devil-linux-develop — developer discussions

Re: [Devil-linux-develop] new testing release

[Diego T. <dt...@co...>]

On Sun, Sep 21, 2003 at 03:58:56PM -0400, Heiko Zuerker wrote:

> Hey, somewhere else you would have to pay for having fun on Sundays. 
> With DL you get it all for free. ;-)

I've been downloading a full fresh dl development kit :) (cvs + packages) 
following the instructions on the ./update_src script. That's, first 
download from a mirror and then update from the main ftp server. When doing
the second update_src, all the files have been downloaded again with the
following msg:

Borrando archivo antiguo `cvs-1.11.6.tar.bz2'
Enviando archivo `cvs-1.11.6.tar.bz2'

(that in english means something like "deleting old archive cvs..." and
"sending file cvs"...

surely this isn't the desired behaviour, but i don't know how to fix it :?

-- 
-- gnupg keyfingerprint -- 48AF 5BF9 8F54 2966 64CC  2327 7CD0 DD91 B09D 5799
-- Use of a keyboard or mouse may be linked to serious injuries or disorders.  
Diego Torres - dt...@co... - Madrid / España

Re: [Devil-linux-develop] $PMAKE

[Bruce S. <bw...@ar...>]

> > BTW, the last time I tried setting the number of parallel build jobs to
> > something greater than one (I set it to two), it aborted compiling while
> > bash.  I changed it back to one and it compiled fine.  FWIW ...
> 
> I enabled it for a bunch of packages and currently have a test compile 
> running with distcc and -j4.
> 
> Some programs have problems with parallel builds (buggy Makefile), then 
> we just have to disable it. In such a case it's probably best to write a 
> comment above the "make ...." line, so nobody enables it again.

OK, I'll try it again (sometime soon).  If anything aborts, I'll remove
the $PMAKE and place a comment in the script.

> > That was a few weeks ago, so maybe it works now.  I'll try it again
> > sometime since I now have a hyperthreaded P4 at home.  :-)
> 
> You got too much money. ;-)

Yeah, I've got so much money, I can't count it all!  ;-)

 - BS

Re: [Devil-linux-develop] new testing release

[Heiko Z. <he...@zu...>]

Dean Nedelman wrote:
>>Hey,
>>
>>I'm currently uploading a new testing release.
>>It should be finished within 1 hour.
> 
> 
> There goes my Sunday!

Hey, somewhere else you would have to pay for having fun on Sundays. 
With DL you get it all for free. ;-)

cya
   Heiko

RE: [Devil-linux-develop] new testing release

[Dean N. <di...@ti...>]

>Hey,
>
>I'm currently uploading a new testing release.
>It should be finished within 1 hour.

There goes my Sunday!

[Devil-linux-develop] new testing release

[Heiko Z. <he...@zu...>]

Hey,

I'm currently uploading a new testing release.
It should be finished within 1 hour.

cya
   Heiko

Re: [Devil-linux-develop] usb boot

[Heiko Z. <he...@zu...>]

Bruce Smith wrote:
>>>>>Do you want me to install grub on DL, and write a install-on-usb
>>>>>script using GRUB with a boot menu and a horned-Tux background?
>>>>>     
>>>>>
>>>>
>>>>Sure! Go ahead, would look much better !
> 
> 
> Have you tried it yet?  How's it look?
> 
> I want to make sure the script runs on a system without grub installed.
> (use the script I uploaded today)

It did not work, I had to execute it within the lfs system.
You need to build grub as a static binary, otherwise you get problems 
with the different glibc versions.

cya
   Heiko

Re: [Devil-linux-develop] $PMAKE

[Heiko Z. <he...@zu...>]

Bruce Smith wrote:
>>This will enable parallell builds for that specific package.
> 
> 
> BTW, the last time I tried setting the number of parallel build jobs to
> something greater than one (I set it to two), it aborted compiling while
> bash.  I changed it back to one and it compiled fine.  FWIW ...

I enabled it for a bunch of packages and currently have a test compile 
running with distcc and -j4.

Some programs have problems with parallel builds (buggy Makefile), then 
we just have to disable it. In such a case it's probably best to write a 
comment above the "make ...." line, so nobody enables it again.


> That was a few weeks ago, so maybe it works now.  I'll try it again
> sometime since I now have a hyperthreaded P4 at home.  :-)
>

You got too much money. ;-)

cya
   Heiko

[Devil-linux-develop] smartcards

[Heiko Z. <he...@zu...>]

Hey,

did anybody of you play with Smartcards under Linux ?
http://www.opensc.org/files/ANNOUNCE-opensc

I think it would be pretty cool and usefull if we support it.

cya
   Heiko

RE: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Dean N. <di...@ti...>]

>>
>> LVM2 was selected over EVMS to replace LVM for the 2.6 kernel.  It
probably
>> shouldn't be considered stable at this point.  In fact, they are still
>> trying to trace down Raid induced file corruption in the 2.6 kernel
series.
>
>i don't think so. in 2.6 there is a device_mapper infraestructure, that
enables
>you to use evms or lvm2 from userspace. There are 3 possibilities, md (the
>usual raid stuff), dm (devicemapper) -> lvm2 || evms.

There were two competing designs - lvm2 and evms.  LVM2 used a device mapper
(dm) to manage its core functionality and evms used a much more kernel
intrusive approach.  In the end it was decided to go with LVM2, at which
point the EVMS group took a step back and completely rewrote their system so
that it could be run on top of LVM2's device mapper system.  But the entire
device_mapper infrastructure -is- LVM2's.

Dean Nedelman
TimeLord Consulting

Re: [Devil-linux-develop] $PMAKE

[Heiko Z. <he...@zu...>]

Bruce Smith wrote:
>>I just fixed a bunch of scripts.
>>When you add a new package to DL, please use
>>make $PMAKE || exit 1
>>
>>This will enable parallell builds for that specific package.
> 
> 
> OK.  I always wondered what that "PMAKE" variable was for.  :-)

If you don't specify parallel builds > 1 then it is just empty.
If you do, it's "-j <number>", i.e. "-j 4"

Fairly simple, but effective.

cya
   Heiko

Re: [Devil-linux-develop] $PMAKE

[Bruce S. <bw...@ar...>]

> This will enable parallell builds for that specific package.

BTW, the last time I tried setting the number of parallel build jobs to
something greater than one (I set it to two), it aborted compiling while
bash.  I changed it back to one and it compiled fine.  FWIW ...

That was a few weeks ago, so maybe it works now.  I'll try it again
sometime since I now have a hyperthreaded P4 at home.  :-)

 - BS

Re: [Devil-linux-develop] $PMAKE

[Bruce S. <bw...@ar...>]

> I just fixed a bunch of scripts.
> When you add a new package to DL, please use
> make $PMAKE || exit 1
> 
> This will enable parallell builds for that specific package.

OK.  I always wondered what that "PMAKE" variable was for.  :-)

 - BS

[Devil-linux-develop] $PMAKE

[Heiko Z. <he...@zu...>]

Hey,

I just fixed a bunch of scripts.
When you add a new package to DL, please use
make $PMAKE || exit 1

This will enable parallell builds for that specific package.

cya
   Heiko

Re: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Diego T. <dt...@co...>]

On Sat, Sep 20, 2003 at 08:38:27PM -0400, Heiko Zuerker wrote:

> >I like that idea !
> 
> The I2C and LM_SENSORS stuff is in CVS now.
> I have to start a clean compile and will provide a new testing release 
> tomorrow when everything works.
> Then you guys can work on the console-status stuff.

i'm going to use it (at least try it) on a vmware machine under win2k. 
hope it works :)

-- 
-- gnupg keyfingerprint -- 48AF 5BF9 8F54 2966 64CC  2327 7CD0 DD91 B09D 5799
-- Use of a keyboard or mouse may be linked to serious injuries or disorders.  
Diego Torres - dt...@co... - Madrid / España

Re: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Diego T. <dt...@co...>]

On Sat, Sep 20, 2003 at 02:02:26PM -0700, Dean Nedelman wrote:
> >>  ftp://ftp.sistina.com/pub/LVM2/device-mapper/device-mapper-latest.tgz
> >>  kernel patches for 2.4.21.
> >
> >Should we switch to LVM2 ? How stable is it?
> 
> LVM2 was selected over EVMS to replace LVM for the 2.6 kernel.  It probably
> shouldn't be considered stable at this point.  In fact, they are still
> trying to trace down Raid induced file corruption in the 2.6 kernel series.

i don't think so. in 2.6 there is a device_mapper infraestructure, that enables
you to use evms or lvm2 from userspace. There are 3 possibilities, md (the
usual raid stuff), dm (devicemapper) -> lvm2 || evms.

-- 
-- gnupg keyfingerprint -- 48AF 5BF9 8F54 2966 64CC  2327 7CD0 DD91 B09D 5799
-- Use of a keyboard or mouse may be linked to serious injuries or disorders.  
Diego Torres - dt...@co... - Madrid / España

Re: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Diego T. <dt...@co...>]

On Sat, Sep 20, 2003 at 10:15:51AM -0400, Heiko Zuerker wrote:
> > ftp://ftp.sistina.com/pub/LVM2/tools/LVM2.0-latest.tgz
> > user space tools
> > 
> > ftp://ftp.sistina.com/pub/LVM2/device-mapper/device-mapper-latest.tgz
> > kernel patches for 2.4.21.
> 
> Should we switch to LVM2 ? How stable is it?

i've been using it for 1'5 years. never had a problem, but my configuration 
is pretty simple, 3 hds on a raid0 (stripped) configuration.

> >http://sourceforge.net/projects/acpi
> > useful with newer or modern pcs, mainly because 2.4.21 is obsolete in 
> > acpi terms...
> 
> True. Add some point we have to support it, the question is do we want 
> to do it in the next release.

maybe better if we wait to the next release, acpi is a strange thing :)

-- 
-- gnupg keyfingerprint -- 48AF 5BF9 8F54 2966 64CC  2327 7CD0 DD91 B09D 5799
-- Use of a keyboard or mouse may be linked to serious injuries or disorders.  
Diego Torres - dt...@co... - Madrid / España

Re: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Heiko Z. <he...@zu...>]

Heiko Zuerker wrote:
> Dean Nedelman wrote:
> 
>>> Diego Torres wrote:
>>>
>>>> this patches could be useful (at least, there are *very* useful for me)
>>>>
>>>> http://secure.netroedge.com/~lm78/index.html
>>>>
>>>> http://secure.netroedge.com/~lm78/archive/i2c-2.8.0.tar.gz
>>>> patches for i2c/lmsensors. don't work for kernels 2.5/2.6. it gives 
>>>> lots
>>
>>
>> of drivers for hardware sensors (cpu temp, cpu fan speed, etc...) 
>> available
>> on today motherboards.
>>
>>>> http://secure.netroedge.com/~lm78/archive/lm_sensors-2.8.0.tar.gz
>>>> userspace applications for installing & configuring the sensors.
>>>
>>>
>>> I would vote for including those.
>>
>>
>>
>> Yes, -IF- we decided to run a monitor on one of the virtual consoles 
>> (maybe
>> #9?), like we today with procinfo.
> 
> 
> I like that idea !

The I2C and LM_SENSORS stuff is in CVS now.
I have to start a clean compile and will provide a new testing release 
tomorrow when everything works.
Then you guys can work on the console-status stuff.

cya
   Heiko

Re: [Devil-linux-develop] clamav

[Heiko Z. <he...@zu...>]

Bruce Smith wrote:
>>>>where should we place the clamav virus signatures?
>>>
>>>Where is the default location for them?
>>
>>/usr/share/clamav
>>But you can easily set this in the config file.
>>
>>
>>>>The virus dbs could get to big to fit in /etc (for the floppy save), but 
>>>>somewhere else it would always be outdated when a system boots.
>>>
>>>How big are they now?
>>
>>linux:/lfssystem/data # l virus*
>>-rw-r--r--    1 root     root       140816 Sep  2 19:26 viruses.db
>>-rw-r--r--    1 root     root        32940 Sep  2 19:26 viruses.db2
>>
>>
>>>>We could add an init script, which fetches the newest files if the user 
>>>>wants to.
>>>
>>>That sounds good.
>>>
>>>
>>>>I personally would place the signatures somewhere under /var.
>>>
>>>Isn't /etc/ & /var/ our only choices for directories not on the CD?
>>
>>Actually yes.
> 
> 
> They seem to me to be more of a data file (than a config file - /etc)
> so, I'd put them in /var. 

agreed

> A script, or cron job, that grabs the new sigs would be a good idea too.
> 
> My current virus scanner (RAV) on my employer's mail gateway comes 
> with a cron job that checks for, and downloads, new sigs ever hour.

I intend to provide the same.
I have postfix + sagator + spamassassin + clamav running, it works just 
fine. The only thing I need are the virus sig updates.

cya
   Heiko

[Devil-linux-develop] [ devil-linux-Bugs-809968 ] clamav virus sig update script

[SourceForge.net <no...@so...>]

Bugs item #809968, was opened at 2003-09-20 19:13
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=410643&aid=809968&group_id=34096

Category: Base System
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Heiko Zuerker (smiley73)
Assigned to: Nobody/Anonymous (nobody)
Summary: clamav virus sig update script

Initial Comment:
for init and cron jobs

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=410643&aid=809968&group_id=34096

Re: [Devil-linux-develop] clamav

[Bruce S. <bw...@ar...>]

> >>where should we place the clamav virus signatures?
> > 
> > Where is the default location for them?
> 
> /usr/share/clamav
> But you can easily set this in the config file.
> 
> >>The virus dbs could get to big to fit in /etc (for the floppy save), but 
> >>somewhere else it would always be outdated when a system boots.
> > 
> > How big are they now?
> 
> linux:/lfssystem/data # l virus*
> -rw-r--r--    1 root     root       140816 Sep  2 19:26 viruses.db
> -rw-r--r--    1 root     root        32940 Sep  2 19:26 viruses.db2
> 
> >>We could add an init script, which fetches the newest files if the user 
> >>wants to.
> > 
> > That sounds good.
> > 
> >>I personally would place the signatures somewhere under /var.
> > 
> > Isn't /etc/ & /var/ our only choices for directories not on the CD?
> 
> Actually yes.

They seem to me to be more of a data file (than a config file - /etc)
so, I'd put them in /var. 

A script, or cron job, that grabs the new sigs would be a good idea too.

My current virus scanner (RAV) on my employer's mail gateway comes 
with a cron job that checks for, and downloads, new sigs ever hour.

 - BS

Re: [Devil-linux-develop] clamav

[Heiko Z. <he...@zu...>]

Bruce Smith wrote:
>>where should we place the clamav virus signatures?
> 
> 
> Where is the default location for them?

/usr/share/clamav
But you can easily set this in the config file.

>>The virus dbs could get to big to fit in /etc (for the floppy save), but 
>>somewhere else it would always be outdated when a system boots.
> 
> 
> How big are they now?

linux:/lfssystem/data # l virus*
-rw-r--r--    1 root     root       140816 Sep  2 19:26 viruses.db
-rw-r--r--    1 root     root        32940 Sep  2 19:26 viruses.db2

>>We could add an init script, which fetches the newest files if the user 
>>wants to.
> 
> 
> That sounds good.
> 
> 
>>I personally would place the signatures somewhere under /var.
> 
> 
> Isn't /etc/ & /var/ our only choices for directories not on the CD?

Actually yes.

Heiko

Re: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Heiko Z. <he...@zu...>]

Dean Nedelman wrote:
>>Diego Torres wrote:
>>
>>>this patches could be useful (at least, there are *very* useful for me)
>>>
>>>http://secure.netroedge.com/~lm78/index.html
>>>
>>> http://secure.netroedge.com/~lm78/archive/i2c-2.8.0.tar.gz
>>> patches for i2c/lmsensors. don't work for kernels 2.5/2.6. it gives lots
> 
> of drivers for hardware sensors (cpu temp, cpu fan speed, etc...) available
> on today motherboards.
> 
>>> http://secure.netroedge.com/~lm78/archive/lm_sensors-2.8.0.tar.gz
>>> userspace applications for installing & configuring the sensors.
>>
>>I would vote for including those.
> 
> 
> Yes, -IF- we decided to run a monitor on one of the virtual consoles (maybe
> #9?), like we today with procinfo.

I like that idea !


>>>http://www.sistina.com/products_lvm.htm
>>>http://tldp.org/HOWTO/LVM-HOWTO/
>>> of course, if someone wants to do serious raid across several
> 
> disks/partitions, he will probably use the device mapper patches + lvm2
> patches to kernel.
> 
>>> ftp://ftp.sistina.com/pub/LVM2/tools/LVM2.0-latest.tgz
>>> user space tools
>>>
>>> ftp://ftp.sistina.com/pub/LVM2/device-mapper/device-mapper-latest.tgz
>>> kernel patches for 2.4.21.
>>
>>Should we switch to LVM2 ? How stable is it?
> 
> 
> LVM2 was selected over EVMS to replace LVM for the 2.6 kernel.  It probably
> shouldn't be considered stable at this point.  In fact, they are still
> trying to trace down Raid induced file corruption in the 2.6 kernel series.

OK, let's stick with LVM1. It works fine for me

>>>http://sourceforge.net/projects/acpi
>>> useful with newer or modern pcs, mainly because 2.4.21 is obsolete in
> 
> acpi terms...
> 
>>True. Add some point we have to support it, the question is do we want
>>to do it in the next release.
> 
> 
> I would vote to ship the V1.0 WITHOUT the ACPI enhancements, in order to
> actually SHIP V1.0!

Yes I agree.

> Now, I am as guilty of feature creap as anybody, but at somepoint we need to
> start looking at putting a stake in the ground with what we have, close up
> the bugs, and mark it V1.0.

I actually was working on closing most of the bugs.
I updated my server at home to the latest CVS release and found quite a 
few problems (all fixed in CVS).

cya
   Heiko

RE: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Dean N. <di...@ti...>]

>Diego Torres wrote:
>> this patches could be useful (at least, there are *very* useful for me)
>>
>> http://secure.netroedge.com/~lm78/index.html
>>
>>  http://secure.netroedge.com/~lm78/archive/i2c-2.8.0.tar.gz
>>  patches for i2c/lmsensors. don't work for kernels 2.5/2.6. it gives lots
of drivers for hardware sensors (cpu temp, cpu fan speed, etc...) available
on today motherboards.
>>
>>  http://secure.netroedge.com/~lm78/archive/lm_sensors-2.8.0.tar.gz
>>  userspace applications for installing & configuring the sensors.
>
>I would vote for including those.

Yes, -IF- we decided to run a monitor on one of the virtual consoles (maybe
#9?), like we today with procinfo.

>>
>> http://www.sistina.com/products_lvm.htm
>> http://tldp.org/HOWTO/LVM-HOWTO/
>>  of course, if someone wants to do serious raid across several
disks/partitions, he will probably use the device mapper patches + lvm2
patches to kernel.
>>
>>  ftp://ftp.sistina.com/pub/LVM2/tools/LVM2.0-latest.tgz
>>  user space tools
>>
>>  ftp://ftp.sistina.com/pub/LVM2/device-mapper/device-mapper-latest.tgz
>>  kernel patches for 2.4.21.
>
>Should we switch to LVM2 ? How stable is it?

LVM2 was selected over EVMS to replace LVM for the 2.6 kernel.  It probably
shouldn't be considered stable at this point.  In fact, they are still
trying to trace down Raid induced file corruption in the 2.6 kernel series.

>> http://sourceforge.net/projects/acpi
>>  useful with newer or modern pcs, mainly because 2.4.21 is obsolete in
acpi terms...
>
>True. Add some point we have to support it, the question is do we want
>to do it in the next release.

I would vote to ship the V1.0 WITHOUT the ACPI enhancements, in order to
actually SHIP V1.0!

Now, I am as guilty of feature creap as anybody, but at somepoint we need to
start looking at putting a stake in the ground with what we have, close up
the bugs, and mark it V1.0.

Just my thoughts....

Dean Nedelman
TimeLord Consulting

[Devil-linux-develop] [ devil-linux-Bugs-802878 ] fix sagator init script

[SourceForge.net <no...@so...>]

Bugs item #802878, was opened at 2003-09-08 20:47
Message generated for change (Settings changed) made by smiley73
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=410643&aid=802878&group_id=34096

Category: Base System
Group: None
>Status: Closed
Resolution: None
Priority: 5
Submitted By: Heiko Zuerker (smiley73)
>Assigned to: Heiko Zuerker (smiley73)
Summary: fix sagator init script

Initial Comment:
already in build/scripts/scripts/sagator , but needs to

be changed for the DL environment

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=410643&aid=802878&group_id=34096

Re: [Devil-linux-develop] useful patches to a 2.4.21 kernel

[Heiko Z. <he...@zu...>]

Diego Torres wrote:
> this patches could be useful (at least, there are *very* useful for me)
> 
> http://secure.netroedge.com/~lm78/index.html
> 
>  http://secure.netroedge.com/~lm78/archive/i2c-2.8.0.tar.gz
>  patches for i2c/lmsensors. don't work for kernels 2.5/2.6. it gives lots of drivers for hardware sensors (cpu temp, cpu fan speed, etc...) available on today motherboards.
> 
>  http://secure.netroedge.com/~lm78/archive/lm_sensors-2.8.0.tar.gz
>  userspace applications for installing & configuring the sensors.

I would vote for including those.

> http://pcmcia-cs.sourceforge.net/
>  as always, getting the latest pcmcia-cs drivers is a must. don't even think in enabling pcmcia in-kernel support.
>  
> http://www.sistina.com/products_lvm.htm
> http://tldp.org/HOWTO/LVM-HOWTO/
>  of course, if someone wants to do serious raid across several disks/partitions, he will probably use the device mapper patches + lvm2 patches to kernel.
> 
>  ftp://ftp.sistina.com/pub/LVM2/tools/LVM2.0-latest.tgz
>  user space tools
>  
>  ftp://ftp.sistina.com/pub/LVM2/device-mapper/device-mapper-latest.tgz
>  kernel patches for 2.4.21.

Should we switch to LVM2 ? How stable is it?

> http://sourceforge.net/projects/acpi
>  useful with newer or modern pcs, mainly because 2.4.21 is obsolete in acpi terms...

True. Add some point we have to support it, the question is do we want 
to do it in the next release.

cya
   Heiko