List Info

Thread: Verbose commands




Verbose commands
user name
2006-05-28 00:11:18
The following constructs will work. This is how it will need
to be 
rendered in the Makefile files. Not sure where to make these
changes in 
the source code.

Example of chapter 6's linux-libc-headers

Old commands were rendered as follows:

(
time { $(CHROOT1) 'cd /jhalfs && source envars
&&
/jhalfs/lfs-commands/chapter06/064-linux-libc-headers
 >>/jhalfs/logs/064-linux-libc-headers 2>&1' ;
} )
2>>logs/064-linux-libc-headers && \

Using 'tee' we can have output on screen as well as in log
files:

(
time { $(CHROOT1) 'cd /jhalfs && source envars
&&
/jhalfs/lfs-commands/chapter06/064-linux-libc-headers | tee
-a
/jhalfs/logs/064-linux-libc-headers 2>&1' ; } ) |
tee -a
logs/064-linux-libc-headers 2>&1 && \

I'll go see if I can find the right locations in the source
to make 
these changes. Will send a patch in case it's deemed an
acceptable way 
of doing things.


-- 
Gerard Beekmans

/* If Linux doesn't have the solution, you have the wrong
problem */

-- 
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
Verbose commands
user name
2006-05-28 01:11:19
There is a side effect that configure script output ends up
a little 
uglier in that the configure script writes info to its own
logfile that 
normally does not end up on stdout. That seems to be echoed
to the 
screen and I'm not sure that that information ends up in
config.log anymore.

That'll have to be verified if this approach is going to be
used. Or 
something similar, changing output redirection in any way
might 
adversely affect it.

Or maybe this has nothing to do with the changes and this
has always 
been going on, but since jhalfs normally lacks output, this
may never 
have been caught before?

-- 
Gerard Beekmans

/* If Linux doesn't have the solution, you have the wrong
problem */

-- 
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
Verbose commands
user name
2006-05-31 18:06:29
El Domingo, 28 de Mayo de 2006 03:11, Gerard Beekmans
escribió:

> Or maybe this has nothing to do with the changes and
this has always
> been going on, but since jhalfs normally lacks output,
this may never
> have been caught before?

Looks like the output mess is due tee, but config.log seems
to be well created 
in all cases.

Thinking on the verbose output, I think that to have the
real output 
redirected to the screen isn't actually required or useful
except to know 
that the build system is alive and doing something, true?

If that is the case, then maybe some type of progression bar
(like a dot 
echoed to the screen each second while make in building a
package) could be 
enough and could be set on by default.

I'm now playing with that...

-- 
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.com
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
Verbose commands
user name
2006-05-31 20:45:44
El Miércoles, 31 de Mayo de 2006 20:06, M.Canales.es
escribió:

> If that is the case, then maybe some type of
progression bar (like a dot
> echoed to the screen each second while make in building
a package) could be
> enough and could be set on by default.

A POC about that has been added to trunk in r2734.

On all my test it works fine, but I'm not running other
make process in 
parallel, then not sure if it's bugs free.


-- 
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.com
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
Verbose commands
user name
2006-05-31 21:41:49
M.Canales.es wrote:
> Thinking on the verbose output, I think that to have
the real output 
> redirected to the screen isn't actually required or
useful except to know 
> that the build system is alive and doing something,
true?

For me personally, yes, you are right about that. It's just
something to 
look at while it's building. I could tail the log file but
you can't 
tail files that don't exist yet unless you write a script
of some sort 
that loops and restarts tail when a new file is present.
Having default 
output is just easier for me is all.

And as always, I can easily modify the master scripts to add
'tee' for 
myself if this doesn't end up in the main code. As usual I
just share 
what I'd like to see and if the default isn't going to do
that, I just 
create a patch. No big deal. 

-- 
Gerard Beekmans

/* If Linux doesn't have the solution, you have the wrong
problem */

-- 
http://linuxfromscratch.org/mailman/listinfo/alfs-discu
ss
FAQ: http://www.linux
fromscratch.org/faq/
Unsubscribe: See the above information page
Verbose commands
user name
2006-06-01 20:24:38
El Miércoles, 31 de Mayo de 2006 22:45, M.Canales.es
escribió:

> On all my test it works fine, but I'm not running
other make process in
> parallel, then not sure if it's bugs free.

If another make is launch wile the build make is running
(for example, to 
render the LFS book) the current progress_bar.sh code will
inherit the PID 
from that new make process :-/

A better test need be done to be sure that we are tracking
the proper make 
process.

-- 
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.com
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
Verbose commands
user name
2006-06-01 21:51:35
El Jueves, 1 de Junio de 2006 22:24, M.Canales.es escribió:

> A better test need be done to be sure that we are
tracking the proper make
> process.

Looks like the bugs was fixed and now progress_bar.sh should
to work as 
expected.

-- 
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.com
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
[1-7]

about | contact  Other archives ( Real Estate discussion Medical topics )