1. 05 Apr, 2019 1 commit
    • Grzegorz Jaszczyk's avatar
      plat: marvell: sip: make sure that comphy init will use correct address · b143fa74
      Grzegorz Jaszczyk authored
      
      
      The argument passed via x1 is used as a base address for comphy related
      routines. Nevertheless validation of this address wasn't good enough and
      allowed some non comphy related addresses to slip over.
      
      To overcome this issue make sure that the address passed via SMC points
      to valid CP range and allow to proceed comphy initializations only with
      correct comphy offset.
      
      This could be fixed in a different way e.g. by passing CP id from the
      caller, but since this API is already used with various Linux, U-Boot
      and UEFI versions it can't be changed.
      Signed-off-by: default avatarGrzegorz Jaszczyk <jaz@semihalf.com>
      Change-Id: Ia74dbc36efcfbefc4a102d31191e6af5808c4a82
      b143fa74
  2. 03 Apr, 2019 3 commits
  3. 02 Apr, 2019 7 commits
  4. 01 Apr, 2019 2 commits
  5. 26 Mar, 2019 1 commit
  6. 21 Mar, 2019 2 commits
  7. 14 Mar, 2019 4 commits
    • Heiko Stuebner's avatar
      rockchip: add an fdt parsing stub for platform param · 7029e806
      Heiko Stuebner authored
      
      
      The Rockchip ATF platform can be entered from both Coreboot and U-Boot.
      While Coreboot does submit the list of linked parameter structs as
      platform param, upstream u-boot actually always provides a pointer
      to a devicetree as parameter.
      This results in current ATF not running at all when started from U-Boot.
      
      To fix this, add a stub that checks if the parameter is a fdt so we
      can at least boot and not get stuck. Later on we can extend this with
      actual parsing of information from the devicetree.
      Signed-off-by: default avatarHeiko Stuebner <heiko@sntech.de>
      Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
      7029e806
    • Louis Mayencourt's avatar
      fvp: Increase the size of the stack for FVP · 01aa5247
      Louis Mayencourt authored
      
      
      When RECLAIM_INIT_CODE is 1, the stack is used to contain the .text.init
      section. This is by default enable on FVP. Due to the size increase of
      the .text.init section, the stack had to be adjusted contain it.
      
      Change-Id: Ia392341970fb86c0426cf2229b1a7295453e2e32
      Signed-off-by: default avatarLouis Mayencourt <louis.mayencourt@arm.com>
      01aa5247
    • Sandrine Bailleux's avatar
      Put Pointer Authentication key value in BSS section · 47102b35
      Sandrine Bailleux authored
      
      
      The dummy implementation of the plat_init_apiakey() platform API uses
      an internal 128-bit buffer to store the initial key value used for
      Pointer Authentication support.
      
      The intent - as stated in the file comments - was for this buffer to
      be write-protected by the MMU. Initialization of the buffer would be
      performed before enabling the MMU, thus bypassing write protection
      checks.
      
      However, the key buffer ended up into its own read-write section by
      mistake due to a typo on the section name ('rodata.apiakey' instead of
      '.rodata.apiakey', note the leading dot). As a result, the linker
      script was not pulling it into the .rodata output section.
      
      One way to address this issue could have been to fix the section
      name. However, this approach does not work well for BL1. Being the
      first image in the boot flow, it typically is sitting in real ROM
      so we don't have the capacity to update the key buffer at any time.
      
      The dummy implementation of plat_init_apiakey() provided at the moment
      is just there to demonstrate the Pointer Authentication feature in
      action. Proper key management and key generation would have to be a
      lot more careful on a production system.
      
      Therefore, the approach chosen here to leave the key buffer in
      writable memory but move it to the BSS section. This does mean that
      the key buffer could be maliciously updated for intalling unintended
      keys on the warm boot path but at the feature is only at an
      experimental stage right now, this is deemed acceptable.
      
      Change-Id: I121ccf35fe7bc86c73275a4586b32d4bc14698d6
      Signed-off-by: default avatarSandrine Bailleux <sandrine.bailleux@arm.com>
      47102b35
    • John Tsichritzis's avatar
      Introduce preliminary support for Neoverse Zeus · a4546e80
      John Tsichritzis authored
      
      
      Change-Id: If56d1e200a31bd716726d7fdc1cc0ae8a63ba3ee
      Signed-off-by: default avatarJohn Tsichritzis <john.tsichritzis@arm.com>
      a4546e80
  8. 13 Mar, 2019 5 commits
  9. 12 Mar, 2019 1 commit
  10. 08 Mar, 2019 4 commits
  11. 07 Mar, 2019 3 commits
  12. 04 Mar, 2019 1 commit
  13. 01 Mar, 2019 2 commits
    • Varun Wadekar's avatar
      Tegra: dummy support for the io_storage backend · 8d56e24b
      Varun Wadekar authored
      
      
      This patch provides dummy macros and platform files to compile
      the io_storage driver backend. This patch is necessary to
      remove the "--unresolved=el3_panic" linker flag from Tegra's
      makefiles and allow us to revert this workaround, previously
      suggested by the ARM toolchain team.
      
      The "--unresolved=el3_panic" flag actually was a big hammer that
      allowed Tegra platforms to work with armlink previously but it
      masks legit errors with the code as well.
      
      Change-Id: I0421d35657823215229f84231896b84167f90548
      Signed-off-by: default avatarVarun Wadekar <vwadekar@nvidia.com>
      8d56e24b
    • Anson Huang's avatar
      imx: make sure GIC redistributor is awake before initialization · e655fefc
      Anson Huang authored
      
      
      GICR_WAKER.ProcessorSleep can only be set to zero when:
      — GICR_WAKER.Sleep bit[0] == 0.
      — GICR_WAKER.Quiescent bit[31] == 0.
      
      On some platforms, when system reboot with GIC in sleep
      mode but with power ON, such as on NXP's i.MX8QM, Linux
      kernel enters suspend but could be requested to reboot,
      and GIC is in sleep mode and it is inside a power domain
      which is ON in this scenario, when CPU reset, the GIC
      driver trys to set CORE's redistributor interface to awake,
      with GICR_WAKER.Sleep bit[0] and GICR_WAKER.Quiescent bit[31]
      both set, the ProcessorSleep bit[1] will never be clear
      and cause system hang.
      
      This patch makes sure GICR_WAKER.Sleep bit[0] and
      GICR_WAKER.Quiescent bit[31] are both zeor before clearing
      ProcessorSleep bit[1].
      Signed-off-by: default avatarAnson Huang <Anson.Huang@nxp.com>
      e655fefc
  14. 28 Feb, 2019 4 commits