On Fri, Sep 12, 2014 at 3:02 PM, Paul E. McKenney <paul...@linux.vnet.ibm.com> wrote: > On Mon, Sep 01, 2014 at 04:44:04PM +0800, Fengguang Wu wrote: >> Greetings, >> >> 0day kernel testing robot got the below dmesg and the first bad commit is >> >> git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master >> commit d860d40327dde251d508a234fa00bd0d90fbb656 >> Author: Shan Wei <davids...@tencent.com> >> AuthorDate: Thu Jun 19 14:12:44 2014 -0700 >> Commit: Paul E. McKenney <paul...@linux.vnet.ibm.com> >> CommitDate: Wed Jul 9 09:15:21 2014 -0700 > > Hearing nothing from Shan or Pranith... > > So, I am not seeing this failure in my testing, but my best guess is > that the problem is due to the fact that force_quiescent_state() is > sometimes invoked with preemption enabled, which breaks __this_cpu_read() > though perhaps with very low probability. The common-case call (from > __call_rcu_core()) -does- have preemption disabled, in fact, it has > interrupts disabled. >> So unless I hear otherwise, I will simply revert this commit.
I missed this report as I was not CC'ed and this is the first time I am seeing this. As Christoph said later in the thread, it really is not clear how this change is triggering the bug. The tracer testing triggers this bug which is a corrupt stack and we see no force_quiescent_state() in the back trace. So may be this is exposing a bug somewhere else? Not really sure how to look at this. > [ 0.420978] Testing tracer branch: > [ 0.421701] BUG: unable to handle kernel NULL pointer dereference at > 000000da > [ 0.422857] IP: [<c1061074>] update_curr+0x1a3/0x2c3 > [ 0.423639] *pdpt = 0000000000000000 *pde = f000ff53f000ff53 > [ 0.424000] Thread overran stack, or stack corrupted -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/