This reverts commit 0e57968a21.
The short version of the original commit is that some i2c devices cannot
be probed via read as they NAK the first cycle, so try and probe via a
write that we abort before it writes to the device. This however is not
allowed by the TRM for any of these parts. The section on I2C_CON
(table 17-35 I2C_CON for am/dm37x for example) says you must not change
the register while STT has been set. On these parts, the unpredictable
behavior that the chip exhibits is not problematic. On OMAP4 however it
results in the chip being in a bad state:
Panda # i2c probe
Valid chip addresses: 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12
13 14 15 16 17 18 19 1A 1B 1C 1D 1E 1F 20 21 22 23 24 25 26 27 28 29 2A
2B 2C 2D 2E 2F 30 31 32 33 34 35 36 37 38 39 3A 3B 3C 3D 3E 3F 40 41 42
43 44 45 46 47 48 49 4A 4B 4C 4D 4E 4F 50 51 52 53 54 55 56 57 58 59 5A
5B 5C 5D 5E 5F 60 61 62 63 64 65 66 67 68 69 6A 6B 6C 6D 6E 6F 70 71 72
73 74 75 76 77 78 79 7A 7B 7C 7D 7E 7F
Panda # i2c md 50 0
timed out in wait_for_pin: I2C_STAT=0
I2C read: I/O error
Error reading the chip.
We must revert the original behavior to bring probe back into line with
the TRM.
Cc: Nick Thompson <nick.thompson@ge.com>
Cc: Heiko Schocher <hs@denx.de>
Signed-off-by: Tom Rini <trini@ti.com>
Acked-by: Heiko Schocher <hs@denx.de>
Previously, only the low 5 bits (NCH) were being transfered
from DDRVTPR to DDRVTPIOCR, the bits 5-9 where zeroed.
VTP_RECAL should be bit 15, not 18.
The only mainline board affected by this change is davinci_sonata.
The other Davinci boards define CONFIG_SKIP_LOWLEVEL_INIT.
However, if the program that loads u-boot on these boards
copied the code from u-boot, they will need fixed as well.
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Please get tested by acks before applying, where tested by
means an overnight memory test.
Thanks
Troy
get_ram_size checks the given memory range for valid ram,
but expects the size of memory to be aligned to the power
of 2. In case of OMAP5 evm board the memory available is
2GB - 16MB(used for TRAP section) = 2032MB.
So always ensure that the size of memory used for testing is
aligned to the power of 2.
Signed-off-by: R Sricharan <r.sricharan@ti.com>
The unmapped entries in tiler space are set with
values 0xFF. So creating a DMM section of
size 16MB at 0xFF000000 with ADDRSPACE set to 0x2.
This way all the unmapped entry accesses to tiler
will be trapped by the EMIF and a error response
is sent to the L3 interconnect. L3 errors are
inturn reported to MPU.
Note that here the tiler trap section is overlapping
with the actual ddr physical space and we lose 16MB
out of the total 2GB.
Signed-off-by: R Sricharan <r.sricharan@ti.com>
The DMM sections can be overlapping with each other, with
sections 3 to 0 having the highest to lowest priority in that
order. There could also be a section that is used trap the
unmapped Tiler entries and this trap section could be
overlapping with the actual sdram area.
So take care of the above scenarios while calculating the
size of the actual ram.
Signed-off-by: R Sricharan <r.sricharan@ti.com>
- change gpio pin settings:
- gpio pin 6[13] (PLC reset) default value low
- gpio pin 6[0] (TPM reset) default value low
- 4 new GPIO pins
pin i/o name
- 3[9] input Board Type
- 2[7] input HW-ID0
- 2[6] input HW-ID1
- 2[3] input HW-ID2
- read board type and hw id from gpio pins on the enbw_cmc board,
and use board type for setting up different gpio pin settings.
- do not pass "davinci_mmc.use_dma=0" to linux, as MMC now
works with DMA.
- update logbuf support:
store post word in RTC scratch register
- add support for configuring KSZ8864RMN switch through
a config file on u-boot startup. For more infos see:
doc/README.switch_config
Signed-off-by: Heiko Schocher <hs@denx.de>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Tom Rini <tom.rini@gmail.com>
Cc: Christian Riesch <christian.riesch@omicron.at>
Cc: Sandeep Paulraj <s-paulraj@ti.com>
Fix the .dts file USB unit addresses not to duplicate each-other.
Fix the board name string to indicate the vendor is Compulab not NVIDIA.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Acked-by: Igor Grinberg <grinberg@compulab.co.il>
Walk the BIT and BCT to find the ODMDATA word in the
CustomerData field and put it into Scratch20 reg for
use by kernel, etc.
Built all Tegra builds OK; Booted on Seaboard and saw
ODMDATA in PMC scratch20 was the same as the value in my
burn-u-boot.sh file (0x300D8011). NOTE: All flash utilities
will have to specify the odmdata (nvflash --odmdata n) on
the command line or via a cfg file, or built in to their
BCT.
Signed-off-by: Tom Warren <twarren@nvidia.com>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
The SPI hardware on Seaboard is too broken to use; it is muxed with the
console UART and requires evil interactions between the SPI and UART
drivers to work even partially. The current code in U-Boot is not
sufficient to make this work correctly; auto boot is aborted due to
corruption in the UART RX channel interrupting it.
Instead, move the environment to eMMC, at the end of the second boot
sector. This should not conflict with any other eMMC usage, irrespective
of whether the board boots from SPI, NAND, or eMMC: if U-Boot is stored
in eMMC, it will be stored well below this location. The kernel only
uses the general area of the eMMC once booted, not the boot sectors.
Boards that are derivatives of Seaboard don't have the muxing issue,
and should/could have a separate U-Boot configuration file that does
enable SPI if desired.
Alternatively, the environment could be stored in NAND flash, but we
currently have no driver for that controller.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Store the environment in eMMC, at the end of the second boot sector.
This should not conflict with any other eMMC usage: U-Boot is stored
well below this location, and the kernel only uses the general area
of the eMMC once booted, not the boot sectors.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
In anticipation of Tegra3 support, continue removing/renaming
Tegra2-specific files. No functional changes (yet).
Updated copyrights to 2012.
Signed-off-by: Tom Warren <twarren@nvidia.com>
In anticipation of Tegra3 support, continue removing/renaming
Tegra2-specific files. No functional changes (yet).
Updated copyrights to 2012.
Signed-off-by: Tom Warren <twarren@nvidia.com>
In anticipation of Tegra3 support, start removing/renaming
Tegra2-specific files. No functional changes (yet).
Also updated copyright to 2012.
Signed-off-by: Tom Warren <twarren@nvidia.com>
Store the environment in eMMC, at the end of the second boot sector.
This should not conflict with any other eMMC usage: U-Boot is stored
well below this location, and the kernel only uses the general area
of the eMMC once booted, not the boot sectors.
Note: This assumes the user plugged the standard 8MB MoviNAND card into
J29/HSMMC/POP. If they didn't, the boot sector layout may be different.
However, use of that particular card is standard practice as far as I
know.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Store the environment in eMMC, at the end of the second boot sector.
This should not conflict with any other eMMC usage: U-Boot is stored
well below this location, and the kernel only uses the general area
of the eMMC once booted, not the boot sectors.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This allows MMC drivers to perform cache flusing on the bufffers
without issue.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Cc: Andy Fleming <afleming@gmail.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Add a device tree for Ventana; the Seaboard file no longer represents
the HW present on Ventana.
Enable USB on Ventana.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
The SMSC95xx series may exist either directly on a main board, or as a USB
to Ethernet dongle. However, dongles containing these chips are very rare.
Hence, remove this config option, except on Harmony where such a chip is
actually present on the board.
The asix option remains, since it's a popular chip, and I actively use a
dongle containing this.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Acked-by: Igor Grinberg <grinberg@compulab.co.il>
Override -march setting for tegra to -march=armv4t for files that are
necessary for low level init on tegra.
The recent change to use -march=armv7-a for armv7 caused a regression
on tegra because tegra starts boot on a arm7tdmi processor before
transferring control to the cortex-a9. While still executing on the
arm7tdmi there are calls to getenv_ulong() and memset() that cause an
illegal instruction exception if compiled for armv7.
Signed-off-by: Allen Martin <amartin@nvidia.com>
Tested-by: Stephen Warren <swarren@wwwdotorg.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Correct this warning seen by Albert:
ap20.c:44:18: warning: array subscript is above array bounds
There is a subtle bug here which currently causes no errors, but might
in future if people use PCI or the 32KHz clock. So take the opportunity
to correct the logic now.
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
... to enable USB host support, which enables Ethernet support.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
... to enable USB host support, which enables Ethernet support.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This is based on top of:
tegra: add alternate UART1 funcmux entry
tegra: add UART1 on GPU funcmux entry
v2: remove enum change
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
CC: Stephen Warren <swarren@wwwdotorg.org>
CC: Tom Warren <twarren@nvidia.com>
CC: Marek Vasut <marex@denx.de>
Signed-off-by: Tom Warren <twarren@nvidia.com>
CONFIG_CMD_PING/NFS aren't required for Whistler to boot.
Add some comments.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This implements a useful bootcmd for Tegra. The boot order is:
* If USB enabled, USB storage
* Internal MMC (SD card or eMMC)
* If networking is enabled, BOOTP/TFTP
When booting from USB or MMC, the boot script is assumed to be in
partition 1 (although this may be overridden via the rootpart variable),
both ext2 and FAT filesystems are supported, the boot script may exist
in either / or /boot, and the boot script may be named boot.scr.uimg or
boot.scr.
When booting over the network, it is assumed that boot.scr.uimg exists
on the TFTP server. There is less flexibility here since those setting
up network booting are expected to need less hand-holding.
In all cases, it is expected that the initial file loaded is a U-Boot
image containing a script that will load the kernel, load any required
initrd, load any required DTB, and finally bootm the kernel.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
console isn't used by anything, and the kernel should be set appropriately
by whatever script is booting the kernel, not imposed by the bootloader.
mem might be useful, but the current value is pretty bogus, since it
includes nvmem options that make no sense for an upstream kernel, and
equally should not be required for any downstream kernel. Either way, this
is also best left to the kernel boot script.
smpflag isn't used by anything, and again was probably intended to be a
kernel command-line option better set by the kernel boot script.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Whistler is a highly configurable Tegra evaluation and development board.
This change adds support for the following specific configuration:
E1120 motherboard
E1108 CPU board
E1116 PMU board
The motherboard configuration switches are set as follows:
SW1=0 SW2=0 SW3=5
S1/S2/S3/S4 all on, except S3 7/8 are off.
Other combinations of daugher boards may work to varying degrees, but will
likely require some SW adjustment.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
(In at least some configurations) Whistler uses UART1 on pingroups
UAA, UAB.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Simon Glass's proposal to fix this on Seaboard was NAK'd, so I
removed his NS16550 references and added a small delay before
SPI/UART muxing. Tested on my Seaboard with large SPI reads/writes
and saw no corruption (crc's matched) and no spurious comm chars.
Signed-off-by: Tom Warren <twarren@nvidia.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Jimmy Zhang <jimmzhang@nvidia.com>
Erratum NMG_CPU_A011 applies to P4080 rev 1.0, 2.0, fixed in rev 3.0.
It also applies to P3041 rev 1.0, 1.1, P2041 rev 1.0, 1.1. It shares the
same workaround as erratum CPU22. Rearrange registers usage in assembly
code to avoid accidental overwriting.
Signed-off-by: York Sun <yorksun@freescale.com>
We don't care E bit of SVR in most cases. Clear E bit for SVR_SOC_VER().
This will simplify the coding. Use IS_E_PROCESSOR() to identify SoC with
encryption. Remove all _E entries from SVR list and CPU list.
Signed-off-by: York Sun <yorksun@freescale.com>
addrmap_phys_to_virt() converts a physical address (phys_addr_t) to a
virtual address, so it should return a pointer instead of an unsigned long.
Its counterpart, addrmap_virt_to_phys(), takes a pointer, so now they're
orthogonal.
The only caller of addrmap_phys_to_virt() converts the return value to
a pointer anyway.
Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
Remove some unused default environment variables (memctl_intlv_ctl,
perf_mode, diuregs, dium, and diuerr), update 'tftpflash' variable,
and add videobootargs as a Linux command line variable (so that we can
easily pass video= to the kernel).
Signed-off-by: Timur Tabi <timur@freescale.com>
Remove some unused macros and remove all #undef macros.
The RTL8139 network adapter is not shipped with the board nor commonly
used, so don't define it by default. The E1000 is still defined.
Add 57,600 baud as an option. For some reason, this baud rate is missing
from many boards.
Signed-off-by: Timur Tabi <timur@freescale.com>
Tom Rini
Troy Kisky
SRICHARAN R
Heiko Schocher
Thomas Weber
Stephen Warren
Tom Warren
Tom Warren
amartin@nvidia.com
Simon Glass
Lucas Stach
York Sun
Timur Tabi
Shengzhou Liu