Alexey Brodkin [Wed, 8 Jun 2016 05:24:54 +0000 (08:24 +0300)]
arc: Update data accessors with use of memory barriers
Memory barriers are proven to be a requirement for both compiler and
real hardware to properly serialize access to critical data.
For example if CPU or data bus it uses may do reordering of data
accesses absence of memory barriers might easily lead to very subtle and
hard to debug data corruptions.
This implementation was heavily borrowed from up to date Linux kernel.
Alexey Brodkin [Wed, 8 Jun 2016 05:19:33 +0000 (08:19 +0300)]
board: axs10x: Flush entire cache after programming reset vector
Now when we have support of IOC (IO-Coherency block) cahce operations
on regions are tuned to not be dummy stubs if IOC was found and enabled
in the core. That makes flush_dcache_range() useless for our purposes
here. And since we do need to flush modified reset vector to at least L2
cache (AKA SLC) so other cores will see it via its L1 instruction cache
we're using always functional flush_dcache_all() here.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com> Cc: Marek Vasut <marex@denx.de>
Alexey Brodkin [Wed, 8 Jun 2016 05:04:03 +0000 (08:04 +0300)]
arc/cache: Flush & invalidate all caches right before enabling IOC
According to ARC HS databook it is required to flush and disable
caches prior programming IOC registers. Otherwise ongoing coherent
memory operations may not observe the coherency protocols as
expected.
But since in ARC HS v2.1 there's no way to disable SLC (AKA L2 cache)
we're doing our best flushing and invalidating it.
Alexey Brodkin [Wed, 8 Jun 2016 04:57:19 +0000 (07:57 +0300)]
arc/cache: really do invalidate_dcache_all() even if IOC exists
invalidate_dcache_all() could be used in different use-cases
and what is especially important most of those cases won't be
related to DMAed data to or from peripherals, i.e. we'll be doing
invalidation of data used purely by CPU cores.
Given that IOC engine only snoops data that goes through DMA
we need to care ourselves about data used only by CPU cores
and so remove dependency on IOC from invalidate_dcache_all()
and always do real invalidation.
Heiko Schocher [Tue, 7 Jun 2016 06:55:45 +0000 (08:55 +0200)]
siemens,am33x: add draco etamin board
In the draco CPU board family, etamin is a new variant
with bigger flash and more RAM. Due to new flash that
uses larger pages (4K) some changes are necessary because
it impacts the MTD partition layout and the ubi mount
parameters.
Signed-off-by: Samuel Egli <samuel.egli@siemens.com> Signed-off-by: Heiko Schocher <hs@denx.de>
[trini: Move BOOTDELAY into defconfig, just always be 3 now] Signed-off-by: Tom Rini <trini@konsulko.com>
Heiko Schocher [Tue, 7 Jun 2016 06:55:44 +0000 (08:55 +0200)]
dfu, nand, ubi: fix erasing after write finish
writting to ubi nand partitions need after write ends an erase
of the remaining sectors. This fail, if dfu write size was not
a multiple of erasesize, example log:
Heiko Schocher [Tue, 7 Jun 2016 06:55:41 +0000 (08:55 +0200)]
nand: add nand mtd concat support
add for nand devices mtd concat support. Generic MTD concat
support is already ported to mainline, and used in the cfi_mtd
driver. This patch adds it similiar for nand devices.
Heiko Schocher [Tue, 7 Jun 2016 06:31:24 +0000 (08:31 +0200)]
doc, spl, am335x: update am335x-network
on the shc board we see when booting in net boot mode,
that the ROM bootloader sends "AM335x ROM" as
vendor-class-identifier. U-Boots doc says "DM814x ROM".
So, add the info to the doc, that there is also
"AM335x ROM" possible.
Signed-off-by: Heiko Schocher <hs@denx.de> Reviewed-by: Tom Rini <trini@konsulko.com>
Heiko Schocher [Tue, 7 Jun 2016 06:31:18 +0000 (08:31 +0200)]
am335x: add some missing CM_CLKMODE_DPLL_SSC macros
add missing CM_CLKMODE_DPLL_SSC_ACK_MASK,
CM_CLKMODE_DPLL_SSC_DOWNSPREAD_MASK and
CM_CLKMODE_DPLL_SSC_TYPE_MASK
defines. Used for enabling spread spectrum.
Signed-off-by: Heiko Schocher <hs@denx.de> Reviewed-by: Tom Rini <trini@konsulko.com>
Heiko Schocher [Tue, 7 Jun 2016 06:31:15 +0000 (08:31 +0200)]
tests: py: disable main_signon check for printenv cmd
if CONFIG_VERSION_VARIABLE is set, the U-Boot environment
contains a "vers" variable with the current U-Boot version
string. If now "printenv" is called, test/py fails as it
detects the main_sign string, which is in this case correct.
So check only the main_sign as an error, if CONFIG_VERSION_VARIABLE
is not set.
Simon Glass [Thu, 9 Jun 2016 02:55:15 +0000 (20:55 -0600)]
tiny-printf: Correct the snprintf() implementation
This current code passes the variable arguments list to sprintf(). This is
not correct. Fix it by calling _vprintf() directly.
This makes firefly-rk3288 boot again.
Fixes: abeb272 ("tiny-printf: Support snprintf()") Reviewed-by: Stefan Roese <sr@denx.de> Acked-by: Marek Vasut <marex@denx.de> Signed-off-by: Simon Glass <sjg@chromium.org>
Masahiro Yamada [Wed, 8 Jun 2016 09:02:32 +0000 (18:02 +0900)]
ARM: uniphier: insert dsb barrier to ensure visibility of store
I noticed secondary CPUs sometimes fail to wake up, and the root
cause is that the sev instruction wakes up slave CPUs before the
preceding the register write is observed by them.
The read-back of the accessed register does not guarantee the order.
In order to ensure the order between the register write and the sev
instruction, a dsb instruction should be executed prior to the sev.
Masahiro Yamada [Tue, 7 Jun 2016 12:03:44 +0000 (21:03 +0900)]
ARM: uniphier: do not overwrite fdt_file environment
This code auto-detects the best-match FDT file name, but it should
respect the user's choice if "fdt_file" environment is found in a
saved set of environments.
Masahiro Yamada [Sat, 4 Jun 2016 13:39:08 +0000 (22:39 +0900)]
ARM: uniphier: support eMMC boot for PH1-LD11 and PH1-LD20
The Boot ROM on PH1-LD11/LD20 exports built-in APIs to load images
from an eMMC device. They are useful to reduce the memory footprint
of SPL, rather than compiling the whole MMC framework.
Andre Przywara [Tue, 31 May 2016 17:45:06 +0000 (10:45 -0700)]
arm/arm64: implement a boot header capability
Some SPL loaders (like Allwinner's boot0, and Broadcom's boot0)
require a header before the actual U-Boot binary to both check its
validity and to find other data to load. Sometimes this header may
only be a few bytes of information, and sometimes this might simply
be space that needs to be reserved for a post-processing tool.
Introduce a config option to allow assembler preprocessor commands
to be inserted into the code at the appropriate location; typical
assembler preprocessor commands might be:
.space 1000
.word 0x12345678
Signed-off-by: Andre Przywara <andre.przywara@arm.com> Signed-off-by: Steve Rae <srae@broadcom.com>
Commit Notes:
Please note that the current code:
start.S (arm64) and
vectors.S (arm)
already jumps over some portion of data already, so this option basically
just increases the size of this region (and the resulting binary).
For use with Allwinner's boot0 blob there is a tool called boot0img[1],
which fills the header to allow booting A64 based boards.
For the Pine64 we need a 1536 byte header (including the branch
instruction) at the moment, so we add this to the defconfig.
[1] https://github.com/apritzel/pine64/tree/master/tools
END Reviewed-by: Tom Rini <trini@konsulko.com>
Chris Packham [Tue, 31 May 2016 08:30:59 +0000 (20:30 +1200)]
lib: make strmhz available in SPL
When setting up a DDR controller it is useful to be able to display
frequencies in a readable form. Make the strmhz() function available in
SPL builds provided there is full vsprintf available.
Reviewed-by: Tony O'Brien <tony.obrien@alliedtelesis.co.nz> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Chris Packham <judge.packham@gmail.com>
Alexander Graf [Fri, 27 May 2016 10:25:03 +0000 (12:25 +0200)]
efi_loader: Don't allocate from memory holes
When a payload calls our memory allocator with the exact address hint, we
happily allocate memory from completely unpopulated regions. Payloads however
expect this to only succeed if they would be allocating from free conventional
memory.
This patch makes the logic behind those checks a bit more obvious and ensures
that we always allocate from known good free conventional memory regions if we
want to allocate ram.
Reported-by: Jonathan Gray <jsg@jsg.id.au> Signed-off-by: Alexander Graf <agraf@suse.de>
Alexander Graf [Thu, 2 Jun 2016 09:38:27 +0000 (11:38 +0200)]
efi_loader: Move to normal debug infrastructure
We introduced special "DEBUG_EFI" defines when the efi loader
support was new. After giving it a bit of thought, turns out
we really didn't have to - the normal #define DEBUG infrastructure
works well enough for efi loader as well.
So this patch switches to the common debug() and #define DEBUG
way of printing debug information.
Alexander Graf [Fri, 20 May 2016 21:28:23 +0000 (23:28 +0200)]
efi_loader: Add exit support
Some times you may want to exit an EFI payload again, for example
to default boot into a PXE installation and decide that you would
rather want to boot from the local disk instead.
This patch adds exit functionality to the EFI implementation, allowing
EFI payloads to exit.
Alexander Graf [Fri, 20 May 2016 21:28:22 +0000 (23:28 +0200)]
arm: Introduce setjmp/longjmp
To quit an EFI application we will need logic to jump to the caller
of a function without returning from the function we called into,
so we need setjmp/longjmp functionality.
This patch introduces a trivial implementation of these that I
verified works on armv7, thumb2 and aarch64.
Ed Swarthout [Wed, 1 Jun 2016 13:11:24 +0000 (08:11 -0500)]
dm: scsi: if_typename should be scsi
Fixes:
=> ext2ls scsi 0:1
** Bad device scsi 0:1 **
for boards which use the scsi legacy driver (such as ls1043ardb).
Signed-off-by: Ed Swarthout <Ed.Swarthout@nxp.com> Tested-by: George McCollister <george.mccollister@gmail.com> Acked-by: Simon Glass <sjg@chromium.org>
Michal Simek [Fri, 22 Apr 2016 12:28:54 +0000 (14:28 +0200)]
ARM64: zynqmp: Add support for standard distro boot commands
Nand and QSPI are not defined now but this will be extended.
Based on selected bootmode boot_targets are rewritten.
Patch also contains detection if variables are saved. If yes don't
rewrite boot_targets variable.
Also move variable setup to the end of file because SCSI needs to be
defined before others macros are using it.
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Reviewed-by: Alexander Graf <agraf@suse.de>
Alexander Graf [Wed, 1 Jun 2016 20:41:54 +0000 (22:41 +0200)]
ARM64: zynqmp: Remove CONFIG_BOOTP_SERVERIP
When the CONFIG_BOOTP_SERVERIP option is set, we ignore all
dhcp values for the tftp server and use our own serverip and
file name instead.
This is usually not what we want and I doubt it's set for a
good reason on ZynqMP. It definitely hurts if we want to support
uEFI PXE boot on it. So just remove the option for now.
Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Correct the nand ecc initialization code
This fixes the issue of incorrect nand ecc
init if no device is found in ecc_matrix then
it endsup ecc init with junk initialization
instead of the most suited one.
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
When no DTB can be matched successfully to the board that's being used
a list of available FIT-embedded DTBs will be output to the console for
diagnostic purposes. But rather than the contents of the "description"
FDT property a non-existent property was accessed and as a result "NULL"
was output instead of the actual name(s) of the DTB(s). Fix this issue
by using the correct property which is also the exact same property
that's used earlier during the actual board matching process.
Signed-off-by: Andreas Dannenberg <dannenberg@ti.com>
Certain GPIO devices have the capability to switch their GPIOs into
open-drain mode, that is, instead of actively driving the output
(Push-pull output), the pin is connected to the collector (for a NPN
transistor) or the drain (for a MOSFET) of a transistor, respectively.
The pin then either forms an open circuit or a connection to ground,
depending on the state of the transistor.
This patch adds functions to the GPIO uclass to switch GPIOs to
open-drain mode on devices that support it.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: York Sun <york.sun@nxp.com>
By default, each bank is assumed to have 32 GPIOs, but the ngpios
setting is honored, so the number of GPIOs for each bank in configurable
to match the actual GPIO count of the SoC (e.g. the 32/32/23 banks of
the P1022 SoC).
The usual functions of GPIO drivers (setting input/output mode and output
value setting) are supported.
The driver has been tested on MPC85XX, but it is likely that other
PowerQUICC III devices will work as well.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: York Sun <york.sun@nxp.com>
Max Krummenacher [Mon, 30 May 2016 14:28:28 +0000 (16:28 +0200)]
nand: fix nand torture to use changed mtd api
The mtd subsystem deprecated and renamed the direct use of the mtd_info
struct's functionpointers. Instead the corresponding mtd_xxx function
should be used.
Scott Wood [Mon, 30 May 2016 18:57:58 +0000 (13:57 -0500)]
mtd: nand: Sync with Linux v4.6
Updates the NAND code to match Linux v4.6. The previous sync was from
Linux v4.1 in commit d3963721d93fafa.
Note that none of the individual NAND drivers tracked Linux closely
enough to be synced themselves, other than manually applying a few
cross-tree changes.
Signed-off-by: Scott Wood <oss@buserror.net> Tested-by: Heiko Schocher <hs@denx.de>
Scott Wood [Mon, 30 May 2016 18:57:57 +0000 (13:57 -0500)]
mtd: nand: Add page argument to write_page() etc.
This change is part of the Linux 4.6 sync. It is being done before the
main sync patch in order to make it easier to address the issue across
all NAND drivers (many/most of which do not closely track their Linux
counterparts) separately from other merge issues.
Scott Wood [Mon, 30 May 2016 18:57:56 +0000 (13:57 -0500)]
mtd: nand: Add+use mtd_to/from_nand and nand_get/set_controller_data
These functions are part of the Linux 4.6 sync. They are being added
before the main sync patch in order to make it easier to address the
issue across all NAND drivers (many/most of which do not closely track
their Linux counterparts) separately from other merge issues.
Scott Wood [Mon, 30 May 2016 18:57:55 +0000 (13:57 -0500)]
nand: Embed mtd_info in struct nand_chip
nand_info[] is now an array of pointers, with the actual mtd_info
instance embedded in struct nand_chip.
This is in preparation for syncing the NAND code with Linux 4.6,
which makes the same change to struct nand_chip. It's in a separate
commit due to the large amount of changes required to accommodate the
change to nand_info[].
Scott Wood [Mon, 30 May 2016 18:57:53 +0000 (13:57 -0500)]
mtd: nand: Remove docg4 driver and palmtreo680 flashing tool
Commit ad4f54ea86b ("arm: Remove palmtreo680 board") removed the only
user of the docg4 driver and the palmtreo680 image flashing tool. This
patch removes them.
Signed-off-by: Scott Wood <oss@buserror.net> Cc: Mike Dunn <mikedunn@newsguy.com> Cc: Simon Glass <sjg@chromium.org>
Marek Vasut [Mon, 30 May 2016 18:57:52 +0000 (13:57 -0500)]
mtd: nand: Remove jz4740 driver
This driver is not used by anyone, remove it.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com> Cc: Paul Burton <paul.burton@imgtec.com> Cc: Scott Wood <oss@buserror.net> Acked-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com> Signed-off-by: Scott Wood <oss@buserror.net>
Correct the nand ecc initialization code
This fixes the issue of incorrect nand ecc
init if no device is found in ecc_matrix then
it endsup ecc init with junk initialization
instead of the most suited one.
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com> Tested-by: Michal Simek <michal.simek@xilinx.com>
Masahiro Yamada [Fri, 3 Jun 2016 22:35:03 +0000 (07:35 +0900)]
usb: move CONFIG_USB_XHCI to Kconfig with renaming
Move CONFIG_USB_XHCI to defconfig files for all boards, renaming it
into CONFIG_USB_XHCI_HCD.
As commented in the help of "config USB_XHCI" entry, this has been
a TODO for a long time; now CONFIG_USB_XHCI_HCD and CONFIG_USB_XHCI
have been unified in favor of the former.
Note:
Some boards define CONFIG_USB_XHCI in their headers without
CONFIG_USB, which does not meet the "depends on" in Kconfig.
I added CONFIG_USB=y for those boards when converting.
Otherwise, they would fail to build.
QorIQ LS1012A FREEDOM (LS1012AFRDM) is a high-performance
development platform, with a complete debugging environment.
The LS1012AFRDM board supports the QorIQ LS1012A processor and is
optimized to support the high-bandwidth DDR3L memory and
a full complement of high-speed SerDes ports.
Signed-off-by: Shengzhou Liu <Shengzhou.Liu@nxp.com> Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Signed-off-by: Pratiyush Mohan Srivastava <pratiyush.srivastava@nxp.com> Signed-off-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
QorIQ LS1012A Reference Design System (LS1012ARDB) is a high-performance
development platform, with a complete debugging environment.
The LS1012ARDB board supports the QorIQ LS1012A processor and is
optimized to support the high-bandwidth DDR3L memory and
a full complement of high-speed SerDes ports.
Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Signed-off-by: Pratiyush Mohan Srivastava <pratiyush.srivastava@nxp.com> Signed-off-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
QorIQ LS1012A Development System (LS1012AQDS) is a high-performance
development platform, with a complete debugging environment.
The LS1012AQDS board supports the QorIQ LS1012A processor and is
optimized to support the high-bandwidth DDR3L memory and
a full complement of high-speed SerDes ports.
Signed-off-by: Calvin Johnson <calvin.johnson@nxp.com> Signed-off-by: Pratiyush Mohan Srivastava <pratiyush.srivastava@nxp.com> Signed-off-by: Abhimanyu Saini <abhimanyu.saini@nxp.com> Signed-off-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
projects / u-boot / log