powerpc/mpc85xx: set L2PE in L2CSR0 before enabling L2 cache
While enabling L2 cache, the value of L2PE (L2 cache parity/ECC
error checking enable) must not be changed while the L2 cache is
enabled.
So, L2PE must be set before enabling L2 cache.
Signed-off-by: Aneesh Bansal <aneesh.bansal@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
Dan Murphy [Mon, 2 May 2016 20:46:02 +0000 (15:46 -0500)]
net: phy: dp83867: Add SGMII helper for configuration
The code assumed that if the interface is not RGMII configured
then it must be SGMII configured. This device has the ability
to support most of the MII interfaces. Therefore add the
helper for SGMII and only configure the device if the interface is
configured for SGMII.
Signed-off-by: Dan Murphy <dmurphy@ti.com> Reviewed-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Michal Simek <michal.simek@xilinx.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Dan Murphy [Mon, 2 May 2016 20:46:01 +0000 (15:46 -0500)]
net: phy: Add phy_interface_is_sgmii to phy.h
Add a helper to phy.h to identify whether the
phy is configured for SGMII all variables.
Signed-off-by: Dan Murphy <dmurphy@ti.com> Reviewed-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Michal Simek <michal.simek@xilinx.com> Tested-by: Mugunthan V N <mugunthanvnm@ti.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Dan Murphy <dmurphy@ti.com> Reviewed-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Michal Simek <michal.simek@xilinx.com> Tested-by: Mugunthan V N <mugunthanvnm@ti.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Dan Murphy [Mon, 2 May 2016 20:45:59 +0000 (15:45 -0500)]
net: phy: ti: Allow the driver to be more configurable
Not all devices use the same internal delay or fifo depth.
Add the ability to set the internal delay for rx or tx and the
fifo depth via the devicetree. If the value is not set in the
devicetree then set the delay to the default.
If devicetree is not used then use the default defines within the
driver.
Signed-off-by: Dan Murphy <dmurphy@ti.com> Tested-by: Mugunthan V N <mugunthanvnm@ti.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Dan Murphy <dmurphy@ti.com> Reviewed-by: Mugunthan V N <mugunthanvnm@ti.com> Tested-by: Mugunthan V N <mugunthanvnm@ti.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Dan Murphy [Mon, 2 May 2016 20:45:57 +0000 (15:45 -0500)]
net: zynq_gem: Add the passing of the phy-handle node
Add the ability to pass the phy-handle node offset
to the phy driver. This allows the phy driver
to access the DT subnode's data and parse accordingly.
Signed-off-by: Dan Murphy <dmurphy@ti.com> Tested-by: Michal Simek <michal.simek@xilinx.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Mugunthan V N [Thu, 28 Apr 2016 10:06:07 +0000 (15:36 +0530)]
drivers: net: cpsw: add support for reading mac address from efuse
Different TI platforms has to read with different combination to
get the mac address from efuse. So add support to read mac address
based on machine/device compatibles.
The code is taken from Linux drivers/net/ethernet/ti/cpsw-common.c
done by Tony Lindgren.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Mugunthan V N [Thu, 28 Apr 2016 10:06:06 +0000 (15:36 +0530)]
drivers: net: cpsw: fix get mdio base and gmii_sel reg from DT
Since dra7x platforms address bus is define as 64 bits to support
LAPE, fdtdec_get_addr() returns a invalid address for mdio based
and gmii_sel register address. Fixing this by using
fdtdec_get_addr_size_auto_noparent() which will derive address
cell and size cell from its parent.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Mugunthan V N [Thu, 28 Apr 2016 10:06:04 +0000 (15:36 +0530)]
drivers: net: cpsw: fix cpsw dp parse when num slaves as 1
On some boards number of slaves can be 1 when only one port
ethernet is pinned out. So do not break when slave_index and
num slaves check fails, instead continue to parse the next
child.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Mugunthan V N [Thu, 28 Apr 2016 10:06:03 +0000 (15:36 +0530)]
ti_omap5_common: eth: do not define DM_ETH for spl
Since omap's spl doesn't support DM currently, do not define
DM_ETH for spl build.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Mugunthan V N [Thu, 28 Apr 2016 10:06:02 +0000 (15:36 +0530)]
drivers: core: device: add support to check dt compatible for a device/machine
Provide an api to check whether the given device or machine is
compatible with the given compat string which helps in making
decisions in drivers based on device or machine compatible.
Idea taken from Linux.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com> Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Kevin Smith [Thu, 31 Mar 2016 19:33:12 +0000 (19:33 +0000)]
net: phy: Add PHY driver for mv88e61xx switches
The previous mv88e61xx driver was a driver for configuring the
switch, but did not integrate with the PHY/networking system, so
it could not be used as a PHY by U-boot. This is a complete
rework to support this device as a PHY.
Signed-off-by: Kevin Smith <kevin.smith@elecsyscorp.com> Acked-by: Prafulla Wadaskar <prafulla@marvell.com> Cc: Albert ARIBAUD <albert.u.boot@aribaud.net> Cc: Joe Hershberger <joe.hershberger@ni.com> Cc: Stefan Roese <sr@denx.de> Cc: Marek Vasut <marex@denx.de> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Kevin Smith [Thu, 31 Mar 2016 19:33:12 +0000 (19:33 +0000)]
net: Remove unused mv88e61xx switch driver
No boards are using this driver. Remove in preparation for a new
driver with integrated PHY support.
Signed-off-by: Kevin Smith <kevin.smith@elecsyscorp.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com> Cc: Prafulla Wadaskar <prafulla@marvell.com> Cc: Albert ARIBAUD <albert.u.boot@aribaud.net> Cc: Stefan Roese <sr@denx.de> Cc: Marek Vasut <marex@denx.de>
Michal Simek [Fri, 20 May 2016 12:59:33 +0000 (14:59 +0200)]
ARM: zynq: Simplify zynq configuration
Extending Kconfig for adding new platform is a lot of work
for nothing. Setting SYS_CONFIG_NAME directly in Kconfig and
remove all dependencies on TARGET_ZYNQ_* options including SPL.
As a side-effect it also remove custom init folder for ps7_init_gpl.*
files. Folder is chosen based on device-tree file.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Michal Simek [Wed, 18 May 2016 12:46:28 +0000 (14:46 +0200)]
phy: marvell: Do not reset phy after negotiation
The patch
"net: phy: do not read configuration register on reset"
(sha1: a058052c358c3ecf5f394ff37def6a45eb26768c)
was causing regression on zynq zc702 board where Marwell 88e1118
phy was resetted after negotiation was setup.
Phy reset is done pretty early in phy_connect_dev() and doens't need to
be called again in phy code.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Michal Simek [Tue, 17 May 2016 12:03:50 +0000 (14:03 +0200)]
image: Add boot_get_fpga() to load fpga with bootm
Add function boot_get_fpga() which find and load bitstream to
programmable logic if fpga entry is present.
Function is supported on Xilinx devices for full and partial bitstreams
in BIN and BIT format.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Remove additional blankline in image.h
and program it via u-boot
tftpb 10000 sd.img
mmcinfo
mmc write 10000 0 $filesize
mmc rescan
mmc part
ls mmc 0
psu_init() function contains low level SoC setup generated for every HW
design by Xilinx design tools. xil_io.h is only supporting file to fix
all dependencies from tools. The same solution was used on Xilinx Zynq.
The patch also change CONFIG_SYS_INIT_SP_ADDR to the end of OCM which
stays at the same location all the time.
Bootrom expects starting address to be at 0xfffc0000 that's why this
address is SPL_TEXT_BASE.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Alexander Graf [Thu, 12 May 2016 11:44:01 +0000 (13:44 +0200)]
ARM64: zynqmp: Align gic ranges for 64k in device tree
The GIC ranges in the zynqmp device tree are only 4kb aligned. Since
commit 12e14066f we automatically deal with aliases GIC regions though,
so we can map them transparently into guests even on 64kb page size
systems.
This patch makes use of that features and sets GICC and GICV to 64kb
aligned and sized regions.
Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Add support for the zynqmpimage to mkimage.
Only basic functionality is supported without encryption and register
initialization with one partition which is filled by U-Boot SPL.
For more detail information look at Xilinx ZynqMP TRM.
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Michal Simek [Fri, 22 Apr 2016 09:48:49 +0000 (11:48 +0200)]
ARM64: zynqmp: Add support for reading MAC from eeprom
Add support for on board eeprom with programmed MAC for using in u-boot
to have uniq address for every board.
Most of the time uniq MAC address is on a label on the board.
If address is not programmed use these command to program it.
ARM: zynq: load u-boot.img whether CONFIG_OF_SEPARATE is defined or not
Since commit ad1ecd2063da ("fdt: Build a U-Boot binary without device
tree"), u-boot-dtb.img is identical to u-boot.img, so SPL can always
load u-boot.img whether CONFIG_OF_SEPARATE is defined or not.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Marek Vasut [Fri, 22 Apr 2016 19:56:21 +0000 (21:56 +0200)]
ARM: sama5d2: Implement boot device autodetection
Implement support for saving ARM register R4 early during boot using
save_boot_params . Implement support for decoding the stored register
R4 value in spl_boot_device() to obtain boot device from which the
SoC booted. This way, the SPL will always load U-Boot from the same
device from which the SPL itself booted instead of using hard-coded
boot device.
This functionality is useful for example when booting sama5d2-xplained
from SD card, where by default the SPL would try loading the U-Boot
from eMMC and fail. This is because eMMC is on SDHCI0 (BOOT_DEVICE_MMC1),
while SD slot is on SDHCI1 (BOOT_DEVICE_MMC2) and the SPL was hard-wired
to always boot from BOOT_DEVICE_MMC1.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Andreas Bießmann <andreas.devel@googlemail.com> Cc: Wenyou Yang <wenyou.yang@atmel.com> Reviewed-by: Andreas Bießmann <andreas@biessmann.org>
Marek Vasut [Mon, 18 Apr 2016 16:30:41 +0000 (18:30 +0200)]
ARM: atmel: Enable FIT image support for SAMA5Dx
Enable the fitImage support for the entire SAMA5Dx lineup of CPUs.
The fitImage is superior image format to uImage and it is useful
to have it available.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Andreas Bießmann <andreas.devel@googlemail.com> Cc: Wenyou Yang <wenyou.yang@atmel.com> Reviewed-by: Andreas Bießmann <andreas@biessmann.org>
[rebase on current ToT] Signed-off-by: Andreas Bießmann <andreas@biessmann.org>
Martin Hejnfelt [Thu, 19 May 2016 07:11:58 +0000 (09:11 +0200)]
omap3: Fix SPI registers on am33xx and am43xx
When the base registers are read from device tree the base is not
0x48030100 as the driver expects, but 0x48030000, resulting in
non functioning SPI. To deal with this, use same idea as how this
is done in the linux kernel (drivers/spi/spi-omap2-mcspi.c) and
add a structure with a field that is used to shift the registers
on these systems.
Tom Rini [Fri, 13 May 2016 14:54:04 +0000 (10:54 -0400)]
kbuild: fixdep: Check fstat(2) return value
Coverity has recently added a check that will find when we don't check
the return code from fstat(2). Copy/paste the checking logic that
print_deps() has with an appropriate re-wording of the perror() message.
Michal Simek [Wed, 4 May 2016 13:14:11 +0000 (15:14 +0200)]
spl: Setup default value for OF_LIST
OF_LIST can't remain empty that's why setup it up to default DTB.
If it is empty u-boot.img is created without FDT partition:
For example:
./tools/mkimage -f auto -A arm -T firmware -C none -O u-boot -a
0x8000000 -e 0 -n "U-Boot 2016.05-rc3 ..." -E -b -d u-boot-nodtb.bin u-boot.img
Can't set 'timestamp' property for '' node (FDT_ERR_NOSPACE)
FIT description: Firmware image with one or more FDT blobs
Created: Wed May 4 15:02:52 2016
Image 0 (firmware@1)
Description: U-Boot 2016.05-rc3-00080-gff2e12ae22a8-dirty for zynqmp
board
Created: Wed May 4 15:02:52 2016
Type: Firmware
Compression: uncompressed
Data Size: unavailable
Architecture: ARM
Load Address: 0x08000000
Default Configuration: 'conf@1'
Configuration 0 (conf@1)
Description: unavailable
Kernel: unavailable
And then image like this doesn't contain description and link to FDT and
can't boot.
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Michal Simek [Wed, 4 May 2016 13:08:00 +0000 (15:08 +0200)]
spl: fit: Print error message when FDT is not present
When FDT is not present in the image user doesn't get any error what's
wrong. Print error message if LIBCOMMON_SUPPORT is enabled.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Seris-cc: uboot Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Simon Glass [Sun, 1 May 2016 19:55:38 +0000 (13:55 -0600)]
mkimage: Add a quiet mode
Some build systems want to be quiet unless there is a problem. At present
mkimage displays quite a bit of information when generating a FIT file. Add
a '-q' flag to silence this.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Simon Glass [Sun, 1 May 2016 19:55:37 +0000 (13:55 -0600)]
image-fit: Don't display an error in fit_set_timestamp()
This function returns an error code and its caller may be able to fix the
error. For example fit_handle_file() expands the device tree to fit if there
is a lack of space.
In this case the caller does not want an error displayed. It is confusing,
since it suggests that something is wrong, when it fact everything is fine.
Drop the error.
Stephen Warren [Mon, 25 Apr 2016 21:55:42 +0000 (15:55 -0600)]
malloc: improve memalign fragmentation fix
Commit 4f144a416469 "malloc: work around some memalign fragmentation
issues" enhanced memalign() so that it can succeed in more cases where
heap fragmentation is present. However, it did not solve as many cases
as it could. This patch enhances the code to cover more cases.
The alignment code works by allocating more space than the user requests,
then adjusting the returned pointer to achieve alignment. In general, one
must allocate "alignment" bytes more than the user requested in order to
guarantee that alignment is possible. This is what the original code does.
The previous enhancement attempted a second allocation if the padded
allocation failed, and succeeded if that allocation just happened to be
aligned; a fluke that happened often in practice. There are still cases
where this could fail, yet where it is still possible to honor the user's
allocation request. In particular, if the heap contains a free region that
is large enough for the user's request, and for leading padding to ensure
alignment, but has no or little space for any trailing padding. In this
case, we can make a third(!) allocation attempt after calculating exactly
the size of the leading padding required to achieve alignment, which is
the minimal over-allocation needed for the overall memalign() operation to
succeed if the third and second allocations end up at the same location.
This patch isn't checkpatch-clean, since it conforms to the existing
coding style in dlmalloc.c, which is different to the rest of U-Boot.
Signed-off-by: Stephen Warren <swarren@nvidia.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Bin Meng [Sun, 22 May 2016 08:45:39 +0000 (01:45 -0700)]
x86: galileo: Override SMBIOS product name
Override the default product name U-Boot reports in the SMBIOS
table, to be compatible with the Intel provided UEFI BIOS, as
Linux kernel drivers (drivers/mfd/intel_quark_i2c_gpio.c and
drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c) make use of
it to do different board level configuration.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:35 +0000 (01:45 -0700)]
x86: broadwell: Correct I/O APIC ID
Currently ID 2 is assgined to broadwell I/O APIC, however per
chromebook_samus.dts 2 is the core#2 LAPIC ID. Now we change
I/O APIC ID to 4 to avoid conflict.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:34 +0000 (01:45 -0700)]
x86: quark: Assign a unique I/O APIC ID
After power-on, both LAPIC and I/O APIC appear with the same APIC ID
zero, which creates an ID conflict. When generating MP table, U-Boot
reports zero as the LAPIC ID in the processor entry, and zero as the
I/O APIC ID in the I/O APIC as well as the I/O interrupt assignment
entries. Such MP table confuses Linux kernel and finally a kernel
panic is seen during boot:
BUG: unable to handle kernel paging request at ffff9000
IP: [<c101d462>] native_io_apic_write+0x22/0x30
*pdpt = 00000000014fb001 *pde = 00000000014ff067 *pte = 0000000000000000
Oops: 0002 [#1]
Modules linked in:
Pid: 1, comm: swapper Tainted: G W 3.8.7 #3 intel galileo/galileo
EIP: 0060:[<c101d462>] EFLAGS: 00010086 CPU: 0
EIP is at native_io_apic_write+0x22/0x30
...
Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:32 +0000 (01:45 -0700)]
x86: Remove SMP limitation in lapic_setup()
At present LAPIC is enabled and configured as virtual wire mode
in lapic_setup() only when CONFIG_SMP is on. This limitation is
however not necessary as for uniprocessor this is still needed.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:31 +0000 (01:45 -0700)]
x86: Don't touch IA32_APIC_BASE MSR on Intel Quark
Intel Quark processor core provides an integrated Local APIC but
does not support the IA32_APIC_BASE MSR. As a result, the Local
APIC is always globally enabled and the Local APIC base address
is fixed at 0xfee00000. Attempting to access the IA32_APIC_BASE
MSR causes a general protection fault.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:27 +0000 (01:45 -0700)]
x86: Add some notes for MRC cache with Intel FSP
MRC cache relies on Intel FSP to produce a special GUID that
contains the MRC cache data. Add such information in the
CONFIG_ENABLE_MRC_CACHE help entry.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:26 +0000 (01:45 -0700)]
x86: crownbay: Disable boot stage support
It is observed that when enabling boot stage support, occasionally
the board reboots during boot over and over again, and eventually
boots to shell. This was seen on my board, but not on Jian's board.
Debugging shows that the TSC timer calibration against PIT fails
as boot stage APIs utilize timer in a very early stage and at that
time TSC/PIT may not be stable enough for the calibration to pass.
Disable it for now.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Cc: Jian Luo <Jian.Luo4@boschrexroth.de> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 22 May 2016 08:45:25 +0000 (01:45 -0700)]
acpi: Clean IASL generated intermediate files
For boards that support ACPI, there are dsdt.aml, dsdt.asl.tmp and
dsdt.c in the board directory after a successful build. These are
intermediate files generated by IASL, and should be removed during
a 'make clean'.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Wed, 11 May 2016 14:45:11 +0000 (07:45 -0700)]
x86: baytrail: Add GPIO ASL description
Since BayTrail, Intel starts to use new GPIO IPs in their chipset.
This adds the GPIO ASL, so that OS can load corresponding drivers
for it. On Linux, this is BayTrail pinctrl driver.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Wed, 11 May 2016 14:45:10 +0000 (07:45 -0700)]
x86: baytrail: Add internal UART ASL description
BayTrail integrates an internal ns15550 compatible UART (PNP0501).
Its IRQ is hardwired to IRQ3 in old revision chipset, but in newer
revision one IRQ4 is being used for ISA compatibility. Handle this
correctly in the ASL file.
Linux does not need this ASL, but Windows need this to correctly
discover a COM port existing in the system so that Windows can
show it in the 'Device Manager' window, and expose this COM port
to any terminal emulation application.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Wed, 11 May 2016 14:45:09 +0000 (07:45 -0700)]
acpi: Quieten IASL output when 'make -s' is used
IASL compiler does not provide a command line option to turn off
its non-warning message. To quieten the output when 'make -s',
redirect its output to /dev/null.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Wed, 11 May 2016 14:45:05 +0000 (07:45 -0700)]
x86: acpi: Remove header length check when writing tables
Before moving 'current' pointer during ACPI table writing, we always
check the table length to see if it is larger than the table header.
Since our purpose is to generate valid tables, the check logic is
always true, which can be avoided.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Wed, 11 May 2016 14:45:03 +0000 (07:45 -0700)]
x86: acpi: Switch to ACPI mode by ourselves instead of requested by OSPM
Per ACPI spec, during ACPI OS initialization, OSPM can determine
that the ACPI hardware registers are owned by SMI (by way of the
SCI_EN bit in the PM1_CNT register), in which case the ACPI OS
issues the ACPI_ENABLE command to the SMI_CMD port. The SCI_EN bit
effectively tracks the ownership of the ACPI hardware registers.
However since U-Boot does not support SMI, we report all 3 fields
in FADT (SMI_CMD, ACPI_ENABLE, ACPI_DISABLE) as zero, by following
the spec who says: these fields are reserved and must be zero on
system that does not support System Management mode.
U-Boot seems to behave in a correct way that the ACPI spec allows,
at least Linux does not complain, but apparently Windows does not
think so. During Windows bring up debugging, it is observed that
even these 3 fields are zero, Windows are still trying to issue SMI
with hardcoded SMI port address and commands, and expecting SCI_EN
to be changed by the firmware. Eventually Windows gives us a BSOD
(Blue Screen of Death) saying ACPI_BIOS_ERROR and refuses to start.
To fix this, turn on the SCI_EN bit by ourselves. With this patch,
now U-Boot can install and boot Windows 8.1/10 successfully with
the help of SeaBIOS using legacy interface (non-UEFI mode).
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
projects / u-boot / log