Messages in this thread | | | Date | Fri, 4 Mar 2016 20:22:29 +0800 | From | zhong jiang <> | Subject | Re: A oops occur when it calls kmem_cache_alloc |
| |
On 2016/3/4 18:45, Tetsuo Handa wrote: > On 2016/03/04 17:36, zhong jiang wrote: >> The vmcore file show the collapse reason that the page had been removed >> when we acqure the page and prepare to remove the page from the slub >> partial list. >> >> The list is protected by the spin_lock from concurrent operation. And I find >> that other core is wating the lock to alloc memory. Therefore , The concurrent >> access should be impossible. >> >> what situatios can happen ? or it is a kernel bug potentially. This question >> almost impossible to produce again. The following is the call statck belonging to >> the module. > No kernel version, no clue. > > > Sorry, The kernel version is 3.4 stable version. And The mainline , by comparison, have no relative modification.
Thanks zhongjiang
| |