In case the iMX8M boot from eMMC boot partition and the primary image
is corrupted, the BootROM is capable of starting a secondary image in
the other eMMC boot partition as a fallback.
However, the BootROM leaves the eMMC BOOT_PARTITION_ENABLE setting as
it was, i.e. pointing to the boot partition containing the corrupted
image, and the BootROM does not provide any indication that this sort
of fallback occured.
According to AN12853 i.MX ROMs Log Events, Rev. 0, May 2020, it is
possible to determine whether fallback event occurred by parsing the
ROM event log. In case ROM event ID 0x51 is present, fallback event
did occur.
This patch implements ROM event log parsing and search for event ID
0x51 for all iMX8M SoCs, and based on that corrects the eMMC boot
partition selection. This way, the SPL loads the remaining boot
components from the same eMMC boot partition from which it was
started, even in case of the fallback.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Faiz Abbas <faiz_abbas@ti.com>
Cc: Harald Seiler <hws@denx.de>
Cc: Lokesh Vutla <lokeshvutla@ti.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Ye Li <ye.li@nxp.com>
Some IPs have their accessible address space restricted by the
interconnect. Let's make sure U-Boot only ever uses the space below
the 4G address boundary (which is 3GiB big), even when the effective
available memory is bigger.
We implement board_get_usable_ram_top() for all i.MX8M SoCs, as the
whole family is affected by this.
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
On ARM, the gd pointer is stored in registers r9 / x18. For this the
-ffixed-r9 / -ffixed-x18 flag is used when compiling, but using global
register variables causes errors when building with LTO, and these
errors are very difficult to overcome.
Richard Biener says [1]:
Note that global register vars shouldn't be used with LTO and if they
are restricted to just a few compilation units the recommended fix is
to build those CUs without -flto.
We cannot do this for U-Boot since all CUs use -ffixed-reg flag.
It seems that with LTO we could in fact store the gd pointer differently
and gain performance or size benefit by allowing the compiler to use
r9 / x18. But this would need more work.
So for now, when building with LTO, go the clang way, and instead of
declaring gd a global register variable, we make it a function call via
macro.
[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68384
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
The OCRAM_S is regular memory, just like the OCRAM, add it to the MMU
tables so it can be used and cached.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
i.MX8MQ B2 is using same value in OCOTP_READ_FUSE_DATA like B1, so
we have to check the ROM verision to distinguish the revision.
As we have checked the B1 rev for sticky bits work around in
secure boot. So it won't apply on B2.
Signed-off-by: Ye Li <ye.li@nxp.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
For dual core and single core iMX8M parts, the thermal node and PMU node
in kernel DTB also needs update to remove the refers to deleted core nodes.
Otherwise both driver will fail to work.
Signed-off-by: Ye Li <ye.li@nxp.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
There are 3 part numbers for 11x11 i.MX8MNano with different core number
configuration: UltraLite Quad/Dual/Solo
Comparing with i.MX8MN Lite parts, they have MIPI DSI disabled. So
checking the MIPI DSI disable fuse to recognize these parts.
Signed-off-by: Ye Li <ye.li@nxp.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Add dedicated defconfigs for iMX8MN low drive mode which set the VDD_SOC
and VDD_DRAM to 0.8v, DDR at 1600MTS (800Mhz clock) and GPU at 200Mhz.
Signed-off-by: Ye Li <ye.li@nxp.com>
Acked-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Historically, the reset_cpu() function had an `addr` parameter which was
meant to pass in an address of the reset vector location, where the CPU
should reset to. This feature is no longer used anywhere in U-Boot as
all reset_cpu() implementations now ignore the passed value. Generic
code has been added which always calls reset_cpu() with `0` which means
this feature can no longer be used easily anyway.
Over time, many implementations seem to have "misunderstood" the
existence of this parameter as a way to customize/parameterize the reset
(e.g. COLD vs WARM resets). As this is not properly supported, the
code will almost always not do what it is intended to (because all
call-sites just call reset_cpu() with 0).
To avoid confusion and to clean up the codebase from unused left-overs
of the past, remove the `addr` parameter entirely. Code which intends
to support different kinds of resets should be rewritten as a sysreset
driver instead.
This transformation was done with the following coccinelle patch:
@@
expression argvalue;
@@
- reset_cpu(argvalue)
+ reset_cpu()
@@
identifier argname;
type argtype;
@@
- reset_cpu(argtype argname)
+ reset_cpu(void)
{ ... }
Signed-off-by: Harald Seiler <hws@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Instead of hard-coding the offset of DRAM entries in MMU table all over
the code, auto-detect the offset. This removes error-prone code which
would break e.g. in case the MMU table is modified.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Ye Li <ye.li@nxp.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Move this out of the common header and include it only where needed. In
a number of cases this requires adding "struct udevice;" to avoid adding
another large header or in other cases replacing / adding missing header
files that had been pulled in, very indirectly. Finally, we have a few
cases where we did not need to include <asm/global_data.h> at all, so
remove that include.
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
the mem_map structure containing the size of SDRAM is used in various
cache functions in cache_v8.c thus we need to update it with the
sdram size the board is configured with as well. Without this
the cache functions do not get setup properly and can hang
in the case where a board reports more SDRAM than defined in
PHYS_SDRAM_SIZE.
Signed-off-by: Tim Harvey <tharvey@gateworks.com>
Latest datasheet revE has removed MIMX8ML7D/5D/7C/5C parts, so
update u-boot to remove decoding and support for those parts.
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Checkpatch reports the following issue:
ERROR: space prohibited before that ',' (ctx:WxW)
#936: FILE: arch/arm/mach-imx/imx8m/soc.c:936:
+ 0, 0 , 0, 0, 0, 0, &res);
Remove the unneeded space.
^
Reported-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
The Linux coding style guide (Documentation/process/coding-style.rst)
clearly says:
It's a **mistake** to use typedef for structures and pointers.
Besides, using typedef for structures is annoying when you try to make
headers self-contained.
Let's say you have the following function declaration in a header:
void foo(bd_t *bd);
This is not self-contained since bd_t is not defined.
To tell the compiler what 'bd_t' is, you need to include <asm/u-boot.h>
#include <asm/u-boot.h>
void foo(bd_t *bd);
Then, the include direcective pulls in more bloat needlessly.
If you use 'struct bd_info' instead, it is enough to put a forward
declaration as follows:
struct bd_info;
void foo(struct bd_info *bd);
Right, typedef'ing bd_t is a mistake.
I used coccinelle to generate this commit.
The semantic patch that makes this change is as follows:
<smpl>
@@
typedef bd_t;
@@
-bd_t
+struct bd_info
</smpl>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Current codes assume the OPTEE address is at the end of first DRAM bank.
Adjust the process to allow OPTEE in the middle of first bank.
When OPTEE memory is removed from first bank, it may split the first bank
to two banks, adjust the MMU table for the split case,
Since the default CONFIG_NR_DRAM_BANKS is 4, it is enough, just enlarge
i.MX8MP evk to default to avoid issue.
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Silvano di Ninno <silvano.dininno@nxp.com>
Tested-by: Silvano di Ninno <silvano.dininno@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
To fused part, we need to disable nodes of dtb to let kernel boot.
To mfgtool, USB issue when using super-speed for mfgtool, temporally
work around the problem to use high-speed only.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
To use one defconfig for all boot device, we have to runtime set
env offset and return env medium according to the boot device.
This patch overrides the env_get_offset and env_get_location to
implement the feature.
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
iMX8MP has 6 fused parts in each qualification tier, with core, VPU,
ISP, NPU or DSP fused respectively.
The configuration tables for enabled modules:
MIMX8ML8DVNLZAA Quad Core, VPU, NPU, ISP, DSP
MIMX8ML7DVNLZAA Quad Core, NPU, ISP
MIMX8ML6DVNLZAA Quad Core, VPU, ISP
MIMX8ML5DVNLZAA Quad Core, VPU
MIMX8ML4DVNLZAA Quad Lite
MIMX8ML3DVNLZAA Dual Core, VPU, NPU, ISP, DSP
Add the support in U-Boot
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
ROM SError happens on two cases:
1. ERR050342, on iMX8MQ HDCP enabled parts ROM writes to GPV1 register, but
when ROM patch lock is fused, this write will cause SError.
2. ERR050350, on iMX8MQ/MM/MN, when the field return fuse is burned, HAB
is field return mode, but the last 4K of ROM is still protected and cause
SError.
Since ROM mask SError until ATF unmask it, so then ATF always meets the
exception. This patch works around the issue in SPL by enabling SPL
Exception vectors table and the SError exception, take the exception
to eret immediately to clear the SError.
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
So far u-boot only load SNSR25C for TMU main probe (probe 0). However,
kernel enables two probes. So it also needs to set default SNSR25C of
TCALIV1 for blank samples.
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
i.MX8MP thermal which has two probes and supports temperature range
from -40 to 125. The driver still uses default 1p HW
calibration at 25C and loads calibration parameters from fuse.
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Like iMX8MM, iMX8MN also needs SW to load TMU TASR and TCALIV registers
value from fuse before enabling TMU calibration. Otherwise the calibration
is not exact.
Reviewed-by: Anson Huang <Anson.Huang@nxp.com>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
On iMX8MM, the default value of TMU registers TCALIV and TASR need
be loaded from fuse. HW won't do this, it expect SW loads them before
using TMU.
Reviewed-by: Bai Ping <ping.bai@nxp.com>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
imx8m has the only implementation of reset_cpu() which does not ignore
the addr parameter and instead gives it some meaning as the base address
of watchdog registers. This breaks convention with the rest of U-Boot
where the parameter is ignored and callers are passing in 0.
Fixes: d2041725e8 ("imx8m: restrict reset_cpu")
Co-developed-by: Harald Seiler <hws@denx.de>
Signed-off-by: Harald Seiler <hws@denx.de>
Signed-off-by: Claudius Heine <ch@denx.de>
The SPL can also be compiled with sysreset drivers just fine, so
update the condition to cater for that option.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: Flavio Suligoi <f.suligoi@asem.it>
Cc: Harald Seiler <hws@denx.de>
Cc: Igor Opaniuk <igor.opaniuk@toradex.com>
Cc: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Cc: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
If the cpu-idle-states is missing from the DT in the first place, do
not fail on removing in. Just move on and do not even print an error,
since not being able to remove something which is not there in the
first place is not an error and surely does not justify failing to
boot.
Turn the surrounding prints into debugs to reduce the useless noise.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Frieder Schrempf <frieder.schrempf@kontron.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
imx8m has the only implementation of `reset_cpu` which does not ignore
the addr parameter and instead gives it some meaning as the base address
of watchdog registers. This breaks convention with the rest of U-Boot
where the parameter is ignored and callers are passing in 0.
Fixes: d2041725e8 ("imx8m: restrict reset_cpu")
Co-Authored-by: Harald Seiler <hws@denx.de>
Signed-off-by: Claudius Heine <ch@denx.de>
Signed-off-by: Harald Seiler <hws@denx.de>
Reviewed-by: Marek Vasut <marex@denx.de>
When doing reset_cpu, in normal case the WDOG_B outputs immediately
after we clean WDA bit. But on mscale, the WDOG_B may be later than
internal reset, and cause PMIC not reset. As we enabled the SD3.0
support, the PMIC must be reset to reset SD card.
Change the reset_cpu to enable the WDOG_B for timeout as well, and set
WDOG timeout to 1s.
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Acked-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Ye Li <ye.li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Acquire ATF commit hash when booting U-Boot to make user easy
to know the ATF version.
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Set the BYPASS ID SWAP bit (GPR10 bit 1) in order for GPU not to
generated AXI bus errors with TZC380 enabled.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Recently the version check was improved to be able to determine that
we're running on SoC revision 2.1. A check for B0 was tightened so
that it now must equal 0x20 instead of being bigger than 0x20. On
some B0 chips the value returned is 0x1020 instead of 0x20. This
means even though it's B0, the check will fail and code relying on
the correct chip revision will make wrong decisions. There is no
documentation of those bits, but it seems that NXP always uses a
byte to encode the revision. Thus remove the upper bits to fix the
regression.
Signed-off-by: Patrick Wildt <patrick@blueri.se>
These functions belong in cpu_func.h. Another option would be cache.h
but that code uses driver model and we have not moved these cache
functions to use driver model. Since they are CPU-related it seems
reasonable to put them here.
Move them over.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
When CONFIG_$(SPL_)CLK not defined, the clock controller device
not exist, so to avoid boot failure for platform not have
CONFIG_$(SPL_)CLK, add a check.
Reviewed-by: Patrick Wildt <patrick@blueri.se>
Tested-by: Patrick Wildt <patrick@blueri.se>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Set the BYPASS ID SWAP bit (GPR10 bit 1) in order for GPU not to
generated AXI bus errors with TZC380 enabled.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
The i.MX8MQ B1 uses OCOTP_HW_OCOTP_READ_FUSE_DATA register for chip id.
It returns a magic number 0xff0055aa. update get_cpu_rev to support it,
and enable ocotp clock to access ocotp.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
CONFIG_SECURE_BOOT is too generic and forbids to use it for cross
architecture purposes. If Secure Boot is required for imx, this means to
enable and use the HAB processor in the soc.
Signed-off-by: Stefano Babic <sbabic@denx.de>
