Some gcc versions do explode if the order of the linker flags are not correct.
One issue was statically building OpenSSL which would require that "-lssl
-lcrypto" be put _before_ the "-lpthread -ldl" flags.

I have not such problem with GCC 10 but does with GCC 9.

Closes #33624

Signed-off-by: David Goulet <dgoulet@torproject.org>

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.

Closes 18426 and 18459.

Because in 95 years, we or our successors will surely care about
enforcing the BSD license terms on this code.  Right?

Fix for 9948; patch from Benedikt Gollatz.

This is the preferred filename to use with Autoconf 2.50 and later.

On some platforms, the linker is perfectly happy to produce binaries
that won't run if you give it the wrong set of flags.  So when not
cross-compiling, try to link-and-run a little test program, rather
than just linking it.

Possible fix for 6173.

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.

This is a matter of making gcc and friends squirm more loudly when
they get an option they don't like (-pedantic) and making clang shut
up with it gets an option it tolerates but doesnt know
(-Qunknown-argument).

Is there no better way?

Also, make the check for whether they're on by default work; there's
no need to mess around with this "$enableval" silliness.

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).

These options only work when using --with-(openssl|libevent)-dir to
explicitly pick a libevent or openssl location.

Tor doesn't use SVN anymore, making $Revision$, $Id$ and $Date$
meaningless. Remove them without replacement.

 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

 Fix typo in acinclude.m4


svn:r16624

svn:r13412

 Oops; left a dquote out of acinclude.m4


svn:r13410

 Write a new autoconf macro to test whether a function is declared. It is suboptimal and possibly buggy in some way, but it seems to work for me.  use it to test for a declaration of malloc_good_size, so we can workaround operating systems (like older OSX) that have the function in their libc but do not deign to declare it in their headers.  Should resolve bug 587.


svn:r13339

 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

 Fix some autoconf bugs, and prevent more from occurring.  Partial backport candidate.


svn:r12767

 Test for ~/.svk/local, not ~/.svk.  Eschew use of test -a and test -o.  Bulletproof more against odd shells.


svn:r10439

 When telling the admin that they will need to install a main package, tell them that they will probably need the dev package too.


svn:r10366

 Oops. we actually _do_ need the extended path; we just do not want to persist it.


svn:r10349

 Ah, maybe _this_ is the patch that will make IRIX builds happy again.


svn:r10348

 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

 Fix bug in 10072 that prevented correct detection of include and lib subdirectories


svn:r10078

 Minimize the libraries that we link things against: there is no reason to link tor-resolve against zlib, openssl, or libevent, for example.


svn:r10072