1. 26 Mar, 2014 3 commits
    • Vikram Kanigiri's avatar
      Initialise UART console in all bootloader stages · 0796fe01
      Vikram Kanigiri authored
      This patch reworks the console driver to ensure that each bootloader stage
      initializes it independently. As a result, both BL3-1 and BL2 platform code
      now calls console_init() instead of relying on BL1 to perform console setup
      
      Fixes ARM-software/tf-issues#120
      
      Change-Id: Ic4d66e0375e40a2fc7434afcabc8bbb4715c14ab
      0796fe01
    • Soby Mathew's avatar
      Move console functions out of pl011.c · c1df3be7
      Soby Mathew authored
      This commit isolates the accessor functions in pl011.c and builds
      a wrapper layer for console functions.
      
      This also modifies the console driver to use the pl011 FIFO.
      
      Fixes ARM-software/tf-issues#63
      
      Change-Id: I3b402171cd14a927831bf5e5d4bb310b6da0e9a8
      c1df3be7
    • Vikram Kanigiri's avatar
      Add standby state support in PSCI cpu_suspend api · d118f9f8
      Vikram Kanigiri authored
      This patch adds support in the generic PSCI implementation to call a
      platform specific function to enter a standby state using an example
      implementation in ARM FVP port
      
      Fixes ARM-software/tf-issues#94
      Change-Id: Ic1263fcf25f28e09162ad29dca954125f9aa8cc9
      d118f9f8
  2. 21 Mar, 2014 3 commits
    • Sandrine Bailleux's avatar
      Semihosting: Fix file mode to load binaries on Windows · 886278e5
      Sandrine Bailleux authored
      Trusted firmware binaries loaded via semihosting used to be
      opened using 'r' mode (i.e. read mode).  This is fine on POSIX
      conforming systems (including Linux) but for Windows it also means
      that the file should be opened in text mode. 'rb' mode must be
      specified instead for binary mode.  On POSIX conforming systems,
      'rb' mode is equivalent to 'r' mode so it does no harm.
      
      Fixes ARM-software/tf-issues#69
      
      Change-Id: Ifa53f2ecfd765f572dea5dd73191f9fe2b2c2011
      886278e5
    • Vikram Kanigiri's avatar
      Remove partially qualified asm helper functions · 6ba0b6d6
      Vikram Kanigiri authored
      Each ARM Trusted Firmware image should know in which EL it is running
      and it should use the corresponding register directly instead of reading
      currentEL and knowing which asm register to read/write
      
      Change-Id: Ief35630190b6f07c8fbb7ba6cb20db308f002945
      6ba0b6d6
    • Vikram Kanigiri's avatar
      Fix the disable_mmu code · 5132060c
      Vikram Kanigiri authored
      Remove the hard coding of all the MMU related registers with 0 and disable MMU
      by clearing the M and C bit in SCTLR_ELx
      
      Change-Id: I4a0b1bb14a604734b74c32eb31315d8504a7b8d8
      5132060c
  3. 10 Mar, 2014 2 commits
    • Jeenu Viswambharan's avatar
      Move architecture timer setup to platform-specific code · 1c297bf0
      Jeenu Viswambharan authored
      At present, bl1_arch_setup() and bl31_arch_setup() program the counter
      frequency using a value from the memory mapped generic timer. The
      generic timer however is not necessarily present on all ARM systems
      (although it is architected to be present on all server systems).
      
      This patch moves the timer setup to platform-specific code and updates
      the relevant documentation. Also, CNTR.FCREQ is set as the specification
      requires the bit corresponding to the counter's frequency to be set when
      enabling. Since we intend to use the base frequency, set bit 8.
      
      Fixes ARM-software/tf-issues#24
      
      Change-Id: I32c52cf882253e01f49056f47c58c23e6f422652
      1c297bf0
    • Jeenu Viswambharan's avatar
      Remove unused 'CPU present' flag · 92a12866
      Jeenu Viswambharan authored
      This patch removes the 'CPU present' flag that's being set but not
      referred or used anywhere else.
      
      Change-Id: Iaf82bdb354134e0b33af16c7ba88eb2259b2682a
      92a12866
  4. 05 Mar, 2014 4 commits
    • Ryan Harkin's avatar
      fvp: plat_io_storage: remove duplicated code · 48e2ca79
      Ryan Harkin authored
      
      
      Fixes ARM-software/tf-issues#41
      
      The policy functions for each file to be loaded were implemented by
      copy/pasting one method and then varying the data checked.
      
      This patch creates a generic function to check the policy based on the
      data stored in a table.
      
      This removes the amount of duplicated code but also makes the code
      simpler and more efficient.
      
      Change-Id: I1c52eacf6f18a1442dabbb33edd03d4bb8bbeae0
      Signed-off-by: default avatarRyan Harkin <ryan.harkin@linaro.org>
      48e2ca79
    • Jon Medhurst's avatar
      fvp: Make use of the generic MMU translation table setup code · 38aa76a8
      Jon Medhurst authored
      
      
      Change-Id: I559c5a4d86cad55ce3f6ad71285b538d3cfd76dc
      Signed-off-by: default avatarJon Medhurst <tixy@linaro.org>
      38aa76a8
    • Jon Medhurst's avatar
      Update Makefiles to get proper dependency checking working. · 6d55d109
      Jon Medhurst authored
      
      
      This change requires all platforms to now specify a list of source files
      rather than object files.
      
      New source files should preferably be specified by using the path as
      well and we should add this in the future for all files so we can remove
      use of vpath. This is desirable because vpath hides issues like the fact
      that BL2 currently pulls in a BL1 file bl1/aarch64/early_exceptions.S
      and if in the future we added bl2/aarch64/early_exceptions.S then it's
      likely only one of the two version would be used for both bootloaders.
      
      This change also removes the 'dump' build target and simply gets
      bootloaders to always generate a dump file. At the same time the -x
      option is added so the section headers and symbols table are listed.
      
      Fixes ARM-software/tf-issues#11
      
      Change-Id: Ie38f7be76fed95756c8576cf3f3ea3b7015a18dc
      Signed-off-by: default avatarJon Medhurst <tixy@linaro.org>
      6d55d109
    • Jon Medhurst's avatar
      Fix implementation and users of gicd_set_ipriorityr() · cf6eeb8a
      Jon Medhurst authored
      
      
      Make gicd_set_ipriorityr() actually write to the priority register.
      
      Also correct callers of this function which want the highest priority
      to use the value zero as this is the highest priority value according
      to the ARM Generic Interrupt Controller Architecture Specification.
      
      To make this easier to get right, we introduce defines for the lowest
      and highest priorities for secure and non-secure interrupts.
      
      Fixes ARM-software/tf-issues#21
      Signed-off-by: default avatarJon Medhurst <tixy@linaro.org>
      cf6eeb8a
  5. 26 Feb, 2014 1 commit
    • Sandrine Bailleux's avatar
      fvp: Initialise UART earlier · 20d284c0
      Sandrine Bailleux authored
      The UART used to be initialised in bl1_platform_setup(). This is too
      late because there are some calls to the assert() macro, which needs
      to print some messages on the console, before that.
      
      This patch moves the UART initialisation code to
      bl1_early_platform_setup().
      
      Fixes ARM-software/tf-issues#49
      
      Change-Id: I98c83a803866372806d2a9c2e1ed80f2ef5b3bcc
      20d284c0
  6. 20 Feb, 2014 8 commits
    • Jeenu Viswambharan's avatar
      Report recoverable errors as warnings · 08c28d53
      Jeenu Viswambharan authored
      At present many recoverable failures are reported as errors. This patch
      modifies all such failures to be reported as warnings instead.
      
      Change-Id: I5141653c82498defcada9b90fdf7498ba496b2f2
      08c28d53
    • Achin Gupta's avatar
      Add Test Secure Payload Dispatcher (TSPD) service · 375f538a
      Achin Gupta authored
      
      
      This patch adds the TSPD service which is responsible for managing
      communication between the non-secure state and the Test Secure Payload
      (TSP) executing in S-EL1.
      
      The TSPD does the following:
      
      1. Determines the location of the TSP (BL3-2) image and passes control
         to it for initialization. This is done by exporting the 'bl32_init()'
         function.
      
      2. Receives a structure containing the various entry points into the TSP
         image as a response to being initialized. The TSPD uses this
         information to determine how the TSP should be entered depending on
         the type of operation.
      
      3. Implements a synchronous mechanism for entering into and returning
         from the TSP image. This mechanism saves the current C runtime
         context on top of the current stack and jumps to the TSP through an
         ERET instruction. The TSP issues an SMC to indicate completion of the
         previous request. The TSPD restores the saved C runtime context and
         resumes TSP execution.
      
      This patch also introduces a Make variable 'SPD' to choose the specific
      SPD to include in the build. By default, no SPDs are included in the
      build.
      
      Change-Id: I124da5695cdc510999b859a1bf007f4d049e04f3
      Co-authored-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      375f538a
    • Achin Gupta's avatar
      Add Test Secure Payload (BL3-2) image · 7c88f3f6
      Achin Gupta authored
      
      
      This patch adds a simple TSP as the BL3-2 image. The secure payload
      executes in S-EL1. It paves the way for the addition of the TSP
      dispatcher runtime service to BL3-1. The TSP and the dispatcher service
      will serve as an example of the runtime firmware's ability to toggle
      execution between the non-secure and secure states in response to SMC
      request from the non-secure state.  The TSP will be replaced by a
      Trusted OS in a real system.
      
      The TSP also exports a set of handlers which should be called in
      response to a PSCI power management event e.g a cpu being suspended or
      turned off. For now it runs out of Secure DRAM on the ARM FVP port and
      will be moved to Secure SRAM later. The default translation table setup
      code assumes that the caller is executing out of secure SRAM. Hence the
      TSP exports its own translation table setup function.
      
      The TSP only services Fast SMCs, is non-reentrant and non-interruptible.
      It does arithmetic operations on two sets of four operands, one set
      supplied by the non-secure client, and the other supplied by the TSP
      dispatcher in EL3. It returns the result according to the Secure Monitor
      Calling convention standard.
      
      This TSP has two functional entry points:
      
      - An initial, one-time entry point through which the TSP is initialized
        and prepares for receiving further requests from secure
        monitor/dispatcher
      
      - A fast SMC service entry point through which the TSP dispatcher
        requests secure services on behalf of the non-secure client
      
      Change-Id: I24377df53399307e2560a025eb2c82ce98ab3931
      Co-authored-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      7c88f3f6
    • Achin Gupta's avatar
      Specify address of UART device to use as a console · 8aa0cd43
      Achin Gupta authored
      This patch adds the ability to specify the base address of a UART
      device for initialising the console. This allows a boot loader stage
      to use a different UART device from UART0 (default) for the console.
      
      Change-Id: Ie60b927389ae26085cfc90d22a564ff83ba62955
      8aa0cd43
    • Achin Gupta's avatar
      Factor out translation table setup in ARM FVP port · a0cd989d
      Achin Gupta authored
      This patch factors out the ARM FVP specific code to create MMU
      translation tables so that it is possible for a boot loader stage to
      create a different set of tables instead of using the default ones.
      The default translation tables are created with the assumption that
      the calling boot loader stage executes out of secure SRAM. This might
      not be true for the BL3_2 stage in the future.
      
      A boot loader stage can define the `fill_xlation_tables()` function as
      per its requirements. It returns a reference to the level 1
      translation table which is used by the common platform code to setup
      the TTBR_EL3.
      
      This patch is a temporary solution before a larger rework of
      translation table creation logic is introduced.
      
      Change-Id: I09a075d5da16822ee32a411a9dbe284718fb4ff6
      a0cd989d
    • Achin Gupta's avatar
      Add support for BL3-2 in BL3-1 · 35ca3511
      Achin Gupta authored
      This patch adds the following support to the BL3-1 stage:
      
      1. BL3-1 allows runtime services to specify and determine the security
         state of the next image after BL3-1. This has been done by adding
         the `bl31_set_next_image_type()` & `bl31_get_next_image_type()`
         apis. The default security state is non-secure. The platform api
         `bl31_get_next_image_info()` has been modified to let the platform
         decide which is the next image in the desired security state.
      
      2. BL3-1 exports the `bl31_prepare_next_image_entry()` function to
         program entry into the target security state. It uses the apis
         introduced in 1. to do so.
      
      3. BL3-1 reads the information populated by BL2 about the BL3-2 image
         into its internal data structures.
      
      4. BL3-1 introduces a weakly defined reference `bl32_init()` to allow
         initialisation of a BL3-2 image. A runtime service like the Secure
         payload dispatcher will define this function if present.
      
      Change-Id: Icc46dcdb9e475ce6575dd3f9a5dc7a48a83d21d1
      35ca3511
    • Achin Gupta's avatar
      Add support for BL3-2 in BL2 · a3050ed5
      Achin Gupta authored
      
      
      This patch adds support for loading a BL3-2 image in BL2. In case a
      BL3-2 image is found, it also passes information to BL3-1 about where it
      is located and the extents of memory available to it. Information about
      memory extents is populated by platform specific code.
      
      The documentation has also been updated to reflect the above changes.
      
      Change-Id: I526b2efb80babebab1318f2b02e319a86d6758b0
      Co-authored-by: default avatarJeenu Viswambharan <jeenu.viswambharan@arm.com>
      a3050ed5
    • Achin Gupta's avatar
      Rework BL2 to BL3-1 hand over interface · e4d084ea
      Achin Gupta authored
      This patch reworks BL2 to BL3-1 hand over interface by introducing a
      composite structure (bl31_args) that holds the superset of information
      that needs to be passed from BL2 to BL3-1.
      
        - The extents of secure memory available to BL3-1
        - The extents of memory available to BL3-2 (not yet implemented) and
          BL3-3
        - Information to execute BL3-2 (not yet implemented) and BL3-3 images
      
      This patch also introduces a new platform API (bl2_get_bl31_args_ptr)
      that needs to be implemented by the platform code to export reference to
      bl31_args structure which has been allocated in platform-defined memory.
      
      The platform will initialize the extents of memory available to BL3-3
      during early platform setup in bl31_args structure. This obviates the
      need for bl2_get_ns_mem_layout platform API.
      
      BL2 calls the bl2_get_bl31_args_ptr function to get a reference to
      bl31_args structure. It uses the 'bl33_meminfo' field of this structure
      to load the BL3-3 image. It sets the entry point information for the
      BL3-3 image in the 'bl33_image_info' field of this structure. The
      reference to this structure is passed to the BL3-1 image.
      
      Also fixes issue ARM-software/tf-issues#25
      
      Change-Id: Ic36426196dd5ebf89e60ff42643bed01b3500517
      e4d084ea
  7. 17 Feb, 2014 6 commits
    • Achin Gupta's avatar
      Increase coherent stack sizes · ca823d2c
      Achin Gupta authored
      This patch increases coherent stack size for both debug and release
      builds in order to accommodate stack-heavy printf() and extended EL3
      functionality
      
      Change-Id: I30ef30530a01517a97e63d703873374828c09f20
      ca823d2c
    • Jeenu Viswambharan's avatar
      Move translation tables into separate section · 74cbb839
      Jeenu Viswambharan authored
      This patch moves the translation tables into their own section. This
      saves space that would otherwise have been lost in padding due to page
      table alignment constraints. The BL31 and BL32 bases have been
      consequently adjusted.
      
      Change-Id: Ibd65ae8a5ce4c4ea9a71a794c95bbff40dc63e65
      74cbb839
    • Harry Liebel's avatar
      Add Firmware Image Package (FIP) driver · 561cd33e
      Harry Liebel authored
      The Firmware Image Package (FIP) driver allows for data to be loaded
      from a FIP on platform storage. The FVP supports loading bootloader
      images from a FIP located in NOR FLASH.
      
      The implemented FVP policy states that bootloader images will be
      loaded from a FIP in NOR FLASH if available and fall back to loading
      individual images from semi-hosting.
      
      NOTE:
      - BL3-3(e.g. UEFI) is loaded into DRAM and needs to be configured
        to run from the BL33_BASE address. This is currently set to
        DRAM_BASE+128MB for the FVP.
      
      Change-Id: I2e4821748e3376b5f9e467cf3ec09509e43579a0
      561cd33e
    • James Morrissey's avatar
      Implement load_image in terms of IO abstraction · 9d72b4ea
      James Morrissey authored
      The modified implementation uses the IO abstraction rather than
      making direct semi-hosting calls.  The semi-hosting driver is now
      registered for the FVP platform during initialisation of each boot
      stage where it is used.  Additionally, the FVP platform includes a
      straightforward implementation of 'plat_get_image_source' which
      provides a generic means for the 'load_image' function to determine
      how to access the image data.
      
      Change-Id: Ia34457b471dbee990c7b3c79de7aee4ceea51aa6
      9d72b4ea
    • James Morrissey's avatar
      Add IO abstraction framework · f2f9bb5e
      James Morrissey authored
      This is intended primarily for use as a storage abstraction.
      It allows operations such as image-loading to be implemented
      in a platform-independent fashion.  Each platform registers
      a set of IO drivers during initialisation.  The platform must
      also provide a function that will return a device and a specifier
      that can be used to access specified content.
      
      Clients of the API will primarily use device and entity handles.
      The term "entity" is deliberately vague, to allow for different
      representations of content accessed using different types of
      specifier, but will often be interpreted as a "file" where the
      specifier will normally be its path.
      
      This commit builds, but is intended to be paired with a sample
      implementation of "load_image" using a semi-hosting driver on FVP.
      
      Change-Id: Id3b52f1c0eb9ce76b44b99fc6b6460803668cc86
      f2f9bb5e
    • James Morrissey's avatar
      Fix asserts appearing in release builds · 40a6f647
      James Morrissey authored
      Also fix warnings generated in release builds when assert code
      is absent.
      
      Change-Id: I45b9173d3888f9e93e98eb5b4fdc06727ba5cbf4
      40a6f647
  8. 23 Jan, 2014 1 commit
    • Achin Gupta's avatar
      fvp: clear a pending cluster power off request · b2187ab9
      Achin Gupta authored
      The last CPU in a cluster is responsible for issuing the cluster power
      down request to the FVP power controller. If another CPU in this
      cluster wakes up before the last CPU enters WFI then the cluster power
      down request remains pending. If this request is not cancelled and the
      newly woken up CPU enters a simple WFI later, the power controller
      powers the cluster down. This leads to unpredictable behaviour.
      
      This patch fixes this issue by ensuring that the first CPU to wake up
      in a cluster writes its MPIDR to the power controller's PPONR. This
      cancels any pending cluster power down request.
      
      Change-Id: I7e787adfd6c9a0bd7308390e3309d46f35c01086
      b2187ab9
  9. 20 Jan, 2014 3 commits
  10. 17 Jan, 2014 4 commits
    • Jeenu Viswambharan's avatar
      Change comments in assembler files to help ctags · 3a4cae05
      Jeenu Viswambharan authored
      Ctags seem to have a problem with generating tags for assembler symbols
      when a comment immediately follows an assembly label.
      
      This patch inserts a single space character between the label
      definition and the following comments to help ctags.
      
      The patch is generated by the command:
      
        git ls-files -- \*.S | xargs sed -i 's/^\([^:]\+\):;/\1: ;/1'
      
      Change-Id: If7a3c9d0f51207ea033cc8b8e1b34acaa0926475
      3a4cae05
    • Harry Liebel's avatar
      Probe for GICv3 re-distributors on core bring-up · eaec590e
      Harry Liebel authored
      The GICv3 distributor can have more ports than CPUs are available in
      the system. Probe all re-distributors and use the matching affinity
      levels as specified by each core and re-distributor to decide which
      re-distributor to use with which CPU core.
      
      If a core cannot be matched with a re-distributor, the core panics and
      is placed in an endless loop.
      
      Change-Id: Ie393cfe07c7449a2383959e3c968664882e18afc
      eaec590e
    • Dan Handley's avatar
      Update year in copyright text to 2014 · e83b0cad
      Dan Handley authored
      Change-Id: Ic7fb61aabae1d515b9e6baf3dd003807ff42da60
      e83b0cad
    • Ian Spray's avatar
      Move GIC setup to a separate file · 8468739c
      Ian Spray authored
      GIC setup code which used to be in bl31_plat_setup.c is now in fvp_gic.c
      to simplify future changes to other bootloader stages.  This patch moves
      code from bl31_plat_setup.c to fvp_gic.c, simplifies the include file
      list for bl31_plat_setup.c, moves GIC declarations from the bl31.h header
      file into the platform.h, and reworks files according to coding style
      guide.
      
      Change-Id: I48d82a4ba33e7114dcc88f9ca98767a06cf8f417
      8468739c
  11. 12 Dec, 2013 2 commits
    • Sandrine Bailleux's avatar
      Make BL31's ns_entry_info a single-cpu area · 93ca221c
      Sandrine Bailleux authored
      ns_entry_info used to be a per-cpu array.  This is a waste of space
      because it is only accessed by the primary CPU on the cold boot path.
      This patch reduces ns_entry_info to a single-cpu area.
      
      Change-Id: I647c70c4e76069560f1aaad37a1d5910f56fba4c
      93ca221c
    • Sandrine Bailleux's avatar
      Remove useless copies of meminfo structures · ee12f6f7
      Sandrine Bailleux authored
      Platform setup code has to reserve some memory for storing the
      memory layout information.  It is populated in early platform setup
      code.
      
      blx_get_sec_mem_layout() functions used to return a copy of this
      structure.  This patch modifies blx_get_sec_mem_layout() functions
      so that they now directly return a pointer to their memory layout
      structure.  It ensures that the memory layout returned by
      blx_get_sec_mem_layout() is always up-to-date and also avoids a
      useless copy of the meminfo structure.
      
      Also rename blx_get_sec_mem_layout() to blx_plat_sec_mem_layout()
      to make it clear those functions are platform specific.
      
      Change-Id: Ic7a6f9d6b6236b14865ab48a9f5eff545ce56551
      ee12f6f7
  12. 05 Dec, 2013 3 commits
    • Achin Gupta's avatar
      remove check on non-secure entrypoint parameter · 2d94d4a0
      Achin Gupta authored
      In fvp_affinst_on/suspend, the non-secure entrypoint is always
      expected to lie in the DRAM. This check will not be valid if
      non-secure code executes directly out of flash e.g. a baremetal
      test. This patch removes this check.
      
      Change-Id: I0436e1138fc394aae8ff1ea59ebe38b46a440b61
      2d94d4a0
    • Achin Gupta's avatar
      psci: fix error due to a non zero context id · c8afc789
      Achin Gupta authored
      In the previous psci implementation, the psci_afflvl_power_on_finish()
      function would run into an error condition if the value of the context
      id parameter in the cpu_on and cpu_suspend psci calls was != 0. The
      parameter was being restored as the return value of the affinity level
      0 finisher function. A non zero context id would be treated as an
      error condition. This would prevent successful wake up of the cpu from
      a power down state. Also, the contents of the general purpose
      registers were not being cleared upon return to the non-secure world
      after a cpu power up. This could potentially allow the non-secure
      world to view secure data.
      
      This patch ensures that all general purpose registers are set to ~0
      prior to the final eret that drops the execution to the non-secure
      world. The context id is used to initialize the general purpose
      register x0 prior to re-entry into the non-secure world and is no
      longer restored as a function return value. A platform helper
      (platform_get_stack()) has been introduced to facilitate this change.
      
      Change-Id: I2454911ffd75705d6aa8609a5d250d9b26fa097c
      c8afc789
    • Achin Gupta's avatar
      clear wakeup enable bit upon resuming from suspend · b127cdb8
      Achin Gupta authored
      The FVP specific code that gets called after a cpu has been physically
      powered on after having been turned off or suspended earlier does not
      clear the PWRC.PWKUPR.WEN bit. Not doing so causes problems if: a cpu
      is suspended, woken from suspend, powered down through a cpu_off call
      & receives a spurious interrupt. Since the WEN bit is not cleared
      after the cpu woke up from suspend, the spurious wakeup will power the
      cpu on. Since the cpu_off call clears the jump address in the mailbox
      this spurious wakeup will cause the cpu to crash.
      
      This patch fixes this issue by clearing the WEN bit whenever a cpu is
      powered up.
      
      Change-Id: Ic91f5dffe1ed01d76bc7fc807acf0ecd3e38ce5b
      b127cdb8