one way to dmesg

one way to dmesg

[Alexey Fisher]

Hallo,
this is sort of continue of discussion about unified way to mount fs
readonly.
I wont to know how about dmesg/printk. I see ext[234] produce different
logs on mount even it use same code. I feel like i have some time to
digg in. My target is easy to parse dmesg.
Firs question: is it really important to print function in ext3_warning?
Or is it just ok to do it ext4_msg way?

Alexey

Re: one way to dmesg

[Greg Freemyer]

On Wed, Nov 4, 2009 at 9:26 AM, Alexey Fisher
<bug-track@fisher-privat.net> wrote:
> Hallo,
> this is sort of continue of discussion about unified way to mount fs
> readonly.
> I wont to know how about dmesg/printk. I see ext[234] produce different
> logs on mount even it use same code. I feel like i have some time to
> digg in. My target is easy to parse dmesg.

Patches for unified mount messages already exist.  Various versions
have been posted in the last couple months.

I don't know if they have been accepted yet or not.

Greg
-- 
Greg Freemyer
Head of EDD Tape Extraction and Processing team
Litigation Triage Solutions Specialist
http://www.linkedin.com/in/gregfreemyer
Preservation and Forensic processing of Exchange Repositories White Paper -
<http://www.norcrossgroup.com/forms/whitepapers/tng_whitepaper_fpe.html>

The Norcross Group
The Intersection of Evidence & Technology
http://www.norcrossgroup.com

Re: one way to dmesg

[Alexey Fisher]

Am Mittwoch, den 04.11.2009, 10:18 -0500 schrieb Greg Freemyer:
> On Wed, Nov 4, 2009 at 9:26 AM, Alexey Fisher
> <bug-track@fisher-privat.net> wrote:
> > Hallo,
> > this is sort of continue of discussion about unified way to mount fs
> > readonly.
> > I wont to know how about dmesg/printk. I see ext[234] produce different
> > logs on mount even it use same code. I feel like i have some time to
> > digg in. My target is easy to parse dmesg.
> 
> Patches for unified mount messages already exist.  Various versions
> have been posted in the last couple months.
> 
> I don't know if they have been accepted yet or not.

Can you point me this patch?
I found only "add a message in remount/umount for ext4"
http://patchwork.ozlabs.org/project/linux-ext4/list/?page=1

Re: one way to dmesg

[Greg Freemyer]

On Wed, Nov 4, 2009 at 10:29 AM, Alexey Fisher
<bug-track@fisher-privat.net> wrote:
> Am Mittwoch, den 04.11.2009, 10:18 -0500 schrieb Greg Freemyer:
>> On Wed, Nov 4, 2009 at 9:26 AM, Alexey Fisher
>> <bug-track@fisher-privat.net> wrote:
>> > Hallo,
>> > this is sort of continue of discussion about unified way to mount fs
>> > readonly.
>> > I wont to know how about dmesg/printk. I see ext[234] produce different
>> > logs on mount even it use same code. I feel like i have some time to
>> > digg in. My target is easy to parse dmesg.
>>
>> Patches for unified mount messages already exist.  Various versions
>> have been posted in the last couple months.
>>
>> I don't know if they have been accepted yet or not.
>
> Can you point me this patch?
> I found only "add a message in remount/umount for ext4"
> http://patchwork.ozlabs.org/project/linux-ext4/list/?page=1

Looks like Toshiyuki Okajima is the author.

http://markmail.org/message/ghyjee2krfmgtybt was just posted in the
last day or two.

That patch only adds the new unified dmesg print to the vfs layer.  I
think there were other patches that removed the scattered prints from
the various individual filesystems.

Greg
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: one way to dmesg

[Alexey Fisher]

Am Mittwoch, den 04.11.2009, 10:46 -0500 schrieb Greg Freemyer:
> On Wed, Nov 4, 2009 at 10:29 AM, Alexey Fisher
> <bug-track@fisher-privat.net> wrote:
> > Am Mittwoch, den 04.11.2009, 10:18 -0500 schrieb Greg Freemyer:
> >> On Wed, Nov 4, 2009 at 9:26 AM, Alexey Fisher
> >> <bug-track@fisher-privat.net> wrote:
> >> > Hallo,
> >> > this is sort of continue of discussion about unified way to mount fs
> >> > readonly.
> >> > I wont to know how about dmesg/printk. I see ext[234] produce different
> >> > logs on mount even it use same code. I feel like i have some time to
> >> > digg in. My target is easy to parse dmesg.
> >>
> >> Patches for unified mount messages already exist.  Various versions
> >> have been posted in the last couple months.
> >>
> >> I don't know if they have been accepted yet or not.
> >
> > Can you point me this patch?
> > I found only "add a message in remount/umount for ext4"
> > http://patchwork.ozlabs.org/project/linux-ext4/list/?page=1
> 
> Looks like Toshiyuki Okajima is the author.
> 
> http://markmail.org/message/ghyjee2krfmgtybt was just posted in the
> last day or two.
> 
> That patch only adds the new unified dmesg print to the vfs layer.  I
> think there were other patches that removed the scattered prints from
> the various individual filesystems.

Great, look really nice :)
but this is not all what i mean.
For example fs specific messages: EXT4-fs blabla, if you look to source
you'll see "EXT3 FS" "EXT II FS" (fat is worse), if you need to grep it
will be nightmare. It seems like ext4 is mostly up to date, ext2/3 need
some work. Are there any work on this?

Alexey

Re: one way to dmesg

[Theodore Tso]

On Wed, Nov 04, 2009 at 04:59:14PM +0100, Alexey Fisher wrote:
> 
> Great, look really nice :)
> but this is not all what i mean.
> For example fs specific messages: EXT4-fs blabla, if you look to source
> you'll see "EXT3 FS" "EXT II FS" (fat is worse), if you need to grep it
> will be nightmare. It seems like ext4 is mostly up to date, ext2/3 need
> some work. Are there any work on this?

What problem are you trying to solve?

Is it just an aesthetics thing?  Is there a specific scenario where
you need some kind of extra functionality?  If so, what is it?

There's some question about whether parsing dmesg is the best way to
get information about error/event reporting.  See the following
presentation from this year's Plumber's conference for a set of
ruminations on this issue:

http://linuxplumbersconf.org/2009/slides/Andi-Kleen-plumbers-error.pdf

						- Ted

Re: one way to dmesg

[Alexey Fisher]

Am Donnerstag, den 05.11.2009, 10:02 -0500 schrieb Theodore Tso:
> On Wed, Nov 04, 2009 at 04:59:14PM +0100, Alexey Fisher wrote:
> > 
> > Great, look really nice :)
> > but this is not all what i mean.
> > For example fs specific messages: EXT4-fs blabla, if you look to source
> > you'll see "EXT3 FS" "EXT II FS" (fat is worse), if you need to grep it
> > will be nightmare. It seems like ext4 is mostly up to date, ext2/3 need
> > some work. Are there any work on this?
> 
> What problem are you trying to solve?
> 
> Is it just an aesthetics thing?  Is there a specific scenario where
> you need some kind of extra functionality?  If so, what is it?

1. reduce information flow if you need to do some critical work with fs.
Easy way to filter it, some thing like 'grep -- "-fs"' (not 'grep --
[EXT,FAT,XFS...]').
2. fs should notify if some opposite to default options are used (or
default options disabled).
3. fs should be silent if default options used
4. fs should say for what bdev this message. 
example: you will 'mount -o ro,xip,reservation -t ext2 /dev/sda[2,4,5]'

in dmesg you'll get:
[ 4893.684892] reservations ON
[ 4893.684896] xip option not supported
[ 4893.684961] EXT2-fs warning: mounting ext3 filesystem as ext2
[ 4893.684964] EXT2-fs warning: maximal mount count reached, running
e2fsck is recommended
[ 4893.684990] EXT II FS: 0.5b, 95/08/09, bs=1024, fs=1024, gc=2,
bpg=8192, ipg=1280, mo=80010]

the question is how to filter it? for what bdev was this message? This
is not only about ext* this is about all fs, i just begun with ext2.

> There's some question about whether parsing dmesg is the best way to
> get information about error/event reporting.  See the following
> presentation from this year's Plumber's conference for a set of
> ruminations on this issue:
> 
> http://linuxplumbersconf.org/2009/slides/Andi-Kleen-plumbers-error.pdf

Thank you for the link, this is god food for brain :)
Please take a look to my initial "[PATCH] Unify log messages in ext2" on
linux-ext4.

Re: one way to dmesg

[Alexey Fisher]

Am Mittwoch, den 04.11.2009, 15:26 +0100 schrieb Alexey Fisher:
> Hallo,
> this is sort of continue of discussion about unified way to mount fs
> readonly.
> I wont to know how about dmesg/printk. I see ext[234] produce different
> logs on mount even it use same code. I feel like i have some time to
> digg in. My target is easy to parse dmesg.
> Firs question: is it really important to print function in ext3_warning?
> Or is it just ok to do it ext4_msg way?
Ok forgot this question, i found ext4_warning. So i will just add
ext2_msg and ext3_msg

Regards,
Alexey