How big a kernel dump can be on 32GB m/c.

[hari krishnan]

Hi kernelnewbies,

I have a system which has 16GB memory +16GB for swap space total space
required in /var/crash is 32GB.
Can I keep less then 32GB for /var/crash ?  many times I see kernel dump
less then 20GB.

My problem is we have 32GB RAM memory in production machine on RHEL5U5 and I
don't have 64GB space for /var/crash.
Could anyone tell me what is the minimum /var/crash we should have to get
dump stored.

Thanks,
Hari

Re: Point of brk()?

[shailesh jain]

Only if stack and heap did not grow in opposite direction. Only if you
did not care for shared libraries memory usage.

On Tue, Oct 5, 2010 at 5:30 PM, javad karabi  wrote:
> What is the point of having an adjustable end to the "data segment"/heap?
>
> Why not just give a process all of the available memory, I.e. set the
> end_of_data variable to the highest possible value, then let paging do the
> rest?
>
> Much love for kernel newbies!



-- 
Shailesh Jain

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Point of brk()?

[javad karabi]

What is the point of having an adjustable end to the "data segment"/heap?

Why not just give a process all of the available memory, I.e. set the
end_of_data variable to the highest possible value, then let paging do the
rest?

Much love for kernel newbies!

Re: Digest of Questions & Solutions

[John Mahoney]

On Tue, Oct 5, 2010 at 4:34 PM, John Mahoney  wrote:
>> John - appreciate you posting that link for me but I guess, I already have
>> it. Infact, if you read my mail once again you will realize that I asked for
>> a compilation of queries posted & solutions offered over the years in the
>> pdf format (if not all the years then probably for all months of this year).
>> However, really appreciate your effort.
>>
>
> Apologies, I misread the statement.
>
>
> you can grab the content locally with something such as
>
> wget \
>     --recursive \
>     --no-clobber \
>     --page-requisites \
>     --html-extension \
>     --convert-links \
>     --restrict-file-names=windows \
>     --domains mail.nl.linux.org \
>     --no-parent \
>         http://mail.nl.linux.org/kernelnewbies/2010-01/
>
> This will grab jan. 2010.  You can probablly find a programto convert
> that to pdf or it can be opened  with a regular browser by opening the
> index,php file in the folder.  This site actually scrapes pretty well.

I meant index.html... I am so used to creating web content with php sorry.

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Digest of Questions & Solutions

[John Mahoney]

> John - appreciate you posting that link for me but I guess, I already have
> it. Infact, if you read my mail once again you will realize that I asked for
> a compilation of queries posted & solutions offered over the years in the
> pdf format (if not all the years then probably for all months of this year).
> However, really appreciate your effort.
>

Apologies, I misread the statement.


you can grab the content locally with something such as

wget \
 --recursive \
 --no-clobber \
 --page-requisites \
 --html-extension \
 --convert-links \
 --restrict-file-names=windows \
 --domains mail.nl.linux.org \
 --no-parent \
 http://mail.nl.linux.org/kernelnewbies/2010-01/

This will grab jan. 2010.  You can probablly find a programto convert
that to pdf or it can be opened  with a regular browser by opening the
index,php file in the folder.  This site actually scrapes pretty well.

Change http://mail.nl.linux.org/kernelnewbies/2010-01/ to
http://mail.nl.linux.org/kernelnewbies/2008-06/ and grab June 2008...
you get the idea.


Best of luck,
John

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Digest of Questions & Solutions

[Manish Katiyar]

On Tue, Oct 5, 2010 at 12:21 PM, Linux "Kernel" Explorer
 wrote:
> Hi James Bond & John Mahoney,
>
> Thanks a ton for replying. I guess, I did my share of search & experiments
> before posting this one out.I am good on Hygiene factors.
>
> John - appreciate you posting that link for me but I guess, I already have
> it. Infact, if you read my mail once again you will realize that I asked for
> a compilation of queries posted & solutions offered over the years in the
> pdf format (if not all the years then probably for all months of this year).
> However, really appreciate your effort.

I haven't tried so I am not sure... but with little scripting and all
those free pdf convertors it should be easy. And then you can post
your script for others too use too :-)

>
> Regards,
>
> Linux "Kernel" Explorer
>
> On Tue, Oct 5, 2010 at 4:24 PM, Bond  wrote:
>>
>> On Sun, Oct 3, 2010 at 2:25 PM, Linux Explorer 
>> wrote:
>> > Hello Everyone,
>> >
>> >
>> Hi explorer things do not work that way as you asked.
>> Do some experiment and then mail here may be some of us can point you
>> to right thing.
>
>



-- 
Thanks -
Manish
==
[$\*.^ -- I miss being one of them
==

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Linux Kernel Development - A Practical Approach

[Robert P. J. Day]

  i've dropped all but the kernelnewbies mailing list from the CC
list.

On Tue, 5 Oct 2010, Tapas Mishra wrote:

> On Tue, Oct 5, 2010 at 12:20 AM, Linux "Kernel" Explorer
>  wrote:
> > Hello Everyone,
> >
> > I am reading 'Linux Kernel Development' by Robert Love these days.
> >
> > This book takes you on a theoretical journey of the linux kernel
> > world.Though the book is good but I do have my share of concerns.

> Exactly even I do have same concerns. www.crashcourse.ca check the
> above website till today this is the only most relevant thing which
> I find easy to begin with.

  thanks, i appreciate the compliment.  my online kernel programming
course still has a number of lessons to be added, i just need the time
to get around to finishing them.  mercifully, i have next week off so
i hope to publish most of what's left in that time.

  as for robert's book, it's not meant as a hard-core kernel
programmer reference; it's more designed to introduce the beginner to
the theory and layout of the kernel, and the principles of kernel
programming.  if you want more driver-oriented content, i'd suggest
"essential linux device drivers":

http://www.amazon.ca/Essential-Device-Drivers-Sreekrishnan-Venkateswaran/dp/0132396556/

robert's book (LKD3) lays the foundation, while ELDD gets into more
specific driver detail.  if you can afford it, you really should buy
both books as they serve very different purposes.

rday

-- 


Robert P. J. Day   Waterloo, Ontario, CANADA
http://crashcourse.ca

Twitter:   http://twitter.com/rpjday
LinkedIn:   http://ca.linkedin.com/in/rpjday


--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Digest of Questions & Solutions

[Linux "Kernel" Explorer]

Hi James Bond & John Mahoney,

Thanks a ton for replying. I guess, I did my share of search & experiments
before posting this one out.I am good on Hygiene factors.

John - appreciate you posting that link for me but I guess, I already have
it. Infact, if you read my mail once again you will realize that I asked for
a compilation of queries posted & solutions offered over the years in the
pdf format (if not all the years then probably for all months of this year).
However, really appreciate your effort.

Regards,

Linux "Kernel" Explorer

On Tue, Oct 5, 2010 at 4:24 PM, Bond  wrote:

> On Sun, Oct 3, 2010 at 2:25 PM, Linux Explorer 
> wrote:
> > Hello Everyone,
> >
> >
> Hi explorer things do not work that way as you asked.
> Do some experiment and then mail here may be some of us can point you
> to right thing.
>

Re: Persistent tap device

[Vimal]

Hi Mulyadi and John,

On 4 October 2010 23:03, John Mahoney  wrote:
>>
>> Most of the times, network related devices aren't represented as file
>> inside /dev. This is...well, you might think,against common sense that
>> "everything is file" under Linux.

@Mulyadi:

Yes, that's right.  I saw the source code of uml_utils, and it has
code to create a tun/tap device and obtain a file descriptor to it,
but not open an existing tun/tap device that has been created using
tunctl.  The example was also there in
Documentation/networking/tuntap.txt.

>
> tun/tap are the exception to the exception to the rule.  In other works
> network drives normally do not appear in /dev/ or as a file, but tun/tap do
> at least in my using them over the past 4 years. OpenVpn is a high profile
> known user of tun/tap, you may want to check there as a start.  Are you sure
> you have the tun/tap driver loaded.

@John:  I expected that, but currently, the only way for a program to
create a tun/tap device is to use an ioctl call which returns a file
descriptor that can be used to read/write on the device.  I was
wondering how to open an already existing tap device..

Thanks all.

-- 
Vimal

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: system call address

[Usman S. Ansari]

>> I have seen kernel symbols (talking in general, not sys_call_table),
>> appear in /proc/kallsyms in some version of kernel and not in others. 
>> Symbol definition is same in both kernel revs.
>>
>> What makes symbol appear or not in /proc/kallsym
>
> It's a configuration option when you compile the kernel: CONFIG_KALLSYMS

Not talking about /proc/kallsym file in general, but individual symbols which 
sometime show up and sometime does not.



--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

a process having different credentials' files

[mohit verma]

hi all,

a process :

current --> files_struct --> file -->cred
||
(* files)  (*fd_array)

it seems a process can have the files having different credentials .  for
example let a process be running in a high privilige level . after setuid (
or any member of family) system call it gains the lower(or higher)
 priority. the process remains same with the previous opened files and there
attached  credentials.

is it right  that  a process can have different ( or same) file opened  with
different credentials ?
if so then it must be a loop hole.

i know i am missing something.please correct me at the places.

Initramfs generated in Fedora 13 but not in Ubuntu 10.04

[Brett Mahar]

Hi all,
I just compiled and installed kernel 2.6.35.4 into Fedora 13_x86 and
noticed that an initramfs file was generated/installed in the boot
folder. Last week, when I unsuccessfully tried to compile and install
the same kernel in Ubuntu 10.04, no initramfs or initrd was generated.
In both cases I had selected "Initial RAM filesystem and RAM disk
(initramfs/initrd) support" while running make menuconfig.
So I am wondering why I ended up with an initramfs with Fedora and not
Ubuntu? I'm pretty new at this so am at a loss as to what could cause
the different outcome. Could the difference be that the default
install of Fedora comes with an initramfs, while Ubuntu doesn't (and
therefore in the kernel build process there is a "template" to work
from)?
Cheers, Brett.

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Digest of Questions & Solutions

[Bond]

On Sun, Oct 3, 2010 at 2:25 PM, Linux Explorer  wrote:
> Hello Everyone,
>
>
Hi explorer things do not work that way as you asked.
Do some experiment and then mail here may be some of us can point you
to right thing.

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Linux Kernel Development - A Practical Approach

[Tapas Mishra]

On Tue, Oct 5, 2010 at 12:20 AM, Linux "Kernel" Explorer
 wrote:
> Hello Everyone,
>
> I am reading 'Linux Kernel Development' by Robert Love these days.
>
> This book takes you on a theoretical journey of the linux kernel
> world.Though the book is good but I do have my share of concerns.
Exactly even I do have same concerns.
www.crashcourse.ca
check the above website till today this is the only most relevant
thing which I find easy to begin with.
You can then go to understand some more howtos on internet that will help.
LKD is no doubt good book but even I do not appreciate that.
Jumping directly to code in your or even  mine type of situation is
not easy having said that I would say you not to
indulge into books.If possible get some one who can easily give you a
hands on experience.
Then you will be able to understand what the text books talk.
A good advice I got from some one who works with processors is rather
than going in manuals
jump into the code you play with it and then you will gradually understand.
There may be different views but I would suggest you not to read books.

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ

Re: Digest of Questions & Solutions

[John Mahoney]

Did you try google for kernel-newbies archive?  This is really basic
knowledge I am not trying to be a hard *ss.

here is one.  I know gmane has most lists archived also as a general rule.

 http://mail.nl.linux.org/kernelnewbies/

--
John

Re: system call address

[luca ellero]

Il 05/10/2010 1.53, Usman S. Ansari wrote:

Not any more, sys_call_table is not exported any more in recent kernel (at
least 2.6)


I have seen kernel symbols (talking in general, not sys_call_table), appear in 
/proc/kallsyms in some version of kernel and not in others. Symbol definition 
is same in both kernel revs.

What makes symbol appear or not in /proc/kallsym




It's a configuration option when you compile the kernel: CONFIG_KALLSYMS

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecar...@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ