Re: [Qemu-devel] [PATCH qemu v6 10/15] spapr_pci_vfio: Remove unnecessary cast to sPAPRPHBVFIOState

[Alexey Kardashevskiy <aik@ozlabs.ru>]

On 04/16/2015 08:31 PM, Thomas Huth wrote:
> Am Sat, 11 Apr 2015 01:24:39 +1000
> schrieb Alexey Kardashevskiy <aik@ozlabs.ru>:
>
>> EEH handlers only need AddressSpace to get to the right VFIO container
>> to call ioctl() so remove unnecessary cast of sPAPRPHBState to
>> sPAPRPHBVFIOState.
>>
>> Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
>> ---
>>   hw/ppc/spapr_pci_vfio.c | 12 ++++--------
>>   1 file changed, 4 insertions(+), 8 deletions(-)
>>
>> diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c
>> index a428166..a5b97d0 100644
>> --- a/hw/ppc/spapr_pci_vfio.c
>> +++ b/hw/ppc/spapr_pci_vfio.c
>> @@ -66,7 +66,6 @@ static void spapr_phb_vfio_reset(DeviceState *qdev)
>>   static int spapr_phb_vfio_eeh_set_option(sPAPRPHBState *sphb,
>>                                            unsigned int addr, int option)
>>   {
>> -    sPAPRPHBVFIOState *svphb = SPAPR_PCI_VFIO_HOST_BRIDGE(sphb);
>>       struct vfio_eeh_pe_op op = { .argsz = sizeof(op) };
>>       int ret;
>>
>> @@ -103,7 +102,7 @@ static int spapr_phb_vfio_eeh_set_option(sPAPRPHBState *sphb,
>>           return RTAS_OUT_PARAM_ERROR;
>>       }
>>
>> -    ret = vfio_container_ioctl(&svphb->phb.iommu_as,
>> +    ret = vfio_container_ioctl(&sphb->iommu_as,
>>                                  VFIO_EEH_PE_OP, &op);
>>       if (ret < 0) {
>>           return RTAS_OUT_HW_ERROR;
>> @@ -114,12 +113,11 @@ static int spapr_phb_vfio_eeh_set_option(sPAPRPHBState *sphb,
>>
>>   static int spapr_phb_vfio_eeh_get_state(sPAPRPHBState *sphb, int *state)
>>   {
>> -    sPAPRPHBVFIOState *svphb = SPAPR_PCI_VFIO_HOST_BRIDGE(sphb);
>>       struct vfio_eeh_pe_op op = { .argsz = sizeof(op) };
>>       int ret;
>>
>>       op.op = VFIO_EEH_PE_GET_STATE;
>> -    ret = vfio_container_ioctl(&svphb->phb.iommu_as,
>> +    ret = vfio_container_ioctl(&sphb->iommu_as,
>>                                  VFIO_EEH_PE_OP, &op);
>>       if (ret < 0) {
>>           return RTAS_OUT_PARAM_ERROR;
>> @@ -131,7 +129,6 @@ static int spapr_phb_vfio_eeh_get_state(sPAPRPHBState *sphb, int *state)
>>
>>   static int spapr_phb_vfio_eeh_reset(sPAPRPHBState *sphb, int option)
>>   {
>> -    sPAPRPHBVFIOState *svphb = SPAPR_PCI_VFIO_HOST_BRIDGE(sphb);
>>       struct vfio_eeh_pe_op op = { .argsz = sizeof(op) };
>>       int ret;
>>
>> @@ -149,7 +146,7 @@ static int spapr_phb_vfio_eeh_reset(sPAPRPHBState *sphb, int option)
>>           return RTAS_OUT_PARAM_ERROR;
>>       }
>>
>> -    ret = vfio_container_ioctl(&svphb->phb.iommu_as,
>> +    ret = vfio_container_ioctl(&sphb->iommu_as,
>>                                  VFIO_EEH_PE_OP, &op);
>>       if (ret < 0) {
>>           return RTAS_OUT_HW_ERROR;
>> @@ -160,12 +157,11 @@ static int spapr_phb_vfio_eeh_reset(sPAPRPHBState *sphb, int option)
>>
>>   static int spapr_phb_vfio_eeh_configure(sPAPRPHBState *sphb)
>>   {
>> -    sPAPRPHBVFIOState *svphb = SPAPR_PCI_VFIO_HOST_BRIDGE(sphb);
>>       struct vfio_eeh_pe_op op = { .argsz = sizeof(op) };
>>       int ret;
>>
>>       op.op = VFIO_EEH_PE_CONFIGURE;
>> -    ret = vfio_container_ioctl(&svphb->phb.iommu_as,
>> +    ret = vfio_container_ioctl(&sphb->iommu_as,
>>                                  VFIO_EEH_PE_OP, &op);
>>       if (ret < 0) {
>>           return RTAS_OUT_PARAM_ERROR;
>
> Patch looks fine to me, so:
>
> Reviewed-by: Thomas Huth <thuth@redhat.com>
>
> But maybe you could even squash this into the patch where you removed the
> iommugroupid parameters (patch #5) ... so that you don't have to touch all
> these lines twice?

May be... Patch #5 enables multiple groups per container, and this one 
prepares for spapr-vfio-phb merge so they solve different issues in general 
and I was told multiple times the keeping patches simpler is a good thing 
(to reasonable extent) :)

Anyway in the next respin I'll remove more of EEH stuff, coming soon!


-- 
Alexey