https://bugzilla.novell.com/show_bug.cgi?id=710361 https://bugzilla.novell.com/show_bug.cgi?id=710361#c0 Summary: bogus cpu time in /proc/*/stat (ps/top/...) Classification: openSUSE Product: openSUSE 11.4 Version: Final Platform: x86-64 OS/Version: Other Status: NEW Severity: Major Priority: P5 - None Component: Kernel AssignedTo: kernel-maintainers@forge.provo.novell.com ReportedBy: koenig@linux.de QAContact: qa@suse.de Found By: --- Blocker: --- some PCs show processes with *very* large cpu times, like these $ uptime 15:10 73 Tage 0:41 an, 19 Benutzer, Durchschnittslast: 4,04, 4,05, 4,07 $ ps up 30598 2046 15116 USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND u1 2046 66.6 0.4 722852 37908 ? Sl May24 69052:36 kdeinit4: kded4 [kdeinit] u1 15116 72.7 0.2 361072 19856 ? Sl May31 68389:03 kdeinit4: kded4 [kdeinit] u1 30598 100 0.4 713360 37252 ? Sl May24 21206660:47 kdeinit4: kded4 [kdeinit] top gives different cpu values for the very huges numbers (21Mh;): $ top | head PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 2046 u1 20 0 705m 37m 22m S 95 0.5 69066:01 kded4 15269 u1 20 0 561m 39m 25m S 92 0.5 68430:20 knotify4 30598 u1 20 0 696m 36m 21m S 92 0.5 5125630h kded4 $ cat /proc/{30598,2046,15116}/stat 30598 (kded4) S 1 30595 30595 0 -1 4202560 8757 269 0 0 94307215484 1750919665747 0 0 20 0 4 0 7571414 730480640 9313 7073566720 1 1 0 0 0 0 0 4096 83177 18446744073709551615 0 0 17 2 0 0 0 0 0 2046 (kded4) S 1 2043 2043 0 -1 4202560 47118 268 0 0 294153632 120160116 0 0 20 0 4 0 9068034 740200448 9477 7073566720 1 1 0 0 0 0 0 4096 83177 18446744073709551615 0 0 17 2 0 0 0 0 0 15116 (kded4) S 1 15112 15112 0 -1 4202560 5951 384 2 0 289908260 120424120 0 0 20 0 2 0 67295382 369737728 4964 7073566720 1 1 0 0 0 0 0 4096 83177 18446744073709551615 0 0 17 1 0 0 5 0 0 or (different PC): $ uptime 15:11 72 Tage 21:01 an, 8 Benutzer, Durchschnittslast: 1,08, 1,09, 1,07 $ ps up 9298 USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND u1 9298 100 0.4 778820 35336 ? Sl Jun01 21206763:41 kdeinit4: kded4 [kdeinit] $ top | head PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 9298 u2 20 0 760m 34m 20m S 100 0.4 5125632h kded4 $ cat /proc/9298/stat 9298 (kded4) S 1 9295 9295 0 -1 4202560 9120 1970 0 0 178494528906 1666732974479 0 1 20 0 4 0 74483073 797511680 8834 7073566720 1 1 0 0 0 0 0 4096 83177 18446744073709551615 0 0 17 2 0 0 0 0 0 both PCs: $ uname -a Linux pc2 2.6.37.6-0.5-desktop #1 SMP PREEMPT 2011-04-25 21:48:33 +0200 x86_64 x86_64 x86_64 GNU/Linux the numbers from "ps" and top for the ~69kh cpu time does not match my manual calculations from the /proc/*/stat numbers, only top display for the 5Mh numbers is "correct" (== matches the proc values) : $ cat /proc/{30598,2046,15116}/stat | cut -d\ -f1,14,15 | while read pid u s ; do echo $pid : $u + $s = $[ ($u + $s) / 100 /3600 ] h ;done 30598 : 94307215484 + 1750919794114 = 5125630 h 2046 : 294244749 + 120196918 = 1151 h 15116 : 289997331 + 120461156 = 1140 h what's wrong with those kernel values for those processes ? why do "ps" and "top" show even weirder numbers then expected for that /proc/*/stat trash ? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.