Re: [edk2-devel] [edk2-stable202402 PATCH 1/2] UefiCpuPkg/PiSmmCpuDxeSmm: distinguish GetSmBase() failure modes

["Laszlo Ersek" <lersek@redhat.com>]

On 2/14/24 04:43, Michael D Kinney wrote:
> Hi Laszlo,
> 
> Thank you for the quick fix.
> 
> I have reviewed the changes.  I agree they fix the issue at hand.
> 
> Reviewed-by: Michael D Kinney <michael.d.kinney@intel.com>
> 
> I have adjusted the commit message with your suggested changes in
> the PR I have prepared:
> 
> https://github.com/tianocore/edk2/pull/5373
> 
> There may be better ways to organize this code in general to make
> it easier to understand and maintain in the future, but we can
> let Ray review that when he returns.  That will also likely be a 
> much bugger change that can be accepted just before a release.
> 
> I also approve this as a critical fix for edk2-stable202402
> 
> I will wait till tomorrow morning my time to see if Gerd and 
> Rahul and Leif can also provide their reviews/approvals and
> to give me some time to run some tests.
> 
> I do not expect Ray Ni or Dun Tan to be available this week.

Thank you all very much for the feedback!
Laszlo

> 
> Best regards,
> 
> Mike
> 
>> -----Original Message-----
>> From: devel@edk2.groups.io <devel@edk2.groups.io> On Behalf Of Laszlo
>> Ersek
>> Sent: Tuesday, February 13, 2024 1:36 PM
>> To: devel@edk2.groups.io
>> Cc: Tan, Dun <dun.tan@intel.com>; Gerd Hoffmann <kraxel@redhat.com>;
>> Kumar, Rahul R <rahul.r.kumar@intel.com>; Ni, Ray <ray.ni@intel.com>
>> Subject: Re: [edk2-devel] [edk2-stable202402 PATCH 1/2]
>> UefiCpuPkg/PiSmmCpuDxeSmm: distinguish GetSmBase() failure modes
>>
>> On 2/13/24 22:09, Laszlo Ersek wrote:
>>> Commit 725acd0b9cc0 ("UefiCpuPkg: Avoid assuming only one
>> smmbasehob",
>>> 2023-12-12) introduced a helper function called GetSmBase(),
>> replacing the
>>> lookup of the first and only "gSmmBaseHobGuid" GUID HOB, with
>> iterated
>>> lookups plus memory allocation.
>>>
>>> This introduced a new failure mode for setting
>> "mCpuHotPlugData.SmBase".
>>> Namely, before commit 725acd0b9cc0, "mCpuHotPlugData.SmBase" would be
>> set
>>> to NULL if and only if the GUID HOB was absent. After the commit, a
>> NULL
>>> assignment would be possible if the GUID HOB was absent, *or* one of
>> the
>>> memory allocations inside GetSmBase() failed.
>>
>> Sorry, these two paragraphs are not precise. A better version:
>>
>> ----------
>> Commit 725acd0b9cc0 ("UefiCpuPkg: Avoid assuming only one smmbasehob",
>> 2023-12-12) introduced a helper function called GetSmBase(), replacing
>> the lookup of the first and only "gSmmBaseHobGuid" GUID HOB and
>> unconditional "mCpuHotPlugData.SmBase" allocation, with iterated
>> lookups
>> plus conditional memory allocation.
>>
>> This introduced a new failure mode for setting
>> "mCpuHotPlugData.SmBase".
>> Namely, before commit 725acd0b9cc0, "mCpuHotPlugData.SmBase" would be
>> allocated regardless of the GUID HOB being absent. After the commit,
>> "mCpuHotPlugData.SmBase" could remain NULL if the GUID HOB was absent,
>> *or* one of the memory allocations inside GetSmBase() failed; and in
>> the
>> former case, we'd even proceed to the rest of PiCpuSmmEntry().
>> ----------
>>
>> Sorry, it's late.
>>
>> If this patch set is accepted otherwise, then Mike or Liming, can you
>> please update the first two paragraphs of the commit message upon
>> merge?
>>
>> Thanks
>> Laszlo
>>
>>>
>>> In relation to this conflation of distinct failure modes, commit
>>> 725acd0b9cc0 actually introduced a NULL pointer dereference. Namely,
>> a
>>> NULL "mCpuHotPlugData.SmBase" is not handled properly at all now.
>> We're
>>> going to fix that NULL pointer dereference in a subsequent patch;
>> however,
>>> as a pre-requisite for that we need to tell apart the failure modes
>> of
>>> GetSmBase().
>>>
>>> For memory allocation failures, return EFI_OUT_OF_RESOURCES. Move the
>>> "assertion" that SMRAM cannot be exhausted happen out to the caller
>>> (PiCpuSmmEntry()). Strengthen the assertion by adding an explicit
>>> CpuDeadLoop() call. (Note: GetSmBase() *already* calls CpuDeadLoop()
>> if
>>> (NumberOfProcessors != MaxNumberOfCpus).)
>>>
>>> For the absence of the GUID HOB, return EFI_NOT_FOUND.
>>>
>>> For good measure, make GetSmBase() STATIC (it should have been STATIC
>> from
>>> the start).
>>>
>>> This is just a refactoring, no behavioral difference is intended
>> (beyond
>>> the explicit CpuDeadLoop() upon SMRAM exhaustion).
>>>
>>> Cc: Dun Tan <dun.tan@intel.com>
>>> Cc: Gerd Hoffmann <kraxel@redhat.com>
>>> Cc: Rahul Kumar <rahul1.kumar@intel.com>
>>> Cc: Ray Ni <ray.ni@intel.com>
>>> Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=4682
>>> Signed-off-by: Laszlo Ersek <lersek@redhat.com>
>>> ---
>>>
>>> Notes:
>>>     context:-U4
>>>
>>>  UefiCpuPkg/PiSmmCpuDxeSmm/PiSmmCpuDxeSmm.c | 40 ++++++++++++++------
>>>  1 file changed, 28 insertions(+), 12 deletions(-)
>>>
>>> diff --git a/UefiCpuPkg/PiSmmCpuDxeSmm/PiSmmCpuDxeSmm.c
>> b/UefiCpuPkg/PiSmmCpuDxeSmm/PiSmmCpuDxeSmm.c
>>> index cd394826ffcf..09382945ddb4 100644
>>> --- a/UefiCpuPkg/PiSmmCpuDxeSmm/PiSmmCpuDxeSmm.c
>>> +++ b/UefiCpuPkg/PiSmmCpuDxeSmm/PiSmmCpuDxeSmm.c
>>> @@ -619,16 +619,23 @@ SmBaseHobCompare (
>>>
>>>  /**
>>>    Extract SmBase for all CPU from SmmBase HOB.
>>>
>>> -  @param[in]  MaxNumberOfCpus   Max NumberOfCpus.
>>> +  @param[in]  MaxNumberOfCpus        Max NumberOfCpus.
>>>
>>> -  @retval SmBaseBuffer          Pointer to SmBase Buffer.
>>> -  @retval NULL                  gSmmBaseHobGuid was not been
>> created.
>>> +  @param[out] AllocatedSmBaseBuffer  Pointer to SmBase Buffer
>> allocated
>>> +                                     by this function. Only set if
>> the
>>> +                                     function returns EFI_SUCCESS.
>>> +
>>> +  @retval EFI_SUCCESS           SmBase Buffer output successfully.
>>> +  @retval EFI_OUT_OF_RESOURCES  Memory allocation failed.
>>> +  @retval EFI_NOT_FOUND         gSmmBaseHobGuid was never created.
>>>  **/
>>> -UINTN *
>>> +STATIC
>>> +EFI_STATUS
>>>  GetSmBase (
>>> -  IN  UINTN  MaxNumberOfCpus
>>> +  IN  UINTN  MaxNumberOfCpus,
>>> +  OUT UINTN  **AllocatedSmBaseBuffer
>>>    )
>>>  {
>>>    UINTN              HobCount;
>>>    EFI_HOB_GUID_TYPE  *GuidHob;
>>> @@ -649,9 +656,9 @@ GetSmBase (
>>>    NumberOfProcessors = 0;
>>>
>>>    FirstSmmBaseGuidHob = GetFirstGuidHob (&gSmmBaseHobGuid);
>>>    if (FirstSmmBaseGuidHob == NULL) {
>>> -    return NULL;
>>> +    return EFI_NOT_FOUND;
>>>    }
>>>
>>>    GuidHob = FirstSmmBaseGuidHob;
>>>    while (GuidHob != NULL) {
>>> @@ -671,11 +678,10 @@ GetSmBase (
>>>      CpuDeadLoop ();
>>>    }
>>>
>>>    SmBaseHobs = AllocatePool (sizeof (SMM_BASE_HOB_DATA *) *
>> HobCount);
>>> -  ASSERT (SmBaseHobs != NULL);
>>>    if (SmBaseHobs == NULL) {
>>> -    return NULL;
>>> +    return EFI_OUT_OF_RESOURCES;
>>>    }
>>>
>>>    //
>>>    // Record each SmmBaseHob pointer in the SmBaseHobs.
>>> @@ -691,9 +697,9 @@ GetSmBase (
>>>    SmBaseBuffer = (UINTN *)AllocatePool (sizeof (UINTN) *
>> (MaxNumberOfCpus));
>>>    ASSERT (SmBaseBuffer != NULL);
>>>    if (SmBaseBuffer == NULL) {
>>>      FreePool (SmBaseHobs);
>>> -    return NULL;
>>> +    return EFI_OUT_OF_RESOURCES;
>>>    }
>>>
>>>    QuickSort (SmBaseHobs, HobCount, sizeof (SMM_BASE_HOB_DATA *),
>> (BASE_SORT_COMPARE)SmBaseHobCompare, &SortBuffer);
>>>    PrevProcessorIndex = 0;
>>> @@ -713,9 +719,10 @@ GetSmBase (
>>>      PrevProcessorIndex += SmBaseHobs[HobIndex]->NumberOfProcessors;
>>>    }
>>>
>>>    FreePool (SmBaseHobs);
>>> -  return SmBaseBuffer;
>>> +  *AllocatedSmBaseBuffer = SmBaseBuffer;
>>> +  return EFI_SUCCESS;
>>>  }
>>>
>>>  /**
>>>    Function to compare 2 MP_INFORMATION2_HOB_DATA pointer based on
>> ProcessorIndex.
>>> @@ -1110,10 +1117,17 @@ PiCpuSmmEntry (
>>>    //
>>>    // Retrive the allocated SmmBase from gSmmBaseHobGuid. If found,
>>>    // means the SmBase relocation has been done.
>>>    //
>>> -  mCpuHotPlugData.SmBase = GetSmBase (mMaxNumberOfCpus);
>>> -  if (mCpuHotPlugData.SmBase != NULL) {
>>> +  mCpuHotPlugData.SmBase = NULL;
>>> +  Status                 = GetSmBase (mMaxNumberOfCpus,
>> &mCpuHotPlugData.SmBase);
>>> +  if (Status == EFI_OUT_OF_RESOURCES) {
>>> +    ASSERT (Status != EFI_OUT_OF_RESOURCES);
>>> +    CpuDeadLoop ();
>>> +  }
>>> +
>>> +  if (!EFI_ERROR (Status)) {
>>> +    ASSERT (mCpuHotPlugData.SmBase != NULL);
>>>      //
>>>      // Check whether the Required TileSize is enough.
>>>      //
>>>      if (TileSize > SIZE_8KB) {
>>> @@ -1125,8 +1139,10 @@ PiCpuSmmEntry (
>>>      }
>>>
>>>      mSmmRelocated = TRUE;
>>>    } else {
>>> +    ASSERT (Status == EFI_NOT_FOUND);
>>> +    ASSERT (mCpuHotPlugData.SmBase == NULL);
>>>      //
>>>      // When the HOB doesn't exist, allocate new SMBASE itself.
>>>      //
>>>      DEBUG ((DEBUG_INFO, "PiCpuSmmEntry: gSmmBaseHobGuid not
>> found!\n"));
>>>
>>>
>>>
>>>
>>>
>>>
>>
>>
>>
>>
>>
> 
> 
> 
> 
> 
> 



-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#115444): https://edk2.groups.io/g/devel/message/115444
Mute This Topic: https://groups.io/mt/104341342/7686176
Group Owner: devel+owner@edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [rebecca@openfw.io]
-=-=-=-=-=-=-=-=-=-=-=-