[lkml]   [2014]   [Jul]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH 4/6] x86-mce: Add spinlocks to prevent duplicated MCP and CMCI reports.
On Thu, Jul 10, 2014 at 11:03:43AM -0700, Havard Skinnemoen wrote:
> For non-shared banks, we might risk some CPUs not being able to
> poll their banks in a long time if they happen to be more or less
> synchronized with a different CPU. This will also get worse with
> shorter polling intervals, and with larger numbers of CPUs.

No, I meant to do something like

if (atomic_dec_and_test(&mce_banks[i].poll))
m.status = mce_rdmsrl(MSR_IA32_MCx_STATUS(i));

atomic_add_unless(&mce_banks[i].poll, 1, 1);

so that you have only one CPU read the status register of mce_banks[i].

For non-shared banks, this will always work because no other CPU will
dec that variable anyway.

Or am I missing something...?


Sent from a fat crate under my desk. Formatting is fine.

 \ /
  Last update: 2014-07-11 04:41    [W:0.116 / U:29.764 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site