|
List Info
Thread: error in step 086 clfs automated
|
|
| error in step 086 clfs automated |

|
2007-08-14 14:41:04 |
hi,
in my mini-ppc (512MB ram) booting from livecd-ppc
6.2.pre
I have download jhalfs-2.3 and last package-set
then I configure CLFS and run MAKE
... but in step 86 error:
Building target 086-changingowner
[/tools/bin/bash: line 3: perl: command not found ] 0
min 0 sec
make: *** [086-changingowner] Error 127
make: *** [mk_SYSTOOLS] Error 2
whatching logs I found 086 log that has a lot of line:
ownwership of '/cross-tools/.......' retained as 0:0
can you suggest me somethings ?
thankyou
sacarde tiscali.it
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  Spain |
2007-08-14 15:27:02 |
El Martes, 14 de Agosto de 2007 21:41, sacarde escribió:
> hi,
> in my mini-ppc (512MB ram) booting from livecd-ppc
6.2.pre
> I have download jhalfs-2.3 and last package-set
> then I configure CLFS and run MAKE
>
> ... but in step 86 error:
>
> Building target 086-changingowner
> [/tools/bin/bash: line 3: perl: command not found ]
0 min 0 sec
I have not build CLFS in several time, but revising the book
and build scripts
looks like CLFS don't install Perl in the cross-tools nor
tools phase, thus
it is not available in the firsts targets run into the
chroot :-/
George, have you build some CLFS system after the chage to
use perl for build
time meassuring?
As a quick fix, you can edit the $BUILD_DIR/jhalfs/Makefile,
go to the
086-changingowner target entry, and change the line
end=$(TIME_MARK) && $(BUILD_TIME)
>>logs/$ &&
to
end=$(TIME_MARK) >>logs/$ &&
Do the same in the following targets up to the one for
temp-perl inclusive.
>
> whatching logs I found 086 log that has a lot of line:
>
> ownwership of '/cross-tools/.......' retained as 0:0
>
That's not good, that might meant that the croos-tools
packages was build as
the root user. Revising the Makefile I can't see how that
could happened,
except if the $LUSER variable, found at the beggining of the
Makefile, is
empty.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfroms
cratch.org
LFS en castellano: http://www.escomp
oslinux.org/lfs-es http://www.lfs-es.info
TLDP-ES: http://es.tldp.org
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  United States |
2007-08-14 16:01:34 |
M.Canales.es wrote:
> El Martes, 14 de Agosto de 2007 21:41, sacarde
escribió:
>> hi,
>> in my mini-ppc (512MB ram) booting from
livecd-ppc 6.2.pre
>> I have download jhalfs-2.3 and last package-set
>> then I configure CLFS and run MAKE
>>
>> ... but in step 86 error:
>>
>> Building target 086-changingowner
>> [/tools/bin/bash: line 3: perl: command not found
] 0 min 0 sec
>
>
> I have not build CLFS in several time, but revising the
book and build scripts
> looks like CLFS don't install Perl in the cross-tools
nor tools phase, thus
> it is not available in the firsts targets run into the
chroot :-/
>
> George, have you build some CLFS system after the chage
to use perl for build
> time meassuring?
No, I have not run testbuilds against CLFS-1.0 since the
change. I
had been working on CLFS-Embedded lately. The Perl timing
issue will not
affect CLFS-sysroot or CLFS-Embedded.
A cheaper workaround for Sacarde would be to disable SBU
calculations.
>
> As a quick fix, you can edit the
$BUILD_DIR/jhalfs/Makefile, go to the
> 086-changingowner target entry, and change the line
>
> end=$(TIME_MARK) && $(BUILD_TIME)
>>logs/$ &&
>
> to
>
> end=$(TIME_MARK) >>logs/$ &&
>
> Do the same in the following targets up to the one for
temp-perl inclusive.
>
>> whatching logs I found 086 log that has a lot of
line:
>>
>> ownwership of '/cross-tools/.......' retained as
0:0
>>
>
> That's not good, that might meant that the croos-tools
packages was build as
> the root user. Revising the Makefile I can't see how
that could happened,
> except if the $LUSER variable, found at the beggining
of the Makefile, is
> empty.
>
Both chroot and boot isolate the final build from the
host. Either we
go back to the bash version or we skip timing of the
temporary Perl
package. (I dislike coding exceptions.) A third option is to
revisit the
timing method.. write a neutral data package
(ticks/milliseconds or what
ever) at the end of the log and do all the processing at the
end of the
build.
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  Spain |
2007-08-14 16:29:15 |
El Martes, 14 de Agosto de 2007 23:01, George Boudreau
escribió:
> No, I have not run testbuilds against CLFS-1.0 since
the change. I
> had been working on CLFS-Embedded lately. The Perl
timing issue will not
> affect CLFS-sysroot or CLFS-Embedded.
Great.
HLFS need be revised also before jhalfs-2.3.1 release, due
the last book
changes. I will try to test it this weekend.
> A cheaper workaround for Sacarde would be to disable
SBU calculations.
That don't help, the time stamp is always added to the
logs.
> Both chroot and boot isolate the final build from
the host. Either we
> go back to the bash version or we skip timing of the
temporary Perl
> package. (I dislike coding exceptions.) A third option
is to revisit the
> timing method.. write a neutral data package
(ticks/milliseconds or what
> ever) at the end of the log and do all the processing
at the end of the
> build.
Well, we need to find a solution that will work with all
books, but at this
momment I can't figure what.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfroms
cratch.org
LFS en castellano: http://www.escomp
oslinux.org/lfs-es http://www.lfs-es.info
TLDP-ES: http://es.tldp.org
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  Spain |
2007-08-14 17:17:02 |
El Miércoles, 15 de Agosto de 2007 00:02, George Boudreau
escribió:
> I will see what I can come up with. It may take some
time as I am
> currently beating on x86_64 wireless problems.
Actually, the most simple solution could be to put
echo "Totalseconds: $SECONDS"
at the end of each build script. That is not very precise on
slow or
overburdened machines, of course, but is simple and don't
relies on external
binaries or built-ins. Tested on a BLFS script it returm the
same time that
is displayed by progress-bar.sh but in seconds.
What do you think?
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfroms
cratch.org
LFS en castellano: http://www.escomp
oslinux.org/lfs-es http://www.lfs-es.info
TLDP-ES: http://es.tldp.org
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  Spain |
2007-08-14 17:36:30 |
El Miércoles, 15 de Agosto de 2007 00:28, George Boudreau
escribió:
> This should provide enough accuracy. My
understanding of the usage of
> SBU was to give builders an approximate build time for
a package. If a
> machine is overloaded then the build will take longer.
I will do a LFS build adding the echo but keeping the
current method to can
compare both results.
If the desviation is not very excesive, or there is the same
desviation on all
packages that could be corrected via create-sbu_du-report.sh
script, we
should be able to do the change.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfroms
cratch.org
LFS en castellano: http://www.escomp
oslinux.org/lfs-es http://www.lfs-es.info
TLDP-ES: http://es.tldp.org
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |

|
2007-08-14 17:57:06 |
On 8/14/07, M.Canales.es <macana macana-es.com> wrote:
> El Miércoles, 15 de Agosto de 2007 00:28, George
Boudreau escribió:
>
> > This should provide enough accuracy. My
understanding of the usage of
> > SBU was to give builders an approximate build time
for a package. If a
> > machine is overloaded then the build will take
longer.
>
> I will do a LFS build adding the echo but keeping the
current method to can
> compare both results.
>
> If the desviation is not very excesive, or there is the
same desviation on all
> packages that could be corrected via
create-sbu_du-report.sh script, we
> should be able to do the change.
I don't know if this is just as silly of a solution, but you
can use
awk instead of perl, and it looks like CLFS-1.0 builds a
cross awk.
$ start=`date +%s.%N`
$ end=`date +%s.%N`
$ echo $start $end | awk '{printf "%.2f
secondsn", $2 - $1}'
3.52 seconds
No escaping, either.
--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  Spain |
2007-08-14 18:15:21 |
El Miércoles, 15 de Agosto de 2007 00:57, Dan Nicholson
escribió:
>
> I don't know if this is just as silly of a solution,
but you can use
> awk instead of perl, and it looks like CLFS-1.0 builds
a cross awk.
>
> $ start=`date +%s.%N`
> $ end=`date +%s.%N`
> $ echo $start $end | awk '{printf "%.2f
secondsn", $2 - $1}'
> 3.52 seconds
>
> No escaping, either.
Yeah, that is another option if the $SECONDS trick offers
not good values.
Thanks.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfroms
cratch.org
LFS en castellano: http://www.escomp
oslinux.org/lfs-es http://www.lfs-es.info
TLDP-ES: http://es.tldp.org
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
| Re: error in step 086 clfs automated |
  Spain |
2007-08-17 11:31:26 |
El Miércoles, 15 de Agosto de 2007 01:15, M.Canales.es
escribió:
> Yeah, that is another option if the $SECONDS trick
offers not good values.
> Thanks.
I finally opted by the $SECONDS solution. I my test the
results are just the
same than using the previous approach, plus have the
advantages that don't
rely on external binaries and allow to remove two lines on
each Makefile's
target entry.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfroms
cratch.org
LFS en castellano: http://www.escomp
oslinux.org/lfs-es http://www.lfs-es.info
TLDP-ES: http://es.tldp.org
--
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
|
|
[1-9]
|
|