1. 25 11月, 2013 5 次提交
  2. 18 11月, 2013 24 次提交
  3. 16 11月, 2013 6 次提交
  4. 15 11月, 2013 2 次提交
    • G
      malta: use unmapped flash base address · 10473d04
      Gabor Juhos 提交于
      The physical base address of the NOR flash is 0x1e000000
      on the Malta boards. The hardware also maps the first 4MiB
      of the flash into the 0x1fc00000-0x1fffffff range.
      
      Currently, U-Boot uses the mapped address to access the
      flash, which does not work in recent qemu versions.
      
      Since commit a427338b222b43197c2776cbc996936df0302f51
      (mips_malta: correct reading MIPS revision at 0x1fc00010)
      writing to the mapped address space causes a CPU exception.
      Due to the exception, U-Boot hangs during boot when it tries
      to detect the CFI flash chip.
      
      Use the correct physical address for the MALTA_FLASH_BASE
      constant to fix the problem. In order to avoid relocation
      problems, also update the CONFIG_SYS_{TEXT,MONITOR}_BASE
      constants.
      
      The change makes it possible to start U-Boot on a Malta
      board emulated with Qemu 1.6.1 and 1.7.0-rc0. It also
      works on older versions (tested with 1.1.1, 1.2.2, 1.4.2,
      1.5.3).
      Signed-off-by: NGabor Juhos <juhosg@openwrt.org>
      Signed-off-by: NDaniel Schwierzeck <daniel.schwierzeck@gmail.com>
      Cc: Paul Burton <paul.burton@imgtec.com>
      10473d04
    • T
      Merge branch 'master' of git://git.denx.de/u-boot-mpc85xx · c3ebb8c3
      Tom Rini 提交于
      c3ebb8c3
  5. 14 11月, 2013 3 次提交
    • L
      powerpc/85xx: fix broken cpu "clock-frequency" property · 51abee64
      Laurentiu TUDOR 提交于
      When indexing freqProcessor[] we use the first
      value in the cpu's "reg" property, which on
      new e6500 cores IDs the threads.
      But freqProcessor[] should be indexed with a
      core index so, when fixing "the clock-frequency"
      cpu node property, access the freqProcessor[]
      with the core index derived from the "reg' property.
      If we don't do this, last half of the "cpu" nodes
      will have broken "clock-frequency" values.
      Signed-off-by: NLaurentiu Tudor <Laurentiu.Tudor@freescale.com>
      Cc: York Sun <yorksun@freescale.com>
      51abee64
    • L
      powerpc/t4240: fix per pci endpoint liodn offsets · 8f9fe660
      Laurentiu TUDOR 提交于
      Update the code that builds the pci endpoint liodn
      offset list so that it doesn't overlap with other
      liodns and doesn't generate negative offsets like:
      
        fsl,liodn-offset-list = <0 0xffffffcd 0xffffffcf
                                   0xffffffd1 0xffffffd3
                                   0xffffffd5 0xffffffd7
                                   0xffffffd9 0xffffffdb>;
      
      The update consists in adding a parameter to the
      function that builds the list to specify the base
      liodn.
      On PCI v2.4 use the old base = 256 and, on PCI 3.0
      where some of the PCIE liodns are larger than 256,
      use a base = 1024. The version check is based on
      the PCI controller's version register.
      Signed-off-by: NLaurentiu Tudor <Laurentiu.Tudor@freescale.com>
      Cc: Scott Wood <scottwood@freescale.com>
      Cc: York Sun <yorksun@freescale.com>
      8f9fe660
    • L
      powerpc/t4240: set pcie liodn in the correct register · b4125a23
      Laurentiu TUDOR 提交于
      The liodn for the T4240's PCIE controller is no longer set
      through a register in the guts register block but with one
      in the PCIE register block itself.
      Use the already existing SET_PCI_LIODN_BASE macro that puts
      the liodn in the correct register.
      Signed-off-by: NLaurentiu Tudor <Laurentiu.Tudor@freescale.com>
      Cc: Scott Wood <scottwood@freescale.com>
      Cc: York Sun <yorksun@freescale.com>
      b4125a23