Re: looking for corrections/enhancements/omissions for LKD3

[Robert P. J. Day]

On Sun, 31 Oct 2010, Greg KH wrote:

> On Sun, Oct 31, 2010 at 08:18:52AM -0400, Robert P. J. Day wrote:
> > > Also, I saw XIP in ext2 filesystem. Quite neat..but again, why?
> >
> >   hmmm ... not sure, i'll look into that.
>
> So you can run Linux on a system with very limited amount of ram and
> your code running in rom or flash.

  sorry, i didn't mean i'd look into *why* XIP exists, i'd look into
whether this was a new feature for ext2.

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: how to efficient use linux caches

[Rajat Sharma]

> what kind of system call will increase the size of cache?

be clear about which cache? Are you interested in dentry cache, inode cache
or page cache? dentry cache and inode cache can grow only by access to
filesystem, access multiple files on your system. For page cache increase,
you can simple copy a very large file say 1G.

> I write a test program as below but from top, the size of cache seems
no change

I didnt get the purpose of your test program, perhaps you want to put memory
pressure on your system. What your test program is doing is just calling
malloc, it does nothing but grow the virtual memory area of your heap, It
wont allocate physical pages until you access the virtual address, try to
write some bytes on each allocated memory such that each 4K page is
accessed. This results in page vault in kernel and since your address is
valid so it will allocate a page for you. Typically called Demand Paging.

If you really want to put memory pressure, try allocating memory in kernel
through kmalloc.

Rajat

On Mon, Nov 1, 2010 at 8:56 AM, loody  wrote:

> Dear all:
> I have some questions about virtual memory (VM) subsystem of the Linux
> kernel.
> 1. From linux doc, there is drop_caches in /proc/ which can let us to
> free the cache that kernel used, such as dentries and inodes, etc.
> I know i can echo 1, 2 or 3 to free different level of caches in the
> linux kernel, but is there any more elegant or programming rule that I
> can follow to let it used more efficiently?
>
> 2. what kind of system call will increase the size of cache?
>  I write a test program as below but from top, the size of cache seems
> no change
>
> #include
> #include
> int main(void)
> {
> int index;
> char input;
> char * point[1024];
>
> for(index=0;index<1024;index++)
> {
> point[index]=malloc(1024*1024);
> printf("please input any charter for going down:\n");
> scanf("%c",&input);
> }
>
> }
>
> the memory usage of test.mips increased but the cached size is always
> 4916k as below:
>
> Mem: 8352K used, 132800K free, 0K shrd, 60K buff, 4916K cached
> Load average: 0.00 0.00 0.00
>   PID USER STATUS   VSZ  PPID %CPU %MEM COMMAND
>   400 0SW   668 1  0.3  0.4 busybox.telnetd
>   498 0RW   584   408  0.3  0.4 top
>   500 0SW 10156   317  0.0  7.1 test.mips
>   317 0SW   596 1  0.0  0.4 sh
>   408 0SW   592   400  0.0  0.4 sh
> 1 0SW   584 0  0.0  0.4 init
> 5 0SW<0 2  0.0  0.0 khelper
>86 0SW<0 2  0.0  0.0 khubd
> 3 0SW<0 2  0.0  0.0 ksoftirqd/0
>76 0SW<0 2  0.0  0.0 kblockd/0
> 4 0SW<0 2  0.0  0.0 events/0
> 2 0SW<0 0  0.0  0.0 kthreadd
>   112 0SW 0 2  0.0  0.0 pdflush
>   113 0SW<0 2  0.0  0.0 kswapd0
> 8 0SW<0 2  0.0  0.0 async/mgr
>   116 0SW<0 2  0.0  0.0 crypto/0
>   252 0SW<0 2  0.0  0.0 mtdblockd
>   254 0SW<0 2  0.0  0.0 nftld
>   255 0SW<0 2  0.0  0.0 inftld
>   286 0SW<0 2  0.0  0.0 rpciod/0
>   253 0SW<0 2  0.0  0.0 ftld
>89 0SW<0 2  0.0  0.0 kseriod
>   111 0SW 0 2  0.0  0.0 pdflush
>   448 0SW<0 2  0.0  0.0 scsi_eh_1
>   449 0SW<0 2  0.0  0.0 usb-storage
>   114 0SW<0 2  0.0  0.0 nfsiod
>   115 0SW<0 2  0.0  0.0 cifsoplockd
>
> Mem: 8352K used, 132800K free, 0K shrd, 60K buff, 4916K cached
> Load average: 0.04 0.01 0.00
>   PID USER STATUS   VSZ  PPID %CPU %MEM COMMAND
>   501 0RW   584   408  0.5  0.4 top
>   400 0SW   668 1  0.3  0.4 busybox.telnetd
>   500 0SW 37912   317  0.0 26.8 test.mips
>   317 0SW   596 1  0.0  0.4 sh
>   408 0SW   592   400  0.0  0.4 sh
> 1 0SW   584 0  0.0  0.4 init
> 5 0SW<0 2  0.0  0.0 khelper
>86 0SW<0 2  0.0  0.0 khubd
> 3 0SW<0 2  0.0  0.0 ksoftirqd/0
>76 0SW<0 2  0.0  0.0 kblockd/0
> 4 0SW<0 2  0.0  0.0 events/0
> 2 0SW<0 0  0.0  0.0 kthreadd
>   112 0SW 0 2  0.0  0.0 pdflush
>   113 0SW<0 2  0.0  0.0 kswapd0
> 8 0SW<0 2  0.0  0.0 async/mgr
>   116 0SW<0 2  0.0  0.0 crypto/0
>   252 0SW<0 2  0.0  0.0 mtdblockd
>   254 0SW<0 2  0.0  0.0 nftld
>   255 0SW<0 2  0.0  0.0 inftld
>   286 0SW<0 2  0.0  0.0 rpciod/0
>   253 0SW<0 2  0.0  0.0 ftld
>89 0SW<  

Re: does get_user_byte and put_user_byte exist

[John Mahoney]

On Sun, Oct 31, 2010 at 10:49 PM, Tapas Mishra wrote:

> I was reading an article
> http://www.linuxjournal.com/article/1221
> They have used get_user_byte and put_user_byte
> functions do they exist?
> I tried to man them but there were no entry.
>

That article is 1996 see:

http://www.linuxgrill.com/anonymous/fire/netfilter/kernel-hacking-HOWTO-4.html



>
> --
> 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: does get_user_byte and put_user_byte exist

[Mulyadi Santosa]

On Mon, Nov 1, 2010 at 09:49, Tapas Mishra  wrote:
> I was reading an article
> http://www.linuxjournal.com/article/1221
> They have used get_user_byte and put_user_byte
> functions do they exist?

Looking from the publication date of the mentioned article, quite
likely it's from 2.0.x or at least the beginning of 2.2.x era. LXR
confirms no such functions exist.

I think it's now called copy_from_user() and copy_to_user()

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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: looking for corrections/enhancements/omissions for LKD3

[Mulyadi Santosa]

Hi Greg...

On Mon, Nov 1, 2010 at 10:21, Greg KH  wrote:
> On Sun, Oct 31, 2010 at 08:18:52AM -0400, Robert P. J. Day wrote:
>> > Also, I saw XIP in ext2 filesystem. Quite neat..but again, why?
>>
>>   hmmm ... not sure, i'll look into that.
>
> So you can run Linux on a system with very limited amount of ram and
> your code running in rom or flash.


Make senses to me...thanks for the explanation greg. Anyway, why ext3
doesn't have similar feature?

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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: how to efficient use linux caches

[Greg Freemyer]

On Sun, Oct 31, 2010 at 11:26 PM, loody  wrote:
> Dear all:
> I have some questions about virtual memory (VM) subsystem of the Linux
> kernel.
> 1. From linux doc, there is drop_caches in /proc/ which can let us to
> free the cache that kernel used, such as dentries and inodes, etc.
> I know i can echo 1, 2 or 3 to free different level of caches in the
> linux kernel, but is there any more elegant or programming rule that I
> can follow to let it used more efficiently?

I think you fail to understand why drop_caches exists.  It is not to
enhance performance.  It is to enhance performance testing.

As such the main userspace api seems sufficient.

If you really want to control the cache, I think you need to implement
it in userspace for your app and do your disk i/o with o_direct.

> 2. what kind of system call will increase the size of cache?
>  I write a test program as below but from top, the size of cache seems
> no change
>
> #include
> #include
> int main(void)
> {
>         int index;
>         char input;
>         char * point[1024];
>
>         for(index=0;index<1024;index++)
>         {
>         point[index]=malloc(1024*1024);
>         printf("please input any charter for going down:\n");
>         scanf("%c",&input);
>         }
>
> }
>
> the memory usage of test.mips increased but the cached size is always
> 4916k as below:
>
> Mem: 8352K used, 132800K free, 0K shrd, 60K buff, 4916K cached
> Load average: 0.00 0.00 0.00
>   PID USER     STATUS   VSZ  PPID %CPU %MEM COMMAND
>   400 0        SW       668     1  0.3  0.4 busybox.telnetd
>   498 0        RW       584   408  0.3  0.4 top
>   500 0        SW     10156   317  0.0  7.1 test.mips
>   317 0        SW       596     1  0.0  0.4 sh
>   408 0        SW       592   400  0.0  0.4 sh
>     1 0        SW       584     0  0.0  0.4 init
>     5 0        SW<        0     2  0.0  0.0 khelper
>    86 0        SW<        0     2  0.0  0.0 khubd
>     3 0        SW<        0     2  0.0  0.0 ksoftirqd/0
>    76 0        SW<        0     2  0.0  0.0 kblockd/0
>     4 0        SW<        0     2  0.0  0.0 events/0
>     2 0        SW<        0     0  0.0  0.0 kthreadd
>   112 0        SW         0     2  0.0  0.0 pdflush
>   113 0        SW<        0     2  0.0  0.0 kswapd0
>     8 0        SW<        0     2  0.0  0.0 async/mgr
>   116 0        SW<        0     2  0.0  0.0 crypto/0
>   252 0        SW<        0     2  0.0  0.0 mtdblockd
>   254 0        SW<        0     2  0.0  0.0 nftld
>   255 0        SW<        0     2  0.0  0.0 inftld
>   286 0        SW<        0     2  0.0  0.0 rpciod/0
>   253 0        SW<        0     2  0.0  0.0 ftld
>    89 0        SW<        0     2  0.0  0.0 kseriod
>   111 0        SW         0     2  0.0  0.0 pdflush
>   448 0        SW<        0     2  0.0  0.0 scsi_eh_1
>   449 0        SW<        0     2  0.0  0.0 usb-storage
>   114 0        SW<        0     2  0.0  0.0 nfsiod
>   115 0        SW<        0     2  0.0  0.0 cifsoplockd
>
> Mem: 8352K used, 132800K free, 0K shrd, 60K buff, 4916K cached
> Load average: 0.04 0.01 0.00
>   PID USER     STATUS   VSZ  PPID %CPU %MEM COMMAND
>   501 0        RW       584   408  0.5  0.4 top
>   400 0        SW       668     1  0.3  0.4 busybox.telnetd
>   500 0        SW     37912   317  0.0 26.8 test.mips
>   317 0        SW       596     1  0.0  0.4 sh
>   408 0        SW       592   400  0.0  0.4 sh
>     1 0        SW       584     0  0.0  0.4 init
>     5 0        SW<        0     2  0.0  0.0 khelper
>    86 0        SW<        0     2  0.0  0.0 khubd
>     3 0        SW<        0     2  0.0  0.0 ksoftirqd/0
>    76 0        SW<        0     2  0.0  0.0 kblockd/0
>     4 0        SW<        0     2  0.0  0.0 events/0
>     2 0        SW<        0     0  0.0  0.0 kthreadd
>   112 0        SW         0     2  0.0  0.0 pdflush
>   113 0        SW<        0     2  0.0  0.0 kswapd0
>     8 0        SW<        0     2  0.0  0.0 async/mgr
>   116 0        SW<        0     2  0.0  0.0 crypto/0
>   252 0        SW<        0     2  0.0  0.0 mtdblockd
>   254 0        SW<        0     2  0.0  0.0 nftld
>   255 0        SW<        0     2  0.0  0.0 inftld
>   286 0        SW<        0     2  0.0  0.0 rpciod/0
>   253 0        SW<        0     2  0.0  0.0 ftld
>    89 0        SW<        0     2  0.0  0.0 kseriod
>   111 0        SW         0     2  0.0  0.0 pdflush
>   448 0        SW<        0     2  0.0  0.0 scsi_eh_1
>   449 0        SW<        0     2  0.0  0.0 usb-storage
>   114 0        SW<        0     2  0.0  0.0 nfsiod
>   115 0        SW<        0     2  0.0  0.0 cifsoplockd
>
> appreciate your kind help,
> miloody

Greg

--
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

how to efficient use linux caches

[loody]

Dear all:
I have some questions about virtual memory (VM) subsystem of the Linux
kernel.
1. From linux doc, there is drop_caches in /proc/ which can let us to
free the cache that kernel used, such as dentries and inodes, etc.
I know i can echo 1, 2 or 3 to free different level of caches in the
linux kernel, but is there any more elegant or programming rule that I
can follow to let it used more efficiently?

2. what kind of system call will increase the size of cache?
 I write a test program as below but from top, the size of cache seems
no change

#include
#include
int main(void)
{
        int index;
        char input;
        char * point[1024];

        for(index=0;index<1024;index++)
        {
        point[index]=malloc(1024*1024);
        printf("please input any charter for going down:\n");
        scanf("%c",&input);
        }

}

the memory usage of test.mips increased but the cached size is always
4916k as below:

Mem: 8352K used, 132800K free, 0K shrd, 60K buff, 4916K cached
Load average: 0.00 0.00 0.00
  PID USER     STATUS   VSZ  PPID %CPU %MEM COMMAND
  400 0        SW       668     1  0.3  0.4 busybox.telnetd
  498 0        RW       584   408  0.3  0.4 top
  500 0        SW     10156   317  0.0  7.1 test.mips
  317 0        SW       596     1  0.0  0.4 sh
  408 0        SW       592   400  0.0  0.4 sh
    1 0        SW       584     0  0.0  0.4 init
    5 0        SW<        0     2  0.0  0.0 khelper
   86 0        SW<        0     2  0.0  0.0 khubd
    3 0        SW<        0     2  0.0  0.0 ksoftirqd/0
   76 0        SW<        0     2  0.0  0.0 kblockd/0
    4 0        SW<        0     2  0.0  0.0 events/0
    2 0        SW<        0     0  0.0  0.0 kthreadd
  112 0        SW         0     2  0.0  0.0 pdflush
  113 0        SW<        0     2  0.0  0.0 kswapd0
    8 0        SW<        0     2  0.0  0.0 async/mgr
  116 0        SW<        0     2  0.0  0.0 crypto/0
  252 0        SW<        0     2  0.0  0.0 mtdblockd
  254 0        SW<        0     2  0.0  0.0 nftld
  255 0        SW<        0     2  0.0  0.0 inftld
  286 0        SW<        0     2  0.0  0.0 rpciod/0
  253 0        SW<        0     2  0.0  0.0 ftld
   89 0        SW<        0     2  0.0  0.0 kseriod
  111 0        SW         0     2  0.0  0.0 pdflush
  448 0        SW<        0     2  0.0  0.0 scsi_eh_1
  449 0        SW<        0     2  0.0  0.0 usb-storage
  114 0        SW<        0     2  0.0  0.0 nfsiod
  115 0        SW<        0     2  0.0  0.0 cifsoplockd

Mem: 8352K used, 132800K free, 0K shrd, 60K buff, 4916K cached
Load average: 0.04 0.01 0.00
  PID USER     STATUS   VSZ  PPID %CPU %MEM COMMAND
  501 0        RW       584   408  0.5  0.4 top
  400 0        SW       668     1  0.3  0.4 busybox.telnetd
  500 0        SW     37912   317  0.0 26.8 test.mips
  317 0        SW       596     1  0.0  0.4 sh
  408 0        SW       592   400  0.0  0.4 sh
    1 0        SW       584     0  0.0  0.4 init
    5 0        SW<        0     2  0.0  0.0 khelper
   86 0        SW<        0     2  0.0  0.0 khubd
    3 0        SW<        0     2  0.0  0.0 ksoftirqd/0
   76 0        SW<        0     2  0.0  0.0 kblockd/0
    4 0        SW<        0     2  0.0  0.0 events/0
    2 0        SW<        0     0  0.0  0.0 kthreadd
  112 0        SW         0     2  0.0  0.0 pdflush
  113 0        SW<        0     2  0.0  0.0 kswapd0
    8 0        SW<        0     2  0.0  0.0 async/mgr
  116 0        SW<        0     2  0.0  0.0 crypto/0
  252 0        SW<        0     2  0.0  0.0 mtdblockd
  254 0        SW<        0     2  0.0  0.0 nftld
  255 0        SW<        0     2  0.0  0.0 inftld
  286 0        SW<        0     2  0.0  0.0 rpciod/0
  253 0        SW<        0     2  0.0  0.0 ftld
   89 0        SW<        0     2  0.0  0.0 kseriod
  111 0        SW         0     2  0.0  0.0 pdflush
  448 0        SW<        0     2  0.0  0.0 scsi_eh_1
  449 0        SW<        0     2  0.0  0.0 usb-storage
  114 0        SW<        0     2  0.0  0.0 nfsiod
  115 0        SW<        0     2  0.0  0.0 cifsoplockd

appreciate your kind help,
miloody

--
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

does get_user_byte and put_user_byte exist

[Tapas Mishra]

I was reading an article
http://www.linuxjournal.com/article/1221
They have used get_user_byte and put_user_byte
functions do they exist?
I tried to man them but there were no entry.

--
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: swap storms on low memory i386, tasks blocked on i386 and amd64 for kernel > 2.6.36-git6

[Tharindu Rukshan Bamunuarachchi]

You can turn off SWAP and try ?
Are you using SWAP partition or file ?

Please search for following "phrase" in mailing list ... could this be
a issue in new release ?
"2.6.36 io bring the system to its knees"

__
Tharindu R Bamunuarachchi.




On Sun, Oct 31, 2010 at 5:50 PM, Mulyadi Santosa
 wrote:
> Hi Arthur..
>
> Interesting to read your (1st? really?) post, but I am not sure how
> far I can help...so I'll just take a shot...
>
> On Sun, Oct 31, 2010 at 18:36, Arthur Marsh
>  wrote:
>> I have a PII-266 with 384 MiB RAM (maximum capacity for the machine) and
>> an AMD64 dual core with 4 GiB RAM, both running Debian unstable plus
>> some packages from experimental and custom kernels.
>>
>> A typical load for the PII-266 is KDE 3.5.10 with konversation, icedove,
>> iceweasel, xmms, lynx, hp-systray, top and aptitude-curses.
>>
>> This works with stock Debian kernels and custom kernels up to and
>> including 2.6.36-git6. Under heavy load, free RAM will hover around 5
>> MiB but audio will still play with a very occasional skip and all
>> applications are responsive.
>
> that's free RAM..how about the buffers and page cache size?
>
> And how big is your swap size? And is it composed as single partition
> only in one disk? or...?
>
>> With the newer kernels, e.g. 2.6.36-git9,10,11 (all with the deadline
>> scheduler) get into a swap storm with over 32 MiB RAM free and kswapd0
>> taking 10 percent or more of CPU time. Kernel 2.6.36-git15 with the cfq
>> scheduler managed to keep smaller applications like xmms and shells
>> running but also had excessive free RAM and kswapd0 taking more than 10
>> percent of CPU time.
>
> have you tried to play with /proc/sys/vm/swappiness?
>
>> On the AMD64 dual core when compiling kernels with CONCURRENCY_LEVEL=4,
>> I would sometimes get the build process pausing with errors like:
>>
>> [ 2880.492025] INFO: task sh:10071 blocked for more than 120 seconds.
>> [ 2880.493165] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
>> disables this message.
>> [ 2880.494299] sh            D 8800cfc53780     0 10071   9092
>> 0x
>> [ 2880.495433]  880128b214a0 0086 880128851b00
>> 88012b67cba0
>> [ 2880.496612]  00013780 00013780 88011969dfd8
>> 88011969c000
>> [ 2880.497765]  880128b21798 880128b217a0 880128b214a0
>> 88011969dfd8
>> [ 2880.498925] Call Trace:
>> [ 2880.500096]  [] ? __switch_to+0x198/0x284
>> [ 2880.501262]  [] ? schedule_timeout+0x2d/0xd7
>> [ 2880.502418]  [] ? need_resched+0x1a/0x23
>> [ 2880.503564]  [] ? schedule+0x5b2/0x5c9
>> [ 2880.504739]  [] ? get_parent_ip+0x9/0x1b
>> [ 2880.505883]  [] ? wait_for_common+0x9d/0x10c
>> [ 2880.507024]  [] ? default_wake_function+0x0/0xf
>> [ 2880.508189]  [] ? stop_one_cpu+0x57/0x6e
>> [ 2880.509288]  [] ? migration_cpu_stop+0x0/0x2a
>> [ 2880.510354]  [] ? sub_preempt_count+0x83/0x94
>> [ 2880.511426]  [] ? sched_exec+0xbe/0xd6
>> [ 2880.512523]  [] ? do_execve+0xd1/0x28e
>> [ 2880.513582]  [] ? sys_execve+0x3f/0x54
>> [ 2880.514616]  [] ? stub_execve+0x6c/0xc0
>> amars...@am64:~$ uname -a
>> Linux am64 2.6.36-git16 #1 SMP PREEMPT Sun Oct 31 15:41:03 CST 2010
>> x86_64 GNU/Linux
>>
>> The process was unblocked by logging another session into the AMD64 machine.
>>
>> The "task foo blocked for more than 120 seconds" has also occurred on
>> the PII-266 uniprocessor machine with some of the 2.6.36-git9 or later
>> kernels. Previously I had not had such a problem on the PII-266 for
>> about 6 months.
>>
>> I can't begin to figure out where this problem was introduced.
>> Git-bisection doesn't always work as it may take a while for the
>> symptoms of a swap storm to appear.
>>
>> Is there any straightforward way to gather more information before
>> reporting this kind of issue upstream?
>
> Hmmm, I just sense, could be that the scheduler hackers once again
> play with child-forked-then-run-first? Once I read it is child which
> runs first, but lately I read it is the parent first...
>
> I also think, it might be something related to a patch Fengguang Wu
> made few weeks ago (IIRC) to reduce latency during high I/O... but I
> can't recall which patch that is. Try to google...
>
> NB: ftrace might help here please kindly read the ftrace tutorial
> posted in lwn.net in few latest editions...
>
>
> --
> regards,
>
> Mulyadi Santosa
> Freelance Linux trainer and consultant
>
> blog: the-hydra.blogspot.com
> training: mulyaditraining.blogspot.com
>
> --
> 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
>
>

--
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: looking for corrections/enhancements/omissions for LKD3

[julie Sullivan]

Hi Rob
Probably you have already thought of this but Jon Corbet outlines some new
features merged for 2.6.37 here:

http://lwn.net/Articles/411590/

some of which might be relevant - this stood out for me for CFS:

"The scheduler also will no longer charge processor time used to handle
interrupts to the process which happened to be running at the time."

Thanks
Julie

Re: looking for corrections/enhancements/omissions for LKD3

[Robert P. J. Day]

On Sun, 31 Oct 2010, julie Sullivan wrote:

>
>     i don't want to get too carried away in the sense of documenting
>   *everything* newer in the kernel, just (for now) new developments that
>   would alter/update the content of the book in some way.
>
>
> Hi Rob and list,

> Are you looking for typos as well? I've noticed one or two that made
> it across from LKD2 to LKD3 but wasn't sure if it was a bit too much
> like nit-picking to comment. Anyone with any intelligence could work
> out what was really meant in these cases anyway. :-)

  as the technical editor of that book, i can assure you there were no
typoes. :-)

  absolutely, i want to know about typoes, no matter how trivial.
thanks.

rday

Re: looking for corrections/enhancements/omissions for LKD3

[julie Sullivan]

>   i don't want to get too carried away in the sense of documenting
> *everything* newer in the kernel, just (for now) new developments that
> would alter/update the content of the book in some way.
>
>
Hi Rob and list,
Are you looking for typos as well? I've noticed one or two that made it
across from LKD2 to LKD3 but wasn't sure if it was a bit too much like
nit-picking to comment. Anyone with any intelligence could work out what was
really meant in these cases anyway. :-)

Regards
Julie

Re: looking for corrections/enhancements/omissions for LKD3

[Alison Chaiken]

I wrote:
>> -- kgdb/kdb

Robert P.J. Day writes:
>  good suggestion but, to help me out, it would be useful to throw in
> a couple lines as to how it relates to the book, page number, any
> additional references and so on

kgdb is mentioned in one paragraph on p. 373, kdb not at all.

>> -- new runtime_pm mechanism
>  again, where would that be incorporated in the current text?

Kernel events layer discussion on p. 361.

-- 
Alison Chaiken
(650) 279-5600  (cell)
             http://www.exerciseforthereader.org/
Standards should be recorded after they've emerged.   -- Jeff Lindsay

--
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: looking for corrections/enhancements/omissions for LKD3

[Mulyadi Santosa]

On Sun, Oct 31, 2010 at 19:18, Robert P. J. Day  wrote:
>    AS is mostly a subset of CFQ, so there's little point in still
>    providing this separate IO scheduler. Hopefully at some point we
>    can get down to one single IO scheduler again, at least this brings
>    us closer by having only one intelligent IO scheduler.
>
>    Signed-off-by: Jens Axboe 
>
> so i'll make a note of that.  thanks, that's exactly the sort of thing
> i'm looking for.

No problem...can I get a free copy of LKD3 plus your original sign
like "RPJD rocks!"? :)

>> Also, I saw XIP in ext2 filesystem. Quite neat..but again, why?
>
>  hmmm ... not sure, i'll look into that.

I just speculate it is due to fact that ext2 is used in embedded
stuffs..thus execute in place reduce page cache usage..especially for
one time running application.
-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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: looking for corrections/enhancements/omissions for LKD3

[Mulyadi Santosa]

Hi Rob...

On Sun, Oct 31, 2010 at 22:20, Robert P. J. Day  wrote:
>  no, since this isn't really a "janitor" type project, it's just to
> collect observations on the book.

OK, understood :)

>  i don't want to get too carried away in the sense of documenting
> *everything* newer in the kernel, just (for now) new developments that
> would alter/update the content of the book in some way.

Hm fine... well, in that case you might look for "something fancy"
that might need attention.

>  if you want to suggest stuff like that, also suggest exactly where
> in the book it would belong.  then it makes sense to include it.  and
> if it's appropriate, a reference to the git commit that introduced it
> would also be cool.

I'll think about it, meanwhile please keep us posted on what you have
found so far..

>  anyway, use your personal judgment and either post under this thread
> or email me, whatever you prefer.

I think I'll keep posting in this kernelnewbies' thread for now..if
everybody with this of course

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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: looking for corrections/enhancements/omissions for LKD3

[Robert P. J. Day]

On Sun, 31 Oct 2010, Alison Chaiken wrote:

> Robert P. J. Day  wrote:
> >> > collecting any observations
> >> > related to kernel features that were introduced so recently, they
> >> > never made it into LKD3, simply being too new.
>
> -- kgdb/kdb

  good suggestion but, to help me out, it would be useful to throw in
a couple lines as to how it relates to the book, page number, any
additional references and so on

> -- new runtime_pm mechanism

  again, where would that be incorporated in the current text?

> -- btrfs

  i'm not sure about this one.  it's simply a new feature in the
kernel. admittedly, but it doesn't immediately fit into the current
text.

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: looking for corrections/enhancements/omissions for LKD3

[Alison Chaiken]

Robert P. J. Day  wrote:
>> > collecting any observations
>> > related to kernel features that were introduced so recently, they
>> > never made it into LKD3, simply being too new.

-- kgdb/kdb

-- new runtime_pm mechanism

-- btrfs

-- 
Alison Chaiken
(650) 279-5600  (cell)
             http://www.exerciseforthereader.org/
Standards should be recorded after they've emerged.   -- Jeff Lindsay

--
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: looking for corrections/enhancements/omissions for LKD3

[Robert P. J. Day]

On Sun, 31 Oct 2010, Mulyadi Santosa wrote:

> Hi Rob...
>
> On Sun, Oct 31, 2010 at 19:34, Robert P. J. Day  wrote:
> >  i'll see what i can do. :-)  anyway, this is where i'm going to
> > track this stuff:
>
> Well, that's just my wishlist, so it's up to you whether you wanna
> do me a favour or not. BTW, have you tried to consult other Linux
> kernel related list like kernel janitor too?

  no, since this isn't really a "janitor" type project, it's just to
collect observations on the book.

> > http://www.crashcourse.ca/wiki/index.php/Updates_to_LKD3
>
> yay, I am the first there :)

  only until someone else finds an update on an earlier page. :-)

> > this is a totally unofficial page, i just want to keep up with
> > changes for when LKD4 comes out. :-)  so i might as well start
> > documenting changes now.  and, as you can see, i'll give credit to
> > all reporters.

> Well, tracking changes are good IMO, but I suggest you just pick the
> most significant ones.
>
> The criterias could be something like:
> - whether it affects latency
> - whether it improves graphic card improvement
> - whether it changes how scheduling works
> etc

  i don't want to get too carried away in the sense of documenting
*everything* newer in the kernel, just (for now) new developments that
would alter/update the content of the book in some way.

> NB: Does LKD3 mention about Kernel Same page Merge (KSM)? No need to
> be deep, but I think it's interesting to mention it since it will
> reduce memory consumption done by virtual machine/hypervisor (KVM,
> for example).

  if you want to suggest stuff like that, also suggest exactly where
in the book it would belong.  then it makes sense to include it.  and
if it's appropriate, a reference to the git commit that introduced it
would also be cool.

  anyway, use your personal judgment and either post under this thread
or email me, whatever you prefer.

rday

-- 


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

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

Re: swap storms on low memory i386, tasks blocked on i386 and amd64 for kernel > 2.6.36-git6

[Mulyadi Santosa]

Hi Arthur..

Interesting to read your (1st? really?) post, but I am not sure how
far I can help...so I'll just take a shot...

On Sun, Oct 31, 2010 at 18:36, Arthur Marsh
 wrote:
> I have a PII-266 with 384 MiB RAM (maximum capacity for the machine) and
> an AMD64 dual core with 4 GiB RAM, both running Debian unstable plus
> some packages from experimental and custom kernels.
>
> A typical load for the PII-266 is KDE 3.5.10 with konversation, icedove,
> iceweasel, xmms, lynx, hp-systray, top and aptitude-curses.
>
> This works with stock Debian kernels and custom kernels up to and
> including 2.6.36-git6. Under heavy load, free RAM will hover around 5
> MiB but audio will still play with a very occasional skip and all
> applications are responsive.

that's free RAM..how about the buffers and page cache size?

And how big is your swap size? And is it composed as single partition
only in one disk? or...?

> With the newer kernels, e.g. 2.6.36-git9,10,11 (all with the deadline
> scheduler) get into a swap storm with over 32 MiB RAM free and kswapd0
> taking 10 percent or more of CPU time. Kernel 2.6.36-git15 with the cfq
> scheduler managed to keep smaller applications like xmms and shells
> running but also had excessive free RAM and kswapd0 taking more than 10
> percent of CPU time.

have you tried to play with /proc/sys/vm/swappiness?

> On the AMD64 dual core when compiling kernels with CONCURRENCY_LEVEL=4,
> I would sometimes get the build process pausing with errors like:
>
> [ 2880.492025] INFO: task sh:10071 blocked for more than 120 seconds.
> [ 2880.493165] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
> disables this message.
> [ 2880.494299] sh            D 8800cfc53780     0 10071   9092
> 0x
> [ 2880.495433]  880128b214a0 0086 880128851b00
> 88012b67cba0
> [ 2880.496612]  00013780 00013780 88011969dfd8
> 88011969c000
> [ 2880.497765]  880128b21798 880128b217a0 880128b214a0
> 88011969dfd8
> [ 2880.498925] Call Trace:
> [ 2880.500096]  [] ? __switch_to+0x198/0x284
> [ 2880.501262]  [] ? schedule_timeout+0x2d/0xd7
> [ 2880.502418]  [] ? need_resched+0x1a/0x23
> [ 2880.503564]  [] ? schedule+0x5b2/0x5c9
> [ 2880.504739]  [] ? get_parent_ip+0x9/0x1b
> [ 2880.505883]  [] ? wait_for_common+0x9d/0x10c
> [ 2880.507024]  [] ? default_wake_function+0x0/0xf
> [ 2880.508189]  [] ? stop_one_cpu+0x57/0x6e
> [ 2880.509288]  [] ? migration_cpu_stop+0x0/0x2a
> [ 2880.510354]  [] ? sub_preempt_count+0x83/0x94
> [ 2880.511426]  [] ? sched_exec+0xbe/0xd6
> [ 2880.512523]  [] ? do_execve+0xd1/0x28e
> [ 2880.513582]  [] ? sys_execve+0x3f/0x54
> [ 2880.514616]  [] ? stub_execve+0x6c/0xc0
> amars...@am64:~$ uname -a
> Linux am64 2.6.36-git16 #1 SMP PREEMPT Sun Oct 31 15:41:03 CST 2010
> x86_64 GNU/Linux
>
> The process was unblocked by logging another session into the AMD64 machine.
>
> The "task foo blocked for more than 120 seconds" has also occurred on
> the PII-266 uniprocessor machine with some of the 2.6.36-git9 or later
> kernels. Previously I had not had such a problem on the PII-266 for
> about 6 months.
>
> I can't begin to figure out where this problem was introduced.
> Git-bisection doesn't always work as it may take a while for the
> symptoms of a swap storm to appear.
>
> Is there any straightforward way to gather more information before
> reporting this kind of issue upstream?

Hmmm, I just sense, could be that the scheduler hackers once again
play with child-forked-then-run-first? Once I read it is child which
runs first, but lately I read it is the parent first...

I also think, it might be something related to a patch Fengguang Wu
made few weeks ago (IIRC) to reduce latency during high I/O... but I
can't recall which patch that is. Try to google...

NB: ftrace might help here please kindly read the ftrace tutorial
posted in lwn.net in few latest editions...


-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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: looking for corrections/enhancements/omissions for LKD3

[Mulyadi Santosa]

Hi Rob...

On Sun, Oct 31, 2010 at 19:34, Robert P. J. Day  wrote:
>  i'll see what i can do. :-)  anyway, this is where i'm going to
> track this stuff:

Well, that's just my wishlist, so it's up to you whether you wanna do
me a favour or not. BTW, have you tried to consult other Linux kernel
related list like kernel janitor too?


> http://www.crashcourse.ca/wiki/index.php/Updates_to_LKD3

yay, I am the first there :)

> this is a totally unofficial page, i just want to keep up with changes
> for when LKD4 comes out. :-)  so i might as well start documenting
> changes now.  and, as you can see, i'll give credit to all reporters.

Well, tracking changes are good IMO, but I suggest you just pick the
most significant ones.

The criterias could be something like:
- whether it affects latency
- whether it improves graphic card improvement
- whether it changes how scheduling works
etc

NB: Does LKD3 mention about Kernel Same page Merge (KSM)? No need to
be deep, but I think it's interesting to mention it since it will
reduce memory consumption done by virtual machine/hypervisor (KVM, for
example).

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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: looking for corrections/enhancements/omissions for LKD3

[Robert P. J. Day]

On Sun, 31 Oct 2010, Mulyadi Santosa wrote:

> On Sat, Oct 30, 2010 at 22:02, Robert P. J. Day  wrote:
> >
> >  as the technical editor for robert love's book "linux kernel
> > development (3rd ed)", i'm interested in collecting any observations
> > related to kernel features that were introduced so recently, they
> > never made it into LKD3, simply being too new.
>
>
> Not sure if it's a feature or not, but I saw in 2.6.36, anticipatory
> I/O scheduler is no longer exist. Is that true? And what could be
> the reason?

  i just checked the git log:

commit 492af6350a5ccf087e4964104a276ed358811458
Author: Jens Axboe 
Date:   Sat Oct 3 09:37:51 2009 +0200

block: remove the anticipatory IO scheduler

AS is mostly a subset of CFQ, so there's little point in still
providing this separate IO scheduler. Hopefully at some point we
can get down to one single IO scheduler again, at least this brings
us closer by having only one intelligent IO scheduler.

Signed-off-by: Jens Axboe 

so i'll make a note of that.  thanks, that's exactly the sort of thing
i'm looking for.


> Also, I saw XIP in ext2 filesystem. Quite neat..but again, why?

  hmmm ... not sure, i'll look into that.

rday

-- 


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

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

Re: looking for corrections/enhancements/omissions for LKD3

[Robert P. J. Day]

On Sun, 31 Oct 2010, Mulyadi Santosa wrote:

> On Sun, Oct 31, 2010 at 19:18, Robert P. J. Day  wrote:
> >    AS is mostly a subset of CFQ, so there's little point in still
> >    providing this separate IO scheduler. Hopefully at some point we
> >    can get down to one single IO scheduler again, at least this brings
> >    us closer by having only one intelligent IO scheduler.
> >
> >    Signed-off-by: Jens Axboe 
> >
> > so i'll make a note of that.  thanks, that's exactly the sort of thing
> > i'm looking for.
>
> No problem...can I get a free copy of LKD3 plus your original sign
> like "RPJD rocks!"? :)

  i'll see what i can do. :-)  anyway, this is where i'm going to
track this stuff:

http://www.crashcourse.ca/wiki/index.php/Updates_to_LKD3

this is a totally unofficial page, i just want to keep up with changes
for when LKD4 comes out. :-)  so i might as well start documenting
changes now.  and, as you can see, i'll give credit to all reporters.

rday

-- 


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

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

Re: T/TCP Implementation variables

[Jayapal Reddy]

Hi Bond,

The below link help in understand w.r.t code

http://www.6test.edu.cn/~lujx/linux_networking/index.html?page=0131777203_ch19lev1sec3.html

- Jayapal reddy

On Sun, Oct 31, 2010 at 4:08 PM, Rajat Sharma  wrote:

> One more:
>
> http://www.amazon.com/Linux-TCP-Networking-Embedded-Systems/dp/1584504811/ref=dp_ob_title_bk
>
> Well I am not a networking guy but just know the names of these two books.
>
> Rajat
>
>
> On Sun, Oct 31, 2010 at 12:56 PM, Kfir Lavi  wrote:
>
>> On Sun, Oct 31, 2010 at 8:21 AM, Rajat Sharma  wrote:
>> > Richard Stevens is a good book, but it was written with only Unix
>> versions
>> > in mind. The kernel implementation can differ from OS to OS. You can
>> refer
>> > to this book specifically for Linux:
>> >
>> > http://oreilly.com/catalog/9780596002558
>> >
>>
>> This book does not go above the IP layer.
>>
>> > Rajat
>> >
>> > On Sun, Oct 31, 2010 at 11:11 AM, Bond 
>> wrote:
>> >>
>> >> Hi,
>> >> I had try to read man page but there are no such entries on my system
>> >> and the Googled links do not seem to give me any useful information.
>> >> I am trying to understand the implementation of T/TCP in kernel.The
>> >> Network Programming Vol 3 of Richard Stevens Chapter 2 of that book
>> >> deals with these.
>> >> It talks of tcp_ccgen,tao_cc,tao_ccsent,I searched on
>> >> http://lxr.linux.no/#linux+v2.6.36/
>> >> for these functions but I got zero results.
>> >> I am trying to understand implementation of T/TCP in linux kernel can
>> >> someone point me to a useful link?
>> >>
>> >>
>> >> --
>> >> http://vger.kernel.org/vger-lists.html
>> >>
>> >> --
>> >> 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
>> >>
>> >
>> >
>>
>
>


-- 
Thanks & Regards,
ఊరడి జయపాల్ రెడ్డి
(Uradi Jayapal reddy)
9866560003

Re: looking for corrections/enhancements/omissions for LKD3

[Mulyadi Santosa]

On Sat, Oct 30, 2010 at 22:02, Robert P. J. Day  wrote:
>
>  as the technical editor for robert love's book "linux kernel
> development (3rd ed)", i'm interested in collecting any observations
> related to kernel features that were introduced so recently, they
> never made it into LKD3, simply being too new.


Not sure if it's a feature or not, but I saw in 2.6.36, anticipatory
I/O scheduler is no longer exist. Is that true? And what could be the
reason?

Also, I saw XIP in ext2 filesystem. Quite neat..but again, why?

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

--
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

swap storms on low memory i386, tasks blocked on i386 and amd64 for kernel > 2.6.36-git6

[Arthur Marsh]

Hi, first post to this list.

I have a PII-266 with 384 MiB RAM (maximum capacity for the machine) and
an AMD64 dual core with 4 GiB RAM, both running Debian unstable plus
some packages from experimental and custom kernels.

A typical load for the PII-266 is KDE 3.5.10 with konversation, icedove,
iceweasel, xmms, lynx, hp-systray, top and aptitude-curses.

This works with stock Debian kernels and custom kernels up to and
including 2.6.36-git6. Under heavy load, free RAM will hover around 5
MiB but audio will still play with a very occasional skip and all
applications are responsive.

With the newer kernels, e.g. 2.6.36-git9,10,11 (all with the deadline
scheduler) get into a swap storm with over 32 MiB RAM free and kswapd0
taking 10 percent or more of CPU time. Kernel 2.6.36-git15 with the cfq
scheduler managed to keep smaller applications like xmms and shells
running but also had excessive free RAM and kswapd0 taking more than 10
percent of CPU time.

On the AMD64 dual core when compiling kernels with CONCURRENCY_LEVEL=4,
I would sometimes get the build process pausing with errors like:

[ 2880.492025] INFO: task sh:10071 blocked for more than 120 seconds.
[ 2880.493165] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
disables this message.
[ 2880.494299] shD 8800cfc53780 0 10071   9092
0x
[ 2880.495433]  880128b214a0 0086 880128851b00
88012b67cba0
[ 2880.496612]  00013780 00013780 88011969dfd8
88011969c000
[ 2880.497765]  880128b21798 880128b217a0 880128b214a0
88011969dfd8
[ 2880.498925] Call Trace:
[ 2880.500096]  [] ? __switch_to+0x198/0x284
[ 2880.501262]  [] ? schedule_timeout+0x2d/0xd7
[ 2880.502418]  [] ? need_resched+0x1a/0x23
[ 2880.503564]  [] ? schedule+0x5b2/0x5c9
[ 2880.504739]  [] ? get_parent_ip+0x9/0x1b
[ 2880.505883]  [] ? wait_for_common+0x9d/0x10c
[ 2880.507024]  [] ? default_wake_function+0x0/0xf
[ 2880.508189]  [] ? stop_one_cpu+0x57/0x6e
[ 2880.509288]  [] ? migration_cpu_stop+0x0/0x2a
[ 2880.510354]  [] ? sub_preempt_count+0x83/0x94
[ 2880.511426]  [] ? sched_exec+0xbe/0xd6
[ 2880.512523]  [] ? do_execve+0xd1/0x28e
[ 2880.513582]  [] ? sys_execve+0x3f/0x54
[ 2880.514616]  [] ? stub_execve+0x6c/0xc0
amars...@am64:~$ uname -a
Linux am64 2.6.36-git16 #1 SMP PREEMPT Sun Oct 31 15:41:03 CST 2010
x86_64 GNU/Linux

The process was unblocked by logging another session into the AMD64 machine.

The "task foo blocked for more than 120 seconds" has also occurred on
the PII-266 uniprocessor machine with some of the 2.6.36-git9 or later
kernels. Previously I had not had such a problem on the PII-266 for
about 6 months.

I can't begin to figure out where this problem was introduced.
Git-bisection doesn't always work as it may take a while for the
symptoms of a swap storm to appear.

Is there any straightforward way to gather more information before
reporting this kind of issue upstream?

Has anyone experienced similar problems or read of similar reports?

Arthur.

--
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: T/TCP Implementation variables

[Rajat Sharma]

One more:
http://www.amazon.com/Linux-TCP-Networking-Embedded-Systems/dp/1584504811/ref=dp_ob_title_bk

Well I am not a networking guy but just know the names of these two books.

Rajat

On Sun, Oct 31, 2010 at 12:56 PM, Kfir Lavi  wrote:

> On Sun, Oct 31, 2010 at 8:21 AM, Rajat Sharma  wrote:
> > Richard Stevens is a good book, but it was written with only Unix
> versions
> > in mind. The kernel implementation can differ from OS to OS. You can
> refer
> > to this book specifically for Linux:
> >
> > http://oreilly.com/catalog/9780596002558
> >
>
> This book does not go above the IP layer.
>
> > Rajat
> >
> > On Sun, Oct 31, 2010 at 11:11 AM, Bond  wrote:
> >>
> >> Hi,
> >> I had try to read man page but there are no such entries on my system
> >> and the Googled links do not seem to give me any useful information.
> >> I am trying to understand the implementation of T/TCP in kernel.The
> >> Network Programming Vol 3 of Richard Stevens Chapter 2 of that book
> >> deals with these.
> >> It talks of tcp_ccgen,tao_cc,tao_ccsent,I searched on
> >> http://lxr.linux.no/#linux+v2.6.36/
> >> for these functions but I got zero results.
> >> I am trying to understand implementation of T/TCP in linux kernel can
> >> someone point me to a useful link?
> >>
> >>
> >> --
> >> http://vger.kernel.org/vger-lists.html
> >>
> >> --
> >> 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: T/TCP Implementation variables

[Kfir Lavi]

On Sun, Oct 31, 2010 at 8:21 AM, Rajat Sharma  wrote:
> Richard Stevens is a good book, but it was written with only Unix versions
> in mind. The kernel implementation can differ from OS to OS. You can refer
> to this book specifically for Linux:
>
> http://oreilly.com/catalog/9780596002558
>

This book does not go above the IP layer.

> Rajat
>
> On Sun, Oct 31, 2010 at 11:11 AM, Bond  wrote:
>>
>> Hi,
>> I had try to read man page but there are no such entries on my system
>> and the Googled links do not seem to give me any useful information.
>> I am trying to understand the implementation of T/TCP in kernel.The
>> Network Programming Vol 3 of Richard Stevens Chapter 2 of that book
>> deals with these.
>> It talks of tcp_ccgen,tao_cc,tao_ccsent,I searched on
>> http://lxr.linux.no/#linux+v2.6.36/
>> for these functions but I got zero results.
>> I am trying to understand implementation of T/TCP in linux kernel can
>> someone point me to a useful link?
>>
>>
>> --
>> http://vger.kernel.org/vger-lists.html
>>
>> --
>> 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
>>
>
>

--
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