Re: [gcv v3 06/35] scheduler: Replace __get_cpu_var uses

[Ingo Molnar]

* Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:

> On Wed, Aug 28, 2013 at 07:48:14PM +0000, Christoph Lameter wrote:
> > 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);
> > 
> 
> This looses a preemption debug check, so NAK
> 
> > 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);
> > 
> 
> This too looses a preemption debug check, NAK
> 
> > 6. Increment/Decrement etc of a per cpu variable
> > 
> > 	DEFINE_PER_CPU(int, y);
> > 	__get_cpu_var(y)++
> > 
> >    Converts to
> > 
> > 	this_cpu_inc(y)
> > 
> 
> Lo and behold.. no preemption checks again.
> 
> 
> Seriously first fix the debug and validation bits of the *this_cpu* 
> stuff.

Note that most of the other 'gcv' patches have these problems as well, so 
it's a NAK from me as well for most of the other patches as well ...

Thanks,

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/