I would prefer not to add error handling here because ComponentName and
ComponentName2 are installed at image handles and the uninstallation operation
would not fail.
Thanks
Feng
From: Simon (Xiang) Lian-SSI [mailto:[email protected]]
Sent: Tuesday, January 07, 2014 06:45
To: Tian, Feng; [email protected]; Kinney, Michael D
Subject: RE: Another bug in NVMe driver
The un-installation of ComponentName and ComponentName2 are assumed to be
always successful. But in case they fail, the unload should return failure as
well. Others look good to me.
Thanks,
Simon
From: Tian, Feng [mailto:[email protected]]
Sent: Sunday, January 05, 2014 5:30 PM
To: [email protected]<mailto:[email protected]>;
Simon (Xiang) Lian-SSI; Kinney, Michael D
Cc: Tian, Feng
Subject: RE: Another bug in NVMe driver
Thanks for reminder, Mike.
So looks like only the For loop in unload() is redundant, Please review the
proposed fix.
Best Regards
Feng
From: Kinney, Michael D
[mailto:[email protected]]<mailto:[mailto:[email protected]]>
Sent: Saturday, January 04, 2014 01:55
To: [email protected]<mailto:[email protected]>;
Simon (Xiang) Lian-SSI; Kinney, Michael D
Subject: Re: [edk2] Another bug in NVMe driver
Feng,
Many UEFI Drivers use the UefiLib to install the driver related protocols and
some of the driver related protocols are optionally installed based on PCD
settings. If the Uninstall attempts to uninstall a protocol that is not
present, then the uninstall will fail and unload will fail. This is likely why
the logic is a slightly more complex than expected in the Unload() function.
The PCDs used by the UefiLib that apply here are:
gEfiMdePkgTokenSpaceGuid.PcdDriverDiagnosticsDisable ## CONSUMES
gEfiMdePkgTokenSpaceGuid.PcdComponentNameDisable ## CONSUMES
gEfiMdePkgTokenSpaceGuid.PcdDriverDiagnostics2Disable ## CONSUMES
gEfiMdePkgTokenSpaceGuid.PcdComponentName2Disable ## CONSUMES
Best regards,
Mike
From: Tian, Feng [mailto:[email protected]]
Sent: Thursday, January 02, 2014 6:33 PM
To: Simon (Xiang) Lian-SSI
Cc: [email protected]<mailto:[email protected]>
Subject: Re: [edk2] Another bug in NVMe driver
Agree, the logic could be simplified like your proposal.
Thanks
Feng
From: Simon (Xiang) Lian-SSI [mailto:[email protected]]
Sent: Friday, January 03, 2014 09:41
To: Tian, Feng
Cc: [email protected]<mailto:[email protected]>
Subject: RE: Another bug in NVMe driver
Hi Feng,
One more thing is that in NvmExpressUnload, the following for loop doesn't seem
to be necessary since we are uninstalling all protocols previously attached to
this specific driver image handle, ie. DriverBinding->ImageHandle==ImageHandle,
all other handles do not even need to be checked.
for (Index = 0; Index < DeviceHandleCount; Index++) {
Status = gBS->HandleProtocol (
DeviceHandleBuffer[Index],
&gEfiDriverBindingProtocolGuid,
(VOID **) &DriverBinding
);
if (EFI_ERROR (Status)) {
continue;
}
if (DriverBinding->ImageHandle != ImageHandle) {
continue;
}
gBS->UninstallProtocolInterface (
ImageHandle,
&gEfiDriverBindingProtocolGuid,
DriverBinding
);
Status = gBS->HandleProtocol (
DeviceHandleBuffer[Index],
&gEfiComponentNameProtocolGuid,
(VOID **) &ComponentName
);
if (!EFI_ERROR (Status)) {
gBS->UninstallProtocolInterface (
ImageHandle,
&gEfiComponentNameProtocolGuid,
ComponentName
);
}
... // Query and uninstall all other protocols
Therefore, can the above code be simply replaced with following?
Status = gBS->UninstallMultipleProtocolInterfaces (
ImageHandle,
&gEfiDriverBindingProtocolGuid,
&gNvmExpressDriverBinding,
&gEfiComponentNameProtocolGuid,
&gNvmExpressComponentName,
&gEfiComponentName2ProtocolGuid,
&gNvmExpressComponentName2,
&gEfiDriverDiagnosticsProtocolGuid,
&gNvmExpressDriverDiagnostics,
&gEfiDriverDiagnostics2ProtocolGuid,
&gNvmExpressDriverDiagnostics2,
&gEfiDriverSupportedEfiVersionProtocolGuid,
&gNvmExpressDriverSupportedEfiVersion,
NULL
);
Thanks,
Simon
From: Tian, Feng [mailto:[email protected]]
Sent: Thursday, January 02, 2014 5:02 PM
To: Simon (Xiang) Lian-SSI
Cc: [email protected]<mailto:[email protected]>;
Tian, Feng
Subject: RE: Another bug in NVMe driver
You are right. I will fix this issue as soon as possible.
Thanks
Feng
From: Simon (Xiang) Lian-SSI
[mailto:[email protected]]<mailto:[mailto:[email protected]]>
Sent: Friday, January 03, 2014 06:41
To: Tian, Feng
Cc: [email protected]<mailto:[email protected]>
Subject: Another bug in NVMe driver
Hi Feng,
There seems to have a minor bug in current NVMe driver that
DriverSupportedEfiVersion protocol hasn't been uninstalled in NvmExpressUnload,
resulting in the driver image handle never gets freed up from the handle
database.
Thanks,
Simon
------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
edk2-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/edk2-devel
