mulyadi may be right......but anothere possibility was that the stracing introduced delays ....which made all the problem go away.....ie, asynchronous operation between reader and writer......which appear as "missing interrupt" (check internet for this - many causes) bottomline is why not introduce some delays...somewhere....from the crash dump stack trace u can possibly narrow it down.... and this is from anupam: On Fri, May 23, 2008 at 10:09 PM, Peter Teoh <htmldeveloper@xxxxxxxxx> wrote: > I like this joke!!!!! it is funny!!!! ahem ! i was kind of serious: http://en.wikipedia.org/wiki/Heisenbug anupam On Thu, May 22, 2008 at 2:05 PM, Mulyadi Santosa <mulyadi.santosa@xxxxxxxxx> wrote: > Hi... > > On Thu, May 22, 2008 at 12:39 PM, Peerless Deepak > <peerlessdriver@xxxxxxxxx> wrote: >> Hi all, >> >> I am wondered whether running a application with strace will make it >> behave differently. >> >> My application on a ARM target board crashed consistently. The odd >> thing was the log file also disappears during the mishap. >> >> So I started the application with strace and captured the output using >> a serial output. >> But now the application which was frequently crashing without strace >> , is not crashing. >> >> No change was there in the environment and the code. Does strace >> affects the way? > > most likely, your application will run slower due to effect of syscall > tracing (before entering kernel space and leaving kernel space back to > user space). > > This slowing effect, sometimes make bugs related to race condition > disappears....or could be worse. > > regards, > > Mulyadi. > > -- > To unsubscribe from this list: send an email with > "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx > Please read the FAQ at http://kernelnewbies.org/FAQ > > -- Regards, Peter Teoh -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
