adjutrix -k shows ~arch as arch

[rbu]

From today's sync of Portage, I get the following behaviour: imlib-1.4.0-r1 is ~arch, but adjutrix -k will display "arch" for all arches:

Keywords for media-libs/imlib2:

| s | | p | | a x | | r 8 | | c 6 | u | a a p s - - | n | l m h i m m p s p f f | u s | p d a p a 6 i p c 3 a b x b | s l | h 6 r p 6 8 p p 6 9 s r s 8 s | e o | a 4 m a 4 k s c 4 0 h c d 6 d | d t

---

1.4.0 | + + + + + + + + + + + ~ | * 0 1.4.0-r1 | + + + + + + + + + + + ~ | 1.4.1.000 | ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ | * 1.4.1.000-r1 | ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ |

$ grep KEYW imlib2* imlib2-1.4.0.ebuild:KEYWORDS="alpha amd64 arm hppa ia64 ~mips ppc ppc64 sh sparc x86 ~x86-fbsd" imlib2-1.4.0-r1.ebuild:KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd" imlib2-1.4.1.000.ebuild:KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd"

$ cvs log imlib2-1.4.0-r1.ebuild

RCS file: /var/cvsroot/gentoo-x86/media-libs/imlib2/imlib2-1.4.0-r1.ebuild,v Working file: imlib2-1.4.0-r1.ebuild head: 1.1 branch: locks: strict access list: symbolic names: keyword substitution: kv total revisions: 1; selected revisions: 1 description:

---

revision 1.1 date: 2008-05-31 07:47:12 +0200; author: vapier; state: Exp; commitid: 48964840e65f4567; Add fixes from upstream for CVE-2008-2426 #223965. (Portage version: 2.2_pre5.spank.spunk)

$ adjutrix -V adjutrix, part of paludis 0.26.1

[rbu]

I should have used Code formatting for the output, redoing for adjutrix -k

Keywords for media-libs/imlib2:

             |                         s     |
             |                         p     |
             |                         a   x |
             |                         r   8 |
             |                         c   6 | u
             | a a             p     s -   - | n
             | l m   h i m m   p s   p f   f | u s
             | p d a p a 6 i p c 3   a b x b | s l
             | h 6 r p 6 8 p p 6 9 s r s 8 s | e o
             | a 4 m a 4 k s c 4 0 h c d 6 d | d t
-------------+-------------------------------+----
1.4.0        | + + + + +   + + +   + +   + ~ | * 0
1.4.0-r1     | + + + + +   + + +   + +   + ~ |
1.4.1.000    | ~ ~ ~ ~ ~   ~ ~ ~   ~ ~   ~ ~ | *
1.4.1.000-r1 | ~ ~ ~ ~ ~   ~ ~ ~   ~ ~   ~ ~ |

[ciaranm]

What does paludis -qM think?

[rbu]

I got the adjutrix results on a CVS checkout that was not actually the tree I was using (for portage). The result does not occur outside of the actual tree: Running "adjutrix -k" in /usr/portage/media-libs/imlib2 works as expected. Running it inside my CVS checkout works weirdly. Diff tells me the directories are completely the same (except for CVS subdirs). So I understand the "paludis" output will not bee too helpful.

rbu@peanut ~/devel/gentoo/gentoo-x86/media-libs/imlib2 $ paludis -qM imlib2
paludis@1212278426: [WARNING portage_environment.dodgy] Use of Portage configuration files will lead to sub-optimal performance and loss of functionality. Full support for Portage configuration formats is not guaranteed; issues should be reported via trac.
* media-libs/imlib2
    gentoo:                  1.4.0 (1.4.0-r1)K (1.4.1.000)K (1.4.1.000-r1)K {:0}
    installed:               1.4.0* {:0}
    DESCRIPTION:             Version 2 of an advanced replacement library for libraries like libXpm
    HOMEPAGE:                http://www.enlightenment.org/Libraries/Imlib2/
    IUSE:                    (X) (bzip2) (-doc) (gif) (jpeg) (-mmx) (mp3) (nls) (png) (tiff) (zlib)
    LICENSE:                 BSD
    INSTALLED_TIME:          Sat Nov 24 08:05:03 2007
    repository:              gentoo
    DEPEND:                  =media-libs/freetype-2* bzip2? ( app-arch/bzip2 ) zlib? ( sys-libs/zlib ) gif? ( >=media-libs/giflib-4.1.0 ) png? ( >=media-libs/libpng-1.2.1 ) jpeg? ( media-libs/jpeg ) tiff? ( >=media-libs/tiff-3.5.5 ) X? ( x11-libs/libXext x11-proto/xextproto ) mp3? ( media-libs/libid3tag ) doc? ( app-doc/doxygen )
    RDEPEND:                 =media-libs/freetype-2* bzip2? ( app-arch/bzip2 ) zlib? ( sys-libs/zlib ) gif? ( >=media-libs/giflib-4.1.0 ) png? ( >=media-libs/libpng-1.2.1 ) jpeg? ( media-libs/jpeg ) tiff? ( >=media-libs/tiff-3.5.5 ) X? ( x11-libs/libXext x11-proto/xextproto ) mp3? ( media-libs/libid3tag ) nls? ( sys-devel/gettext )
    PDEPEND:
    SRC_URI:                 http://enlightenment.freedesktop.org/files/imlib2-1.4.0.tar.gz mirror://sourceforge/enlightenment/imlib2-1.4.0.tar.gz
    VDBDIR:                  /var/db/pkg/media-libs/imlib2-1.4.0
    USE:                     (3dnow) (3dnowex) (X) (X509) (a52) (aac) (aalib) (acl) (acpi) (alsa) (alsa_cards_ali5451) (alsa_cards_als4000) (alsa_cards_atiixp) (alsa_cards_atiixp-modem) (alsa_cards_bt87x) (alsa_cards_ca0106) (alsa_cards_cmipci) (alsa_cards_emu10k1x) (alsa_cards_ens1370) (alsa_cards_ens1371) (alsa_cards_es1938) (alsa_cards_es1968) (alsa_cards_fm801) (alsa_cards_hda-intel) (alsa_cards_intel8x0) (alsa_cards_intel8x0m) (alsa_cards_maestro3) (alsa_cards_trident) (alsa_cards_usb-audio) (alsa_cards_via82xx) (alsa_cards_via82xx-modem) (alsa_cards_ymfpci) (alsa_pcm_plugins_adpcm) (alsa_pcm_plugins_alaw) (alsa_pcm_plugins_asym) (alsa_pcm_plugins_copy) (alsa_pcm_plugins_dmix) (alsa_pcm_plugins_dshare) (alsa_pcm_plugins_dsnoop) (alsa_pcm_plugins_empty) (alsa_pcm_plugins_extplug) (alsa_pcm_plugins_file) (alsa_pcm_plugins_hooks) (alsa_pcm_plugins_iec958) (alsa_pcm_plugins_ioplug) (alsa_pcm_plugins_ladspa) (alsa_pcm_plugins_lfloat) (alsa_pcm_plugins_linear) (alsa_pcm_plugins_meter) (alsa_pcm_plugins_mulaw) (alsa_pcm_plugins_multi) (alsa_pcm_plugins_null) (alsa_pcm_plugins_plug) (alsa_pcm_plugins_rate) (alsa_pcm_plugins_route) (alsa_pcm_plugins_share) (alsa_pcm_plugins_shm) (alsa_pcm_plugins_softvol) (amd64) (apache2) (arts) (asf) (audiofile) (avahi) (bash-completion) (berkdb) (bitmap-fonts) (bl) (bluetooth) (bzip2) (bzlib) (cairo) (cdr) (cli) (cracklib) (crypt) (cups) (curl) (dbus) (dga) (divx) (dri) (dts) (dv) (dvd) (dvdr) (dvdread) (eds) (elibc_glibc) (emboss) (encode) (esd) (evo) (exif) (fam) (ffmpeg) (firefox) (flac) (foomaticdb) (fortran) (ftp) (gcj) (gdbm) (ggi) (gif) (gimpprint) (glibc-omitfp) (glitz) (gmedia) (gnome) (gpm) (gstreamer) (gtk) (gtk2) (hal) (http) (iconv) (idn) (ieee1394) (imagemagick) (imlib) (input_devices_evdev) (input_devices_keyboard) (input_devices_mouse) (input_devices_wacom) (ipv6) (isdnlog) (jack) (java) (jpeg) (kde) (kerberos) (kernel_linux) (kqemu) (lcd_devices_luise) (lcd_devices_ncurses) (lcd_devices_picolcd) (ldap) (libcaca) (libnotify) (linguas_de) (linguas_en) (live) (logitech-mouse) (lzo) (mad) (midi) (mikmod) (mmx2) (mmxext) (mozilla) (mp3) (mpeg) (mudflap) (musepack) (nas) (ncurses) (network) (networkmanager) (nls) (nptl) (nptlonly) (nsplugin) (nvidia) (ogg) (openal) (opengl) (openmp) (oss) (pam) (pcre) (pdf) (perl) (php) (pic) (png) (ppds) (pppd) (pulseaudio) (python) (qemu) (qt3) (qt3support) (qt4) (quicktime) (readline) (realmedia) (reflection) (rss) (samba) (sasl) (sdl) (session) (slang) (sndfile) (speex) (spell) (spl) (sse) (sse2) (ssl) (subversion) (svg) (tcltk) (tcpd) (tetex) (theora) (threads) (tiff) (truetype) (truetype-fonts) (type1) (type1-fonts) (unicode) (urandom) (usb) (userland_GNU) (v4l) (v4l2) (vcd) (video_cards_i810) (video_cards_vesa) (vorbis) (wma) (wmf) (wmp) (x264) (xanim) (xen) (xine) (xml) (xorg) (xspf) (xv) (xvid) (xvmc) (zlib)
    INHERITED:               enlightenment eutils libtool multilib portability toolchain-funcs
    CBUILD:                  x86_64-pc-linux-gnu
    CFLAGS:                  -march=nocona -mtune=nocona -O2 -pipe -mmmx -msse -msse2 -msse3 -mfpmath=sse -fweb -frename-registers -fomit-frame-pointer -fno-ident
    CHOST:                   x86_64-pc-linux-gnu
    CXXFLAGS:                -march=nocona -mtune=nocona -O2 -pipe -mmmx -msse -msse2 -msse3 -mfpmath=sse -fweb -frename-registers -fomit-frame-pointer -fno-ident
    LDFLAGS:                 -Wl,--as-needed -Wl,-O1

Key to mask reasons:

* K: keywords

[rbu]

s/The result does not occur outside of the actual tree/The result does not occur inside of the actual tree/

[ciaranm]

Try adding in some 1>&2 output in builtin_metadata.bash, see what keys are being passed from bash to adjutrix.

[ciaranm]

I can't reproduce this. If anyone can, please indicate:

• whether it's for all packages, just one package or what
• what builtin_metadata.bash is showing for the keywords
• what keywords_graph.cc is showing for the keywords

[rbu]

I hope it's appropriate to reopen. I added a

        echo "${key}=\""${k}"\"" 1>&2

to /usr/libexec/paludis/0/builtin_metadata.bash, and the output is as follows (cut):

imlib2-1.4.0-r1> DEPEND="=media-libs/freetype-2* bzip2? ( app-arch/bzip2 ) zlib? ( sys-libs/zlib ) gif? ( >=media-libs/giflib-4.1.0 ) png? ( >=media-libs/libpng-1.2.1 ) jpeg? ( media-libs/jpeg ) tiff? ( >=media-libs/tiff-3.5.5 ) X? ( x11-libs/libXext x11-proto/xextproto ) mp3? ( media-libs/libid3tag ) doc? ( app-doc/doxygen ) "
imlib2-1.4.0-r1> RDEPEND="=media-libs/freetype-2* bzip2? ( app-arch/bzip2 ) zlib? ( sys-libs/zlib ) gif? ( >=media-libs/giflib-4.1.0 ) png? ( >=media-libs/libpng-1.2.1 ) jpeg? ( media-libs/jpeg ) tiff? ( >=media-libs/tiff-3.5.5 ) X? ( x11-libs/libXext x11-proto/xextproto ) mp3? ( media-libs/libid3tag ) nls? ( sys-devel/gettext ) "
imlib2-1.4.0-r1> PDEPEND=" "
imlib2-1.4.0-r1> IUSE="X bzip2 gif jpeg mmx mp3 png tiff zlib nls doc "
imlib2-1.4.0-r1> SRC_URI="http://enlightenment.freedesktop.org/files/imlib2-1.4.0.tar.gz mirror://sourceforge/enlightenment/imlib2-1.4.0.tar.gz"
imlib2-1.4.0-r1> RESTRICT=""
imlib2-1.4.0-r1> LICENSE="BSD"
imlib2-1.4.0-r1> KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd alpha amd64 arm hppa ia64 mips ppc ppc64 sh sparc x86 ~x86-fbsd "
imlib2-1.4.0-r1> INHERITED=" toolchain-funcs multilib portability eutils libtool enlightenment"
imlib2-1.4.0-r1> PROVIDE=""
imlib2-1.4.0-r1> HOMEPAGE="http://www.enlightenment.org/"
imlib2-1.4.0-r1> DESCRIPTION="Version 2 of an advanced replacement library for libraries like libXpm"
imlib2-1.4.0-r1> DEPENDENCIES=""
imlib2-1.4.0-r1> E_IUSE=" nls doc "
imlib2-1.4.0-r1> E_DEPEND=" doc? ( app-doc/doxygen ) "
imlib2-1.4.0-r1> E_RDEPEND=" nls? ( sys-devel/gettext ) "
imlib2-1.4.0-r1> E_PDEPEND=" "
imlib2-1.4.0-r1> E_KEYWORDS=" alpha amd64 arm hppa ia64 mips ppc ppc64 sh sparc x86 ~x86-fbsd "
imlib2-1.4.0-r1> PLATFORMS=""
imlib2-1.4.0-r1> E_PLATFORMS=""
imlib2-1.4.0-r1> MYOPTIONS=""
imlib2-1.4.0-r1> E_MYOPTIONS=""
imlib2-1.4.0-r1> E_DEPENDENCIES=""
imlib2-1.4.0-r1> BINARY_KEYWORDS=""
imlib2-1.4.0-r1> BINARY_URI=""
imlib2-1.4.0-r1> GENERATED_USING=""
imlib2-1.4.0-r1> GENERATED_TIME=""
imlib2-1.4.0-r1> BINARY_PLATFORMS=""
imlib2-1.4.0-r1> SLOT="0"
imlib2-1.4.0-r1> EAPI="0"
imlib2-1.4.0-r1> OPTIONS=""
imlib2-1.4.0-r1> USE=""
1.4.0-r1     | + + + + +   + + +   + +   + ~ |

Again, in the same directory: $ grep KEYW *1.4.0-r1* KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd"

Please note this is in a tree checkout, not the actual tree. The portage provided tree has stable keywords, but even if I revert those, the situation stays. /usr/portage/media-libs/imlib2 $ grep KEYW *1.4.0-r1* KEYWORDS="alpha amd64 arm hppa ia64 ~mips ppc ppc64 sh sparc x86 ~x86-fbsd"

No, it does not happen for all packages (for instance, it does not happen for other versions in the same directory).

If I add the following to ebuild_f_metadata()

    echo "${KEYWORDS}" 1>&2

I get this:

Keywords for media-libs/imlib2:

imlib2-1.4.1.000-r1> ~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd       ~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd
imlib2-1.4.1.000> ~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd       ~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd
imlib2-1.4.0-r1> ~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sh ~sparc ~x86 ~x86-fbsd       alpha amd64 arm hppa ia64 mips ppc ppc64 sh sparc x86 ~x86-fbsd
imlib2-1.4.0> alpha amd64 arm hppa ia64 ~mips ppc ppc64 sh sparc x86 ~x86-fbsd       alpha amd64 arm hppa ia64 mips ppc ppc64 sh sparc x86 ~x86-fbsd
             |                         s     |
             |                         p     |
             |                         a   x |
             |                         r   8 |
             |                         c   6 | u
             | a a             p     s -   - | n
             | l m   h i m m   p s   p f   f | u s
             | p d a p a 6 i p c 3   a b x b | s l
             | h 6 r p 6 8 p p 6 9 s r s 8 s | e o
             | a 4 m a 4 k s c 4 0 h c d 6 d | d t
-------------+-------------------------------+----
1.4.0        | + + + + +   + + +   + +   + ~ | * 0
1.4.0-r1     | + + + + +   + + +   + +   + ~ |
1.4.1.000    | ~ ~ ~ ~ ~   ~ ~ ~   ~ ~   ~ ~ | *
1.4.1.000-r1 | ~ ~ ~ ~ ~   ~ ~ ~   ~ ~   ~ ~ |

If this is unexpected, please let me know what else to change to help debug.

[ciaranm]

That's weird. The bash side is getting two sets of KEYWORDS. Are you absolutely sure there's no eclass weirdness, or ebuild KEYWORDS manipulation that your grep is missing?

[ciaranm]

Such as from that horrible 'enlightenment' eclass, for example?

[rbu]

Yes, it's caused by the enlightenment eclass, which resets KEYWORDS. Paludis merges the ebuild's ~arch and eclasses arch. Portage reacts with overwriting it, keeping only ~arch. Thanks, I guess it's not primarily a bug in Paludis then.

[dleverton]

*mutters something about ciaranm and his silly hacks*

Yes, the eclass is crazy, but Paludis shouldn't be doing that.

[dleverton]

8d653af