Re: [edk2] 答复: [ScsiDiskDxe] Implict type conversion bug

2017-04-11 Thread Wu, Hao A 
Hi Gary,

Thanks for the reporting, I will take a look at all the cases you mentioned.

Best Regards,
Hao Wu


> -Original Message-
> From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On Behalf Of
> Guoheyi
> Sent: Tuesday, April 11, 2017 3:05 PM
> To: Zeng, Star; edk2-devel@lists.01.org
> Cc: Tian, Feng
> Subject: [edk2] 答复: [ScsiDiskDxe] Implict type conversion bug
> 
> Hi Star,
> 
> I searched " LastLba3" in EDK2 code and found there were some other code
> similar to this one, could you please check whether they need to be fixed as
> well?
> 
> IntelFrameworkModulePkg/Bus/Pci/IdeBusDxe/Atapi.c:IdeDev-
> >BlkIo.Media->LastBlock = (Data.LastLba3 << 24) |
> IntelFrameworkModulePkg/Bus/Pci/IdeBusDxe/Atapi.c:  IdeDev-
> >BlkIo.Media->LastBlock =  (FormatData.LastLba3 << 24) |
> MdeModulePkg/Bus/Ufs/UfsBlockIoPei/UfsBlockIoPei.c:Private-
> >Media[DeviceIndex].LastBlock  = (Capacity16.LastLba3 << 24) |
> (Capacity16.LastLba2 << 16) | (Capacity16.LastLba1 << 8) |
> Capacity16.LastLba0;
> MdeModulePkg/Bus/Ufs/UfsBlockIoPei/UfsBlockIoPei.c:Private-
> >Media[DeviceIndex].LastBlock  = (Capacity.LastLba3 << 24) |
> (Capacity.LastLba2 << 16) | (Capacity.LastLba1 << 8) | Capacity.LastLba0;
> MdeModulePkg/Bus/Pci/IdeBusPei/AtapiPeim.c:  MediaInfo->LastBlock=
> (Data.LastLba3 << 24) | (Data.LastLba2 << 16) | (Data.LastLba1 << 8) |
> Data.LastLba0;
> MdeModulePkg/Bus/Pci/IdeBusPei/AtapiPeim.c:MediaInfo->LastBlock =
> (FormatData.LastLba3 << 24) |
> 
> (I think the two lines below should be OK for LastBlock is defined as UINT32)
> MdeModulePkg/Bus/Usb/UsbBotPei/PeiAtapi.c:  LastBlock = (Data.LastLba3 <<
> 24) | (Data.LastLba2 << 16) | (Data.LastLba1 << 8) | Data.LastLba0;
> MdeModulePkg/Bus/Usb/UsbBotPei/PeiAtapi.c:LastBlock =
> (FormatData.LastLba3 << 24) | (FormatData.LastLba2 << 16) |
> (FormatData.LastLba1 << 8) | FormatData.LastLba0;
> 
> 
> Thanks and regards,
> 
> Gary (Heyi Guo)
> 
> 
> -邮件原件-
> 发件人: Guoheyi
> 发送时间: 2017年4月11日 12:25
> 收件人: 'Zeng, Star'; edk2-devel@lists.01.org
> 抄送: Tian, Feng
> 主题: 答复: [edk2] [ScsiDiskDxe] Implict type conversion bug
> 
> Thanks for your quick fix :)
> 
> -邮件原件-
> 发件人: Zeng, Star [mailto:star.z...@intel.com]
> 发送时间: 2017年4月10日 17:48
> 收件人: Guoheyi; edk2-devel@lists.01.org
> 抄送: Tian, Feng; Zeng, Star
> 主题: RE: [edk2] [ScsiDiskDxe] Implict type conversion bug
> 
> Hao has sent patch for this issue.
> You can search emails with title "[edk2] [PATCH] MdeModulePkg/ScsiDiskDxe:
> Fix potential implicit sign extension".
> 
> Thanks,
> Star
> -Original Message-
> From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On Behalf Of
> Guoheyi
> Sent: Monday, April 10, 2017 5:40 PM
> To: edk2-devel@lists.01.org
> Cc: Tian, Feng <feng.t...@intel.com>
> Subject: [edk2] [ScsiDiskDxe] Implict type conversion bug
> 
> Hi folks,
> 
> In MdeModulePkg/Bus/Scsi/ScsiDiskDxe/ScsiDisk.c, Line 2757, we have the
> code as below:
> 
> ScsiDiskDevice->BlkIo.Media->LastBlock =  (Capacity10->LastLba3 << 24) |
>   (Capacity10->LastLba2 << 16) |
>   (Capacity10->LastLba1 << 8)  |
>Capacity10->LastLba0;
> 
> However, we tested on VS2008 and found that the intermediate result would be
> convert to INT32 then INT64 implicitly, so the result will be sign extended 
> and
> LastBlock may not be assigned with what is expected. We can simply test
> LastLba3 >= 0x80 to see what will happen.
> 
> Please let me know your comments.
> 
> Thanks and regards,
> 
> Gary (Heyi Guo)
> ___
> edk2-devel mailing list
> edk2-devel@lists.01.org
> https://lists.01.org/mailman/listinfo/edk2-devel
> ___
> edk2-devel mailing list
> edk2-devel@lists.01.org
> https://lists.01.org/mailman/listinfo/edk2-devel
___
edk2-devel mailing list
edk2-devel@lists.01.org
https://lists.01.org/mailman/listinfo/edk2-devel

[edk2] 答复: [ScsiDiskDxe] Implict type conversion bug

2017-04-11 Thread Guoheyi 
Hi Star,

I searched " LastLba3" in EDK2 code and found there were some other code 
similar to this one, could you please check whether they need to be fixed as 
well?

IntelFrameworkModulePkg/Bus/Pci/IdeBusDxe/Atapi.c:
IdeDev->BlkIo.Media->LastBlock = (Data.LastLba3 << 24) |
IntelFrameworkModulePkg/Bus/Pci/IdeBusDxe/Atapi.c:  
IdeDev->BlkIo.Media->LastBlock =  (FormatData.LastLba3 << 24) |
MdeModulePkg/Bus/Ufs/UfsBlockIoPei/UfsBlockIoPei.c:
Private->Media[DeviceIndex].LastBlock  = (Capacity16.LastLba3 << 24) | 
(Capacity16.LastLba2 << 16) | (Capacity16.LastLba1 << 8) | Capacity16.LastLba0;
MdeModulePkg/Bus/Ufs/UfsBlockIoPei/UfsBlockIoPei.c:
Private->Media[DeviceIndex].LastBlock  = (Capacity.LastLba3 << 24) | 
(Capacity.LastLba2 << 16) | (Capacity.LastLba1 << 8) | Capacity.LastLba0;
MdeModulePkg/Bus/Pci/IdeBusPei/AtapiPeim.c:  MediaInfo->LastBlock= 
(Data.LastLba3 << 24) | (Data.LastLba2 << 16) | (Data.LastLba1 << 8) | 
Data.LastLba0;
MdeModulePkg/Bus/Pci/IdeBusPei/AtapiPeim.c:MediaInfo->LastBlock = 
(FormatData.LastLba3 << 24) |

(I think the two lines below should be OK for LastBlock is defined as UINT32)
MdeModulePkg/Bus/Usb/UsbBotPei/PeiAtapi.c:  LastBlock = (Data.LastLba3 << 24) | 
(Data.LastLba2 << 16) | (Data.LastLba1 << 8) | Data.LastLba0;
MdeModulePkg/Bus/Usb/UsbBotPei/PeiAtapi.c:LastBlock = (FormatData.LastLba3 
<< 24) | (FormatData.LastLba2 << 16) | (FormatData.LastLba1 << 8) | 
FormatData.LastLba0;


Thanks and regards,

Gary (Heyi Guo)


-邮件原件-
发件人: Guoheyi 
发送时间: 2017年4月11日 12:25
收件人: 'Zeng, Star'; edk2-devel@lists.01.org
抄送: Tian, Feng
主题: 答复: [edk2] [ScsiDiskDxe] Implict type conversion bug

Thanks for your quick fix :)

-邮件原件-
发件人: Zeng, Star [mailto:star.z...@intel.com] 
发送时间: 2017年4月10日 17:48
收件人: Guoheyi; edk2-devel@lists.01.org
抄送: Tian, Feng; Zeng, Star
主题: RE: [edk2] [ScsiDiskDxe] Implict type conversion bug

Hao has sent patch for this issue.
You can search emails with title "[edk2] [PATCH] MdeModulePkg/ScsiDiskDxe: Fix 
potential implicit sign extension".

Thanks,
Star
-Original Message-
From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On Behalf Of Guoheyi
Sent: Monday, April 10, 2017 5:40 PM
To: edk2-devel@lists.01.org
Cc: Tian, Feng 
Subject: [edk2] [ScsiDiskDxe] Implict type conversion bug

Hi folks,

In MdeModulePkg/Bus/Scsi/ScsiDiskDxe/ScsiDisk.c, Line 2757, we have the code as 
below:

ScsiDiskDevice->BlkIo.Media->LastBlock =  (Capacity10->LastLba3 << 24) |
  (Capacity10->LastLba2 << 16) |
  (Capacity10->LastLba1 << 8)  |
   Capacity10->LastLba0;

However, we tested on VS2008 and found that the intermediate result would be 
convert to INT32 then INT64 implicitly, so the result will be sign extended and 
LastBlock may not be assigned with what is expected. We can simply test 
LastLba3 >= 0x80 to see what will happen.

Please let me know your comments.

Thanks and regards,

Gary (Heyi Guo)
___
edk2-devel mailing list
edk2-devel@lists.01.org
https://lists.01.org/mailman/listinfo/edk2-devel
___
edk2-devel mailing list
edk2-devel@lists.01.org
https://lists.01.org/mailman/listinfo/edk2-devel

[edk2] 答复: [ScsiDiskDxe] Implict type conversion bug

2017-04-10 Thread Guoheyi 
Thanks for your quick fix :)

-邮件原件-
发件人: Zeng, Star [mailto:star.z...@intel.com] 
发送时间: 2017年4月10日 17:48
收件人: Guoheyi; edk2-devel@lists.01.org
抄送: Tian, Feng; Zeng, Star
主题: RE: [edk2] [ScsiDiskDxe] Implict type conversion bug

Hao has sent patch for this issue.
You can search emails with title "[edk2] [PATCH] MdeModulePkg/ScsiDiskDxe: Fix 
potential implicit sign extension".

Thanks,
Star
-Original Message-
From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On Behalf Of Guoheyi
Sent: Monday, April 10, 2017 5:40 PM
To: edk2-devel@lists.01.org
Cc: Tian, Feng 
Subject: [edk2] [ScsiDiskDxe] Implict type conversion bug

Hi folks,

In MdeModulePkg/Bus/Scsi/ScsiDiskDxe/ScsiDisk.c, Line 2757, we have the code as 
below:

ScsiDiskDevice->BlkIo.Media->LastBlock =  (Capacity10->LastLba3 << 24) |
  (Capacity10->LastLba2 << 16) |
  (Capacity10->LastLba1 << 8)  |
   Capacity10->LastLba0;

However, we tested on VS2008 and found that the intermediate result would be 
convert to INT32 then INT64 implicitly, so the result will be sign extended and 
LastBlock may not be assigned with what is expected. We can simply test 
LastLba3 >= 0x80 to see what will happen.

Please let me know your comments.

Thanks and regards,

Gary (Heyi Guo)
___
edk2-devel mailing list
edk2-devel@lists.01.org
https://lists.01.org/mailman/listinfo/edk2-devel
___
edk2-devel mailing list
edk2-devel@lists.01.org
https://lists.01.org/mailman/listinfo/edk2-devel