[Skiboot] [PATCH] core/cpu: Initialize all cpu thread areas to avoid invalid memory access.
Mahesh Jagannath Salgaonkar
mahesh at linux.vnet.ibm.com
Thu Sep 6 16:43:12 AEST 2018
On 09/06/2018 10:55 AM, Vasant Hegde wrote:
>> I'm still a little concerned that skiboot is putting junk data in the
>> unused cpu_thread structures at all. If it is then I'd expect
>> cpu_for_each*() to be broken at broken at runtime even if it boots
>> successfully from a cold IPL. I'll take a look at it sometime.
> Yesterday we spent sometime to track down the real issue. It turned out
> that Hostboot
> copied 16MB HDAT data to OPAL memory instead of 8MB.. That corrupted our
> TCE space
> and CPU stack.
Thanks Vasant for tracking this down.
> Hostboot upstream has required commit... that needs to be backported to
> relevant FSP builds.
So with the hostboot fix mentioned below we should not have any
corruption in CPU stack. With that I am wondering whether this patch is
required at all.
> FYI : HB commit :
> commit 3d3d39d62a94da9dc9bc2da73474c9c3400762c4
> Author: Mike Baiocchi <mbaiocch at us.ibm.com>
> Date: Thu May 3 09:02:02 2018 -0500
More information about the Skiboot