public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: Amit kumar <akamit91@hotmail.com>
To: Andrew Fish <afish@apple.com>
Cc: TVKR <tresko1@gmail.com>,
	"edk2-devel@lists.01.org" <edk2-devel@lists.01.org>
Subject: Re: DisconnectController API not working.
Date: Wed, 28 Jun 2017 19:37:04 +0000	[thread overview]
Message-ID: <DM5PR11MB15780CD866011EF25FCC04A2DCDD0@DM5PR11MB1578.namprd11.prod.outlook.com> (raw)
In-Reply-To: <F9C3C076-F75D-43C0-B71C-86A0947EDD87@apple.com>

Just a final check,
You used this

    gBS->DisconnectController (
     ConvertHandleIndexToHandle (0x175),
     ConvertHandleIndexToHandle (0x176),
     NULL);

not this

    gBS->DisconnectController (
     ConvertHandleIndexToHandle (175),
     ConvertHandleIndexToHandle (176),
     NULL);

right ?

Amit

On Jun 29, 2017, at 1:04 AM, Andrew Fish <afish@apple.com<mailto:afish@apple.com>> wrote:


On Jun 28, 2017, at 12:27 PM, TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com>> wrote:

Tried it, still no luck :(

There is something in the Shell that makes the in-built disconnect without
any issues but when the same DisconnectController is called from an
application is fails.


It is more likely you are doing something wrong.

Using HandleIndex is risky as it is more about User Interface than reality. Have you tried printing out the EFI_HANDLE values you are passing into gBS->DisconnectController()? The should match the values you get when you dump a specific handle in the shell. The EFI_HANDLE value is not going to change, it is only going to get removed if all protocols on that handle are uninstalled.

So for example.
176: 7219E798

If you used 178 at the Shell you would expect to see 7219E798 printed out as the handle value.

Thanks,

Andrew Fish

Thanks

On Wed, Jun 28, 2017 at 1:23 PM, Amit kumar <akamit91@hotmail.com<mailto:akamit91@hotmail.com>> wrote:



EFI_STATUS
EFIAPI
UefiMain (
IN EFI_HANDLE        ImageHandle,
IN EFI_SYSTEM_TABLE  *SystemTable
)
{
  EFI_STATUS Status;


    gBS->DisconnectController (
     ConvertHandleIndexToHandle (0x175),
     ConvertHandleIndexToHandle (0x176),
     NULL);
 return (Status);

}


------------------------------
*From:* edk2-devel <edk2-devel-bounces@lists.01.org<mailto:edk2-devel-bounces@lists.01.org>> on behalf of Amit
kumar <akamit91@hotmail.com<mailto:akamit91@hotmail.com>>
*Sent:* Wednesday, June 28, 2017 11:41:53 PM
*To:* TVKR
*Cc:* edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
*Subject:* Re: [edk2] DisconnectController API not working.

Try this; just a quick thing




EFI_STATUS
EFIAPI
UefiMain (
IN EFI_HANDLE        ImageHandle,
IN EFI_SYSTEM_TABLE  *SystemTable
)
{
  EFI_STATUS Status;


    gBS->DisconnectController (
     ConvertHandleIndexToHandle (175),
     ConvertHandleIndexToHandle (176),
     NULL);
 return (Status);

}


________________________________
From: TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com>>
Sent: Wednesday, June 28, 2017 11:31:46 PM
To: Amit kumar
Cc: edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
Subject: Re: [edk2] DisconnectController API not working.

Yes, the shell's in-built command disconnect (disconnect 175 176) worked
but DisconnectController call from the application still did not work.

On Wed, Jun 28, 2017 at 12:56 PM, Amit kumar <akamit91@hotmail.com<mailto:akamit91@hotmail.com><mailto:
akamit91@hotmail.com<mailto:akamit91@hotmail.com>>> wrote:

okay, os tell me , did you try  "disconnect 175 176" in UEFI shell ?

________________________________
From: TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com><mailto:tresko1@gmail.com>>
Sent: Wednesday, June 28, 2017 11:21:34 PM

To: Amit kumar
Cc: edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org><mailto:edk2-devel@lists.01.org>
Subject: Re: [edk2] DisconnectController API not working.

Sorry, I wasn't clear in my last email. No it did not hang. The 'reconnect
-r' command completed successfully (no errors reported on screen).

On Wed, Jun 28, 2017 at 12:50 PM, Amit kumar <akamit91@hotmail.com<mailto:akamit91@hotmail.com><mailto:
akamit91@hotmail.com<mailto:akamit91@hotmail.com>>> wrote:

Did it hang ?

________________________________
From: TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com><mailto:tresko1@gmail.com>>
Sent: Wednesday, June 28, 2017 10:45:13 PM
To: Amit kumar
Cc: edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org><mailto:edk2-devel@lists.01.org>

Subject: Re: [edk2] DisconnectController API not working.

Thanks for the suggestion Amit. I tried the reconnect -r but it doesn't
seem to help.

On Wed, Jun 28, 2017 at 10:29 AM, Amit kumar <akamit91@hotmail.com<mailto:akamit91@hotmail.com><mailto:
akamit91@hotmail.com<mailto:akamit91@hotmail.com>>> wrote:
Hi Tresko,
I looks like there is no problem with your implementation, as was the case
with me.
I can suggest you a quick trick here, if it won’t work I will help you
further which might take a while.
Follow the steps below.
1. Go to refi shell
2. Reconnect -r
3. If the sysytem hangs.
4. Try to update your BIOS/UEFI to latest (as I had noticed the eariler
BIOS/UEFI had some some problem which normally gets fixed with updates).
5. After update once again try  Reconnect -r.
6. If theres a no hang restart the system and now you can try your app or
just give disconnect <ControllerHandeIndex> in uefi shell.
7. If still I doesn’t work you can ping me I might suggest the other way
around.

Amit

On Jun 28, 2017, at 8:05 PM, TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com><mailto:tres
ko1@gmail.com<mailto:ko1@gmail.com>>> wrote:

Hi Amit,

Here is the output. It is the call 'gBS->DisconnectController
(ControllerHandle, NULL, NULL)' (where ControllerHandle=7219EA98) that
doesn't seem to correctly work.

FS0:\> drivers
Output Truncated..
176 07042200 B Y Y   1   1 Intel(R) PRO/1000 7.3.18 PCI-E
PciRoot(0x0)/Pci(0x2,0x3)/Pci(0x0,0x0)/Offset(0x1D200,0x3E9FF)
178 07042200 B Y Y   1   1 Intel(R) PRO/1000 7.3.18 PCI-E
PciRoot(0x0)/Pci(0x2,0x3)/Pci(0x0,0x1)/Offset(0x1D200,0x3E9FF)

FS0:\> dh 176 -d -v
176: 7219E798
DriverVersion
0x00020028
EfiDriverHealthProtocolGuid
DriverConfiguration
DriverDiagnostics2
ComponentName2
DriverDiagnostics
ComponentName
DriverBinding
ImageDevicePath
LoadedImage
   Revision......: 0x00001000
   ParentHandle..: 72B70A18
   SystemTable...: 788FEF18
   DeviceHandle..: 7219EA98
   FilePath......: 7219E998
   OptionsSize...: 0
   LoadOptions...: <null string>
   ImageBase.....: 776A0000
   ImageSize.....: 555E0
   CodeType......: EfiBootServicesCode
   DataType......: EfiBootServicesData
   Unload........: 776A0708

   Child[176]        : Intel(R) PRO/1000 7.3.18 PCI-E
 Driver Image Name  : Offset(0x1D200,0x3E9FF)
 Driver Version     : 07031800
 Driver Type        : Bus
 Configuration      : YES
 Diagnostics        : YES
 Child Controllers  :
   Ctrl[175]         : PciRoot(0x0)/Pci(0x2,0x3)/Pci(0x0,0x0)
   Child[219]        : Intel Ethernet Server Adapter I350-T2


FS0:\> dh 175 -d -v
175: 7219EA98
UnknownDevice
UnknownDevice
BusSpecificDriverOverride
LoadFile2
PCIIO
DevicePath
PciRoot(0x0)/Pci(0x2,0x3)/Pci(0x0,0x0)
 Controller Name    : PciRoot(0x0)/Pci(0x2,0x3)/Pci(0x0,0x0)
 Device Path        : PciRoot(0x0)/Pci(0x2,0x3)/Pci(0x0,0x0)
 Controller Type    : BUS
 Configuration      : NO
 Diagnostics        : NO
 Managed by         :
   Drv[176]          : Intel(R) PRO/1000 7.3.18 PCI-E
 Parent Controllers :
   Parent[75]       : PciRoot(0x0)
 Child Controllers  :
   Child[219]        : Intel Ethernet Server Adapter I350-T2


Thanks

On Wed, Jun 28, 2017 at 9:07 AM, Amit kumar <akamit91@hotmail.com<mailto:akamit91@hotmail.com><mailto:a
kamit91@hotmail.com<mailto:kamit91@hotmail.com>>> wrote:

Hi Tresko,

Hi Amit,

I am seeing the exact same issue on my system with DisconnectController.
The shell command 'disconnect' works but when I use the
gBS->DisconnectController >call from my application it doesn't seem to
really disconnect the same controller (even though the call returns
SUCCESS) and it still appears to be managed by >driver. I even checked the
ControllerHandle passed and it looks correct. Were you able to resolve this
eventually?


Yes i was able to resolve it. It might be helpful if you can tell me the
output of dh -d <handle> command.

Thanks

Amit


________________________________
From: TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com><mailto:tresko1@gmail.com>>
Sent: Wednesday, June 28, 2017 6:06:11 PM
To: edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org><mailto:edk2-devel@lists.01.org>;
akamit91@hotmail.com<mailto:akamit91@hotmail.com><mailto:akamit91@hotmail.com>

Subject: Re: [edk2] DisconnectController API not working.

BTW, I ported the disconnect code (under ShellPkg\Library\UefiShellDriver1CommandsLib)
to a standalone application and it fails too. Looks like there is something
else in the shell app that is taking care of properly disconnecting the
controller.

Thanks

On Tue, Jun 27, 2017 at 5:54 PM, TVKR <tresko1@gmail.com<mailto:tresko1@gmail.com><mailto:tres
ko1@gmail.com<mailto:ko1@gmail.com>>> wrote:
Hi Amit,

I am seeing the exact same issue on my system with DisconnectController.
The shell command 'disconnect' works but when I use the
gBS->DisconnectController call from my application it doesn't seem to
really disconnect the same controller (even though the call returns
SUCCESS) and it still appears to be managed by driver. I even checked the
ControllerHandle passed and it looks correct. Were you able to resolve this
eventually?


Thanks

-----Original Message-----
From: edk2-devel [mailto:edk2-devel-bounces at lists.01.org<http://lists.01.org><
https://lists.01.org/mailman/listinfo/edk2-devel>] On Behalf Of Amit kumar
Sent: Tuesday, April 4, 2017 12:07 AM
To: Tian, Feng <feng.tian at intel.com<http://intel.com><https://lists.01.
org/mailman/listinfo/edk2-devel>>
Cc: edk2-devel at lists.01.org<http://lists.01.org><https://lists.01.org/mailman/listinfo/edk2-
devel>
Subject: Re: [edk2] DisconnectController API not working.

No i am not trying to disconnect USB controller in use.

________________________________
From: Tian, Feng <feng.tian at intel.com<http://intel.com><https://lists.01.
org/mailman/listinfo/edk2-devel>>
Sent: Wednesday, April 5, 2017 8:21:36 AM
To: Andrew Fish; Amit kumar
Cc: edk2-devel at lists.01.org<http://lists.01.org><https://lists.01.org/mailman/listinfo/edk2-
devel>; Tian, Feng
Subject: RE: [edk2] DisconnectController API not working.

Kumar,

Do "map -r" at first please.


PS: for the issue you encounter, are you trying to disconnect the usb
controller in use? Please note at this time you are reading the file from
usb key. It means the disconnect will never succeed.

Thanks
Feng

-----Original Message-----
From: edk2-devel [mailto:edk2-devel-bounces at lists.01.org<http://lists.01.org><
https://lists.01.org/mailman/listinfo/edk2-devel>] On Behalf Of Andrew
Fish
Sent: Tuesday, April 4, 2017 12:07 AM
To: Amit kumar <akamit91 at hotmail.com<http://hotmail.com><https://lists.01.
org/mailman/listinfo/edk2-devel>>
Cc: edk2-devel at lists.01.org<http://lists.01.org><https://lists.01.org/mailman/listinfo/edk2-
devel>
Subject: Re: [edk2] DisconnectController API not working.


On Apr 3, 2017, at 2:58 AM, Amit kumar <akamit91 at hotmail.com<http://hotmail.com><
https://lists.01.org/mailman/listinfo/edk2-devel>> wrote:

Hi Andrew ,

I did some testing and found out, same code works fine on Server boards
but it fails on desktop boards.


Servers tend to connect less things as it can take a long time.

I Have double checked the ControllerHandle by printing its handle index
using ConvertHandleToHandleIndex(ControllerHandle), the  ControllerHandle
looks correct.


I have one new query;

i have made a startup.nsh file in which i am issuing load driver.efi , i
am getting file not found.

i have also tried using load fs0:driver.efi, and again i get  file not
found.  I have been trying to run this script from internal shell as well
as usb shell.


Does that command work if you boot the shell?

Thanks,

Andrew Fish


Regards

Amit Kumar

________________________________
From: afish at apple.com<http://apple.com><https://lists.01.org/mailman/listinfo/edk2-
devel> <afish at apple.com<http://apple.com><https://lists.01.org/mailman/listinfo/edk2-
devel>> on behalf of Andrew Fish
<afish at apple.com<http://apple.com><https://lists.01.org/mailman/listinfo/edk2-devel>>
Sent: Friday, March 31, 2017 10:42:08 PM
To: Amit kumar
Cc: edk2-devel at lists.01.org<http://lists.01.org><https://lists.01.
org/mailman/listinfo/edk2-devel>
Subject: Re: [edk2] DisconnectController API not working.


On Mar 31, 2017, at 5:26 AM, Amit kumar <akamit91 at hotmail.com<http://hotmail.com><
https://lists.01.org/mailman/listinfo/edk2-devel><mailto:akamit91 at
hotmail.com<http://hotmail.com><https://lists.01.org/mailman/listinfo/edk2-devel>>> wrote:


Hi ,

I am trying to disconnect a controller from the usb mass storage
driver, for which i am using

Status = gBS->DisconnectController (
  ControllerHandle,
  NULL,
  NULL
  );

after the call i get Status = SUCCESS , but when i run drivers command
in shell, i find it still being manged by the same old driver handle.
Although when i run disconnect command from shell (e.g disconnect 163
...output disconnect(163,0,0):Status = SUCCESS) it works as expected and
detaches the controller from the driver.
Can some body point me out what could be the reason.

Amit,

It is always good to double check you are using the right
ControllerHandle.

Sequence of events matters. There are boot flows and Shell commands that
do ConnectController.
~/work/src/edk2/ShellPkg(master)>git grep ConnectController
Library/UefiShellDebug1CommandsLib/LoadPciRom.c:402:
gBS->ConnectController (HandleBuffer[Index], NULL, NULL, TRUE);
Library/UefiShellDriver1CommandsLib/Connect.c:54:        Status =
gBS->ConnectController (Handle, NULL, RemainingDevicePath, FALSE);
Library/UefiShellDriver1CommandsLib/Connect.c:95:
gBS->ConnectController (RootBridgeHandleBuffer[RootBridgeIndex], NULL,
NULL, FALSE);
Library/UefiShellDriver1CommandsLib/Connect.c:116:ConnectControllers (
Library/UefiShellDriver1CommandsLib/Connect.c:150:  // This is where we
call the gBS->ConnectController function.
Library/UefiShellDriver1CommandsLib/Connect.c:158:      Status =
gBS->ConnectController(*HandleWalker, DriverHandleList, NULL, Recursive);
Library/UefiShellDriver1CommandsLib/Connect.c:167:    Status =
gBS->ConnectController(ControllerHandle, DriverHandleList, NULL,
Recursive);
Library/UefiShellDriver1CommandsLib/Connect.c:298:
Status = gBS->ConnectController (
Library/UefiShellDriver1CommandsLib/Connect.c:354:ConvertAndConnectCon
trollers (
Library/UefiShellDriver1CommandsLib/Connect.c:391:  return
(ConnectControllers(Handle1, Handle2, Recursive, Output, (BOOLEAN)(Handle2
!= NULL && Handle1 != NULL)));
Library/UefiShellDriver1CommandsLib/Connect.c:536:          Status =
ConvertAndConnectControllers(Handle1, Handle2, ShellCommandLineGetFlag(Package,
L"-r"), (BOOLEAN)(Count!=0));
Library/UefiShellDriver1CommandsLib/DrvCfg.c:492:  EFI_HANDLE
ConnectControllerContextOverride[2];
Library/UefiShellDriver1CommandsLib/DrvCfg.c:514:
ConnectControllerContextOverride[0]  = DriverImageHandle;
Library/UefiShellDriver1CommandsLib/DrvCfg.c:515:
ConnectControllerContextOverride[1]  = NULL;
Library/UefiShellDriver1CommandsLib/DrvCfg.c:516:
gBS->ConnectController (ControllerHandle, ConnectControllerContextOverride,
NULL, TRUE);
Library/UefiShellLevel2CommandsLib/Load.c:52:    Status =
gBS->ConnectController (HandleBuffer[Index], NULL, NULL, TRUE);


Thanks,

Andrew Fish

_______________________________________________
edk2-devel mailing list
edk2-devel at lists.01.org<http://lists.01.org><https://lists.01.org/mailman/listinfo/edk2-
devel><mailto:edk2-devel at lists.01.org<http://lists.01.org><https://lists.01.
org/mailman/listinfo/edk2-devel>>
https://lists.01.org/mailman/listinfo/edk2-devel

_______________________________________________
edk2-devel mailing list
edk2-devel at lists.01.org<https://lists.01.org/mailman/listinfo/edk2-
devel>
https://lists.01.org/mailman/listinfo/edk2-devel







_______________________________________________
edk2-devel mailing list
edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
https://lists.01.org/mailman/listinfo/edk2-devel

_______________________________________________
edk2-devel mailing list
edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
https://lists.01.org/mailman/listinfo/edk2-devel


  reply	other threads:[~2017-06-28 19:35 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-06-27 22:54 DisconnectController API not working TVKR
2017-06-28 12:36 ` TVKR
2017-06-28 14:07   ` Amit kumar
2017-06-28 14:35     ` TVKR
2017-06-28 15:29       ` Amit kumar
2017-06-28 17:15         ` TVKR
2017-06-28 17:50           ` Amit kumar
2017-06-28 17:51             ` TVKR
2017-06-28 17:56               ` Amit kumar
2017-06-28 18:01                 ` TVKR
2017-06-28 18:02                   ` TVKR
2017-06-28 18:11                   ` Amit kumar
2017-06-28 18:23                     ` Amit kumar
2017-06-28 19:27                       ` TVKR
2017-06-28 19:34                         ` Andrew Fish
2017-06-28 19:37                           ` Amit kumar [this message]
2017-06-28 19:53                             ` TVKR
2017-06-29 19:58                               ` TVKR
  -- strict thread matches above, loose matches on Subject: below --
2017-03-31 12:26 Amit kumar
2017-03-31 17:12 ` Andrew Fish
2017-04-03  9:58   ` Amit kumar
2017-04-03 16:07     ` Andrew Fish
2017-04-05  2:51       ` Tian, Feng
2017-04-05  3:38         ` Amit kumar

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-list from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=DM5PR11MB15780CD866011EF25FCC04A2DCDD0@DM5PR11MB1578.namprd11.prod.outlook.com \
    --to=devel@edk2.groups.io \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox