The desired outcome of this change is that we'll set
-Wl,--version-script based on linker kind and not on the output of
$LINKER -v.
This is a cheap way to address a simple problem that has a complicated
ideal solution. The underlying issue is that in some situations, when
targeting Android, a macOS system ld is interrogated to determine if
a cross-compiling linker "is GNU ld" and a particular linker feature
is set in that situation. The macOS system ld doesn't pass the "is
GNU ld" test, and the linker feature isn't set; that causes link
failures, even though the actual linker has nothing to do with the
system ld.
The ideal solution is to test for linker capabilities dynamically. We
do a lot of that in old-configure.in, and we don't do any of that in
toolchain.configure. Rather than start testing in
toolchain.configure, we hard-code: a cheap solution to the immediate
problem.
MinGW suffers somewhat from the opposite problem: the linker "is GNU
ld" (compatible), but the linker checks don't happen at all. We hard-code
for MinGW based on the C compiler instead.
Differential Revision: https://phabricator.services.mozilla.com/D8471
--HG--
extra : moz-landing-system : lando
The desired outcome of this change is that we'll set
`-Wl,--version-script` based on linker kind and not on the output of
`$LINKER -v`.
This is a cheap way to address a simple problem that has a complicated
ideal solution. The underlying issue is that in some situations, when
targeting Android, a macOS system `ld` is interrogated to determine if
a cross-compiling linker "is GNU ld" and a particular linker feature
is set in that situation. The macOS system `ld` doesn't pass the "is
GNU ld" test, and the linker feature isn't set; that causes link
failures, even though the actual linker has nothing to do with the
system `ld`.
The ideal solution is to test for linker capabilities dynamically. We
do a lot of that in old-configure.in, and we don't do any of that in
toolchain.configure. Rather than start testing in
toolchain.configure, we hard-code: a cheap solution to the immediate
problem.
Differential Revision: https://phabricator.services.mozilla.com/D8471
--HG--
extra : moz-landing-system : lando
The GNU_AS check in old-configure depended on running with the value
of $AS before it gets reset to just be the C compiler, which breaks when
we move setting AS into moz.configure.
This patch moves the GNU_AS check to toolchain.configure and changes it
so that it works when the assembler is the C compiler. We do have to
fix things slightly for clang, because the previous check was
succeeding, but not because of clang: it was detecting the presence of
"GNU" in the output for GNU ld/gold and a message about the GNU GPL.
This is a fairly straightforward port of the AS tool checks from old-configure
to toolchain.configure. AS is a little quirky in that we currently do a
normal-looking check for it, but then override that value to be the C compiler
for non-Windows builds, and ml[64]/armasm64 for Windows builds.
After migrating those checks, the only things left in the MOZ_DEFAULT_COMPILER
macro in compiler-opts.m4 were some unused bits, so I removed them:
* Setting of CPP/CXXCPP, which are set in toolchain.configure now
* Setting HOST_LDFLAGS to empty, which doesn't seem particularly useful.
There was also a quirky old test that the assembler was ml[64] when js-ctypes
is enabled that I removed, I don't think it provides any value since this
patch will ensure that we're using the right assembler for Windows builds.
BUILD_TOOLS was only ever used for things that another variable provides
equally well. Removing BUILD_TOOLS means that we can remove win_srcdir
and WIN_TOP_SRC as well.
We were already always enabling it on builds with frame pointers on
Android. We now extend this to builds without frame pointers, helping
with client-side stacktracing for crashes.
It's also the default on many platforms, such as x86-64, whether frame
pointer are enabled or not and on x86 with GCC. Somehow, it's not the
case for x86 with clang, so this makes clang builds the same as GCC
builds in that regard.
On ARM, we were always omitting frame pointers because of a GCC bug, but
we've not been using GCC for Android builds for a long time (and the GCC
bug was actually fixed in 4.7 and backported to 4.6, making it fixed in
all versions of GCC we support).
Differential Revision: https://phabricator.services.mozilla.com/D6110
--HG--
extra : moz-landing-system : lando
Last attempt, a few years ago, blatantly failed because nautilus (the
GNOME file manager) can't start PIE executables, which look like shared
libraries, and that it thus considers not being executables.
Downstreams don't actually have the problem, because users won't be
launching Firefox from a file manager, but for mozilla.org builds, it is
a problem because users would download, then extract, and then likely
try to run the Firefox executable from a file manager.
So for mozilla.org builds, we still need to find a way around the
nautilus problem.
A .desktop file could be a solution, but .desktop files have not
actually been designed for this use case, which leads to:
- having to use an awful one-liner shell wrapper to derive the path
to the executable from that of the .desktop file,
- not even being able to associate an icon,
- the .desktop file not being copiable to a location where .desktop
files would normally go, because it would then fail to find the
executable.
Another possibility is to go back to using a shell wrapper, but that's
not entirely appealing.
What we chose here is similar, where we have a small `firefox` wrapper
that launches the real `firefox-bin` (which is still leftover from those
old times where we had a shell wrapper, for reasons).
The small `firefox` wrapper is a minimalist C executable that just
finds the path to the `firefox-bin` executable and executes it with the
same args it was called with. The wrapper is only enabled when the
MOZ_NO_PIE_COMPAT environment variable is set, which we only take into
account on Linux. The variable is only really meant to be used for
mozilla.org builds, for the nautilus problem. Downstreams will just pick
the default, which is changed to build PIE.
On other platforms, PIE was already enabled by default, so we just
remove the --enable-pie configure flag.
Differential Revision: https://phabricator.services.mozilla.com/D5109
When building executables as PIE, and because we use -Bsymbolic, which
symbols are exported from an executable varies based on the libraries
it's directly linked against, to fulfil their symbol needs. So when a
library depends on e.g. ASAN runtime symbols, and the linker finds that,
it will keep those ASAN symbols in the executable for the library. And
drop the other, unused symbols.
But when the executable then dlopen()s a library (e.g. shlibsign loading
libfreebl) that uses another set of ASAN symbols, including symbols that
none of the direct dependencies of the executable need, dlopen() fails
because of the missing symbols.
It's not currently an apparent problem because we don't enable PIE, and
we build Gecko executables with -rdynamic already (for mozjemalloc). But
we don't build non-Gecko executables this way (like shlibsign).
Differential Revision: https://phabricator.services.mozilla.com/D5108
At least, when that combination doesn't work. This requires passing the
LTO flags to the compiler invocations for the test, which in turn
requires actually calling the linker directly on Windows, instead of
directly through the compiler, as when actually building.
Differential Revision: https://phabricator.services.mozilla.com/D4739
--HG--
extra : moz-landing-system : lando
Back before expandlibs was removed in bug 1429875, not using list files
was supported, but nowadays, it just leads to errors like the following:
*** No rule to make target 'libfake_so.list', needed by 'libfake.so'. Stop.
We might as well error out during configure, rather than much later
during the build.
Differential Revision: https://phabricator.services.mozilla.com/D4738
Back before expandlibs was removed in bug 1429875, not using list files
was supported, but nowadays, it just leads to errors like the following:
*** No rule to make target 'libfake_so.list', needed by 'libfake.so'. Stop.
We might as well error out during configure, rather than much later
during the build.
Differential Revision: https://phabricator.services.mozilla.com/D4738
yasm doesn't support aarch64, and trying to use GNU as with an MSVC
build seems like sadness waiting to happen. Instead, we'll generate our
own assembly file that armasm64 will accept.
The media/libpng/moz.build file overrides the C standard used via
-std=c89, per bug 1371266, which conflicts with the use of the
arm_neon.h header: compilation fails on the inline keyword, which didn't
exist in C89. We thus "bump" to the GNU89 standard, which is C89+GNU
extensions, including inline.
--HG--
extra : rebase_source : fe93a13e3bef8888e1874d2e94a6d8ef396aaf83
It's the *compile*SdkVersion that needs to match the installed Android SDK plat-
form in order to be able to build an app, whereas the *target*SdkVersion is
merely a compatibility flag.
Since the received wisdom is that targetSdkVersion should be <= compileSdk-
Version and Android Studio is also showing a warning to that effect if you
modify the build.gradle of a small sample app accordingly, I've also added a
corresponding configure check of our own to enforce this.
MozReview-Commit-ID: F2RZemChFrm
--HG--
extra : rebase_source : cf4f6256baa4446d673b94d97f9497f93d7917ff
This forces the linker ot use libatomic with clang on x86 (not only Android)
and forwards the intent to use libatomic to the shell's moz.build.
--HG--
extra : rebase_source : 0c803a3e11efcad3f17a462c2d38e85ec6cb556a
The moz.build Java JAR definitions are, of course, broken, but they
will be removed soon enough.
MozReview-Commit-ID: KIxqLDwd9I7
--HG--
extra : rebase_source : 8312b3f125793f73d3e835d1c0a5c7cabd4ebc0c
I choose to clean a bunch of ANDROID_* moz.build cruft here, too,
since it's just passing dependencies between moz.build and
Makefile.in. The replacement for all of this is to just use
GENERATED_FILES in moz.build, but it'll still take some work to get to
that. (Why does this stuff exist? GENERATED_FILES didn't exist and
was resisted when I built this stuff.)
MozReview-Commit-ID: D3GJqJNL0Ih
--HG--
extra : rebase_source : 07351f9d3702cfc42c58bd317885d07882c45c3a
NDK headers are grouped into a "sysroot" directory, which doesn't
contain architecture-specific bits, and a "system" directory, which
contains only the architecture-specific bits. Previously, both
directories are the same, under platforms/android-*/arch-*/. However,
with unified headers in NDK r16, the two are different, so we need to
support that in the Android build scripts.
Unified headers also rely on the build system defining the
__ANDROID_API__ macro, so we add support for that as well.
MozReview-Commit-ID: 9zBNQC3BRFl
--HG--
extra : rebase_source : 9ded282343df64d9cc4abcf7d7c6b03ac3423ff0
NDK headers are grouped into a "sysroot" directory, which doesn't
contain architecture-specific bits, and a "system" directory, which
contains only the architecture-specific bits. Previously, both
directories are the same, under platforms/android-*/arch-*/. However,
with unified headers in NDK r16, the two are different, so we need to
support that in the Android build scripts.
Unified headers also rely on the build system defining the
__ANDROID_API__ macro, so we add support for that as well.
MozReview-Commit-ID: 9zBNQC3BRFl
--HG--
extra : rebase_source : 36c9d4d5625fecbbe2485668811f85f9d94372d5
Will also address Bug 1377553 and part of Bug 1419607
MozReview-Commit-ID: AUCqBxEGpAl
--HG--
extra : rebase_source : 5547e2c8fbf4e2e87182b8720d8352c131e4ec65
We no longer support Android/armv6 and we requires NEON for Android/arm, so
we can remove armv6 support for Android.
MozReview-Commit-ID: Hh17BTyE0wR
--HG--
extra : rebase_source : 57e043ecb1bb57a026c0b656b82768b899ddae78
Bug 1256642 introduced magic at the emitter level to determine whether a
binary contains C++ sources and should be linked with the C compiler or
the C++ compiler.
Unfortunately, the Binary() moz.build template always adds C++ OS
libraries on Android (through STLPORT_LIBS), and C++ libraries on Linux
(stdc++compat).
The latter only ends up forcing every Binary() to be linked with the C++
linker, which is unfortunate, but doesn't cause much problems. The
former, however, involving OS libraries, the magic from bug 1256642
doesn't kick in, so we end up trying to link C++ OS libraries with the C
linker. Which ends up failing, because the libraries in STLPORT_LIBS
require -lm, which, while it's added by the C++ compiler when linking,
is not when the linkage is driven by the C compiler.
Because the fallible library, linked to all GeckoBinary()s is a C++
library, we still ended up linking with the C++ compiler on Android, so
this wasn't actually causing any problem... until I tried to remove that
fallible library in bug 1423803.
Anyways, the core problem is that moz.build evaluation is happening too
early to know whether any C++ sources are being linked together, so
there is no way the Binary() template can do the right thing. So this
change moves the logic to the emitter.
This also changes the type of STLPORT_LIBS to a list.
--HG--
extra : rebase_source : a70ddf7a132f94dc10e7e1db94ae80fb8d7a269f
Will also address Bug 1377553 and part of Bug 1419607
MozReview-Commit-ID: AUCqBxEGpAl
--HG--
extra : rebase_source : f7582d7089f0f4582a02aeaef090dc0701df994d
According to https://developer.android.com/ndk/guides/abis.html,
androideabi-v7a must support vfpv3-d16. So we should use it for fpu flag.
MozReview-Commit-ID: 3rhmRTekmwD
--HG--
extra : rebase_source : c5ffa22d8712fc7b8006cc340175a9586d77f49b
Stylo's bindgen is configured partially through a .toml.in file that
substitutes the value of a configure variable (BINDGEN_CFLAGS) into a
TOML list. We can debate whether this is a good thing to do some other
time; the reality is that the current moz.configure code that provides
the set_config for BINDGEN_CFLAGS needs to perform all the quoting
itself.
We want, however, to define the substituted variable in old-configure.in
land (some of the values that will go into BINDGEN_CFLAGS are only
defined in old-configure.in, and are not trivially ported to
moz.configure), which means that we need to have quoting logic in
m4/Python when we generate config.status. This patch adds an
appropriate macro for doing so.
The various AC_SUBST macros generate AC_SUBST_*FOO macros for holding the
values to substitute. The macros also cross-check the AC_SUBST_* macros
generated by other variants to make sure that you don't try to do
something like AC_SUBST(FOO) and AC_SUBST_SET(FOO). However, the check
in AC_SUBST_SET for AC_SUBST_LIST duplicate is missing an underscore:
the AC_SUBST_LIST macro generates another macro starting with
AC_SUBST_LIST_, but the AC_SUBST_SET macro checks for the prefix
AC_SUBST_LIST, which is missing the trailing underscore.
As we're going to be adding yet another AC_SUBST_* macro variant, and
therefore adding more checks to all existing macros, let's clean this up
before we start.
NDK r15+ clang changed the code generation strategy for the __atomic_*
intrinics such that using them with 64-bit types now requires linking
with libatomic. Our current configure tests for libatomic doesn't catch
this, because the std::atomic implementation is such that it doesn't
require an external library, even for 64-bit types, whereas the
__atomic_* intrinsics do. The safest thing to do is to force this
configure check to always return true when we are compiling for
x86/Android with clang.
The NDK clang needs to be informed about the existence of a GCC
toolchain, so important programs like the linker can be located. With
this change, we're starting to use command-line options that are
incompatible with GCC, so we also add a check to inform the user about
the non-support for this configuration.
While we're here, provide a reference to unique_list as defined in
moz.configure when executing config.data to avoid its redefinition
in m4.
MozReview-Commit-ID: AI6XhoYR0Ye
Although there is emulator command in android_sdk_root/tools even if Android
Tools 26, it doesn't work well because emulator64-arm is missing in same
directory.
Valild emulator command moves from android_sdk_root/tools to
android_sdk_root/emulator from Android Tools 26 (updated by sdkmanager).
MozReview-Commit-ID: 4IGHK3YcZXD
--HG--
extra : rebase_source : cc0c67f24c32a55ac40e65a5a036d5072dff6ee4
extra : histedit_source : 9718746e56dd9ec16a540533acfe909c319c234b
To build sytlo, we have to set compiler flags via BINDGEN_CFLAGS. Since we
pass stlport flags to clang, I would like to move STLPORT_CPPFLAGS to
moz.configure.
MozReview-Commit-ID: 26jvUqUvwTY
--HG--
extra : rebase_source : 5568627368fbf2dce02904918e50a241713d0a85
libstdc++ support is broken after moving to moz.configure. No one uses this option and NDK will remove GCC, so we should remove this and --with-android-cxx-stl option.
MozReview-Commit-ID: 3mqyHoRCE00
--HG--
extra : rebase_source : 35aa911a69e159e67f624ab5ab9aea8af4c5342f
cpu_arch_dir isn't absolute path, we cannot always detect thumb library for c++ library. Our build tool still uses NDK r11c, so we should check thumb library correctly.
MozReview-Commit-ID: 912ZDNeUcKc
--HG--
extra : rebase_source : 8b69348ef3280f32b7a228d66462070413a25e73
Some gaia-related code was also found and removed as part of the
cleanup.
MozReview-Commit-ID: DEjVSljzzu1
--HG--
extra : rebase_source : 58c4c34df44a258d90029853f29ea01338bd142b
This patch does two things:
- add a Gradle-only ANDROID_COMPILE_SDK_VERSION substitution;
- uses it while uniformizing all of the Gradle Android SDK version
configurations.
The approach is fairly standard (and we were using it already); see,
for example
https://medium.com/@ali.muzaffar/gradle-configure-variables-for-all-android-project-modules-in-one-place-5a6e56cd384e
This will make bumping the Gradle configuration versions forward
easier.
MozReview-Commit-ID: 1j5siCvR5qt
--HG--
extra : rebase_source : 07afb00de0e4a72af4026eb19ff4f2530c119336
This is just a small piece in the larger project of minimizing the set
of global AC_DEFINEs. None of these are relevant to C/C++ code;
they're mobile/android specific, so we should get them out of
mozilla-config.h.
In addition, this replaces hard-coded values with values from the
build system throughout mobile/android.
MozReview-Commit-ID: 9IJlsm38LFK
--HG--
extra : rebase_source : 0e55c0967cbb7f0df0d297eb330fc615e5759346
The -fsanitize=integer analysis from UBSan can be helpful to detect signed and unsigned integer overflows in the codebase. Unfortunately, those occur very frequently, making it impossible to test anything with it without the use of a huge blacklist. This patch includes a blacklist that is broad enough to silence everything that would drain performance too much. But even with this blacklist, neither tests nor fuzzing is "clean". We can however in the future combine this with static analysis to limit ourselves to interesting places to look at, or improve the dynamic analysis to omit typical benign overflows.
It also adds another attribute that can be used on functions. It is not used right now because it was initially easier to add things to the compile-time blacklist to get started.
Finally, it includes a runtime suppression list and patches various parts in the test harnesses to support that. It is currently empty and it should not be used on frequent overflows because it is expensive. However, it has the advantage that it can be used to differentiate between signed and unsigned overflows while the compile-time blacklist cannot do that. So it can be used to e.g. silence unsigned integer overflows on a file or function while still reporting signed issues. We can also use this suppression list for any other UBSan related suppressions, should we ever want to use other features from that sanitizer.
MozReview-Commit-ID: C5ofhfJdpCS
--HG--
extra : rebase_source : 952043a441b41b2f58ec4abc51ac15fa71fc142f
For parts of configuring Stylo, we need information about the library
extensions on all of our platforms, and this change is a reasonable way
to get at that information without duplicating it in two places. Plus
moving more things to moz.configure is more better.
This intentionally allows to set MOZ_INSTALL_TRACKING without
reference to the milestone being release or beta. That is, we
separate the default value (which depends on release or beta) from the
value specified, making life easier for developers.
MozReview-Commit-ID: 3vPF7KO7fEX
--HG--
extra : rebase_source : 8d5764104b5322a32e4a048bfd3222f62fed73bb
Also remove bits of a comment, now that we support only macOS 10.9+, most of the
comment isn't relevant.
While PIE is enabled by default on macOS, this isn't true of clang on Linux.
--enable-pie can now be used with clang on Linux.
r=froydnj
MozReview-Commit-ID: rc6zJiWzLo
--HG--
extra : rebase_source : 3745175e106ea8c6be9271d8135d43ba359434c7
It looks like Google decided to split these jars out a bit, so we need to piece
them all back together.
We could probably just query the sdk version instead, but I'm not 100% sure
know when this setup changed - moreover we don't know when (if?) the paths
are likely to change again. SDK 26.0 still has lint 25.3.1, so the SDK and
lint versions don't appear to be tied.
It seems that only the lint* jars are needed to compile 'build/annotationProcessor',
however we need all the remaining jars in the classpath when running that code
in 'widget/android/bindings'.
MozReview-Commit-ID: GAKwMrVXW55
--HG--
extra : rebase_source : 4e790aaccae8ccc3f151c39bf1ef4404b2581d7a
When using NDK r13+ with part 1, the following build error still occurs when using cmath.
0:21.01 /mozilla/android-ndk-r14b/sources/cxx-stl/llvm-libc++/include/math.h:661:105: error: 'acosl' was not declared in this scope
0:21.01 inline _LIBCPP_INLINE_VISIBILITY long double acos(long double __lcpp_x) _NOEXCEPT {return acosl(__lcpp_x);}
To fix this, we need change the order of include path.
MozReview-Commit-ID: AZ3ebx3pAil
--HG--
extra : rebase_source : ca3ef90372a9c2b84f4a9bed3581be0f21e17d77
The inner directory in these projects was removed in r13.
MozReview-Commit-ID: AxWgxXyOKJ4
--HG--
extra : rebase_source : 1290e8c709930f28f80bc201d7e72b53b8063634
This ensures that we correctly link against /usr/lib/libc++.1.dylib.
With this flag left in LDFLAGS, the linker would find libc++.1.dylib in
clang/lib, which would cause us to link against @rpath/libc++.1.dylib.
This ensures that we correctly link against /usr/lib/libc++.1.dylib.
With this flag left in LDFLAGS, the linker would find libc++.1.dylib in
clang/lib, which would cause us to link against @rpath/libc++.1.dylib.
CLOSED TREE
Backed out changeset 01cfc71ce542 (bug 1322735)
Backed out changeset 84c729c41230 (bug 1322735)
Backed out changeset b419aaefae95 (bug 1322735)
A side effect of ffi.m4 was that it would un-set CFLAGS when compiling with MSC.
This incidentally worked around the fact that if CFLAGS are set in jemalloc's
configure, it will not set certain variables that are necessary to configure
jemalloc, but are unlikely to be set manually by someone building Firefox.
MozReview-Commit-ID: 2l54YhSAUaL
CLOSED TREE
--HG--
extra : amend_source : 1f0c7bbb5aa8a3dab38f0785e13e32f59e8f8c79
extra : histedit_source : ca99420cac7019a4b6fd6aab781b93151092a8bc%2C0ef091317a27688c734f20417875406726e35de7
I recently discovered that build-tools 23.0.3 doesn't seem to exist on Fedora. Although
I fixed mozboot to download 23.0.1 instead, the builds were still failing because of
configure required 23.0.3.
This seems like an artificial limitation, as building with 23.0.1 seems to work just fine.
This patch will allow either 23.0.1 or 23.0.3.
Note: It would probably better to check for some "minimum" version of build-tools I think
GNU sort has a -V option we could use, but I don't know how cross-platform this is.
MozReview-Commit-ID: 8W0P3yyAHu1
--HG--
extra : rebase_source : 166c1b66962cac6af92c4ce6d3cc1780d9049800
aarch64 doesn't require compiler flag when using NEON. To use NEON on aarch64 with BUILD_ARM_NEON, we should move NEON's flags to NEON_FLAGS like SSE2_FLAGS.
MozReview-Commit-ID: PGDjcHdTfH
--HG--
extra : rebase_source : d21bd6f743f363fdd183bda9b9f308fcf100186b
Nick Alexander
Cosmin Sabou
David Major
Guillaume GARDET
shindli
Ted Mielczarek
Nathan Froyd
Thomas P.
Andreea Pavel
arthur.iakab
Mike Hommey
Jim Chen
Tom Ritter
Sebastian Hengst
Jan Beich
Henri Sivonen
Mike Hommey
Andi-Bogdan Postelnicu
Jan Henning
Benjamin Bouvier
Jeff Walden
Matías Zúñiga
Makoto Kato
Bill McCloskey
Chris Manchester
Philip Chimento
Carsten "Tomcat" Book
Sebastian Kaspari
Gregory Szorc
mozilla@mozilla-23773
Sylvestre Ledru
Christian Holler
André Bargull
Alex Gaynor
Jim Blandy
Andrzej Hunt
Dan Albert
Jesse Schwartzentruber
Ehsan Akhgari
Iris Hsiao
Eric Rahm
Ryan VanderMeulen
Phil Ringnalda
Daniel Stenberg
Ting-Yu Chou
Mike Shal
Jonathan Kew
James Willcox
Tom Schuster
Kris Maglione
Andrew Halberstadt