summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* use -l to set the macro dir when using autoconf-2.13 #448986 by Ryan SleeviMike Frysinger2013-01-281-4/+9
|
* Avoid unsafe nested multijob_finish_one in eautoreconf for bug #426512.Zac Medico2013-01-051-3/+9
|
* Be more strict about eautoreconfig on aclocal.m4. Fixes bug#424763 and ↵Tomas Chvatal2012-09-201-2/+2
| | | | bug#420631.
* _elibtoolize: properly set LIBTOOLIZE in case glibtoolize exists in the ↵Fabian Groffen2012-07-241-3/+2
| | | | system, bug #418465
* add automake-1.12 to the list so people with only that version installed can ↵Mike Frysinger2012-06-081-3/+3
| | | | build things #419459 by Oleg
* run autotools in subdirs in parallel with new multiprocessing eclassMike Frysinger2012-06-071-9/+20
|
* allow gtkdoc steps to be skipped if the tools are not installed #419979 -- ↵Mike Frysinger2012-06-071-6/+12
| | | | if packages want these steps, they can DEPEND on the pkg in question; similarly, give a blanket pass to autopoint for embedded/USE=-nls systems that do not install the gettext package #417641
* Support other GNOME-related tools in eautoreconf: gtk-doc, gnome-doc & ↵Michał Górny2012-06-061-9/+18
| | | | glib-gettext.
* Avoid type -P output for glibtoolize, bug #419641Fabian Groffen2012-06-051-2/+2
|
* do not require aclocal.m4 to trace macros as simpler packages might not end ↵Mike Frysinger2012-05-241-5/+5
| | | | up generating one, yet will expand autoconf macros all the time that might be the ones we are looking for #417275 by Rick Farina
* also scan for older AC_PROG_INTLTOOL since not all packages have updated yetMike Frysinger2012-05-211-3/+3
|
* run intltoolize when IT_PROG_INTLTOOL is in use #416789 by Markus OehmeMike Frysinger2012-05-211-4/+8
|
* re-arch external tool handling to make it scalable for adding more toolsMike Frysinger2012-05-211-15/+53
|
* change eautomake over to autotools_check_macro now that we cache the results ↵Mike Frysinger2012-05-201-6/+2
| | | | and it should be quick
* optimize autotools_check_macro to cache the result of the autoconf trace as ↵Mike Frysinger2012-05-201-15/+41
| | | | we can end up calling this code multiple times when doing `eautoreconf`
* allow autotools_check_macro_val to take multiple macros when the values can ↵Mike Frysinger2012-05-201-10/+16
| | | | safely be merged by the caller (like macros that implicitly create dirs, and we just want a list of all those dirs)
* make eliding of --install transparent to the caller so that we can drop ↵Mike Frysinger2012-05-201-14/+14
| | | | duplicated variable expansion and simplify higher code flows
* we don't use anything from eutils eclass now, so drop the inheritMike Frysinger2012-05-201-2/+2
|
* always depend on automake even when WANT_AUTOMAKE=none #401605Mike Frysinger2012-05-201-4/+5
|
* eautoreconf: automatically run eautopoint when the code base uses itMike Frysinger2012-05-201-1/+4
|
* automatically create AC_CONFIG_MACRO_DIR dir tooMike Frysinger2012-05-201-4/+6
|
* drop AT_NOE{ACLOCAL,AUTOCONF,AUTOHEADER} as people should not be disabling ↵Mike Frysinger2012-03-221-25/+9
| | | | these things, and document AT_NOEAUTOMAKE as a hack rather than a normally acceptable solution
* handle commented out AM_INIT_AUTOMAKE #398743Mike Frysinger2012-03-221-2/+4
|
* restore functionality lost in previous commitMike Frysinger2012-03-211-13/+17
|
* Go back to use a variable rather than an array for _LATEST_AUTOMAKE, use ↵Diego Elio Pettenò2012-03-211-15/+18
| | | | versionator eclass to go back to the full version instead, so that it's clear what's going on and other developers don't misread the code. It's only perfect for EAPI >= 0 but it's not broken on EAPI=0 anyway.
* Fix automake depend atom to not force downgrade from 1.11.2 or 1.11.2 to ↵Samuli Suominen2012-03-211-2/+2
| | | | 1.11.1 wrt #409083
* Raise _AUTOMAKE_LATEST from 1.11 to 1.11.1 wrt #408981Mike Frysinger2012-03-211-6/+8
|
* Revert for a minute.Samuli Suominen2012-03-201-2/+2
|
* Raise _AUTOMAKE_LATEST from 1.11 to 1.11.1 wrt #408981Samuli Suominen2012-03-201-2/+2
|
* Remove my WANT_AUTOMAKE=none overloading in eautomake and instead introduce ↵Robin H. Johnson2012-02-201-13/+30
| | | | AT_NOEACLOCAL, AT_NOEAUTOCONF, AT_NOEAUTOMAKE that work the same as AT_NOELIBTOOLIZE. Should fix bug #404555.
* require autoconf-2.68 since that is the latest stable now #403035Mike Frysinger2012-02-131-2/+2
|
* update comments to reflect reality of func namesMike Frysinger2012-02-131-3/+3
|
* Provide a way to run eautoreconf without automake by using WANT_AUTOMAKE=none.Robin H. Johnson2012-02-121-2/+8
|
* add a helper for updating config.rpath files for projects that dont use gettextMike Frysinger2012-01-061-1/+20
|
* always run automake if the configure file calls AM_INIT_AUTOMAKE #397697 by ↵Mike Frysinger2012-01-041-6/+4
| | | | My Th
* fix thinko with array handling when injecting args into the cmdlineMike Frysinger2011-12-141-6/+6
|
* push down AT_M4DIR handling into autotools_run_tool to keep the callers ↵Mike Frysinger2011-12-141-13/+28
| | | | simpler, and change the ordering with AT_SYS_M4DIR so that it comes after the ebuild/package settings rather than in the middle
* drop old M4DIR support since we have AT_M4DIR and the tree uses only that, ↵Mike Frysinger2011-12-141-3/+5
| | | | and note GNU m4 as a requirement
* initial AT_SYS_M4DIR for weird system integrators ↵Mike Frysinger2011-12-131-7/+13
| | | | (cross-compiling/prefix/etc...)
* unify duplicated code for extracting values from traced variablesMike Frysinger2011-12-131-26/+13
|
* avoid multiple inclusions when possible to speed caching upMike Frysinger2011-12-131-1/+6
|
* add a --force flag to elibtoolize so that eautoreconf can make sure it runs ↵Mike Frysinger2011-11-141-2/+2
| | | | even in face of poor interactions with earlier ebuilds/eclasses code #232820
* also check for README files for --foreign (reported by Mr_Bones_)Mike Frysinger2011-10-051-2/+2
|
* drop comment that bleeds into man pageMike Frysinger2011-09-231-3/+1
|
* handle multiline ACLOCAL_AMFLAGS #383525 by Mikael MorinMike Frysinger2011-09-231-14/+28
|
* use `die` rather than invalid depend atoms since council ok-ed it some time agoMike Frysinger2011-08-221-3/+3
|
* add a WANT_LIBTOOL knob so people can disable the dep if they really know ↵Mike Frysinger2011-08-221-4/+18
| | | | what they are doing
* make sure log output is unique #378057 by Martin von GagernMike Frysinger2011-08-071-2/+7
|
* setup env before loading ACLOCAL_AMFLAGS #365401 by Dennis SchriddeMike Frysinger2011-05-161-6/+13
|
* Make sure to use AT_M4DIR values for autoheader and autoconf calls as well.Diego Elio Pettenò2011-04-061-19/+25
| | | | This is required to re-build autotools on app-admin/sudo-1.8*