From: "Kun Qin" <kuqin12@gmail.com>
To: Pierre Gondois <pierre.gondois@arm.com>,
Sami Mujawar <sami.mujawar@arm.com>,
devel@edk2.groups.io
Cc: Joe Lopez <joelopez@microsoft.com>, "nd@arm.com" <nd@arm.com>
Subject: Re: [edk2-devel] [PATCH v3 5/6] DynamicTablesPkg: AcpiSsdtPcieLibArm: Added function to reserve ECAM space
Date: Wed, 10 Aug 2022 15:37:01 -0700 [thread overview]
Message-ID: <68fdf6f0-3059-ea56-9e96-758cccb46c01@gmail.com> (raw)
In-Reply-To: <8e9707a8-b200-5727-46bb-58f99c74f501@arm.com>
Hi Pierre/Sami,
Thanks for your feedback. We modified the routine to be based on
`CM_ARM_PCI_CONFIG_SPACE_INFO` and sanity checked the table
outputs from UEFI shell and verified the system bootabilty on VDK
based ARM platform.
The new patch can be found here:
https://edk2.groups.io/g/devel/message/92317
Looking forward to your further feedback.
Thanks,
Kun
On 8/10/2022 1:51 AM, Pierre Gondois wrote:
>
>
> On 8/8/22 17:39, Sami Mujawar wrote:
>> Hi Kun,
>>
>> I have just tried testing your patch with Kvmtool guest firmware and
>> think this patch may need some modifications.
>>
>> Also, the patch 4/6 may need some adjustment, which I will reply back
>> on that patch separately.
>>
>> Regards,
>>
>> Sami Mujawar
>>
>> On 08/08/2022 02:22 pm, Sami Mujawar wrote:
>>> Hi Kun,
>>>
>>> Thank you for this patch.
>>>
>>> These changes look good to me.
>>>
>>> Reviewed-by: Sami Mujawar <sami.mujawar@arm.com>
>>>
>>> Regards,
>>>
>>> Sami Mujawar
>>>
>>> On 31/07/2022 06:37 am, Kun Qin via groups.io wrote:
>>>> REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3998
>>>>
>>>> Certain OSes will complain if the ECAM config space is not reserved in
>>>> the ACPI namespace.
>>>>
>>>> This change adds a function to reserve PNP motherboard resources for a
>>>> given PCI node.
>>>>
>>>> Co-authored-by: Joe Lopez <joelopez@microsoft.com>
>>>> Signed-off-by: Kun Qin <kuqin12@gmail.com>
>>>> Reviewed-by: Pierre Gondois <pierre.gondois@arm.com>
>>>> ---
>>>>
>>>> Notes:
>>>> v2:
>>>> - Only create RES0 after config space checking [Pierre]
>>>> v3:
>>>> - Updated function names and descriptions [Pierre]
>>>> - Moved translation calculation to CONFIG case [Pierre]
>>>>
>>>> DynamicTablesPkg/Library/Acpi/Arm/AcpiSsdtPcieLibArm/SsdtPcieGenerator.c
>>>> | 171 ++++++++++++++++++++
>>>> 1 file changed, 171 insertions(+)
>>>>
>>>> diff --git
>>>> a/DynamicTablesPkg/Library/Acpi/Arm/AcpiSsdtPcieLibArm/SsdtPcieGenerator.c
>>>> b/DynamicTablesPkg/Library/Acpi/Arm/AcpiSsdtPcieLibArm/SsdtPcieGenerator.c
>>>>
>>>> index ceffe2838c03..658a089c8f1f 100644
>>>> ---
>>>> a/DynamicTablesPkg/Library/Acpi/Arm/AcpiSsdtPcieLibArm/SsdtPcieGenerator.c
>>>> +++
>>>> b/DynamicTablesPkg/Library/Acpi/Arm/AcpiSsdtPcieLibArm/SsdtPcieGenerator.c
>>>> @@ -616,6 +616,169 @@ GeneratePciCrs (
>>>> return Status;
>>>>
>>>> }
>>>>
>>>>
>>>> +/** Generate a RES0 device node to reserve PNP motherboard resources
>>>>
>>>> + for a given PCI node.
>>>>
>>>> +
>>>>
>>>> + @param [in] PciNode Parent PCI node handle of the generated
>>>>
>>>> + resource object.
>>>>
>>>> + @param [out] CrsNode CRS node of the AML tree to populate.
>>>>
>>>> +
>>>>
>>>> + @retval EFI_SUCCESS The function completed
>>>> successfully.
>>>>
>>>> + @retval EFI_INVALID_PARAMETER Invalid input parameter.
>>>>
>>>> + @retval EFI_OUT_OF_RESOURCES Could not allocate memory.
>>>>
>>>> +**/
>>>>
>>>> +STATIC
>>>>
>>>> +EFI_STATUS
>>>>
>>>> +EFIAPI
>>>>
>>>> +GenerateMotherboardDevice (
>>>>
>>>> + IN AML_OBJECT_NODE_HANDLE PciNode,
>>>>
>>>> + OUT AML_OBJECT_NODE_HANDLE *CrsNode
>>>>
>>>> + )
>>>>
>>>> +{
>>>>
>>>> + EFI_STATUS Status;
>>>>
>>>> + UINT32 EisaId;
>>>>
>>>> + AML_OBJECT_NODE_HANDLE ResNode;
>>>>
>>>> +
>>>>
>>>> + if (CrsNode == NULL) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return EFI_INVALID_PARAMETER;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + // ASL: Device (RES0) {}
>>>>
>>>> + Status = AmlCodeGenDevice ("RES0", PciNode, &ResNode);
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + // ASL: Name (_HID, EISAID ("PNP0C02"))
>>>>
>>>> + Status = AmlGetEisaIdFromString ("PNP0C02", &EisaId); /* PNP
>>>> Motherboard Resources */
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + Status = AmlCodeGenNameInteger ("_HID", EisaId, ResNode, NULL);
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + // ASL: Name (_CRS, ResourceTemplate () {})
>>>>
>>>> + Status = AmlCodeGenNameResourceTemplate ("_CRS", ResNode, CrsNode);
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + return Status;
>>>>
>>>> +}
>>>>
>>>> +
>>>>
>>>> +/** Reserves ECAM space for PCI config space
>>>>
>>>> +
>>>>
>>>> + @param [in] Generator The SSDT Pci generator.
>>>>
>>>> + @param [in] CfgMgrProtocol Pointer to the Configuration
>>>> Manager
>>>>
>>>> + Protocol interface.
>>>>
>>>> + @param [in] PciInfo Pci device information.
>>>>
>>>> + @param [in, out] PciNode RootNode of the AML tree to
>>>> populate.
>>>>
>>>> +
>>>>
>>>> + @retval EFI_SUCCESS The function completed
>>>> successfully.
>>>>
>>>> + @retval EFI_INVALID_PARAMETER Invalid parameter.
>>>>
>>>> + @retval EFI_OUT_OF_RESOURCES Could not allocate memory.
>>>>
>>>> +**/
>>>>
>>>> +STATIC
>>>>
>>>> +EFI_STATUS
>>>>
>>>> +EFIAPI
>>>>
>>>> +ReserveEcamSpace (
>>>>
>>>> + IN ACPI_PCI_GENERATOR *Generator,
>>>>
>>>> + IN CONST EDKII_CONFIGURATION_MANAGER_PROTOCOL *CONST
>>>> CfgMgrProtocol,
>>>>
>>>> + IN CONST CM_ARM_PCI_CONFIG_SPACE_INFO *PciInfo,
>>>>
>>>> + IN OUT AML_OBJECT_NODE_HANDLE PciNode
>>>>
>>>> + )
>>>>
>>>> +{
>>>>
>>>> + EFI_STATUS Status;
>>>>
>>>> + AML_OBJECT_NODE_HANDLE CrsNode;
>>>>
>>>> + BOOLEAN Translation;
>>>>
>>>> + UINT32 Index;
>>>>
>>>> + CM_ARM_OBJ_REF *RefInfo;
>>>>
>>>> + UINT32 RefCount;
>>>>
>>>> + CM_ARM_PCI_ADDRESS_MAP_INFO *AddrMapInfo;
>>>>
>>>> + BOOLEAN IsPosDecode;
>>>>
>>>> +
>>>>
>>>> + // Get the array of CM_ARM_OBJ_REF referencing the
>>>>
>>>> + // CM_ARM_PCI_ADDRESS_MAP_INFO objects.
>>>>
>>>> + Status = GetEArmObjCmRef (
>>>>
>>>> + CfgMgrProtocol,
>>>>
>>>> + PciInfo->AddressMapToken,
>>>>
>>>> + &RefInfo,
>>>>
>>>> + &RefCount
>>>>
>>>> + );
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + for (Index = 0; Index < RefCount; Index++) {
>>>>
>>>> + // Get CM_ARM_PCI_ADDRESS_MAP_INFO structures one by one.
>>>>
>>>> + Status = GetEArmObjPciAddressMapInfo (
>>>>
>>>> + CfgMgrProtocol,
>>>>
>>>> + RefInfo[Index].ReferenceToken,
>>>>
>>>> + &AddrMapInfo,
>>>>
>>>> + NULL
>>>>
>>>> + );
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>> [SAMI] Sorry for missing this earlier in the review. However, the
>> ECAM memory space is described byCM_ARM_PCI_CONFIG_SPACE_INFO. So, I
>> think that needs to be used here.
>>
>> The CM_ARM_PCI_CONFIG_SPACE_INFO structure does not include the
>> length of the configuration space and would probably need to be updated.
>>
>> Which platform are you testing these changes on? I would like to
>> understand more about your use case. Is it possible to share some
>> more details, please?
>>
>> [/SAMI]
>
> [Pierre]
>
> Yes indeed, CM_ARM_PCI_CONFIG_SPACE_INFO should contain the configuration
> address space, it should not be described in CM_ARM_PCI_ADDRESS_MAP_INFO.
> So there should be no need to search through the
> CM_ARM_PCI_ADDRESS_MAP_INFO
> objects. Sorry for missing this earlier.
>
> The length of the address space could be computed as:
> length = (end_bus - start_bus + 1) × 32 devices × 8 functions × 4 KB
>
> [/Pierre]
>
>>
>>>> +
>>>>
>>>> + switch (AddrMapInfo->SpaceCode) {
>>>>
>>>> + case PCI_SS_CONFIG:
>>>>
>>>> + Translation = (AddrMapInfo->CpuAddress !=
>>>> AddrMapInfo->PciAddress);
>>>>
>>>> + if (AddrMapInfo->CpuAddress >= AddrMapInfo->PciAddress) {
>>>>
>>>> + IsPosDecode = TRUE;
>>>>
>>>> + } else {
>>>>
>>>> + IsPosDecode = FALSE;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + Status = GenerateMotherboardDevice (PciNode, &CrsNode);
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + break;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + Status = AmlCodeGenRdQWordMemory (
>>>>
>>>> + FALSE,
>>>>
>>>> + IsPosDecode,
>>>>
>>>> + TRUE,
>>>>
>>>> + TRUE,
>>>>
>>>> + FALSE, // non-cacheable
>>>>
>>>> + TRUE,
>>>>
>>>> + 0,
>>>>
>>>> + AddrMapInfo->PciAddress,
>>>>
>>>> + AddrMapInfo->PciAddress +
>>>> AddrMapInfo->AddressSize - 1,
>>>>
>>>> + Translation ? AddrMapInfo->CpuAddress -
>>>> AddrMapInfo->PciAddress : 0,
>>>>
>>>> + AddrMapInfo->AddressSize,
>>>>
>>>> + 0,
>>>>
>>>> + NULL,
>>>>
>>>> + 0,
>>>>
>>>> + TRUE,
>>>>
>>>> + CrsNode,
>>>>
>>>> + NULL
>>>>
>>>> + );
>>>>
>>>> + break;
>>>>
>>>> + default:
>>>>
>>>> + break;
>>>>
>>>> + } // switch
>>>>
>>>> +
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> + return Status;
>>>>
>>>> +}
>>>>
>>>> +
>>>>
>>>> /** Generate a Pci device.
>>>>
>>>>
>>>> @param [in] Generator The SSDT Pci generator.
>>>>
>>>> @@ -702,9 +865,17 @@ GeneratePciDevice (
>>>> return Status;
>>>>
>>>> }
>>>>
>>>>
>>>> + // Add the PNP Motherboard Resources Device to reserve ECAM space
>>>>
>>>> + Status = ReserveEcamSpace (Generator, CfgMgrProtocol, PciInfo,
>>>> PciNode);
>>>>
>>>> + if (EFI_ERROR (Status)) {
>>>>
>>>> + ASSERT (0);
>>>>
>>>> + return Status;
>>>>
>>>> + }
>>>>
>>>> +
>>>>
>>>> // Add the template _OSC method.
>>>>
>>>> Status = AddOscMethod (PciInfo, PciNode);
>>>>
>>>> ASSERT_EFI_ERROR (Status);
>>>>
>>>> +
>>>>
>>>> return Status;
>>>>
>>>> }
>>>>
>>>>
next prev parent reply other threads:[~2022-08-10 22:37 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-07-31 5:37 [PATCH v3 0/6] Enhance DynamicTablesPkg modules Kun Qin
2022-07-31 5:37 ` [PATCH v3 1/6] DynamicTablesPkg: DynamicPlatRepoLib: Added MemoryAllocationLib to inf Kun Qin
2022-07-31 5:37 ` [PATCH v3 2/6] DynamicTablesPkg: DynamicPlatRepoLib: Fix incorrect dereferencing Kun Qin
2022-07-31 5:37 ` [PATCH v3 3/6] DynamicTablesPkg: DynamicPlatRepoLib: Adding more token fixers Kun Qin
2022-07-31 5:37 ` [PATCH v3 4/6] DynamicTablesPkg: DynamicTableManagerDxe: Added check for installed tables Kun Qin
2022-08-08 13:05 ` Sami Mujawar
2022-08-08 15:39 ` Sami Mujawar
2022-08-08 23:43 ` Kun Qin
2022-08-09 9:01 ` Sami Mujawar
2022-08-10 22:31 ` Kun Qin
2022-07-31 5:37 ` [PATCH v3 5/6] DynamicTablesPkg: AcpiSsdtPcieLibArm: Added function to reserve ECAM space Kun Qin
2022-08-08 13:22 ` [edk2-devel] " Sami Mujawar
2022-08-08 15:39 ` Sami Mujawar
2022-08-10 8:51 ` PierreGondois
2022-08-10 22:37 ` Kun Qin [this message]
2022-07-31 5:37 ` [PATCH v3 6/6] DynamicTablesPkg: AcpiSsdtPcieLibArm: Added case handling for PCI config Kun Qin
2022-08-08 13:22 ` Sami Mujawar
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=68fdf6f0-3059-ea56-9e96-758cccb46c01@gmail.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