On Thu, Jan 10, 2008 at 10:58:57AM +0100, Andi Kleen wrote:
quoted text
> > It seems the correct solution would be not to hijack __cpuinit
> > (as your patch does), but to create a new annotation.
> 
> The rationale is that after suspend the CPU has to be reinitialized.
> That is because it is essentially like a reboot. All the previous
> CPU state is gone.
>...

But your patch does:

+config PM_CPUINIT
+       bool
+       depends on PM
+       default y

As an example, even plain ACPI support without any suspend support in 
the kernel at all requires CONFIG_PM and therefore forces all __cpuinit 
code to be non-__init after your patch.

And if the dependency was corrected to PM_SLEEP it will still make 
the UP kernel use more memory since we currently have __cpuinit code 
that gets discarded after boot but suspend/resume is apparently working.

Plus my other point that it seems to be wrong to do whatever change only 
for x86.

quoted text
> -Andi

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

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