T1024RDB is a Freescale Reference Design Board that hosts the T1024 SoC.

T1024RDB board Overview
-----------------------
- T1024 SoC integrating two 64-bit e5500 cores up to 1.4GHz
- CoreNet fabric supporting coherent and noncoherent transactions with
  prioritization and bandwidth allocation
- 32-/64-bit DDR3L SDRAM memory controller with ECC and interleaving support
- Accelerator: DPAA components consist of FMan, BMan, QMan, DCE and SEC
- Ethernet interfaces:
  - Two 10M/100M/1G RGMII ports on-board
  - one 10Gbps XFI interface
- PCIe: Three PCIe controllers: one PCIe Slot and two Mini-PCIe connectors.
- SerDes: 4 lanes up to 10.3125GHz
- IFC: 128MB NOR Flash, 512MB NAND Flash and CPLD
- eSPI: 64MB N25Q512 SPI flash.
- Deep Sleep power implementaion (wakeup from GPIO/Timer/Ethernet/USB)
- USB: Two  Type-A USB2.0 ports with internal PHY
- eSDHC: Support SD, SDHC, SDXC and MMC/eMMC
- I2C: Four I2C controllers
- UART: Two UART serial ports
Signed-off-by: NShengzhou Liu <Shengzhou.Liu@freescale.com>
[York Sun: Fix ft_board_setup() type, fix MAINTAINERS for SECURE_BOOT
	   Fix Kconfig by adding SUPPORT_SPL]
Reviewed-by: NYork Sun <yorksun@freescale.com>

T1024QDS is a high-performance computing evaluation, development and
test platform for T1024 QorIQ Power Architecture processor.

T1024QDS board Overview
-----------------------
- T1024 SoC integrating two 64-bit e5500 cores up to 1.4GHz
- CoreNet fabric supporting coherent and noncoherent transactions with
  prioritization and bandwidth allocation
- 32-/64-bit DDR3L/DDR4 SDRAM memory controller with ECC and interleaving support
- Accelerator: DPAA components consist of FMan, BMan, QMan, DCE and SEC
- Ethernet interfaces:
  - Two 10M/100M/1G RGMII ports on-board
  - Three 1G/2.5Gbps SGMII ports
  - Four 1Gbps QSGMII ports
  - one 10Gbps XFI or 10Base-KR interface
- SerDes: 4 lanes up to 10.3125GHz Supporting SGMII/QSGMII, XFI, PCIe, SATA and Aurora
- PCIe: Three PCI Express controllers with five PCIe slots.
- IFC: 128MB NOR Flash, 2GB NAND Flash, PromJet debug port and Qixis FPGA
- Video: DIU supports video up to 1280x1024x32 bpp.
  - Chrontel CH7201 for HDMI connection.
  - TI DS90C387R for direct LCD connection.
  - Raw (not encoded) video connector for testing or other encoders.
- QUICC Engine block
  - 32-bit RISC controller for flexible support of the communications peripherals
  - Serial DMA channel for receive and transmit on all serial channels
  - Two universal communication controllers, supporting TDM, HDLC, and UART
- Deep Sleep power implementaion (wakeup from GPIO/Timer/Ethernet/USB)
- eSPI: Three SPI flash devices.
- SATA: one SATA 2.O.
- USB: Two USB2.0 ports with internal PHY (one Type-A + one micro Type mini-AB)
- eSDHC: Support SD, SDHC, SDXC and MMC/eMMC.
- I2C: Four I2C controllers.
- UART: Two UART on board.
Signed-off-by: NShengzhou Liu <Shengzhou.Liu@freescale.com>
[York Sun: Fix ft_board_setup() type, fix MAINTAINERS for SECURE_BOOT
	   Fix Kconfig by adding SUPPORT_SPL]
Reviewed-by: NYork Sun <yorksun@freescale.com>

T2080 v1.1 requires different MEM_PLL_RAT from previous v1.0,
and also update core frequency to 1.8GHz for v1.1.
We reserve the support for T2080 v1.0 and enable v1.1 by default.
Signed-off-by: NShengzhou Liu <Shengzhou.Liu@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

In B4860QDS board SerDes2 lanes EFGH either go to SFP or AMC riser card
slot2 so either DTSEC3/DTSEC4 or TGEC1/TGEC2 should be accessible. This
Patch enables DTSEC3/DTSEC4 or TGEC1/TGEC2 on bases of user specified
string fsl_b4860_serdes2:sfp_amc=amc or fsl_b4860_serdes2:sfp_amc=sfp
respectively in hwconfig.
Signed-off-by: NSuresh Gupta <suresh.gupta@freescale.com>
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

XFI does not work stable on current board, it's due to heat sink issue,
to make it work stable the board needs additional heat sink, enable two
XFI lanes only. Right now we do not have such an erratum for the issue,
so use a define CONFIG_SYS_FSL_B4860QDS_XFI_ERR to identify it.
The workaround will only be used in XFI protocols and only if the
hwconfig indicates that XFI is prefered.

A new VSC3308 config function is used instead of re-use the original
function, to avoid making the function complex and ugly.
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

SerDes2 lanes EFGH either go to SFP or AMC riser card slot2.
By default AMC will be configured even if no hwconfig is specified.

To enable XFI via SFP use the below hwconfig:
	fsl_b4860_serdes2:sfp_amc=sfp
Signed-off-by: NSuresh Gupta <suresh.gupta@freescale.com>
Signed-off-by: NPoonam Aggrwal <poonam.aggrwal@ffeescale.com>
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Since xfi has no phy, we delete the property "phy-handle" and use
a "fixed-link" property for a xfi port.
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

We need following changes to make xfi work on B4:
1. set cross-point switch VSC3308 to use sfp config when running xfi;
2. add 10G interface check for xfi;
3. set phy address for xfi so the 10G ports can be registered by mdio;
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Add the support of newly added LC VCO SerDes protocols
for configuration of IDT and VSC crossbar
Signed-off-by: NShaveta Leekha <shaveta@freescale.com>
Signed-off-by: NPoonam Aggrwal <poonam.aggrwal@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

- Enable SGMII support for 0x8d Serdes 2 protocol.
    - Correct Phy address for DTSECx for 0x8d/0xb2 Serdes 2 protocol.
    - Updated debug statement
    - Add Alternate LC VCO protocols(0x8d-->0x8c, 0xb2-->0xb1)
    - Rename onboard PHY address defines for more readability
    - Add these new Defines in B4860QDS.h file
Signed-off-by: NShaveta Leekha <shaveta@freescale.com>
Signed-off-by: NPoonam Aggrwal <poonam.aggrwal@freescale.com>
Signed-off-by: NSuresh Gupta <suresh.gupta@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

B4860, B4440, B4420 and B4220 have MAPLE, so enable law creation
for them only. Remove static LAW creation for MAPLE.
Signed-off-by: NShaveta Leekha <shaveta@freescale.com>
Signed-off-by: NSandeep Singh <Sandeep@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

The fuse status register provides the values from on-chip
voltage ID efuses programmed at the factory.
These values define the voltage requirements for
the chip. u-boot reads FUSESR and translates the values
into the appropriate commands to set the voltage output
value of an external voltage regulator.
Signed-off-by: NYing Zhang <b40530@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Workaround of Errata A-008044 was implemented without errata number and it is
enabled by default. Errata A-008044 is only valid for T1040 Rev 1.0.

So put errata number and make it conditional.
Signed-off-by: NPrabhakar Kushwaha <prabhakar@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

When device is configured to load RCW from NAND flash IFC_A[16:31] are driven
low after RCW loading. Hence Devices connected on IFC_CS[1:7] and using
IFC_A[16:31] lines are not accessible.

Workaround is already in-place.
Put the errata number to adhere errata handling framework.
Signed-off-by: NPrabhakar Kushwaha <prabhakar@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Adjust timing for dual-rank UDIMM, verified on M3CQ-8GHS3C0E for speed of
1066, 1333, 1600, 1866MT/s. The 1866 timing is copied to 2133 timing in
case such DIMM comes available.

Also update single-rank 1866 timing. Enable interactive debugging as well.
Signed-off-by: NYork Sun <yorksun@freescale.com>
CC: Shengzhou Liu <Shengzhou.Liu@freescale.com>

EC1 pins in RCW can be selected as RGMII1, GPIO3, CAN1/2, FTM1 or
SAI1/2. There is a bug that EC3 RGMII could not work when selecting EC1
as other functionality except RGMII. The workaround is to select
ge2_clk125 for eTSEC clock muxing in register SCFG_ETSECCMCR.
Signed-off-by: NAlison Wang <alison.wang@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

SystemID information could be read through I2C1 from EEPROM
on LS1021ATWR board.

As LS1 is a little-endian processor, getting the version ID by
be32_to_cpu() is wrong. Fix it by using e.version directly.
This change will be compatible for both ARM and PowerPC.

As there is an errata that I2C1 could not work in SD boot,
reading EEPROM through I2C1 is disabled too in SD boot.
Signed-off-by: NAlison Wang <alison.wang@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

SCFG_SCFGREVCR is SCFG bit reverse register. This register
must be written with 0xFFFFFFFF before writing to any other
SCFG register. Then other SCFG register could be written in
big-endian mode.

Address: 157_0000h base + 200h offset = 157_0200h
Bit   0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15|16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
W/R                                   SCFGREV
Reset 0 0 0 0 0 0 0 0 0 0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0
0-31
SCFGREV SCFG Bit Reverse Control Filed
32'h 0000_0000 - No bit reverse is applied
32'h FFFF_FFFF - Bit reverse is applied; so 31:0 will be stored/read as
0:31

This patch removes the bit reversing for SCFG registers in
u-boot. It will be implemented through PBI commands in RCW
.pbi
write 0x570200, 0xffffffff
.end
So other SCFG register could be written in big-endian mode
in u-boot or kernel directly.
Signed-off-by: NAlison Wang <alison.wang@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Disable the snoop for slave interface 0, 1 and 2
to avoid the interleaving on the CCI400 BUS.
Signed-off-by: NJason Jin <Jason.Jin@freescale.com>
Signed-off-by: NMinghuan Lian <Minghuan.Lian@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

U-Boot has never cared about the type when we get max/min of two
values, but Linux Kernel does.  This commit gets min, max, min3, max3
macros synced with the kernel introducing type checks.

Many of references of those macros must be fixed to suppress warnings.
We have two options:
 - Use min, max, min3, max3 only when the arguments have the same type
   (or add casts to the arguments)
 - Use min_t/max_t instead with the appropriate type for the first
   argument
Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: NPavel Machek <pavel@denx.de>
Acked-by: NLukasz Majewski <l.majewski@samsung.com>
Tested-by: NLukasz Majewski <l.majewski@samsung.com>
[trini: Fixup arch/blackfin/lib/string.c]
Signed-off-by: NTom Rini <trini@ti.com>

When an invalid USDHC port is passed we should return -EINVAL instead of 0.

Also, return the error immediately on fsl_esdhc_initialize() failure.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

When an invalid USDHC port is passed we should return -EINVAL instead of 0.

Also, return the error immediately on fsl_esdhc_initialize() failure.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

When an invalid USDHC port is passed we should return -EINVAL instead of 0.

Also, return the error immediately on fsl_esdhc_initialize() failure.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

When an invalid USDHC port is passed we should return -EINVAL instead of 0.

Also, return the error immediately on fsl_esdhc_initialize() failure.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

When an invalid USDHC port is passed we should return -EINVAL instead of 0.

Also, return the error immediately on fsl_esdhc_initialize() failure.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

Use these new functions where possible. They default to a value of 2 so we
cannot use them in some places where we need a default value of 1.
Signed-off-by: NSimon Glass <sjg@chromium.org>
Reviewed-by: NTom Rini <trini@ti.com>

This function can fail if the device tree runs out of space. Rather than
silently booting with an incomplete device tree, allow the failure to be
detected.

Unfortunately this involves changing a lot of places in the code. I have
not changed behvaiour to return an error where one is not currently
returned, to avoid unexpected breakage.

Eventually it would be nice to allow boards to register functions to be
called to update the device tree. This would avoid all the many functions
to do this. However it's not clear yet if this should be done using driver
model or with a linker list. This work is left for later.
Signed-off-by: NSimon Glass <sjg@chromium.org>
Acked-by: NAnatolij Gustschin <agust@denx.de>

Add the pfuze100 initialization in power_init_board for imx6q/dl
sabreauto board.
Signed-off-by: NYe.Li <B37916@freescale.com>

Modify the pfuze init for mx6sxsabresd to use the shared
"pfuze_common_init" function. And move this initialization to
power_init_board.
Signed-off-by: NYe.Li <B37916@freescale.com>

Modify the pfuze init for mx6sabresd to use the shared
"pfuze_common_init" function. And move this initialization to
power_init_board.
Signed-off-by: NYe.Li <B37916@freescale.com>

Since the Pfuze initializations are similar on various mx6 SABRE
boards. Factorize the initialization to a common function in file
board/freescale/common/pfuze.c. So that all SABRE boards BSP can
share the function.
Signed-off-by: NYe.Li <B37916@freescale.com>

Add full support for USDHC2, USDHC3, USDHC4 on mx6sx sabresd board.
The default boot socket is USDHC4, so the MMC environment device and
mmcdev variable are set to this device.
Signed-off-by: NYe.Li <B37916@freescale.com>

In U-boot it is preferred to access the register via structure pointer, so
convert it such style.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

Let's add mx6sabresd_spl_defconfig entry into MAINTAINERS, so that we avoid
getting a warning that the mx6sabresd_spl is not maintained.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

We can simply the return the value from enable_fec_anatop_clock() to make the
code smaller and simpler.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

We can simply the return the value from enable_fec_anatop_clock() to make the
code smaller and simpler.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

Make clear that current SPL code only supports the mx6q variant.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

When an invalid USDHC port is passed we should return -EINVAL instead of 0.

Also, return the error immediately on fsl_esdhc_initialize() failure.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

mx6sabreauto boards come with 32 MiB of parallel NOR flash.

Add support for it:

U-Boot 2015.01-rc1-18107-g1543636-dirty (Nov 14 2014 - 11:11:04)

CPU:   Freescale i.MX6Q rev1.2 at 792 MHz
Reset cause: POR
Board: MX6Q-Sabreauto revA
I2C:   ready
DRAM:  2 GiB
Flash: 32 MiB
NAND:  0 MiB

Due to pin conflict with I2C3, only define configure I2C3 IOMUX when flash is
not used.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>

Many boards use a minimal .cfg file in the SPL case.

Introduce spl_sd.cfg so that we can reuse it.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Acked-by: NStefano Babic <sbabic@denx.de>