On Mon, Mar 07, 2016 at 12:38:16PM -0500, Chris Metcalf wrote:
> On 03/07/2016 04:48 AM, Peter Zijlstra wrote:
> I'm a little skeptical that a single percpu write is going to add much
> measurable overhead to this path.
So that write is almost guaranteed to be a cacheline miss, those things
hurt
On 03/07/2016 04:48 AM, Peter Zijlstra wrote:
On Tue, Mar 01, 2016 at 11:01:42AM -0500, Chris Metcalf wrote:
+++ b/kernel/sched/idle.c
@@ -52,15 +52,25 @@ static int __init cpu_idle_nopoll_setup(char *__unused)
__setup("hlt", cpu_idle_nopoll_setup);
#endif
+static DEFINE_PER_CPU(bool, cpu_id
On 03/07/2016 03:26 AM, Daniel Thompson wrote:
Chris Metcalf wrote:
+static DEFINE_PER_CPU(bool, cpu_idling);
+
+/* Was the cpu was in the low-level idle code when interrupted? */
+bool in_cpu_idle(void)
+{
+return this_cpu_read(cpu_idling);
I think we continue to need the code to identify
On Tue, Mar 01, 2016 at 11:01:42AM -0500, Chris Metcalf wrote:
> +++ b/kernel/sched/idle.c
> @@ -52,15 +52,25 @@ static int __init cpu_idle_nopoll_setup(char *__unused)
> __setup("hlt", cpu_idle_nopoll_setup);
> #endif
> +static DEFINE_PER_CPU(bool, cpu_idling);
> +
> +/* Was the cpu was in the l
On 01/03/16 23:01, Chris Metcalf wrote:
(+PeterZ, Rafael, and Daniel Lezcano for cpuidle and scheduler)
On 03/01/2016 09:23 AM, Daniel Thompson wrote:
On 29/02/16 21:40, Chris Metcalf wrote:
When doing an nmi backtrace of many cores, and most of them are idle,
the output is a little overwhelmi
(+PeterZ, Rafael, and Daniel Lezcano for cpuidle and scheduler)
On 03/01/2016 09:23 AM, Daniel Thompson wrote:
On 29/02/16 21:40, Chris Metcalf wrote:
When doing an nmi backtrace of many cores, and most of them are idle,
the output is a little overwhelming and very uninformative. Suppress
messa
On 29/02/16 21:40, Chris Metcalf wrote:
When doing an nmi backtrace of many cores, and most of them are idle,
the output is a little overwhelming and very uninformative. Suppress
messages for cpus that are idling when they are interrupted and
just emit one line, "NMI backtrace for N skipped: idl
When doing an nmi backtrace of many cores, and most of them are idle,
the output is a little overwhelming and very uninformative. Suppress
messages for cpus that are idling when they are interrupted and
just emit one line, "NMI backtrace for N skipped: idle".
Signed-off-by: Chris Metcalf
---
li
8 matches
Mail list logo
