Re: [PATCH] MdeModulePkg/NvmExpressDxe: fix error status override

[Heyi Guo <heyi.guo@linaro.org>]

在 12/4/2017 9:32 AM, Wu, Hao A 写道:
> Comments below.
>
> Best Regards,
> Hao Wu
>
>
>> -----Original Message-----
>> From: Zeng, Star
>> Sent: Monday, December 04, 2017 9:17 AM
>> To: Heyi Guo; linaro-uefi@lists.linaro.org; edk2-devel@lists.01.org
>> Cc: Heyi Guo; Dong, Eric; Wu, Hao A; Ni, Ruiyu; Zeng, Star
>> Subject: RE: [edk2] [PATCH] MdeModulePkg/NvmExpressDxe: fix error status
>> override
>>
>> Cc Hao and Ruiyu.
>>
>> Thanks,
>> Star
>> -----Original Message-----
>> From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of Heyi
>> Guo
>> Sent: Saturday, December 2, 2017 5:22 PM
>> To: linaro-uefi@lists.linaro.org; edk2-devel@lists.01.org
>> Cc: Heyi Guo <guoheyi@huawei.com>; Heyi Guo <heyi.guo@linaro.org>; Dong,
>> Eric <eric.dong@intel.com>; Zeng, Star <star.zeng@intel.com>
>> Subject: [edk2] [PATCH] MdeModulePkg/NvmExpressDxe: fix error status
>> override
>>
>> Commit f6b139b added return status handling to PciIo->Mem.Write.
>> However, the second status handling will override EFI_DEVICE_ERROR returned
>> in this branch:
>>
>>    //
>>    // Check the NVMe cmd execution result
>>    //
>>    if (Status != EFI_TIMEOUT) {
>>      if ((Cq->Sct == 0) && (Cq->Sc == 0)) {
>>        Status = EFI_SUCCESS;
>>      } else {
>>        Status = EFI_DEVICE_ERROR;
>>                 ^^^^^^^^^^^^^^^^
>>
>> Since PciIo->Mem.Write will probably return SUCCESS, it causes
>> NvmExpressPassThru to return SUCCESS even when DEVICE_ERROR occurs.
>> Callers of NvmExpressPassThru will then continue executing which may cause
>> further unexpected results, e.g. DiscoverAllNamespaces couldn't break out the
>> loop.
>>
>> So we add a | (bit-or) to combine the return status together.
>>
>> Contributed-under: TianoCore Contribution Agreement 1.1
>> Signed-off-by: Heyi Guo <guoheyi@huawei.com>
>> Cc: Star Zeng <star.zeng@intel.com>
>> Cc: Eric Dong <eric.dong@intel.com>
>> ---
>>   MdeModulePkg/Bus/Pci/NvmExpressDxe/NvmExpressPassthru.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/MdeModulePkg/Bus/Pci/NvmExpressDxe/NvmExpressPassthru.c
>> b/MdeModulePkg/Bus/Pci/NvmExpressDxe/NvmExpressPassthru.c
>> index c33038f..2698b27 100644
>> --- a/MdeModulePkg/Bus/Pci/NvmExpressDxe/NvmExpressPassthru.c
>> +++ b/MdeModulePkg/Bus/Pci/NvmExpressDxe/NvmExpressPassthru.c
>> @@ -831,7 +831,7 @@ NvmExpressPassThru (
>>     }
>>
>>     Data = ReadUnaligned32 ((UINT32*)&Private->CqHdbl[QueueId]);
>> -  Status = PciIo->Mem.Write (
>> +  Status |= PciIo->Mem.Write (
>>                  PciIo,
>>                  EfiPciIoWidthUint32,
>>                  NVME_BAR,
> I searched the whole edk2 code base and did not find a similar case like:
> Status |= Foo();
>
> I also think using the above style might cause the function returning
> confusing status to the caller.
That's true.
>
> So how about introducing a new variable?
>    EFI_STATUS   PreviousStatus;
>    ...
>    Data = ReadUnaligned32 ((UINT32*)&Private->CqHdbl[QueueId]);
>    PreviousStatus = Status;
>    Status = PciIo->Mem.Write (...);
>    Status = EFI_ERROR (PreviousStatus) ? PreviousStatus : Status;
Looks good to me. I will send out v2 patch. Thanks for your comments.

Regards,

Gary (Heyi Guo)

>
>
>> --
>> 2.7.2.windows.1
>>
>> _______________________________________________
>> edk2-devel mailing list
>> edk2-devel@lists.01.org
>> https://lists.01.org/mailman/listinfo/edk2-devel