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compilers automatically, and so the AC_PROG_CC_C99 macro is
obsolete.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detect recoverably at autoconf-time whether a macro is obsolete or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.
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>
From c618c4f279, 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>
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>
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
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)
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>
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.
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
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
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>
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>
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.
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.
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.
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.
