[Bugme-janitors] [Bug 10656] dmesg timestamps to small and jumping back and forth. Regression from 2.6.24

[bugme-daemon at bugzilla.kernel.org]

http://bugzilla.kernel.org/show_bug.cgi?id=10656





------- Comment #5 from spamtrap at knobisoft.de  2008-05-09 06:07 -------
It seemes that the following change to kernel/printk.c causes the problems:

@@ -680,7 +742,7 @@
                                        loglev_char = default_message_loglevel
                                                + '0';
                                }
-                               t = printk_clock();
+                               t = cpu_clock(printk_cpu);
                                nanosec_rem = do_div(t, 1000000000);
                                tlen = sprintf(tbuf,
                                                "<%c>[%5lu.%06lu] ",


 In 2.6.24, printk_clock() translates into sched_clock(). If I call
sched_clock() in 2.6.25 instead of cpu_clock(), the dmesg timestamps are back
to the expected behaviour. So, something in cpu_clock() seems to be broken.

 Hope this helps.

Martin
PS: Ignore the NFS comment on the 2.6.24 config. It was ment for the 2.6.24
dmesg :-(


-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.