devil-linux-discuss — General Mailing List for all discussions

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

[Andrzej O. <an...@ma...>]

Heiko Zuerker wrote:
> Maybe we're missing a directory or some other command we need to execute for
> /dev to be proper?

Perhaps something did not go well in the previous compilation. Next 
compilation gave good results, with no udev errors. Yesterday I put this build 
(32-bit) on the main router and on all my satelite routers and on MTA gateway 
machine. Everything so far is working properly. mysql too.

Best Regards

-- 
Andrzej Odyniec

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

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

> -----Original Message-----
> From: Dominic Raferd [mailto:dl...@ed...]
> Sent: Friday, March 16, 2012 6:37 AM
> To: dev...@li...
> Subject: Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer
> 
> 
> 
> On 15/03/2012 14:24, Dominic Raferd wrote:
> > On 15/03/2012 13:21, Heiko Zuerker wrote:
> >> Quoting Dominic Raferd<dl...@ed...>:
> >>
> >>> On 15/03/2012 13:03, Heiko Zuerker wrote:
> >>>> Hey,
> >>>>
> >>>> is anybody using 1.6.0-RC2 (or the current CVS version) together with
> mysql?
> >>> yes (1.6.0-RC2-server x86)
> >>>
> >>>> If yes, is your mysql server starting on boot?
> >>> yes, but I had to run mysql_upgrade (try: man mysql_upgrade)...
> >> I did that.
> >> I get the following error in the mysql.err file:
> >> -----------------------------------------
> >> 120315 08:18:15 mysqld_safe mysqld from pid file /var/run/mysql.pid
> >> ended
> >> 120315 08:18:57 mysqld_safe Starting mysqld daemon with databases
> >> from /var/mysql
> >> 120315  8:18:58 [ERROR] Fatal error: Please read "Security" section
> >> of the manual to find out how to run mysqld as root!
> >>
> >> 120315  8:18:58 [ERROR] Aborting
> >> -----------------------------------------
> >>
> >> When I start it on the command line via the init script, then it
> >> works fine and runs under the  mysql user.
> >>
> >> I have this specified in the config file:
> >> MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid --
> datadir=/var/mysql"
> >>
> > Hmm, it works for me. I have:
> > MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid
> > --datadir=/home/mysql --tmpdir=/home/bigtmp"
> >
> > So I guess in your case, but not mine, mysqld_safe is not picking up
> > $MYSQL_PARM correctly? Has there been a change in CVS to
> > /etc/init.d/mysqld or to /usr/bin/mysqld_safe?
> 
> I think I caught this one. The reason for the problem is that variable
$USER is
> not defined when mysqld starts at boot time (it should be 'root'). If you
use
> the standard my.cnf (at /etc-cds/my.cnf and thence
> /etc/my.cnf) this doesn't matter because it defines user=mysql in [mysqld]
> section. But if you have a bespoke /etc/my.cnf without this definition,
then
> mysqld is started up without any specified user and so it falls back to
using
> the current user (root), and this then throws the error. [I confirmed this
by
> testing.]
> 
> If having $USER undefined at boot is 'by design' or unavoidable then
> /usr/bin/mysqld_safe can be patched to work thus:
> 
> sed -i 's/"$USER" = "root"/`id -u` -eq 0/' mysqld_safe
> 
> I've tested this and it works. Otherwise, if (as seems better to me) $USER
can
> be defined before /etc/init.d/mysqld runs at boot time (as presumably it
> used to be before 1.6.0?), this patch should be unnecessary.

I like your fix better actually.
Applied to CVS. Thanks ! :)
 
> At the end of working this out, I foolishly did rm /etc/init.d/mysqld
(instead
> of rm /shm/etc-mods/init.d/mysqld) and now I can't restore the correct
link-
> type to the underlying file at /etc-cd/init.d/mysqld (I've used a symbolic
link
> but this isn't right I know). Can anyone help me do this?

Honestly I have no clue... :(
I'd just copy the one from /etc-cd/ over. It'll get fixed during the next
upgrade.

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

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

> -----Original Message-----
> From: Andrzej Odyniec [mailto:an...@ma...]
> Sent: Thursday, March 15, 2012 1:29 PM
> To: dev...@li...
> Subject: Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer
> 
> Heiko Zuerker wrote:
> > is anybody using 1.6.0-RC2 (or the current CVS version) together with
> mysql?
> > If yes, is your mysql server starting on boot?
> 
> I did a review of my servers and I have one, bare metal, which starts
mysql -
> for some time unnecessarily. He was needed by the php forum on Apache,
> but now this forum does not exist. On this server was working compilation
> with kernel 3.2.6. However, mysql started on boot every time and did not
> fall.
> 
> > Starting Apache2: [  OK  ]
> > Starting MYSQL Server: [  OK  ]
> 
> After this is running on ps list:
> 
> > root@Front:~ # ps -ef|grep mysql
> > root      1806     1  0 18:39 ?        00:00:00 /bin/sh
/usr/bin/mysqld_safe --
> user=mysql --pid-file=/var/run/mysql.pid --datadir=/var/mysql
> > mysql     2124  1806  0 18:39 ?        00:00:00 /usr/sbin/mysqld
--basedir=/usr --
> datadir=/var/mysql --plugin-dir=/usr/lib/mysql/plugin --log-
> error=/var/mysql/Front.err --pid-file=/var/run/mysql.pid --
> socket=/tmp/mysql.sock --port=3306
> 
> Today I compiled new image from last CVS (this with changed network
script)
> and additionally patch and grsecurity 3.2.11, because I have had yesterday
> strange issue with 3.2.9 on one of satelite routers -- this router did not
> translated one (just one!) of internal addresses via NAT (but before
change
> to
> 3.2.9 all was ok). So I try 3.2.11...
> 
> Under this compilation on the mentioned server mysql is starting correctly
> too. But in both cases, there is no questions to mysql server, so daemon
only
> waits. But waits stable.
> 
> > I'm having trouble with 2 of my mysql servers and want to make sure
> > it's not related to my configuration.
> > (This is holding up the final 1.6 release).
> 
> Maybe this is configuration problem?
> 
> BTW. On this mentioned server I have very old nontypical SCSI disk with
> Adaptec controller. After upgrade from 3.2.6 (old udevd) to 3.2.11 (new
> udevd), this is the first time, I see on console this messages from udevd:
> 
> > Mounting other file systems...
> > /dev/sda1 on /var type ext4 (rw)
> > /dev/sde1 on /var/mailer type ext4 (rw)
> > udevd[700]: failed to create queue file: No such file or directory
> >
> > udevd[700]: failed to create queue file: No such file or directory
> >
> > udevd[700]: failed to create queue file: No such file or directory
> >
> > /dev/sdf4 on /var/iso type ext3 (rw)
> 
> and this messages are later repeated. BTW. this disk is workin correctly.
What
> do you think about it after your successful battle with the new udevd?

>From what I found on google, this points towards the devtmpfs not being
mounted.
Maybe we're missing a directory or some other command we need to execute for
/dev to be proper?
 
> As for now, I will return to 3.2.6.

I'll include .11 in the next update

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

[Dominic R. <dl...@ed...>]

On 15/03/2012 14:24, Dominic Raferd wrote:
> On 15/03/2012 13:21, Heiko Zuerker wrote:
>> Quoting Dominic Raferd<dl...@ed...>:
>>
>>> On 15/03/2012 13:03, Heiko Zuerker wrote:
>>>> Hey,
>>>>
>>>> is anybody using 1.6.0-RC2 (or the current CVS version) together with mysql?
>>> yes (1.6.0-RC2-server x86)
>>>
>>>> If yes, is your mysql server starting on boot?
>>> yes, but I had to run mysql_upgrade (try: man mysql_upgrade)...
>> I did that.
>> I get the following error in the mysql.err file:
>> -----------------------------------------
>> 120315 08:18:15 mysqld_safe mysqld from pid file /var/run/mysql.pid ended
>> 120315 08:18:57 mysqld_safe Starting mysqld daemon with databases from
>> /var/mysql
>> 120315  8:18:58 [ERROR] Fatal error: Please read "Security" section of
>> the manual to find out how to run mysqld as root!
>>
>> 120315  8:18:58 [ERROR] Aborting
>> -----------------------------------------
>>
>> When I start it on the command line via the init script, then it works
>> fine and runs under the  mysql user.
>>
>> I have this specified in the config file:
>> MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid --datadir=/var/mysql"
>>
> Hmm, it works for me. I have:
> MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid
> --datadir=/home/mysql --tmpdir=/home/bigtmp"
>
> So I guess in your case, but not mine, mysqld_safe is not picking up
> $MYSQL_PARM correctly? Has there been a change in CVS to
> /etc/init.d/mysqld or to /usr/bin/mysqld_safe?

I think I caught this one. The reason for the problem is that variable 
$USER is not defined when mysqld starts at boot time (it should be 
'root'). If you use the standard my.cnf (at /etc-cds/my.cnf and thence 
/etc/my.cnf) this doesn't matter because it defines user=mysql in 
[mysqld] section. But if you have a bespoke /etc/my.cnf without this 
definition, then mysqld is started up without any specified user and so 
it falls back to using the current user (root), and this then throws the 
error. [I confirmed this by testing.]

If having $USER undefined at boot is 'by design' or unavoidable then 
/usr/bin/mysqld_safe can be patched to work thus:

sed -i 's/"$USER" = "root"/`id -u` -eq 0/' mysqld_safe

I've tested this and it works. Otherwise, if (as seems better to me) 
$USER can be defined before /etc/init.d/mysqld runs at boot time (as 
presumably it used to be before 1.6.0?), this patch should be unnecessary.

At the end of working this out, I foolishly did rm /etc/init.d/mysqld 
(instead of rm /shm/etc-mods/init.d/mysqld) and now I can't restore the 
correct link-type to the underlying file at /etc-cd/init.d/mysqld (I've 
used a symbolic link but this isn't right I know). Can anyone help me do 
this?

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

[Andrzej O. <an...@ma...>]

Heiko Zuerker wrote:
> is anybody using 1.6.0-RC2 (or the current CVS version) together with mysql?
> If yes, is your mysql server starting on boot?

I did a review of my servers and I have one, bare metal, which starts mysql - 
for some time unnecessarily. He was needed by the php forum on Apache, but now 
this forum does not exist. On this server was working compilation with kernel 
3.2.6. However, mysql started on boot every time and did not fall.

> Starting Apache2: [  OK  ]
> Starting MYSQL Server: [  OK  ]

After this is running on ps list:

> root@Front:~ # ps -ef|grep mysql
> root      1806     1  0 18:39 ?        00:00:00 /bin/sh /usr/bin/mysqld_safe --user=mysql --pid-file=/var/run/mysql.pid --datadir=/var/mysql
> mysql     2124  1806  0 18:39 ?        00:00:00 /usr/sbin/mysqld --basedir=/usr --datadir=/var/mysql --plugin-dir=/usr/lib/mysql/plugin --log-error=/var/mysql/Front.err --pid-file=/var/run/mysql.pid --socket=/tmp/mysql.sock --port=3306

Today I compiled new image from last CVS (this with changed network script) 
and additionally patch and grsecurity 3.2.11, because I have had yesterday 
strange issue with 3.2.9 on one of satelite routers -- this router did not 
translated one (just one!) of internal addresses via NAT (but before change to 
3.2.9 all was ok). So I try 3.2.11...

Under this compilation on the mentioned server mysql is starting correctly 
too. But in both cases, there is no questions to mysql server, so daemon only 
waits. But waits stable.

> I'm having trouble with 2 of my mysql servers and want to make sure  
> it's not related to my configuration.
> (This is holding up the final 1.6 release).

Maybe this is configuration problem?

BTW. On this mentioned server I have very old nontypical SCSI disk with 
Adaptec controller. After upgrade from 3.2.6 (old udevd) to 3.2.11 (new 
udevd), this is the first time, I see on console this messages from udevd:

> Mounting other file systems...
> /dev/sda1 on /var type ext4 (rw)
> /dev/sde1 on /var/mailer type ext4 (rw)
> udevd[700]: failed to create queue file: No such file or directory
> 
> udevd[700]: failed to create queue file: No such file or directory
> 
> udevd[700]: failed to create queue file: No such file or directory
> 
> /dev/sdf4 on /var/iso type ext3 (rw)

and this messages are later repeated. BTW. this disk is workin correctly. What 
do you think about it after your successful battle with the new udevd?

As for now, I will return to 3.2.6.

Best Regards

-- 
Andrzej Odyniec

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

[Dominic R. <dl...@ed...>]

On 15/03/2012 13:21, Heiko Zuerker wrote:
> Quoting Dominic Raferd<dl...@ed...>:
>
>> On 15/03/2012 13:03, Heiko Zuerker wrote:
>>> Hey,
>>>
>>> is anybody using 1.6.0-RC2 (or the current CVS version) together with mysql?
>> yes (1.6.0-RC2-server x86)
>>
>>> If yes, is your mysql server starting on boot?
>> yes, but I had to run mysql_upgrade (try: man mysql_upgrade)...
> I did that.
> I get the following error in the mysql.err file:
> -----------------------------------------
> 120315 08:18:15 mysqld_safe mysqld from pid file /var/run/mysql.pid ended
> 120315 08:18:57 mysqld_safe Starting mysqld daemon with databases from
> /var/mysql
> 120315  8:18:58 [ERROR] Fatal error: Please read "Security" section of
> the manual to find out how to run mysqld as root!
>
> 120315  8:18:58 [ERROR] Aborting
> -----------------------------------------
>
> When I start it on the command line via the init script, then it works
> fine and runs under the  mysql user.
>
> I have this specified in the config file:
> MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid --datadir=/var/mysql"
>

Hmm, it works for me. I have:
MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid 
--datadir=/home/mysql --tmpdir=/home/bigtmp"

So I guess in your case, but not mine, mysqld_safe is not picking up 
$MYSQL_PARM correctly? Has there been a change in CVS to 
/etc/init.d/mysqld or to /usr/bin/mysqld_safe?

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

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

Quoting Dominic Raferd <dl...@ed...>:

> On 15/03/2012 13:03, Heiko Zuerker wrote:
>> Hey,
>>
>> is anybody using 1.6.0-RC2 (or the current CVS version) together with mysql?
>
> yes (1.6.0-RC2-server x86)
>
>> If yes, is your mysql server starting on boot?
>
> yes, but I had to run mysql_upgrade (try: man mysql_upgrade)...

I did that.
I get the following error in the mysql.err file:
-----------------------------------------
120315 08:18:15 mysqld_safe mysqld from pid file /var/run/mysql.pid ended
120315 08:18:57 mysqld_safe Starting mysqld daemon with databases from  
/var/mysql
120315  8:18:58 [ERROR] Fatal error: Please read "Security" section of  
the manual to find out how to run mysqld as root!

120315  8:18:58 [ERROR] Aborting
-----------------------------------------

When I start it on the command line via the init script, then it works  
fine and runs under the  mysql user.

I have this specified in the config file:
MYSQL_PARM="--user=mysql --pid-file=/var/run/mysql.pid --datadir=/var/mysql"

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

Re: [Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

[Dominic R. <dl...@ed...>]

On 15/03/2012 13:03, Heiko Zuerker wrote:
> Hey,
>
> is anybody using 1.6.0-RC2 (or the current CVS version) together with mysql?

yes (1.6.0-RC2-server x86)

> If yes, is your mysql server starting on boot?

yes, but I had to run mysql_upgrade (try: man mysql_upgrade)...

> I'm having trouble with 2 of my mysql servers and want to make sure
> it's not related to my configuration.
> (This is holding up the final 1.6 release).

HTH

Dominic

[Devil-Linux-discuss] mysql on 1.6.0-RC2 and newer

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

Hey,

is anybody using 1.6.0-RC2 (or the current CVS version) together with mysql?
If yes, is your mysql server starting on boot?

I'm having trouble with 2 of my mysql servers and want to make sure  
it's not related to my configuration.
(This is holding up the final 1.6 release).

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

Re: [Devil-Linux-discuss] Hardening Openvpn security

[Frank W. <Fra...@ct...>]

Hi, 
I use the following (on DL)

in openvpn's server conf:

...
auth-user-pass-verify /etc/openvpn/auth/auth.pl via-file
tmp-dir /var/tmp
username-as-common-name
client-connect /etc/openvpn/auth/connect-script.pl
client-disconnect /etc/openvpn/auth/disconnect-script.pl
...



in the auth.pl file:

....
# Get username/password from file

if ($ARG = shift @ARGV) {
    if (!open (UPFILE, "<$ARG")) {
        print "Could not open username/password file: $ARG\n";
        exit 1;
    }
} else {
    print "No username/password file specified on command line\n";
    exit 1;
}

$username = <UPFILE>;
$password = <UPFILE>;

...


>From here I do an LDAP verification... 


Hope this helps,

Frank

-- 
_______________________________________________
Centre de Technologie de l'Education
29 avenue John F. Kennedy
L-1855 Luxembourg-Kirchberg
email: Fra...@ct...
tél.: +352 247-85973
fax: +352 333797
_______________________________________________



On Tue, 2012-03-06 at 11:18 +0100, Paolo Spizzirri wrote:
> The director of my company is paranoid about VPN, so i decided to
> improve openvpn security adding username/password authentication.
> Is there some script in devil distro to check username and password ?
> At the moment i want to store username and password on file in devil
> linux machine; What type of format of data is required for the file ?
> Thanks for your cooperation,
> Paolo 
> -- 
> 
> Paolo Spizzirri
> 
> Software Developer
> 
> Mobile: +39 349 7784465
> Email: pao...@gm...
> 
> ------------------------------------------------------------------------------
> Keep Your Developer Skills Current with LearnDevNow!
> The most comprehensive online learning library for Microsoft developers
> is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
> Metro Style Apps, more. Free future releases when you subscribe now!
> http://p.sf.net/sfu/learndevnow-d2d
> _______________________________________________ Devil-linux-discuss mailing list Dev...@li... https://lists.sourceforge.net/lists/listinfo/devil-linux-discuss

[Devil-Linux-discuss] Hardening Openvpn security

[Paolo S. <pao...@gm...>]

The director of my company is paranoid about VPN, so i decided to improve openvpn security adding username/password
authentication.
Is there some script in devil distro to check username and password ? At the moment i want to store username and
password on file in devil linux machine; What type of format of data is required for the file ?
Thanks for your cooperation,
Paolo
-- 

*Paolo Spizzirri*

Software Developer

Mobile: +39 349 7784465
Email: pao...@gm...

Re: [Devil-Linux-discuss] deprecated oom_adj in udevd

[Serge L. <ser...@gm...>]

Sure. Disable it for now - I'll try to adapt the rules and re-enable it then.

Serge

On 03/05/2012 05:19 PM, Heiko Zuerker wrote:
> I now have a working version with the new udev.
> Unfortunately it doesn't like our aoe rule set and even complains about some
> of its own.
> I don't think this is worth potentially delaying the final 1.6... Thoughts?
>

Re: [Devil-Linux-discuss] deprecated oom_adj in udevd

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

I now have a working version with the new udev.
Unfortunately it doesn't like our aoe rule set and even complains about some
of its own.
I don't think this is worth potentially delaying the final 1.6... Thoughts?

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

> -----Original Message-----
> From: Heiko Zuerker [mailto:he...@zu...]
> Sent: Monday, March 05, 2012 12:32 PM
> To: dev...@li...
> Subject: Re: [Devil-Linux-discuss] deprecated oom_adj in udevd
> 
> Serge,
> 
> Quoting Serge Leschinsky <ser...@gm...>:
> > I have a working udev-173. It seems to be the last "old" udev (no
> > mandatory /run, do devfs, no kmod etc) - if you don't mind I'll update
> > DL to this version.
> 
> Yes that would be great.
> I'll keep looking into the "new" udev, but only have limited time
available.
> 
> --
> 
> Regards
>    Heiko Zuerker
>    http://www.devil-linux.org
> 
> 
> 
>
----------------------------------------------------------------------------
--
> Try before you buy = See our experts in action!
> The most comprehensive online learning library for Microsoft developers is
> just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro
> Style Apps, more. Free future releases when you subscribe now!
> http://p.sf.net/sfu/learndevnow-dev2
> _______________________________________________
> Devil-linux-discuss mailing list
> Dev...@li...
> https://lists.sourceforge.net/lists/listinfo/devil-linux-discuss

Re: [Devil-Linux-discuss] deprecated oom_adj in udevd

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

Serge,

Quoting Serge Leschinsky <ser...@gm...>:
> I have a working udev-173. It seems to be the last "old" udev (no mandatory
> /run, do devfs, no kmod etc) - if you don't mind I'll update DL to  
> this version.

Yes that would be great.
I'll keep looking into the "new" udev, but only have limited time available.

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

Re: [Devil-Linux-discuss] deprecated oom_adj in udevd

[Serge L. <ser...@gm...>]

Heiko,

I have a working udev-173. It seems to be the last "old" udev (no mandatory 
/run, do devfs, no kmod etc) - if you don't mind I'll update DL to this version.

Serge

On 03/05/2012 06:55 AM, Heiko Zuerker wrote:
>
> Quoting Heiko Zuerker<he...@zu...>:
>
>> Quoting Boursin Olivier<oli...@in...>:
>>
>>> Hello,
>>> I just migrate from Devil-Linux 1.4.2-i686 to Devil-Linux 1.6.0-RC2-i686
>>> and It seem that there is an error message concerning udevd at the
>>> beginning.
>>>
>>> udevd (925): /proc/925/oom_adj is deprecated, please use
>>> /proc/925/oom_score_adj instead.
>>>
>>> It is not blocking anything for me but I report this as it is still a
>>> RC.
>>> Does anyone else have this message ?
>>
>> The next DL release will have the latest udev, hopefully this will
>> resolve your issue.
>
> Adding the new udev has turned out to be much more complicated than I
> anticipated. I'll play around more, but don't have an ETA yet on when
> a newer version will be included.
>

Re: [Devil-Linux-discuss] deprecated oom_adj in udevd

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

Quoting Heiko Zuerker <he...@zu...>:

> Quoting Boursin Olivier <oli...@in...>:
>
>> Hello,
>> I just migrate from Devil-Linux 1.4.2-i686 to Devil-Linux 1.6.0-RC2-i686
>> and It seem that there is an error message concerning udevd at the
>> beginning.
>>
>> udevd (925): /proc/925/oom_adj is deprecated, please use
>> /proc/925/oom_score_adj instead.
>>
>> It is not blocking anything for me but I report this as it is still a
>> RC.
>> Does anyone else have this message ?
>
> The next DL release will have the latest udev, hopefully this will
> resolve your issue.

Adding the new udev has turned out to be much more complicated than I  
anticipated. I'll play around more, but don't have an ETA yet on when  
a newer version will be included.

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

Re: [Devil-Linux-discuss] deprecated oom_adj in udevd

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

Quoting Boursin Olivier <oli...@in...>:

> Hello,
> I just migrate from Devil-Linux 1.4.2-i686 to Devil-Linux 1.6.0-RC2-i686
> and It seem that there is an error message concerning udevd at the
> beginning.
>
> udevd (925): /proc/925/oom_adj is deprecated, please use
> /proc/925/oom_score_adj instead.
>
> It is not blocking anything for me but I report this as it is still a
> RC.
> Does anyone else have this message ?

The next DL release will have the latest udev, hopefully this will  
resolve your issue.

-- 

Regards
   Heiko Zuerker
   http://www.devil-linux.org

[Devil-Linux-discuss] deprecated oom_adj in udevd

[Boursin O. <oli...@in...>]

Hello,
I just migrate from Devil-Linux 1.4.2-i686 to Devil-Linux 1.6.0-RC2-i686
and It seem that there is an error message concerning udevd at the
beginning.

udevd (925): /proc/925/oom_adj is deprecated, please use
/proc/925/oom_score_adj instead.

It is not blocking anything for me but I report this as it is still a
RC.
Does anyone else have this message ?

O. Boursin

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Serge L. <ser...@gm...>]

On 02/27/2012 10:51 AM, Serge Leschinsky wrote:

> well, probably the hw was detected, but an appropriate device was not created.
> In initrd we don't use mdev - udev analog. Hmm.... Let's check udev events -
> edit /etc/init.d/boot and add "udevadm monitor" just after udevd launch. It's
> better to redirect output somewhere. syslog is not started yet.

please "udevadm monitor --env" to see all information

Serge

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Serge L. <ser...@gm...>]

Hi Martin,

On 02/27/2012 07:49 AM, Martin Glazer wrote:

> Thanks for the instructions, but even after replacing the initrd, I had no luck
> getting the mSATA drive recognised on boot.
>
> I also added pci_hotplug to the /etc/sysconfig/config INITRD parameter without
> success.
>
> Was there anything else I needed to do to get this recognised (if it was the
> hotplug module) or any other way of finding what module is required?

well, probably the hw was detected, but an appropriate device was not created. 
In initrd we don't use mdev - udev analog. Hmm.... Let's check udev events -
edit /etc/init.d/boot and add "udevadm monitor" just after udevd launch. It's 
better to redirect output somewhere. syslog is not started yet.


Serge

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Martin G. <sou...@gl...>]

Hi Serge,

On 02/24/2012 11:34 PM, Serge Leschinsky wrote:
> I assume you are in DL.
>
> cd /tmp/
> cp /boot/initrd.gz ./
> gzip -d initrd.gz
> mkdir 123
> cd ./123
> cpio -id<  ../initrd
> mkdir lib/modules/3.2.4-grsec/kernel/drivers/pci
> cp -ar /lib/modules/3.2.4-grsec/kernel/drivers/pci/hotplug
> ./lib/modules/3.2.4-grsec/kernel/drivers/pci/
> find . | cpio -H newc -o | gzip -9>  ../initrd.gz
>
> then replace old initrd.gz file on your boot media (/dev/sdb) by new one

Thanks for the instructions, but even after replacing the initrd, I had 
no luck getting the mSATA drive recognised on boot.

I also added pci_hotplug to the /etc/sysconfig/config INITRD parameter 
without success.

Was there anything else I needed to do to get this recognised (if it was 
the hotplug module) or any other way of finding what module is required?

Thanks

Martin

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Serge L. <ser...@gm...>]

On 02/24/2012 10:14 PM, Martin Glazer wrote:

> I think it must be the hotplug modules that are missing
>
> this is my output from lsmod
> root@Devil:~ # lsmod
> Module Size Used by
> ipv6 274019 12
> dm_mod 66521 0
> md_mod 100338 0
> i2c_isch 3124 0
> i2c_core 18870 1 i2c_isch
> uas 8089 0
> ahci 19843 0
> libahci 19102 1 ahci
> e1000e 138681 0
> libata 163311 2 ahci,libahci
> pch_gbe 32970 0
> mii 3787 1 pch_gbe
> pch_phub 6506 0
> lpc_sch 2464 0
> shpchp 25273 0
> pci_hotplug 22206 1 shpchp
> cn 3905 0
> loop 67443 2
...
> Can you please let me know how I can add the missing modules to initrd?
>

I assume you are in DL.

cd /tmp/
cp /boot/initrd.gz ./
gzip -d initrd.gz
mkdir 123
cd ./123
cpio -id < ../initrd
mkdir lib/modules/3.2.4-grsec/kernel/drivers/pci
cp -ar /lib/modules/3.2.4-grsec/kernel/drivers/pci/hotplug 
./lib/modules/3.2.4-grsec/kernel/drivers/pci/
find . | cpio -H newc -o | gzip -9 > ../initrd.gz

then replace old initrd.gz file on your boot media (/dev/sdb) by new one


Serge

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Martin G. <sou...@gl...>]

Hi Serge,

On 02/24/2012 06:51 PM, Serge Leschinsky wrote:
> Hmm... Am I correct that this device (/dev/sdb) has PCI-E interface? Interesting...

That is what I thought

> However, I don't see any problem with the _drive_ modules, all ata* drivers are
> in initrd. Most probably there is a pci discovery specific and you need pci
> hotplug modules ( pci_hotplug, shpchp) which are apparently not in initrd. If
> you want, I can send you instruction how to rebuild/repack it and we check what
> exactly modules are necessary for correct mSATA drive detection.
> Probably we need to add them by default....

I think it must be the hotplug modules that are missing

this is my output from lsmod
root@Devil:~ # lsmod
Module                  Size  Used by
ipv6                  274019  12
dm_mod                 66521  0
md_mod                100338  0
i2c_isch                3124  0
i2c_core               18870  1 i2c_isch
uas                     8089  0
ahci                   19843  0
libahci                19102  1 ahci
e1000e                138681  0
libata                163311  2 ahci,libahci
pch_gbe                32970  0
mii                     3787  1 pch_gbe
pch_phub                6506  0
lpc_sch                 2464  0
shpchp                 25273  0
pci_hotplug            22206  1 shpchp
cn                      3905  0
loop                   67443  2


additionally, in dmesg, this is what I get from when the mSATA drive is 
detected.
...
scsi2 : ahci
ata1: SATA max UDMA/133 abar m1024@0xa0004400 port 0xa0004500 irq 52
ata2: SATA max UDMA/133 abar m1024@0xa0004400 port 0xa0004580 irq 52
ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata1.00: ATA-7: TS2GMSA500, 20101212, max UDMA/133
ata1.00: 3865680 sectors, multi 0: LBA
ata1.00: configured for UDMA/133
scsi 1:0:0:0: Direct-Access     ATA      TS2GMSA500       2010 PQ: 0 ANSI: 5
sd 1:0:0:0: Attached scsi generic sg1 type 0
sd 1:0:0:0: [sdb] 3865680 512-byte logical blocks: (1.97 GB/1.84 GiB)
sd 1:0:0:0: [sdb] Write Protect is off
sd 1:0:0:0: [sdb] Mode Sense: 00 3a 00 00
sd 1:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't 
support DPOO
  or FUA
  sdb: sdb1 sdb2
sd 1:0:0:0: [sdb] Attached SCSI disk
ata2: SATA link down (SStatus 0 SControl 300)
...

Can you please let me know how I can add the missing modules to initrd?

Thanks

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Serge L. <ser...@gm...>]

Hi Martin,

On 02/24/2012 12:06 AM, Martin Glazer wrote:
> Hi,
>
> Thanks for that tip.
>
> I followed your instructions and it looks like it is using the pcieport module
....
> Kernel driver in use: pcieport
Hmm... Am I correct that this device (/dev/sdb) has PCI-E interface? Interesting...

However, I don't see any problem with the _drive_ modules, all ata* drivers are 
in initrd. Most probably there is a pci discovery specific and you need pci 
hotplug modules ( pci_hotplug, shpchp) which are apparently not in initrd. If 
you want, I can send you instruction how to rebuild/repack it and we check what 
exactly modules are necessary for correct mSATA drive detection.
Probably we need to add them by default....

Serge

Re: [Devil-Linux-discuss] Installing on mSATA drive

[Martin G. <sou...@gl...>]

Hi,

Thanks for that tip.

I followed your instructions and it looks like it is using the pcieport 
module

root@Devil:~ # ll /sys/class/block/sdb
lrwxrwxrwx 1 root root 0 Mar  3 06:24 /sys/class/block/sdb -> 
../../devices/pci0000:00/0000:00:17.0/0000:01:00.0/00b

root@Devil:~ # lspci -v
...

00:17.0 PCI bridge: Intel Corporation Atom Processor E6xx PCI Express 
Port 1 (prog-if 00 [Normal decode])
         Flags: bus master, fast devsel, latency 0
         Bus: primary=00, secondary=01, subordinate=02, sec-latency=0
         I/O behind bridge: 00001000-00001fff
         Memory behind bridge: a0000000-a0ffffff
         Prefetchable memory behind bridge: 80000000-801fffff
         Capabilities: [40] Express Root Port (Slot-), MSI 00
         Capabilities: [90] Subsystem: Gammagraphx, Inc. (or missing ID) 
Device 0000
         Capabilities: [a0] Power Management version 2
         Kernel driver in use: pcieport

Although doing a lsmod, I don't see that module loaded,

root@Devil:~ # lsmod
Module                  Size  Used by
ipv6                  274019  12
dm_mod                 66521  0
md_mod                100338  0
i2c_isch                3124  0
i2c_core               18870  1 i2c_isch
ahci                   19843  1
libahci                19102  1 ahci
libata                163311  2 ahci,libahci
uas                     8089  0
e1000e                138681  0
pch_gbe                32970  0
pch_phub                6506  0
mii                     3787  1 pch_gbe
lpc_sch                 2464  0
shpchp                 25273  0
pci_hotplug            22206  1 shpchp
cn                      3905  0
loop                   67443  2





On 02/23/2012 08:06 PM, Serge Leschinsky wrote:
> Hi Martin,
>
> On 02/23/2012 06:05 PM, Martin Glazer wrote:
>
>> Here is the output of fdisk -l (the USB device is seen as /dev/sda)
> ...
>> Any other thoughts?
> find the PCI address for the device
>
> example:
>
> root@hyper:/sys/class/block # ls -l /sys/class/block/sdb
> lrwxrwxrwx 1 root root 0 Feb 22 18:09 /sys/class/block/sdb ->
> ../../devices/pci0000:00/0000:00:1a.7/usb1/1-1/1-1:1.0/host0/target0:0:0/0:0:0:0/block/sdb
>
> (address is 00:1a.7)
>
> then find what modules the PCI device uses.
>
> example:
> 00:1a.7 USB controller: Intel Corporation 82801JI (ICH10 Family) USB2 EHCI
> Controller #2 (prog-if 20 [EHCI])
>           Subsystem: Super Micro Computer Inc Device 0006
>           Flags: bus master, medium devsel, latency 0, IRQ 18
>           Memory at fbeda000 (32-bit, non-prefetchable) [size=1K]
>           Capabilities: [50] Power Management version 2
>           Capabilities: [58] Debug port: BAR=1 offset=00a0
>           Capabilities: [98] PCI Advanced Features
>           Kernel driver in use: ehci_hcd
>
> (kernel driver is ehci_hcd).
>
>
> If it doesn't work for some reason, just post "lspci -v | grep "Kernel driver in
> use:" I'll try to find what driver is absent in initrd.
>
>
> Serge