Re: OOM in v4.8

[Aaron Lu <aaron.lu@intel.com>]

[-- Attachment #1: Type: text/plain, Size: 1209 bytes --]

On Wed, Oct 12, 2016 at 10:43:42AM +0200, Michal Hocko wrote:
> On Wed 12-10-16 16:24:47, Aaron Lu wrote:
> > On 10/12/2016 04:00 PM, Michal Hocko wrote:
> [...]
> > > which is an atomic high order request that failed which is not all that
> > > unexpected when the system is low on memory. The allocation failure
> > > report is hard to read because of unexpected end-of-lines but I suspect
> > 
> > Sorry about that, I'll try to find out why dmesg is saved so ugly on
> > that test box.
> 
> Not your fault. This seems to be 4bcc595ccd80 ("printk: reinstate
> KERN_CONT for printing continuation lines")
> 
> > > that again we are not able to allocate because of the CMA standing in
> > > the way. I wouldn't call the above failure critical though.
> >  
> > I'll test that commit and v4.8 again with cma=0 added to cmdline.
> 
> Thanks!

With cma=0:
1 on v4.8, 8 tests OOMed out of 10 tests;
2 on 101105b1717f, 1 test OOMed out of 10 tests as before.

It seems to be worse for v4.8, previouslly it's 6 failures.
For 101105b1717f, it's the same case: perf requested a order 2 atomic
allocation and failed, no OOM killer is invoked.

Both dmesgs are attached.

Thanks,
Aaron

[-- Attachment #2: v4.8.xz --]
[-- Type: application/x-xz, Size: 26780 bytes --]

[-- Attachment #3: 101105b1717f.xz --]
[-- Type: application/x-xz, Size: 21928 bytes --]