Re: Regression on ia64 with cpu masks: optimize and clean up cpumask_of_cpu()

[Mike Travis <travis@sgi.com>]

Check linux-next, I believe the exact same fix is in there.  Apparently,
as Stephen and others have pointed out earlier, that syntax isn't evaluated 
correctly in earlier gcc's (at least earlier then the 4. series).

Thanks,
Mike

Simon Horman wrote:
> On Tue, Jul 29, 2008 at 01:45:01PM +1000, Simon Horman wrote:
>> Hi,
>>
>> I haven't had a time to look into this closeley yet,
>> but "cpu masks: optimize and clean up cpumask_of_cpu()"
>> aka e56b3bc7942982ac2589c942fb345e38bc7a341a causes
>> a build failure for me on ia64.
>>
>> # ia64-unknown-linux-gnu-gcc --version
>> ia64-unknown-linux-gnu-gcc (GCC) 3.4.5
>> Copyright (C) 2004 Free Software Foundation, Inc.
>> This is free software; see the source for copying conditions.  There is NO
>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>>
>> # make
>> make[1]: `include/asm-ia64/nr-irqs.h' is up to date.
>>   CHK     include/linux/version.h
>>   CHK     include/linux/utsrelease.h
>>   CALL    scripts/checksyscalls.sh
>>   CHK     include/linux/compile.h
>>   CC      drivers/acpi/processor_throttling.o
>> drivers/acpi/processor_throttling.c: In function `acpi_processor_get_throttling':
>> drivers/acpi/processor_throttling.c:841: error: invalid lvalue in unary `&'
>> drivers/acpi/processor_throttling.c: In function `acpi_processor_set_throttling':
>> drivers/acpi/processor_throttling.c:1028: error: invalid lvalue in unary `&'
>> drivers/acpi/processor_throttling.c:1059: error: invalid lvalue in unary `&'
>> make[2]: *** [drivers/acpi/processor_throttling.o] Error 1
>> make[1]: *** [drivers/acpi] Error 2
>> make: *** [drivers] Error 2
> 
> I'm pretty sure that I am missing something subtle here,
> but something that is working for me is:
> 
> Index: linux-2.6/include/linux/cpumask.h
> ===================================================================
> --- linux-2.6.orig/include/linux/cpumask.h	2008-07-29 19:03:06.000000000 +1000
> +++ linux-2.6/include/linux/cpumask.h	2008-07-29 19:03:16.000000000 +1000
> @@ -287,7 +287,7 @@ static inline const cpumask_t *get_cpu_m
>   * gcc optimizes it out (it's a constant) and there's no huge stack
>   * variable created:
>   */
> -#define cpumask_of_cpu(cpu) ({ *get_cpu_mask(cpu); })
> +#define cpumask_of_cpu(cpu) (*get_cpu_mask(cpu))
>  
>  
>  #define CPU_MASK_LAST_WORD BITMAP_LAST_WORD_MASK(NR_CPUS)