On Gen5, when the FPGA is loaded and there was some prior interaction
between the HPS and the FPGA via bridges (e.g. Linux was running and
using some of the IPs in the FPGA) followed by warm reset, it has been
observed that there might be outstanding unfinished transactions. This
leads to an obscure misbehavior of the bridge.
When the bridge is enabled again in U-Boot and there are outstanding
transactions, a read from within the bridge address range would return
a result of the previous read instead. Example:
=> bridge enable ; md 0xff200000 1
ff200000: 1234abcd
=> bridge enable ; md 0xff200010 1
ff200010: 5678dcba <------- this is in fact a value which is stored in
a memory at 0xff200000
=> bridge enable ; md 0xff200000 1
ff200000: 90effe09 <------- this is in fact a value which is stored in
a memory at 0xff200010
and so it continues. Issuing a write does lock the system up completely.
This patch opens the FPGA bridges in 'bridge enable' command, the tears
them down again, and then opens them again. This allows these outstanding
transactions to complete and makes this misbehavior go away.
However, it is not entirely clear whether this is the correct solution.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dalon Westergreen <dwesterg@gmail.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Ley Foon Tan <ley.foon.tan@intel.com>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Reviewed-by: Ley Foon Tan <ley.foon.tan@intel.com>
On Gen5, the 'bridge disable' command write 0x0 to brgmodrst register,
which releases all bridges from reset, instead of putting all bridges
into reset. Fix this by inverting the mask and actually putting the
bridges into reset.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dalon Westergreen <dwesterg@gmail.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Ley Foon Tan <ley.foon.tan@intel.com>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Reviewed-by: Ley Foon Tan <ley.foon.tan@intel.com>
Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
The file arch/arm/mach-socfpga/include/mach/misc.h used the same include
guard as the global include/misc.h.
Fix this by giving the arch file an arch prefix.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Reviewed-by: Ley Foon Tan <ley.foon.tan@intel.com>
The default implementation of ERR_PTR/PTR_ERR maps errno values at the
and of the address range (e.g. -EINVAL/-22 gets 0xFFFFFFEA).
For socfpga gen5 SPL, this doesn't really work, as the heap is nearly
at the end of the 32 bit address range.
This patch adjusts the ERR_PTR_OFFSET to map errno values into the range
of the Boot ROM, which should not be used for valid pointers.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
The company Samtec was merged into Softing, migrate the board over to
the new name and update copyright headers.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Silvio Fricke <silvio.fricke@softing.com>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
This is currently a decimal value which is not as convenient or
meaningful. Also U-Boot tends to use hex everywhere.
Convert this option to hex and add a comment for the size_check macro.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Acked-by: Bin Meng <bmeng.cn@gmail.com>
[bmeng: correct the typo in the commit title]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
At present there is only one control for this and it is used for both SPL
and TPL. But SPL might have a lot more space than TPL so the extra cost of
a full printf() might be acceptable.
Split the option into two, providing separate SPL and TPL controls. The
TPL setting defaults to the same as SPL.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The socfpga gen5 SPL manually zeroed bss in board_init_f(). Now that the
DDR driver does not use bss any more, bss is not used before board_init_r()
and we can remove this hack.
bss is normally zeroed by crt0.S, but after board_init_f(), before
board_init_r(). socfpga just had this double-zeroing because it invalidly
used bss in board_init_f() already (during DDR initialization).
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
This moves sysreset support for socfgpa from ad-hoc code in mach-socfpga
to a UCLASS_SYSRESET based dm driver.
A side effect is that gen5 and a10 can now select between cold and warm
reset.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
This adds a define for the bit in rstmgr's ctrl regiser that issues
a cold reset (we had a define for the warm reset bit only) in preparation
for a proper sysrese driver.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Series changes: 2
- separate this patch to the register descriptions from the actual
sysreset driver patch
This provides an SPL_SIZE_LIMIT that makes the build check that the SPL
binary loaded from flash fits into the SRAM (64 KiB) and leaves enough
room for global data, heap and stack (512 bytes assumed stack usage).
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
On SoCFPGA A10 systems, it can rarely happen that a reboot from Linux
will result in stale data in PL310 L2 cache controller. Even if the L2
cache controller is disabled via the CTRL register CTRL_EN bit, those
data can interfere with operation of devices using DMA, like e.g. the
DWMMC controller. This can in turn cause e.g. SPL to fail reading data
from SD/MMC.
The obvious solution here would be to fully reset the L2 cache controller
via the reset manager MPUMODRST L2 bit, however this causes bus hang even
if executed entirely from L1 I-cache to avoid generating any bus traffic
through the L2 cache controller.
This patch thus configures and enables the L2 cache controller very early
in the SPL boot process, clears the L2 cache and disables the L2 cache
controller again.
The reason for doing it in SPL is because we need to avoid accessing any
of the potentially stale data in the L2 cache, and we are certain any of
the stale data will be below the OCRAM address range. To further reduce
bus traffic during the L2 cache invalidation, we enable L1 I-cache and
run the invalidation code entirely out of the L1 I-cache.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dalon Westergreen <dwesterg@gmail.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Pull the PL310 clearing code into common code, so it can be reused
by Arria10.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dalon Westergreen <dwesterg@gmail.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
While converting CONFIG_SYS_[DI]CACHE_OFF to Kconfig, there are instances
where these configuration items are conditional on SPL. This commit adds SPL
variants of these configuration items, uses CONFIG_IS_ENABLED(), and updates
the configurations as required.
Acked-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Trevor Woerner <trevor@toganlabs.com>
[trini: Make the default depend on the setting for full U-Boot, update
more zynq hardware]
Signed-off-by: Tom Rini <trini@konsulko.com>
Re-add support for Aries Embedded MCV SoM, which is CycloneV based
and the associated MCVEVK and MCVEVP baseboard. The board can boot
from eMMC. Ethernet and USB is supported.
The Aries Embedded boards have been removed with commit 03b54997d5
("board/aries: Remove"). I will now take care of them.
The device-tree files are from mainline Linux commit e93c9c99a629
("Linux v5.1)".
Signed-off-by: Wolfgang Grandegger <wg@aries-embedded.de>
CC: Marek Vasut <marex@denx.de>
CC: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
The stratix 10 reset manager ad-hoc code in arch/arm contains an unused
function 'reset_deassert_peripherals_handoff' that has been added from
the beginning. As this is probably a result of copying the gen5 reset
manager and this function has never been used, remove it.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
commit c5de2b7eae ("arm: socfpga: implement proper peripheral reset")
has removed the call to 'reset_deassert_peripherals_handoff()' from
socfpga gen5 SPL since the reset driver now handles resets. However,
commit c1d4b464c8 ("ARM: socfpga: Disable bridges in SPL unless booting from FPGA")
has re-added this ad-hoc reset code, so that all peripherals were now
again enabled instead of letting the drivers enable them by request.
While at it, remove this function for gen5 as it should not be used.
Fixes: commit c1d4b464c8 ("ARM: socfpga: Disable bridges in SPL unless booting from FPGA")
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Add support for loading FPGA bitstream to get DDR up running before
U-Boot is loaded into DDR. Boot device initialization, generic firmware
loader and SPL FAT support are required for this whole mechanism to work.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
Add FPGA driver to support program FPGA with FPGA bitstream loading from
filesystem. The driver are designed based on generic firmware loader
framework. The driver can handle FPGA program operation from loading FPGA
bitstream in flash to memory and then to program FPGA.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
Send CONFIG_STATUS and RECONFIG_STATUS mailbox commands to Secure
Device Manager (SDM) to get the status of FPGA and make sure the
FPGA is in user mode before enable the bridge.
Signed-off-by: Ang, Chee Hong <chee.hong.ang@intel.com>
Software must never reset FPGA2SOC bridge. This bridge must only be
reset by POR/COLD/WARM reset. Asserting the FPGA2SOC reset by software
can cause the SoC to lock-up if there are traffics being drived into
FPGA2SOC bridge.
Signed-off-by: Ang, Chee Hong <chee.hong.ang@intel.com>
Convert Stratix 10 SDRAM driver to device model.
Get rid of call to socfpga_per_reset() and use reset
framework.
SPL is changed from calling function in SDRAM driver
directly to just probing UCLASS_RAM.
Move sdram_s10.h from arch to driver/ddr/altera directory.
Signed-off-by: Ley Foon Tan <ley.foon.tan@intel.com>
Find the UCLASS_CACHE driver to configure the cache controller's
settings.
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
The usage of socfpga_sdram_apply_static_cfg() seems rather dubious and
is confirmed to lead to a rare system hang when enabling bridges. This
patch removes the socfpga_sdram_apply_static_cfg() altogether, because
it's use seems unjustified and problematic.
The socfpga_sdram_apply_static_cfg() triggers write to SDRAM staticcfg
register to set the applycfg bit, which according to old vendor U-Boot
sources can only be written when there is no traffic between the SDRAM
controller and the rest of the system. Empirical measurements confirm
this, setting the applycfg bit when there is traffic between the SDRAM
controller and CPU leads to the SDRAM controller accesses being blocked
shortly after.
Altera originally solved this by moving the entire code which sets the
staticcfg register to OCRAM [1]. The commit message claims that the
applycfg bit needs to be set after write to fpgaportrst register. This
is however inverted by Altera shortly after in [2], where the order
becomes the exact opposite of what commit message [1] claims to be the
required order. The explanation points to a possible problem in AMP
use-case, where the FPGA might be sending transactions through the F2S
bridge.
However, the AMP is only the tip of the iceberg here. Any of the other
L2, L3 or L4 masters can trigger transactions to the SDRAM. It becomes
rather non-trivial to guarantee there are no transactions to the SDRAM
controller.
The SoCFPGA SDRAM driver always writes the applycfg bit in SPL. Thus,
writing the applycfg again in bridge enable code seems redundant and
can presumably be dropped.
[1] 75905816ec
[2] 8ba6986b04
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Add optional "mask" argument to the SoCFPGA bridge command, to select
which bridges should be enabled/disabled. This allows the user to avoid
enabling bridges which are not connected into the FPGA fabric. Default
behavior is to enable/disable all bridges.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Instead of just putting the bridges into reset, fully remove the bridges
from the L3 main bridge space when disabling them by clearing bits in
NIC-301 remap register. Moreover, only touch the 3 LSbits in brgmodrst
register as the rest of the bits are undefined.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Disable bridges between L3 Main switch and FPGA unless booting
from FPGA and keep them disabled to prevent glitches and possible
hangs of the L3 Main switch.
The current version of the code could have enabled the bridges
between the L3 Main switch and FPGA for a short period of time
in board_init_f() in case the FPGA was programmed and then again
disable them at the end of board_init_f(). Replace this with a
code which only sets up the handoff registers and let the user
enable the bridges later on.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
Factor out the code for programming preloader handoff register values,
the ISWGRP Handoff 0 and 1. These registers later control which bridges
are enabled by the "bridge" command on Gen5 devices.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
A mailbox command contains 1-u32 header + arguments. The "len" variable
only contains the length of the arguments, but not the 1-u32 header.
Include the length of header when checking the ring buffer space to
prevent off-by-one error.
Signed-off-by: Ley Foon Tan <ley.foon.tan@intel.com>
Signed-off-by: Chee Hong Ang <chee.hong.ang@intel.com>
By enabling debug prints in malloc_simple, we can see that SPL for socfpga
gen5 does by far not need the 8 KiB malloc pool currently allocated for
SPL in pre-reloc phase.
On socfpga_socrates, 1304 bytes are currently used (and this increases by
~200 bytes only for the sdram/reset fixes in socfpga-next).
To prevent wasting precious SRAM space, let's reduce the initial heap used
for SPL to 2 KiB. This is still some hundred bytes more than currently
used. Also, the gen5 SPL enables stack and heap in DDR memory pretty
early. Only the initial uclass/dm parsing, serial console and DDR
initialization is done in the initial heap, so these 2 KiB should be
enough for all boards.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Acked-by: Marek Vasut <marex@denx.de>
This commit moves common config options used in all socfpga boards
to select/imply in Kconfig. This both cleans up the defconfig files
as well as makes future changes easier.
Options implied/defaulted for all sub-arches:
- SPL, SPL_DM, USE_TINY_PRINTF, NR_DRAM_BANKS
Options implied/defaulted for implied for A10 & gen5:
- FPGA_SOCFPGA, SYS_MALLOC_F_LEN, SYS_TEXT_BASE
Options implied/defaulted for gen5:
- SPL_STACK_R, SPL_SYS_MALLOC_SIMPLE, SPL_STACK_R_ADDR
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Acked-by: Marek Vasut <marex@denx.de>
Scrub memory content if ECC is enabled and it is not
from warm reset boot.
Enable icache and dcache before scrub memory
and use "DC ZVA" instruction to clear memory
to zeros. This instruction writes a cache line
at a time and it can prevent false ECC error
trigger if write cache line partially.
Signed-off-by: Ley Foon Tan <ley.foon.tan@intel.com>
Move SDRAM size check to SDRAM driver. sdram_calculate_size()
is called in SDRAM initialization already, avoid calling
twice in size check function.
Signed-off-by: Ley Foon Tan <ley.foon.tan@intel.com>
This commit removes ad-hoc reset handling for peripheral resets from SPL
for socfpga gen5.
This is done because as U-Boot drivers support reset handling by now.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
To clean up reset handling for socfpga gen5, port the DDR driver to DM
using UCLASS_RAM and implement proper reset handling.
This gets us rid of one ad-hoc call to socfpga_per_reset().
The gen5 driver is implemented in 2 distinct files. One of it (containing
the calibration training) is not touched much and is kept at using
hard coded addresses since the code grows even more otherwise.
SPL is changed from calling hard into the DDR driver code to just
probing UCLASS_RESET and UCLASS_RAM. It is happy after finding a RAM
driver after that.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
The bootrom seems to leave the D-cache in messed up state, make sure
the SPL disables it so it can not interfere with operation.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Chin Liang See <chin.liang.see@intel.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Cc: Tien Fong Chee <tien.fong.chee@intel.com>
The values for the data and tag latency settings on the PL310 caches
controller is an (n-1). For example, the "arm,tag-latency" is specified
as <1 1 1>, so the values that should be written to register should be
0x000. And for the "arm,data-latency" specified as <2 1 1>, the register
value should be 0x010.
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
On SoCFPGA Gen5 systems, it can rarely happen that a reboot from Linux
will result in stale data in PL310 L2 cache controller. Even if the L2
cache controller is disabled via the CTRL register CTRL_EN bit, those
data can interfere with operation of devices using DMA, like e.g. the
DWMMC controller. This can in turn cause e.g. SPL to fail reading data
from SD/MMC.
The obvious solution here would be to fully reset the L2 cache controller
via the reset manager MPUMODRST L2 bit, however this causes bus hang even
if executed entirely from L1 I-cache to avoid generating any bus traffic
through the L2 cache controller.
This patch thus configures and enables the L2 cache controller very early
in the SPL boot process, clears the L2 cache and disables the L2 cache
controller again.
The reason for doing it in SPL is because we need to avoid accessing any
of the potentially stale data in the L2 cache, and we are certain any of
the stale data will be below the OCRAM address range. To further reduce
bus traffic during the L2 cache invalidation, we enable L1 I-cache and
run the invalidation code entirely out of the L1 I-cache.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Dalon Westergreen <dwesterg@gmail.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Configure the PL310 tag and data latency registers, which slightly
improves performance and aligns the behavior with Linux.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Dalon Westergreen <dwesterg@gmail.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
The 'dwmac_socfpga' ETH driver can now get the MACs out of reset
via the socfpga reset driver and can set PHY mode via syscon.
This means we can now remove the ad-hoc code to do this from
arch/arm/mach-socfpga.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Replace CONFIG_SPL_EXT_SUPPORT to CONFIG_SPLY_FS_EXT4 so both
obj-$(CONFIG_$(SPL_)FS_EXT4) and CONFIG_IS_ENABLED(FS_EXT4) can be
used to control the build in both SPL and U-Boot.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Replace CONFIG_SPL_FAT_SUPPORT with CONFIG_SPL_FS_FAT so
obj-$(CONFIG_$(SPL_)FS_FAT) can be used to control the build in both
SPL and U-Boot.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Enable 'fpga' command in u-boot. User will be able to use the FPGA
command to program the FPGA on Stratix10 SoC.
Signed-off-by: Ang, Chee Hong <chee.hong.ang@intel.com>
Add a generic mailbox API for FPGA reconfig status which can be
called by others. This new function accepts 2 different mailbox
commands: CONFIG_STATUS or RECONFIG_STATUS.
Signed-off-by: Ang, Chee Hong <chee.hong.ang@intel.com>
