On 10/31/23 02:41, Wu, Hao A wrote:
> (Add MdePkg maintainers for their feedbacks)
>
> Sorry that I do not have strong opinion on this one.
>
> Some of my thoughts are:
> * If you find the to-be-added APIs can be used in serveral places to reduce
> repetative codes, then it will be worthwhile to add new library APIs.
>
> * TimerLib has many instance implementations, my take is that many post-mem
> instances use library level global variable to store information like timer
> frequency to avoid calculating it every time.
> For pre-mem instances, such caching is not feasible. I think it will be the
> API consumer's responsibility to limit the usage of these APIs for
> performance
> consideration.
I wouldn't recommend extending the TimerLib class. TimerLib has a huge
amount of instances (implementations), and adding the (effectively) same
calculation to each is a waste of developer effort, reviewer effort, and
long-term maintenance (lots of duplications).
Introducing a separate TimerTickDiffLib class is what I proposed in
2017. It can be small and simple; do one thing and do it well. It would
be widely used (*many* sites need tick differences). Yet keeping it
separate from TimerLib is absolutely justified, as TimerTickDiffLib
would be independent of both timer hardware specifics *and* of firmware
phase too (RAM or no RAM).
The functions -- or more likely, the single function -- in
TimerTickDiffLib can easily take the performance counter characteristics
as parameters. Then callers in RAM-based modules may pass those args
from global variables in which they cache the perf counter
characteristics themselves. No need to move the caching into a
particular (RAM-bound) TimerTickDiffLib instance. And callers that XIP
from flash can fill in those args from repeated / separate
GetPerformanceCounterProperties() calls, or -- in the luckier case --
cache the output of GetPerformanceCounterProperties() for each *call
site*, in local variables.
A futher API optimization may be to introduce a new (small) structure
type in TimerTickDiffLib, for collecting the outputs of
GetPerformanceCounterProperties(). Then the sole function in
TimerTickDiffLib would take a pointer to a *constant* structure of this
type.
Thanks
Laszlo
>
> Best Regards,
> Hao Wu
>
>> -----Original Message-----
>> From: devel@edk2.groups.io <devel@edk2.groups.io> On Behalf Of Henz,
>> Patrick
>> Sent: Tuesday, October 31, 2023 4:36 AM
>> To: Wu, Hao A <hao.a...@intel.com>; devel@edk2.groups.io; Michael Brown
>> <mc...@ipxe.org>
>> Subject: Re: [edk2-devel] [PATCH] MdeModulePkg/XhciDxe: Use Performance
>> Timer for XHCI Timeouts
>>
>> My changes have been in for a little over a month now, I've been looking into
>> updating the TimerLib API but I'm questioning if it still makes sense to go
>> down
>> that path. The functions I had implemented, XhcGetElapsedTicks () and
>> XhcConvertTimeToTicks (), now rely heavily on the use of global variables for
>> caching return values from GetPerformanceCounterProperties. As-is these
>> functions won't work for all instances of TimerLib, specifically if the
>> instance is
>> used before memory (RAM) is available. I could implement the functions as
>> they originally were, but I wouldn't be able to replace the Xhc functions
>> without adding additional parameters to said TimerLib functions (e.g. adding
>> Frequency, StartValue, StopValue to ConvertTimeToTicks), which feels clunky
>> to me. Would a new library make sense here? Something that sits between
>> the driver and the TimerLib library? Or do we just leave things as they
>> currently are?
>>
>> Thanks,
>> Patrick Henz
>>
>> -----Original Message-----
>> From: Wu, Hao A <hao.a...@intel.com>
>> Sent: Thursday, August 10, 2023 8:43 PM
>> To: Henz, Patrick <patrick.h...@hpe.com>; devel@edk2.groups.io; Michael
>> Brown <mc...@ipxe.org>
>> Subject: RE: [edk2-devel] [PATCH] MdeModulePkg/XhciDxe: Use Performance
>> Timer for XHCI Timeouts
>>
>> My personal preference is to do this as two steps.
>> Address the issue in XhciDxe first. Then add new TimerLib API to replace all
>> driver/library internal implementations.
>>
>> Best Regards,
>> Hao Wu
>>
>>> -----Original Message-----
>>> From: Henz, Patrick <patrick.h...@hpe.com>
>>> Sent: Friday, August 11, 2023 6:45 AM
>>> To: devel@edk2.groups.io; Wu, Hao A <hao.a...@intel.com>; Michael
>>> Brown <mc...@ipxe.org>
>>> Subject: RE: [edk2-devel] [PATCH] MdeModulePkg/XhciDxe: Use
>>> Performance Timer for XHCI Timeouts
>>>
>>> I can certainly make that change.
>>>
>>> For what it's worth I have been working on adding the new function,
>>> GetElapsedTicks, to the various implementations of TimerLib. I've
>>> finished up testing, I would just need to clean up the commits for a
>>> patch. Should I move forward with that, or would we rather I just add
>>> XhcGetElapsedTime to XhciDxe for the time being?
>>>
>>> Thanks,
>>> Patrick Henz
>>>
>>> -----Original Message-----
>>> From: devel@edk2.groups.io <devel@edk2.groups.io> On Behalf Of Wu,
>> Hao
>>> A
>>> Sent: Sunday, July 30, 2023 9:57 PM
>>> To: devel@edk2.groups.io; Henz, Patrick <patrick.h...@hpe.com>;
>>> Michael Brown <mc...@ipxe.org>
>>> Subject: Re: [edk2-devel] [PATCH] MdeModulePkg/XhciDxe: Use
>>> Performance Timer for XHCI Timeouts
>>>
>>> For the 2 occurrences of in XhcWaitOpRegBit & XhcExecTransfer:
>>> TimeoutTime = XHC_MICROSECOND_TO_NANOSECOND (Timeout *
>>> XHC_1_MILLISECOND); How about changing them to:
>>> TimeoutTime = XHC_MICROSECOND_TO_NANOSECOND ((UINT64)
>> Timeout
>>> * XHC_1_MILLISECOND); to address possible overflow during "Timeout *
>>> XHC_1_MILLISECOND"?
>>>
>>> For extending XhcGetElapsedTime as a TimerLib API, I am fine to put it
>>> in XhciDxe at this moment.
>>> If package maintainers suggest to make it as a public library API, my
>>> take is that this should be done in a separate commit.
>>>
>>> Best Regards,
>>> Hao Wu
>>>
>>>> -----Original Message-----
>>>> From: devel@edk2.groups.io <devel@edk2.groups.io> On Behalf Of Henz,
>>>> Patrick
>>>> Sent: Thursday, July 6, 2023 4:16 AM
>>>> To: devel@edk2.groups.io
>>>> Cc: Henz, Patrick <patrick.h...@hpe.com>
>>>> Subject: [edk2-devel] [PATCH] MdeModulePkg/XhciDxe: Use Performance
>>>> Timer for XHCI Timeouts
>>>>
>>>> REF:INVALID URI REMOVED
>>>> bu
>>>> g.cgi?id=2948__;!!NpxR!kyFQM5IkKYAG9CRBO4xphwBnzi_jhb2oU-
>>> NKbMjOV-lctg5
>>>> _B3K1Lcta452Gx-1twRt8At3cueAYDq_n$
>>>>
>>>> XhciDxe uses the timer functionality provided by the boot services
>>>> table to detect timeout conditions. This breaks the driver's
>>>> ExitBootServices call back, as CoreExitBootServices halts the timer
>>>> before signaling the ExitBootServices event. If the host controller
>>>> fails to halt in the call back, the timeout condition will never
>>>> occur and the boot gets stuck in an indefinite spin loop. Use the
>>>> free running timer provided by TimerLib to calculate timeouts,
>>>> avoiding the potential hang.
>>>>
>>>> Signed-off-by: Patrick Henz <patrick.h...@hpe.com>
>>>> Reviewed-by:
>>>> ---
>>>> MdeModulePkg/Bus/Pci/XhciDxe/Xhci.c | 56 +++++++++++++++++++
>>>> MdeModulePkg/Bus/Pci/XhciDxe/Xhci.h | 22 ++++++++
>>>> MdeModulePkg/Bus/Pci/XhciDxe/XhciDxe.inf | 2 +
>>>> MdeModulePkg/Bus/Pci/XhciDxe/XhciReg.c | 67 +++++++++--------------
>>>> MdeModulePkg/Bus/Pci/XhciDxe/XhciSched.c | 68
>>>> +++++++++---------------
>>>> 5 files changed, 129 insertions(+), 86 deletions(-)
>>>>
>>>> diff --git a/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.c
>>>> b/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.c
>>>> index 62682dd27c..1dcbe20512 100644
>>>> --- a/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.c
>>>> +++ b/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.c
>>>> @@ -1,6 +1,7 @@
>>>> /** @file
>>>>
>>>> The XHCI controller driver.
>>>>
>>>>
>>>>
>>>> +(C) Copyright 2023 Hewlett Packard Enterprise Development LP<BR>
>>>>
>>>> Copyright (c) 2011 - 2022, Intel Corporation. All rights
>>>> reserved.<BR>
>>>>
>>>> SPDX-License-Identifier: BSD-2-Clause-Patent
>>>>
>>>>
>>>>
>>>> @@ -2294,3 +2295,58 @@ XhcDriverBindingStop (
>>>>
>>>>
>>>> return EFI_SUCCESS;
>>>>
>>>> }
>>>>
>>>> +
>>>>
>>>> +/**
>>>>
>>>> + Computes and returns the elapsed time in nanoseconds since
>>>>
>>>> + PreviousTick. The value of PreviousTick is overwritten with the
>>>>
>>>> + current performance counter value.
>>>>
>>>> +
>>>>
>>>> + @param PreviousTick Pointer to PreviousTick count.
>>>>
>>>> + @return The elapsed time in nanoseconds since PreviousCount.
>>>>
>>>> + PreviousCount is overwritten with the current performance
>>>>
>>>> + counter value.
>>>>
>>>> +**/
>>>>
>>>> +UINT64
>>>>
>>>> +XhcGetElapsedTime (
>>>>
>>>> + IN OUT UINT64 *PreviousTick
>>>>
>>>> + )
>>>>
>>>> +{
>>>>
>>>> + UINT64 StartValue;
>>>>
>>>> + UINT64 EndValue;
>>>>
>>>> + UINT64 CurrentTick;
>>>>
>>>> + UINT64 Delta;
>>>>
>>>> +
>>>>
>>>> + GetPerformanceCounterProperties (&StartValue, &EndValue);
>>>>
>>>> +
>>>>
>>>> + CurrentTick = GetPerformanceCounter ();
>>>>
>>>> +
>>>>
>>>> + //
>>>>
>>>> + // Determine if the counter is counting up or down
>>>>
>>>> + //
>>>>
>>>> + if (StartValue < EndValue) {
>>>>
>>>> + //
>>>>
>>>> + // Counter counts upwards, check for an overflow condition
>>>>
>>>> + //
>>>>
>>>> + if (*PreviousTick > CurrentTick) {
>>>>
>>>> + Delta = (EndValue - *PreviousTick) + CurrentTick;
>>>>
>>>> + } else {
>>>>
>>>> + Delta = CurrentTick - *PreviousTick;
>>>>
>>>> + }
>>>>
>>>> + } else {
>>>>
>>>> + //
>>>>
>>>> + // Counter counts downwards, check for an underflow condition
>>>>
>>>> + //
>>>>
>>>> + if (*PreviousTick < CurrentTick) {
>>>>
>>>> + Delta = (StartValue - CurrentTick) + *PreviousTick;
>>>>
>>>> + } else {
>>>>
>>>> + Delta = *PreviousTick - CurrentTick;
>>>>
>>>> + }
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + //
>>>>
>>>> + // Set PreviousTick to CurrentTick
>>>>
>>>> + //
>>>>
>>>> + *PreviousTick = CurrentTick;
>>>>
>>>> +
>>>>
>>>> + return GetTimeInNanoSecond (Delta);
>>>>
>>>> +}
>>>>
>>>> diff --git a/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.h
>>>> b/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.h
>>>> index ca223bd20c..77feb66647 100644
>>>> --- a/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.h
>>>> +++ b/MdeModulePkg/Bus/Pci/XhciDxe/Xhci.h
>>>> @@ -2,6 +2,7 @@
>>>>
>>>>
>>>> Provides some data structure definitions used by the XHCI host
>>>> controller driver.
>>>>
>>>>
>>>>
>>>> +(C) Copyright 2023 Hewlett Packard Enterprise Development LP<BR>
>>>>
>>>> Copyright (c) 2011 - 2017, Intel Corporation. All rights
>>>> reserved.<BR>
>>>>
>>>> Copyright (c) Microsoft Corporation.<BR>
>>>>
>>>> SPDX-License-Identifier: BSD-2-Clause-Patent
>>>>
>>>> @@ -26,6 +27,7 @@ SPDX-License-Identifier: BSD-2-Clause-Patent
>>>> #include <Library/UefiLib.h>
>>>>
>>>> #include <Library/DebugLib.h>
>>>>
>>>> #include <Library/ReportStatusCodeLib.h>
>>>>
>>>> +#include <Library/TimerLib.h>
>>>>
>>>>
>>>>
>>>> #include <IndustryStandard/Pci.h>
>>>>
>>>>
>>>>
>>>> @@ -37,6 +39,10 @@ typedef struct _USB_DEV_CONTEXT
>>> USB_DEV_CONTEXT;
>>>> #include "ComponentName.h"
>>>>
>>>> #include "UsbHcMem.h"
>>>>
>>>>
>>>>
>>>> +//
>>>>
>>>> +// Converts a count from microseconds to nanoseconds
>>>>
>>>> +//
>>>>
>>>> +#define XHC_MICROSECOND_TO_NANOSECOND(Time) ((UINT64)(Time)
>> *
>>>> 1000)
>>>>
>>>> //
>>>>
>>>> // The unit is microsecond, setting it as 1us.
>>>>
>>>> //
>>>>
>>>> @@ -720,4 +726,20 @@ XhcAsyncIsochronousTransfer (
>>>> IN VOID *Context
>>>>
>>>> );
>>>>
>>>>
>>>>
>>>> +/**
>>>>
>>>> + Computes and returns the elapsed time in nanoseconds since
>>>>
>>>> + PreviousTick. The value of PreviousTick is overwritten with the
>>>>
>>>> + current performance counter value.
>>>>
>>>> +
>>>>
>>>> + @param PreviousTick Pointer to PreviousTick count.
>>>>
>>>> + before calling this function.
>>>>
>>>> + @return The elapsed time in nanoseconds since PreviousCount.
>>>>
>>>> + PreviousCount is overwritten with the current performance
>>>>
>>>> + counter value.
>>>>
>>>> +**/
>>>>
>>>> +UINT64
>>>>
>>>> +XhcGetElapsedTime (
>>>>
>>>> + IN OUT UINT64 *PreviousTick
>>>>
>>>> + );
>>>>
>>>> +
>>>>
>>>> #endif
>>>>
>>>> diff --git a/MdeModulePkg/Bus/Pci/XhciDxe/XhciDxe.inf
>>>> b/MdeModulePkg/Bus/Pci/XhciDxe/XhciDxe.inf
>>>> index 5865d86822..18ef87916a 100644
>>>> --- a/MdeModulePkg/Bus/Pci/XhciDxe/XhciDxe.inf
>>>> +++ b/MdeModulePkg/Bus/Pci/XhciDxe/XhciDxe.inf
>>>> @@ -3,6 +3,7 @@
>>>> # It implements the interfaces of monitoring the status of all
>>>> ports and transferring
>>>>
>>>> # Control, Bulk, Interrupt and Isochronous requests to those
>>>> attached usb LS/FS/HS/SS devices.
>>>>
>>>> #
>>>>
>>>> +# (C) Copyright 2023 Hewlett Packard Enterprise Development LP<BR>
>>>>
>>>> # Copyright (c) 2011 - 2018, Intel Corporation. All rights
>>>> reserved.<BR>
>>>>
>>>> #
>>>>
>>>> # SPDX-License-Identifier: BSD-2-Clause-Patent
>>>>
>>>> @@ -54,6 +55,7 @@
>>>> BaseMemoryLib
>>>>
>>>> DebugLib
>>>>
>>>> ReportStatusCodeLib
>>>>
>>>> + TimerLib
>>>>
>>>>
>>>>
>>>> [Guids]
>>>>
>>>> gEfiEventExitBootServicesGuid ## SOMETIMES_CONSUMES ##
>>>> Event
>>>>
>>>> diff --git a/MdeModulePkg/Bus/Pci/XhciDxe/XhciReg.c
>>>> b/MdeModulePkg/Bus/Pci/XhciDxe/XhciReg.c
>>>> index 5700fc5fb8..2499698715 100644
>>>> --- a/MdeModulePkg/Bus/Pci/XhciDxe/XhciReg.c
>>>> +++ b/MdeModulePkg/Bus/Pci/XhciDxe/XhciReg.c
>>>> @@ -2,6 +2,7 @@
>>>>
>>>>
>>>> The XHCI register operation routines.
>>>>
>>>>
>>>>
>>>> +(C) Copyright 2023 Hewlett Packard Enterprise Development LP<BR>
>>>>
>>>> Copyright (c) 2011 - 2017, Intel Corporation. All rights
>>>> reserved.<BR>
>>>>
>>>> SPDX-License-Identifier: BSD-2-Clause-Patent
>>>>
>>>>
>>>>
>>>> @@ -417,15 +418,14 @@ XhcClearOpRegBit (
>>>> Wait the operation register's bit as specified by Bit
>>>>
>>>> to become set (or clear).
>>>>
>>>>
>>>>
>>>> - @param Xhc The XHCI Instance.
>>>>
>>>> - @param Offset The offset of the operation register.
>>>>
>>>> - @param Bit The bit of the register to wait for.
>>>>
>>>> - @param WaitToSet Wait the bit to set or clear.
>>>>
>>>> - @param Timeout The time to wait before abort (in
>>>> millisecond,
>>>> ms).
>>>>
>>>> + @param Xhc The XHCI Instance.
>>>>
>>>> + @param Offset The offset of the operation register.
>>>>
>>>> + @param Bit The bit of the register to wait for.
>>>>
>>>> + @param WaitToSet Wait the bit to set or clear.
>>>>
>>>> + @param Timeout The time to wait before abort (in millisecond, ms).
>>>>
>>>>
>>>>
>>>> - @retval EFI_SUCCESS The bit successfully changed by host
>>> controller.
>>>>
>>>> - @retval EFI_TIMEOUT The time out occurred.
>>>>
>>>> - @retval EFI_OUT_OF_RESOURCES Memory for the timer event could
>> not
>>>> be allocated.
>>>>
>>>> + @retval EFI_SUCCESS The bit successfully changed by host controller.
>>>>
>>>> + @retval EFI_TIMEOUT The time out occurred.
>>>>
>>>>
>>>>
>>>> **/
>>>>
>>>> EFI_STATUS
>>>>
>>>> @@ -437,54 +437,35 @@ XhcWaitOpRegBit (
>>>> IN UINT32 Timeout
>>>>
>>>> )
>>>>
>>>> {
>>>>
>>>> - EFI_STATUS Status;
>>>>
>>>> - EFI_EVENT TimeoutEvent;
>>>>
>>>> -
>>>>
>>>> - TimeoutEvent = NULL;
>>>>
>>>> + UINT64 TimeoutTime;
>>>>
>>>> + UINT64 ElapsedTime;
>>>>
>>>> + UINT64 TimeDelta;
>>>>
>>>> + UINT64 CurrentTick;
>>>>
>>>>
>>>>
>>>> if (Timeout == 0) {
>>>>
>>>> return EFI_TIMEOUT;
>>>>
>>>> }
>>>>
>>>>
>>>>
>>>> - Status = gBS->CreateEvent (
>>>>
>>>> - EVT_TIMER,
>>>>
>>>> - TPL_CALLBACK,
>>>>
>>>> - NULL,
>>>>
>>>> - NULL,
>>>>
>>>> - &TimeoutEvent
>>>>
>>>> - );
>>>>
>>>> -
>>>>
>>>> - if (EFI_ERROR (Status)) {
>>>>
>>>> - goto DONE;
>>>>
>>>> - }
>>>>
>>>> -
>>>>
>>>> - Status = gBS->SetTimer (
>>>>
>>>> - TimeoutEvent,
>>>>
>>>> - TimerRelative,
>>>>
>>>> - EFI_TIMER_PERIOD_MILLISECONDS (Timeout)
>>>>
>>>> - );
>>>>
>>>> -
>>>>
>>>> - if (EFI_ERROR (Status)) {
>>>>
>>>> - goto DONE;
>>>>
>>>> - }
>>>>
>>>> + TimeoutTime = XHC_MICROSECOND_TO_NANOSECOND (Timeout *
>>>> XHC_1_MILLISECOND);
>>>>
>>>> + ElapsedTime = 0;
>>>>
>>>> + CurrentTick = GetPerformanceCounter ();
>>>>
>>>>
>>>>
>>>> do {
>>>>
>>>> if (XHC_REG_BIT_IS_SET (Xhc, Offset, Bit) == WaitToSet) {
>>>>
>>>> - Status = EFI_SUCCESS;
>>>>
>>>> - goto DONE;
>>>>
>>>> + return EFI_SUCCESS;
>>>>
>>>> }
>>>>
>>>>
>>>>
>>>> gBS->Stall (XHC_1_MICROSECOND);
>>>>
>>>> - } while (EFI_ERROR (gBS->CheckEvent (TimeoutEvent)));
>>>>
>>>> -
>>>>
>>>> - Status = EFI_TIMEOUT;
>>>>
>>>> + TimeDelta = XhcGetElapsedTime (&CurrentTick);
>>>>
>>>> + // Ensure that ElapsedTime is always incremented to avoid
>>>> + indefinite
>>>> hangs
>>>>
>>>> + if (TimeDelta == 0) {
>>>>
>>>> + TimeDelta = XHC_MICROSECOND_TO_NANOSECOND
>>>> (XHC_1_MICROSECOND);
>>>>
>>>> + }
>>>>
>>>>
>>>>
>>>> -DONE:
>>>>
>>>> - if (TimeoutEvent != NULL) {
>>>>
>>>> - gBS->CloseEvent (TimeoutEvent);
>>>>
>>>> - }
>>>>
>>>> + ElapsedTime += TimeDelta;
>>>>
>>>> + } while (ElapsedTime < TimeoutTime);
>>>>
>>>>
>>>>
>>>> - return Status;
>>>>
>>>> + return EFI_TIMEOUT;
>>>>
>>>> }
>>>>
>>>>
>>>>
>>>> /**
>>>>
>>>> diff --git a/MdeModulePkg/Bus/Pci/XhciDxe/XhciSched.c
>>>> b/MdeModulePkg/Bus/Pci/XhciDxe/XhciSched.c
>>>> index 298fb88b81..5177886e52 100644
>>>> --- a/MdeModulePkg/Bus/Pci/XhciDxe/XhciSched.c
>>>> +++ b/MdeModulePkg/Bus/Pci/XhciDxe/XhciSched.c
>>>> @@ -2,6 +2,7 @@
>>>>
>>>>
>>>> XHCI transfer scheduling routines.
>>>>
>>>>
>>>>
>>>> +(C) Copyright 2023 Hewlett Packard Enterprise Development LP<BR>
>>>>
>>>> Copyright (c) 2011 - 2020, Intel Corporation. All rights
>>>> reserved.<BR>
>>>>
>>>> Copyright (c) Microsoft Corporation.<BR>
>>>>
>>>> Copyright (C) 2022 Advanced Micro Devices, Inc. All rights
>>>> reserved.<BR>
>>>>
>>>> @@ -1273,15 +1274,14 @@ EXIT:
>>>> /**
>>>>
>>>> Execute the transfer by polling the URB. This is a synchronous
>>>> operation.
>>>>
>>>>
>>>>
>>>> - @param Xhc The XHCI Instance.
>>>>
>>>> - @param CmdTransfer The executed URB is for cmd transfer or
>> not.
>>>>
>>>> - @param Urb The URB to execute.
>>>>
>>>> - @param Timeout The time to wait before abort, in
>>>> millisecond.
>>>>
>>>> + @param Xhc The XHCI Instance.
>>>>
>>>> + @param CmdTransfer The executed URB is for cmd transfer or not.
>>>>
>>>> + @param Urb The URB to execute.
>>>>
>>>> + @param Timeout The time to wait before abort, in millisecond.
>>>>
>>>>
>>>>
>>>> - @return EFI_DEVICE_ERROR The transfer failed due to transfer
>> error.
>>>>
>>>> - @return EFI_TIMEOUT The transfer failed due to time out.
>>>>
>>>> - @return EFI_SUCCESS The transfer finished OK.
>>>>
>>>> - @retval EFI_OUT_OF_RESOURCES Memory for the timer event could
>> not
>>>> be allocated.
>>>>
>>>> + @return EFI_DEVICE_ERROR The transfer failed due to transfer error.
>>>>
>>>> + @return EFI_TIMEOUT The transfer failed due to time out.
>>>>
>>>> + @return EFI_SUCCESS The transfer finished OK.
>>>>
>>>>
>>>>
>>>> **/
>>>>
>>>> EFI_STATUS
>>>>
>>>> @@ -1296,12 +1296,14 @@ XhcExecTransfer (
>>>> UINT8 SlotId;
>>>>
>>>> UINT8 Dci;
>>>>
>>>> BOOLEAN Finished;
>>>>
>>>> - EFI_EVENT TimeoutEvent;
>>>>
>>>> + UINT64 TimeoutTime;
>>>>
>>>> + UINT64 ElapsedTime;
>>>>
>>>> + UINT64 TimeDelta;
>>>>
>>>> + UINT64 CurrentTick;
>>>>
>>>> BOOLEAN IndefiniteTimeout;
>>>>
>>>>
>>>>
>>>> Status = EFI_SUCCESS;
>>>>
>>>> Finished = FALSE;
>>>>
>>>> - TimeoutEvent = NULL;
>>>>
>>>> IndefiniteTimeout = FALSE;
>>>>
>>>>
>>>>
>>>> if (CmdTransfer) {
>>>>
>>>> @@ -1319,34 +1321,14 @@ XhcExecTransfer (
>>>>
>>>>
>>>> if (Timeout == 0) {
>>>>
>>>> IndefiniteTimeout = TRUE;
>>>>
>>>> - goto RINGDOORBELL;
>>>>
>>>> - }
>>>>
>>>> -
>>>>
>>>> - Status = gBS->CreateEvent (
>>>>
>>>> - EVT_TIMER,
>>>>
>>>> - TPL_CALLBACK,
>>>>
>>>> - NULL,
>>>>
>>>> - NULL,
>>>>
>>>> - &TimeoutEvent
>>>>
>>>> - );
>>>>
>>>> -
>>>>
>>>> - if (EFI_ERROR (Status)) {
>>>>
>>>> - goto DONE;
>>>>
>>>> }
>>>>
>>>>
>>>>
>>>> - Status = gBS->SetTimer (
>>>>
>>>> - TimeoutEvent,
>>>>
>>>> - TimerRelative,
>>>>
>>>> - EFI_TIMER_PERIOD_MILLISECONDS (Timeout)
>>>>
>>>> - );
>>>>
>>>> -
>>>>
>>>> - if (EFI_ERROR (Status)) {
>>>>
>>>> - goto DONE;
>>>>
>>>> - }
>>>>
>>>> -
>>>>
>>>> -RINGDOORBELL:
>>>>
>>>> XhcRingDoorBell (Xhc, SlotId, Dci);
>>>>
>>>>
>>>>
>>>> + TimeoutTime = XHC_MICROSECOND_TO_NANOSECOND (Timeout *
>>>> XHC_1_MILLISECOND);
>>>>
>>>> + ElapsedTime = 0;
>>>>
>>>> + CurrentTick = GetPerformanceCounter ();
>>>>
>>>> +
>>>>
>>>> do {
>>>>
>>>> Finished = XhcCheckUrbResult (Xhc, Urb);
>>>>
>>>> if (Finished) {
>>>>
>>>> @@ -1354,22 +1336,22 @@ RINGDOORBELL:
>>>> }
>>>>
>>>>
>>>>
>>>> gBS->Stall (XHC_1_MICROSECOND);
>>>>
>>>> - } while (IndefiniteTimeout || EFI_ERROR (gBS->CheckEvent
>>>> (TimeoutEvent)));
>>>>
>>>> + TimeDelta = XhcGetElapsedTime (&CurrentTick);
>>>>
>>>> + // Ensure that ElapsedTime is always incremented to avoid
>>>> + indefinite
>>>> hangs
>>>>
>>>> + if (TimeDelta == 0) {
>>>>
>>>> + TimeDelta = XHC_MICROSECOND_TO_NANOSECOND
>>>> (XHC_1_MICROSECOND);
>>>>
>>>> + }
>>>>
>>>>
>>>>
>>>> -DONE:
>>>>
>>>> - if (EFI_ERROR (Status)) {
>>>>
>>>> - Urb->Result = EFI_USB_ERR_NOTEXECUTE;
>>>>
>>>> - } else if (!Finished) {
>>>>
>>>> + ElapsedTime += TimeDelta;
>>>>
>>>> + } while (IndefiniteTimeout || ElapsedTime < TimeoutTime);
>>>>
>>>> +
>>>>
>>>> + if (!Finished) {
>>>>
>>>> Urb->Result = EFI_USB_ERR_TIMEOUT;
>>>>
>>>> Status = EFI_TIMEOUT;
>>>>
>>>> } else if (Urb->Result != EFI_USB_NOERROR) {
>>>>
>>>> Status = EFI_DEVICE_ERROR;
>>>>
>>>> }
>>>>
>>>>
>>>>
>>>> - if (TimeoutEvent != NULL) {
>>>>
>>>> - gBS->CloseEvent (TimeoutEvent);
>>>>
>>>> - }
>>>>
>>>> -
>>>>
>>>> return Status;
>>>>
>>>> }
>>>>
>>>>
>>>>
>>>> --
>>>> 2.34.1
>>>>
>>>>
>>>>
>>>> -=-=-=-=-=-=
>>>> Groups.io Links: You receive all messages sent to this group.
>>>> View/Reply Online (#106668):
>>>> INVALID URI REMOVED
>>>> 06
>>>> 668__;!!NpxR!kyFQM5IkKYAG9CRBO4xphwBnzi_jhb2oU-NKbMjOV-
>>> lctg5_B3K1Lcta4
>>>> 52Gx-1twRt8At3cuWxQsO18$ Mute This Topic:
>>>> https://groups.io/mt/99972791/1768737
>>>> NpxR!kyFQM5IkKYAG9CRBO4xphwBnzi_jhb2oU-NKbMjOV-
>>> lctg5_B3K1Lcta452Gx-1tw
>>>> Rt8At3cubh2HDyk$ Group Owner: devel+ow...@edk2.groups.io
>>>> Unsubscribe:
>>>> https://edk2.groups.io/g/devel/unsub
>>>> !N
>>>> pxR!kyFQM5IkKYAG9CRBO4xphwBnzi_jhb2oU-NKbMjOV-
>>> lctg5_B3K1Lcta452Gx-1twR
>>>> t8At3cuWYMMXVU$ [hao.a...@intel.com] -=-=-=-=-=-=
>>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>>
>>
>>
>>
>
>
>
>
>
>
-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#110403): https://edk2.groups.io/g/devel/message/110403
Mute This Topic: https://groups.io/mt/99972791/21656
Group Owner: devel+ow...@edk2.groups.io
Unsubscribe:
https://edk2.groups.io/g/devel/leave/9847357/21656/1706620634/xyzzy
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
