On 3/7/2014 3:57 PM, Marcy Cortes wrote:
weird, could the combo of the two of them be to blame?
I'm sure they got it right. Especially since that bit of code has been effectively reverted in upstream as well at later kernel versions and why I wasn't able to make it happen in a 3.13.5 kernel.
The method I used to try and figure out the cause probably changed things enough where the real problem was simply bypassed in some way. I can honestly say that the vmscan code is WAY over my head, so I was really just taking shots in the dark. Anything to try and push for a resolution to the problem. My method for recreating the problem was a little C program that simply allocated a bunch of memory over and over while I'm dropping the caches in another ssh session. Worked very reliably. But, with the patch reverted, I am no longer able to recreate the issue after many attempts. Thanks again, Leland ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
