ShellPkg: Difference in behavior of 'dh' between old shell and new

ShellPkg: Difference in behavior of 'dh' between old shell and new

[Jeff Westfahl]

I noticed a difference in behavior of 'dh' between the old shell and the 
new shell. With the old shell, 'dh -v' for a LoadedImage handle shows the 
following:

Handle D3 (3A537F98)
    Image (3A532818)   File:MicrocodeUpdate
      ParentHandle..: 3A64F118
      SystemTable...: 3D2A8F18
      DeviceHandle..: 3B1B2098
      FilePath......: FvFile(F3331DE6-4A55-44E4-B767-7453F7A1A021)
      ImageBase.....: 3D650000 - 3D655540
      ImageSize.....: 5540
      CodeType......: RT_code
      DataType......: RT_data

With the new shell, I get this for the same LoadedImage handle:

D3: 3A537F98
LoadedImage
      Revision......: 0x00001000
      ParentHandle..: 3A64F118
      SystemTable...: 3D2A8F18
      DeviceHandle..: 3B1B2098
      FilePath......: 3A539018
      OptionsSize...: 0
      LoadOptions...: 0
      ImageBase.....: 3D650000
      ImageSize.....: 5540
      CodeType......: EfiRuntimeServicesCode
      DataType......: EfiRuntimeServicesData
      Unload........: 0

The old shell shows the name of the file associated with the LoadedImage, 
which seems like useful information. Is this omission intentional or an 
oversight? Would a patch adding the file name to ShellPkg be welcomed?

Regards,
Jeff

Re: ShellPkg: Difference in behavior of 'dh' between old shell and new

[Carsey, Jaben]

That would be welcome; I don't know how it was missed.  A caveat is that the shell should use UEFI protocols only.  No PI protocols can be required (optional is ok).

Only the specific output controlled by "-sfo" is severely constrained by the spec.

> -----Original Message-----
> From: Jeff Westfahl [mailto:jeff.westfahl@ni.com]
> Sent: Wednesday, April 19, 2017 10:36 AM
> To: edk2-devel@lists.01.org
> Cc: Carsey, Jaben <jaben.carsey@intel.com>; Ni, Ruiyu <ruiyu.ni@intel.com>
> Subject: [edk2] ShellPkg: Difference in behavior of 'dh' between old shell and
> new
> Importance: High
> 
> I noticed a difference in behavior of 'dh' between the old shell and the new shell.
> With the old shell, 'dh -v' for a LoadedImage handle shows the
> following:
> 
> Handle D3 (3A537F98)
>     Image (3A532818)   File:MicrocodeUpdate
>       ParentHandle..: 3A64F118
>       SystemTable...: 3D2A8F18
>       DeviceHandle..: 3B1B2098
>       FilePath......: FvFile(F3331DE6-4A55-44E4-B767-7453F7A1A021)
>       ImageBase.....: 3D650000 - 3D655540
>       ImageSize.....: 5540
>       CodeType......: RT_code
>       DataType......: RT_data
> 
> With the new shell, I get this for the same LoadedImage handle:
> 
> D3: 3A537F98
> LoadedImage
>       Revision......: 0x00001000
>       ParentHandle..: 3A64F118
>       SystemTable...: 3D2A8F18
>       DeviceHandle..: 3B1B2098
>       FilePath......: 3A539018
>       OptionsSize...: 0
>       LoadOptions...: 0
>       ImageBase.....: 3D650000
>       ImageSize.....: 5540
>       CodeType......: EfiRuntimeServicesCode
>       DataType......: EfiRuntimeServicesData
>       Unload........: 0
> 
> The old shell shows the name of the file associated with the LoadedImage,
> which seems like useful information. Is this omission intentional or an oversight?
> Would a patch adding the file name to ShellPkg be welcomed?
> 
> Regards,
> Jeff

Re: ShellPkg: Difference in behavior of 'dh' between old shell and new

[Jeff Westfahl]

Hi Jaben,

No PI protocols can be required means that I can try to use them? And 
gracefully handle the case where they aren't found?

Regards,
Jeff

On Wed, 19 Apr 2017, Carsey, Jaben wrote:

> That would be welcome; I don't know how it was missed.  A caveat is that the shell should use UEFI protocols only.  No PI protocols can be required (optional is ok).
>
> Only the specific output controlled by "-sfo" is severely constrained by the spec.
>
>> -----Original Message-----
>> From: Jeff Westfahl [mailto:jeff.westfahl@ni.com]
>> Sent: Wednesday, April 19, 2017 10:36 AM
>> To: edk2-devel@lists.01.org
>> Cc: Carsey, Jaben <jaben.carsey@intel.com>; Ni, Ruiyu <ruiyu.ni@intel.com>
>> Subject: [edk2] ShellPkg: Difference in behavior of 'dh' between old shell and
>> new
>> Importance: High
>>
>> I noticed a difference in behavior of 'dh' between the old shell and the new shell.
>> With the old shell, 'dh -v' for a LoadedImage handle shows the
>> following:
>>
>> Handle D3 (3A537F98)
>>     Image (3A532818)   File:MicrocodeUpdate
>>       ParentHandle..: 3A64F118
>>       SystemTable...: 3D2A8F18
>>       DeviceHandle..: 3B1B2098
>>       FilePath......: FvFile(F3331DE6-4A55-44E4-B767-7453F7A1A021)
>>       ImageBase.....: 3D650000 - 3D655540
>>       ImageSize.....: 5540
>>       CodeType......: RT_code
>>       DataType......: RT_data
>>
>> With the new shell, I get this for the same LoadedImage handle:
>>
>> D3: 3A537F98
>> LoadedImage
>>       Revision......: 0x00001000
>>       ParentHandle..: 3A64F118
>>       SystemTable...: 3D2A8F18
>>       DeviceHandle..: 3B1B2098
>>       FilePath......: 3A539018
>>       OptionsSize...: 0
>>       LoadOptions...: 0
>>       ImageBase.....: 3D650000
>>       ImageSize.....: 5540
>>       CodeType......: EfiRuntimeServicesCode
>>       DataType......: EfiRuntimeServicesData
>>       Unload........: 0
>>
>> The old shell shows the name of the file associated with the LoadedImage,
>> which seems like useful information. Is this omission intentional or an oversight?
>> Would a patch adding the file name to ShellPkg be welcomed?
>>
>> Regards,
>> Jeff
>

Re: ShellPkg: Difference in behavior of 'dh' between old shell and new

[Carsey, Jaben]

Pretty much. 

The shell can absolutely perform better by using them, but it must not generate errors due to not finding them.

> -----Original Message-----
> From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of
> Jeff Westfahl
> Sent: Thursday, May 04, 2017 8:00 AM
> To: Carsey, Jaben <jaben.carsey@intel.com>
> Cc: Ni, Ruiyu <ruiyu.ni@intel.com>; edk2-devel@lists.01.org
> Subject: Re: [edk2] ShellPkg: Difference in behavior of 'dh' between old shell
> and new
> Importance: High
> 
> Hi Jaben,
> 
> No PI protocols can be required means that I can try to use them? And
> gracefully handle the case where they aren't found?
> 
> Regards,
> Jeff
> 
> On Wed, 19 Apr 2017, Carsey, Jaben wrote:
> 
> > That would be welcome; I don't know how it was missed.  A caveat is that
> the shell should use UEFI protocols only.  No PI protocols can be required
> (optional is ok).
> >
> > Only the specific output controlled by "-sfo" is severely constrained by the
> spec.
> >
> >> -----Original Message-----
> >> From: Jeff Westfahl [mailto:jeff.westfahl@ni.com]
> >> Sent: Wednesday, April 19, 2017 10:36 AM
> >> To: edk2-devel@lists.01.org
> >> Cc: Carsey, Jaben <jaben.carsey@intel.com>; Ni, Ruiyu
> <ruiyu.ni@intel.com>
> >> Subject: [edk2] ShellPkg: Difference in behavior of 'dh' between old shell
> and
> >> new
> >> Importance: High
> >>
> >> I noticed a difference in behavior of 'dh' between the old shell and the
> new shell.
> >> With the old shell, 'dh -v' for a LoadedImage handle shows the
> >> following:
> >>
> >> Handle D3 (3A537F98)
> >>     Image (3A532818)   File:MicrocodeUpdate
> >>       ParentHandle..: 3A64F118
> >>       SystemTable...: 3D2A8F18
> >>       DeviceHandle..: 3B1B2098
> >>       FilePath......: FvFile(F3331DE6-4A55-44E4-B767-7453F7A1A021)
> >>       ImageBase.....: 3D650000 - 3D655540
> >>       ImageSize.....: 5540
> >>       CodeType......: RT_code
> >>       DataType......: RT_data
> >>
> >> With the new shell, I get this for the same LoadedImage handle:
> >>
> >> D3: 3A537F98
> >> LoadedImage
> >>       Revision......: 0x00001000
> >>       ParentHandle..: 3A64F118
> >>       SystemTable...: 3D2A8F18
> >>       DeviceHandle..: 3B1B2098
> >>       FilePath......: 3A539018
> >>       OptionsSize...: 0
> >>       LoadOptions...: 0
> >>       ImageBase.....: 3D650000
> >>       ImageSize.....: 5540
> >>       CodeType......: EfiRuntimeServicesCode
> >>       DataType......: EfiRuntimeServicesData
> >>       Unload........: 0
> >>
> >> The old shell shows the name of the file associated with the
> LoadedImage,
> >> which seems like useful information. Is this omission intentional or an
> oversight?
> >> Would a patch adding the file name to ShellPkg be welcomed?
> >>
> >> Regards,
> >> Jeff
> >
> _______________________________________________
> edk2-devel mailing list
> edk2-devel@lists.01.org
> https://lists.01.org/mailman/listinfo/edk2-devel