| View previous topic :: View next topic |
| Author |
Message |
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
 Posted: Sun Nov 07, 2004 4:57 am Post subject: Stage 1 Install: Fatal Error: scripts/bootstrap.sh: Perl |
 |
|
I'm trying to do a Stage 1 install on a P3-800 system. The bootstrap terminates with the following fatal error:
Bootstrap command: | Code: | | scripts/bootstrap.sh |
Display output following fatal error: | Code: | 20:50:23 (45.58 KB/s) - '/usr/portage/distfiles/gettext-1.03.tar.gz' saved [4064/4064]
>>> md5 src_uri ;-) gettext-1.03.tar.gz
/usr/portage/eclass/perl-module.eclass: line 168: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 171: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 174: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 177: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 180: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 183: perl: command not found
>>> Unpacking source...
>>> Unpacking gettext-1.03.tar.gz to /var/tmp/portage/Locale-gettext-1.03/work
>>> Source unpacked.
/usr/portage/eclass/perl-module.eclass: line 168: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 171: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 174: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 177: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 180: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 183: perl: command not found
/usr/portage/eclass/perl-module.eclass: line 56: perl: command not found
make: *** No targets specified and no makefile found. Stop.
!!! ERROR: dev-perl-molule_src_compile, Line 65, Exitcode 2
!!! Function perl-module_src_compile, Line 65, Exitcode 2
!!! compilation failed
!!! If you need support, post the tompost build error, NOT this status message |
i have been told that this may be a dependency bug in Portage related to perl not compiling as a dependancy of emerge system. i've been recommended to perform the following commands: | Code: | | emerge libperl && emerge libperl && emerge system |
this problem came up in a thread in the Documentation Tips and Tricks forum, but since that isn't a support forum, i thought I'd repeat my support question here:
the fatal error in the bootstrap script was caused by the failure to compile the perl module. although "emerge libperl" will solve the problem related to the absense of perl, i'm unclear as to whether there may be any subsequent operations in the bootstrap script that are being left unperformed.
are there any other steps in the "scripts/bootstrap.sh" script that remain unperformed/incomplete as a result of the perl compilation fatal error? that is to say, if compiling perl is the last step in bootstrapping, its simple enough to resume with the aforementioned commands. but if other bootstrap operations were supposed follow perl compilation but remain unperformed as a result of the fatal error, is it necessary to re-enter the bootstrap operation at its actual termination point, rather than resuming with the following step, building the system with "emerge system"? (if that is the case i am clueless about how to proceed.)
the reason i ask is because i have already tried the recommendation to "emerge libperl && emerge libperl && emerge system", but this alternative method resulted in a later fatal error during "emerge system", suggesting that there could be a problem related to an incomplete bootstrap. i am wondering if the later fatal error has occurred as a result of the libperl fatal error bringing the "scripts/bootstrap.sh" operation to a premature halt.
TIA!
Edit: here is the URL for the thread referenced earlier in this message:
Installing Gentoo: The Developers Method: Stage 1 and NPTL
https://forums.gentoo.org/viewtopic.php?t=189250
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks
Last edited by Bob P on Mon Nov 08, 2004 6:57 pm; edited 1 time in total |
|
| Back to top |
|
 |
donjuan
l33t
Joined: 11 May 2004
Posts: 760
Location: At Uni
|
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Sun Nov 07, 2004 7:00 am Post subject: |
|
|
/etc/make.conf:
| Code: | CHOST="i686-pc-linux-gnu"
CFLAGS="-march=pentium3 -O3 -pipe"
CXXFLAGS="${CFLAGS}"
ACCEPT_KEYWORDS="~x86"
PORTAGE_TMPDIR=/var/tmp
PORTDIR=/usr/portage
DISTDIR=${PORTDIR}/distfiles
PKGDIR=${PORTDIR}/packages
PORT_LOGDIR=/var/log/portage
SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage"
RSYNC_RETRIES="3"
RSYNC_TIMEOUT=180
MAKEOPTS="-j3"
AUTOCLEAN="yes"
FEATURES="sandbox ccache"
CCACHE_SIZE="2G"
USE="aac acl acpi activefilter adns aim amd apache2 \
audiofile avantgo bcmath bdf bidi bluetooth bzlib cap cddb cdinstall \
cdparanoia cdrom cgi chroot cjk client cpdflib ctype curl \
curlwrappers dba dio divx4linux djbfft dnd droproot dv dvb dvd dvdr \
dvdread edl eds emacs emacs-w3 erandom exif ext-png ext-zlib \
extensions faac faad fastcgi fbcon ffmpeg flash fluidsynth fmod \
freetype freewnn ftp gb gcl gd \
geoip gimp gimpprint ginac glade glgd glut gmp gmtfull gnomedb \
gnuplot gps graphviz gs gsl gtkhtml hal hardenedphp hbci hdf iconv \
icq ide idl ieee1394 image imap imlib2 innodb ipv6 irda irmc j2ee \
jabber java javacomm javadoc javamail javascript jbig jboss jce jms \
jmx joystick jp2 jpeg2k jta jython kadu-modules kadu-voice kakasi \
kerberos krb4 latex lcms leim libdsk libgd libgda libsamplerate live \
lm_sensors lua lufsusermount lzo lzw lzw-tiff m17n-lib maildir \
matroska mbox mcal mdb memlimit mhash migemo mime mixer mmap mmx mmx2 \
mng mozcalendar mozdomi mozp3p mozsvg mozxmlterm mp3 mpeg4 mpi \
mplayer msn mysql mule music nagios-dns nagios-ntp nagios-ping \
nagios-ssh native net network nis nptl ntlm nvidia nviz oav ocaml \
offensive ofx openal opie oscar ospfapi parse-clocks pcap pcntl pcre \
pdf pg-hier pg-intdatetime pg-vacuumdelay php physfs pic plotutils \
portaudio posix postgresql ppds print pthreads quotes rhino rogue \
rplay S3TC sasl servlet-2.3 servlet-2.4 session sftplogging \
shaper shared sharedmem sheep silc simplexml skk slp smime sndfile \
snmp sockets socks5 softquota sox stencil-buffer stroke svg sysvipc \
szip t1lib tga theora tidy transcode unicode usb v4l v4l2 vda vhosts \
videos virus-scan wifi wmf wxwin wxwindows X509 xatrix xchatdccserver \
xchattext xemacs xgetdefault xine xmlrpc xosd xpm xrandr xscreensaver \
xsl xvid xvmc yahoo yv12 zvbi" |
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
donjuan
l33t
Joined: 11 May 2004
Posts: 760
Location: At Uni
|
| Posted: Sun Nov 07, 2004 7:42 am Post subject: |
|
|
If you read over that bug you'll see that it mentions that what you have there is probably the source of your problems. You probably have many more use flags than you'll ever need, and they are probably also causing problems. But if you want to keep things this way, make a backup of the config, then use one without most of those use flags and the ~x86 keyword for bootstrapping.
Also, if you want to resume from where you left off, you might try using the --resume option with the script. I think it is necessary to resume bootstrapping since it crapped out there.
Edit: It seems that those guys didn't know what they were talking about with the ~x86 bootstrapping, so neither did I. So you can forget about that, but I wouldn't be surprised if your USE flags are still causing some problems. However, since I saw you filed a bug about it I'll let the developers take care of the problem.
_________________
Command-line ACCEPT_KEYWORDS is considered harmful, use the package.* files.
The Stage 1 on 3 Install |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Sun Nov 07, 2004 5:31 pm Post subject: |
|
|
| donjuan wrote: | | Also, if you want to resume from where you left off, you might try using the --resume option with the script. I think it is necessary to resume bootstrapping since it crapped out there. |
thanks. for your help. i looked over that bug report, and tried bootstrapping without "~x86" but that didn't work either.
i didn't know about the --resume parameter. THAT will come in handy.
regarding the many USE flags... its beginning to look like that may be a problem, but i was just following the recommendations in the "build like a developer" thread in Documentation Tips & Tricks. maybe that wasn't a good idea. we'll see.
thanks again!
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 4:49 pm Post subject: |
|
|
tried it again... very reproduceable results.
i have submitted this to bugzilla, but it doesn't seem to have been noticed yet. i know i'm probably just being impatient, but how long does it take for the typical bugzilla report to get noticed? i'm wondering, as some bug reports on topics similar to mine seem to have never received any response.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 5:15 pm Post subject: |
|
|
Hi to everyone.
I am having the exact same problem during the last 3 days.
I have tried bootstrapping my system more than 6 times with no success whether i was using "~x86" of just "x86" or even reducing the number of USE flags (i m folloing the same tutorial). I ve even tried to emerge libperl before bootstrpaping (this is probably not a sollution but i was desperate).
I think i am gonna give it one more try just using USE="nptl" in make.conf.
Anyway i hope someone will come with a solution soon! |
|
| Back to top |
|
|
mstyer
n00b
Joined: 29 Oct 2004
Posts: 52
Location: Philly
|
| Posted: Mon Nov 08, 2004 5:42 pm Post subject: |
|
|
| donjuan wrote: | | Edit: It seems that those guys didn't know what they were talking about with the ~x86 bootstrapping, ... |
What do you mean?  |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 6:40 pm Post subject: |
|
|
| aldo_a wrote: | I am having the exact same problem during the last 3 days.
I have tried bootstrapping my system more than 6 times with no success whether i was using "~x86" of just "x86" or even reducing the number of USE flags (i m folloing the same tutorial). I ve even tried to emerge libperl before bootstrpaping (this is probably not a sollution but i was desperate). |
if you've run the installation process using the "super duper one-liner," its not intuitive to the n00b exactly where the problem occurs. to make it obvious, i parsed out the individual commands from ali3nx's one-liner and executed each command individually. as you probably know, the error occurs as a result of the "scripts/bootstrap.sh" command. the really bad news for me is that this error occurs 6 to 6.5 hours into the execution of the bootstrap script, which makes troubleshooting very cumbersome! 
FYI i have repoorted this problem to bugzilla. the bug number is 70325 and the URL for the bug report is:
https://bugs.gentoo.org/show_bug.cgi?id=70325
you may want to take a look at the bug and see if my description is complete and accurate so that it reflects your actual experiences. you may also want to post your information to that thread, as it may help the developers.
since i posted my earlier query about how long it takes for a bug to be addressed, i have seen no responses to the bug report, but i do see that it appears to have been assigned a "P2" priority level, suggesting that the developers see this as a significant problem.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 6:55 pm Post subject: |
|
|
in response to the suggestion on the other thread that the problem could be ameliorated by issuing the command "emerge libperl && emerge libperl && emerge system":
i am uncertain as to whether this advice was given based on the assumption that the fatal error occurs as a result of "emerge system" rather than as a result of "scripts/bootstrap.sh". if the fatal error occurs during the bootstrapping script, i would think that after emerging libperl, one would have to resume the bootstrap script in order to complete the bootstrap, before proceeding to "emerge system." (but then I am a n00b and this could be a prime example of the blind leading the blind!  )
if you try the "emerge libperl" solution, watch the display. if you don't, you'll miss an error message that is generated that indicates that threading will not be fully supported as a result of emerging with the "emerge libperl" command. according to the error report, to fully support threading the following command must be used:
| Code: | | USE="ithreads" emerge libperl |
i've just completed that step, and it appears that lthe step was successful.
now comes the difficult part -- resuming the bootstrap. unfortunately, the bootstrap.sh script appears to be merciless to the user in the respect that it starts over at square 1 and repeats the entire bootstrap process from beginning to end, without regard for the fact that the vast majority of the operations were successfully completed before the fatal error occurred. for me, this means that once i restart the bootstrap script on my P3-800, i'll have to wait 6 more hours before i learn whether the workaround was successful. it sure would be a nice enhancement to the system if the bootstrap scripts were able to pick up where they left off, rather than always starting from the beginning.
so if indeed the error in the compilation of gettext -1.03 fails as a result of perl not being on the system, maybe this workaround will work. i am uncertain whether our forcibly emerged version of perl (the one using the USE="ithreads" statement) is likely to be over-written later in the bootstrapping process. if that's the case, we should probably add "ithreads" to the USE statements in our /etc/make.conf files.
advice anyone?
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks
Last edited by Bob P on Mon Nov 08, 2004 7:01 pm; edited 1 time in total |
|
| Back to top |
|
|
hielvc
Advocate
Joined: 19 Apr 2002
Posts: 2805
Location: Oceanside, Ca
|
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 7:09 pm Post subject: |
|
|
thanks for the suggestion and the link to your other post. i have to admit, as a n00b i am now on the steep portion of the learning curve. can you suggest what portions of this script need to be edited out? (i would copy the script here, but unfortunately i cannot move the data from one PC to the other.)
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
hielvc
Advocate
Joined: 19 Apr 2002
Posts: 2805
Location: Oceanside, Ca
|
| Posted: Mon Nov 08, 2004 7:24 pm Post subject: |
|
|
| Code: | einfo "Using baselayout : ${myBASELAYOUT}"
einfo "Using portage : ${myPORTAGE}"
einfo "Using os-headers : ${myOS_HEADERS}"
einfo "Using binutils : ${myBINUTILS}"
einfo "Using gcc : ${myGCC}"
[ "${myGETTEXT}" != "" ] && einfo "Using gettext : ${myGETTEXT}"
einfo "Using libc : ${myLIBC}"
einfo "Using texinfo : ${myTEXINFO}"
einfo "Using zlib : ${myZLIB}"
einfo "Using ncurses : ${myNCURSES |
So if youre failing at gettext comment out all 3, the last being the "emerge gcc".
_________________
An A-Z Index of the Linux BASH command line |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 8:08 pm Post subject: |
|
|
| Bob P wrote: |
if you've run the installation process using the "super duper one-liner," its not intuitive to the n00b exactly where the problem occurs. to make it obvious, i parsed out the individual commands from ali3nx's one-liner and executed each command individually. as you probably know, the error occurs as a result of the "scripts/bootstrap.sh" command. the really bad news for me is that this error occurs 6 to 6.5 hours into the execution of the bootstrap script, which makes troubleshooting very cumbersome!
FYI i have repoorted this problem to bugzilla. the bug number is 70325 and the URL for the bug report is:
https://bugs.gentoo.org/show_bug.cgi?id=70325
you may want to take a look at the bug and see if my description is complete and accurate so that it reflects your actual experiences. you may also want to post your information to that thread, as it may help the developers.
since i posted my earlier query about how long it takes for a bug to be addressed, i have seen no responses to the bug report, but i do see that it appears to have been assigned a "P2" priority level, suggesting that the developers see this as a significant problem. |
Hi again!
About the "super duper one-liner" i was lucky enough to give command by command from the very beggining so i found imediately that the prob was with the bootstrap script.(Its the second time i am using this tutorial - the first time before 3 weeks i had a perfectlly installed system but i wanted to do the whole procces again because of some inconsistency in ld.so - by that time i was using the old bootstrap-cascade.sh script )
I took a look at the bug you had posted and i have to agree tha we r doing the very same steps. The only difference is tha i am using :
CFLAGS="-march=athlon64 -O3 -pipe -fomit-frame-pointer" for my system
Right now i am trying to bootstrap the system again hoping to find a workaround. One thing i pointed out (probably it is not the problem) is that in the make.conf we both use MAKEOPTS="-j3" when we sould use MAKEOPTS="-j2" for one cpu system.
So now i am gonna try bootstrapping again using:
ACCEPT_KEYWORDS="x86"
MAKEOPTS="-j2"
USE="3dnow 3dnowext mmx mmx2 sse sse2 nptl"
Hopefully i would have some results in about 2 to 3 hours so as soon as i have smth i ll let u know!! |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 8:26 pm Post subject: |
|
|
| aldo_a wrote: | Right now i am trying to bootstrap the system again hoping to find a workaround. One thing i pointed out (probably it is not the problem) is that in the make.conf we both use MAKEOPTS="-j3" when we sould use MAKEOPTS="-j2" for one cpu system.
So now i am gonna try bootstrapping again using:
ACCEPT_KEYWORDS="x86"
MAKEOPTS="-j2"
USE="3dnow 3dnowext mmx mmx2 sse sse2 nptl"
Hopefully i would have some results in about 2 to 3 hours so as soon as i have smth i ll let u know!! |
fwiw, after the initial post i have already run the bootstrap with "MAKEOPTS="-j2". no improvement.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 8:29 pm Post subject: |
|
|
| aldo_a wrote: |
I took a look at the bug you had posted and i have to agree tha we r doing the very same steps. The only difference is tha i am using :
CFLAGS="-march=athlon64 -O3 -pipe -fomit-frame-pointer" for my system
|
i would post that information in the bug thread so that the devlopers know about it. they may not be reading this thread, and it would help for them to know that you've got the same problems.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 8:31 pm Post subject: |
|
|
| hielvc wrote: | | Code: | einfo "Using baselayout : ${myBASELAYOUT}"
einfo "Using portage : ${myPORTAGE}"
einfo "Using os-headers : ${myOS_HEADERS}"
einfo "Using binutils : ${myBINUTILS}"
einfo "Using gcc : ${myGCC}"
[ "${myGETTEXT}" != "" ] && einfo "Using gettext : ${myGETTEXT}"
einfo "Using libc : ${myLIBC}"
einfo "Using texinfo : ${myTEXINFO}"
einfo "Using zlib : ${myZLIB}"
einfo "Using ncurses : ${myNCURSES |
So if youre failing at gettext comment out all 3, the last being the "emerge gcc". |
i think you've lost me here. that code above doesn't look like the code that was commented out in the hyperlinked text:
| Code: | [# disable collision-protection
export FEATURES="${FEATURES} -collision-protect"
.>>>>>>>> PUT A # IN FRONT OF THE LOWER STATMENT<<<<<<<<<
USE="-* build bootstrap ${STAGE1_USE}" emerge ${STRAP_EMERGE_OPTS} ${myPORTAGE} || cleanup 1
echo -------------------------------------------------------------------------------
# We can't unmerge headers which may or may not exist yet. If your
# trying to use nptl, it may be needed to flush out any old headers
# before fully bootstrapping.
#emerge ${STRAP_EMERGE_OPTS} -C virtual/os-headers || cleanup 1
>>>>>>>>>>>AND HERE<<<<<<<<<<<<<<<<<<<<
emerge ${STRAP_EMERGE_OPTS} ${myOS_HEADERS} ${myTEXINFO} ${myGETTEXT} ${myBINUTILS} || cleanup 1
>>>>>>>>>>>>AND HERE<<<<<<<<<<<<<<<<<<<<<<<
emerge ${STRAP_EMERGE_OPTS} ${myGCC} || cleanup 1
echo ------------------------------------------------------------------------------- |
in the second code sample, are you referring to doing this?
| Code: | #USE="-* build bootstrap ${STAGE1_USE}" emerge ${STRAP_EMERGE_OPTS} ${myPORTAGE} || cleanup 1
echo -------------------------------------------------------------------------------
|
i think that i'm wading into a very dangerous area. as a n00b left to my own devices in trying to find a workaround for this bug, i think i'm likely to make some changes that could end up doing alot more harm than good... 
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 9:47 pm Post subject: |
|
|
Have good news i succeded bootstraping the system with:
ACCEPT_KEYWORDS="x86"
MAKEOPTS="-j2"
USE="3dnow 3dnowext mmx mmx2 sse sse2 nptl"
Though bootstrping finished quite quickly i had a alook at /var/log/emerge.log and it ends with "...exiting succesfuly" so that must be it!!
As i saw "gettext-0.12.1-r2" is installed but Locale-gettext isnt so i am sure that when i ll try to emerge system i ll fail again but that may be solved by "emerge libperl && emerge libperl && emerge system"
I ll try out and i ll keep in touch if u have any questions ask 4 the next few hours i ll be online.. |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 9:56 pm Post subject: |
|
|
Just a tip
"3dnow 3dnowext mmx mmx2 sse sse2" are supported by amd64 and i am not sure which of them will be supported by your cpu (3dnow 3dnowext as far as i know are not)
So u might consider using just USE="nptl" for bootstraping |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 10:41 pm Post subject: |
|
|
Just one more observation :
I am ready to start "emerge system"
I replaced in make.conf the USE="3dnow 3dnowext mmx mmx2 sse sse2" which i used during bootstraping with
USE="aac acl acpi activefilter adns aim amd apache2 \
audiofile avantgo bcmath bdf bidi bluetooth bzlib cap cddb cdinstall \
cdparanoia cdrom cgi chroot cjk client cpdflib ctype curl \
curlwrappers dba dio divx4linux djbfft dnd droproot dv dvb dvd dvdr \
dvdread edl eds emacs emacs-w3 erandom exif ext-png ext-zlib \
extensions faac faad fastcgi fbcon ffmpeg flash fluidsynth fmod \
freetype freewnn ftp gb gcl gd \
geoip gimp gimpprint ginac glade glgd glut gmp gmtfull gnomedb \
gnuplot gps graphviz gs gsl gtkhtml hal hardenedphp hbci hdf iconv \
icq ide idl ieee1394 image imap imlib2 innodb ipv6 irda irmc j2ee \
jabber java javacomm javadoc javamail javascript jbig jboss jce jms \
jmx joystick jp2 jpeg2k jta jython kadu-modules kadu-voice kakasi \
kerberos krb4 latex lcms leim libdsk libgd libgda libsamplerate live \
lm_sensors lua lufsusermount lzo lzw lzw-tiff m17n-lib maildir \
matroska mbox mcal mdb memlimit mhash migemo mime mixer mmap mmx mmx2 \
mng mozcalendar mozdomi mozp3p mozsvg mozxmlterm mp3 mpeg4 mpi \
mplayer msn mysql mule music nagios-dns nagios-ntp nagios-ping \
nagios-ssh native net network nis nptl ntlm nvidia nviz oav ocaml \
offensive ofx openal opie oscar ospfapi parse-clocks pcap pcntl pcre \
pdf pg-hier pg-intdatetime pg-vacuumdelay php physfs pic plotutils \
portaudio posix postgresql ppds print pthreads quotes rhino rogue \
rplay S3TC sasl servlet-2.3 servlet-2.4 session sftplogging \
shaper shared sharedmem sheep silc simplexml skk slp smime sndfile \
snmp sockets socks5 softquota sox stencil-buffer stroke svg sysvipc \
szip t1lib tga theora tidy transcode unicode usb v4l v4l2 vda vhosts \
videos virus-scan wifi wmf wxwin wxwindows X509 xatrix xchatdccserver \
xchattext xemacs xgetdefault xine xmlrpc xosd xpm xrandr xscreensaver \
xsl xvid xvmc yahoo yv12 zvbi"
they exist in make.conf from ali3nx.
If i leave ACCEPT_KEYWORDS="x86" (which i used during boostrap) and do an "emerge system --pretend" everything seems to be fine...libperl is followed by perl... and Locale-gettext is nowhere. even if add in USE="" the "nls" flag Locale-gettext is still nowhere.
If i set ACCEPT_KEYWORDS="~x86" then the problem appers because Locate-gettext is quite before libperl and perl in the installtion order. Even if i try to add in Locale-gettext ebuild DEPEND="sys-devel/libperl dev-lang/perl" perl is still under Locale-gettext - libperl goes just one possition above Locale..
So i am just wondering if there is some problem withe the dependencies of the packages that precede Locale-gettext in the case where :
_ACCEPT_KEYWORDS="~x86" is used
_no modifications are made to the ebuild of any file |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 10:43 pm Post subject: |
|
|
| aldo_a wrote: | | So u might consider using just USE="nptl" for bootstraping |
one nagging question remains -- how much "optimization" and compatability am i giving up if i follow that course and throw away 99% of the USE statements? does that essentially make the system more USE-less?
sure, if you eliminate everything from the USE statements, the bootstrap time requirement will go way down. but aren't we trading off quite a bit of system functinality and performance by not building much of anything into the base layout?
i'm willing to spend the time on bootstrapping, as long as it works. its the redundancy of performing the same operation over and over, yet still not having it work because of a persistent error, that is growing old. a little expertise on how to proceed from the true experts would be greatly appreciated. 
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 10:49 pm Post subject: |
|
|
The packages that precede Locale-gettext for my USE configuration are:
app/shells/bash-3.0-r7
sys-apps/gawk-3.1.3-r1
sys-apps/sysvinit-2.85-r1
dev-python-fchksum-1.7.1
sys-libs/readline-5.0-r1
sys-libs/db-1.85-r2
sys-libs/gdbm-1.8.3-r1
sys-devel/flex-2.5.4a-r5
sys-devel/bc-1.06-r5
I dont know if i am wright but i ll check the above ebuilds and i ll post again |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 10:54 pm Post subject: |
|
|
| aldo_a wrote: | If i leave ACCEPT_KEYWORDS="x86" (which i used during boostrap) and do an "emerge system --pretend" everything seems to be fine...libperl is followed by perl... and Locale-gettext is nowhere. even if add in USE="" the "nls" flag Locale-gettext is still nowhere.
If i set ACCEPT_KEYWORDS="~x86" then the problem appers because Locate-gettext is quite before libperl and perl in the installtion order. Even if i try to add in Locale-gettext ebuild DEPEND="sys-devel/libperl dev-lang/perl" perl is still under Locale-gettext - libperl goes just one possition above Locale..
So i am just wondering if there is some problem withe the dependencies of the packages that precede Locale-gettext in the case where :
_ACCEPT_KEYWORDS="~x86" is used
_no modifications are made to the ebuild of any file |
since you've made it through the bootstrap, i think that you could quite easily avoid the "gettext dependency on perl during emerge system" issue by simply emerging perl before you perform the emerge system. that may solve the problem even if you do use the "`x86" keyword.
if i were as far as you are in the installation, i would emerge libperl and then repeat your emerge system pretends to see if you get the same results with the "~x86" keyword as you do with "x86".
WARNING: The Blind are Leading the Blind.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
aldo_a
n00b
Joined: 08 Nov 2004
Posts: 37
Location: I am out there!
|
| Posted: Mon Nov 08, 2004 11:10 pm Post subject: |
|
|
| Bob P wrote: |
WARNING: The Blind are Leading the Blind. |
Lol u r right!!!!
I ve just got obsessed to understand where the err is. Thats why i dont want to emerge libperl and perl yet.
What i saw is that using ~x86 installs sys-apps/gawk-3.1.3-r1 in the second possition whether using x86 does it quite a few possitions above.
So the prob might be with the bash's dependencies.
~x86 installs app/shells/bash-3.0-r7
x86 installs app/shells/bash-2.05b-r9 |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Nov 08, 2004 11:13 pm Post subject: |
|
|
| aldo_a wrote: | | Bob P wrote: |
WARNING: The Blind are Leading the Blind. |
Lol u r right!!!! |
i'm still hoping that we get a knowledgeable person to come along and help us... but i'd settle for a seeing-eye dog.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
|
Installing Gentoo
