Re: [Qemu-devel] [PATCH 10/12] linux-user: clock_nanosleep errno Handling on PPC

[Tom Musta]

On 8/4/2014 12:43 PM, Peter Maydell wrote:
> On 4 August 2014 17:45, Tom Musta <address@hidden> wrote:
>> The clock_nanosleep syscall is unusual in that it returns positive
>> numbers in error handling situations, versus returning -1 and setting
>> errno, or returning a negative errno value.  On POWER, the kernel will
>> set the SO bit of CR0 to indicate failure in a syscall.  QEMU has
>> generic handling to do this for syscalls with standard return values.
>>
>> Add special case code for clock_nanosleep to handle CR0 properly.
>>
>> Signed-off-by: Tom Musta <address@hidden>
>>
>> diff --git a/linux-user/syscall.c b/linux-user/syscall.c
>> index 95cee0b..5660520 100644
>> --- a/linux-user/syscall.c
>> +++ b/linux-user/syscall.c
>> @@ -8993,6 +8993,14 @@ abi_long do_syscall(void *cpu_env, int num, abi_long 
>> arg1,
>>          ret = get_errno(clock_nanosleep(arg1, arg2, &ts, arg4 ? &ts : 
>> NULL));
>>          if (arg4)
>>              host_to_target_timespec(arg4, &ts);
>> +
>> +#if defined(TARGET_PPC) || defined(TARGET_PPC64)
> 
> ...isn't TARGET_PPC always defined if TARGET_PPC64 is?
> (ie second condition in the || is unnecessary)
> 
> -- PMM
> 

You are correct.  I was thinking "TARGET_PPC" was a euphemism for 32 bit Power
(since --target=ppc-softmmy,ppc-linux-user are used to configure 32 bit
implementations).  But I see now in config-target.h that this is not the case.