__get_cpu_var() is used for multiple purposes in the kernel source. One of them is
address calculation via the form &__get_cpu_var(x). This calculates the address for
the instance of the percpu variable of the current processor based on an offset.
Other use cases are for storing and retrieving data from the current processors percpu area.
__get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store and retrieve operations
could use a segment prefix (or global register on other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use
optimized assembly code to read and write per cpu variables.
This patch converts __get_cpu_var into either an explicit address calculation using this_cpu_ptr()
or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided
and less registers are used when code is generated.
At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too.
The patchset includes passes over all arches as well. Once these operations are used throughout then
specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by
f.e. using a global register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu variable.
DEFINE_PER_CPU(int, u);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(this_cpu_ptr(&x), y, sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
this_cpu_inc(y)
Acked-by: Vineet Gupta <[email protected]>
Signed-off-by: Christoph Lameter <[email protected]>
Index: linux/arch/arc/kernel/kprobes.c
===================================================================
--- linux.orig/arch/arc/kernel/kprobes.c 2013-08-26 13:28:53.000000000 -0500
+++ linux/arch/arc/kernel/kprobes.c 2013-08-26 13:29:19.285103021 -0500
@@ -87,13 +87,13 @@ static void __kprobes save_previous_kpro
static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
{
- __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
+ __this_cpu_write(current_kprobe, kcb->prev_kprobe.kp);
kcb->kprobe_status = kcb->prev_kprobe.status;
}
static inline void __kprobes set_current_kprobe(struct kprobe *p)
{
- __get_cpu_var(current_kprobe) = p;
+ __this_cpu_write(current_kprobe, p);
}
static void __kprobes resume_execution(struct kprobe *p, unsigned long addr,
@@ -237,7 +237,7 @@ int __kprobes arc_kprobe_handler(unsigne
return 1;
} else if (kprobe_running()) {
- p = __get_cpu_var(current_kprobe);
+ p = __this_cpu_read(current_kprobe);
if (p->break_handler && p->break_handler(p, regs)) {
setup_singlestep(p, regs);
kcb->kprobe_status = KPROBE_HIT_SS;
Index: linux/arch/arc/kernel/time.c
===================================================================
--- linux.orig/arch/arc/kernel/time.c 2013-08-26 13:28:53.000000000 -0500
+++ linux/arch/arc/kernel/time.c 2013-08-26 13:29:19.285103021 -0500
@@ -206,7 +206,7 @@ static DEFINE_PER_CPU(struct clock_event
static irqreturn_t timer_irq_handler(int irq, void *dev_id)
{
- struct clock_event_device *clk = &__get_cpu_var(arc_clockevent_device);
+ struct clock_event_device *clk = this_cpu_ptr(&arc_clockevent_device);
arc_timer_event_ack(clk->mode == CLOCK_EVT_MODE_PERIODIC);
clk->event_handler(clk);
On 08/29/2013 01:18 AM, Christoph Lameter wrote:
> Acked-by: Vineet Gupta <[email protected]>
> Signed-off-by: Christoph Lameter <[email protected]>
Applied to for-next.
Thx,
-Vineet
On Thu, 29 Aug 2013, Vineet Gupta wrote:
> On 08/29/2013 01:18 AM, Christoph Lameter wrote:
> > Acked-by: Vineet Gupta <[email protected]>
> > Signed-off-by: Christoph Lameter <[email protected]>
>
> Applied to for-next.
Ok dropping it from patchset.
Hi Christoph,
On 08/29/2013 10:13 PM, Christoph Lameter wrote:
> On Thu, 29 Aug 2013, Vineet Gupta wrote:
>
>> On 08/29/2013 01:18 AM, Christoph Lameter wrote:
>>> Acked-by: Vineet Gupta <[email protected]>
>>> Signed-off-by: Christoph Lameter <[email protected]>
>>
>> Applied to for-next.
>
> Ok dropping it from patchset.
>
Considering other discussions on this thread, shall I drop this from my for-curr
for this merge window ? I don't see any other arch changes to that effect in
latest linux-next.
-Vineet
On Wed, 4 Sep 2013, Vineet Gupta wrote:
> Considering other discussions on this thread, shall I drop this from my for-curr
> for this merge window ? I don't see any other arch changes to that effect in
> latest linux-next.
The required fix of the macros was merged yesterday into upstream.
On 09/04/2013 07:44 PM, Christoph Lameter wrote:
> On Wed, 4 Sep 2013, Vineet Gupta wrote:
>
>> Considering other discussions on this thread, shall I drop this from my for-curr
>> for this merge window ? I don't see any other arch changes to that effect in
>> latest linux-next.
>
> The required fix of the macros was merged yesterday into upstream.
>
Yes I do see the macro fix in mainline but no discussions related to missing
preemption debug checks after NAK from Peter/Ingo.
I'm gonna move this one out of my 3.12 merge window queue to my for-next - so not
dropping it (you can rest assured) but it doesn't look merge ready too. OK ?
-Vineet
On Thu, 5 Sep 2013, Vineet Gupta wrote:
> Yes I do see the macro fix in mainline but no discussions related to missing
> preemption debug checks after NAK from Peter/Ingo.
v4 was posted yesterday with the preemption bug checks.
On 09/05/2013 07:49 PM, Christoph Lameter wrote:
> On Thu, 5 Sep 2013, Vineet Gupta wrote:
>
>> Yes I do see the macro fix in mainline but no discussions related to missing
>> preemption debug checks after NAK from Peter/Ingo.
> v4 was posted yesterday with the preemption bug checks.
I see that. IMHO it would be better if all of this gets merged into mainline in
one cycle.
I'm a relatively new maintainer so I'm not sure how to handle this. Andrew, Tejun,
Steven what say you ?
-Vineet