On Thu, Jan 28, 2010 at 05:46, <[email protected]> wrote:
> Revision 8250 Author gyang Date 2010-01-28 05:46:55 -0500 (Thu, 28 Jan 2010)
>
> Log Message
>
> [!no_src_qa!]Task [#5433], implement cpu_freq with cpuvoltage, and merge
> with hotplug_cpu
>
> Modified Paths
>
> trunk/arch/blackfin/include/asm/cdef_misc.h
> trunk/arch/blackfin/include/asm/dpmc.h
> trunk/arch/blackfin/include/asm/smp.h
> trunk/arch/blackfin/mach-bf561/hotplug.c
> trunk/arch/blackfin/mach-bf561/secondary.S
> trunk/arch/blackfin/mach-common/cpufreq.c
> trunk/arch/blackfin/mach-common/dpmc.c
>
> Diff
>
> Modified: trunk/arch/blackfin/include/asm/cdef_misc.h (8249 => 8250)
>
> --- trunk/arch/blackfin/include/asm/cdef_misc.h 2010-01-28 09:58:21 UTC
> (rev
> 8249)
> +++ trunk/arch/blackfin/include/asm/cdef_misc.h 2010-01-28 10:46:55 UTC
> (rev
> 8250)
> @@ -33,58 +33,78 @@
>
> #ifndef __ASSEMBLY__
> #include <asm/irq.h>
> -/* Writing to PLL_CTL initiates a PLL relock sequence. */
> -static __inline__ void bfin_write_PLL_CTL(unsigned int val)
> +
> +#define SUPPLE_0_WAKEUP ((IRQ_SUPPLE_0 - (IRQ_CORETMR + 1)) % 32)
> +
> +static __inline__ void bfin_iwr_set_pll(unsigned long *iwr0,
> + unsigned long *iwr1, unsigned long *iwr2,
> + unsigned long off)
> {
> - unsigned long flags = 0;
> #ifdef SIC_IWR0
> - unsigned long iwr0;
> + *iwr0 = bfin_read32(SIC_IWR0 + off);
> ...
>
> +/* Writing to PLL_CTL initiates a PLL relock sequence. */
> +static __inline__ void bfin_write_PLL_CTL(unsigned int val)
> +{
> + unsigned long flags = 0;
> + unsigned long iwr0, iwr1, iwr2;
> + unsigned long off = (bfin_read_DSPID() & 0xff) ? 0x1000 : 0;
this is cool for SMP ports, but it's no good for UP. the DSPID is
going to have a static value on all UP parts but this is going to be
calculated at runtime instead of compile time.
> static int
> vreg_cpufreq_notifier(struct notifier_block *nb, unsigned long val, void
> *data)
> {
> struct cpufreq_freqs *freq = data;
> + unsigned int cpu;
> + unsigned int this_cpu = smp_processor_id();
> + cpumask_t mask = cpu_online_map;
> + cpu_clear(this_cpu, mask);
>
> + if (((struct cpufreq_freqs *)data)->cpu != CPUFREQ_CPU)
this cast makes no sense as you already have a local variable "freq"
that is of the right type.
-mike
_______________________________________________
Linux-kernel-commits mailing list
[email protected]
https://blackfin.uclinux.org/mailman/listinfo/linux-kernel-commits
