qemu-s390x
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v4 6/6] s390x: kvm: Make kvm_sclp_service_call void


From: David Hildenbrand
Subject: Re: [PATCH v4 6/6] s390x: kvm: Make kvm_sclp_service_call void
Date: Wed, 27 Nov 2019 19:17:10 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.1.1

On 27.11.19 19:07, David Hildenbrand wrote:
> On 27.11.19 18:50, Janosch Frank wrote:
>> It defaults to returning 0 anyway and that return value is not
>> necessary, as 0 is also the default rc that the caller would return.
>>
>> While doing that we can simplify the logic a bit and return early if
>> we inject a PGM exception. Also we always set a 0 cc, so let's not
>> base it on the rc of the sclp emulation functions.
>>
>> Signed-off-by: Janosch Frank <address@hidden>
>> ---
>>  target/s390x/kvm.c | 12 +++++-------
>>  1 file changed, 5 insertions(+), 7 deletions(-)
>>
>> diff --git a/target/s390x/kvm.c b/target/s390x/kvm.c
>> index 0c9d14b4b1..08bb1edca0 100644
>> --- a/target/s390x/kvm.c
>> +++ b/target/s390x/kvm.c
>> @@ -1159,13 +1159,13 @@ void kvm_s390_access_exception(S390CPU *cpu, 
>> uint16_t code, uint64_t te_code)
>>      kvm_s390_vcpu_interrupt(cpu, &irq);
>>  }
>>  
>> -static int kvm_sclp_service_call(S390CPU *cpu, struct kvm_run *run,
>> +static void kvm_sclp_service_call(S390CPU *cpu, struct kvm_run *run,
>>                                   uint16_t ipbh0)
>>  {
>>      CPUS390XState *env = &cpu->env;
>>      uint64_t sccb;
>>      uint32_t code;
>> -    int r = 0;
>> +    int r;
>>  
>>      sccb = env->regs[ipbh0 & 0xf];
>>      code = env->regs[(ipbh0 & 0xf0) >> 4];
>> @@ -1173,11 +1173,9 @@ static int kvm_sclp_service_call(S390CPU *cpu, struct 
>> kvm_run *run,
>>      r = sclp_service_call(env, sccb, code);
>>      if (r < 0) {
>>          kvm_s390_program_interrupt(cpu, -r);
>> -    } else {
>> -        setcc(cpu, r);
>> +        return;
>>      }
>> -
>> -    return 0;
>> +    setcc(cpu, 0);
> 
> For now, sclp_service_call will return <= 0 ... but don't we actually
> have the option to return a cc? What does the spec say? Always set to 0?
> 
> At least also the TCG implementation sets the CC to whatever is returned
> here .... and Claudio's unit tests have code to handle cc != 0 ... and
> the kernel as well (drivers/s390/char/sclp.h:sclp_service_call())
> 

FWIW, I'd keep this as

setcc(cpu, r);

if we ever have to set a cc != 0.

-- 
Thanks,

David / dhildenb




reply via email to

[Prev in Thread] Current Thread [Next in Thread]