kernel thread as real threads..

[Julian Elischer]

Kris Kennaway wrote:

>On Thu, Jan 19, 2006 at 06:56:17PM -0800, Julian Elischer wrote:
>  
>
>>some progrsss..
>>as the first few lines show, it's not quite perfect yet but it's most of 
>>the way there..
>>(Like proc 1 isn't init)
>>
>>SMP: AP CPU #1 Launched!
>>SMP: AP CPU #2 Launched!
>>SMP: AP CPU #3 Launched!
>>panic: blockable sleep lock (sleep mutex) buffer daemon lock @ 
>>/usr/src/sys/kern/vfs_bio.c:2048
>>cpuid = 2
>>KDB: enter: panic
>>[thread pid 0 tid 100051 ]
>>Stopped at      kdb_enter+0x2b: nop    
>>db> ps
>> pid   proc     uid  ppid  pgrp  flag   stat  wmesg    wchan  cmd
>>   1 c7cb1000    0     0     0 0000200 [SLPQ g_waitidle 
>>0xc07239ec][SLP] swapper
>>   0 c0723c40    0     0     0 0000288 (threaded)  swapper
>>  thread 0xc7d6a340 ksegrp 0xc7cb0960 [RUNQ] schedcpu
>>  thread 0xc7d6a4e0 ksegrp 0xc7cb09c0 [RUNQ] nfsiod 3
>>    
>>
>
>What about that threads don't show CPU usage or accumulate CPU time?
>This is annoying enough for user threads but would be a pretty serious
>usability limitation if it happened for kernel threads too.
>
>Kris
>  
>

the example I showed was the 'ps' from ddb which of course doesn't show 
any stats anyhow.

Well, stats are collected. It is just a case of getting them out and 
displayed..
It's kind of meaningless in KSE type threads but for libthr and kernel 
threads it would have meaning.

Anynow I just want to see what it takes to not have 40 extra proc 
structs hanging around.