From: "Abner Chang" <abner.chang@hpe.com>
To: Jonathan Cameron <jonathan.cameron@huawei.com>
Cc: "devel@edk2.groups.io" <devel@edk2.groups.io>
Subject: Re: [edk2-devel] [edk2-staging/RISC-V PATCH v1 1/14]: BaseTools: Update EDK2 build tool for RISC-V platform
Date: Wed, 28 Aug 2019 09:08:09 +0000 [thread overview]
Message-ID: <CS1PR8401MB119284595F65C793D9F246AAFFA30@CS1PR8401MB1192.NAMPRD84.PROD.OUTLOOK.COM> (raw)
In-Reply-To: <20190828095955.00002835@huawei.com>
> -----Original Message-----
> From: Jonathan Cameron [mailto:jonathan.cameron@huawei.com]
> Sent: Wednesday, August 28, 2019 5:00 PM
> To: Chang, Abner (HPS SW/FW Technologist) <abner.chang@hpe.com>
> Cc: devel@edk2.groups.io
> Subject: Re: [edk2-devel] [edk2-staging/RISC-V PATCH v1 1/14]: BaseTools:
> Update EDK2 build tool for RISC-V platform
>
> On Wed, 28 Aug 2019 08:43:10 +0000
> "Chang, Abner (HPS SW/FW Technologist)" <abner.chang@hpe.com> wrote:
>
> > Thanks Jonathan, the cover letter and 3/14 patch were just sent.
>
> Great. So am I right in thinking these apply to:
> INVALID URI REMOVED
> 3A__github.com_tianocore_edk2-2Dstaging_tree_RISC-
> 2DV&d=DwICAg&c=C5b8zRQO1miGmBeVZ2LFWg&r=_SN6FZBN4Vgi4Ulkskz6
> qU3NYRO03nHp9P7Z5q59A3E&m=4xGwArhbyMMHxteOpv4UP8PJWPKah5H
> fR_X2Dl0H8vE&s=sCxw_Qm3DjjrX9vULVZYYw6jqhmQWygJP-B0Dltr6kc&e= ?
>
> Any plans to rebase that tree as it seems to be a few years old now?
Yes. we have plan to rebase it as you said this one is too old, will do it while edk2 devel review those patches ( includes edk2-platform/devel-rsicv patches)
Thanks
>
> Thanks,
>
> Jonathan
>
> >
> > Abner
> >
> > > -----Original Message-----
> > > From: Jonathan Cameron [mailto:jonathan.cameron@huawei.com]
> > > Sent: Wednesday, August 28, 2019 4:18 PM
> > > To: Chang, Abner (HPS SW/FW Technologist) <abner.chang@hpe.com>
> > > Cc: devel@edk2.groups.io
> > > Subject: Re: [edk2-devel] [edk2-staging/RISC-V PATCH v1 1/14]:
> BaseTools:
> > > Update EDK2 build tool for RISC-V platform
> > >
> > > Hi Abner,
> > >
> > > Just noticed in passing that this series doesn't seem to have a 3/14?
> > >
> > > Speaking personally it would be useful to have a cover letter with a
> > > quick summary of the series content.
> > >
> > > Thanks,
> > >
> > > Jonathan
> > >
> > > On Tue, 27 Aug 2019 14:00:19 +0800
> > > Abner Chang <abner.chang@hpe.com> wrote:
> > >
> > > > Elf64Convert.c
> > > > - Relocation process to hadnle below opcodes,
> > > > * PCRELHI20
> > > > * PCRELLO12
> > > > * ADD32
> > > > * SUB32
> > > >
> > > > GenFvInternalLib.c
> > > > - This atches jump instrcution at the position of first
> > > > instrcution fetched by
> > > RISC-V processor after Zeroth Stage Boot Loader (ZSBL).
> > > >
> > > > Contributed-under: TianoCore Contribution Agreement 1.0
> > > > Signed-off-by: Abner Chang <abner.chang@hpe.com>
> > > > ---
> > > > BaseTools/Source/C/GenFv/GenFvInternalLib.c | 311
> > > > ++++++++++++------
> > > ----------
> > > > BaseTools/Source/C/GenFw/Elf64Convert.c | 68 ++++++
> > > > 2 files changed, 197 insertions(+), 182 deletions(-)
> > > >
> > > > diff --git a/BaseTools/Source/C/GenFv/GenFvInternalLib.c
> > > > b/BaseTools/Source/C/GenFv/GenFvInternalLib.c
> > > > index 01da00c..92abb7c 100644
> > > > --- a/BaseTools/Source/C/GenFv/GenFvInternalLib.c
> > > > +++ b/BaseTools/Source/C/GenFv/GenFvInternalLib.c
> > > > @@ -1956,157 +1956,6 @@ Returns:
> > > > return EFI_UNSUPPORTED;
> > > > }
> > > >
> > > > -EFI_STATUS
> > > > -UpdateRiscvResetVectorIfNeeded (
> > > > - MEMORY_FILE *FvImage,
> > > > - FV_INFO *FvInfo,
> > > > - EFI_FFS_FILE_HEADER *VtfFileImage
> > > > - )
> > > > -/*++
> > > > -
> > > > -Routine Description:
> > > > - This parses the FV looking for SEC and patches that address
> > > > into the
> > > > - beginning of the FV header.
> > > > -
> > > > - For RISC-V ISA, the reset vector is at 0xfff~ff00h or 200h
> > > > -
> > > > -Arguments:
> > > > - FvImage Memory file for the FV memory image/
> > > > - FvInfo Information read from INF file.
> > > > - VtfFileImage Instance of VTF file.
> > > > -
> > > > -Returns:
> > > > -
> > > > - EFI_SUCCESS Function Completed successfully.
> > > > - EFI_ABORTED Error encountered.
> > > > - EFI_INVALID_PARAMETER A required parameter was NULL.
> > > > - EFI_NOT_FOUND PEI Core file not found.
> > > > -
> > > > ---*/
> > > > -{
> > > > - EFI_FFS_FILE_HEADER *PeiCoreFile;
> > > > - EFI_FFS_FILE_HEADER *SecCoreFile;
> > > > - EFI_STATUS Status;
> > > > - EFI_FILE_SECTION_POINTER Pe32Section;
> > > > - UINT32 EntryPoint;
> > > > - UINT32 BaseOfCode;
> > > > - UINT16 MachineType;
> > > > - EFI_PHYSICAL_ADDRESS PeiCorePhysicalAddress;
> > > > - EFI_PHYSICAL_ADDRESS SecCorePhysicalAddress;
> > > > - EFI_PHYSICAL_ADDRESS TrapAddress;
> > > > -
> > > > - //
> > > > - // Verify input parameters
> > > > - //
> > > > - if (FvImage == NULL || FvInfo == NULL) {
> > > > - Error (NULL, 0, 3000, "Invalid", "FvImage or FvInfo is NULL");
> > > > - return EFI_INVALID_PARAMETER;
> > > > - }
> > > > - //
> > > > - // Initialize FV library
> > > > - //
> > > > - InitializeFvLib (FvImage->FileImage, FvInfo->Size);
> > > > -
> > > > - //
> > > > - // Find the Sec Core
> > > > - //
> > > > - Status = GetFileByType (EFI_FV_FILETYPE_SECURITY_CORE, 1,
> > > > &SecCoreFile);
> > > > - if (EFI_ERROR (Status) || SecCoreFile == NULL) {
> > > > - //
> > > > - // Maybe hardware does SEC job and we only have PEI Core?
> > > > - //
> > > > -
> > > > - //
> > > > - // Find the PEI Core. It may not exist if SEC loads DXE core directly
> > > > - //
> > > > - PeiCorePhysicalAddress = 0;
> > > > - Status = GetFileByType (EFI_FV_FILETYPE_PEI_CORE, 1,
> &PeiCoreFile);
> > > > - if (!EFI_ERROR(Status) && PeiCoreFile != NULL) {
> > > > - //
> > > > - // PEI Core found, now find PE32 or TE section
> > > > - //
> > > > - Status = GetSectionByType (PeiCoreFile, EFI_SECTION_PE32, 1,
> > > &Pe32Section);
> > > > - if (Status == EFI_NOT_FOUND) {
> > > > - Status = GetSectionByType (PeiCoreFile, EFI_SECTION_TE, 1,
> > > &Pe32Section);
> > > > - }
> > > > -
> > > > - if (EFI_ERROR (Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "could not find either a PE32 or a TE
> > > section in PEI core file!");
> > > > - return EFI_ABORTED;
> > > > - }
> > > > -
> > > > - Status = GetPe32Info (
> > > > - (VOID *) ((UINTN) Pe32Section.Pe32Section +
> > > GetSectionHeaderLength(Pe32Section.CommonHeader)),
> > > > - &EntryPoint,
> > > > - &BaseOfCode,
> > > > - &MachineType
> > > > - );
> > > > -
> > > > - if (EFI_ERROR (Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "could not get the PE32 entry point
> for
> > > the PEI core!");
> > > > - return EFI_ABORTED;
> > > > - }
> > > > - //
> > > > - // Physical address is FV base + offset of PE32 + offset of the entry
> > > point
> > > > - //
> > > > - PeiCorePhysicalAddress = FvInfo->BaseAddress;
> > > > - PeiCorePhysicalAddress += (UINTN) Pe32Section.Pe32Section +
> > > GetSectionHeaderLength(Pe32Section.CommonHeader) - (UINTN)
> FvImage-
> > > >FileImage;
> > > > - PeiCorePhysicalAddress += EntryPoint;
> > > > - DebugMsg (NULL, 0, 9, "PeiCore physical entry point address",
> > > "Address = 0x%llX", (unsigned long long) PeiCorePhysicalAddress);
> > > > - RiscvPatchVtf (VtfFileImage, (UINT32)PeiCorePhysicalAddress);
> > > > - }
> > > > - return EFI_SUCCESS;
> > > > - }
> > > > -
> > > > - //
> > > > - // Sec Core found, now find PE32 section
> > > > - //
> > > > - Status = GetSectionByType (SecCoreFile, EFI_SECTION_PE32, 1,
> > > > &Pe32Section);
> > > > - if (Status == EFI_NOT_FOUND) {
> > > > - Status = GetSectionByType (SecCoreFile, EFI_SECTION_TE, 1,
> > > &Pe32Section);
> > > > - }
> > > > -
> > > > - if (EFI_ERROR (Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "could not find a PE32 section in the
> SEC
> > > core file.");
> > > > - return EFI_ABORTED;
> > > > - }
> > > > -
> > > > - Status = GetPe32Info (
> > > > - (VOID *) ((UINTN) Pe32Section.Pe32Section +
> > > GetSectionHeaderLength(Pe32Section.CommonHeader)),
> > > > - &EntryPoint,
> > > > - &BaseOfCode,
> > > > - &MachineType
> > > > - );
> > > > - if (EFI_ERROR (Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "could not get the PE32 entry point for
> > > the SEC core.");
> > > > - return EFI_ABORTED;
> > > > - }
> > > > -
> > > > - if ((MachineType != EFI_IMAGE_MACHINE_RISCV32) &&
> > > (MachineType != EFI_IMAGE_MACHINE_RISCV64)) {
> > > > - //
> > > > - // If SEC is not RISC-V we have nothing to do
> > > > - //
> > > > - return EFI_SUCCESS;
> > > > - }
> > > > -
> > > > - //
> > > > - // Physical address is FV base + offset of PE32 + offset of the
> > > > entry point
> > > > - //
> > > > - SecCorePhysicalAddress = FvInfo->BaseAddress;
> > > > - SecCorePhysicalAddress += (UINTN) Pe32Section.Pe32Section +
> > > > GetSectionHeaderLength(Pe32Section.CommonHeader) - (UINTN)
> > > > FvImage->FileImage;
> > > > - SecCorePhysicalAddress += EntryPoint;
> > > > - DebugMsg (NULL, 0, 0x14, "SecCore physical entry point
> > > > address", "Address = 0x%llX", (unsigned long long)
> > > > SecCorePhysicalAddress);
> > > > - RiscvPatchVtf (VtfFileImage, (UINT32)SecCorePhysicalAddress);
> > > > - //
> > > > - // Update RISC-V trap handler.
> > > > - //
> > > > - TrapAddress = (UINTN) Pe32Section.Pe32Section +
> > > > GetSectionHeaderLength(Pe32Section.CommonHeader) + EntryPoint;
> > > > - TrapAddress -= 40;
> > > > - RiscvPatchVtfTrapHandler (VtfFileImage, TrapAddress);
> > > > -
> > > > - DebugMsg (NULL, 0, 9, "Update Reset vector in FV Header",
> > > > NULL);
> > > > - return EFI_SUCCESS;
> > > > -}
> > > >
> > > > EFI_STATUS
> > > > FindCorePeSection(
> > > > @@ -2581,6 +2430,106 @@ Returns:
> > > > }
> > > >
> > > > EFI_STATUS
> > > > +UpdateRiscvResetVectorIfNeeded (
> > > > + MEMORY_FILE *FvImage,
> > > > + FV_INFO *FvInfo
> > > > + )
> > > > +/*++
> > > > +
> > > > +Routine Description:
> > > > + This parses the FV looking for SEC and patches that address
> > > > +into the
> > > > + beginning of the FV header.
> > > > +
> > > > + For RISC-V ISA, the reset vector is at 0xfff~ff00h or 200h
> > > > +
> > > > +Arguments:
> > > > + FvImage Memory file for the FV memory image/
> > > > + FvInfo Information read from INF file.
> > > > +
> > > > +Returns:
> > > > +
> > > > + EFI_SUCCESS Function Completed successfully.
> > > > + EFI_ABORTED Error encountered.
> > > > + EFI_INVALID_PARAMETER A required parameter was NULL.
> > > > + EFI_NOT_FOUND PEI Core file not found.
> > > > +
> > > > +--*/
> > > > +{
> > > > + EFI_STATUS Status;
> > > > + UINT16 MachineType;
> > > > + EFI_FILE_SECTION_POINTER SecPe32;
> > > > + EFI_PHYSICAL_ADDRESS SecCoreEntryAddress;
> > > > +
> > > > + UINT32 bSecCore;
> > > > + UINT32 tmp;
> > > > +
> > > > +
> > > > + //
> > > > + // Verify input parameters
> > > > + //
> > > > + if (FvImage == NULL || FvInfo == NULL) {
> > > > + return EFI_INVALID_PARAMETER; } // // Initialize FV
> > > > + library // InitializeFvLib (FvImage->FileImage, FvInfo->Size);
> > > > +
> > > > + //
> > > > + // Find the Sec Core
> > > > + //
> > > > + Status = FindCorePeSection(FvImage->FileImage, FvInfo->Size,
> > > > + EFI_FV_FILETYPE_SECURITY_CORE, &SecPe32);
> > > > + if(EFI_ERROR(Status)) {
> > > > + printf("skip because Secutiry Core not found\n");
> > > > + return EFI_SUCCESS;
> > > > + }
> > > > +
> > > > + DebugMsg (NULL, 0, 9, "Update SEC core in FV Header", NULL);
> > > > +
> > > > + Status = GetCoreMachineType(SecPe32, &MachineType);
> > > > + if(EFI_ERROR(Status)) {
> > > > + Error(NULL, 0, 3000, "Invalid", "Could not get the PE32
> > > > + machine type for
> > > SEC core.");
> > > > + return EFI_ABORTED;
> > > > + }
> > > > +
> > > > + if ((MachineType != EFI_IMAGE_MACHINE_RISCV32) &&
> > > (MachineType != EFI_IMAGE_MACHINE_RISCV64)) {
> > > > + Error(NULL, 0, 3000, "Invalid", "Could not update SEC core
> > > > + because
> > > Machine type is not RiscV.");
> > > > + return EFI_ABORTED;
> > > > + }
> > > > +
> > > > + Status = GetCoreEntryPointAddress(FvImage->FileImage, FvInfo,
> > > > + SecPe32, &SecCoreEntryAddress);
> > > > + if(EFI_ERROR(Status)) {
> > > > + Error(NULL, 0, 3000, "Invalid", "Could not get the PE32 entry
> > > > + point
> > > address for SEC Core.");
> > > > + return EFI_ABORTED;
> > > > + }
> > > > +
> > > > + VerboseMsg("SecCore entry point Address = 0x%llX", (unsigned
> > > > + long
> > > > + long) SecCoreEntryAddress); VerboseMsg("BaseAddress = 0x%llX",
> > > > + (unsigned long long) FvInfo->BaseAddress); bSecCore =
> > > > + (SecCoreEntryAddress - FvInfo->BaseAddress); VerboseMsg("offset
> > > > + = 0x%llX", bSecCore);
> > > > +
> > > > + if(bSecCore > 0x0fffff) {
> > > > + Error(NULL, 0, 3000, "Invalid", "SEC Entry point must be
> > > > + within 1MB of
> > > start of the FV");
> > > > + return EFI_ABORTED;
> > > > + }
> > > > +
> > > > + tmp = bSecCore;
> > > > + bSecCore = 0;
> > > > + //J-type
> > > > + bSecCore = (tmp&0x100000)<<11; //imm[20] at bit[31]
> > > > + bSecCore |= (tmp&0x0007FE)<<20; //imm[10:1] at bit[30:21]
> > > > + bSecCore |= (tmp&0x000800)<<9; //imm[11] at bit[20]
> > > > + bSecCore |= (tmp&0x0FF000); //imm[19:12] at bit[19:12]
> > > > + bSecCore |= 0x6F; //JAL opcode
> > > > +
> > > > + memcpy(FvImage->FileImage, &bSecCore, sizeof(bSecCore));
> > > > +
> > > > + return EFI_SUCCESS;
> > > > +}
> > > > +
> > > > +
> > > > +
> > > > +EFI_STATUS
> > > > GetPe32Info (
> > > > IN UINT8 *Pe32,
> > > > OUT UINT32 *EntryPoint,
> > > > @@ -3037,7 +2986,6 @@ Returns:
> > > > FvHeader->Checksum = 0;
> > > > FvHeader->Checksum = CalculateChecksum16 ((UINT16 *)
> FvHeader,
> > > FvHeader->HeaderLength / sizeof (UINT16));
> > > > }
> > > > -
> > > > //
> > > > // Add files to FV
> > > > //
> > > > @@ -3069,39 +3017,22 @@ Returns:
> > > > goto Finish;
> > > > }
> > > >
> > > > - if (mRiscV) {
> > > > + if (!mArm && !mRiscV) {
> > > > //
> > > > - // Update RISCV reset vector.
> > > > + // Update reset vector (SALE_ENTRY for IPF)
> > > > + // Now for IA32 and IA64 platform, the fv which has bsf
> > > > + file must have
> > > the
> > > > + // EndAddress of 0xFFFFFFFF. Thus, only this type fv needs
> > > > + to update
> > > the
> > > > + // reset vector. If the PEI Core is found, the VTF file will probably
> get
> > > > + // corrupted by updating the entry point.
> > > > //
> > > > - DebugMsg (NULL, 0, INFO_LOG_LEVEL, "Update RISCV reset
> vector",
> > > NULL);
> > > > - Status = UpdateRiscvResetVectorIfNeeded (&FvImageMemoryFile,
> > > &mFvDataInfo, VtfFileImage);
> > > > - if (EFI_ERROR (Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "Could not update the reset vector
> for
> > > RISC-V.");
> > > > + if ((mFvDataInfo.BaseAddress + mFvDataInfo.Size) ==
> > > FV_IMAGES_TOP_ADDRESS) {
> > > > + Status = UpdateResetVector (&FvImageMemoryFile,
> > > > + &mFvDataInfo,
> > > VtfFileImage);
> > > > + if (EFI_ERROR(Status)) {
> > > > + Error (NULL, 0, 3000, "Invalid", "Could not update the
> > > > + reset vector.");
> > > > goto Finish;
> > > > - }
> > > > - //
> > > > - // Update Checksum for FvHeader
> > > > - //
> > > > - FvHeader->Checksum = 0;
> > > > - FvHeader->Checksum = CalculateChecksum16 ((UINT16 *)
> FvHeader,
> > > FvHeader->HeaderLength / sizeof (UINT16));
> > > > - } else {
> > > > - if (!mArm) {
> > > > - //
> > > > - // Update reset vector (SALE_ENTRY for IPF)
> > > > - // Now for IA32 and IA64 platform, the fv which has bsf file must
> have
> > > the
> > > > - // EndAddress of 0xFFFFFFFF. Thus, only this type fv needs to
> update
> > > the
> > > > - // reset vector. If the PEI Core is found, the VTF file will probably
> get
> > > > - // corrupted by updating the entry point.
> > > > - //
> > > > - if ((mFvDataInfo.BaseAddress + mFvDataInfo.Size) ==
> > > FV_IMAGES_TOP_ADDRESS) {
> > > > - Status = UpdateResetVector (&FvImageMemoryFile,
> &mFvDataInfo,
> > > VtfFileImage);
> > > > - if (EFI_ERROR(Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "Could not update the reset
> > > vector.");
> > > > - goto Finish;
> > > > - }
> > > > - DebugMsg (NULL, 0, 9, "Update Reset vector in VTF file", NULL);
> > > > - }
> > > > }
> > > > + DebugMsg (NULL, 0, 9, "Update Reset vector in VTF file", NULL);
> > > > + }
> > > > }
> > > > }
> > > >
> > > > @@ -3119,6 +3050,22 @@ Returns:
> > > > FvHeader->Checksum = CalculateChecksum16 ((UINT16 *)
> > > > FvHeader,
> > > FvHeader->HeaderLength / sizeof (UINT16));
> > > > }
> > > >
> > > > + if (mRiscV) {
> > > > + //
> > > > + // Update RISCV reset vector.
> > > > + //
> > > > + Status = UpdateRiscvResetVectorIfNeeded (&FvImageMemoryFile,
> > > &mFvDataInfo);
> > > > + if (EFI_ERROR (Status)) {
> > > > + Error (NULL, 0, 3000, "Invalid", "Could not update the
> > > > + reset vector for
> > > RISC-V.");
> > > > + goto Finish;
> > > > + }
> > > > + //
> > > > + // Update Checksum for FvHeader
> > > > + //
> > > > + FvHeader->Checksum = 0;
> > > > + FvHeader->Checksum = CalculateChecksum16 ((UINT16 *)
> > > > + FvHeader,
> > > > + FvHeader->HeaderLength / sizeof (UINT16)); }
> > > > +
> > > > //
> > > > // Update FV Alignment attribute to the largest alignment of
> > > > all the FFS
> > > files in the FV
> > > > //
> > > > @@ -3853,7 +3800,7 @@ Returns:
> > > > ImageContext.DestinationAddress = NewPe32BaseAddress;
> > > > Status = PeCoffLoaderRelocateImage (&ImageContext);
> > > > if (EFI_ERROR (Status)) {
> > > > - Error (NULL, 0, 3000, "Invalid", "RelocateImage() call failed on
> rebase of
> > > %s", FileName);
> > > > + Error (NULL, 0, 3000, "Invalid", "RelocateImage() call
> > > > + failed on rebase of %s Status=%d", FileName, Status);
> > > > free ((VOID *) MemoryImagePointer);
> > > > return Status;
> > > > }
> > > > diff --git a/BaseTools/Source/C/GenFw/Elf64Convert.c
> > > > b/BaseTools/Source/C/GenFw/Elf64Convert.c
> > > > index 4857485..77b4d53 100644
> > > > --- a/BaseTools/Source/C/GenFw/Elf64Convert.c
> > > > +++ b/BaseTools/Source/C/GenFw/Elf64Convert.c
> > > > @@ -946,8 +946,60 @@ WriteSections64 (
> > > > RiscvSymSecIndex = 0;
> > > > break;
> > > >
> > > > + case R_RISCV_PCREL_HI20:
> > > > + RiscvHi20Targ = Targ;
> > > > + RiscvHi20Sym = SymShdr;
> > > > + RiscvSymSecIndex = Sym->st_shndx;
> > > > +
> > > > + Value = (UINT32)(RV_X(*(UINT32 *)RiscvHi20Targ, 12, 20));
> > > > + printf("PCREL_HI20 Sym:[%s] value:0x%x
> > > > + SymShdr->sh_addr:0x%lx
> > > mCoffSectionOffset:%x \n", GetSymName(Sym), Value, SymShdr-
> >sh_addr,
> > > mCoffSectionsOffset[Sym->st_shndx]);
> > > > + break;
> > > > + case R_RISCV_PCREL_LO12_I:
> > > > + if (RiscvHi20Targ != NULL && RiscvHi20Sym != NULL &&
> > > RiscvSymSecIndex != 0) {
> > > > + int i;
> > > > + Value2 = (UINT32)(RV_X(*(UINT32 *)RiscvHi20Targ, 12, 20));
> > > > + Value = (UINT32)(RV_X(*(UINT32 *)Targ, 20, 12));
> > > > + if(Value & (RISCV_IMM_REACH/2)) {
> > > > + Value |= ~(RISCV_IMM_REACH-1);
> > > > + }
> > > > + printf("PCREL_LO12_I Sym:[%s] value:0x%x SymShdr-
> > > >sh_addr:0x%lx mCoffSectionOffset:%x \n", GetSymName(Sym), Value,
> > > SymShdr->sh_addr, mCoffSectionsOffset[Sym->st_shndx]);
> > > > + Value = Value - RiscvHi20Sym->sh_addr +
> > > mCoffSectionsOffset[RiscvSymSecIndex];
> > > > + if(-2048 > (INT32)Value) {
> > > > + i = (-Value / 4096);
> > > > + //Error (NULL, 0, 3000, "Invalid", "WriteSections64():
> > > PCREL_LO12_I relocation out of range. %d i=%d", Value, i);
> > > > + printf("WriteSections64(): PCREL_LO12_I relocation out of
> range.
> > > Value:%d Value2:%d i=%d\n", Value, Value2, i);
> > > > + Value2 -= i;
> > > > + Value += 4096 * i;
> > > > + if(-2048 > (INT32)Value) {
> > > > + Value2 -= 1;
> > > > + Value += 4096;
> > > > + }
> > > > + }
> > > > + else if( 2047 < (INT32)Value) {
> > > > + i = (Value / 4096);
> > > > + //Error (NULL, 0, 3000, "Invalid", "WriteSections64():
> > > PCREL_LO12_I relocation out of range. %d i=%d", Value, i);
> > > > + printf("WriteSections64(): PCREL_LO12_I relocation out of
> range.
> > > Value:%d Value2:%d i=%d\n", Value, Value2, i);
> > > > + Value2 += i;
> > > > + Value -= 4096 * i;
> > > > + if(2047 < (INT32)Value) {
> > > > + Value2 += 1;
> > > > + Value -= 4096;
> > > > + }
> > > > + }
> > > > +
> > > > + *(UINT32 *)Targ = (RV_X(Value, 0, 12) << 20) |
> > > (RV_X(*(UINT32*)Targ, 0, 20));
> > > > + *(UINT32 *)RiscvHi20Targ = (RV_X(Value2, 0,
> > > > + 20)<<12) |
> > > (RV_X(*(UINT32 *)RiscvHi20Targ, 0, 12));
> > > > + printf("PCREL_LO12_I Sym:[%s] relocated
> > > > + value:0x%x(%d)
> > > value2:0x%x(%d) SymShdr->sh_addr:0x%lx mCoffSectionOffset:%x \n",
> > > GetSymName(Sym), Value, Value, Value2, Value2, SymShdr->sh_addr,
> > > mCoffSectionsOffset[Sym->st_shndx]);
> > > > + }
> > > > + RiscvHi20Sym = NULL;
> > > > + RiscvHi20Targ = NULL;
> > > > + RiscvSymSecIndex = 0;
> > > > + break;
> > > > +
> > > > case R_RISCV_ADD64:
> > > > case R_RISCV_SUB64:
> > > > + case R_RISCV_ADD32:
> > > > + case R_RISCV_SUB32:
> > > > case R_RISCV_BRANCH:
> > > > case R_RISCV_JAL:
> > > > case R_RISCV_GPREL_I:
> > > > @@ -1120,6 +1172,20 @@ WriteRelocations64 (
> > > > EFI_IMAGE_REL_BASED_ABSOLUTE);
> > > > break;
> > > >
> > > > + case R_RISCV_ADD32:
> > > > + CoffAddFixup(
> > > > + (UINT32) ((UINT64) mCoffSectionsOffset[RelShdr->sh_info]
> > > > + + (Rel->r_offset - SecShdr->sh_addr)),
> > > > + EFI_IMAGE_REL_BASED_ABSOLUTE);
> > > > + break;
> > > > +
> > > > + case R_RISCV_SUB32:
> > > > + CoffAddFixup(
> > > > + (UINT32) ((UINT64) mCoffSectionsOffset[RelShdr->sh_info]
> > > > + + (Rel->r_offset - SecShdr->sh_addr)),
> > > > + EFI_IMAGE_REL_BASED_ABSOLUTE);
> > > > + break;
> > > > +
> > > > case R_RISCV_BRANCH:
> > > > CoffAddFixup(
> > > > (UINT32) ((UINT64)
> > > > mCoffSectionsOffset[RelShdr->sh_info]
> > > > @@ -1145,6 +1211,8 @@ WriteRelocations64 (
> > > > case R_RISCV_SET8:
> > > > case R_RISCV_SET16:
> > > > case R_RISCV_SET32:
> > > > + case R_RISCV_PCREL_HI20:
> > > > + case R_RISCV_PCREL_LO12_I:
> > > > break;
> > > >
> > > > default:
> > >
> >
>
next prev parent reply other threads:[~2019-08-28 9:08 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-27 6:00 [edk2-staging/RISC-V PATCH v1 1/14]: BaseTools: Update EDK2 build tool for RISC-V platform Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 2/14]: BaseTools/Conf: Update build flags for RISC-V RV64 Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 4/14]: MdePkg/Include: Update SmBios header file Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 5/14]: RiscVPkg/Include: Add/Update header files of RISC-V CPU package Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 6/14]: RiscVPkg/opesbi: Add opensbi-HOWTO.txt Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 7/14]: RiscVPkg/RealTimeClockRuntimeDxe: Add RISC-V RTC Runtime Driver Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 8/14]: RiscVPkg/Universal: Remove stale moudles Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 9/14]: RiscVPkg/CpuDxe: Use RISC-V platform level timer library Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 10/14]: RiscVPkg/SmbiosDxe: RISC-V platform generic SMBIOS DXE driver Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 11/14]: RiscVPkg: Updates for supporting RISC-V OpenSBI Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 12/14]: RiscVVirtPkg: Remove RISC-V virtual package Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 13/14]: RiscVPkg/Library: Add/Update/Remove Library instances for RISC-V platform Abner Chang
2019-08-27 6:00 ` [edk2-staging/RISC-V PATCH v1 14/14]: MdeModulePkg/DxeIplPeim: Abstract platform DXEIPL on " Abner Chang
2019-08-28 8:17 ` [edk2-devel] [edk2-staging/RISC-V PATCH v1 1/14]: BaseTools: Update EDK2 build tool for " jonathan.cameron
2019-08-28 8:43 ` Abner Chang
2019-08-28 8:59 ` Jonathan Cameron
2019-08-28 9:08 ` Abner Chang [this message]
[not found] ` <15BF0B00F4581767.2982@groups.io>
2019-09-04 11:18 ` Abner Chang
2019-09-04 14:32 ` Jonathan Cameron
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=CS1PR8401MB119284595F65C793D9F246AAFFA30@CS1PR8401MB1192.NAMPRD84.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