summaryrefslogtreecommitdiff
path: root/ArmPkg
AgeCommit message (Collapse)Author
2016-05-12ArmPkg/PlatformBootManagerLib: implement new generic versionArd Biesheuvel
This implements the platform glue for the new generic BDS implementation. It is based on the ArmVirtQemu version, with the QEMU references removed. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-12ArmPkg/ArmLib: don't invalidate entire I-cache on range operationArd Biesheuvel
Instead of cleaning the data cache to the PoU by virtual address and subsequently invalidating the entire I-cache, invalidate only the range that we just cleaned. This way, we don't invalidate other cachelines unnecessarily. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/AArch64Mmu: don't let table entries inherit XN permission bitsArd Biesheuvel
When we split a block entry into a table entry, the UXN/PXN/XN permission attributes are inherited both by the new table entry and by the new block entries at the next level down. Unlike the NS bit, which only affects the next level of lookup, the XN table bits supersede the permissions of the final translation, and setting the permissions at multiple levels is not only redundant, it also prevents us from lifting XN restrictions on a subregion of the original block entry by simply clearing the appropriate bits at the lowest level. So drop the code that sets the UXN/PXN/XN bits on the table entries. Reported-by: "Oliyil Kunnil, Vishal" <vishalo@qti.qualcomm.com> Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/ArmDmaLib: assert that consistent mappings are uncachedArd Biesheuvel
DmaMap () only allows uncached mappings to be used for creating consistent mappings with operation type MapOperationBusMasterCommonBuffer. However, if the buffer passed to DmaMap () happens to be aligned to the CWG, there is no need for a bounce buffer, and we perform the cache maintenance directly without ever checking if the memory attributes of the buffer adhere to the API. So add some debug code that asserts that the operation type and the memory attributes are consistent. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Tested-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/ArmDmaLib: do not remap arbitrary memory regions as uncachedArd Biesheuvel
In the DmaMap () operation, if the region to be mapped happens to be aligned to the Cache Writeback Granule (CWG) (whose value is typically 64 or 128 bytes and 2 KB maximum), we remap the memory as uncached. Since remapping memory occurs at page granularity, while the buffer and the CWG may be much smaller, there is no telling what other memory we affect by doing this, especially since the operation is not reverted in DmaUnmap(). So remove the remapping call. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/ArmDmaLib: reject consistent DMA mappings of cached memoryArd Biesheuvel
DmaMap () operations of type MapOperationBusMasterCommonBuffer should return a mapping that is coherent between the CPU and the device. For this reason, the API only allows DmaMap () to be called with this operation type if the memory to be mapped was allocated by DmaAllocateBuffer (), which in this implementation guarantees the coherency by using uncached mappings on the CPU side. This means that, if we encounter a cached mapping in DmaMap () with this operation type, the code is either broken, or someone is violating the API, but simply proceeding with a double buffer makes no sense at all, and can only cause problems. So instead, actively reject this operation type for cached memory mappings. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/ArmDmaLib: interpret GCD attributes as a bit fieldArd Biesheuvel
Comparing a GCD attribute field directly against EFI_MEMORY_UC and EFI_MEMORY_WT is incorrect, since it may have other bits set as well which are not related to the cacheability of the region. So instead, test explicitly against the flags EFI_MEMORY_WB and EFI_MEMORY_WT, which must be set if the region may be mapped with cacheable attributes. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/ArmDmaLib: consistently use 'gCacheAlignment - 1' as alignment maskArd Biesheuvel
We manage to use both an AND operation with 'gCacheAlignment - 1' and a modulo operation with 'gCacheAlignment' in the same compound if statement. Since gCacheAlignment is a global of which the compiler cannot guarantee that it is a power of two, simply use the AND version in both cases. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-10ArmPkg/ArmDmaLib: deal with NULL return value of UncachedAllocatePages ()Ard Biesheuvel
The allocation function UncachedAllocatePages () may return NULL, in which case our implementation of DmaAllocateBuffer () should return EFI_OUT_OF_RESOURCES rather than silently ignoring the NULL value and returning EFI_SUCCESS. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-09ArmPkg/DefaultExceptionHandlerLib: fix typoArd Biesheuvel
Replace : with ; which was changes accidentally. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-05-09ArmPkg/DefaultExceptionHandlerLib: add stack dump to exception handling codeArd Biesheuvel
This adds a partial stack dump (256 bytes at either side of the stack pointer) to the CPU state dumping routine that is invoked when taking an unexpected exception. Since dereferencing the stack pointer may itself fault, ensure that we don't enter the dumping routine recursively. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-05-09ArmPkg/DefaultExceptionHandlerLib: use deadloop rather than ASSERTArd Biesheuvel
The default exception handler, which is essentially the one that is invoked for unexpected exceptions, ends with an ASSERT (FALSE), to ensure that execution halts after dumping the CPU state. However, ASSERTs are compiled out in RELEASE builds, and since we simply return to wherever the ELR is pointing, we will not make any progress in case of synchronous aborts, and the same exception will be taken again immediately, resulting in the string 'Exception at 0x....' to be printed over and over again. So use an explicit deadloop instead. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-04-29ArmPkg: implement CpuIo2 protocol driver specific for PCIArd Biesheuvel
The CpuIo2 protocol is required by the generic PciHostBridgeDxe driver, which relies on it to back its own I/O and MMIO operations. Since ARM has no native I/O port equivalent, such accesses can only originate from PCI drivers, and the PCI I/O space is translated to MMIO in this case. So we can implement this protocol using MMIO operations only, and take the PCI I/O translation offset into account when performing I/O port accesses. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-04-29ArmPlatformPkg: move PCI related PCD definitions to ArmPkgArd Biesheuvel
The PCI related PCDs are not platform specific, and architectural protocols such as CpuIo2 are based on PCI provided MMIO to IO translation, so these PCDs belong in ArmPkg not ArmPlatformPkg. NOTE: this *WILL* break some out-of-tree platforms, the fix is changing all consumers of gArmPlatformTokenSpaceGuid.PcdPci* to gArmTokenSpaceGuid.PcdPci* Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-04-25ArmPkg/Gic: Fix boundary checking bugHeyi Guo
mGicNumInterrupts is the total number of interrupts, so the interrupt ID equal to mGicNumInterrupts is also invalid. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Heyi Guo <heyi.guo@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-04-14ArmPkg/AArch64Mmu: disable MMU during page table manipulationsArd Biesheuvel
On ARM, manipulating live page tables is cumbersome since the architecture mandates the use of break-before-make, i.e., replacing a block entry with a table entry requires an intermediate step via an invalid entry, or TLB conflicts may occur. Since it is not generally feasible to decide in the page table manipulation routines whether such an invalid entry will result in those routines themselves to become unavailable, use a function that is callable with the MMU off (i.e., a leaf function that does not access the stack) to perform the change of a block entry into a table entry. Note that the opposite should never occur, i.e., table entries are never coalesced into block entries. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Mark Rutland <mark.rutland@arm.com>
2016-04-14ArmPkg/AArch64Mmu: Fix XN attribute for device memoryHeyi Guo
Now XN attribute will be set automatically if the region is declared as device memory. However, the function ArmMemoryAttributeToPageAttribute is to get attribute for block and page descriptors, not for table descriptors, so attribute TT_TABLE_*XN does not really take effect. Need to use TT_*XN_MASK instead. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Heyi Guo <heyi.guo@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-04-01ArmPkg/ArmArchTimerLib: correct typosEvan Lloyd
Some minor typographical problems were noticed during previous commits. This change corrects those, and contains no functional modifications. The changes are in comments, and one diagnostic message. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <evan.lloyd@arm.com> Reviewed-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-04-01ArmPkg/ArmArchTimerLib: fix unused variable in RELEASE buildsSami Mujawar
The TimerFreq variable in the TimerConstructor() is unused in RELEASE builds since ASSERTs are then disabled. The only use of the variable (in the ASSERT) is replaced by a direct invocation of the function previously used to set it. NOTE: The build tools suppress warnings of this using compiler options eg. -Wno-unused-but-set-variable for GCC toolchain or --diag_suppress=550 for RVCT toolchain. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <evan.lloyd@arm.com> Reviewed-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-04-01ArmPkg/ArmArchTimerLib: add GetTimeInNanoSecond() to ArmArchTimerLibSami Mujawar
FirmwarePerformanceDxe.c utilizes the Timer Library function GetTimeInNanoSecond() which was not implemented by the ArmArchTimerLib. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <evan.lloyd@arm.com> Reviewed-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-04-01ArmPkg/ArchArmTimerLib: refactor MultU64xN and TimerFreq definitionsSami Mujawar
This refactors some timer code to define MultU64xN as a preprocessor symbol rather than a function pointer, and to factor out the code that obtains the timer frequency into GetPlatformTimerFreq (). Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <evan.lloyd@arm.com> Reviewed-by: Ryan Harkin <ryan.harkin@linaro.org> Contributed-under: TianoCore Contribution Agreement 1.0 [ard.biesheuvel: split off from 'add GetTimeInNanoSecond() to ArmArchTimerLib'] Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-03-25ArmPkg|EmbeddedPkg: make PcdCpuVectorBaseAddress 64 bits wideLeendert van Doorn
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Leo Duran <leo.duran@amd.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-03-25ArmPkg: apply Cortex-A57 errataLeendert van Doorn
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Leo Duran <leo.duran@amd.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-03-22ArmPkg/AArch64Mmu: use correct AP[] bits in ArmClearMemoryRegionReadOnlyArd Biesheuvel
The function ArmClearMemoryRegionReadOnly() was supposed to undo the effect of ArmSetMemoryRegionReadOnly(), but instead, it sets the permissions to EL0-no access, EL1-read-only. Since the EL0 bit should be 1 to align with EL2/3 (where the bit is SBO), use TT_AP_RW_RW instead, which makes the entry read-write for EL0 when executing at EL1, and read-write for all other levels. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-03-22ArmPkg/ArmExceptionLib: reimplement register stack/unstack routinesArd Biesheuvel
This replaces the somewhat opaque preprocessor based stack/unstack macros with open coded ldp/stp sequences to preserve the interrupted context before handing over to the exception handler in C. This removes various arithmetic operations on the stack pointer, and reduces the exception return critical section to its minimum size (i.e., the bare minimum required to populate the ELR and SPSR registers and invoke the eret). Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-22ArmPkg/ArmExceptionLib: avoid indirect call if using vector table in placeArd Biesheuvel
If we are using the vector table in place, there is no need to make an indirect call to the common handler routine from the vector table entries, so just use a straight branch instruction in that case. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-22ArmPkg/ArmExceptionLib: make build time define visible to the compilerArd Biesheuvel
The global gArmRelocateVectorTable is a build time constant, but due to its external linkage and lack of constness, the compiler does not see that. So turn it into a static boolean, and at the same time, make the function CopyExceptionHandlers() (which is only called if gArmRelocateVectorTable is set) static as well, so that the compiler can eliminate it completely if we are using the vector table in place. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-22ArmPkg/ArmExceptionLib: don't restore ESR and FAR upon exception returnArd Biesheuvel
ESR and FAR are populated by the hardware upon exception entry, and describe the exception, not the interrupted context. So there is no point in restoring their values before returning from the exception. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-22ArmPkg/ArmExceptionLib: stack FPSR on common pathArd Biesheuvel
We have three code paths to stack/unstack the exception context, one for each of EL3, EL2 and EL1. However, they all access the same copy of FPSR so move that access to the common path. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-22ArmPkg/ArmExceptionLib: fold exception handler prologue into vector tableArd Biesheuvel
Unlike the AArch32 vector table, which has room for a single instruction for each exception type, the AArch64 exception table has 128 byte slots, which can easily hold the shared prologues that are emitted out of line. So refactor this code into a single macro, and expand it into each vector table slot. Since the address of the command handler entry point is no longer patched in by the C code, we can just emit the literal into each vector entry directly. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-22ArmPkg/AsmMacroIoLibV8: remove undocumented assumption from ELx macrosArd Biesheuvel
The macros EL1_OR_EL2() and EL1_OR_EL2_OR_EL3() allow conditional execution of assembly sequences based on the current exception level, by jumping to caller supplied labels 1f, 2f or 3f. However, the jump to 1f is actually a fallthrough, which means the EL1 code needs to follow right after the macro invocation, and the 1f label is ignored. So let's fix this by making all jumps explicit. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> Reviewed-by: Eugene Cohen <eugene@hp.com>
2016-03-17ArmPkg: update CpuDxe to use CpuExceptionHandlerLibCohen, Eugene
Use the new ARM/AArch64 implementation of the base CpuExceptionHandlerLib library from CpuDxe to centralize exception handling. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Eugene Cohen <eugene@hp.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-03-16ArmPkg: ARM/AArch64 implementation of CpuExceptionHandlerLibCohen, Eugene
Introduce ARM and AArch64 instances of the CpuExceptionHandlerLib which provides exception handling and registration of handlers regardless of execution phase. Two variants of the ArmExceptionLib are provided: one where exception handlers reside within the module (meeting appropriate architectural alignment requirements for the vector table) and another one that will relocate a copy of thee xception handlers to an address specified by PcdCpuVectorBaseAddress. The ArmRelocateExceptionLib is intended for use in cases where ArmExceptionLib is too large for the application (uncompressed XIP images) as driven by the vector table alignment padding. The AArch64 build of this library supports execution at EL1, EL2, and EL3 exception levels. Tested on ARM, and AArch64 with SEC, DXE Core, and CpuDxe modules. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Eugene Cohen <eugene@hp.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-03-16ArmPkg/ArmLib: add ArmReadHcr to enable read-modify-write of HCRCohen, Eugene
Add ArmReadHcr() to ArmLib to enable read-modify-write of the HCR system register. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Eugene Cohen <eugene@hp.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-03-03ArmPkg: Configure TTBCR registerEvan Lloyd
Architecturally, the TTBCR register value is undefined at reset for Non-Secure. On some platforms the reset value for TTBCR is not zero and this causes a data abort exception once the MMU is enabled. This patch configures the TTBCR register to enable translation table walk using TTBR0. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <evan.lloyd@arm.com>
2016-02-23ArmPkg: CpuDxe: don't track interrupt state in a global variableCohen, Eugene
Update the CpuDxe driver to remove an assumption that it is the only component modifying interrupt state since this can be done through BaseLib as well. Instead of using a global variable for last interrupt state we now check the current PSTATE value directly. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Eugene Cohen <eugene@hp.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-02-23ArmPkg: CpuDxe: fix AArch64 interrupt read masksCohen, Eugene
The AArch64 DAIF bits are different for reading (mrs) versus writing (msr). The bitmask definitions assumed they were the same causing incorrect results when trying to determine the current interrupt state through ArmGetInterruptState. The logic for interpreting the DAIF read data using the csel instruction was also incorrect and is fixed. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Eugene Cohen <eugene@hp.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-02-15ArmPkg: DefaultExceptionHandler fixes for use with DxeCoreCohen, Eugene
Modify the DefaultExceptionHandler (uefi-variant) so it can be used by DxeCore (via CpuExceptionHandlerLib) where the debug info table is not yet published at library constructor time. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Eugene Cohen <eugene@hp.com> Tested-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2016-02-03ArmPkg: Add isb when setting SCREvan Lloyd
Some updates to SCR can cause a problem which manifests as an undefined opcode exception. This may be when a speculative secure instruction fetch happens after the NS bit is set. An isb is required to make the register change take effect fully. Contributed-under: Tianocore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <Evan.Lloyd@arm.com> Reviewed-by: Sami Mujawar <Sami.Mujawar@arm.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2016-01-30ArmPkg: Rectify file modesEvan Lloyd
Problems have been encountered because some of the source files have execute permission set. This can cause git to report them as changed when they are checked out onto a file system with inherited permissions. This has been seen using Cygwin, MinGW and PowerShell Git. This patch makes no change to source file content, and only aims to correct the file modes/permissions. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Evan Lloyd <evan.lloyd@arm.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19778 6f19259b-4bc3-4df7-8a09-765794883524
2016-01-20ArmPkg: LinuxLoader: fix typo in help stringRyan Harkin
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19698 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-17ArmPkg/ArmSoftfloatLib: add missing entry points for RVCTArd Biesheuvel
The RVCT compiler may emit calls to the various __aeabi_c?cmp?? functions, which return their results via the CPU condition flags C and Z. According to ARM doc IHI 0043D 'Run-time ABI for the ARM architecture': The 3-way comparison functions c*cmple, c*cmpeq and c*rcmple return their results in the CPSR Z and C flags. C is clear only if the operands are ordered and the first operand is less than the second. Z is set only when the operands are ordered and equal. Add implementations for the double and float variants of the above. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19327 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-16ArmPkg: rewrite vector table population macrosArd Biesheuvel
Unfortunately, Clang does not support the use of symbol references in .org directives, and bails with the following error message when it encounters them: <...>:error: expected assembly-time absolute expression .org DebugAgentVectorTable + 0x000 So replace the .org arguments with absolute values, and move the whole vector table into a subsection with the appropriate alignment, and starting at .org 0x0. This gives the same protection with respect to entries that exceed 128 bytes, in a way that Clang supports as well. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Mark Rutland <mark.rutland@arm.com> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19303 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg/ArmV7Mmu: prefer non shareable memory on non-coherent hardwareArd Biesheuvel
Commit SVN r18778 made all mappings of normal memory (inner) shareable, even on hardware that implements shareability as uncached accesses. The original concerns that prompted the change, regarding coherent DMA and virt guests migrating between CPUs, do not apply to such hardware, so revert to the original behavior in that case. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19285 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg/ArmSoftfloatLib: restrict -fno-tree-vrp option to GCC46 and GCC47Ard Biesheuvel
The -fno-tree-vrp option is not required for GCC 4.8 or later, and is not supported by CLANG. So restrict its use to GCC 4.6 and 4.7, which are the oldest versions we support for ARM. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19283 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg/ArmV7Lib: add CLANG alternative for FPEXC accessArd Biesheuvel
The open coded access to co-processor #10 to set FPEXC is not supported by the CLANG assembler, but the architecturally correct VMSR instruction is not supported by older binutils. So keep the former unless __clang__ is defined. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19282 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg/CompilerIntrinsicsLib: add GCC version of __aeabi_memset()Ard Biesheuvel
CLANG for ARM may emit calls to __aeabi_memset(), which is subtly different from the default memset() [arguments 2 and 3 are reversed] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19281 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg: use unified asm syntax for CLANGArd Biesheuvel
The CLANG assembler does not support the legacy, non-unified assembler syntax, i.e., it does not support the reordering of the condition suffixes with the increment/decrement before/after or byte/word suffixes, and it does not recognize the 'empty descending' (ED) suffix at all. So move to the unified syntax, and replace 'empty descending' with 'decrement after' or 'increment before' as appropriate. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19280 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg/ArmGic: fix bug in GICv3 distributor configurationArd Biesheuvel
In the function ArmGicEnableDistributor (), the Affinity Routing Enable (ARE) bit, which essentially defines whether the GIC runs in v2 or v3 mode, is inadvertently cleared when enabling the GIC distributor if it is running in v3 mode. So fix that. Reported-by: Supreeth Venkatesh <Supreeth.Venkatesh@arm.com> Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19274 6f19259b-4bc3-4df7-8a09-765794883524
2015-12-15ArmPkg/CpuDxe: drop ARMv4 exception handling codeArd Biesheuvel
Since we do not support anything below ARMv7, let's promote the ARMv6 exception handling code in CpuDxe to the only version we provide for ARM. This means we can drop the unused ARMv4 version. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19273 6f19259b-4bc3-4df7-8a09-765794883524