/tcp.h>]])
dnl KIST needs SIOCOUTQNSD to exist for an ioctl call.
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [
#include <linux/sockios.h>
#ifndef SIOCOUTQNSD
#error
#endif
])], have_siocoutqnsd=yes, have_siocoutqnsd=no)
if test "x$have_siocoutqnsd" = "xyes"; then
if test "x$ac_cv_member_struct_tcp_info_tcpi_unacked" = "xyes"; then
if test "x$ac_cv_member_struct_tcp_info_tcpi_snd_mss" = "xyes"; then
have_kist_support=yes
fi
fi
fi
])
dnl Now, trigger the check.
CHECK_KIST_SUPPORT
AS_IF([test "x$have_kist_support" = "xyes"],
[AC_DEFINE(HAVE_KIST_SUPPORT, 1, [Defined if KIST scheduler is supported
on this system])],
[AC_MSG_NOTICE([KIST scheduler can't be used. Missing support.])])
Stop looking at session->ciphers when possible If the OpenSSL team accepts my patch to add an SSL_get_client_ciphers function, this patch will make Tor use it when available, thereby working better with openssl 1.1.
2015-05-26 17:05:36 +02:00
LIBS="$save_LIBS"
LDFLAGS="$save_LDFLAGS"
CPPFLAGS="$save_CPPFLAGS"
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
dnl ------------------------------------------------------
dnl Where do you live, zlib? And how do we call you?
Fix configure.in zlib package names.
2011-02-15 02:15:42 +01:00
tor_zlib_pkg_redhat="zlib"
tor_zlib_pkg_debian="zlib1g"
tor_zlib_devpkg_redhat="zlib-devel"
tor_zlib_devpkg_debian="zlib1g-dev"
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
TOR_SEARCH_LIBRARY(zlib, $tryzlibdir, [-lz],
[#include <zlib.h>],
[const char * zlibVersion(void);],
Remove spurious exit() calls from configure.ac We added these a while ago, but they do no actual good, and cause implicit declaration warnings in some situations. Rather than just adding stdint.h, it's easier to remove the exit() calls as redundant. Fixes bug 18626; bugfix from "cypherpunks"
2016-03-25 21:26:13 +01:00
[zlibVersion();], [--with-zlib-dir],
r12351@catbus: nickm | 2007-04-11 12:09:46 -0400 More autoconf hacking: use same machinery to find zlib as for openssl and libevent. Have unified library finder include setup for --with variable. Start trying to suggest to the user what packages they should install if finding the library fails. svn:r9945
2007-04-11 18:28:44 +02:00
[/opt/zlib])
Add --enable-static-zlib option Works like the --enable-static-openssl/libevent options. Requires --with-zlib-dir to be set. Note that other dependencies might still pull in a dynamicly linked zlib, if you don't link them in statically too.
2010-04-14 16:48:05 +02:00
if test "$enable_static_zlib" = "yes"; then
if test "$tor_cv_library_zlib_dir" = "(system)"; then
AC_MSG_ERROR("You must specify an explicit --with-zlib-dir=x option when
using --enable-static-zlib")
else
TOR_ZLIB_LIBS="$TOR_LIBDIR_zlib/libz.a"
fi
else
TOR_ZLIB_LIBS="-lz"
fi
AC_SUBST(TOR_ZLIB_LIBS)
Add --enable-lzma to our configure script. This patch adds support for enabling support for LZMA to our configure script. By default, the --enable-lzma option is set to "auto" which means if liblzma is available we'll build Tor with LZMA support. See: https://bugs.torproject.org/21662
2017-04-18 13:56:02 +02:00
dnl ------------------------------------------------------
dnl Where we do we find lzma?
AC_ARG_ENABLE(lzma,
Fix help message for --enable-lzma Fixes #22111.
2017-05-01 17:15:49 +02:00
AS_HELP_STRING(--enable-lzma, [enable support for the LZMA compression scheme.]),
Add --enable-lzma to our configure script. This patch adds support for enabling support for LZMA to our configure script. By default, the --enable-lzma option is set to "auto" which means if liblzma is available we'll build Tor with LZMA support. See: https://bugs.torproject.org/21662
2017-04-18 13:56:02 +02:00
[case "${enableval}" in
configure: Fix variable name typos The "if-present" clause of AC_ARG_ENABLE() sets enable_{lzma,zstd}, so we don't need to set it. The "if-not-present" clause needs to set enable_{lzma,zstd}, rather than just {lzma,zstd}. The rest of the script only ever checks for enable_{lzma,zstd} = no, so these changes do not actually change the behaviour of configure. Fixes bug 32401.
2019-11-06 05:02:22 +01:00
"yes") ;;
"no") ;;
Add --enable-lzma to our configure script. This patch adds support for enabling support for LZMA to our configure script. By default, the --enable-lzma option is set to "auto" which means if liblzma is available we'll build Tor with LZMA support. See: https://bugs.torproject.org/21662
2017-04-18 13:56:02 +02:00
* ) AC_MSG_ERROR(bad value for --enable-lzma) ;;
configure: Fix variable name typos The "if-present" clause of AC_ARG_ENABLE() sets enable_{lzma,zstd}, so we don't need to set it. The "if-not-present" clause needs to set enable_{lzma,zstd}, rather than just {lzma,zstd}. The rest of the script only ever checks for enable_{lzma,zstd} = no, so these changes do not actually change the behaviour of configure. Fixes bug 32401.
2019-11-06 05:02:22 +01:00
esac], [enable_lzma=auto])
Add --enable-lzma to our configure script. This patch adds support for enabling support for LZMA to our configure script. By default, the --enable-lzma option is set to "auto" which means if liblzma is available we'll build Tor with LZMA support. See: https://bugs.torproject.org/21662
2017-04-18 13:56:02 +02:00
if test "x$enable_lzma" = "xno"; then
have_lzma=no;
else
PKG_CHECK_MODULES([LZMA],
[liblzma],
have_lzma=yes,
have_lzma=no)
if test "x$have_lzma" = "xno" ; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
AC_MSG_WARN([Unable to find liblzma, $pkg_config_user_action, or set LZMA_CFLAGS and LZMA_LIBS.])
Add --enable-lzma to our configure script. This patch adds support for enabling support for LZMA to our configure script. By default, the --enable-lzma option is set to "auto" which means if liblzma is available we'll build Tor with LZMA support. See: https://bugs.torproject.org/21662
2017-04-18 13:56:02 +02:00
fi
fi
if test "x$have_lzma" = "xyes"; then
AC_DEFINE(HAVE_LZMA,1,[Have LZMA])
TOR_LZMA_CFLAGS="${LZMA_CFLAGS}"
TOR_LZMA_LIBS="${LZMA_LIBS}"
fi
AC_SUBST(TOR_LZMA_CFLAGS)
AC_SUBST(TOR_LZMA_LIBS)
Add --enable-zstd to our configure script. This patch adds support for enabling support for Zstandard to our configure script. By default, the --enable-zstd option is set to "auto" which means if libzstd is available we'll build Tor with Zstandard support. See: https://bugs.torproject.org/21662
2017-04-18 17:54:59 +02:00
dnl ------------------------------------------------------
dnl Where we do we find zstd?
AC_ARG_ENABLE(zstd,
AS_HELP_STRING(--enable-zstd, [enable support for the Zstandard compression scheme.]),
[case "${enableval}" in
configure: Fix variable name typos The "if-present" clause of AC_ARG_ENABLE() sets enable_{lzma,zstd}, so we don't need to set it. The "if-not-present" clause needs to set enable_{lzma,zstd}, rather than just {lzma,zstd}. The rest of the script only ever checks for enable_{lzma,zstd} = no, so these changes do not actually change the behaviour of configure. Fixes bug 32401.
2019-11-06 05:02:22 +01:00
"yes") ;;
"no") ;;
Add --enable-zstd to our configure script. This patch adds support for enabling support for Zstandard to our configure script. By default, the --enable-zstd option is set to "auto" which means if libzstd is available we'll build Tor with Zstandard support. See: https://bugs.torproject.org/21662
2017-04-18 17:54:59 +02:00
* ) AC_MSG_ERROR(bad value for --enable-zstd) ;;
configure: Fix variable name typos The "if-present" clause of AC_ARG_ENABLE() sets enable_{lzma,zstd}, so we don't need to set it. The "if-not-present" clause needs to set enable_{lzma,zstd}, rather than just {lzma,zstd}. The rest of the script only ever checks for enable_{lzma,zstd} = no, so these changes do not actually change the behaviour of configure. Fixes bug 32401.
2019-11-06 05:02:22 +01:00
esac], [enable_zstd=auto])
Add --enable-zstd to our configure script. This patch adds support for enabling support for Zstandard to our configure script. By default, the --enable-zstd option is set to "auto" which means if libzstd is available we'll build Tor with Zstandard support. See: https://bugs.torproject.org/21662
2017-04-18 17:54:59 +02:00
if test "x$enable_zstd" = "xno"; then
have_zstd=no;
else
PKG_CHECK_MODULES([ZSTD],
Check for libzstd >= 1.1 The consensus compression code depends on a streaming compression API that is new in libzstd-1.1. Fixes #22413.
2017-05-30 19:00:54 +02:00
[libzstd >= 1.1],
Add --enable-zstd to our configure script. This patch adds support for enabling support for Zstandard to our configure script. By default, the --enable-zstd option is set to "auto" which means if libzstd is available we'll build Tor with Zstandard support. See: https://bugs.torproject.org/21662
2017-04-18 17:54:59 +02:00
have_zstd=yes,
have_zstd=no)
if test "x$have_zstd" = "xno" ; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
AC_MSG_WARN([Unable to find libzstd, $pkg_config_user_action, or set ZSTD_CFLAGS and ZSTD_LIBS.])
Add --enable-zstd to our configure script. This patch adds support for enabling support for Zstandard to our configure script. By default, the --enable-zstd option is set to "auto" which means if libzstd is available we'll build Tor with Zstandard support. See: https://bugs.torproject.org/21662
2017-04-18 17:54:59 +02:00
fi
fi
if test "x$have_zstd" = "xyes"; then
AC_DEFINE(HAVE_ZSTD,1,[Have Zstd])
TOR_ZSTD_CFLAGS="${ZSTD_CFLAGS}"
TOR_ZSTD_LIBS="${ZSTD_LIBS}"
Use autoconf to check for optional zstd functionality. Fixes a bug in our zstd-static code. Bug not in any released version of Tor.
2018-02-18 22:19:43 +01:00
dnl now check for zstd functions
save_LIBS="$LIBS"
save_CFLAGS="$CFLAGS"
LIBS="$LIBS $ZSTD_LIBS"
CFLAGS="$CFLAGS $ZSTD_CFLAGS"
AC_CHECK_FUNCS(ZSTD_estimateCStreamSize \
ZSTD_estimateDCtxSize)
LIBS="$save_LIBS"
CFLAGS="$save_CFLAGS"
Add --enable-zstd to our configure script. This patch adds support for enabling support for Zstandard to our configure script. By default, the --enable-zstd option is set to "auto" which means if libzstd is available we'll build Tor with Zstandard support. See: https://bugs.torproject.org/21662
2017-04-18 17:54:59 +02:00
fi
AC_SUBST(TOR_ZSTD_CFLAGS)
AC_SUBST(TOR_ZSTD_LIBS)
Add ability to keep the CAP_NET_BIND_SERVICE capability on Linux This feature allows us to bind low ports when starting as root and switching UIDs. Based on code by David Goulet. Implement feature 8195
2015-11-06 19:12:44 +01:00
dnl ----------------------------------------------------------------------
dnl Check if libcap is available for capabilities.
tor_cap_pkg_debian="libcap2"
tor_cap_pkg_redhat="libcap"
tor_cap_devpkg_debian="libcap-dev"
tor_cap_devpkg_redhat="libcap-devel"
AC_CHECK_LIB([cap], [cap_init], [],
AC_MSG_NOTICE([Libcap was not found. Capabilities will not be usable.])
)
AC_CHECK_FUNCS(cap_set_proc)
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
dnl ---------------------------------------------------------------------
dnl Now that we know about our major libraries, we can check for compiler
dnl and linker hardening options. We need to do this with the libraries known,
dnl since sometimes the linker will like an option but not be willing to
dnl use it with a build of a library.
added automake/autoconf support. When in doubt, "aclocal && autoconf && autoheader && automake" from the top dir. svn:r10
2002-06-29 01:26:42 +02:00
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
all_ldflags_for_check="$TOR_LDFLAGS_zlib $TOR_LDFLAGS_openssl $TOR_LDFLAGS_libevent"
22839: Build tor with rust enabled on win - make tor_util static library name configurable - fix Rust libary dependency order for Windows
2017-08-10 10:08:41 +02:00
all_libs_for_check="$TOR_ZLIB_LIBS $TOR_LIB_MATH $TOR_LIBEVENT_LIBS $TOR_OPENSSL_LIBS $TOR_SYSTEMD_LIBS $TOR_LIB_WS32 $TOR_LIB_GDI $TOR_LIB_USERENV $TOR_CAP_LIBS"
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
Add -ftrapv to gcc-hardening ... mostly! We know there are overflows in curve25519-donna-c32, so we'll have to have that one be fwrapv. Only apply the asan, ubsan, and trapv options to the code that does not need to run in constant time. Those options introduce branches to the code they instrument. (These introduced branches should never actually be taken, so it might _still_ be constant time after all, but branch predictors are complicated enough that I'm not really confident here. Let's aim for safety.) Closes 17983.
2016-04-13 00:59:40 +02:00
CFLAGS_FTRAPV=
CFLAGS_FWRAPV=
CFLAGS_ASAN=
CFLAGS_UBSAN=
Add libfuzzer support.
2016-12-19 21:17:42 +01:00
Only use -Qunused-arguments when building with clang. fix on 5210 fix.
2012-06-13 22:37:23 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [
#if !defined(__clang__)
#error
#endif])], have_clang=yes, have_clang=no)
Add an --enable-pic option for people who hack Tor into libraries Closes ticket 23846
2018-09-12 16:10:29 +02:00
if test "x$enable_pic" = "xyes"; then
TOR_CHECK_CFLAGS(-fPIC)
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$enable_gcc_hardening" != "xno"; then
Undefine _FORTIFY_SOURCE before defining it. This makes our compilation options checks in autoconf work better on systems that already define _FORTIFY_SOURCE. Fixes at least one case of bug 18841; bugfix on 0.2.3.17-beta. Patch from "trudokal".
2016-05-11 18:15:37 +02:00
CFLAGS="$CFLAGS -U_FORTIFY_SOURCE -D_FORTIFY_SOURCE=2"
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$have_clang" = "xyes"; then
Only use -Qunused-arguments when building with clang. fix on 5210 fix.
2012-06-13 22:37:23 +02:00
TOR_CHECK_CFLAGS(-Qunused-arguments)
fi
Give a better warning when stack protection breaks linking. Fix for 9948; patch from Benedikt Gollatz.
2013-10-21 19:07:47 +02:00
TOR_CHECK_CFLAGS(-fstack-protector-all, also_link)
AS_VAR_PUSHDEF([can_compile], [tor_cv_cflags_-fstack-protector-all])
AS_VAR_PUSHDEF([can_link], [tor_can_link_-fstack-protector-all])
Fix configure script build with autoconf < 2.63 We added some AS_VAR_IF-based checks to detect whether we have managed to compile (but not link) with stack-protector. On autoconf before 2.63, we don't have AS_VAR_IF, so we just have to let the user get a compile error rather than a helpful "find libssp" error. Fixes bug 12693; bugfix on 0.2.5.2-alpha (commit 21ac292820da578)
2014-08-13 18:01:58 +02:00
m4_ifdef([AS_VAR_IF],[
Give a better warning when stack protection breaks linking. Fix for 9948; patch from Benedikt Gollatz.
2013-10-21 19:07:47 +02:00
AS_VAR_IF(can_compile, [yes],
AS_VAR_IF(can_link, [yes],
[],
AC_MSG_ERROR([We tried to build with stack protection; it looks like your compiler supports it but your libc does not provide it. Are you missing libssp? (You can --disable-gcc-hardening to ignore this error.)]))
Fix configure script build with autoconf < 2.63 We added some AS_VAR_IF-based checks to detect whether we have managed to compile (but not link) with stack-protector. On autoconf before 2.63, we don't have AS_VAR_IF, so we just have to let the user get a compile error rather than a helpful "find libssp" error. Fixes bug 12693; bugfix on 0.2.5.2-alpha (commit 21ac292820da578)
2014-08-13 18:01:58 +02:00
)])
Give a better warning when stack protection breaks linking. Fix for 9948; patch from Benedikt Gollatz.
2013-10-21 19:07:47 +02:00
AS_VAR_POPDEF([can_link])
AS_VAR_POPDEF([can_compile])
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
TOR_CHECK_CFLAGS(-Wstack-protector)
TOR_CHECK_CFLAGS(--param ssp-buffer-size=1)
oss_fuzz also requires no PIE
2017-01-09 17:26:27 +01:00
if test "$bwin32" = "false" && test "$enable_libfuzzer" != "yes" && test "$enable_oss_fuzz" != "yes"; then
Add an --enable-pic option for people who hack Tor into libraries Closes ticket 23846
2018-09-12 16:10:29 +02:00
if test "$enable_pic" != "yes"; then
Explain PIE/PIC rationale for 23846.
2018-09-13 20:53:20 +02:00
# If we have already enabled -fPIC, then we don't also need to
# compile with -fPIE...
Add an --enable-pic option for people who hack Tor into libraries Closes ticket 23846
2018-09-12 16:10:29 +02:00
TOR_CHECK_CFLAGS(-fPIE)
fi
Explain PIE/PIC rationale for 23846.
2018-09-13 20:53:20 +02:00
# ... but we want to link our executables with -pie in any case, since
# they're executables, not a library.
Do not try to use -pie on windows; it appears to break badly and weirdly
2012-06-11 18:22:48 +02:00
TOR_CHECK_LDFLAGS(-pie, "$all_ldflags_for_check", "$all_libs_for_check")
fi
Check linking of hardening options, give better warnings if it fails. Previously we'd only check whether the hardening options succeeded at the compile step. Now we'll try to link with them too, and tell the user in advance if something seems likely to go wrong. Closes ticket 18895.
2016-05-16 15:37:27 +02:00
TOR_TRY_COMPILE_WITH_CFLAGS(-fwrapv, also_link, CFLAGS_FWRAPV="-fwrapv", true)
Check whether gcc-hardening is runnable, and log an error if not Closes ticket 27530.
2019-05-29 22:08:33 +02:00
AC_MSG_CHECKING([whether we can run hardened binaries])
AC_RUN_IFELSE([AC_LANG_PROGRAM([], [return 0;])],
[AC_MSG_RESULT([yes])],
[AC_MSG_RESULT([no])
AC_MSG_ERROR([dnl
We can link with compiler hardening options, but we can't run with them.
That's a bad sign! If you must, you can pass --disable-gcc-hardening to
configure, but it would be better to figure out what the underlying problem
is.])],
[AC_MSG_RESULT([cross])])
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
fi
added automake/autoconf support. When in doubt, "aclocal && autoconf && autoheader && automake" from the top dir. svn:r10
2002-06-29 01:26:42 +02:00
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
if test "$fragile_hardening" = "yes"; then
TROVE-2017-001 : move -ftrapv back into --expensive-hardening.
2017-01-22 17:32:54 +01:00
TOR_TRY_COMPILE_WITH_CFLAGS(-ftrapv, also_link, CFLAGS_FTRAPV="-ftrapv", true)
if test "$tor_cv_cflags__ftrapv" = "yes" && test "$tor_can_link__ftrapv" != "yes"; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
TROVE-2017-001 : move -ftrapv back into --expensive-hardening.
2017-01-22 17:32:54 +01:00
AC_MSG_WARN([The compiler supports -ftrapv, but for some reason I was not able to link with -ftrapv. Are you missing run-time support? Run-time hardening will not work as well as it should.])
fi
Check linking of hardening options, give better warnings if it fails. Previously we'd only check whether the hardening options succeeded at the compile step. Now we'll try to link with them too, and tell the user in advance if something seems likely to go wrong. Closes ticket 18895.
2016-05-16 15:37:27 +02:00
if test "$tor_cv_cflags__ftrapv" != "yes"; then
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
AC_MSG_ERROR([You requested fragile hardening, but the compiler does not seem to support -ftrapv.])
Check linking of hardening options, give better warnings if it fails. Previously we'd only check whether the hardening options succeeded at the compile step. Now we'll try to link with them too, and tell the user in advance if something seems likely to go wrong. Closes ticket 18895.
2016-05-16 15:37:27 +02:00
fi
TOR_TRY_COMPILE_WITH_CFLAGS([-fsanitize=address], also_link, CFLAGS_ASAN="-fsanitize=address", true)
if test "$tor_cv_cflags__fsanitize_address" = "yes" && test "$tor_can_link__fsanitize_address" != "yes"; then
Make configure log messages less confusing during static builds Closes #24558.
2017-12-12 02:43:53 +01:00
AC_MSG_ERROR([The compiler supports -fsanitize=address, but for some reason I was not able to link when using it. Are you missing run-time support? With GCC you need libubsan.*, and with Clang you need libclang_rt.ubsan*])
Check linking of hardening options, give better warnings if it fails. Previously we'd only check whether the hardening options succeeded at the compile step. Now we'll try to link with them too, and tell the user in advance if something seems likely to go wrong. Closes ticket 18895.
2016-05-16 15:37:27 +02:00
fi
TOR_TRY_COMPILE_WITH_CFLAGS([-fsanitize=undefined], also_link, CFLAGS_UBSAN="-fsanitize=undefined", true)
if test "$tor_cv_cflags__fsanitize_address" = "yes" && test "$tor_can_link__fsanitize_address" != "yes"; then
Make configure log messages less confusing during static builds Closes #24558.
2017-12-12 02:43:53 +01:00
AC_MSG_ERROR([The compiler supports -fsanitize=undefined, but for some reason I was not able to link when using it. Are you missing run-time support? With GCC you need libasan.*, and with Clang you need libclang_rt.ubsan*])
Check linking of hardening options, give better warnings if it fails. Previously we'd only check whether the hardening options succeeded at the compile step. Now we'll try to link with them too, and tell the user in advance if something seems likely to go wrong. Closes ticket 18895.
2016-05-16 15:37:27 +02:00
fi
Silence unneeded clang warns that triggered in prob distr tests. See https://trac.torproject.org/projects/tor/ticket/29528#comment:3 and https://bugs.llvm.org/show_bug.cgi?id=19535
2019-02-25 14:29:41 +01:00
TOR_TRY_COMPILE_WITH_CFLAGS([-fno-sanitize=float-divide-by-zero], also_link, CFLAGS_UBSAN="-fno-sanitize=float-divide-by-zero", true)
if test "$tor_cv_cflags__fno_sanitize_float_divide_by_zero" = "yes" && test "$tor_can_link__fno_sanitize_float_divide_by_zero" != "yes"; then
AC_MSG_ERROR([The compiler supports -fno-sanitize=float-divide-by-zero, but for some reason I was not able to link when using it. Are you missing run-time support? With GCC you need libasan.*, and with Clang you need libclang_rt.ubsan*])
fi
Check linking of hardening options, give better warnings if it fails. Previously we'd only check whether the hardening options succeeded at the compile step. Now we'll try to link with them too, and tell the user in advance if something seems likely to go wrong. Closes ticket 18895.
2016-05-16 15:37:27 +02:00
TOR_CHECK_CFLAGS([-fno-omit-frame-pointer])
New --enable-expensive-hardening option It turns on -fsanitize=address and -fsanitize=ubsan if they work. Most relays won't want this. Some clients may. Ticket 11477.
2014-04-14 23:10:34 +02:00
fi
Add -ftrapv to gcc-hardening ... mostly! We know there are overflows in curve25519-donna-c32, so we'll have to have that one be fwrapv. Only apply the asan, ubsan, and trapv options to the code that does not need to run in constant time. Those options introduce branches to the code they instrument. (These introduced branches should never actually be taken, so it might _still_ be constant time after all, but branch predictors are complicated enough that I'm not really confident here. Let's aim for safety.) Closes 17983.
2016-04-13 00:59:40 +02:00
CFLAGS_BUGTRAP="$CFLAGS_FTRAPV $CFLAGS_ASAN $CFLAGS_UBSAN"
CFLAGS_CONSTTIME="$CFLAGS_FWRAPV"
Fix "conditional "ADD_MULODI4" was never defined".
2016-05-19 22:38:24 +02:00
mulodi_fixes_ftrapv=no
Include __mulodi4 in libor_ctime when it fixes clang -m32 -ftrapv We use a pretty specific pair of autoconf tests here to make sure that we only add this code when: a) a 64-bit signed multiply fails to link, AND b) the same 64-bit signed multiply DOES link correctly when __mulodi4 is defined. Closes ticket 19079.
2016-05-18 15:44:05 +02:00
if test "$have_clang" = "yes"; then
saved_CFLAGS="$CFLAGS"
CFLAGS="$CFLAGS $CFLAGS_FTRAPV"
AC_MSG_CHECKING([whether clang -ftrapv can link a 64-bit int multiply])
AC_LINK_IFELSE([
AC_LANG_SOURCE([[
#include <stdint.h>
#include <stdlib.h>
int main(int argc, char **argv)
{
int64_t x = ((int64_t)atoi(argv[1])) * (int64_t)atoi(argv[2])
add missing atoi for cosmetic reasons
2016-05-19 22:01:25 +02:00
* (int64_t)atoi(argv[3]);
Include __mulodi4 in libor_ctime when it fixes clang -m32 -ftrapv We use a pretty specific pair of autoconf tests here to make sure that we only add this code when: a) a 64-bit signed multiply fails to link, AND b) the same 64-bit signed multiply DOES link correctly when __mulodi4 is defined. Closes ticket 19079.
2016-05-18 15:44:05 +02:00
return x == 9;
} ]])],
[ftrapv_can_link=yes; AC_MSG_RESULT([yes])],
[ftrapv_can_link=no; AC_MSG_RESULT([no])])
if test "$ftrapv_can_link" = "no"; then
AC_MSG_CHECKING([whether defining __mulodi4 fixes that])
AC_LINK_IFELSE([
AC_LANG_SOURCE([[
#include <stdint.h>
#include <stdlib.h>
int64_t __mulodi4(int64_t a, int64_t b, int *overflow) {
*overflow=0;
return a;
}
int main(int argc, char **argv)
{
int64_t x = ((int64_t)atoi(argv[1])) * (int64_t)atoi(argv[2])
add missing atoi for cosmetic reasons
2016-05-19 22:01:25 +02:00
* (int64_t)atoi(argv[3]);
Include __mulodi4 in libor_ctime when it fixes clang -m32 -ftrapv We use a pretty specific pair of autoconf tests here to make sure that we only add this code when: a) a 64-bit signed multiply fails to link, AND b) the same 64-bit signed multiply DOES link correctly when __mulodi4 is defined. Closes ticket 19079.
2016-05-18 15:44:05 +02:00
return x == 9;
} ]])],
[mulodi_fixes_ftrapv=yes; AC_MSG_RESULT([yes])],
[mulodi_fixes_ftrapv=no; AC_MSG_RESULT([no])])
fi
CFLAGS="$saved_CFLAGS"
fi
Fix "conditional "ADD_MULODI4" was never defined".
2016-05-19 22:38:24 +02:00
AM_CONDITIONAL(ADD_MULODI4, test "$mulodi_fixes_ftrapv" = "yes")
Add -ftrapv to gcc-hardening ... mostly! We know there are overflows in curve25519-donna-c32, so we'll have to have that one be fwrapv. Only apply the asan, ubsan, and trapv options to the code that does not need to run in constant time. Those options introduce branches to the code they instrument. (These introduced branches should never actually be taken, so it might _still_ be constant time after all, but branch predictors are complicated enough that I'm not really confident here. Let's aim for safety.) Closes 17983.
2016-04-13 00:59:40 +02:00
dnl These cflags add bunches of branches, and we haven't been able to
dnl persuade ourselves that they're suitable for code that needs to be
dnl constant time.
AC_SUBST(CFLAGS_BUGTRAP)
Fix typos. Typos found with codespell. Please keep in mind that this should have impact on actual code and must be carefully evaluated: src/core/or/lttng_circuit.inc - ctf_enum_value("CONTROLER", CIRCUIT_PURPOSE_CONTROLLER) + ctf_enum_value("CONTROLLER", CIRCUIT_PURPOSE_CONTROLLER)
2020-11-11 12:38:19 +01:00
dnl These cflags are variant ones suitable for code that needs to be
Add -ftrapv to gcc-hardening ... mostly! We know there are overflows in curve25519-donna-c32, so we'll have to have that one be fwrapv. Only apply the asan, ubsan, and trapv options to the code that does not need to run in constant time. Those options introduce branches to the code they instrument. (These introduced branches should never actually be taken, so it might _still_ be constant time after all, but branch predictors are complicated enough that I'm not really confident here. Let's aim for safety.) Closes 17983.
2016-04-13 00:59:40 +02:00
dnl constant-time.
AC_SUBST(CFLAGS_CONSTTIME)
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$enable_linker_hardening" != "xno"; then
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
TOR_CHECK_LDFLAGS(-z relro -z now, "$all_ldflags_for_check", "$all_libs_for_check")
fi
First implementation of tor-fw-helper. tor-fw-helper is a command-line tool to wrap and abstract various firewall port-forwarding tools. This commit matches the state of Jacob's tor-fw-helper branch as of 23 September 2010. (commit msg by Nick)
2010-04-17 02:45:12 +02:00
On Linux (and some other systems) we need -rdynamic for backtraces
2013-07-19 20:18:43 +02:00
# For backtrace support
TOR_CHECK_LDFLAGS(-rdynamic)
First implementation of tor-fw-helper. tor-fw-helper is a command-line tool to wrap and abstract various firewall port-forwarding tools. This commit matches the state of Jacob's tor-fw-helper branch as of 23 September 2010. (commit msg by Nick)
2010-04-17 02:45:12 +02:00
dnl ------------------------------------------------------
Build donna32 with -fomit-frame-pointer
2013-01-30 19:08:04 +01:00
dnl Now see if we have a -fomit-frame-pointer compiler option.
saved_CFLAGS="$CFLAGS"
TOR_CHECK_CFLAGS(-fomit-frame-pointer)
New --enable-expensive-hardening option It turns on -fsanitize=address and -fsanitize=ubsan if they work. Most relays won't want this. Some clients may. Ticket 11477.
2014-04-14 23:10:34 +02:00
F_OMIT_FRAME_POINTER=''
Build donna32 with -fomit-frame-pointer
2013-01-30 19:08:04 +01:00
if test "$saved_CFLAGS" != "$CFLAGS"; then
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
if test "$fragile_hardening" = "yes"; then
New --enable-expensive-hardening option It turns on -fsanitize=address and -fsanitize=ubsan if they work. Most relays won't want this. Some clients may. Ticket 11477.
2014-04-14 23:10:34 +02:00
F_OMIT_FRAME_POINTER='-fomit-frame-pointer'
fi
Build donna32 with -fomit-frame-pointer
2013-01-30 19:08:04 +01:00
fi
CFLAGS="$saved_CFLAGS"
AC_SUBST(F_OMIT_FRAME_POINTER)
pass our compiler -fasynchronous-unwind-tables by default This should make more platforms (in particular, ones with compilers where -fomit-frame-pointer is on by default but table generation isn't) support backtrace generation. Thanks to cypherpunks for this one. Fixes bug 11047; bugfix on 0.2.5.2-alpha.
2014-02-24 17:41:53 +01:00
dnl ------------------------------------------------------
dnl If we are adding -fomit-frame-pointer (or if the compiler's doing it
dnl for us, as GCC 4.6 and later do at many optimization levels), then
dnl we should try to add -fasynchronous-unwind-tables so that our backtrace
dnl code will work.
TOR_CHECK_CFLAGS(-fasynchronous-unwind-tables)
Add a basic seccomp2 syscall filter on Linux It's controlled by the new Sandbox argument. Right now, it's rather coarse-grained, it's Linux-only, and it may break some features.
2013-06-17 12:07:14 +02:00
dnl ============================================================
dnl Check for libseccomp
New --disable-seccomp option to turn off support for seccomp. Fixes 11628.
2014-04-28 17:11:11 +02:00
if test "x$enable_seccomp" != "xno"; then
AC_CHECK_HEADERS([seccomp.h])
AC_SEARCH_LIBS(seccomp_init, [seccomp])
fi
Add a basic seccomp2 syscall filter on Linux It's controlled by the new Sandbox argument. Right now, it's rather coarse-grained, it's Linux-only, and it may break some features.
2013-06-17 12:07:14 +02:00
More generic passphrase hashing code, including scrypt support Uses libscrypt when found; otherwise, we don't have scrypt and we only support openpgp rfc2440 s2k hashing, or pbkdf2. Includes documentation and unit tests; coverage around 95%. Remaining uncovered code is sanity-checks that shouldn't be reachable fwict.
2014-08-28 23:38:22 +02:00
dnl ============================================================
dnl Check for libscrypt
if test "x$enable_libscrypt" != "xno"; then
AC_CHECK_HEADERS([libscrypt.h])
AC_SEARCH_LIBS(libscrypt_scrypt, [scrypt])
Make sure that libscrypt_scrypt actually exists before using it. Previously, if the header was present, we'd proceed even if the function wasn't there. Easy fix for bug 19161. A better fix would involve trying harder to find libscrypt_scrypt.
2016-05-24 16:28:31 +02:00
AC_CHECK_FUNCS([libscrypt_scrypt])
More generic passphrase hashing code, including scrypt support Uses libscrypt when found; otherwise, we don't have scrypt and we only support openpgp rfc2440 s2k hashing, or pbkdf2. Includes documentation and unit tests; coverage around 95%. Remaining uncovered code is sanity-checks that shouldn't be reachable fwict.
2014-08-28 23:38:22 +02:00
fi
Add a wrapper around, and test and build support for, curve25519. We want to use donna-c64 when we have a GCC with support for 64x64->uint128_t multiplying. If not, we want to use libnacl if we can, unless it's giving us the unsafe "ref" implementation. And if that isn't going to work, we'd like to use the portable-and-safe-but-slow 32-bit "donna" implementation. We might need more library searching for the correct libnacl, especially once the next libnacl release is out -- it's likely to have bunches of better curve25519 implementations. I also define a set of curve25519 wrapper functions, though it really shouldn't be necessary. We should eventually make the -donna*.c files get build with -fomit-frame-pointer, since that can make a difference.
2012-12-03 21:44:21 +01:00
dnl ============================================================
dnl We need an implementation of curve25519.
dnl set these defaults.
build_curve25519_donna=no
build_curve25519_donna_c64=no
use_curve25519_donna=no
use_curve25519_nacl=no
CURVE25519_LIBS=
Remove configure option to disable curve25519 By now, support in the network is widespread and it's time to require more modern crypto on all Tor instances, whether they're clients or servers. By doing this early in 0.2.6, we can be sure that at some point all clients will have reasonable support.
2014-10-27 14:37:50 +01:00
dnl The best choice is using curve25519-donna-c64, but that requires
dnl that we
AC_CACHE_CHECK([whether we can use curve25519-donna-c64],
tor_cv_can_use_curve25519_donna_c64,
[AC_RUN_IFELSE(
[AC_LANG_PROGRAM([dnl
#include <stdint.h>
typedef unsigned uint128_t __attribute__((mode(TI)));
int func(uint64_t a, uint64_t b) {
uint128_t c = ((uint128_t)a) * b;
int ok = ((uint64_t)(c>>96)) == 522859 &&
(((uint64_t)(c>>64))&0xffffffffL) == 3604448702L &&
(((uint64_t)(c>>32))&0xffffffffL) == 2351960064L &&
(((uint64_t)(c))&0xffffffffL) == 0;
return ok;
}
], [dnl
int ok = func( ((uint64_t)2000000000) * 1000000000,
((uint64_t)1234567890) << 24);
return !ok;
])],
[tor_cv_can_use_curve25519_donna_c64=yes],
[tor_cv_can_use_curve25519_donna_c64=no],
[AC_LINK_IFELSE(
[AC_LANG_PROGRAM([dnl
#include <stdint.h>
typedef unsigned uint128_t __attribute__((mode(TI)));
int func(uint64_t a, uint64_t b) {
uint128_t c = ((uint128_t)a) * b;
int ok = ((uint64_t)(c>>96)) == 522859 &&
(((uint64_t)(c>>64))&0xffffffffL) == 3604448702L &&
(((uint64_t)(c>>32))&0xffffffffL) == 2351960064L &&
(((uint64_t)(c))&0xffffffffL) == 0;
return ok;
}
], [dnl
int ok = func( ((uint64_t)2000000000) * 1000000000,
((uint64_t)1234567890) << 24);
return !ok;
])],
[tor_cv_can_use_curve25519_donna_c64=cross],
[tor_cv_can_use_curve25519_donna_c64=no])])])
AC_CHECK_HEADERS([crypto_scalarmult_curve25519.h \
nacl/crypto_scalarmult_curve25519.h])
AC_CACHE_CHECK([for nacl compiled with a fast curve25519 implementation],
tor_cv_can_use_curve25519_nacl,
[tor_saved_LIBS="$LIBS"
LIBS="$LIBS -lnacl"
AC_LINK_IFELSE(
[AC_LANG_PROGRAM([dnl
#ifdef HAVE_CRYPTO_SCALARMULT_CURVE25519_H
#include <crypto_scalarmult_curve25519.h>
#elif defined(HAVE_NACL_CRYPTO_SCALARMULT_CURVE25519_H)
#include <nacl/crypto_scalarmult_curve25519.h>
#endif
#ifdef crypto_scalarmult_curve25519_ref_BYTES
#error Hey, this is the reference implementation! That's not fast.
#endif
], [
unsigned char *a, *b, *c; crypto_scalarmult_curve25519(a,b,c);
])], [tor_cv_can_use_curve25519_nacl=yes],
[tor_cv_can_use_curve25519_nacl=no])
LIBS="$tor_saved_LIBS" ])
dnl Okay, now we need to figure out which one to actually use. Fall back
dnl to curve25519-donna.c
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$tor_cv_can_use_curve25519_donna_c64" != "xno"; then
Remove configure option to disable curve25519 By now, support in the network is widespread and it's time to require more modern crypto on all Tor instances, whether they're clients or servers. By doing this early in 0.2.6, we can be sure that at some point all clients will have reasonable support.
2014-10-27 14:37:50 +01:00
build_curve25519_donna_c64=yes
use_curve25519_donna=yes
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
elif test "x$tor_cv_can_use_curve25519_nacl" = "xyes"; then
Remove configure option to disable curve25519 By now, support in the network is widespread and it's time to require more modern crypto on all Tor instances, whether they're clients or servers. By doing this early in 0.2.6, we can be sure that at some point all clients will have reasonable support.
2014-10-27 14:37:50 +01:00
use_curve25519_nacl=yes
CURVE25519_LIBS=-lnacl
else
build_curve25519_donna=yes
use_curve25519_donna=yes
fi
Add a wrapper around, and test and build support for, curve25519. We want to use donna-c64 when we have a GCC with support for 64x64->uint128_t multiplying. If not, we want to use libnacl if we can, unless it's giving us the unsafe "ref" implementation. And if that isn't going to work, we'd like to use the portable-and-safe-but-slow 32-bit "donna" implementation. We might need more library searching for the correct libnacl, especially once the next libnacl release is out -- it's likely to have bunches of better curve25519 implementations. I also define a set of curve25519 wrapper functions, though it really shouldn't be necessary. We should eventually make the -donna*.c files get build with -fomit-frame-pointer, since that can make a difference.
2012-12-03 21:44:21 +01:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$use_curve25519_donna" = "xyes"; then
Add a wrapper around, and test and build support for, curve25519. We want to use donna-c64 when we have a GCC with support for 64x64->uint128_t multiplying. If not, we want to use libnacl if we can, unless it's giving us the unsafe "ref" implementation. And if that isn't going to work, we'd like to use the portable-and-safe-but-slow 32-bit "donna" implementation. We might need more library searching for the correct libnacl, especially once the next libnacl release is out -- it's likely to have bunches of better curve25519 implementations. I also define a set of curve25519 wrapper functions, though it really shouldn't be necessary. We should eventually make the -donna*.c files get build with -fomit-frame-pointer, since that can make a difference.
2012-12-03 21:44:21 +01:00
AC_DEFINE(USE_CURVE25519_DONNA, 1,
[Defined if we should use an internal curve25519_donna{,_c64} implementation])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$use_curve25519_nacl" = "xyes"; then
Add a wrapper around, and test and build support for, curve25519. We want to use donna-c64 when we have a GCC with support for 64x64->uint128_t multiplying. If not, we want to use libnacl if we can, unless it's giving us the unsafe "ref" implementation. And if that isn't going to work, we'd like to use the portable-and-safe-but-slow 32-bit "donna" implementation. We might need more library searching for the correct libnacl, especially once the next libnacl release is out -- it's likely to have bunches of better curve25519 implementations. I also define a set of curve25519 wrapper functions, though it really shouldn't be necessary. We should eventually make the -donna*.c files get build with -fomit-frame-pointer, since that can make a difference.
2012-12-03 21:44:21 +01:00
AC_DEFINE(USE_CURVE25519_NACL, 1,
[Defined if we should use a curve25519 from nacl])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
AM_CONDITIONAL(BUILD_CURVE25519_DONNA,
test "x$build_curve25519_donna" = "xyes")
AM_CONDITIONAL(BUILD_CURVE25519_DONNA_C64,
test "x$build_curve25519_donna_c64" = "xyes")
Add a wrapper around, and test and build support for, curve25519. We want to use donna-c64 when we have a GCC with support for 64x64->uint128_t multiplying. If not, we want to use libnacl if we can, unless it's giving us the unsafe "ref" implementation. And if that isn't going to work, we'd like to use the portable-and-safe-but-slow 32-bit "donna" implementation. We might need more library searching for the correct libnacl, especially once the next libnacl release is out -- it's likely to have bunches of better curve25519 implementations. I also define a set of curve25519 wrapper functions, though it really shouldn't be necessary. We should eventually make the -donna*.c files get build with -fomit-frame-pointer, since that can make a difference.
2012-12-03 21:44:21 +01:00
AC_SUBST(CURVE25519_LIBS)
Make our compiler-hardening checks robust against MinGW First, specify -Werror when we are testing each option; if it causes a warning to appear, we shouldn't be adding it. Second, do not attempt to add these options until after we have found the libraries we want. Previously, I would hit a bug where the linker hardening options worked fine when we weren't linking anything, but failed completely once we added openssl or libevent.
2012-06-11 17:00:48 +02:00
dnl Make sure to enable support for large off_t if available.
Use autoconf to enable largefile support where necessary. Use ftello where available, since ftell can fail at 2GB. svn:r2806
2004-11-12 06:05:41 +01:00
AC_SYS_LARGEFILE
Remove all use of the assert.h header Nothing in Tor has actually called assert() for some while.
2018-06-20 16:39:07 +02:00
AC_CHECK_HEADERS([errno.h \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
fcntl.h \
signal.h \
string.h \
sys/capability.h \
sys/fcntl.h \
sys/stat.h \
sys/time.h \
sys/types.h \
time.h \
unistd.h \
arpa/inet.h \
crt_externs.h \
execinfo.h \
gnu/libc-version.h \
grp.h \
ifaddrs.h \
inttypes.h \
limits.h \
linux/types.h \
Code for anonymous mappings via mmap() or CreateFileMapping(). Using an anonymous mmap() is a good way to get pages that we can set kernel-level flags on, like minherit() or madvise() or mlock(). We're going to use that so that we can make uninheritable locked pages to store PRNG data.
2019-02-05 22:58:56 +01:00
mach/vm_inherit.h \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
machine/limits.h \
malloc.h \
netdb.h \
netinet/in.h \
netinet/in6.h \
pwd.h \
readpassphrase.h \
stdatomic.h \
sys/eventfd.h \
sys/file.h \
sys/ioctl.h \
sys/limits.h \
sys/mman.h \
sys/param.h \
sys/prctl.h \
Merge branch 'osx_sierra_028'
2016-09-24 22:33:09 +02:00
sys/random.h \
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
sys/resource.h \
sys/select.h \
sys/socket.h \
sys/statvfs.h \
sys/syscall.h \
sys/sysctl.h \
sys/time.h \
sys/types.h \
sys/un.h \
sys/utime.h \
sys/wait.h \
syslog.h \
Add support for patterns on %include #25140 Also adds generic tor_glob function to expand globs.
2020-06-03 23:09:42 +02:00
utime.h \
glob.h])
Fix bug 161: when --disable-threads is set, do not search for or link against pthreads libraries svn:r4545
2005-07-13 19:31:52 +02:00
Fix some more FreeBSD4 issues (based on a patch from grarpamp) Apparently, freebsd 4 doesn't like malloc.h, needs sys/param.h for MIN/MAX, and doesn't have a SIZE_MAX. For bug 3894.
2012-05-16 20:31:36 +02:00
AC_CHECK_HEADERS(sys/param.h)
r12061@catbus: nickm | 2007-03-04 14:41:10 -0500 Run configure.in through autoupdate svn:r9729
2007-03-04 20:49:04 +01:00
AC_CHECK_HEADERS(net/if.h, net_if_found=1, net_if_found=0,
r8704@Kushana: nickm | 2006-09-02 01:32:34 -0400 correct includes for net/if.h and net/pfvar.h on openbsd (from Tup) svn:r8316
2006-09-02 07:33:11 +02:00
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
r7323@Kushana: nickm | 2006-08-10 22:04:57 -0700 Add sys/socket.h prereq for net/if.h and net/pfvar.h svn:r7019
2006-08-11 09:09:09 +02:00
#include <sys/socket.h>
#endif])
r12061@catbus: nickm | 2007-03-04 14:41:10 -0500 Run configure.in through autoupdate svn:r9729
2007-03-04 20:49:04 +01:00
AC_CHECK_HEADERS(net/pfvar.h, net_pfvar_found=1, net_pfvar_found=0,
r8704@Kushana: nickm | 2006-09-02 01:32:34 -0400 correct includes for net/if.h and net/pfvar.h on openbsd (from Tup) svn:r8316
2006-09-02 07:33:11 +02:00
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
r7323@Kushana: nickm | 2006-08-10 22:04:57 -0700 Add sys/socket.h prereq for net/if.h and net/pfvar.h svn:r7019
2006-08-11 09:09:09 +02:00
#include <sys/socket.h>
r8704@Kushana: nickm | 2006-09-02 01:32:34 -0400 correct includes for net/if.h and net/pfvar.h on openbsd (from Tup) svn:r8316
2006-09-02 07:33:11 +02:00
#endif
#ifdef HAVE_NET_IF_H
#include <net/if.h>
Include netinet/in.h (if detected) in check for net/pfvar.h Patch from rubiate; fixes bug 17551.
2015-11-25 15:27:52 +01:00
#endif
#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
r7323@Kushana: nickm | 2006-08-10 22:04:57 -0700 Add sys/socket.h prereq for net/if.h and net/pfvar.h svn:r7019
2006-08-11 09:09:09 +02:00
#endif])
Use the appropriate call to getsockopt for IPv6 sockets The original call to getsockopt to know the original address on transparently proxyed sockets using REDIRECT in iptables failed with IPv6 addresses because it assumed all sockets used IPv4. This patch fixes this by using the appropriate options and adding the headers containing the needed definitions for these. This patch is released under the same license as the original file as long as the author iscredited. Signed-off-by: Francisco Blas Izquierdo Riera (klondike) <klondike@gentoo.org>
2014-12-23 16:51:33 +01:00
AC_CHECK_HEADERS(linux/if.h,[],[],
[
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
])
r7313@Kushana: nickm | 2006-08-10 02:13:35 -0700 Add missing .h to header file name. My bad. svn:r7013
2006-08-10 11:14:57 +02:00
AC_CHECK_HEADERS(linux/netfilter_ipv4.h,
r12061@catbus: nickm | 2007-03-04 14:41:10 -0500 Run configure.in through autoupdate svn:r9729
2007-03-04 20:49:04 +01:00
linux_netfilter_ipv4=1, linux_netfilter_ipv4=0,
r11852@Kushana: nickm | 2007-01-05 18:34:27 -0500 More missing headers in configure.in in order to solve previous openbsd in6_addr issue and prevent related issues from arising. svn:r9273
2007-01-06 00:34:31 +01:00
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
r7315@Kushana: nickm | 2006-08-10 02:30:13 -0700 add missing "test" and missing include. svn:r7015
2006-08-10 11:30:25 +02:00
#include <sys/socket.h>
Patch from coderman: detect transparent proxy header correctly on some linux versions. svn:r15726
2008-07-07 17:09:26 +02:00
#endif
Oops. On recent linux kernels we were not detecting the linux/netfilter_ipv4.h header properly, since we needed to have in_addr and in6_addr and __u32 defined before trying to include it. Fix that. svn:r16460
2008-08-07 21:39:52 +02:00
#ifdef HAVE_LIMITS_H
#include <limits.h>
#endif
#ifdef HAVE_LINUX_TYPES_H
#include <linux/types.h>
#endif
#ifdef HAVE_NETINET_IN6_H
#include <netinet/in6.h>
#endif
Patch from coderman: detect transparent proxy header correctly on some linux versions. svn:r15726
2008-07-07 17:09:26 +02:00
#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
r7315@Kushana: nickm | 2006-08-10 02:30:13 -0700 add missing "test" and missing include. svn:r7015
2006-08-10 11:30:25 +02:00
#endif])
r7300@Kushana: nickm | 2006-08-10 01:36:40 -0700 Distinguish netfilter vs pf at configure time based on headers, not on OS. svn:r7008
2006-08-10 11:01:46 +02:00
Use the appropriate call to getsockopt for IPv6 sockets The original call to getsockopt to know the original address on transparently proxyed sockets using REDIRECT in iptables failed with IPv6 addresses because it assumed all sockets used IPv4. This patch fixes this by using the appropriate options and adding the headers containing the needed definitions for these. This patch is released under the same license as the original file as long as the author iscredited. Signed-off-by: Francisco Blas Izquierdo Riera (klondike) <klondike@gentoo.org>
2014-12-23 16:51:33 +01:00
AC_CHECK_HEADERS(linux/netfilter_ipv6/ip6_tables.h,
linux_netfilter_ipv6_ip6_tables=1, linux_netfilter_ipv6_ip6_tables=0,
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
#ifdef HAVE_LIMITS_H
#include <limits.h>
#endif
#ifdef HAVE_LINUX_TYPES_H
#include <linux/types.h>
#endif
#ifdef HAVE_NETINET_IN6_H
#include <netinet/in6.h>
#endif
#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
#endif
#ifdef HAVE_LINUX_IF_H
#include <linux/if.h>
#endif])
Fix #19449 Remove --disable-transparent configure option
2016-06-22 20:34:56 +02:00
transparent_ok=0
if test "x$net_if_found" = "x1" && test "x$net_pfvar_found" = "x1"; then
transparent_ok=1
fi
if test "x$linux_netfilter_ipv4" = "x1"; then
transparent_ok=1
fi
if test "x$linux_netfilter_ipv6_ip6_tables" = "x1"; then
transparent_ok=1
fi
if test "x$transparent_ok" = "x1"; then
AC_DEFINE(USE_TRANSPARENT, 1, "Define to enable transparent proxy support")
else
AC_MSG_NOTICE([Transparent proxy support enabled, but missing headers.])
r7300@Kushana: nickm | 2006-08-10 01:36:40 -0700 Distinguish netfilter vs pf at configure time based on headers, not on OS. svn:r7008
2006-08-10 11:01:46 +02:00
fi
r9478@Kushana: nickm | 2006-11-05 10:52:29 -0500 add includes to AC_CHECK_MEMBERS test for timeval.tv_sec. svn:r8906
2006-11-05 16:52:33 +01:00
AC_CHECK_MEMBERS([struct timeval.tv_sec], , ,
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif])
It is possible to have struct timeval but not gettimeofday. svn:r2125
2004-08-04 01:31:22 +02:00
autoconf around missing stdint.h svn:r376
2003-08-11 22:40:21 +02:00
AC_CHECK_SIZEOF(char)
AC_CHECK_SIZEOF(short)
AC_CHECK_SIZEOF(int)
Compile time check for being able to fit {unsigned} int inside void pointer
2019-03-22 11:57:58 +01:00
AC_CHECK_SIZEOF(unsigned int)
autoconf around missing stdint.h svn:r376
2003-08-11 22:40:21 +02:00
AC_CHECK_SIZEOF(long)
AC_CHECK_SIZEOF(long long)
AC_CHECK_SIZEOF(__int64)
Add intptr_t and uintptr_t to torint.h svn:r1794
2004-05-05 22:26:35 +02:00
AC_CHECK_SIZEOF(void *)
Finish implementing GETINFO; make it easy to query address maps. svn:r3732
2005-03-03 07:37:54 +01:00
AC_CHECK_SIZEOF(time_t)
r8750@totoro: nickm | 2006-09-27 20:52:01 -0400 Fix some warnings on mingw; hopefully this should let us build on mingw without warnings. svn:r8509
2006-09-28 02:53:02 +02:00
AC_CHECK_SIZEOF(size_t)
Compile (with warnings) with mingw64 Patch from yayooo for bug 7260, forward-ported to 0.2.4.
2012-11-02 18:48:29 +01:00
AC_CHECK_SIZEOF(pid_t)
Patch from Michael Mohr to fix cross-compilation. Backport candidate. Tweaked to use sensible defaults for NULL_REP_IS_ZERO_BYTES and TIME_T_IS_SIGNED. svn:r6474
2006-05-23 10:38:18 +02:00
r18362@catbus: nickm | 2008-02-21 22:29:14 -0500 Add missing part of last patch svn:r13671
2008-02-22 04:29:17 +01:00
AC_CHECK_TYPES([uint, u_char, ssize_t])
r8757@Kushana: nickm | 2006-09-07 13:07:46 -0400 Fix more compile warnings on Woody. svn:r8349
2006-09-07 21:00:51 +02:00
Use : rather than /bin/true in configure.ac Some platforms have true at different locations, like /usr/bin/true.
2015-09-10 09:33:59 +02:00
AC_PC_FROM_UCONTEXT([:])
Use pc_from_ucontext.m4 from Google Performance Tools This M4 module lets us learn the right way (out of at least 18 possibilities) to extract the current PC for stack-trace-fixup-in-signal purposes. The Google Performance Tools license is 3-clause BSD.
2013-07-26 12:46:17 +02:00
Remove a few no-longer-used checks from configure.in svn:r17604
2008-12-11 22:56:22 +01:00
dnl used to include sockaddr_storage, but everybody has that.
AC_CHECK_TYPES([struct in6_addr, struct sockaddr_in6, sa_family_t], , ,
r11852@Kushana: nickm | 2007-01-05 18:34:27 -0500 More missing headers in configure.in in order to solve previous openbsd in6_addr issue and prevent related issues from arising. svn:r9273
2007-01-06 00:34:31 +01:00
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_NETINET_IN_H
r11850@Kushana: nickm | 2007-01-05 17:12:08 -0500 Look in netinet/in6.h header when looking for struct in6_addr. (Issue found by saeftl.) svn:r9272
2007-01-05 23:12:15 +01:00
#include <netinet/in.h>
#endif
#ifdef HAVE_NETINET_IN6_H
#include <netinet/in6.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
r13907@Kushana: nickm | 2007-08-02 10:29:13 -0700 Try (once again) to fix mingw build wrt detecting ipv6 types. svn:r11037
2007-08-02 19:30:42 +02:00
#endif
Fix straggling MS_WINDOWS issues; add a changes file There was one MS_WINDOWS that remained because it wasn't on a macro line; a few remaining uses (and the definition!) in configure.in; and a now-nonsensical stanza of eventdns_tor.h that previously defined 'WIN32' if it didn't exist.
2012-01-31 17:04:27 +01:00
#ifdef _WIN32
Bump _WIN32_WINNT to 0x0501 throughout the code This tells the windows headers to give us definitions that didn't exist before XP -- like the ones that we need for IPv6 support. See bug #5861. We didn't run into this issue with mingw, since mingw doesn't respect _WIN32_WINNT as well as it should for some of its definitions.
2012-05-14 19:46:37 +02:00
#define _WIN32_WINNT 0x0501
r13907@Kushana: nickm | 2007-08-02 10:29:13 -0700 Try (once again) to fix mingw build wrt detecting ipv6 types. svn:r11037
2007-08-02 19:30:42 +02:00
#define WIN32_LEAN_AND_MEAN
#include <winsock2.h>
#include <ws2tcpip.h>
#endif
])
Only set sin_len/sin6_len when they exist. svn:r17851
2009-01-02 21:57:10 +01:00
AC_CHECK_MEMBERS([struct in6_addr.s6_addr32, struct in6_addr.s6_addr16, struct sockaddr_in.sin_len, struct sockaddr_in6.sin6_len], , ,
r15161@tombo: nickm | 2007-12-05 11:30:37 -0500 Fix bug reported by Steve Murphy on or-talk: detect the s6_addr32 and s6_addr16 fields via autoconf. svn:r12679
2007-12-05 17:30:52 +01:00
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
#endif
#ifdef HAVE_NETINET_IN6_H
#include <netinet/in6.h>
#endif
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
Fix straggling MS_WINDOWS issues; add a changes file There was one MS_WINDOWS that remained because it wasn't on a macro line; a few remaining uses (and the definition!) in configure.in; and a now-nonsensical stanza of eventdns_tor.h that previously defined 'WIN32' if it didn't exist.
2012-01-31 17:04:27 +01:00
#ifdef _WIN32
Bump _WIN32_WINNT to 0x0501 throughout the code This tells the windows headers to give us definitions that didn't exist before XP -- like the ones that we need for IPv6 support. See bug #5861. We didn't run into this issue with mingw, since mingw doesn't respect _WIN32_WINNT as well as it should for some of its definitions.
2012-05-14 19:46:37 +02:00
#define _WIN32_WINNT 0x0501
r15161@tombo: nickm | 2007-12-05 11:30:37 -0500 Fix bug reported by Steve Murphy on or-talk: detect the s6_addr32 and s6_addr16 fields via autoconf. svn:r12679
2007-12-05 17:30:52 +01:00
#define WIN32_LEAN_AND_MEAN
#include <winsock2.h>
#include <ws2tcpip.h>
#endif
])
r11536@Kushana: nickm | 2006-12-11 22:48:21 -0500 Add ipv6 functionality to evdns. svn:r9089
2006-12-12 04:48:46 +01:00
r14647@tombo: nickm | 2007-11-02 10:48:37 -0400 Use rlim_t instead of unsigned long to manipulate rlimit values. svn:r12339
2007-11-02 15:50:37 +01:00
AC_CHECK_TYPES([rlim_t], , ,
Make it build on OpenBSD again. Patch from tup. svn:r12489
2007-11-12 21:42:57 +01:00
[#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
r17340@catbus: nickm | 2007-12-22 12:30:07 -0500 Try to resolve bug 553 by adding a sys/time.h include to the autoconf test for rlim_t. svn:r12931
2007-12-22 18:30:12 +01:00
#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif
Make it build on OpenBSD again. Patch from tup. svn:r12489
2007-11-12 21:42:57 +01:00
#ifdef HAVE_SYS_RESOURCE_H
r14647@tombo: nickm | 2007-11-02 10:48:37 -0400 Use rlim_t instead of unsigned long to manipulate rlimit values. svn:r12339
2007-11-02 15:50:37 +01:00
#include <sys/resource.h>
#endif
])
Use a nicely written autoconf macro to determine the sign of a type This beats our old implementation, which wouldn't work when cross-compiling
2013-02-07 22:23:48 +01:00
AX_CHECK_SIGN([time_t],
Remove support for unsigned time_t We've never actually tested this support, and we should probably assume it's broken. To the best of my knowledge, only OpenVMS has this, and even on OpenVMS it's a compile-time option to disable it. And I don't think we build on openvms anyway. (Everybody else seems to be working around the 2038 problem by using a 64-bit time_t, which won't expire for roughly 292 billion years.) Closes ticket 18184.
2016-01-29 15:18:59 +01:00
[ : ],
Use a nicely written autoconf macro to determine the sign of a type This beats our old implementation, which wouldn't work when cross-compiling
2013-02-07 22:23:48 +01:00
[ : ], [
r11852@Kushana: nickm | 2007-01-05 18:34:27 -0500 More missing headers in configure.in in order to solve previous openbsd in6_addr issue and prevent related issues from arising. svn:r9273
2007-01-06 00:34:31 +01:00
#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif
#ifdef HAVE_TIME_H
#include <time.h>
#endif
Fix logic to set TIME_T_MAX; apparently, everybody had thought of the prospect of a signed time_t but me. svn:r4675
2005-07-25 20:10:08 +02:00
])
r15646@tombo: nickm | 2007-12-23 13:17:33 -0500 Handle cross-compilation more sanely: avoid most uses of the (never-defined!) CROSS_COMPILE variable; in the one place where we cannot help it, use the correct cross_compiling. svn:r12945
2007-12-23 19:20:22 +01:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$ax_cv_decl_time_t_signed" = "no"; then
Remove support for unsigned time_t We've never actually tested this support, and we should probably assume it's broken. To the best of my knowledge, only OpenVMS has this, and even on OpenVMS it's a compile-time option to disable it. And I don't think we build on openvms anyway. (Everybody else seems to be working around the 2038 problem by using a 64-bit time_t, which won't expire for roughly 292 billion years.) Closes ticket 18184.
2016-01-29 15:18:59 +01:00
AC_MSG_ERROR([You have an unsigned time_t; Tor does not support that. Please tell the Tor developers about your interesting platform.])
Warn at configure time when time_t is unsigned Inspired by #8042. As far as I know, OpenVMS is the only place you're likely to hit an unsigned time_t these days, and Tor's VMS support is... lacking. Still worth letting people know about it, though.
2013-03-11 20:33:44 +01:00
fi
Use a nicely written autoconf macro to determine the sign of a type This beats our old implementation, which wouldn't work when cross-compiling
2013-02-07 22:23:48 +01:00
AX_CHECK_SIGN([size_t],
[ tor_cv_size_t_signed=yes ],
[ tor_cv_size_t_signed=no ], [
Detect signed size_t and report an error at configure time.
2011-01-03 22:54:57 +01:00
#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$ax_cv_decl_size_t_signed" = "yes"; then
Detect signed size_t and report an error at configure time.
2011-01-03 22:54:57 +01:00
AC_MSG_ERROR([You have a signed size_t; that's grossly nonconformant.])
fi
Have autoconf check whether enums are signed. Fixes bug 7727; fix on 0.2.4.10-alpha.
2013-02-07 22:29:32 +01:00
AX_CHECK_SIGN([enum always],
[ AC_DEFINE(ENUM_VALS_ARE_SIGNED, 1, [Define if enum is always signed]) ],
[ : ], [
enum always { AAA, BBB, CCC };
])
Make Tor compile with no warnings with gcc4.0 on OSX svn:r4184
2005-05-07 07:55:06 +02:00
AC_CHECK_SIZEOF(socklen_t, , [AC_INCLUDES_DEFAULT()
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
])
autoconf around missing stdint.h svn:r376
2003-08-11 22:40:21 +02:00
Initial IRIX compatibility fixes: check for inttypes.h; check for platform cell_t. svn:r3648
2005-02-22 05:26:45 +01:00
# We want to make sure that we _don't_ have a cell_t defined, like IRIX does.
AC_CHECK_SIZEOF(cell_t)
Detect broken stdatomic.h, and pretend that it isn't there at all I hope that the debian clang maintainers will look at debian bug 903709 soon. But until they do, this should keep our users and our CI happy on sid with clang. Closes ticket 26779.
2018-08-08 21:41:40 +02:00
# Let's see if stdatomic works. (There are some debian clangs that screw it
# up; see Tor bug #26779 and debian bug 903709.)
AC_CACHE_CHECK([whether C11 stdatomic.h actually works],
tor_cv_stdatomic_works,
[AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
#include <stdatomic.h>
struct x { atomic_size_t y; };
void try_atomic_init(struct x *xx)
{
atomic_init(&xx->y, 99);
atomic_fetch_add(&xx->y, 1);
}
]])], [tor_cv_stdatomic_works=yes], [tor_cv_stdatomic_works=no])])
if test "$tor_cv_stdatomic_works" = "yes"; then
AC_DEFINE(STDATOMIC_WORKS, 1, [Set to 1 if we can compile a simple stdatomic example.])
elif test "$ac_cv_header_stdatomic_h" = "yes"; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
Detect broken stdatomic.h, and pretend that it isn't there at all I hope that the debian clang maintainers will look at debian bug 903709 soon. But until they do, this should keep our users and our CI happy on sid with clang. Closes ticket 26779.
2018-08-08 21:41:40 +02:00
AC_MSG_WARN([Your compiler provides the stdatomic.h header, but it doesn't seem to work. I'll pretend it isn't there. If you are using Clang on Debian, maybe this is because of https://bugs.debian.org/903709 ])
fi
Add an autoconf test to make sure that memset(&ptr, 0, sizeof(ptr)) and ptr=NULL are equivalent svn:r2536
2004-10-16 22:26:12 +02:00
# Now make sure that NULL can be represented as zero bytes.
AC_CACHE_CHECK([whether memset(0) sets pointers to NULL], tor_cv_null_is_zero,
[AC_RUN_IFELSE([AC_LANG_SOURCE(
[[#include <stdlib.h>
Fix check for NULL representation to include string.h; this makes fixes building on Irix, I hope. svn:r3625
2005-02-15 01:40:36 +01:00
#include <string.h>
More fixes for IRIX compilation: make very sure we have NULL in configure.in. This may work now. svn:r3627
2005-02-15 20:48:07 +01:00
#include <stdio.h>
#ifdef HAVE_STDDEF_H
#include <stddef.h>
#endif
build: fix -Wstrict-prototypes (Clang 16) Clang 16 warns on -Wstrict-prototypes in preparation for C23 which can among other things, lead to some configure tests silently failing/returning the wrong result. Fixes this error: ``` -ignoreme: warning: a function declaration without a prototype is deprecated in all versions of C [-Wstrict-prototypes] +ignoreme: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] main () ``` For more information, see LWN.net [0] or LLVM's Discourse [1], gentoo-dev@ [2], or the (new) c-std-porting mailing list [3]. [0] https://lwn.net/Articles/913505/ [1] https://discourse.llvm.org/t/configure-script-breakage-with-the-new-werror-implicit-function-declaration/65213 [2] https://archives.gentoo.org/gentoo-dev/message/dd9f2d3082b8b6f8dfbccb0639e6e240 [3] hosted at lists.linux.dev. Bug: https://bugs.gentoo.org/879747 Signed-off-by: Sam James <sam@gentoo.org> ---
2022-11-08 07:42:59 +01:00
int main (void) { char *p1,*p2; p1=NULL; memset(&p2,0,sizeof(p2));
Paranoid bulletproofing for exit code svn:r2537
2004-10-16 22:37:54 +02:00
return memcmp(&p1,&p2,sizeof(char*))?1:0; }]])],
Add an autoconf test to make sure that memset(&ptr, 0, sizeof(ptr)) and ptr=NULL are equivalent svn:r2536
2004-10-16 22:26:12 +02:00
[tor_cv_null_is_zero=yes],
[tor_cv_null_is_zero=no],
[tor_cv_null_is_zero=cross])])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_null_is_zero" = "cross"; then
Patch from Michael Mohr to fix cross-compilation. Backport candidate. Tweaked to use sensible defaults for NULL_REP_IS_ZERO_BYTES and TIME_T_IS_SIGNED. svn:r6474
2006-05-23 10:38:18 +02:00
# Cross-compiling; let's hope that the target isn't raving mad.
AC_MSG_NOTICE([Cross-compiling: we'll assume that NULL is represented as a sequence of 0-valued bytes.])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_null_is_zero" != "no"; then
Add an autoconf test to make sure that memset(&ptr, 0, sizeof(ptr)) and ptr=NULL are equivalent svn:r2536
2004-10-16 22:26:12 +02:00
AC_DEFINE([NULL_REP_IS_ZERO_BYTES], 1,
[Define to 1 iff memset(0) sets pointers to NULL])
fi
Detect platforms where memset(0) doesn't set doubles to 0.0. This is allowed by the C statndard, which permits you to represent doubles any way you like, but in practice we have some code that assumes that memset() clears doubles in structs. Noticed as part of 7802 review; see 8081 for more info.
2013-01-29 23:38:15 +01:00
AC_CACHE_CHECK([whether memset(0) sets doubles to 0.0], tor_cv_dbl0_is_zero,
[AC_RUN_IFELSE([AC_LANG_SOURCE(
[[#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#ifdef HAVE_STDDEF_H
#include <stddef.h>
#endif
build: fix -Wstrict-prototypes (Clang 16) Clang 16 warns on -Wstrict-prototypes in preparation for C23 which can among other things, lead to some configure tests silently failing/returning the wrong result. Fixes this error: ``` -ignoreme: warning: a function declaration without a prototype is deprecated in all versions of C [-Wstrict-prototypes] +ignoreme: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] main () ``` For more information, see LWN.net [0] or LLVM's Discourse [1], gentoo-dev@ [2], or the (new) c-std-porting mailing list [3]. [0] https://lwn.net/Articles/913505/ [1] https://discourse.llvm.org/t/configure-script-breakage-with-the-new-werror-implicit-function-declaration/65213 [2] https://archives.gentoo.org/gentoo-dev/message/dd9f2d3082b8b6f8dfbccb0639e6e240 [3] hosted at lists.linux.dev. Bug: https://bugs.gentoo.org/879747 Signed-off-by: Sam James <sam@gentoo.org> ---
2022-11-08 07:42:59 +01:00
int main (void) { double d1,d2; d1=0; memset(&d2,0,sizeof(d2));
Detect platforms where memset(0) doesn't set doubles to 0.0. This is allowed by the C statndard, which permits you to represent doubles any way you like, but in practice we have some code that assumes that memset() clears doubles in structs. Noticed as part of 7802 review; see 8081 for more info.
2013-01-29 23:38:15 +01:00
return memcmp(&d1,&d2,sizeof(d1))?1:0; }]])],
[tor_cv_dbl0_is_zero=yes],
[tor_cv_dbl0_is_zero=no],
[tor_cv_dbl0_is_zero=cross])])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_dbl0_is_zero" = "cross"; then
Detect platforms where memset(0) doesn't set doubles to 0.0. This is allowed by the C statndard, which permits you to represent doubles any way you like, but in practice we have some code that assumes that memset() clears doubles in structs. Noticed as part of 7802 review; see 8081 for more info.
2013-01-29 23:38:15 +01:00
# Cross-compiling; let's hope that the target isn't raving mad.
AC_MSG_NOTICE([Cross-compiling: we'll assume that 0.0 can be represented as a sequence of 0-valued bytes.])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_dbl0_is_zero" != "no"; then
Detect platforms where memset(0) doesn't set doubles to 0.0. This is allowed by the C statndard, which permits you to represent doubles any way you like, but in practice we have some code that assumes that memset() clears doubles in structs. Noticed as part of 7802 review; see 8081 for more info.
2013-01-29 23:38:15 +01:00
AC_DEFINE([DOUBLE_0_REP_IS_ZERO_BYTES], 1,
[Define to 1 iff memset(0) sets doubles to 0.0])
fi
Malloc and friends are critical-path: Thus, add an it-wont-happen branch prediction for NULL returns, and skip the malloc(0) check on platforms where malloc(0) returns a pointer. svn:r8431
2006-09-20 00:36:48 +02:00
# And what happens when we malloc zero?
AC_CACHE_CHECK([whether we can malloc(0) safely.], tor_cv_malloc_zero_works,
[AC_RUN_IFELSE([AC_LANG_SOURCE(
[[#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#ifdef HAVE_STDDEF_H
#include <stddef.h>
#endif
build: fix -Wstrict-prototypes (Clang 16) Clang 16 warns on -Wstrict-prototypes in preparation for C23 which can among other things, lead to some configure tests silently failing/returning the wrong result. Fixes this error: ``` -ignoreme: warning: a function declaration without a prototype is deprecated in all versions of C [-Wstrict-prototypes] +ignoreme: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] main () ``` For more information, see LWN.net [0] or LLVM's Discourse [1], gentoo-dev@ [2], or the (new) c-std-porting mailing list [3]. [0] https://lwn.net/Articles/913505/ [1] https://discourse.llvm.org/t/configure-script-breakage-with-the-new-werror-implicit-function-declaration/65213 [2] https://archives.gentoo.org/gentoo-dev/message/dd9f2d3082b8b6f8dfbccb0639e6e240 [3] hosted at lists.linux.dev. Bug: https://bugs.gentoo.org/879747 Signed-off-by: Sam James <sam@gentoo.org> ---
2022-11-08 07:42:59 +01:00
int main (void) { return malloc(0)?0:1; }]])],
Malloc and friends are critical-path: Thus, add an it-wont-happen branch prediction for NULL returns, and skip the malloc(0) check on platforms where malloc(0) returns a pointer. svn:r8431
2006-09-20 00:36:48 +02:00
[tor_cv_malloc_zero_works=yes],
[tor_cv_malloc_zero_works=no],
[tor_cv_malloc_zero_works=cross])])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_malloc_zero_works" = "cross"; then
Malloc and friends are critical-path: Thus, add an it-wont-happen branch prediction for NULL returns, and skip the malloc(0) check on platforms where malloc(0) returns a pointer. svn:r8431
2006-09-20 00:36:48 +02:00
# Cross-compiling; let's hope that the target isn't raving mad.
AC_MSG_NOTICE([Cross-compiling: we'll assume that we need to check malloc() arguments for 0.])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_malloc_zero_works" = "yes"; then
Malloc and friends are critical-path: Thus, add an it-wont-happen branch prediction for NULL returns, and skip the malloc(0) check on platforms where malloc(0) returns a pointer. svn:r8431
2006-09-20 00:36:48 +02:00
AC_DEFINE([MALLOC_ZERO_WORKS], 1,
[Define to 1 iff malloc(0) returns a pointer])
fi
r8844@totoro: nickm | 2006-10-02 16:32:44 -0400 Resolve an XXXX: make sure that we are actually on 2s-complement hardware. svn:r8576
2006-10-02 23:00:20 +02:00
# whether we seem to be in a 2s-complement world.
AC_CACHE_CHECK([whether we are using 2s-complement arithmetic], tor_cv_twos_complement,
[AC_RUN_IFELSE([AC_LANG_SOURCE(
build: fix -Wstrict-prototypes (Clang 16) Clang 16 warns on -Wstrict-prototypes in preparation for C23 which can among other things, lead to some configure tests silently failing/returning the wrong result. Fixes this error: ``` -ignoreme: warning: a function declaration without a prototype is deprecated in all versions of C [-Wstrict-prototypes] +ignoreme: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] main () ``` For more information, see LWN.net [0] or LLVM's Discourse [1], gentoo-dev@ [2], or the (new) c-std-porting mailing list [3]. [0] https://lwn.net/Articles/913505/ [1] https://discourse.llvm.org/t/configure-script-breakage-with-the-new-werror-implicit-function-declaration/65213 [2] https://archives.gentoo.org/gentoo-dev/message/dd9f2d3082b8b6f8dfbccb0639e6e240 [3] hosted at lists.linux.dev. Bug: https://bugs.gentoo.org/879747 Signed-off-by: Sam James <sam@gentoo.org> ---
2022-11-08 07:42:59 +01:00
[[int main (void) { int problem = ((-99) != (~99)+1);
r8844@totoro: nickm | 2006-10-02 16:32:44 -0400 Resolve an XXXX: make sure that we are actually on 2s-complement hardware. svn:r8576
2006-10-02 23:00:20 +02:00
return problem ? 1 : 0; }]])],
[tor_cv_twos_complement=yes],
[tor_cv_twos_complement=no],
[tor_cv_twos_complement=cross])])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_twos_complement" = "cross"; then
r8844@totoro: nickm | 2006-10-02 16:32:44 -0400 Resolve an XXXX: make sure that we are actually on 2s-complement hardware. svn:r8576
2006-10-02 23:00:20 +02:00
# Cross-compiling; let's hope that the target isn't raving mad.
r10016@danube: weasel | 2006-10-07 03:25:51 +0200 Fix a configure notice on crosscompile svn:r8625
2006-10-07 03:26:41 +02:00
AC_MSG_NOTICE([Cross-compiling: we'll assume that negative integers are represented with two's complement.])
r8844@totoro: nickm | 2006-10-02 16:32:44 -0400 Resolve an XXXX: make sure that we are actually on 2s-complement hardware. svn:r8576
2006-10-02 23:00:20 +02:00
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_twos_complement" != "no"; then
r8844@totoro: nickm | 2006-10-02 16:32:44 -0400 Resolve an XXXX: make sure that we are actually on 2s-complement hardware. svn:r8576
2006-10-02 23:00:20 +02:00
AC_DEFINE([USING_TWOS_COMPLEMENT], 1,
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
[Define to 1 iff we represent negative integers with
two's complement])
r8844@totoro: nickm | 2006-10-02 16:32:44 -0400 Resolve an XXXX: make sure that we are actually on 2s-complement hardware. svn:r8576
2006-10-02 23:00:20 +02:00
fi
Add a data-independent variant of memcmp and a d-i memeq function. The tor_memcmp code is by Robert Ransom, and the tor_memeq code is by me. Both incorporate some ideas from DJB's stuff.
2011-05-10 00:39:23 +02:00
# What does shifting a negative value do?
AC_CACHE_CHECK([whether right-shift on negative values does sign-extension], tor_cv_sign_extend,
[AC_RUN_IFELSE([AC_LANG_SOURCE(
build: fix -Wstrict-prototypes (Clang 16) Clang 16 warns on -Wstrict-prototypes in preparation for C23 which can among other things, lead to some configure tests silently failing/returning the wrong result. Fixes this error: ``` -ignoreme: warning: a function declaration without a prototype is deprecated in all versions of C [-Wstrict-prototypes] +ignoreme: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] main () ``` For more information, see LWN.net [0] or LLVM's Discourse [1], gentoo-dev@ [2], or the (new) c-std-porting mailing list [3]. [0] https://lwn.net/Articles/913505/ [1] https://discourse.llvm.org/t/configure-script-breakage-with-the-new-werror-implicit-function-declaration/65213 [2] https://archives.gentoo.org/gentoo-dev/message/dd9f2d3082b8b6f8dfbccb0639e6e240 [3] hosted at lists.linux.dev. Bug: https://bugs.gentoo.org/879747 Signed-off-by: Sam James <sam@gentoo.org> ---
2022-11-08 07:42:59 +01:00
[[int main (void) { int okay = (-60 >> 8) == -1; return okay ? 0 : 1; }]])],
Add a data-independent variant of memcmp and a d-i memeq function. The tor_memcmp code is by Robert Ransom, and the tor_memeq code is by me. Both incorporate some ideas from DJB's stuff.
2011-05-10 00:39:23 +02:00
[tor_cv_sign_extend=yes],
[tor_cv_sign_extend=no],
[tor_cv_sign_extend=cross])])
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_sign_extend" = "cross"; then
Add a data-independent variant of memcmp and a d-i memeq function. The tor_memcmp code is by Robert Ransom, and the tor_memeq code is by me. Both incorporate some ideas from DJB's stuff.
2011-05-10 00:39:23 +02:00
# Cross-compiling; let's hope that the target isn't raving mad.
AC_MSG_NOTICE([Cross-compiling: we'll assume that right-shifting negative integers causes sign-extension])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_sign_extend" != "no"; then
Add a data-independent variant of memcmp and a d-i memeq function. The tor_memcmp code is by Robert Ransom, and the tor_memeq code is by me. Both incorporate some ideas from DJB's stuff.
2011-05-10 00:39:23 +02:00
AC_DEFINE([RSHIFT_DOES_SIGN_EXTEND], 1,
[Define to 1 iff right-shifting a negative value performs sign-extension])
fi
Ensure that uint8_t is unsigned char Many places in our code assume that uint8_t is the same type as unsigned char. Test this assumption in the configure script. This is important because of the privileged aliasing properties of character types in C. Fixes #22410.
2017-05-26 23:04:52 +02:00
# Is uint8_t the same type as unsigned char?
AC_CACHE_CHECK([whether uint8_t is the same type as unsigned char], tor_cv_uint8_uchar,
[AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
#include <stdint.h>
extern uint8_t c;
unsigned char c;]])],
[tor_cv_uint8_uchar=yes],
[tor_cv_uint8_uchar=no],
[tor_cv_uint8_uchar=cross])])
if test "$tor_cv_uint8_uchar" = "cross"; then
AC_MSG_NOTICE([Cross-compiling: we'll assume that uint8_t is the same type as unsigned char])
fi
if test "$tor_cv_uint8_uchar" = "no"; then
AC_MSG_ERROR([We assume that uint8_t is the same type as unsigned char, but your compiler disagrees.])
fi
r14240@tombo: nickm | 2008-02-17 14:26:52 -0500 Add a --with-tcmalloc argument to configure. svn:r13546
2008-02-17 20:26:55 +01:00
AC_ARG_WITH(tcmalloc,
configure: Interpret --with-tcmalloc=no correctly Fixes bug 32124; bugfix on 0.2.0.20-rc.
2019-10-18 09:16:32 +02:00
AS_HELP_STRING(--with-tcmalloc, [use tcmalloc memory allocation library. Deprecated; see --with-malloc]))
r14240@tombo: nickm | 2008-02-17 14:26:52 -0500 Add a --with-tcmalloc argument to configure. svn:r13546
2008-02-17 20:26:55 +01:00
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
default_malloc=system
r14240@tombo: nickm | 2008-02-17 14:26:52 -0500 Add a --with-tcmalloc argument to configure. svn:r13546
2008-02-17 20:26:55 +01:00
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
if test "x$enable_openbsd_malloc" = "xyes" ; then
AC_MSG_NOTICE([The --enable-openbsd-malloc argument is deprecated; use --with-malloc=openbsd instead.])
default_malloc=openbsd
Remove tor_malloc_roundup(). This function never actually did us any good, and it added a little complexity. See the changes file for more info.
2012-08-13 19:27:32 +02:00
fi
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
configure: Interpret --with-tcmalloc=no correctly Fixes bug 32124; bugfix on 0.2.0.20-rc.
2019-10-18 09:16:32 +02:00
if test "x$with_tcmalloc" = "xyes"; then
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
AC_MSG_NOTICE([The --with-tcmalloc argument is deprecated; use --with-malloc=tcmalloc instead.])
default_malloc=tcmalloc
Remove tor_malloc_roundup(). This function never actually did us any good, and it added a little complexity. See the changes file for more info.
2012-08-13 19:27:32 +02:00
fi
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
AC_ARG_WITH(malloc,
AS_HELP_STRING([--with-malloc=[system,jemalloc,tcmalloc,openbsd]],
[select special malloc implementation [system]]),
[ malloc="$with_malloc" ], [ malloc="$default_malloc" ])
AS_CASE([$malloc],
[tcmalloc], [
PKG_CHECK_MODULES([TCMALLOC],
[libtcmalloc],
have_tcmalloc=yes,
have_tcmalloc=no)
if test "x$have_tcmalloc" = "xno" ; then
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
AC_MSG_ERROR([Unable to find tcmalloc requested by --with-malloc, $pkg_config_user_action, or set TCMALLOC_CFLAGS and TCMALLOC_LIBS.])
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
fi
CFLAGS="$CFLAGS $TCMALLOC_CFLAGS"
LIBS="$TCMALLOC_LIBS $LIBS"
],
[jemalloc], [
PKG_CHECK_MODULES([JEMALLOC],
[jemalloc],
have_jemalloc=yes,
have_jemalloc=no)
configure: Stop failing when jemalloc is requested, but tcmalloc is not found Fixes bug 32124; bugfix on 0.3.5.1-alpha.
2019-10-18 09:18:52 +02:00
if test "x$have_jemalloc" = "xno" ; then
configure: List the env vars needed if pkg-config doesn't work Part of 31922.
2019-10-03 08:53:12 +02:00
AC_MSG_ERROR([Unable to find jemalloc requested by --with-malloc, $pkg_config_user_action, or set JEMALLOC_CFLAGS and JEMALLOC_LIBS.])
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
fi
CFLAGS="$CFLAGS $JEMALLOC_CFLAGS"
LIBS="$JEMALLOC_LIBS $LIBS"
using_custom_malloc=yes
],
[openbsd], [
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
Update AC_WARN->AC_MSG_WARN. #23878.
2018-08-04 12:53:21 +02:00
AC_MSG_WARN([The openbsd malloc port is deprecated in Tor 0.3.5 and will be removed in a future version.])
Add a --with-malloc option. Based on a patch from Hello71 on ticket 20424. This patch additionally fixes openbsd-malloc support, switches our tcmalloc support to use pkgconfig, and tells the compiler to omit system malloc implementations as appropriate. squash! Add a --with-malloc option. Edit changelog file to fix a typo and credit Alex Xu in preferred format.
2018-07-11 15:28:23 +02:00
enable_openbsd_malloc=yes
],
[system], [
# handle this later, including the jemalloc fallback
],
[AC_MSG_ERROR([--with-malloc=`$with_malloc' not supported, see --help])
])
AM_CONDITIONAL(USE_OPENBSD_MALLOC, test "x$enable_openbsd_malloc" = "xyes")
if test "$malloc" != "system"; then
# Tell the C compiler not to use the system allocator functions.
TOR_CHECK_CFLAGS([-fno-builtin-malloc -fno-builtin-realloc -fno-builtin-calloc -fno-builtin-free])
Remove tor_malloc_roundup(). This function never actually did us any good, and it added a little complexity. See the changes file for more info.
2012-08-13 19:27:32 +02:00
fi
When setting a nonstandard malloc, disable the system malloc. Closes ticket 20424.
2018-07-11 15:36:24 +02:00
if test "$using_custom_malloc" = "yes"; then
# Tell the C compiler not to use the system allocator functions.
TOR_CHECK_CFLAGS([-fno-builtin-malloc -fno-builtin-realloc -fno-builtin-calloc -fno-builtin-free])
fi
Remove tor_malloc_roundup(). This function never actually did us any good, and it added a little complexity. See the changes file for more info.
2012-08-13 19:27:32 +02:00
Fix compilation with with bionic libc. This fixes bug 1147: bionic doesn't have an actual implementation of mlockall(); mlockall() is merely in the headers but not actually in the library. This prevents Tor compilation with the bionic libc for Android handsets.
2009-11-14 22:43:22 +01:00
# By default, we're going to assume we don't have mlockall()
Fix compilation on OSX 10.3. On this OSX version, there is a stub mlockall() function that doesn't work, *and* the declaration for it is hidden by an '#ifdef _P1003_1B_VISIBLE'. This would make autoconf successfully find the function, but our code fail to build when no declaration was found. This patch adds an additional test for the declaration.
2009-11-20 19:28:16 +01:00
# bionic and other platforms have various broken mlockall subsystems.
# Some systems don't have a working mlockall, some aren't linkable,
# and some have it but don't declare it.
Fix compilation with with bionic libc. This fixes bug 1147: bionic doesn't have an actual implementation of mlockall(); mlockall() is merely in the headers but not actually in the library. This prevents Tor compilation with the bionic libc for Android handsets.
2009-11-14 22:43:22 +01:00
AC_CHECK_FUNCS(mlockall)
Fix compilation on OSX 10.3. On this OSX version, there is a stub mlockall() function that doesn't work, *and* the declaration for it is hidden by an '#ifdef _P1003_1B_VISIBLE'. This would make autoconf successfully find the function, but our code fail to build when no declaration was found. This patch adds an additional test for the declaration.
2009-11-20 19:28:16 +01:00
AC_CHECK_DECLS([mlockall], , , [
#ifdef HAVE_SYS_MMAN_H
#include <sys/mman.h>
#endif])
Fix compilation with with bionic libc. This fixes bug 1147: bionic doesn't have an actual implementation of mlockall(); mlockall() is merely in the headers but not actually in the library. This prevents Tor compilation with the bionic libc for Android handsets.
2009-11-14 22:43:22 +01:00
Apply Matt Ghali's --with-syslog-facility patch svn:r5883
2006-02-01 03:15:44 +01:00
# Allow user to specify an alternate syslog facility
AC_ARG_WITH(syslog-facility,
Be consistent with capitalization, quoting, and spacing of help strings.
2015-02-19 09:34:14 +01:00
AS_HELP_STRING(--with-syslog-facility=LOG, [syslog facility to use (default=LOG_DAEMON)]),
Apply Matt Ghali's --with-syslog-facility patch svn:r5883
2006-02-01 03:15:44 +01:00
syslog_facility="$withval", syslog_facility="LOG_DAEMON")
AC_DEFINE_UNQUOTED(LOGFACILITY,$syslog_facility,[name of the syslog facility])
AC_SUBST(LOGFACILITY)
Patch from Jacob Appelbaum and me to make User option more robust, properly set supplementary groups, deprecated the Group option, and log more information on credential switching svn:r17200
2008-11-07 03:06:12 +01:00
# Check if we have getresuid and getresgid
AC_CHECK_FUNCS(getresuid getresgid)
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
# Check for gethostbyname_r in all its glorious incompatible versions.
# (This logic is based on that in Python's configure.in)
AH_TEMPLATE(HAVE_GETHOSTBYNAME_R,
[Define this if you have any gethostbyname_r()])
AC_CHECK_FUNC(gethostbyname_r, [
AC_MSG_CHECKING([how many arguments gethostbyname_r() wants])
OLD_CFLAGS=$CFLAGS
CFLAGS="$CFLAGS $MY_CPPFLAGS $MY_THREAD_CPPFLAGS $MY_CFLAGS"
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
#include <netdb.h>
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]], [[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
char *cp1, *cp2;
struct hostent *h1, *h2;
int i1, i2;
(void)gethostbyname_r(cp1,h1,cp2,i1,&h2,&i2);
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
]])],[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
AC_DEFINE(HAVE_GETHOSTBYNAME_R)
AC_DEFINE(HAVE_GETHOSTBYNAME_R_6_ARG, 1,
[Define this if gethostbyname_r takes 6 arguments])
AC_MSG_RESULT(6)
], [
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
#include <netdb.h>
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]], [[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
char *cp1, *cp2;
struct hostent *h1;
int i1, i2;
(void)gethostbyname_r(cp1,h1,cp2,i1,&i2);
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]])], [
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
AC_DEFINE(HAVE_GETHOSTBYNAME_R)
AC_DEFINE(HAVE_GETHOSTBYNAME_R_5_ARG, 1,
[Define this if gethostbyname_r takes 5 arguments])
AC_MSG_RESULT(5)
], [
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
#include <netdb.h>
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]], [[
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
char *cp1;
struct hostent *h1;
struct hostent_data hd;
(void) gethostbyname_r(cp1,h1,&hd);
Replace obsolete macros with modern equivalents
2016-03-28 21:59:35 +02:00
]])], [
Snarf some logic from python, adapted to our own needs, to handle gethostbyname_r correctly across platforms. svn:r3649
2005-02-22 05:50:31 +01:00
AC_DEFINE(HAVE_GETHOSTBYNAME_R)
AC_DEFINE(HAVE_GETHOSTBYNAME_R_3_ARG, 1,
[Define this if gethostbyname_r takes 3 arguments])
AC_MSG_RESULT(3)
], [
AC_MSG_RESULT(0)
])
])
])
CFLAGS=$OLD_CFLAGS
])
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
AC_CACHE_CHECK([whether the C compiler supports __func__],
r11592@catbus: nickm | 2007-01-29 18:09:16 -0500 Oops. Apparently, we weren't supposed to call our autoconf cached variables ac_cv_*; these are reserved. svn:r9460
2007-01-30 00:09:26 +01:00
tor_cv_have_func_macro,
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
AC_COMPILE_IFELSE([AC_LANG_SOURCE([
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
#include <stdio.h>
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
int main(int c, char **v) { puts(__func__); }])],
r11592@catbus: nickm | 2007-01-29 18:09:16 -0500 Oops. Apparently, we weren't supposed to call our autoconf cached variables ac_cv_*; these are reserved. svn:r9460
2007-01-30 00:09:26 +01:00
tor_cv_have_func_macro=yes,
tor_cv_have_func_macro=no))
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
AC_CACHE_CHECK([whether the C compiler supports __FUNC__],
r11592@catbus: nickm | 2007-01-29 18:09:16 -0500 Oops. Apparently, we weren't supposed to call our autoconf cached variables ac_cv_*; these are reserved. svn:r9460
2007-01-30 00:09:26 +01:00
tor_cv_have_FUNC_macro,
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
AC_COMPILE_IFELSE([AC_LANG_SOURCE([
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
#include <stdio.h>
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
int main(int c, char **v) { puts(__FUNC__); }])],
r11592@catbus: nickm | 2007-01-29 18:09:16 -0500 Oops. Apparently, we weren't supposed to call our autoconf cached variables ac_cv_*; these are reserved. svn:r9460
2007-01-30 00:09:26 +01:00
tor_cv_have_FUNC_macro=yes,
tor_cv_have_FUNC_macro=no))
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
AC_CACHE_CHECK([whether the C compiler supports __FUNCTION__],
r11592@catbus: nickm | 2007-01-29 18:09:16 -0500 Oops. Apparently, we weren't supposed to call our autoconf cached variables ac_cv_*; these are reserved. svn:r9460
2007-01-30 00:09:26 +01:00
tor_cv_have_FUNCTION_macro,
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
AC_COMPILE_IFELSE([AC_LANG_SOURCE([
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
#include <stdio.h>
Fix warnings with new versions of autoconf It looks like autoconf 2.68 introduced a bunch of new warnings when it didn't like the syntax you used or forgot to use AC_LANG_(SOURCE|PROGRAM).
2010-10-11 11:55:30 +02:00
int main(int c, char **v) { puts(__FUNCTION__); }])],
r11592@catbus: nickm | 2007-01-29 18:09:16 -0500 Oops. Apparently, we weren't supposed to call our autoconf cached variables ac_cv_*; these are reserved. svn:r9460
2007-01-30 00:09:26 +01:00
tor_cv_have_FUNCTION_macro=yes,
tor_cv_have_FUNCTION_macro=no))
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
Revert "Stop checking whether environ is declared." This reverts commit 954eeda619a59dae76144ad69967f0ed7341b564. Apparently, OpenBSD is what expects you to declare environ yourself. So 19142 is a wontfix.
2016-12-16 18:16:52 +01:00
AC_CACHE_CHECK([whether we have extern char **environ already declared],
tor_cv_have_environ_declared,
AC_COMPILE_IFELSE([AC_LANG_SOURCE([
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <stdlib.h>
int main(int c, char **v) { char **t = environ; }])],
tor_cv_have_environ_declared=yes,
tor_cv_have_environ_declared=no))
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_have_func_macro" = "yes"; then
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
AC_DEFINE(HAVE_MACRO__func__, 1, [Defined if the compiler supports __func__])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_have_FUNC_macro" = "yes"; then
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
AC_DEFINE(HAVE_MACRO__FUNC__, 1, [Defined if the compiler supports __FUNC__])
fi
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$tor_cv_have_FUNCTION_macro" = "yes"; then
Instead of listing a set of compilers that prefers __func__ to __FUNCTION__, use autoconf. Also, prefer __func__ in our own code: __func__ is a C99 standard, whereas __FUNCTION__ is not. [Fixes bug 254.] svn:r6144
2006-03-13 01:25:36 +01:00
AC_DEFINE(HAVE_MACRO__FUNCTION__, 1,
[Defined if the compiler supports __FUNCTION__])
fi
Revert "Stop checking whether environ is declared." This reverts commit 954eeda619a59dae76144ad69967f0ed7341b564. Apparently, OpenBSD is what expects you to declare environ yourself. So 19142 is a wontfix.
2016-12-16 18:16:52 +01:00
if test "$tor_cv_have_environ_declared" = "yes"; then
AC_DEFINE(HAVE_EXTERN_ENVIRON_DECLARED, 1,
[Defined if we have extern char **environ already declared])
fi
- cause configure to create a tor.sh which will have directories set correctly based on how configure was run - cause tor to guess the location of torrc more intelligently - cause cause src/config/torrc and src/conf/sample-server-torrc to be generated with contents that are correct for the way configure was run - cause "make install" to put torrc, sample-server-torrc, and dirservers somewhere intelligent svn:r587
2003-10-15 09:19:38 +02:00
# $prefix stores the value of the --prefix command line option, or
# NONE if the option wasn't set. In the case that it wasn't set, make
# it be the default, so that we can use it to expand directories now.
if test "x$prefix" = "xNONE"; then
prefix=$ac_default_prefix
fi
# and similarly for $exec_prefix
if test "x$exec_prefix" = "xNONE"; then
exec_prefix=$prefix
fi
Fix running unit tests from outside of the build directory (fixes bug #2051) Currently the unit tests test_util_spawn_background_* assume that they are run from the Tor build directory. This is not the case when running make distcheck, so the test will fail. This problem is fixed by autoconf setting BUILDDIR to be the root of the Tor build directory, and this preprocessor variable being used to specify the absolute path to test-child. Also, in test-child, do not print out argv[0] because this will no longer be predictable. Found by Sebastian Hahn.
2010-10-12 00:22:52 +02:00
if test "x$BUILDDIR" = "x"; then
BUILDDIR=`pwd`
fi
AC_SUBST(BUILDDIR)
AH_TEMPLATE([BUILDDIR],[tor's build directory])
AC_DEFINE_UNQUOTED(BUILDDIR,"$BUILDDIR")
Add a test for geoip_load_file().
2018-03-15 15:21:34 +01:00
if test "x$SRCDIR" = "x"; then
Define SRCDIR more correctly in configure.ac Without this fix, we were just getting SRCDIR=`pwd`, which naturally was breaking out-of-tree builds. Bugfix on becae4c943969a4f4f14423cc897d39f41af7773; bug not in any released Tor.
2018-03-23 17:38:27 +01:00
SRCDIR=$(cd "$srcdir"; pwd)
Add a test for geoip_load_file().
2018-03-15 15:21:34 +01:00
fi
AH_TEMPLATE([SRCDIR],[tor's sourcedir directory])
AC_DEFINE_UNQUOTED(SRCDIR,"$SRCDIR")
Build correctly for use on OS X platforms with case-sensitive filesystems svn:r8487
2006-09-24 23:45:02 +02:00
if test "x$CONFDIR" = "x"; then
CONFDIR=`eval echo $sysconfdir/tor`
fi
- cause configure to create a tor.sh which will have directories set correctly based on how configure was run - cause tor to guess the location of torrc more intelligently - cause cause src/config/torrc and src/conf/sample-server-torrc to be generated with contents that are correct for the way configure was run - cause "make install" to put torrc, sample-server-torrc, and dirservers somewhere intelligent svn:r587
2003-10-15 09:19:38 +02:00
AC_SUBST(CONFDIR)
Fix configure.in to not produce broken configure files with more recent versions of autoconf. Thanks to Clint for his auto* voodoo. svn:r6632
2006-06-16 04:04:04 +02:00
AH_TEMPLATE([CONFDIR],[tor's configuration directory])
- cause configure to create a tor.sh which will have directories set correctly based on how configure was run - cause tor to guess the location of torrc more intelligently - cause cause src/config/torrc and src/conf/sample-server-torrc to be generated with contents that are correct for the way configure was run - cause "make install" to put torrc, sample-server-torrc, and dirservers somewhere intelligent svn:r587
2003-10-15 09:19:38 +02:00
AC_DEFINE_UNQUOTED(CONFDIR,"$CONFDIR")
r16663@catbus: nickm | 2007-11-13 21:51:28 -0500 Back out parts of r12376: replace file locations in scripts. svn:r12490
2007-11-14 03:55:12 +01:00
BINDIR=`eval echo $bindir`
AC_SUBST(BINDIR)
LOCALSTATEDIR=`eval echo $localstatedir`
AC_SUBST(LOCALSTATEDIR)
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$bwin32" = "true"; then
Enable ASLR and permanent DEP for Windows executables Fix for #2358
2011-01-07 06:08:27 +01:00
# Test if the linker supports the --nxcompat and --dynamicbase options
# for Windows
save_LDFLAGS="$LDFLAGS"
LDFLAGS="-Wl,--nxcompat -Wl,--dynamicbase"
AC_MSG_CHECKING([whether the linker supports DllCharacteristics])
AC_LINK_IFELSE([AC_LANG_PROGRAM([])],
[AC_MSG_RESULT([yes])]
[save_LDFLAGS="$save_LDFLAGS $LDFLAGS"],
[AC_MSG_RESULT([no])]
)
LDFLAGS="$save_LDFLAGS"
fi
Use a stricter set of warnings; make them all pass. svn:r2645
2004-11-02 04:02:17 +01:00
# Set CFLAGS _after_ all the above checks, since our warnings are stricter
# than autoconf's macros like.
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$GCC" = "yes"; then
When building with GCC, use -fno-strict-aliasing GCC's interpretation of the C99 aliasing rules, to be charitable, creates a dialect of C intended for a better programmers than I am certain of my ability to be in all times. I just spent 2 hours tracking down a platform-hyperspecific libevent bug that turned out to be because of this, and darned if I ever want to do *that* again. One of Linus's recent rants will give you a picture of why GCC's behavior here can lead to fun surprises in your binaries: http://lwn.net/Articles/316126/ svn:r18351
2009-01-31 08:51:02 +01:00
# Disable GCC's strict aliasing checks. They are an hours-to-debug
# accident waiting to happen.
Better workaround for CFLAGS issues from #14072; fixes #14162 When I applied patch fcc78e5f8a3249eadfea31db, I somehow broke stack trace symbols on Linux. I'll leave it to others to figure out why that happens. This should be better. Really. Fixes bug 14162; bug not in any released version of Tor.
2015-01-10 23:21:11 +01:00
CFLAGS="$CFLAGS -Wall -fno-strict-aliasing"
On non-gcc compilers, use -g -O instead of -Wall -g -O2: resolve bug 273. svn:r6245
2006-03-27 02:43:39 +02:00
else
Use package-config output for -lsystemd correctly In systemd 209, they deprecated -lsystemd-daemon in favor of -lsystemd. So we'd better actually look at the pkg-config output, or we'll get warnings on newer distributions. For some as-yet-unknown-to-me reason, setting CFLAGS so early makes it so -O2 -g doesn't get added to it later. So, adding it myself later. Perhaps a better fix here can be found. Fixes 14072; bugfix on 0.2.6.2-alpha. Based on a patch from h.venev
2015-01-06 20:05:35 +01:00
# Override optimization level for non-gcc compilers
Remove superfluous -g -O2 compiler argument Autoconf adds -g -O2 by default, so adding it ourselves is not required. It also caused a warning with clang for every source file, so remove it here. Fixes last issue of ticket 2696.
2011-03-18 17:04:01 +01:00
CFLAGS="$CFLAGS -O"
r8948@totoro: nickm | 2006-10-08 17:04:53 -0400 Add an --enable-gcc-warnings flag to configure.in svn:r8655
2006-10-08 23:05:11 +02:00
enable_gcc_warnings=no
New configure option: --enable-gcc-warnings-advisory the new configure option --enable-gcc-warnings-advisory enables all the gcc flags that --enable-gcc-warnings does with the exception of -Werror.
2010-01-24 19:32:41 +01:00
enable_gcc_warnings_advisory=no
On non-gcc compilers, use -g -O instead of -Wall -g -O2: resolve bug 273. svn:r6245
2006-03-27 02:43:39 +02:00
fi
r8948@totoro: nickm | 2006-10-08 17:04:53 -0400 Add an --enable-gcc-warnings flag to configure.in svn:r8655
2006-10-08 23:05:11 +02:00
Make advisory-warnings on by default. Add --enable-fatal-warnings to control -Werror. Closes ticket 19044.
2016-05-23 20:39:56 +02:00
# Warnings implies advisory-warnings and -Werror.
if test "$enable_gcc_warnings" = "yes"; then
enable_gcc_warnings_advisory=yes
enable_fatal_warnings=yes
fi
Ignore deprecation warnings on OS X Starting with Lion, Apple decided to deprecate the system openssl. We can start requiring users to install their own openssl once OS X doesn't ship with it anymore.
2011-08-10 19:22:41 +02:00
# OS X Lion started deprecating the system openssl. Let's just disable
Use dead_strip to reduce binary size on OS X This option seems to be supported all the way back to at least 10.4, so enabling it for OS X in general should be fine. If not, someone will yell. With no libs statically linked, that's a 3% win in binary size, with just libevent linked statically, this gives us an advantage of 5% in terms of binary size, and with libevent and openssl statically linked, we gain over 18% or over 500KB. Implements ticket 2915.
2012-01-20 23:13:35 +01:00
# all deprecation warnings on OS X. Also, to potentially make the binary
# a little smaller, let's enable dead_strip.
Ignore deprecation warnings on OS X Starting with Lion, Apple decided to deprecate the system openssl. We can start requiring users to install their own openssl once OS X doesn't ship with it anymore.
2011-08-10 19:22:41 +02:00
case "$host_os" in
darwin*)
CFLAGS="$CFLAGS -Wno-deprecated-declarations"
Use dead_strip to reduce binary size on OS X This option seems to be supported all the way back to at least 10.4, so enabling it for OS X in general should be fine. If not, someone will yell. With no libs statically linked, that's a 3% win in binary size, with just libevent linked statically, this gives us an advantage of 5% in terms of binary size, and with libevent and openssl statically linked, we gain over 18% or over 500KB. Implements ticket 2915.
2012-01-20 23:13:35 +01:00
LDFLAGS="$LDFLAGS -dead_strip" ;;
Ignore deprecation warnings on OS X Starting with Lion, Apple decided to deprecate the system openssl. We can start requiring users to install their own openssl once OS X doesn't ship with it anymore.
2011-08-10 19:22:41 +02:00
esac
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
TOR_WARNING_FLAGS=""
r13885@Kushana: nickm | 2007-08-01 09:10:55 -0700 Stop saying "cvs". Noted by Jens. svn:r11014
2007-08-01 18:43:27 +02:00
# Add some more warnings which we use in development but not in the
Fix "implicit declaration of ftello" warning: add AC_FUNC_FSEEKO in configure.in and always include orconfig.h before anything else svn:r2854
2004-11-14 18:21:32 +01:00
# released versions. (Some relevant gcc versions can't handle these.)
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
#
# Note that we have to do this near the end of the autoconf process, or
# else we may run into problems when these warnings hit on the testing C
# programs that autoconf wants to build.
Make advisory-warnings on by default. Add --enable-fatal-warnings to control -Werror. Closes ticket 19044.
2016-05-23 20:39:56 +02:00
if test "x$enable_gcc_warnings_advisory" != "xno"; then
r8987@totoro: nickm | 2006-10-10 19:46:08 -0400 Remove -Winline from standard list, add -Wswitch-enum. svn:r8687
2006-10-11 01:46:51 +02:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
case "$host" in
Add autoconf magic to support Bitrig Bitrig is an openbsd fork. Patch from dhill. Ticket 6982.
2012-10-04 16:56:33 +02:00
*-*-openbsd* | *-*-bitrig*)
Use -Wno-system-headers on openbsd to resolve 2nd case of bug1848
2010-08-24 19:43:53 +02:00
# Some OpenBSD versions (like 4.8) have -Wsystem-headers by default.
# That's fine, except that the headers don't pass -Wredundant-decls.
# Therefore, let's disable -Wsystem-headers when we're building
# with maximal warnings on OpenBSD.
CFLAGS="$CFLAGS -Wno-system-headers" ;;
esac
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
CFLAGS_NOWARNINGS="$CFLAGS"
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
# GCC4.3 users once report trouble with -Wstrict-overflow=5. GCC5 users
# have it work better.
# CFLAGS="$CFLAGS -Wstrict-overflow=1"
# This warning was added in gcc 4.3, but it appears to generate
# spurious warnings in gcc 4.4. I don't know if it works in 4.5.
#CFLAGS="$CFLAGS -Wlogical-op"
m4_foreach_w([warning_flag], [
-Waddress
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Waddress-of-array-temporary
-Waddress-of-temporary
-Wambiguous-macro
-Wanonymous-pack-parens
-Warc
-Warc-abi
-Warc-bridge-casts-disallowed-in-nonarc
-Warc-maybe-repeated-use-of-weak
-Warc-performSelector-leaks
-Warc-repeated-use-of-weak
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Warray-bounds
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Warray-bounds-pointer-arithmetic
-Wasm
-Wasm-operand-widths
-Watomic-properties
-Watomic-property-with-user-defined-accessor
-Wauto-import
-Wauto-storage-class
-Wauto-var-id
-Wavailability
-Wbackslash-newline-escape
-Wbad-array-new-length
-Wbind-to-temporary-copy
-Wbitfield-constant-conversion
-Wbool-conversion
-Wbool-conversions
-Wbuiltin-requires-header
-Wchar-align
-Wcompare-distinct-pointer-types
-Wcomplex-component-init
-Wconditional-type-mismatch
-Wconfig-macros
-Wconstant-conversion
-Wconstant-logical-operand
-Wconstexpr-not-const
-Wcustom-atomic-properties
-Wdangling-field
-Wdangling-initializer-list
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wdate-time
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wdelegating-ctor-cycles
-Wdeprecated-implementations
-Wdeprecated-register
-Wdirect-ivar-access
-Wdiscard-qual
-Wdistributed-object-modifiers
-Wdivision-by-zero
-Wdollar-in-identifier-extension
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wdouble-promotion
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wduplicate-decl-specifier
-Wduplicate-enum
-Wduplicate-method-arg
-Wduplicate-method-match
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wduplicated-cond
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wdynamic-class-memaccess
-Wembedded-directive
-Wempty-translation-unit
-Wenum-conversion
-Wexit-time-destructors
-Wexplicit-ownership-type
-Wextern-initializer
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wextra
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wextra-tokens
-Wflexible-array-extensions
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wfloat-conversion
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wformat-non-iso
-Wfour-char-constants
-Wgcc-compat
-Wglobal-constructors
-Wgnu-array-member-paren-init
-Wgnu-designator
-Wgnu-static-float-init
-Wheader-guard
-Wheader-hygiene
-Widiomatic-parentheses
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wignored-attributes
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wimplicit-atomic-properties
-Wimplicit-conversion-floating-point-to-bool
-Wimplicit-exception-spec-mismatch
-Wimplicit-fallthrough
-Wimplicit-fallthrough-per-function
-Wimplicit-retain-self
-Wimport-preprocessor-directive-pedantic
-Wincompatible-library-redeclaration
-Wincompatible-pointer-types-discards-qualifiers
-Wincomplete-implementation
-Wincomplete-module
-Wincomplete-umbrella
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Winit-self
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wint-conversions
-Wint-to-void-pointer-cast
-Winteger-overflow
-Winvalid-constexpr
-Winvalid-iboutlet
-Winvalid-noreturn
-Winvalid-pp-token
-Winvalid-source-encoding
-Winvalid-token-paste
-Wknr-promoted-parameter
-Wlarge-by-value-copy
-Wliteral-conversion
-Wliteral-range
-Wlocal-type-template-args
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wlogical-op
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wloop-analysis
-Wmain-return-type
-Wmalformed-warning-check
-Wmethod-signatures
-Wmicrosoft
-Wmicrosoft-exists
-Wmismatched-parameter-types
-Wmismatched-return-types
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wmissing-field-initializers
-Wmissing-format-attribute
-Wmissing-noreturn
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wmissing-selector-name
-Wmissing-sysroot
Add -Wmissing-variable-declarations, with attendant fixes This is a big-ish patch, but it's very straightforward. Under this clang warning, we're not actually allowed to have a global variable without a previous extern declaration for it. The cases where we violated this rule fall into three roughly equal groups: * Stuff that should have been static. * Stuff that was global but where the extern was local to some other C file. * Stuff that was only global when built for the unit tests, that needed a conditional extern in the headers. The first two were IMO genuine problems; the last is a wart of how we build tests.
2016-06-02 15:46:12 +02:00
-Wmissing-variable-declarations
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wmodule-conflict
-Wnested-anon-types
-Wnewline-eof
-Wnon-literal-null-conversion
-Wnon-pod-varargs
-Wnonportable-cfstrings
Switch -Wnormalized=id to -Wnormalized=nfkc Modified -Wnormalized flag to nfkc option in configure.ac to avoid source code identifier confusion. Fixes #24467 Signed-off-by: Fernando Fernandez Mancera <ffernandezmancera@gmail.com>
2017-11-29 17:04:41 +01:00
-Wnormalized=nfkc
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wnull-arithmetic
-Wnull-character
-Wnull-conversion
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wnull-dereference
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wout-of-line-declaration
-Wover-aligned
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Woverlength-strings
-Woverride-init
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Woverriding-method-mismatch
-Wpointer-type-mismatch
-Wpredefined-identifier-outside-function
-Wprotocol-property-synthesis-ambiguity
-Wreadonly-iboutlet-property
-Wreadonly-setter-attrs
-Wreceiver-expr
-Wreceiver-forward-class
-Wreceiver-is-weak
-Wreinterpret-base-class
-Wrequires-super-attribute
-Wreserved-user-defined-literal
-Wreturn-stack-address
-Wsection
-Wselector-type-mismatch
-Wsentinel
-Wserialized-diagnostics
Fix all -Wshadow warnings on Linux This is a partial fix for 18902.
2016-07-05 20:19:31 +02:00
-Wshadow
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wshift-count-negative
-Wshift-count-overflow
-Wshift-negative-value
-Wshift-overflow=2
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wshift-sign-overflow
-Wshorten-64-to-32
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wsizeof-array-argument
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wsource-uses-openmp
-Wstatic-float-init
-Wstatic-in-inline
-Wstatic-local-in-inline
TROVE-2017-001 : move -ftrapv back into --expensive-hardening.
2017-01-22 17:32:54 +01:00
-Wstrict-overflow=1
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wstring-compare
Add clang's -Wstring-conversion, and fix the one place it hits
2016-06-02 16:05:03 +02:00
-Wstring-conversion
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wstrlcpy-strlcat-size
-Wstrncat-size
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wsuggest-attribute=format
-Wsuggest-attribute=noreturn
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wsuper-class-method-mismatch
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wswitch-bool
-Wsync-nand
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wtautological-constant-out-of-range-compare
-Wtentative-definition-incomplete-type
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wtrampolines
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wtype-safety
-Wtypedef-redefinition
-Wtypename-missing
-Wundefined-inline
-Wundefined-internal
-Wundefined-reinterpret-cast
-Wunicode
-Wunicode-whitespace
-Wunknown-warning-option
-Wunnamed-type-template-args
-Wunneeded-member-function
-Wunsequenced
-Wunsupported-visibility
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wunused-but-set-parameter
-Wunused-but-set-variable
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wunused-command-line-argument
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wunused-const-variable=2
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wunused-exception-parameter
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wunused-local-typedefs
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wunused-member-function
-Wunused-sanitize-argument
-Wunused-volatile-lvalue
-Wuser-defined-literals
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
-Wvariadic-macros
Add all the clang-only warnings that do not trigger now
2016-06-01 22:20:17 +02:00
-Wvector-conversion
-Wvector-conversions
-Wvexing-parse
-Wvisibility
-Wvla-extension
-Wzero-length-array
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
], [ TOR_TRY_COMPILE_WITH_CFLAGS(warning_flag, [],
[TOR_WARNING_FLAGS="$TOR_WARNING_FLAGS warning_flag" CFLAGS="$CFLAGS warning_flag"], true)
])
version: Bump version to 0.4.8.7-dev Signed-off-by: David Goulet <dgoulet@torproject.org>
2023-09-25 16:44:06 +02:00
added check for compiler support of Wextra-semi update AX_CHECK_COMPILE_FLAG function call
2022-10-14 00:35:19 +02:00
AX_CHECK_COMPILE_FLAG([-Wextra-semi], [CFLAGS="$CFLAGS -Wextra-semi"], [], [-Werror])
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
TROVE-2017-001 : move -ftrapv back into --expensive-hardening.
2017-01-22 17:32:54 +01:00
dnl We should re-enable this in some later version. Clang doesn't
dnl mind, but it causes trouble with GCC.
dnl -Wstrict-overflow=2
Disable -Wthread-safety. See changes file; closes ticket 20110.
2016-09-08 15:37:40 +02:00
dnl These seem to require annotations that we don't currently use,
dnl and they give false positives in our pthreads wrappers. (Clang 4)
dnl -Wthread-safety
dnl -Wthread-safety-analysis
dnl -Wthread-safety-attributes
dnl -Wthread-safety-beta
dnl -Wthread-safety-precise
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
W_FLAGS="$W_FLAGS -W -Wfloat-equal -Wundef -Wpointer-arith"
W_FLAGS="$W_FLAGS -Wstrict-prototypes -Wmissing-prototypes -Wwrite-strings"
W_FLAGS="$W_FLAGS -Wredundant-decls -Wchar-subscripts -Wcomment -Wformat=2"
W_FLAGS="$W_FLAGS -Wwrite-strings"
W_FLAGS="$W_FLAGS -Wnested-externs -Wbad-function-cast -Wswitch-enum"
W_FLAGS="$W_FLAGS -Waggregate-return -Wpacked -Wunused"
W_FLAGS="$W_FLAGS -Wunused-parameter "
Set our autoconf-breaking options last, not before we check for others
2016-06-01 22:13:41 +02:00
# These interfere with building main() { return 0; }, which autoconf
# likes to use as its default program.
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
W_FLAGS="$W_FLAGS -Wold-style-definition -Wmissing-declarations"
TOR_WARNING_FLAGS="$TOR_WARNING_FLAGS $W_FLAGS"
CFLAGS="$CFLAGS $W_FLAGS"
Set our autoconf-breaking options last, not before we check for others
2016-06-01 22:13:41 +02:00
Use autoconf, not gcc version, to decide which warnings we have This gives more accurate results under Clang, which can only help us detect more warnings in more places. Fixes bug 19216; bugfix on 0.2.0.1-alpha
2016-06-01 19:38:36 +02:00
if test "$tor_cv_cflags__Wnull_dereference" = "yes"; then
AC_DEFINE([HAVE_CFLAG_WNULL_DEREFERENCE], 1, [True if we have -Wnull-dereference])
r18347@catbus: nickm | 2008-02-21 15:54:40 -0500 Make autoconf autodetect -Wshorten-64-to-32 when --enable-gcc-warnings is given to the configure script. svn:r13661
2008-02-21 21:54:45 +01:00
fi
Whoops -- this got lost in the merge.
2016-06-11 18:58:24 +02:00
if test "$tor_cv_cflags__Woverlength_strings" = "yes"; then
AC_DEFINE([HAVE_CFLAG_WOVERLENGTH_STRINGS], 1, [True if we have -Woverlength-strings])
fi
Don't disable an unsupported compiler warning Conditionalize the pragma that temporarily disables -Wunused-const-variable. Some versions of gcc don't support it. We need to do this because of an apparent bug in some libzstd headers. Fixes bug 26785; bugfix on 0.3.2.11.
2018-07-26 19:32:34 +02:00
if test "$tor_cv_cflags__warn_unused_const_variable_2" = "yes"; then
AC_DEFINE([HAVE_CFLAG_WUNUSED_CONST_VARIABLE], 1, [True if we have -Wunused-const-variable])
fi
r18347@catbus: nickm | 2008-02-21 15:54:40 -0500 Make autoconf autodetect -Wshorten-64-to-32 when --enable-gcc-warnings is given to the configure script. svn:r13661
2008-02-21 21:54:45 +01:00
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
CFLAGS="$CFLAGS_NOWARNINGS"
Make advisory-warnings on by default. Add --enable-fatal-warnings to control -Werror. Closes ticket 19044.
2016-05-23 20:39:56 +02:00
if test "x$enable_fatal_warnings" = "xyes"; then
# I'd like to use TOR_CHECK_CFLAGS here, but I can't, since the
# default autoconf programs are full of errors.
CFLAGS="$CFLAGS -Werror"
fi
Only use -Wlogical-op with GCC 4.6 and later It generates an apparently spurious warning with gcc 4.4 in debian; we haven't tested 4.5.
2013-10-31 16:57:50 +01:00
r8948@totoro: nickm | 2006-10-08 17:04:53 -0400 Add an --enable-gcc-warnings flag to configure.in svn:r8655
2006-10-08 23:05:11 +02:00
fi
Use a stricter set of warnings; make them all pass. svn:r2645
2004-11-02 04:02:17 +01:00
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
AC_SUBST(TOR_WARNING_FLAGS)
echo "$TOR_WARNING_FLAGS">warning_flags
TOR_TRY_COMPILE_WITH_CFLAGS([@warning_flags], [],
CFLAGS="$CFLAGS @warning_flags",
CFLAGS="$CFLAGS $TOR_WARNING_FLAGS")
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "$enable_coverage" = "yes" && test "$have_clang" = "no"; then
Coverage support: build with --enable-coverage to have tests run with gcov If you pass the --enable-coverage flag on the command line, we build our testing binaries with appropriate options eo enable coverage testing. We also build a "tor-cov" binary that has coverage enabled, for integration tests. On recent OSX versions, test coverage only works with clang, not gcc. So we warn about that. Also add a contrib/coverage script to actually run gcov with the appropriate options to generate useful .gcov files. (Thanks to automake, the .o files will not have the names that gcov expects to find.) Also, remove generated gcda and gcno files on clean.
2013-07-09 18:37:11 +02:00
case "$host_os" in
darwin*)
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
fix an error message in a configure warning string
2016-07-26 17:33:02 +02:00
AC_MSG_WARN([Tried to enable coverage on OSX without using the clang compiler. This might not work! If coverage fails, use CC=clang when configuring with --enable-coverage.])
Coverage support: build with --enable-coverage to have tests run with gcov If you pass the --enable-coverage flag on the command line, we build our testing binaries with appropriate options eo enable coverage testing. We also build a "tor-cov" binary that has coverage enabled, for integration tests. On recent OSX versions, test coverage only works with clang, not gcc. So we warn about that. Also add a contrib/coverage script to actually run gcov with the appropriate options to generate useful .gcov files. (Thanks to automake, the .o files will not have the names that gcov expects to find.) Also, remove generated gcda and gcno files on clean.
2013-07-09 18:37:11 +02:00
esac
fi
r18347@catbus: nickm | 2008-02-21 15:54:40 -0500 Make autoconf autodetect -Wshorten-64-to-32 when --enable-gcc-warnings is given to the configure script. svn:r13661
2008-02-21 21:54:45 +01:00
ext: build equix and hashx using automake This replaces the sketchy cmake invocation we had inside configure The libs are always built and always used in unit tests, but only included in libtor and tor when --enable-gpl is set. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-10 04:33:41 +01:00
# These HashX parameter definitions are needed in CPPFLAGS when compiling
# the equix and hashx ext modules, but elsewhere in tor we can use orconfig.h
m4_define([equix_hashx_size], [8])
[HASHX_SIZE=]equix_hashx_size
AC_SUBST([HASHX_SIZE])
AC_DEFINE([HASHX_SIZE], equix_hashx_size,
[Output size in bytes for the internal customization of HashX])
AC_DEFINE([HASHX_STATIC], [1], [We statically link with HashX])
AC_DEFINE([EQUIX_STATIC], [1], [We statically link with EquiX])
r12984@catbus: nickm | 2007-05-25 19:59:41 -0400 Do not alter CPPFLAGS until we are actually done with the rest of configure.in: this way, we notice if there are library paths that we are missing. svn:r10345
2007-05-26 01:59:47 +02:00
CPPFLAGS="$CPPFLAGS $TOR_CPPFLAGS_libevent $TOR_CPPFLAGS_openssl $TOR_CPPFLAGS_zlib"
Split long lines in configure.in and Makefile.am files Having very long single lines with lots and lots of things in them tends to make files hard to diff and hard to merge. Since our tools are one-line-at-a-time, we should try to construct lists that way too, within reason. This incidentally turned up a few headers in configure.in that we were for some reason searching for twice.
2010-11-11 20:22:48 +01:00
AC_CONFIG_FILES([
Tabify lists in configure.ac
2018-07-11 02:15:23 +02:00
Doxyfile
Makefile
contrib/operator-tools/tor.logrotate
src/config/torrc.sample
src/config/torrc.minimal
scripts/maint/checkOptionDocs.pl
Try @warning_flags to avoid bloating verbose make logs We now accumulate warning flags in a separate variable, "TOR_WARNING_FLAGS", and write it to a "warning_flags" file. Then we test whether the compiler will accept "@warning_flags": if so, we put "@warning_flags" in the CFLAGS; if not, we copy the contents of "$TOR_WARNING_FLAGS" into the CFLAGS. Closes ticket 28924.
2018-12-21 15:49:33 +01:00
warning_flags
Split long lines in configure.in and Makefile.am files Having very long single lines with lots and lots of things in them tends to make files hard to diff and hard to merge. Since our tools are one-line-at-a-time, we should try to construct lists that way too, within reason. This incidentally turned up a few headers in configure.in that we were for some reason searching for twice.
2010-11-11 20:22:48 +01:00
])
Merge remote branch 'origin/maint-0.2.2' Conflicts: configure.in
2010-11-11 19:59:18 +01:00
Quotes added on configure script's string comparison. Solves #17744 The syntax has been generalized too, so there are only double quotes. Signed-off-by: zerosion <zerosion@protonmail.ch>
2016-02-04 02:28:38 +01:00
if test "x$asciidoc" = "xtrue" && test "$ASCIIDOC" = "none"; then
doc: Move manpages into doc/man/ Closes #40044 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-07-14 15:02:03 +02:00
regular_mans="doc/man/tor doc/man/tor-gencert doc/man/tor-resolve doc/man/torify"
Config check for asciidoc and generated manpages. If USE_ASCIIDOC is enabled but asciidoc isn't present and manpages aren't already generated, it'll throw a warning during configure. Works with the current git / tarball split. Caveat: regular_mans are listed in the configure.ac See #6506
2013-05-14 19:18:28 +02:00
for file in $regular_mans ; do
double-quotes and changes file for 6506.
2013-05-24 20:43:57 +02:00
if ! [[ -f "$srcdir/$file.1.in" ]] || ! [[ -f "$srcdir/$file.html.in" ]] ; then
Config check for asciidoc and generated manpages. If USE_ASCIIDOC is enabled but asciidoc isn't present and manpages aren't already generated, it'll throw a warning during configure. Works with the current git / tarball split. Caveat: regular_mans are listed in the configure.ac See #6506
2013-05-14 19:18:28 +02:00
echo "==================================";
echo;
Improving error message.
2014-09-28 17:09:25 +02:00
echo "Building Tor has failed since manpages cannot be built.";
echo;
echo "You need asciidoc installed to be able to build the manpages.";
Config check for asciidoc and generated manpages. If USE_ASCIIDOC is enabled but asciidoc isn't present and manpages aren't already generated, it'll throw a warning during configure. Works with the current git / tarball split. Caveat: regular_mans are listed in the configure.ac See #6506
2013-05-14 19:18:28 +02:00
echo "To build without manpages, use the --disable-asciidoc argument";
echo "when calling configure.";
echo;
echo "==================================";
exit 1;
fi
done
fi
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
if test "$fragile_hardening" = "yes"; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
AC_MSG_WARN([
============
Warning! Building Tor with --enable-fragile-hardening (also known as
--enable-expensive-hardening) makes some kinds of attacks harder, but makes
other kinds of attacks easier. A Tor instance build with this option will be
somewhat less vulnerable to remote code execution, arithmetic overflow, or
out-of-bounds read/writes... but at the cost of becoming more vulnerable to
denial of service attacks. For more information, see
Update remaining trac.tpo wiki urls
2020-09-14 17:02:38 +02:00
https://gitlab.torproject.org/tpo/core/team/-/wikis/TorFragileHardening
Rename --enable-expensive-hardening configure option It is renamed to --enable-fragile-hardening. TROVE-2017-001 was triggerable only through the expensive hardening which is making the tor daemon abort when the issue is detected. Thus, it makes tor more at risk of remote crashes but safer against RCE or heartbleed bug category. Fixes #21290. Signed-off-by: David Goulet <dgoulet@torproject.org>
2017-02-01 16:08:40 +01:00
============
])
fi
r12061@catbus: nickm | 2007-03-04 14:41:10 -0500 Run configure.in through autoupdate svn:r9729
2007-03-04 20:49:04 +01:00
AC_OUTPUT
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
if test "$openssl_ver_mismatch" = "yes"; then
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
tor_incr_n_warnings
configure.ac: Give a warning if openssl headers don't match library. We don't look at the patchlevel, since that tends not to have any API changes, and sometimes gets out of sync when distributors are careless. We only give the warning when the test program compiles but gives a nonzero exit status: sadly, autoconf doesn't give us an easy way to distinguish these. Fixes #40138
2020-09-23 15:52:43 +02:00
AC_MSG_WARN([
============
Warning! The version OpenSSL headers we get from compiling with
"${TOR_CPPFLAGS_OPENSSL:-(no extra options)}"
do not match version of the OpenSSL library we get when linking with
"$TOR_LDFLAGS_OPENSSL $TOR_OPENSSL_LIBS".
This might cause compilation to fail. Try using --with-openssl-dir to specify
the exact OpenSSL path you want.
============
])
fi
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
#
# Mini-report on what will be built.
#
PPRINT_INIT
PPRINT_SET_INDENT(1)
PPRINT_SET_TS(65)
AS_ECHO
AS_ECHO("Tor Version: ${PPRINT_COLOR_BLDBLU}Tor $PACKAGE_VERSION $PPRINT_COLOR_RST")
AS_ECHO
PPRINT_SUBTITLE([Build Features])
PPRINT_PROP_STRING([Compiler], [$CC])
PPRINT_PROP_STRING([Host OS], [$host_os])
configure: Add --enable-gpl option This change on its own doesn't use the option for anything, but it includes support for configure and a message in 'tor --version' Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-08 18:08:23 +01:00
PPRINT_PROP_STRING([License Option], [$license_option])
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
AS_ECHO
test "x$enable_fatal_warnings" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([Warnings are fatal (--enable-fatal-warnings)], $value)
test "x$enable_android" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([Android support (--enable-android)], $value)
AS_ECHO
PPRINT_SUBTITLE([Static Build])
test "x$enable_static_tor" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([tor (--enable-static-tor)], $value)
if test "x$enable_static_libevent" = "xyes"; then
PPRINT_PROP_STRING([libevent], [$TOR_LIBDIR_libevent])
else
PPRINT_PROP_BOOL([libevent (--enable-static-libevent)], $value)
fi
if test "x$enable_static_openssl" = "xyes"; then
PPRINT_PROP_STRING([libssl], [$TOR_LIBDIR_openssl])
else
PPRINT_PROP_BOOL([libssl (--enable-static-openssl)], $value)
fi
if test "x$enable_static_zlib" = "xyes"; then
PPRINT_PROP_STRING([zlib1g], [$TOR_LIBDIR_zlib])
else
PPRINT_PROP_BOOL([zlib1g (--enable-static-zlib)], $value)
fi
AS_ECHO
PPRINT_SUBTITLE([Optional Libraries])
test "x$enable_nss" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([libnss (--enable-nss)], $value)
test "x$enable_seccomp" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([libseccomp (--disable-seccomp)], $value)
test "x$enable_libscrypt" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([libscrypt (--disable-libscrypt)], $value)
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
test "x$enable_systemd" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([Systemd support (--enable-systemd)], $value)
Bugfix for 40236 Make configure script minireport accurate automatic detection and usage of liblzma and libzstd
2021-01-18 20:37:26 +01:00
test "x$have_lzma" = "xyes" && value=1 || value=0
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
PPRINT_PROP_BOOL([liblzma (--enable-lzma)], $value)
Bugfix for 40236 Make configure script minireport accurate automatic detection and usage of liblzma and libzstd
2021-01-18 20:37:26 +01:00
test "x$have_zstd" = "xyes" && value=1 || value=0
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
PPRINT_PROP_BOOL([libzstd (--enable-zstd)], $value)
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
AS_ECHO
PPRINT_SUBTITLE([Hardening])
test "x$enable_gcc_hardening" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([Compiler Hardening (--disable-gcc-hardening)], $value)
test "x$enable_linker_hardening" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([Linker Hardening (--disable-linker-hardening)], $value)
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
test "x$fragile_hardening" = "xyes" && value=1 || value=0
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
PPRINT_PROP_BOOL([Fragile Hardening (--enable-fragile-hardening, dev only)], $value)
AS_ECHO
PPRINT_SUBTITLE([Modules])
hs_pow: Make proof-of-work support optional in configure This adds a new "pow" module for the user-visible proof of work support in ./configure, and this disables src/feature/hs/hs_pow at compile-time. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-09 00:44:55 +01:00
# Modules have documentation hints indicating how they can be enabled
# or disabled, and those hints can select which version of our message
# to show based on variables at configure-time.
#
# Each "module_option_hints(<name>)" macro, if it exists, must
# visit exactly one HINT_* macro using shell conditionals.
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
m4_foreach_w([mname], MODULES,
[
configure: Fix enabled module variable expansion Fixes #33646 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-03-31 14:21:12 +02:00
AM_COND_IF(m4_join([], [BUILD_MODULE_], m4_toupper([]mname[])), value=1, value=0)
hs_pow: Make proof-of-work support optional in configure This adds a new "pow" module for the user-visible proof of work support in ./configure, and this disables src/feature/hs/hs_pow at compile-time. Signed-off-by: Micah Elizabeth Scott <beth@torproject.org>
2023-03-09 00:44:55 +01:00
m4_pushdef([HINT_OPT], [PPRINT_PROP_BOOL](mname ($1), [[$value]]))
m4_pushdef([HINT_NONE], [PPRINT_PROP_BOOL](mname, [[$value]]))
m4_ifdef([module_option_hints](mname),
[m4_indir([module_option_hints](mname))],
[HINT_NONE])
m4_popdef([HINT_OPT], [HINT_NONE])
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
]
)
configure: Add --disable-module-relay to summary Closes #32266 Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-10-24 15:08:36 +02:00
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
AS_ECHO
PPRINT_SUBTITLE([Documentation])
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
test "x$enable_asciidoc" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([AsciiDoc (--disable-asciidoc)], $value)
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
test "x$enable_manpage" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([Man Pages (--disable-manpage)], $value)
test "x$enable_html_manual" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([HTML Manual (--disable-html-manual)], $value)
AS_ECHO
PPRINT_SUBTITLE([Tests])
test "x$enable_unittests" != "xno" && value=1 || value=0
PPRINT_PROP_BOOL([Unit tests (--disable-unittests)], $value)
test "x$enable_asserts_in_tests" = "xno" && value=1 || value=0
configure: Fix and improve summary pretty printing Fixes #32230 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-30 17:18:34 +01:00
PPRINT_PROP_BOOL([assert()s enabled (--enable-asserts-in-tests, dev only)], $value)
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
test "x$enable_coverage" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([Code Coverage (--enable-coverage)], $value)
test "x$enable_libfuzzer" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([libFuzzer support (--enable-libfuzzer)], $value)
test "x$enable_oss_fuzz" = "xyes" && value=1 || value=0
PPRINT_PROP_BOOL([OSS-Fuzz support (--enable-oss-fuzz)], $value)
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
AS_ECHO
trace: Comments and configure fix Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-15 16:55:07 +01:00
PPRINT_SUBTITLE([Tracing (--enable-tracing-instrumentation-<type>)])
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
test "x$enable_tracing_instrumentation_log_debug" = "xyes" && value=1 || value=0
trace: Comments and configure fix Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-15 16:55:07 +01:00
PPRINT_PROP_BOOL([Tracepoints to log_debug() (log-debug)], $value)
trace: Better structure lib/trace and configure options In the next commits, we'll add more tracing options for instrumentation and specific tracer. This rename follows a more meaningful naming standard. It also adds a catch all "HAVE_TRACING" define that indicate in the code that we have tracing enabled. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-09 18:17:49 +01:00
trace: Add USDT probes generation support This commit adds both configure options and probe generation for tracepoints. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-14 20:58:09 +01:00
test "x$enable_tracing_instrumentation_usdt" = "xyes" && value=1 || value=0
trace: Comments and configure fix Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-15 16:55:07 +01:00
PPRINT_PROP_BOOL([USDT Instrumentation (usdt)], $value)
trace: Add USDT probes generation support This commit adds both configure options and probe generation for tracepoints. Part of #32910 Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-14 20:58:09 +01:00
trace: Add LTTng-UST interface support No probes at this point. They are per subsystem and thus in later commits. Part of #32910
2020-02-11 17:26:04 +01:00
test "x$enable_tracing_instrumentation_lttng" = "xyes" && value=1 || value=0
trace: Comments and configure fix Signed-off-by: David Goulet <dgoulet@torproject.org>
2020-01-15 16:55:07 +01:00
PPRINT_PROP_BOOL([LTTng Instrumentation (lttng)], $value)
trace: Add LTTng-UST interface support No probes at this point. They are per subsystem and thus in later commits. Part of #32910
2020-02-11 17:26:04 +01:00
Print summary at the end of the configure script Signed-off-by: David Goulet <dgoulet@torproject.org>
2019-08-08 16:08:02 +02:00
AS_ECHO
PPRINT_SUBTITLE([Install Directories])
report_mandir="`eval eval echo $mandir`"
PPRINT_PROP_STRING([Binaries], [$BINDIR])
PPRINT_PROP_STRING([Configuration], [$CONFDIR])
PPRINT_PROP_STRING([Man Pages], [$report_mandir])
AS_ECHO
AS_ECHO(["Configure Line: ./configure $configure_flags"])
configure.ac: Summarize how many times we warned. It's way too easy to have a warning in the middle of the configure process that the user never notices.
2020-09-23 16:11:36 +02:00
if test "$tor_ac_n_warnings" != "0"; then
AS_ECHO
PPRINT_WARN([
Encountered $tor_ac_n_warnings warning(s). See messages above for more info.
])
fi
Reference in New Issue Copy Permalink