Re: chromium port causing massive I/O faults

From: Alexander Best <arundel_at_freebsd.org>
Date: Wed, 27 Jul 2011 09:38:57 +0000
On Wed Jul 27 11, René Ladan wrote:
> 2011/7/27 Alexander Best <arundel_at_freebsd.org>:
> > On Wed Jul 27 11, René Ladan wrote:
> >> 2011/7/27 Gleb Kurtsou <gleb.kurtsou_at_gmail.com>:
> >> > On (27/07/2011 00:48), Alexander Best wrote:
> >> >> On Mon Jul 25 11, Matthias Andree wrote:
> >> >> > Am 25.07.2011 09:21, schrieb Alexander Best:
> >> >> > > On Mon Jul 25 11, Adrian Chadd wrote:
> >> >> > >> Is it perhaps doing disk IO using mmap?
> >> >> > >
> >> >> > > how can i check, whether that's the case or not?
> >> >> >
> >> >> > Use truss(1) for instance.
> >> >> >
> >> >> > However, unless there are *practical* problems, a high number of page
> >> >> > faults is not an indication for problems.  Although it may sound scary,
> >> >> > page faults are a feature of the memory management.
> >> >>
> >> >> unfortunately truss(1) is crashing chromium :( i opened up a new thread
> >> >> reagarding this issue on freebsd-current_at_.
> >> > Could you try ktrace? It works for me
> >> >
> >> >> another thing i noticed is the increase in system calls caused by chromium.
> >> >> let's have a look at hub.freebsd.org:
> >> >>
> >> >> uptime = 149 days
> >> >>
> >> >> and 'vmstat -s' reports:
> >> >>
> >> >> 2168697753 cpu context switches
> >> >> 2266220366 device interrupts
> >> >> 2902880931 software interrupts
> >> >> 3779075897 traps
> >> >> 902107847 system calls
> >> >>
> >> >> now on my box:
> >> >>
> >> >> uptime = 2 days
> >> >>
> >> >> and 'vmstat -s' reports:
> >> >>
> >> >> 1155995386 cpu context switches
> >> >> 164577882 device interrupts
> >> >> 189456976 software interrupts
> >> >> 137007580 traps
> >> >> 2178434582 system calls
> >> > About 2.5k syscalls with chrome + a lot of other stuff running. 1.5k
> >> > without chrome.
> >> >
> >> > Looks like there is a lot of clock_gettime and gettimeofday syscalls.
> >> > ~ % kdump -m 1 -f ktrace.out | grep 'CALL .*gettime' | wc -l
> >> >   24343
> >> >
> >> > ~ % kdump -E -m 1 -f ktrace.out | grep 'CALL .*gettime' | tail -20
> >> >  12747 chrome   15.077376 CALL  gettimeofday(0x7fffff7f9630,0x7fffff7f9640)
> >> >  12747 chrome   15.077396 CALL  clock_gettime(0x4,0x7fffffbfb6f0)
> >> >  12747 chrome   15.077497 CALL  gettimeofday(0x7fffffbfb650,0x7fffffbfb660)
> >> >  12747 chrome   15.077609 CALL  gettimeofday(0x7fffffbfb650,0x7fffffbfb660)
> >> >  12747 chrome   15.077723 CALL  gettimeofday(0x7fffffbfb650,0)
> >> >  12747 chrome   15.077845 CALL  clock_gettime(0,0x7fffffbfb2b0)
> >> >  12747 chrome   15.078337 CALL  clock_gettime(0x4,0x7fffff9fa630)
> >> >  12747 chrome   15.078544 CALL  clock_gettime(0x4,0x7fffff9fa650)
> >> >  12747 chrome   15.078587 CALL  clock_gettime(0x4,0x7fffff9fa650)
> >> >  12747 chrome   15.078632 CALL  clock_gettime(0x4,0x7fffff9fa650)
> >> >  12747 chrome   15.078674 CALL  clock_gettime(0x4,0x7fffff9fa650)
> >> >  12747 chrome   15.082803 CALL  gettimeofday(0x7ffffedd3630,0x7ffffedd3640)
> >> >  12747 chrome   15.084644 CALL  gettimeofday(0x7fffffbfb650,0x7fffffbfb660)
> >> >  12747 chrome   15.084746 CALL  clock_gettime(0x4,0x7fffffbfb670)
> >> >  12747 chrome   15.084815 CALL  clock_gettime(0x4,0x7fffffbfb670)
> >> >  12747 chrome   15.086620 CALL  gettimeofday(0x7ffffefd4650,0x7ffffefd4660)
> >> >  12747 chrome   15.086736 CALL  clock_gettime(0x4,0x7ffffefd4670)
> >> >  12747 chrome   15.086815 CALL  clock_gettime(0x4,0x7ffffefd4670)
> >> >  12747 chrome   15.098315 CALL  gettimeofday(0x7fffffffafe0,0x7fffffffaff0)
> >> >  12747 chrome   15.098680 CALL  clock_gettime(0x4,0x7fffffffb250)
> >> >
> >> > Some work was done by kib_at_ to create a kernel page strong current time
> >> > and other misc info to eliminate gettimeofday kind syscalls.  Bits of it
> >> > were commited but I'm not sure if it was finished.
> >> > But anyway calling gettimeofday hundreds of times per second is a chrome
> >> > bug.

...also the number of context switches is very high. the following 'vmstat -s'
output was taken after only 32 minutes of uptime and chromium running for ~ 10
minutes:

 39775038 cpu context switches
  1716910 device interrupts
  1707161 software interrupts
  1764371 traps
 57319358 system calls
       15 kernel threads created
     2120  fork() calls
       11 vfork() calls
       25 rfork() calls
        0 swap pager pageins
        0 swap pager pages paged in
        0 swap pager pageouts
        0 swap pager pages paged out
    71184 vnode pager pageins
   102181 vnode pager pages paged in
    13321 vnode pager pageouts
    67437 vnode pager pages paged out
        0 page daemon wakeups
        0 pages examined by the page daemon
     4662 pages reactivated
    93964 copy-on-write faults
      274 copy-on-write optimized faults
   358563 zero fill pages zeroed
      319 zero fill pages prezeroed
      302 intransit blocking page faults
   740518 total VM faults taken
        0 pages affected by kernel thread creation
  1130760 pages affected by  fork()
    17316 pages affected by vfork()
    22319 pages affected by rfork()
     7162 pages cached
   693935 pages freed
        0 pages freed by daemon
   396060 pages freed by exiting processes
    34690 pages active
    88551 pages inactive
      164 pages in VM cache
    76703 pages wired down
   301738 pages free
     4096 bytes per page
   426219 total name lookups
          cache hits (87% pos + 2% neg) system 0% per-directory
          deletions 2%, falsehits 0%, toolong 0%

with chromium running:

otaku% vmstat -s|grep "context switches"; sleep 1; vmstat -s|grep "context switches"
 39604188 cpu context switches
 39610679 cpu context switches

without:

otaku% vmstat -s|grep "context switches"; sleep 1; vmstat -s|grep "context switches"
 39722188 cpu context switches
 39722998 cpu context switches

cheers.
alex

> >> >
> >> > FreeBSD 9.0-CURRENT #2 r224003+777e962: Thu Jul 14 13:04:55 EEST 2011
> >> > chromium-11.0.696.57_1
> >>                  ^^^^^^^^^^^^^
> >> Can you retry with an up-to-date version of www/chromium?  The
> >> codebase of chromium
> >> changes quite fast so not using the latest version in ports might
> >> render obsolete (and
> >> upstream unsupported) results.
> >
> > my tests were done with chromium-12.0.742.124 btw.
> >
> Ok, I'll do some tests with the beta version from the chruetertee
> repository (13.0.782.99).
> >>
> >> René
> >> --
> >> http://www.rene-ladan.nl/
> >
Received on Wed Jul 27 2011 - 07:38:57 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:16 UTC