Messages in this thread |  | | | Date | Sat, 27 Feb 2010 09:10:24 +0100 (CET) | | From | Thomas Gleixner <> | | Subject | Re: [ANNOUNCE] 2.6.33-rc8-rt2 |
| |
On Fri, 26 Feb 2010, Dan Carpenter wrote: > kernel/trace/latency_hist.c > 341 static ssize_t > 342 latency_hist_reset(struct file *file, const char __user *a, > 343 size_t size, loff_t *off) > 344 { > 345 int cpu; > 346 struct hist_data *hist = NULL; > 347 #if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ > 348 defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) > 349 struct maxlatproc_data *mp = NULL; > 350 #endif > 351 off_t latency_type = (off_t) file->private_data; > 352 > 353 for_each_online_cpu(cpu) { > 354 > 355 switch (latency_type) { > > It seems like this switch statement should have a default case. > > I haven't looked into it very closely but it seems like if you had all > those config options turned off then "hist" would be still be NULL.
When all config options are turned off then the file is not compiled at all.
Thanks,
tglx
|  |