oweals/u-boot.git
8 years agoARM: OMAP3_LOGIC: Update MTD Partition Table
Adam Ford [Sun, 13 Nov 2016 04:02:37 +0000 (22:02 -0600)]
ARM: OMAP3_LOGIC: Update MTD Partition Table

The previous partition table did not support a separate device tree
and the kernel size was limited to 4MB.  This update shows the
location of the device tree (labeled as spl-os) for those who
want to use Falcon Mode or use U-Boot to store the Flattened
Device Tree (FDT) to NAND without appending it to the kernel.

This also grows the kernel to 6MB since 4MB was becomming tight

Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoARM: OMAP3_LOGIC: Remove FIT Support
Adam Ford [Sun, 13 Nov 2016 03:53:31 +0000 (21:53 -0600)]
ARM: OMAP3_LOGIC: Remove FIT Support

Commit ("2cd1ff84037a: OMAP3_LOGIC: Setup defconfig to enable
SPL and NAND booting") accidentally enabled FIT support.

This patch removes the FIT support.

Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoARM: OMAP3_LOGIC: Fix SPL Memory Map for Falcon Mode
Adam Ford [Sun, 13 Nov 2016 03:43:37 +0000 (21:43 -0600)]
ARM: OMAP3_LOGIC: Fix SPL Memory Map for Falcon Mode

The memory map defined in commit ("49c7303f0e52: OMAP3: Enable SPL
on omap3_logic) was used by a copy-paste of another board without
fully understanding how the map works in Falcon mode.  This patch
undoes the customization and uses the default SPL Memory Map
for OMAP3.

When building the uImage, set LOADADDR=0x82000000 and Falcon
mode should properly load.

Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agorpi: passthrough of the firmware provided FDT blob
Cédric Schieli [Fri, 11 Nov 2016 10:59:07 +0000 (11:59 +0100)]
rpi: passthrough of the firmware provided FDT blob

Raspberry firmware used to pass a FDT blob at a fixed address (0x100),
but this is not true anymore. The address now depends on both the
memory size and the blob size [1].

If one wants to passthrough this FDT blob to the kernel, the most
reliable way is to save its address from the r2/x0 register in the
U-Boot entry point and expose it in a environment variable for
further processing.

This patch just does this:
- save the provided address in the global variable fw_dtb_pointer
- expose it in ${fdt_addr} if it points to a a valid FDT blob

There are many different ways to use it. One can, for example, use
the following script which will extract from the tree the command
line built by the firmware, then hand over the blob to a previously
loaded kernel:

fdt addr ${fdt_addr}
fdt get value bootargs /chosen bootargs
bootz ${kernel_addr_r} - ${fdt_addr}

Alternatively, users relying on sysboot/pxe can simply omit any FDT
statement in their extlinux.conf file, U-Boot will automagically pick
${fdt_addr} and pass it to the kernel.

[1] https://www.raspberrypi.org/forums//viewtopic.php?f=107&t=134018

Signed-off-by: Cédric Schieli <cschieli@gmail.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
8 years agoarm: add save_boot_params for ARM1176
Cédric Schieli [Fri, 11 Nov 2016 10:59:06 +0000 (11:59 +0100)]
arm: add save_boot_params for ARM1176

Implement a hook to allow boards to save boot-time CPU state for later
use. When U-Boot is chain-loaded by another bootloader, CPU registers may
contain useful information such as system configuration information. This
feature mirrors the equivalent ARMv7 feature.

Signed-off-by: Cédric Schieli <cschieli@gmail.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
8 years agoimage: Combine image_sig_algo with image_sign_info
Andrew Duda [Tue, 8 Nov 2016 18:53:41 +0000 (18:53 +0000)]
image: Combine image_sig_algo with image_sign_info

Remove the need to explicitly add SHA/RSA pairings. Invalid SHA/RSA
pairings will still fail on verify operations when the hash length is
longer than the key length.

Follow the same naming scheme "checksum,crytpo" without explicitly
defining the string.

Indirectly adds support for "sha1,rsa4096" signing/verification.

Signed-off-by: Andrew Duda <aduda@meraki.com>
Signed-off-by: aduda <aduda@meraki.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
8 years agoimage: Add crypto_algo struct for RSA info
Andrew Duda [Tue, 8 Nov 2016 18:53:41 +0000 (18:53 +0000)]
image: Add crypto_algo struct for RSA info

Cut down on the repetition of algorithm information by defining separate
checksum and crypto structs. image_sig_algos are now simply pairs of
unique checksum and crypto algos.

Signed-off-by: Andrew Duda <aduda@meraki.com>
Signed-off-by: aduda <aduda@meraki.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
8 years agorsa: Verify RSA padding programatically
Andrew Duda [Tue, 8 Nov 2016 18:53:40 +0000 (18:53 +0000)]
rsa: Verify RSA padding programatically

Padding verification was done against static SHA/RSA pair arrays which
take up a lot of static memory, are mostly 0xff, and cannot be reused
for additional SHA/RSA pairings. The padding can be easily computed
according to PKCS#1v2.1 as:

  EM = 0x00 || 0x01 || PS || 0x00 || T

where PS is (emLen - tLen - 3) octets of 0xff and T is DER encoding
of the hash.

Store DER prefix in checksum_algo and create rsa_verify_padding
function to handle verification of a message for any SHA/RSA pairing.

Signed-off-by: Andrew Duda <aduda@meraki.com>
Signed-off-by: aduda <aduda@meraki.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
8 years agorsa: cosmetic: rename pad_len to key_len
Andrew Duda [Tue, 8 Nov 2016 18:53:39 +0000 (18:53 +0000)]
rsa: cosmetic: rename pad_len to key_len

checksum_algo's pad_len field isn't actually used to store the length of
the padding but the total length of the RSA key (msg_len + pad_len)

Signed-off-by: Andrew Duda <aduda@meraki.com>
Signed-off-by: aduda <aduda@meraki.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
8 years agoTI: Remove CONFIG_OMAP_COMMON in favor of CONFIG_ARCH_OMAP2
Tom Rini [Tue, 8 Nov 2016 02:34:55 +0000 (21:34 -0500)]
TI: Remove CONFIG_OMAP_COMMON in favor of CONFIG_ARCH_OMAP2

With the move to arch/arm/mach-omap2 there are now very few uses of
CONFIG_OMAP_COMMON and further they can all be replaced with
CONFIG_ARCH_OMAP2, so do so.

Signed-off-by: Tom Rini <trini@konsulko.com>
8 years agoarm: Introduce arch/arm/mach-omap2 for OMAP2 derivative platforms
Tom Rini [Tue, 8 Nov 2016 02:34:54 +0000 (21:34 -0500)]
arm: Introduce arch/arm/mach-omap2 for OMAP2 derivative platforms

This moves what was in arch/arm/cpu/armv7/omap-common in to
arch/arm/mach-omap2 and moves
arch/arm/cpu/armv7/{am33xx,omap3,omap4,omap5} in to arch/arm/mach-omap2
as subdirectories.  All refernces to the former locations are updated to
the current locations.  For the logic to decide what our outputs are,
consolidate the tests into a single config.mk rather than including 4.

Signed-off-by: Tom Rini <trini@konsulko.com>
8 years agoarm: Introduce ARCH_OMAP2
Tom Rini [Tue, 8 Nov 2016 02:34:53 +0000 (21:34 -0500)]
arm: Introduce ARCH_OMAP2

To start consolidating various TI-related code, introduce the ARCH_OMAP2
symbol.  While we have removed omap2-specific boards some time ago,
matching up with the kernel naming here will help overall.

Signed-off-by: Tom Rini <trini@konsulko.com>
8 years agofs-test.sh: Update expected results
Stefan Brüns [Sun, 6 Nov 2016 00:32:15 +0000 (01:32 +0100)]
fs-test.sh: Update expected results

After the latest changes, ext4 no longer has any fails.

Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
8 years agoext4: Allow reading files with non-zero offset, clamp read len
Stefan Brüns [Sun, 6 Nov 2016 17:33:57 +0000 (18:33 +0100)]
ext4: Allow reading files with non-zero offset, clamp read len

Support was already implemented, but not hooked up. This fixes several
fails in the test cases.

Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
8 years agoext4: Fix handling of sparse files
Stefan Brüns [Sat, 5 Nov 2016 21:17:14 +0000 (22:17 +0100)]
ext4: Fix handling of sparse files

A sparse file may have regions not mapped by any extents, at the start
or at the end of the file, or anywhere between, thus not finding a
matching extent region is never an error.

Found by python filesystem tests.

Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
8 years agotest/py: expose config and log as session scoped fixture
Stefan Brüns [Sat, 5 Nov 2016 16:45:32 +0000 (17:45 +0100)]
test/py: expose config and log as session scoped fixture

If a test uses a fixture which is expensive to setup, the fixture can
possibly created with session or module scope. As u_boot_console has
function scope, it can not be used in this case.

Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
8 years agogpio: dwapb: Add support for port B
Phil Edworthy [Thu, 3 Nov 2016 11:05:12 +0000 (11:05 +0000)]
gpio: dwapb: Add support for port B

The IP supports two ports, A and B, each providing up to 32 gpios.
The driver already creates a 2nd gpio bank by reading the 2nd node
from DT, so this is quite a simple change to support the 2nd bank.

Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
Acked-by: Marek Vasut <marex@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
8 years agoarm: dra7xx: Unify Android partition table
Semen Protsenko [Mon, 31 Oct 2016 17:53:46 +0000 (19:53 +0200)]
arm: dra7xx: Unify Android partition table

Make Android partition table the same as for AM57x EVM.

  1. Make "bootloader" partition start from 0x300 sectors offset, so
     DRA7 is bootable in Android mode (see
     CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR option).
  2. Increase "bootloader" partition size, because size of u-boot.img is
     about 632 KiB (when building DT defconfig, with FIT image enabled).
  3. Specify "reserved" partition explicitly, rather than specifying
     "efs" partition start. Reserved area will be used to store U-Boot
     environment on eMMC. It's convenient to have it exposed explicitly
     so we can read/write U-Boot environment.
  4. Keep all Android partitions locations intact, by reducing
     "reserved" partition size. CONFIG_ENV_SIZE is considered.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
8 years agoarm: am57xx: Enable 8-bit eMMC access
Semen Protsenko [Mon, 31 Oct 2016 17:53:42 +0000 (19:53 +0200)]
arm: am57xx: Enable 8-bit eMMC access

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
8 years agoarm: am57xx: Define Android partition table
Semen Protsenko [Mon, 31 Oct 2016 17:53:38 +0000 (19:53 +0200)]
arm: am57xx: Define Android partition table

"fastboot oem format" command reuses "gpt write" command, which in turn
requires correct partitions defined in $partitions variable. This patch
adds such definition of Android partitions for DRA7XX EVM board.

By default $partitions variable contains Linux partition table. In order
to prepare Android environment one can run next commands from U-Boot
shell:

    => env set partitions $partitions_android
    => env save

After those operations one can go to fastboot mode and perform
"fastboot oem format" to create Android partition table.

While at it, enable CONFIG_RANDOM_UUID to spare user from providing
UUIDs for each partition manually.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
8 years agoomap3_beagle: use config_distro_bootcmd
Guillaume GARDET [Tue, 25 Oct 2016 16:50:32 +0000 (18:50 +0200)]
omap3_beagle: use config_distro_bootcmd

Add support for distro_bootcmd on MMC and fall back to prior
behavior if distro_bootcmd fails.

Tested on Beagleboad xM to boot GRUB2 (and then Linux kernel) in EFI mode
from MMC.

Signed-off-by: Guillaume GARDET <guillaume.gardet@free.fr>
Cc: Tom Rini <trini@konsulko.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoconfigs: am57xx: Enable fastboot
Semen Protsenko [Mon, 24 Oct 2016 15:41:13 +0000 (18:41 +0300)]
configs: am57xx: Enable fastboot

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoconfigs: am57xx: Enable download gadget
Semen Protsenko [Mon, 24 Oct 2016 15:41:12 +0000 (18:41 +0300)]
configs: am57xx: Enable download gadget

Enable USB download gadget (needed for fastboot support) and all
dependencies.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoti_omap5_common: Respect USB controller number in fastboot
Semen Protsenko [Mon, 24 Oct 2016 15:41:11 +0000 (18:41 +0300)]
ti_omap5_common: Respect USB controller number in fastboot

On "fastboot reboot-bootloader" we check "dofastboot" variable and do
"fastboot 0" command in U-Boot if it's 1. But there are boards which have
USB controller number other than 0, so it should be respected when
performing "fastboot" command.

This patch reuses CONFIG_FASTBOOT_USB_DEV option toprovide correct USB
controller number to "fastboot" command.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agofastboot: Add CONFIG_FASTBOOT_USB_DEV option
Semen Protsenko [Mon, 24 Oct 2016 15:41:10 +0000 (18:41 +0300)]
fastboot: Add CONFIG_FASTBOOT_USB_DEV option

Some boards (like AM57x EVM) has USB OTG controller other than 0. So in
order to use correct controller number in compiled environment we should
define CONFIG_FASTBOOT_USB_DEV option.

For example, when doing "fastboot reboot-bootloader" we want to enter
fastboot mode automatically. But to do so we need to provide controller
number to "fastboot" command. If this procedure is defined in some config
which is common to bunch of boards, and boards have different USB
controller numbers, we can't just hardcode "fastboot 0" in the
environment. We need to use configurable option, which this patch adds.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoboard: ti: amx3xx: Remove multiple EEPROM reads
Lokesh Vutla [Fri, 14 Oct 2016 05:05:25 +0000 (10:35 +0530)]
board: ti: amx3xx: Remove multiple EEPROM reads

Detect the board very early and avoid reading eeprom multiple times.

Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoARM: AMx3xx: Centralize early clock initialization
Lokesh Vutla [Fri, 14 Oct 2016 05:05:24 +0000 (10:35 +0530)]
ARM: AMx3xx: Centralize early clock initialization

This is similar to Commit 93e6253d11030 ("ARM: OMAP4/5: Centralize
early clock initialization") that was done for OMAP4+, reflecting the same
for AM33xx and AM43xx SoCs to centralize clock initialization.

Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
[trini: Add setup_early_clocks that calls setup_clocks_for_console for
        ti81xx]
Signed-off-by: Tom Rini <trini@konsulko.com>
8 years agoARM: AMx3xx: Allow arch specific code to use early DM
Lokesh Vutla [Fri, 14 Oct 2016 05:05:23 +0000 (10:35 +0530)]
ARM: AMx3xx: Allow arch specific code to use early DM

Early system initialization is being done before initf_dm is being called
in U-Boot. Then system will fail to boot if any of the DM enabled driver
is being called in this system initialization code. So, rearrange the
code a bit so that DM enabled drivers can be called during early system
initialization. This is inspired by commit e850ed82bce8 ("ARM: OMAP4+: Allow
arch specific code to use early DM")

Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agospl: Remove CONFIG_SYS_U_BOOT_MAX_SIZE_SECTORS
Semen Protsenko [Wed, 16 Nov 2016 17:19:06 +0000 (19:19 +0200)]
spl: Remove CONFIG_SYS_U_BOOT_MAX_SIZE_SECTORS

This option isn't used for anything, so get rid of it.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
8 years agospl: Convert CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR to Kconfig
Semen Protsenko [Wed, 16 Nov 2016 17:19:05 +0000 (19:19 +0200)]
spl: Convert CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR to Kconfig

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
[trini: Fix sniper and kc1 migration]
Signed-off-by: Tom Rini <trini@konsulko.com>
8 years agoMerge branch 'master' of http://git.denx.de/u-boot-mmc
Tom Rini [Thu, 17 Nov 2016 16:46:56 +0000 (11:46 -0500)]
Merge branch 'master' of git.denx.de/u-boot-mmc

8 years agoMerge tag 'signed-efi-next' of git://github.com/agraf/u-boot
Tom Rini [Thu, 17 Nov 2016 16:46:45 +0000 (11:46 -0500)]
Merge tag 'signed-efi-next' of git://github.com/agraf/u-boot

Patch queue for efi - 2016-11-17

Highlights this time around:

  - x86 efi_loader support
  - hello world efi test case
  - network device name is now representative
  - terminal output reports modes correctly
  - fix psci reset for ls1043/ls1046
  - fix efi_add_runtime_mmio definition for x86
  - efi_loader support for ls2080

8 years agols2080ardb: Convert to distro boot
Alexander Graf [Thu, 17 Nov 2016 00:03:02 +0000 (01:03 +0100)]
ls2080ardb: Convert to distro boot

Most new systems in U-Boot these days make use of the generic "distro"
framework which allows a user to have U-Boot scan for a bootable OS
on all available media types.

This patch extends the LS2080ARDB board to use that framework if the
hard coded NOR flash location does not contain a bootable image.

Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoarmv8: fsl-layerscape: Add support for efi_loader RTS reset
Alexander Graf [Thu, 17 Nov 2016 00:03:01 +0000 (01:03 +0100)]
armv8: fsl-layerscape: Add support for efi_loader RTS reset

When implementing efi loader support, we can expose runtime services
for payloads. One such service is CPU reset.

This patch implements RTS CPU reset support for layerscape systems.

Signed-off-by: Alexander Graf <agraf@suse.de>
Reviewed-by: York Sun <york.sun@nxp.com>
8 years agoarmv8: ls2080a: Declare spin tables as reserved for efi loader
Alexander Graf [Thu, 17 Nov 2016 00:03:00 +0000 (01:03 +0100)]
armv8: ls2080a: Declare spin tables as reserved for efi loader

The efi loader code has its own memory map, so it needs to be aware where
the spin tables are located, to ensure that no code writes into those
regions.

Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agols2080ardb: Reserve DP-DDR RAM
Alexander Graf [Thu, 17 Nov 2016 00:02:59 +0000 (01:02 +0100)]
ls2080ardb: Reserve DP-DDR RAM

The DP-DDR shouldn't be exposed as conventional memory to an OS, so let's
rather claim it's a reserved region in the EFI memory map

Signed-off-by: Alexander Graf <agraf@suse.de>
Reviewed-by: York Sun <york.sun@nxp.com>
8 years agols2080: Exit dpaa only right before exiting U-Boot
Alexander Graf [Thu, 17 Nov 2016 00:02:57 +0000 (01:02 +0100)]
ls2080: Exit dpaa only right before exiting U-Boot

On ls2080 we have a separate network fabric component which we need to
shut down before we enter Linux (or any other OS). Along with that also
comes configuration of the fabric using a description file.

Today we always stop and configure the fabric in the boot script and
(again) exit it on device tree generation. This works ok for the normal
booti case, but with bootefi the payload we're running may still want to
access the network.

So let's add a new fsl_mc command that defers configuration and stopping
the hardware to when we actually exit U-Boot, so that we can still use
the fabric from an EFI payload.

For existing boot scripts, nothing should change with this patch.

Signed-off-by: Alexander Graf <agraf@suse.de>
Reviewed-by: York Sun <york.sun@nxp.com>
[agraf: Fix x86 build]

8 years agoefi_loader: Fix efi_add_runtime_mmio definition
Alexander Graf [Thu, 17 Nov 2016 00:02:56 +0000 (01:02 +0100)]
efi_loader: Fix efi_add_runtime_mmio definition

The efi_add_runtime_mmio prototype for disabled CONFIG_EFI_LOADER
was different from the enabled one. Sync them.

Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi_loader: Disable PSCI reset for ls1043 and ls1046
Alexander Graf [Thu, 17 Nov 2016 00:02:55 +0000 (01:02 +0100)]
efi_loader: Disable PSCI reset for ls1043 and ls1046

The NXP ls1043 and ls1046 systems do not (yet) have PSCI enablement
for reset. Don't enable generic PSCI reset code on them.

Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi_loader: AArch64: Run EFI payloads in EL2 if U-Boot runs in EL3
Alexander Graf [Thu, 17 Nov 2016 00:02:58 +0000 (01:02 +0100)]
efi_loader: AArch64: Run EFI payloads in EL2 if U-Boot runs in EL3

Some boards decided not to run ATF or other secure firmware in EL3, so
they instead run U-Boot there. The uEFI spec doesn't know what EL3 is
though - it only knows about EL2 and EL1. So if we see that we're running
in EL3, let's get into EL2 to make payloads happy.

Signed-off-by: Alexander Graf <agraf@suse.de>
Reviewed-by: York Sun <york.sun@nxp.com>
8 years agosunxi: sina33: Enable the LCD
Maxime Ripard [Fri, 4 Nov 2016 15:18:11 +0000 (16:18 +0100)]
sunxi: sina33: Enable the LCD

The SinA33 comes with an optional 7" display. Enable it in the
configuration.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
8 years agosunxi: sina33: Enable the eMMC
Maxime Ripard [Fri, 4 Nov 2016 15:18:10 +0000 (16:18 +0100)]
sunxi: sina33: Enable the eMMC

The SinA33 has an 4GB Toshiba eMMC connected to the MMC2 controller.
Enable it.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Reviewed-by: Chen-Yu Tsai <wens@csie.org>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
8 years agommc: sunxi: Enable 8bits bus width for sun8i
Maxime Ripard [Fri, 4 Nov 2016 15:18:09 +0000 (16:18 +0100)]
mmc: sunxi: Enable 8bits bus width for sun8i

The sun8i SoCs also have a 8 bits capable MMC2 controller. Enable the
support for those too.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Reviewed-by: Chen-Yu Tsai <wens@csie.org>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
8 years agommc: Retry the switch command
Maxime Ripard [Fri, 4 Nov 2016 15:18:08 +0000 (16:18 +0100)]
mmc: Retry the switch command

Some eMMC will fail at the first switch, but would succeed in a subsequent
one.

Make sure we try several times to cover those cases. The number of retries
(and the behaviour) is currently what is being used in Linux.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
8 years agoARM64: zynqmp: Adding prefetchable memory space to pcie
Bharat Kumar Gogada [Tue, 2 Aug 2016 15:04:13 +0000 (20:34 +0530)]
ARM64: zynqmp: Adding prefetchable memory space to pcie

Adding prefetchable memory space to pcie device tree node.
Shifting configuration space to 64-bit address space.
Removing pcie device tree node from amba as it requires size-cells=<2>
in order to access 64-bit address space.

Signed-off-by: Bharat Kumar Gogada <bharatku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add clocks for LPDDMA
Kedareswara rao Appana [Fri, 30 Sep 2016 05:04:59 +0000 (10:34 +0530)]
ARM64: zynqmp: Add clocks for LPDDMA

Zynqmp DMA driver expects two clocks (main clock and apb clock)
For LPDDMA channels the two clocks are missing in the
Dma node resulting probe failure.

xilinx-zynqmp-dma ffa80000.dma: main clock not found.
xilinx-zynqmp-dma ffa80000.dma: Probing channel failed
xilinx-zynqmp-dma: probe of ffa80000.dma failed with error -2

This patch fixes this issue.

Signed-off-by: Kedareswara rao Appana <appanad@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add description for LPDDMA channel usage
Kedareswara rao Appana [Fri, 9 Sep 2016 07:06:01 +0000 (12:36 +0530)]
ARM64: zynqmp: Add description for LPDDMA channel usage

LPDDMA default allows only secured access.
inorder to enable these dma channels,
one should ensure that it allows non secure access.
This patch updates the same.

Reported-by: Sai Pavan Boddu <saipava@xilinx.com>
Signed-off-by: Kedareswara rao Appana <appanad@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Use 64bit size cell format for main amba bus
Michal Simek [Thu, 11 Feb 2016 06:19:06 +0000 (07:19 +0100)]
ARM64: zynqmp: Use 64bit size cell format for main amba bus

Use 64bit size cell for main amba bus instead of 32bit because PCIe
node requires it Change 64bit sizes also for all others IPs.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add ocm node in dtsi
Naga Sureshkumar Relli [Wed, 18 May 2016 06:53:13 +0000 (12:23 +0530)]
ARM64: zynqmp: Add ocm node in dtsi

This patch adds ocm controller node in zynqmp.dtsi.
needed for OCM edac support.

Signed-off-by: Naga Sureshkumar Relli <nagasure@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add device tree properties for ZynqMP GT core
Anurag Kumar Vulisha [Tue, 17 May 2016 11:19:01 +0000 (16:49 +0530)]
ARM64: zynqmp: Add device tree properties for ZynqMP GT core

This patch adds the ZynqMP GT core device-tree properties for
zynqmp.dtsi file.

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Tested-by: Hyun Kwon <hyunk@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoRevert "ARM64: zynqmp: Added broken-tuning property to SD, eMMC nodes"
Michal Simek [Thu, 20 Oct 2016 08:36:05 +0000 (10:36 +0200)]
Revert "ARM64: zynqmp: Added broken-tuning property to SD, eMMC nodes"

This reverts commit bd750e7a6c515c081b72d4ef108a2bfa691a3fd1

Implemented the new workaround for auto tuning based on
zynqmp compatible string, so removed the 'broken-tuning'
property.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: change sdhci compatible string.
Sai Krishna Potthuri [Tue, 16 Aug 2016 09:11:35 +0000 (14:41 +0530)]
ARM64: zynqmp: change sdhci compatible string.

This patch changes the compatible string for sdhci node,
adds "xlnx,device_id" and "xlnx,mio_bank" property to sdhci node.

Signed-off-by: Sai Krishna Potthuri <lakshmis@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: List all SMMU ids
Michal Simek [Wed, 6 Apr 2016 08:43:23 +0000 (10:43 +0200)]
ARM64: zynqmp: List all SMMU ids

Add SMMU description for all tested IPs.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add support for zynqmp fpga manager
Nava kishore Manne [Sat, 20 Aug 2016 18:47:52 +0000 (00:17 +0530)]
ARM64: zynqmp: Add support for zynqmp fpga manager

Add support for zynqmp fpga manager.

Signed-off-by: Nava kishore Manne <navam@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add cortexa53 edac node
Naga Sureshkumar Relli [Mon, 20 Jun 2016 10:18:30 +0000 (15:48 +0530)]
ARM64: zynqmp: Add cortexa53 edac node

This patch adds edac node for arm cortexa53 to report
errors on L1 and L2 caches.

Signed-off-by: Naga Sureshkumar Relli <nagasure@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoRevert "ARM64: zynqmp: Add serdes address space dp driver"
Michal Simek [Thu, 20 Oct 2016 08:38:16 +0000 (10:38 +0200)]
Revert "ARM64: zynqmp: Add serdes address space dp driver"

This reverts commit 786db82bd5bf09cc8f78c8b14445e843d7566b1c.

Since we are using serdes driver , no need of mapping serdes register
space into DP driver.

Signed-off-by: Anurag Kumar Vulisha <anuragku@xilinx.com>
Tested-by: Hyun Kwon <hyunk@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: drm: Add DMA index
Hyun Kwon [Fri, 15 Jul 2016 00:42:44 +0000 (17:42 -0700)]
ARM64: zynqmp: drm: Add DMA index

Each plane can be associated with multiple DMA channels. So add
index for each DMA channel.

Signed-off-by: Hyun Kwon <hyun.kwon@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Sync gpio node properties
Michal Simek [Thu, 20 Oct 2016 08:26:13 +0000 (10:26 +0200)]
ARM64: zynqmp: Sync gpio node properties

Keep dtsi in sync with mainline kernel.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Remove xlnx,id property
Michal Simek [Tue, 9 Aug 2016 13:06:58 +0000 (15:06 +0200)]
ARM64: zynqmp: Remove xlnx,id property

Remove unused xlnx,id property because it is not the part of
DT binding.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: pci: Updating device tree as per upstream
Bharat Kumar Gogada [Tue, 19 Jul 2016 15:19:29 +0000 (20:49 +0530)]
ARM64: zynqmp: pci: Updating device tree as per upstream

Updating required device tree changes as per mainlined driver
from 4.6 kernel.

Signed-off-by: Bharat Kumar Gogada <bharatku@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Support for multiple PM IDs assigned to a PM domain
Filip Drazic [Mon, 29 Aug 2016 17:32:59 +0000 (19:32 +0200)]
ARM64: zynqmp: Support for multiple PM IDs assigned to a PM domain

Previously, it was assumed that there is a 1:1 mapping between
PM ID defined in the platform firmware and a PM domain. However, there
can be a situation where multiple PM IDs belong to a single PM domain
(e.g. PM IDs for GPU and two pixel processors correspond to a single
PM domain).

This patch adds support for assigning more than one PM ID to
a single PM domain.

Updated documentation accordingly.

Assigned pixel processors PM IDs to GPU PM domain.

Signed-off-by: Filip Drazic <filip.drazic@aggios.com>
Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: DT: Add PM domains for GPU and PCIE
Filip Drazic [Mon, 29 Aug 2016 17:32:56 +0000 (19:32 +0200)]
ARM64: zynqmp: DT: Add PM domains for GPU and PCIE

Signed-off-by: Filip Drazic <filip.drazic@aggios.com>
Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: DT: Remove unused PM domains for PLL
Filip Drazic [Thu, 25 Aug 2016 16:58:51 +0000 (18:58 +0200)]
ARM64: zynqmp: DT: Remove unused PM domains for PLL

Signed-off-by: Filip Drazic <filip.drazic@aggios.com>
Acked-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: DT: Remove unused DDR PM domain
Filip Drazic [Thu, 25 Aug 2016 16:58:49 +0000 (18:58 +0200)]
ARM64: zynqmp: DT: Remove unused DDR PM domain

DDR power states are handled by the PM firmware, so this domain is
redundant. Also, since there is no device using this PM domain,
it will be powered off during boot, which is wrong.

Signed-off-by: Filip Drazic <filip.drazic@aggios.com>
Acked-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Remove note about level shifter on zcu102
Michal Simek [Wed, 19 Oct 2016 14:07:58 +0000 (16:07 +0200)]
ARM64: zynqmp: Remove note about level shifter on zcu102

i2c device is just level shifter. Remove reference from dts.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add dcc port to dtsi
Michal Simek [Fri, 9 Sep 2016 06:46:39 +0000 (08:46 +0200)]
ARM64: zynqmp: Add dcc port to dtsi

Add dcc to dtsi for supporting system without serial port.
DCC is enabled by default on ZynqMP.
Adding dcc to zcu100 and zcu102 which were tested.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add gpio-keys for zcu102
Michal Simek [Wed, 25 May 2016 18:09:35 +0000 (20:09 +0200)]
ARM64: zynqmp: Add gpio-keys for zcu102

There is gpio push button on MIO22. Add it to DTS to have full board
description.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Enable gpio-led as heartbeat on zcu102
Michal Simek [Wed, 20 Apr 2016 11:12:25 +0000 (13:12 +0200)]
ARM64: zynqmp: Enable gpio-led as heartbeat on zcu102

Show user that Linux is alive on the board.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Enable can1 for ep108
Naga Sureshkumar Relli [Tue, 12 Apr 2016 06:16:11 +0000 (11:46 +0530)]
ARM64: zynqmp: Enable can1 for ep108

This patch enables can1 for ep108.

Signed-off-by: Naga Sureshkumar Relli <nagasure@xilinx.com>
Reviewed-by: Kedareswara rao Appana <appanad@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Added clocks to DT for ep108
VNSL Durga [Mon, 11 Apr 2016 12:13:47 +0000 (17:43 +0530)]
ARM64: zynqmp: Added clocks to DT for ep108

Added clks for ep108 platform.

Signed-off-by: VNSL Durga <vnsldurg@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add clocks for LPDDMA
Kedareswara rao Appana [Fri, 9 Sep 2016 07:06:00 +0000 (12:36 +0530)]
ARM64: zynqmp: Add clocks for LPDDMA

Zynqmp DMA driver expects two clocks (main clock and apb clock)
LPDDMA clock cofiguration is missing for the same in the
zynqmp-clk.dtsi file.

This patch updates for the same.

Reported-by: Sai Pavan Boddu <saipava@xilinx.com>
Signed-off-by: Kedareswara rao Appana <appanad@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Remove DTC 1.4.2 warnings
Michal Simek [Fri, 11 Nov 2016 12:21:04 +0000 (13:21 +0100)]
ARM64: zynqmp: Remove DTC 1.4.2 warnings

DTC 1.4.2 reports these warnings:
Warning (unit_address_vs_reg): Node /amba_apu has a reg or ranges
property, but no unit name
Warning (unit_address_vs_reg): Node /amba has a reg or ranges property,
but no unit name
Warning (unit_address_vs_reg): Node /amba/usb@fe200000 has a unit name,
but no reg property
Warning (unit_address_vs_reg): Node /amba/usb@fe300000 has a unit name,
but no reg property
Warning (unit_address_vs_reg): Node
/amba/dma@fd4c0000/dma-video0channel@fd4c0000 has a unit name, but no
reg property
Warning (unit_address_vs_reg): Node
/amba/dma@fd4c0000/dma-video1channel@fd4c0000 has a unit name, but no
reg property
Warning (unit_address_vs_reg): Node
/amba/dma@fd4c0000/dma-video2channel@fd4c0000 has a unit name, but no
reg property
Warning (unit_address_vs_reg): Node
/amba/dma@fd4c0000/dma-graphicschannel@fd4c0000 has a unit name, but no
reg property
Warning (unit_address_vs_reg): Node
/amba/dma@fd4c0000/dma-audio0channel@fd4c0000 has a unit name, but no
reg property
Warning (unit_address_vs_reg): Node
/amba/dma@fd4c0000/dma-audio1channel@fd4c0000 has a unit name, but no
reg property
Warning (unit_address_vs_reg): Node /memory has a reg or ranges
property, but no unit name

This patch is fixing them.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM: zynq: Remove DTC 1.4.2 warnings
Michal Simek [Fri, 11 Nov 2016 12:11:37 +0000 (13:11 +0100)]
ARM: zynq: Remove DTC 1.4.2 warnings

DTC 1.4.2 reports these warnings:
Warning (unit_address_vs_reg): Node /memory has a reg or ranges
property, but no unit name
Warning (unit_address_vs_reg): Node /pmu has a reg or ranges property,
but no unit name
Warning (unit_address_vs_reg): Node /fixedregulator@0 has a unit name,
but no reg property

This patch is fixing them.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Correct the sdhci minimum frequency for ep108
Siva Durga Prasad Paladugu [Tue, 1 Nov 2016 18:19:53 +0000 (23:49 +0530)]
ARM64: zynqmp: Correct the sdhci minimum frequency for ep108

Correct the sdhci minimum frequency for ep platform.
It should be right shift instead of left shift operand.

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Ignore warnings from autogenerated files
Michal Simek [Thu, 14 Jul 2016 09:13:04 +0000 (11:13 +0200)]
ARM64: zynqmp: Ignore warnings from autogenerated files

Autogenerated files contain casting issues and missing function
declaration and even usleep implementation. Suppress them for now
till these files are fixed.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Fix secondary bootmode enabling
Michal Simek [Tue, 25 Oct 2016 09:43:02 +0000 (11:43 +0200)]
ARM64: zynqmp: Fix secondary bootmode enabling

Do not setup use_alt bit which copy alternative boot mode to
boot mode. The reason is that this bit is cleared after POR
but not after any software reset which will cause
that after SW reset bootrom will look for different boot image.

This patch setups alternative boot mode selection (purely SW
handling) and extends code to read this alternative boot mode first and
use it if it is setup.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Add support for SD1 with level shifters bootmode
Siva Durga Prasad Paladugu [Wed, 21 Sep 2016 06:15:05 +0000 (11:45 +0530)]
ARM64: zynqmp: Add support for SD1 with level shifters bootmode

Add support for SD1 with level shifters bootmode.

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Record board name as serial number for DFU/FASTBOOT
Michal Simek [Thu, 1 Sep 2016 09:27:32 +0000 (11:27 +0200)]
ARM64: zynqmp: Record board name as serial number for DFU/FASTBOOT

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM64: zynqmp: Adjust to new SMC interface to get silicon version
Soren Brinkmann [Thu, 29 Sep 2016 18:44:41 +0000 (11:44 -0700)]
ARM64: zynqmp: Adjust to new SMC interface to get silicon version

The new FW interface returns the IDCODE and version register, leaving
extracting bitfields to the caller.

Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM: zynq: Add support for Zynq 7000S 7007s/7012s/7014s devices
Michal Simek [Tue, 18 Oct 2016 14:10:25 +0000 (16:10 +0200)]
ARM: zynq: Add support for Zynq 7000S 7007s/7012s/7014s devices

Zynq 7000S (Single A9 core) devices is using different ID code.
This patch adds this new codes and assign them.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agonet: zynq_gem: Correct SGMII enable bit setting
Siva Durga Prasad Paladugu [Mon, 16 May 2016 10:01:38 +0000 (15:31 +0530)]
net: zynq_gem: Correct SGMII enable bit setting

Correct the SGMII enable bit position to 27 instead
of 31.

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
8 years agonet: zynq_gem: Modify the nwcfg bit definitions
Siva Durga Prasad Paladugu [Mon, 16 May 2016 10:01:37 +0000 (15:31 +0530)]
net: zynq_gem: Modify the nwcfg bit definitions

Modify the nwcfg bit definitions to have 32-bit
by removing the extra nibble.

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
8 years agonand: arasan_nfc: Clear ecc on bit while sending read command
Siva Durga Prasad Paladugu [Thu, 25 Aug 2016 10:30:04 +0000 (16:00 +0530)]
nand: arasan_nfc: Clear ecc on bit while sending read command

Clear ecc ON bit while sending read command as all types
of read command(like reading spare) doesnt need ECC to be
enabled. It has been anyway taken care in other places
whereever required using arasan_nand_enable_ecc().

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agozynq: nand: Runtime detection of nand buswidth through slcr
Michal Simek [Wed, 26 Oct 2016 08:49:37 +0000 (10:49 +0200)]
zynq: nand: Runtime detection of nand buswidth through slcr

This patch adds support to check the buswidth on nand flash
at runtime based on nand MIO configurations done by FSBL.

User needs to correctly configure the MIO's based on the
buswidth supported by the nand flash which is present on the board.

Added nand8 and nand16 @periph names on slcr driver.

Signed-off-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agozynq: nand: Enable Nand flash controller driver a zynq board
Siva Durga Prasad Paladugu [Tue, 27 Sep 2016 05:25:47 +0000 (10:55 +0530)]
zynq: nand: Enable Nand flash controller driver a zynq board

Enable zynq Nand flash controller driver for a zynq ZC770
XM011(dc2) board.

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agomtd: nand: zynq_nand: Add nand driver support for zynq
Siva Durga Prasad Paladugu [Tue, 27 Sep 2016 05:25:46 +0000 (10:55 +0530)]
mtd: nand: zynq_nand: Add nand driver support for zynq

Add nand flash controller driver support for zynq SoC.

Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM: zynq: Add support for the topic-miami system-on-modules and carrier boards
Mike Looijmans [Fri, 30 Sep 2016 06:13:13 +0000 (08:13 +0200)]
ARM: zynq: Add support for the topic-miami system-on-modules and carrier boards

The topic-miami SoMs contain a Zynq xc7z015 or xc7z030 SoC, 1GB DDR3L RAM,
32MB QSPI NOR flash and 256MB NAND flash.

The topic-miamiplus SoMs contain a Zynq xc7z035, xc7z045 or xc7z100 SoC,
2x 1GB DDR3L RAM, 64MB dual-parallel QSPI flash, clock sources
and a fan controller.

The "Florida" carrier boards add SD, USB, ethernet and other interfaces.

Signed-off-by: Mike Looijmans <mike.looijmans@topic.nl>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agoARM: zynq: Make SYS_VENDOR configurable
Mike Looijmans [Wed, 28 Sep 2016 05:46:30 +0000 (07:46 +0200)]
ARM: zynq: Make SYS_VENDOR configurable

Add a string description for SYS_VENDOR to allow configuring boards from
other vendors than just "xilinx".

Signed-off-by: Mike Looijmans <mike.looijmans@topic.nl>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agotools: mkimage: Check if file is regular file
Michal Simek [Fri, 21 Oct 2016 11:16:13 +0000 (13:16 +0200)]
tools: mkimage: Check if file is regular file

Current Makefile.spl passes -R parameter which is not empty
and pointing to ./ folder.
"./tools/mkimage -T zynqmpimage -R ./"" -d spl/u-boot-spl.bin
spl/boot.bin"
That's why mkimage is trying to parse ./ file and generate
register init which is wrong.
Check that passed filename is regular file. If not do not work with it.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agotools: mkimage: Add support for initialization table for Zynq and ZynqMP
Mike Looijmans [Tue, 20 Sep 2016 09:37:24 +0000 (11:37 +0200)]
tools: mkimage: Add support for initialization table for Zynq and ZynqMP

The Zynq/ZynqMP boot.bin file contains a region for register initialization
data. Filling in proper values in this table can reduce boot time
(e.g. about 50ms faster on QSPI boot) and also reduce the size of
the SPL binary.

The table is a simple text file with register+data on each line. Other
lines are simply skipped. The file can be passed to mkimage using the
"-R" parameter.

It is recommended to add reg init file to board folder.
For example:
CONFIG_BOOT_INIT_FILE="board/xilinx/zynqmp/xilinx_zynqmp_zcu102/reg.int

Signed-off-by: Mike Looijmans <mike.looijmans@topic.nl>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
8 years agox86: Enable EFI loader support
Simon Glass [Mon, 7 Nov 2016 15:47:16 +0000 (08:47 -0700)]
x86: Enable EFI loader support

Enable this so that EFI applications (notably grub) can be run under U-Boot
on x86 platforms.

At present the 'hello world' EFI application is not supported for the
qemu-x86_efi_payload64 board. That board builds a payload consisting of a
64-bit header and a 32-bit U-Boot, which is incompatible with the way the
EFI loader builds its EFI application. The following error is obtained:

x86_64-linux-ld.bfd: i386 architecture of input file
   `lib/efi_loader/helloworld.o' is incompatible with i386:x86-64 output

This could be corrected with additional Makefile rules. For now, this
feature is disabled for that board.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
[agraf: drop hello kconfig bits]
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi: x86: Adjust EFI files support efi_loader
Simon Glass [Mon, 7 Nov 2016 15:47:15 +0000 (08:47 -0700)]
efi: x86: Adjust EFI files support efi_loader

Add compiler flags and make a few minor adjustments to support the efi
loader.

Signed-off-by: Simon Glass <sjg@chromium.org>
[agraf: Add Kconfig dep]
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agox86: Move efi .S files into the 'lib' directory
Simon Glass [Mon, 7 Nov 2016 15:47:14 +0000 (08:47 -0700)]
x86: Move efi .S files into the 'lib' directory

These files now need to be in a standard place so that they can be located
by generic Makefile rules. Move them to the 'lib' directory.

Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agox86: Move efi .lds files into the 'lib' directory
Simon Glass [Mon, 7 Nov 2016 15:47:13 +0000 (08:47 -0700)]
x86: Move efi .lds files into the 'lib' directory

These files now need to be in a standard place so that they can be located
by generic Makefile rules. Move them to the 'lib' directory.

Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi: arm: Add aarch64 EFI app support
Simon Glass [Mon, 7 Nov 2016 15:47:11 +0000 (08:47 -0700)]
efi: arm: Add aarch64 EFI app support

Add support for EFI apps on aarch64. This includes start-up and relocation
code plus a link script.

Signed-off-by: Simon Glass <sjg@chromium.org>
[agraf: add kconfig dep]
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi: arm: Add EFI app support
Simon Glass [Mon, 7 Nov 2016 15:47:10 +0000 (08:47 -0700)]
efi: arm: Add EFI app support

Add support for EFI apps on ARM. This includes start-up and relocation
code, plus a link script and some compiler setting changes.

Signed-off-by: Simon Glass <sjg@chromium.org>
[agraf: Remove whitespace change, add kconfig dep]
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoelf: arm: Add a few ARM relocation types
Simon Glass [Mon, 7 Nov 2016 15:47:09 +0000 (08:47 -0700)]
elf: arm: Add a few ARM relocation types

Rather than hard-coding the relocation type, add it to the ELF header file
and use it from there.

Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi: Add support for a hello world test program
Simon Glass [Mon, 7 Nov 2016 15:47:08 +0000 (08:47 -0700)]
efi: Add support for a hello world test program

It is useful to have a basic sanity check for EFI loader support. Add a
'bootefi hello' command which loads HelloWord.efi and runs it under U-Boot.

Signed-off-by: Simon Glass <sjg@chromium.org>
[agraf: Fix documentation, add unfulfilled kconfig dep]
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agoefi: Makefile: Export variables for use with EFI
Simon Glass [Mon, 7 Nov 2016 15:47:07 +0000 (08:47 -0700)]
efi: Makefile: Export variables for use with EFI

When building an EFI app we need three things:

   - start-up code
   - relocation code
   - link script

These are all different for each architecture. We also need special
compiler flags in some cases.

Add top-level Makefile variables for these along with documentation.

Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
8 years agox86: Tidy up selection of building the EFI stub
Simon Glass [Mon, 7 Nov 2016 15:47:06 +0000 (08:47 -0700)]
x86: Tidy up selection of building the EFI stub

At present we use a CONFIG option in efi.h to determine whether we are
building the EFI stub or not. This means that the same header cannot be
used for EFI_LOADER support. The CONFIG option will be enabled for the
whole build, even when not building the stub.

Use a different define instead, set up just for the files that make up the
stub.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Signed-off-by: Alexander Graf <agraf@suse.de>