Is the adb shell's version of 'ps' that different from the standard? The standard 'ps' shows -t being for 'terminal', not for 'thread'. 'ps -eLf' is the standard way to get thread info.
On Sep 4, 7:17 pm, Ubuntu guy <[email protected]> wrote: > adb shell ps -t <browser_pid> should give thread info > > On Sep 2, 6:21 am, Jack Harvard <[email protected]> wrote: > > > Thanks a lot, unfortunately we use BusyBox as the filesystem, the 'top' > > command options are slightly different, 'top -b' got me something as > > follows, but would like to get something as you did (suppose you use ps > > command, i couldn't get exactly the same output as yours, would be nice to > > have the function names printed out). > > > Mem: 189324K used, 65772K free, 0K shrd, 864K buff, 78104K cached > > CPU: 50.0% usr 0.0% sys 0.0% nic 50.0% idle 0.0% io 0.0% irq 0.0% sirq > > Load average: 0.18 0.17 0.09 4/298 1085 > > PID PPID USER STAT VSZ %MEM CPU %CPU COMMAND > > 728 727 1000 S 136m 54.7 2 0.0 system_server > > 1027 727 10030 R 128m 51.2 3 0.0 com.android.browser > > 786 727 1001 S 86028 33.6 0 0.0 com.android.phone > > 868 727 10005 S 83568 32.6 0 0.0 android.process.acore > > 960 727 10029 S 83516 32.6 1 0.0 com.android.mms > > 779 727 10012 S 78344 30.6 0 0.0 com.android.inputmethod.latin > > 787 727 1000 S 76668 29.9 1 0.0 com.android.systemui > > 788 727 10020 S 76120 29.7 1 0.0 com.android.launcher > > 932 727 10016 S 73868 28.8 0 0.0 com.android.email > > 1002 727 10027 S 73584 28.7 2 0.0 com.cooliris.media > > 905 727 10003 S 73176 28.5 3 0.0 android.process.media > > 916 727 10006 S 72168 28.1 1 0.0 com.android.deskclock > > 945 727 10025 S 72040 28.1 3 0.0 com.android.providers.calendar > > 985 727 10013 S 71856 28.0 0 0.0 com.android.quicksearchbox > > 899 727 10002 S 71448 27.9 1 0.0 com.android.bluetooth > > 977 727 10008 S 71384 27.8 0 0.0 com.android.music > > 993 727 10014 S 70820 27.6 0 0.0 com.android.protips > > > On 2 Sep 2011, at 00:11, Dianne Hackborn wrote: > > > > Many of those probably aren't even browser threads, but threads for > > > Dalvik and the application framework. For example, here is a list of > > > threads in my browser: > > > > 1 7035 native 79 21 main > > > *2 7038 vmwait 10 0 GC > > > *3 7040 vmwait 0 0 Signal Catcher > > > *4 7041 running 0 5 JDWP > > > *5 7042 vmwait 13 7 Compiler > > > *6 7043 wait 0 0 ReferenceQueueDaemon > > > *7 7044 wait 0 0 FinalizerDaemon > > > *8 7045 timed-wait 0 0 FinalizerWatchdogDaemon > > > 9 7046 native 0 0 Binder Thread #1 > > > 10 7047 native 0 0 Binder Thread #2 > > > 11 7057 wait 0 0 pool-2-thread-1 > > > 12 7050 native 0 0 CookieSyncManager > > > 13 7056 timed-wait 2 0 pool-1-thread-2 > > > 14 7058 native 322 30 WebViewCoreThread > > > 15 7053 native 2 1 BackgroundHandler > > > 16 7054 timed-wait 0 0 pool-1-thread-1 > > > 17 7059 wait 1 0 AsyncTask #1 > > > 18 7062 native 8 1 TexturesGenerator > > > *19 7063 wait 10 1 Thread-183 > > > 20 7067 native 0 0 Thread-185 > > > 21 7066 native 6 0 Thread-186 > > > 22 7070 wait 10 1 AsyncTask #2 > > > 23 7074 native 18 51 Thread-188 > > > 24 7076 native 0 0 Thread-189 > > > 25 7077 wait 0 0 AsyncTask #3 > > > 26 7078 wait 10 1 AsyncTask #4 > > > 27 7079 wait 11 0 AsyncTask #5 > > > > WebViewCoreThread is I believe the main browser thread. You can easily > > > see what threads are actually using CPU with "adb shell top -m 10 -t" > > > while you are using the browser. > > > > On Thu, Sep 1, 2011 at 12:15 PM, Jack Harvard <[email protected]> > > > wrote: > > > Thanks for your reply. I'm on the research side, rather than development. > > > I observed 17-22 threads from com.android.browser, suppose only one of > > > those thread is doing the real work of rendering the page. > > > > On 31 Aug 2011, at 22:58, Dianne Hackborn wrote: > > > > > I really don't know what you are trying to get at. As I said, as far > > > > as I know browser CPU use is generally pretty single-threaded -- the > > > > DOM rendering and JavaScript interpreter all need to run in the same > > > > thread. Just creating more threads doesn't magically allow this work > > > > to be spread across them. > > > > > And again, Gingerbread was not written for multi-core CPUs. In fact if > > > > you try to run stock GB on a multi-core CPU you will be extremely > > > > unstable because there are many things not working correctly for SMP, > > > > especially ARM SMP. So it doesn't make sense to ask questions about > > > > how stock GB runs on multi-core, because it just doesn't. > > > > > If you are wondering about a particular phone someone has released with > > > > GB, you will need to talk with that manufacturer, because they will > > > > have needed to do some work to have GB run on a multi-core CPU. > > > > > On Wed, Aug 31, 2011 at 2:09 PM, Jack Harvard <[email protected]> > > > > wrote: > > > > Thanks for your reply, Dianne. If I run the Android browser as my only > > > > application in Gingerbread on a quad-core smartphone, and the page to > > > > be loaded is stored locally in memory (no network delay etc, in this > > > > case, the browser is single-threaded only as no network connection?), I > > > > wouldn't see any performance scaling (or worse, as the OS may move the > > > > threads to another core less busy in a circle) from a single-core > > > > smartphone, everything else being the same? > > > > > On 31 Aug 2011, at 21:00, Dianne Hackborn wrote: > > > > > > Gingerbread isn't designed for multi-core CPUs. The browser doesn't > > > > > try to create threads to use multiple CPUs, it creates threads to > > > > > have multiple network connections. > > > > > > Besides networking, browsers are pretty intrinsically single-threaded > > > > > anyway. The complicated stuff all needs to run in one thread: DOM > > > > > rendering, JavaScript, etc. > > > > > > On Wed, Aug 31, 2011 at 12:08 PM, Jack Harvard > > > > > <[email protected]> wrote: > > > > > I want to understand how many threads Gingerbread generates in order > > > > > to understand how the browser performance scale with multicores? For > > > > > example, if "a few " means 2, that means 4 cores wouldn't scale the > > > > > browser performance by anywhere close to 4 times. I understand that > > > > > Gingerbread doesn't officially support SMP, sometimes browser does > > > > > crash on multicore platforms, it's not a problem for me though. > > > > > > On 31 Aug 2011, at 18:57, Dianne Hackborn wrote: > > > > > > > Oh and Gingerbread wasn't designed to run on anything but single > > > > > > CPU systems, so that's irrelevant to the question of the base > > > > > > platform. Of course people shipping Android modify it in various > > > > > > ways, and the browser is often one of the things they modify. > > > > > > > On Wed, Aug 31, 2011 at 10:57 AM, Dianne Hackborn > > > > > > <[email protected]> wrote: > > > > > > A few. Why? :) > > > > > > > On Wed, Aug 31, 2011 at 10:40 AM, Jack Harvard > > > > > > <[email protected]> wrote: > > > > > > Wondering how many threads does android browser generate? Suppose > > > > > > it's Gingerbread running on a dual-core or quad-core ARM processor. > > > > > > > -- > > > > > > You received this message because you are subscribed to the Google > > > > > > Groups "Android Developers" group. > > > > > > To post to this group, send email to > > > > > > [email protected] > > > > > > To unsubscribe from this group, send email to > > > > > > [email protected] > > > > > > For more options, visit this group at > > > > > >http://groups.google.com/group/android-developers?hl=en > > > > > > > -- > > > > > > Dianne Hackborn > > > > > > Android framework engineer > > > > > > [email protected] > > > > > > > Note: please don't send private questions to me, as I don't have > > > > > > time to provide private support, and so won't reply to such > > > > > > e-mails. All such questions should be posted on public forums, > > > > > > where I and others can see and answer them. > > > > > > > -- > > > > > > Dianne Hackborn > > > > > > Android framework engineer > > > > > > [email protected] > > > > > > > Note: please don't send private questions to me, as I don't have > > > > > > time to provide private support, and so won't reply to such > > > > > > e-mails. All such questions should be posted on public forums, > > > > > > where I and others can see and answer them. > > > > > > > -- > > > > > > You received this message because you are subscribed to the Google > > > > > > Groups "Android Developers" group. > > > > > > To post to this group, send email to > > > > > > [email protected] > > > > > > To unsubscribe from this group, send email to > > > > > > [email protected] > > > > > > For more options, visit this group at > > > > > >http://groups.google.com/group/android-developers?hl=en > > > > > > -- > > > > > You received this message because you are subscribed to the Google > > > > > Groups "Android Developers" group. > > > > > To post to this group, send email to > > > > > [email protected] > > > > > To unsubscribe from this group, send email to > > > > > [email protected] > > > > > For more options, visit this group at > > > > >http://groups.google.com/group/android-developers?hl=en > > > > > > -- > > > > > Dianne Hackborn > > > > > Android framework engineer > > > > > [email protected] > > > > > > Note: please don't send > > ... > > read more » -- You received this message because you are subscribed to the Google Groups "Android Developers" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-developers?hl=en
