Instead of just making our own "check-local" and calling the binaries
ourselves, just append them into "TESTS" variable. Then they run after
all check_PROGRAMS are compiled.
The reasons for changing are:
1) If we change the test and call "make check" the test is not
compiled again -- and the only way to compile it is to "make clean".
2) There's no need to reinvent the wheel here.
With a recent version of Automake, the test output is redirected to log
files. This is good but unexpected for whom was used to the previous
way. So, be warned.
SVN revision: 82841
This patch was brought to you by:
sed -i'' 's/svnversion/LC_ALL=C svnversion/' configure.ac
Anybody who has the entire trunk and branches around is welcome to apply an analogue patches to it.
SVN revision: 82822
Instead of -I$(top_srcdir)... -I$(top_builddir)... and then do it for
the .la, use the EFL_ macros to generate the contents to be used in
automake files.
There is a nasty bit that libtool will parse Makefile*.am and will not
get _DEPENDENCIES from _LIBADD and _LDADD if these are in
@REPLACEMENT@. To solve this we must explicitly set _DEPENDENCIES. The
contents of this is almost the same as _LIBADD or _LDADD with the
"_INTERNAL_" replacement name.
I hope the code will be result will be shorter and consistent as there
is less places to change when we add/remove dependencies.
Statistics are quite impressive (diffstat):
{{{
37 files changed, 663 insertions(+), 1599 deletions(-)
}}}
SVN revision: 82785
This improves the tests on freebsd, which doesn't have clearenv() (
though this commit was untested on such environement). clearenv was
rejected from POSIX so we better aim to stop its usage. However since
tests use setenv() this may lead to leaks, but it's better a leak than a
failing test.
On Linux all tests that were passing continue to pass.
SVN revision: 82779
it was reported that AC_CHECK_FUNCS() couldn't work on netbsd because
dirfd is a macro and it would not properly evaluate int that
traditional test.
SVN revision: 82736
- they have no 'all' rule, keep out of SUBDIRS
- they depend on 'all-am', the non-recursive target that builds everything.
- they do not need a directory on its own to declare nothing.x
SVN revision: 82689
This one was a painful bitch. The edbus2 port was quite broken, mainly
leaking eina_stringshare and also not adding the '\0' to the strings
that are represented as bytearray (paths cannot be utf8 to avoid
translations).
Emotion plugin was also quite bogus and the video thumbnail as edje
(animated) is not working yet due bug in Edje_Edit api -- someone
needs to investigate this, seems strange.
Emotion plugin also had a bug that it was deleting the object from
inside object callback.
Now it seems to work. Please report if it does not.
SVN revision: 82675
buffer is lightweight and dependency for many engines, merge it back
into core.
extn is a module on its own, and it's the only one linking to
ecore_ipc, no need to add that to ecore_evas.
minor cosmetic changes to configure to make output consistent.
SVN revision: 82648
* strings.h: is very useful, very old and very standard. The only place
that checks for that is simple_xml_parser and I have no idea why
SVN revision: 82576
this one was quite a huge work, but hopefully it's correct.
NOTES:
* removed vlc generic module, it should go into a separate package.
* gstreamer is enabled by default (see --disable-gstreamer)
* xine is disabled by default (see --enable-gstreamer)
* generic is always built statically if supported
* gstreamer and xine can't be configured as static (just lacks command line options, build system supports it)
* v4l2 is enabled by default on linux if eeze is built (see --disable-v4l2)
* emotion_test moved to src/tests/emotion and depends on EFL_ENABLE_TESTS (--with-tests), but is still installed if enabled.
TODO (need your help!):
* fix warnings with gstreamer and xine engine
* call engine shutdown functions if building as static
* remove direct usage of PACKAGE_*_DIR and use eina_prefix
* add eina_prefix checkme file as evas and others
* add support for $EFL_RUN_IN_TREE
* create separate package for emotion_generic_modules
* check docs hierarchy (doxygen is segv'in here)
SVN revision: 82501
* Split out ecore_imf_xim to do its own check
* Fixed problem with xcb's makekeys, no rule for
$(top_builddir)/src/utils/ecore/makekeys$(EXEEXT) exists so make
used an implicit rule (ignoring any cflags of course)
* Fixed gl_x11 engine to build with either Xlib or XCB (xcb flags were
missing)
* Added EFL_FIND_X and replace any used of AC_PATH_X{,TRA}. First
looks for Xorg pkg-config files then if those arn't found it falls
back to old AC_PATH_X. Also generalized common header and lib
checks. Could probably use some polishing (the AC_CACHE_VAL cruft
especially) but this is what I have time for tonight.
Now X11 should be found on non-standard locations by means of xmkmf,
--x-includes/--x-libraries and also pkg-config.
SVN revision: 82475
Weirdly enough raster turned PROGRAM cache on to fix a bug saying that
the other path was not tested, but from IN-EFL/edje/configure.ac CALC
cache was enabled and PROGRAM was disabled.
I kept the PROGRAM on and enabled CALC. Added fixed point as well.
Cedric, Raster: should we keep PROGRAM and CALC cache options if the
other path is not tested? Can't we assume they are always on? How much
memory these cache add? Can't this be reduced?
SVN revision: 82410
Tests are disabled by default, but they should be mandatory for
distcheck. So should be examples, we should have them to at least
compile.
As there is no way to force "make distcheck" to run "make examples",
I've created --enable-always-build-examples to make examples regular
noinst_PROGRAMS, being always compiled. distcheck will use that
option.
SVN revision: 82276
this is still in progress, mostly the multisense stuff is pending.
it seems that when we merge ecore_audio in edje the libremix and
similar are gone, at least from Edje, and will be in ecore_audio
itself (or pulseaudio).
Changes:
* __UNUSED__ to EINA_UNUSED
* binaries (epp, embryo_cc, edje_cc) now consider EFL_RUN_IN_TREE and
will assume the binaries are still not installed, running from
build tree location (needs more testing, maybe doesn't work with
srcdir != builddir, still doesn't solve cross compile builds)
SVN revision: 82139
Changes also in this commit:
* fix missing EAPI in symbols used by modules
* removed old libudev and libmount support as agreed by discomfitor/zmike
* replaced __UNUSED__ with EINA_UNUSED
* fixed docs hierarchy
SVN revision: 82100
va_list may be typedef'ed not only to array and pointer but also to a
plain struct. It could be made to work this way, but it's a lot simpler
a safer to not depend on it. To deal with the array corner case we copy
the va_list from the function arguments to the stack and call the "real"
function passing it.
SVN revision: 82017
As reported by Doug Newgard it was causing problems and seems this is
the reason. No idea why it works for me in the same
distro/autoconf/automake.
SVN revision: 81959
- fchmod() was isolated by HAVE_CHMOD, which was always present
before... then fchmod() is also present as no errors were reported
since its introduction.
- fchmod() is POSIX for a while now.
- lstat() is POSIX for a while now.
- setxattr is supported by EFL_CHECK_FUNCS() as is used by eina.
- splice() check added to EFL_CHECK_FUNCS()
SVN revision: 81938
some functions were being checked but their IFDEF not being handled,
others (getpwnam and getgrnam) were isolated in a code that is now
Windows, then no need to check them as well as these are POSIX.
SVN revision: 81934
okay, remove that print as it will always print sections removed from
PIC leftovers, etc.
keep the --gc-sections, it's useful to shrink binaries. Developers
willing to delete dead code can --print-gc-sections themselves.
SVN revision: 81920
seems that automake will parse LDFLAGS for -module and if it's not
present it will complain about name not starting with 'lib'.
seems my last try was without NOCONFIGURE=1 and autogen continued to
the old ./configure, that printed lots of messages and the error went
unnoticed
SVN revision: 81917
- remove EFL_LIBS and EFL_CFLAGS, use per-lib values that inherit
from EFL (general)
- add NAME_LDFLAGS and EFL_LDFLAGS for linker flags.
- LDADD (binaries) now use NAME_LDFLAGS instead of NAME_LIBS, as they
link to libname.la and that will pull in the libtool dependencies
SVN revision: 81915
* -fvisibility=hidden will hide symbols not marked with EAPI in the
final binary (so/executable).
* -ffunction-sections and -fdata-sections will split those into
independent elf sections, then we can -Wl,--gc-sections for those
that are unused. During development --Wl,--print-gc-sections will
warn us of functions that are left unused and were collected.
SVN revision: 81906
note: dns.c were already emitting warnings regarding shadow, so I
removed the flag to quiet override-init. It will a bit louder, but
then we remember it would be better to rewrite this dns in proper EFL.
SVN revision: 81905
added couple of macros:
- EFL_LIB_START(PKG): setup variables and replacements (AC_SUBST),
prints started checks
- EFL_LIB_END(PKG): prints ended checks
- EFL_LIB_START_OPTIONAL(PKG, TEST): wraps EFL_LIB_START() with a
conditional test. defines AC_DEFINE(HAVE_PKG).
- EFL_LIB_END_OPTIONAL(PKG): wraps EFL_LIB_END() with a conditional
test, defines AM_CONDITIONAL(HAVE_PKG).
SVN revision: 81903
in some cases libtool would build 2 versions, pic for shared and
no-pic for static. Make sure we just build one as pic is good enough.
NOTE: this is only used on supported platforms.
SVN revision: 81902
weird enough to build with memcheck.h you just need valgrind's CFLAGS,
not its libraries as they are not supposed to be used like that,
throwing many bgPlain_ errors (vgPlain_tl_pre_clo_init,
vgPlain_free...) from libcoregrind-x86-linux.a
SVN revision: 81901
From Autobook: http://www.sourceware.org/autobook/autobook/autobook_85.html
libtool will build executables suitable for copying into their
respective installation destinations, obviating the need for
relinking them on those hosts which would have required
it. Whenever libtool links an executable which uses shared
libraries, it also creates a wrapper script which ensures that the
environment is correct for loading the correct libraries, See
section 10.5 Executing Uninstalled Binaries. On those hosts which
require it, the wrapper script will also relink the executable in
the build tree if you attempt to run it from there before
installation.
Sometimes this behaviour is not what you want, particularly if you
are developing the package and not installing between test
compilations. By passing `--disable-fast-install', the default
behaviour is reversed; executables will be built so that they can
be run from the build tree without relinking, but during
installation they may be relinked.
The user may still override this default, depending on platform
support, by specifying --enable-fast-install to configure.
NOTE: In my measurements it made no difference in install time either.
SVN revision: 81900
Most systems will never ever need the static builds of EFL, and it
doubles the compilation time (4m36s x 2m48s on my laptop). Then
disable it by default.
The user may still override this default by specifying --enable-static
to configure.
SVN revision: 81898
- moved up, closer to wayland, so it's before ecore_evas (which is
large and still messy)
- pulseaudio and sndfile checks are strict
- pulseaudio is enabled by default
- sndfile is mandatory
SVN revision: 81812
what the hell pass_through has to do with enabling a default mempool I
have no idea, so make it closer to the config.h variable name.
SVN revision: 81802
It's better than nothing, I don't know what the proper fix is.
FRIBIDI_CFLAGS just doesn't exist. Doesn't that macro we use call
PKG_CHECK_MODULES?
SVN revision: 81385
Disabled by default, enable with --enable-audio
ALSA support is disabled as it is not there yet. Pulseaudio should work
though.
Support for .ogg and .wav is there as well (.mp3 is not)
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
SVN revision: 81000
instead of -DHAVE_ECORE_X_XCB in the pc, which is a super-hack, add a
proper compile-time generated ecore_x_version.h that is included by
Ecore_X.h
This file contains HAVE_ECORE_X_XLIB or HAVE_ECORE_X_XCB defined and
will also define ECORE_X_VERSION_MAJOR and ECORE_X_VERSION_MINOR as
per other EFL (Ecore), also defining the runtime version in
Ecore_X_Version.
SVN revision: 80524
Many changes in this commit, but overall it should work. If not let me
know.
* ECORE_CHECK_X_EXTENSION() will now fail if Xlib extension is not found.
* remove usage of ECORE_EVAS_CHECK_MODULE(), using the simplified
ECORE_EVAS_MODULE()
* EVAS_CHECK_ENGINE() will not declare an option anymore.
* fixed --with-x11 and --with-opengl regarding defaults.
* fixed typo in EVAS_STATIC_BUILD_SOFTWARE_X11 -> XLIB conditional
* simplified Xi2 and Xgestures support under --enable-gesture. Let me
know if these should be checked separately.
* remove every other X extension option, they are all forced by
now. If no major showstopper appears with this we may remove all
the #ifdef in the code!
If there is something broken, let me know.
Next on the list for the weekend is to finish the option checks for
SDL, Wayland-EGL, Cocoa and similar.
SVN revision: 80505
still lots to do, but some improvements:
- ecore_evas does not inherit pkg-config from modules since modules are SO
- renamed internal ecore evas define from SOFTWARE_BUFFER to BUFFER,
to make consistent.
SVN revision: 80473
* --with-x11 takes one of:
* xlib: use standard Xlib (used if not windows or ps3)
* xcb: use new xcb library
* none: do not build x11 (used if windows or ps3)
* --with-opengl takes one of:
* full: complete OpenGL support (used if not windows or ps3, ???)
* es: reduced version (OpenGL-ES)
* none: do not build OpenGL (used if windows or ps3, ???)
* --enable-wayland: disabled by default
These are not used as they should yet, will take more work in the
engines to make it happen.
SVN revision: 80467
remove the option and logic, but still leave the framework in there if
some day a rare case of engine that does not rely on software generic
shows.
SVN revision: 80464
platform_pc defines requirements for every platform, each can be
addressed individually by platform_pc_evil, platform_pc_escape or
platform_pc_exotic.
SVN revision: 80454
instead of replicating common patterns all around, let's use the
common patterns as m4 macros.
EFL_DEPEND_PKG() will use pkg-config and add the dependency.
EFL_OPTIONAL_DEPEND_PKG() will use the former, but also defines
have_XYZ and does the if statement for you.
SVN revision: 80402
Instead of distributing system/platform features all around, many are
shared among modules, summarize these things into one single line
using +/- notation.
SVN revision: 80394
* a single option --with-evas-dither-mask=TYPE (big, small, line or none).
* make a wise decision to fallback to small dither mask for
conversions that do not support "no-dither" or "line". Before if
you did not specify it would fallback to big (128x128).
SVN revision: 80383
instead of the previous mess, just define the functions with common
names and call the backend that was compiled in, similar to what eio
does.
also do not be silent on errors, use eina_safety_checks to issue warnings.
SVN revision: 80360
Another try to make inotify checks more common.
This time uses AC_CHECK_HEADERS() as for others, that already define
HAVE_SYS_INOTIFY_H, then uses that.
I still kept AM_CONDITIONAL([HAVE_INOTIFY]) because I plan to convert
ecore_file to the same, smarter, method that is used in eio (compiling
the file depending on the backend.
SVN revision: 80358
* remove the option, as it shouldn't really exist.
* keep code, as requested by Cedric, someone may fix it someday. And
the default is the DISABLED, as the enable was broken.
* Add instructions on how to solve the problem, if a brave folk is up
to it.
SVN revision: 80352
* rename USE_GNUTLS and USE_OPENSSL to HAVE_GNUTLS and HAVE_OPENSSL
in ecore_con, to match other modules such as Eet.
* define requirements_pc_crypto, requirements_pc_deps_crypto and
requirements_libs_crypto so modules can use that.
* move to a common check section.
SVN revision: 80288
* remove EVAS_ prefix as it may be used by other libs some day.
* SSE3 is detected at runtime if x86.
* remove AC_SUBST([altivec_cflags]) as it was not being used anywhere.
* moved to top of file (maybe position is not optimal, let's wait
vtorri to review)
* simplified single-line summary that is as informative as before.
SVN revision: 80284
Actually, this variable wasn't being set on any other backend, so I just
removed it from wayland-egl. I guess it was just a copy & paste typo
when merging ecore into the EFL tree.
SVN revision: 80266
"pkg-config --exists <none>" will return false, thus failing the
configure phase. Not sure if building ecore-con without curl makes sense
though.
SVN revision: 80265
vtorri: I told you these names are very very error prone. I did the
mistake while reading it before, raster did the same today and even
yourself did a typo. The "pc", "libs" and similar in the middle of a
long variable name is easy to miss! Even more if the prefix is
"requirements_" :-(
SVN revision: 80234
After agreement in the mail list, core developers agree to remove this
engine that was not being supported for a long time.
Given that most operations Evas uses are not accelerated in DirectFB,
or at least hardware that exclusively supports DirectFB, it's better
for those people to just use Evas/Ecore software (buffer) rendering
and expose DirectFB's framebuffer as destination surface.
SVN revision: 80232
* remove magic debug output in evas part
* always use version for pc file, it's actually safe
* fix compilation of gl-sdl
* avoid circular dependency of libevas on itself
SVN revision: 78935
* gl engines were checking for eet module, which does not
exist when we install first the efl package.
* fix pkgconfig values for static linking
* add Evas output to configure
SVN revision: 78918
I've tested make -j 3 install and it works nicely
I've tested expedite with software and opengl xlib,
and it works. Not tested other engines, so please
report any problems (engines or other) on the ML.
TODO: examples and tests, I'll add them later
ISSUE: Eina_Unicode size check. It indirectly depends on
eina_config.h, which is created at the end of the
configure script. So its size is always 0. I don't
know how that size is used, so I can't do a lot,
for now.
SVN revision: 78895
Please check.
note1: Only lib and bin for now, but should be extended to other stuff
note2: distcheck does not work because eo_suite is failing.
SVN revision: 78758
seems the debug of threads and stringshare could cause major
slowdowns, then add another profile to produce debug. Changes:
* dev: unlimited log.
* debug: same as dev with debug malloc, threads and stringshare.
SVN revision: 78120
if the distro still wants to support the SUPER OLD pkg-config, then it
will have to fix the generated .pc before installing them:
{{{
sed -i 's/^Requires.private:/Requires:/g' pc/*.pc
}}}
SVN revision: 77818
it's still not merged into EFL single tree, but once it is we should
use "HAVE_EXOTIC" define to work with it.
Before we had: EINA_HAVE_EXOTIC (not used), HAVE_EXOTIC_H (used by
eina_module.c). Since the other libs are all defined as HAVE_LIBNAME,
we're using HAVE_EXOTIC everywhere now.
SVN revision: 77816
it's pointless to have this and may cause problems or complex
configure.ac when we have libraries that use 'eet' during its compile
phase (like elementary generates its profile/configuration).
SVN revision: 77815
Now we always build the following memory pools statically:
- pass_through: calls malloc/free directly, useful to debug.
- chained_pool: default for ages.
- one_big: used by some embedded systems (should we remove?)
Removed:
- ememoa_fixed and ememoa_unknown: depends on a separate lib, not supported?
- buddy: nobody uses it?
NOTE: we do not need the src/modules/eina/mp/*/Makefile.am anymore
since they are statically built. But I'll keep these and the
references in src/modules/eina/mp/Makefile.am
SVN revision: 77792
Introduce a new --with-crypto={openssl,gnutls,none} that will allow
one to choose the EFL cryptographic system.
If set to gnutls or openssl, cipher and signature will be
enabled. Otherwise it's disabled.
NOTE: gnutls is trying to cope with old API and that sucks. Should we
just drop the support for old gnutls and bump the required version?
SVN revision: 77789
Introduced --with-profile={dev,release} that will simplify how to set
build options of EFL.
NOTE-1: specific e17 benchmark is now gone, it will always be built
and is the default benchmark for eina. If we want to have a faster
benchmark in the future, just add a command line option for
eina_suite.
NOTE-2: valgrind build is broken as it needs -fPIC. Will get to it
later. Likely someone needs to revisit the eina mempools for valgrind
and other basic tools (eo? likely evas).
SVN revision: 77771
Vincent, please check those and take some action. I'm not sure about
the windows stuff, but we should find a way to make it more automatic
than what it is right now.
SVN revision: 77769
Remove every mention to libraries that are still not in the
single-tree build such as emotion, efreet and ethumb.
Keep it simple and just add them back when they are merged.
Added similar variables for all of evil, eina, eo and eet. Keep them
in the same order for now.
SVN revision: 77767