Re: pdflush preemption

[John Sigler <linux.kernel@free.fr>]

Daniel Walker wrote:

> John Sigler wrote:
> 
>> Why does pdflush kick in to ruin my party? :-)
>> The expected latency is ~600 µs.
>> http://linux.kernel.free.fr/latency/pdflush.trace
>> Does ide_inb mean I'm reading from the disk?
> 
> Does your real time application lock its memory, or allow itself to be
> swapped?

Now that you mention it, I do remember reading this page:
http://rt.wiki.kernel.org/index.php/HOWTO:_Build_an_RT-application

There is no swap configured on the system, and 90% of the 256 MB of RAM 
are free, according to /proc/meminfo, and the system is otherwise idle.

I don't think this is a major page fault. Am I wrong?

# /bin/time ./check_dektec_output2 4140063957 38000000 2632 30
0.02user 0.31system 0:20.02elapsed 1%CPU
(0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (0major+270minor)pagefaults 0swaps
                   ^^^^^^

# cat /proc/vmstat
nr_free_pages 59351
nr_inactive 1465
nr_active 1830
nr_anon_pages 486
nr_mapped 569
nr_file_pages 2813
nr_dirty 0
nr_writeback 0
nr_slab_reclaimable 449
nr_slab_unreclaimable 663
nr_page_table_pages 41
nr_unstable 0
nr_bounce 0
nr_vmscan_write 0
pgpgin 8197
pgpgout 113
pswpin 0
pswpout 0
pgalloc_dma 0
pgalloc_normal 27260
pgfree 86682
pgactivate 1372
pgdeactivate 0
pgfault 77490
pgmajfault 121
pgrefill_dma 0
pgrefill_normal 0
pgsteal_dma 0
pgsteal_normal 0
pgscan_kswapd_dma 0
pgscan_kswapd_normal 0
pgscan_direct_dma 0
pgscan_direct_normal 0
pginodesteal 0
slabs_scanned 0
kswapd_steal 0
kswapd_inodesteal 0
pageoutrun 0
allocstall 0
pgrotated 0

Regards.