1. 01 Nov, 2018 1 commit
  2. 29 Oct, 2018 1 commit
    • Soby Mathew's avatar
      PIE: Position Independant Executable support for BL31 · 931f7c61
      Soby Mathew authored
      
      
      This patch introduces Position Independant Executable(PIE) support
      in TF-A. As a initial prototype, only BL31 can support PIE. A trivial
      dynamic linker is implemented which supports fixing up Global Offset
      Table(GOT) and Dynamic relocations(.rela.dyn). The fixup_gdt_reloc()
      helper function implements this linker and this needs to be called
      early in the boot sequence prior to invoking C functions. The GOT is
      placed in the RO section of BL31 binary for improved security and the
      BL31 linker script is modified to export the appropriate symbols
      required for the dynamic linker.
      
      The C compiler always generates PC relative addresses to linker symbols
      and hence referencing symbols exporting constants are a problem when
      relocating the binary. Hence the reference to the
      `__PERCPU_TIMESTAMP_SIZE__` symbol in PMF is removed and is now calculated
      at runtime based on start and end addresses.
      
      Change-Id: I1228583ff92cf432963b7cef052e95d995cca93d
      Signed-off-by: default avatarSoby Mathew <soby.mathew@arm.com>
      931f7c61
  3. 16 Oct, 2018 1 commit
  4. 04 Oct, 2018 1 commit
  5. 03 Oct, 2018 1 commit
  6. 28 Sep, 2018 1 commit
  7. 18 Sep, 2018 1 commit
    • Antonio Nino Diaz's avatar
      BL31: Fix warning about BL32 init function · 74ad948f
      Antonio Nino Diaz authored
      
      
      The expected value for failure is 0, so the warning only has to be shown
      in that case. This is the way the TSPD has done it since it was
      introduced, and the way SPM and OP-TEE do it.
      
      Trusty wrongly returns 0 on success.
      
      In the case of TLK, the return value of tlkd_init() is passed from the
      secure world in register X1 when calling the SMC TLK_ENTRY_DONE.
      
      Change-Id: I39106d67631ee57f109619f8830bf4b9d96155e6
      Signed-off-by: default avatarAntonio Nino Diaz <antonio.ninodiaz@arm.com>
      74ad948f
  8. 06 Sep, 2018 1 commit
  9. 30 Aug, 2018 1 commit
  10. 20 Aug, 2018 2 commits
    • Jeenu Viswambharan's avatar
      EHF: MISRA fixes · 03b645ed
      Jeenu Viswambharan authored
      
      
      These changes address most of the required MISRA rules. In the process,
      some from generic code are also fixed.
      
      No functional changes.
      
      Change-Id: I19786070af7bc5e1f6d15bdba93e22a4451d8fe9
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      03b645ed
    • Jeenu Viswambharan's avatar
      AArch64: Enable MPAM for lower ELs · 5f835918
      Jeenu Viswambharan authored
      
      
      Memory Partitioning And Monitoring is an Armv8.4 feature that enables
      various memory system components and resources to define partitions.
      Software running at various ELs can then assign themselves to the
      desired partition to control their performance aspects.
      
      With this patch, when ENABLE_MPAM_FOR_LOWER_ELS is set to 1, EL3 allows
      lower ELs to access their own MPAM registers without trapping to EL3.
      This patch however doesn't make use of partitioning in EL3; platform
      initialisation code should configure and use partitions in EL3 if
      required.
      
      Change-Id: I5a55b6771ccaa0c1cffc05543d2116b60cbbcdcd
      Co-authored-by: default avatarJames Morse <james.morse@arm.com>
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      5f835918
  11. 17 Aug, 2018 1 commit
  12. 19 Jul, 2018 3 commits
    • Jeenu Viswambharan's avatar
      RAS: Introduce handler for EL3 EAs · eaeaa4d0
      Jeenu Viswambharan authored
      
      
      External Aborts while executing in EL3 is fatal in nature. This patch
      allows for the platform to define a handler for External Aborts received
      while executing in EL3. A default implementation is added which falls
      back to platform unhandled exception.
      
      Change-Id: I466f2c8113a33870f2c7d2d8f2bf20437d9fd354
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      eaeaa4d0
    • Jeenu Viswambharan's avatar
      RAS: Introduce handler for Double Faults · d5a23af5
      Jeenu Viswambharan authored
      
      
      Double fault is when the PE receives another error whilst one is being
      handled. To detect double fault condition, a per-CPU flag is introduced
      to track the status of error handling. The flag is checked/modified
      while temporarily masking external aborts on the PE.
      
      This patch routes double faults to a separate platform-defined handler.
      
      Change-Id: I70e9b7ba4c817273c55a0af978d9755ff32cc702
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      d5a23af5
    • Jeenu Viswambharan's avatar
      RAS: Introduce handler for Uncontainable errors · b56dc2a9
      Jeenu Viswambharan authored
      
      
      Uncontainable errors are the most severe form of errors, which typically
      mean that the system state can't be trusted any more. This further means
      that normal error recovery process can't be followed, and an orderly
      shutdown of the system is often desirable.
      
      This patch allows for the platform to define a handler for Uncontainable
      errors received. Due to the nature of Uncontainable error, the handler
      is expected to initiate an orderly shutdown of the system, and therefore
      is not expected to return. A default implementation is added which falls
      back to platform unhandled exception.
      
      Also fix ras_arch.h header guards.
      
      Change-Id: I072e336a391a0b382e77e627eb9e40729d488b55
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      b56dc2a9
  13. 17 Jul, 2018 2 commits
  14. 11 Jul, 2018 2 commits
    • Roberto Vargas's avatar
      Add end_vector_entry assembler macro · a9203eda
      Roberto Vargas authored
      
      
      Check_vector_size checks if the size of the vector fits
      in the size reserved for it. This check creates problems in
      the Clang assembler. A new macro, end_vector_entry, is added
      and check_vector_size is deprecated.
      
      This new macro fills the current exception vector until the next
      exception vector. If the size of the current vector is bigger
      than 32 instructions then it gives an error.
      
      Change-Id: Ie8545cf1003a1e31656a1018dd6b4c28a4eaf671
      Signed-off-by: default avatarRoberto Vargas <roberto.vargas@arm.com>
      a9203eda
    • Roberto Vargas's avatar
      Use ALIGN instead of NEXT in linker scripts · 5629b2b1
      Roberto Vargas authored
      
      
      Clang linker doesn't support NEXT. As we are not using the MEMORY command
      to define discontinuous memory for the output file in any of the linker
      scripts, ALIGN and NEXT are equivalent.
      
      Change-Id: I867ffb9c9a76d4e81c9ca7998280b2edf10efea0
      Signed-off-by: default avatarRoberto Vargas <roberto.vargas@arm.com>
      5629b2b1
  15. 27 Jun, 2018 1 commit
    • Jeenu Viswambharan's avatar
      DynamIQ: Enable MMU without using stack · 64ee263e
      Jeenu Viswambharan authored
      
      
      Having an active stack while enabling MMU has shown coherency problems.
      This patch builds on top of translation library changes that introduces
      MMU-enabling without using stacks.
      
      Previously, with HW_ASSISTED_COHERENCY, data caches were disabled while
      enabling MMU only because of active stack. Now that we can enable MMU
      without using stack, we can enable both MMU and data caches at the same
      time.
      
      NOTE: Since this feature depends on using translation table library v2,
      disallow using translation table library v1 with HW_ASSISTED_COHERENCY.
      
      Fixes ARM-software/tf-issues#566
      
      Change-Id: Ie55aba0c23ee9c5109eb3454cb8fa45d74f8bbb2
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      64ee263e
  16. 21 Jun, 2018 2 commits
    • Jeenu Viswambharan's avatar
      SDEI: Make dispatches synchronous · cdb6ac94
      Jeenu Viswambharan authored
      
      
      SDEI event dispatches currently only sets up the Non-secure context
      before returning to the caller. The actual dispatch only happens upon
      exiting EL3 next time.
      
      However, for various error handling scenarios, it's beneficial to have
      the dispatch happen synchronously. I.e. when receiving SDEI interrupt,
      or for a successful sdei_dispatch_event() call, the event handler is
      executed; and upon the event completion, dispatcher execution resumes
      after the point of dispatch. The jump primitives introduced in the
      earlier patch facilitates this feature.
      
      With this patch:
      
        - SDEI interrupts and calls to sdei_dispatch_event prepares the NS
          context for event dispatch, then sets a jump point, and immediately
          exits EL3. This results in the client handler executing in
          Non-secure.
      
        - When the SDEI client completes the dispatched event, the SDEI
          dispatcher does a longjmp to the jump pointer created earlier. For
          the caller of the sdei_dispatch_event() in particular, this would
          appear as if call returned successfully.
      
      The dynamic workaround for CVE_2018_3639 is slightly shifted around as
      part of related minor refactoring. It doesn't affect the workaround
      functionality.
      
      Documentation updated.
      
      NOTE: This breaks the semantics of the explicit dispatch API, and any
      exiting usages should be carefully reviewed.
      
      Change-Id: Ib9c876d27ea2af7fb22de49832e55a0da83da3f9
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      cdb6ac94
    • Jeenu Viswambharan's avatar
      BL31: Introduce jump primitives · e7b9473e
      Jeenu Viswambharan authored
      
      
      This patch introduces setjmp() and ongjmp() primitives to enable
      standard setjmp/longjmp style execution. Both APIs parameters take a
      pointer to struct jmpbuf type, which hosts CPU registers saved/restored
      during jump.
      
      As per the standard usage:
      
        - setjmp() return 0 when a jump is setup; and a non-zero value when
          returning from jump.
      
        - The caller of setjmp() must not return, or otherwise update stack
          pointer since.
      
      Change-Id: I4af1d32e490cfa547979631b762b4cba188d0551
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      e7b9473e
  17. 13 Jun, 2018 1 commit
    • Sandrine Bailleux's avatar
      SPM: Treat SP xlat tables the same as others · d801a1d0
      Sandrine Bailleux authored
      The translation tables allocated for the Secure Partition do not need
      to be treated as a special case. They can be put amongst the other
      tables mapping BL31's general purpose memory. They will be mapped with
      the same attributes as them, which is fine.
      
      The explicit alignment constraint in BL31's linker script to pad the
      last page of memory allocated to the Secure Partition's translation
      tables is useless too, as page tables are per se pages, thus their
      end address is naturally aligned on a page-boundary.
      
      In fact, this patch does not change the existing behaviour. Since
      patch 22282bb6
      
       ("SPM: Move all SP-related info to SP context
      struct"), the secure_partition.c file has been renamed into sp_xlat.c
      but the linker script has not been properly updated. As a result, the
      SP translation tables are not specifically put at the start of the
      xlat_table linker section, the __SP_IMAGE_XLAT_TABLES_START__/_END__
      symbols have the same value, the size of the resulting mmap_region
      covering these xlat tables is 0 and so it is ignored.
      
      Change-Id: I4cf0a4cc090298811cca53fc9cee74df0f2b1512
      Signed-off-by: default avatarSandrine Bailleux <sandrine.bailleux@arm.com>
      d801a1d0
  18. 23 May, 2018 1 commit
  19. 04 May, 2018 3 commits
    • Jeenu Viswambharan's avatar
      AArch64: Introduce RAS handling · 14c6016a
      Jeenu Viswambharan authored
      
      
      RAS extensions are mandatory for ARMv8.2 CPUs, but are also optional
      extensions to base ARMv8.0 architecture.
      
      This patch adds build system support to enable RAS features in ARM
      Trusted Firmware. A boolean build option RAS_EXTENSION is introduced for
      this.
      
      With RAS_EXTENSION, an Exception Synchronization Barrier (ESB) is
      inserted at all EL3 vector entry and exit. ESBs will synchronize pending
      external aborts before entering EL3, and therefore will contain and
      attribute errors to lower EL execution. Any errors thus synchronized are
      detected via. DISR_EL1 register.
      
      When RAS_EXTENSION is set to 1, HANDLE_EL3_EA_FIRST must also be set to 1.
      
      Change-Id: I38a19d84014d4d8af688bd81d61ba582c039383a
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      14c6016a
    • Jeenu Viswambharan's avatar
      AArch64: Introduce External Abort handling · 76454abf
      Jeenu Viswambharan authored
      
      
      At present, any External Abort routed to EL3 is reported as an unhandled
      exception and cause a panic. This patch enables ARM Trusted Firmware to
      handle External Aborts routed to EL3.
      
      With this patch, when an External Abort is received at EL3, its handling
      is delegated to plat_ea_handler() function. Platforms can provide their
      own implementation of this function. This patch adds a weak definition
      of the said function that prints out a message and just panics.
      
      In order to support handling External Aborts at EL3, the build option
      HANDLE_EA_EL3_FIRST must be set to 1.
      
      Before this patch, HANDLE_EA_EL3_FIRST wasn't passed down to
      compilation; this patch fixes that too.
      
      Change-Id: I4d07b7e65eb191ff72d63b909ae9512478cd01a1
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      76454abf
    • Jeenu Viswambharan's avatar
      AArch64: Refactor GP register restore to separate function · ef653d93
      Jeenu Viswambharan authored
      
      
      At present, the function that restores general purpose registers also
      does ERET. Refactor the restore code to restore general purpose
      registers without ERET to complement the save function.
      
      The macro save_x18_to_x29_sp_el0 was used only once, and is therefore
      removed, and its contents expanded inline for readability.
      
      No functional changes, but with this patch:
      
        - The SMC return path will incur an branch-return and an additional
          register load.
      
        - The unknown SMC path restores registers x0 to x3.
      
      Change-Id: I7a1a63e17f34f9cde810685d70a0ad13ca3b7c50
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      ef653d93
  20. 23 Apr, 2018 1 commit
    • Antonio Nino Diaz's avatar
      Add support for the SMC Calling Convention 2.0 · 2f370465
      Antonio Nino Diaz authored
      
      
      Due to differences in the bitfields of the SMC IDs, it is not possible
      to support SMCCC 1.X and 2.0 at the same time.
      
      The behaviour of `SMCCC_MAJOR_VERSION` has changed. Now, it is a build
      option that specifies the major version of the SMCCC that the Trusted
      Firmware supports. The only two allowed values are 1 and 2, and it
      defaults to 1. The value of `SMCCC_MINOR_VERSION` is derived from it.
      
      Note: Support for SMCCC v2.0 is an experimental feature to enable
      prototyping of secure partition specifications. Support for this
      convention is disabled by default and could be removed without notice.
      
      Change-Id: I88abf9ccf08e9c66a13ce55c890edea54d9f16a7
      Signed-off-by: default avatarAntonio Nino Diaz <antonio.ninodiaz@arm.com>
      2f370465
  21. 14 Mar, 2018 1 commit
    • Dimitris Papastamos's avatar
      Fixup `SMCCC_ARCH_FEATURES` semantics · a205a56e
      Dimitris Papastamos authored
      
      
      When querying `SMCCC_ARCH_WORKAROUND_1` through `SMCCC_ARCH_FEATURES`,
      return either:
        * -1 to indicate the PE on which `SMCCC_ARCH_FEATURES` is called
          requires firmware mitigation for CVE-2017-5715 but the mitigation
          is not compiled in.
        * 0 to indicate that firmware mitigation is required, or
        * 1 to indicate that no firmware mitigation is required.
      
      This patch complies with v1.2 of the firmware interfaces
      specification (ARM DEN 0070A).
      
      Change-Id: Ibc32d6620efdac6c340758ec502d95554a55f02a
      Signed-off-by: default avatarDimitris Papastamos <dimitris.papastamos@arm.com>
      a205a56e
  22. 01 Mar, 2018 1 commit
    • Dan Handley's avatar
      Suppress spurious deprecated declaration warnings · 97924e45
      Dan Handley authored
      
      
      Some generic compatibility functions emit deprecated declaration warnings
      even when platforms do not use the deprecated functions directly. This
      can be confusing. Suppress these warnings by using:
      `#pragma GCC diagnostic ignored "-Wdeprecated-declarations"`
      
      Also emit a runtime warning if the weak plat/common implemntation of
      plat_get_syscnt_freq2() is used, as this implies the platform has not
      migrated from plat_get_syscnt_freq(). The deprecated  declaration warnings
      only help detect when platforms are calling deprecated functions, not when
      they are defining deprecated functions.
      
      Fixes ARM-software/tf-issues#550
      
      Change-Id: Id14a92279c2634c1e76db8ef210da8affdbb2a5d
      Signed-off-by: default avatarDan Handley <dan.handley@arm.com>
      97924e45
  23. 28 Feb, 2018 2 commits
  24. 27 Feb, 2018 2 commits
    • Jeenu Viswambharan's avatar
      EHF: Introduce preempted return code parameter to ehf_allow_ns_preemption() · af34cd72
      Jeenu Viswambharan authored
      
      
      When a Yielding SMC is preempted, it's possible that Non-secure world is
      resumed afterwards. In this case, Non-secure execution would find itself
      in a state where the SMC has returned. However, the dispatcher might not
      get an opportunity to populate the corrected return code for having
      been preempted, and therefore the caller of the Yielding SMC cannot
      reliably determine whether the SMC had successfully completed or had
      been preempted.
      
      To solve this, this patch introduces a new parameter to the
      ehf_allow_ns_preemption() API. An SPD, through this parameter, would
      provide the expected error code when a Yielding SMC is preempted. EHF
      can then populate the specified value in x0 of the Non-secure context so
      that the caller of the Yielding SMC correctly identifies the SMC return
      as a preemption.
      
      Documentation updates to follow.
      
      Change-Id: Ia9c3f8f03f9d72d81aa235eaae2ee0374b972e1e
      Signed-off-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      af34cd72
    • Antonio Nino Diaz's avatar
      Add comments about mismatched TCR_ELx and xlat tables · 883d1b5d
      Antonio Nino Diaz authored
      
      
      When the MMU is enabled and the translation tables are mapped, data
      read/writes to the translation tables are made using the attributes
      specified in the translation tables themselves. However, the MMU
      performs table walks with the attributes specified in TCR_ELx. They are
      completely independent, so special care has to be taken to make sure
      that they are the same.
      
      This has to be done manually because it is not practical to have a test
      in the code. Such a test would need to know the virtual memory region
      that contains the translation tables and check that for all of the
      tables the attributes match the ones in TCR_ELx. As the tables may not
      even be mapped at all, this isn't a test that can be made generic.
      
      The flags used by enable_mmu_xxx() have been moved to the same header
      where the functions are.
      
      Also, some comments in the linker scripts related to the translation
      tables have been fixed.
      
      Change-Id: I1754768bffdae75f53561b1c4a5baf043b45a304
      Signed-off-by: default avatarAntonio Nino Diaz <antonio.ninodiaz@arm.com>
      883d1b5d
  25. 26 Feb, 2018 1 commit
    • Soby Mathew's avatar
      Introduce the new BL handover interface · a6f340fe
      Soby Mathew authored
      
      
      This patch introduces a new BL handover interface. It essentially allows
      passing 4 arguments between the different BL stages. Effort has been made
      so as to be compatible with the previous handover interface. The previous
      blx_early_platform_setup() platform API is now deprecated and the new
      blx_early_platform_setup2() variant is introduced. The weak compatiblity
      implementation for the new API is done in the `plat_bl_common.c` file.
      Some of the new arguments in the new API will be reserved for generic
      code use when dynamic configuration support is implemented. Otherwise
      the other registers are available for platform use.
      
      Change-Id: Ifddfe2ea8e32497fe1beb565cac155ad9d50d404
      Signed-off-by: default avatarSoby Mathew <soby.mathew@arm.com>
      a6f340fe
  26. 15 Feb, 2018 1 commit
    • Antonio Nino Diaz's avatar
      Redefine SMC_UNK as -1 instead of 0xFFFFFFFF · 4abd7fa7
      Antonio Nino Diaz authored
      
      
      According to the SMC Calling Convention (ARM DEN0028B):
      
          The Unknown SMC Function Identifier is a sign-extended value of
          (-1) that is returned in R0, W0 or X0 register.
      
      The value wasn't sign-extended because it was defined as a 32-bit
      unsigned value (0xFFFFFFFF).
      
      SMC_PREEMPT has been redefined as -2 for the same reason.
      
      NOTE: This might be a compatibility break for some AArch64 platforms
      that don't follow the previous version of the SMCCC (ARM DEN0028A)
      correctly. That document specifies that only the bottom 32 bits of the
      returned value must be checked. If a platform relies on the top 32 bits
      of the result being 0 (so that SMC_UNK is 0x00000000FFFFFFFF), it will
      have to fix its code to comply with the SMCCC.
      
      Change-Id: I7f7b109f6b30c114fe570aa0ead3c335383cb54d
      Signed-off-by: default avatarAntonio Nino Diaz <antonio.ninodiaz@arm.com>
      4abd7fa7
  27. 12 Feb, 2018 1 commit
  28. 29 Jan, 2018 2 commits
  29. 11 Jan, 2018 1 commit