As part of disabling caches MMU as well gets disabled. But MMU is not
available on all armv7 cores like R5F. So disable MMU only if it is
available.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
'make tests' on a 32bit ARM system leads to
In file included from ../common/cli_hush.c:79:
../include/malloc.h:364:7: error: conflicting types for ‘memset’
void* memset(void*, int, size_t);
^~~~~~
In file included from ../include/compiler.h:126,
from ../include/env.h:12,
from ../common/cli_hush.c:78:
../include/linux/string.h:103:15:
note: previous declaration of ‘memset’ was here
extern void * memset(void *,int,__kernel_size_t);
^~~~~~
In file included from ../common/cli_hush.c:79:
../include/malloc.h:365:7: error: conflicting types for ‘memcpy’
void* memcpy(void*, const void*, size_t);
^~~~~~
In file included from ../include/compiler.h:126,
from ../include/env.h:12,
from ../common/cli_hush.c:78:
../include/linux/string.h:106:15:
note: previous declaration of ‘memcpy’ was here
extern void * memcpy(void *,const void *,__kernel_size_t);
^~~~~~
According to the U-Boot coding style guide common.h should be the first
include.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When full malloc is enabled and SYS_MALLOC_F is also enabled, the simple
pre-reloc heap is used before relocation. In this case, calloc() uses
the MALLOC_ZERO macro to zero out the allocated memory. However, since
this macro is specially crafted for the dlmalloc implementation, it
does not always work for simple malloc.
For example, when allocating 16 bytes via simple malloc, only the first
12 bytes get zeroed out. The last 4 bytes will remain untouched.
This is a problem for DM drivers that are allocated before relocation:
memory allocated via 'platdata_auto_alloc_size' might not be set to
zero, resulting in bogus behaviour.
To fix this, use 'memset' instead of 'MALLOC_ZERO' to zero out memory
that compes from simple malloc.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
As SPL_SPI_FLASH_SUPPORT cannot work without SPL_SPI_SUPPORT, fix
dependencies to prevent enabling SPI flash support without basic SPI
support.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Reviewed-by: Lukasz Majewski <lukma@denx.de>
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>
This patch changes ERR_PTR/PTR_ERR to use CONFIG_ERR_PTR_OFFSET to map
errno values into a pointer region that cannot contain valid pointers.
IS_ERR and IS_ERR_OR_NULL have to be converted to use PTR_ERR, too,
for this to work.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Some U-Boot pointers have redundant information, so we can use a scheme
where we can return either an error code or a pointer with the same
return value. The default implementation just casts the pointer to a
number, however, this may fail on platforms where the end of the address
range is used for valid pointers (e.g. 0xffffff00 is a valid heap pointer
in socfpga SPL). For such platforms, this value provides an upper range
of those error pointer values - up to 'MAX_ERRNO' bytes below this value
must be unused/invalid addresses.
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Provide optimized memcpy_{from,to}io() and memset_io(). This is required
when moving large amount of data to and from IO regions such as IP
registers or accessing memory mapped flashes.
Code is borrowed from Linux Kernel v5.4.
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
API_dev_write(va_list ap) is currently lacking the write support
to storage devices because, historically, those devices did not
implement block_write()
The solution has been tested by loading and booting a (patched)
GRUB instance in a QEMU vexpress-a9 environment. The disk write
operations were triggered with GRUB's save_env command.
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@gmail.com>
Now that we have again fixed the problems that building with clang
exposes, enable these tests on Azure and GitLab-CI as well.
Signed-off-by: Tom Rini <trini@konsulko.com>
When using the OVERRIDE variable we need to pass -O to buildman as well
to use the "override" option to buildman.
Fixed: e9500f49ea ("travis: Use buildman for building with clang")
Signed-off-by: Tom Rini <trini@konsulko.com>
Newer versions of LLVM-7 will provide an unused function warning over
console_doenv() in the case of SYS_CONSOLE_IS_IN_ENV not being enabled
as can be the case in SPL. Add guards around this function.
Signed-off-by: Tom Rini <trini@konsulko.com>
imx targets are defined in arch/arm/mach-imx/Makefile.
Some of them are dependent on targets defined in main
Makefile. For the Makefile in arch/arm/mach-imx this
targets must be finished before the imx targets are
build, if not you get for example the error:
make -f /home/hs/abb/mainlining/u-boot/scripts/Makefile.build obj=arch/arm/mach-imx u-boot-dtb.imx
make[2]: *** No rule to make target 'u-boot-fit-dtb.bin', needed by 'u-boot-dtb.imx'. Stop.
make[1]: *** [/home/hs/abb/mainlining/u-boot/Makefile:1123: u-boot-dtb.imx] Error 2
make[1]: *** Waiting for unfinished jobs....
make[1]: Leaving directory '/work/hs/compile/u-boot/aristainetos2_defconfig'
make: *** [Makefile:148: sub-make] Error 2
compile failed
In above case of CONFIG_MULTI_DTB_FIT is defined, the
u-boot-dtb.imx is dependent on the u-boot-fit-dtb.bin
which may is not build yet ...
I could reproduce this error on a travis build also if
I build an out-of-tree build on a local machine with a
build directory on a "slow" slow storage device. If
building the same source target with a build dir on a
fast storage device, the build works.
I found no solution to tell the arch/arm/mach-imx/Makefile
to find the targets in main Makefile, if there is a way
this would be the better fix.
I solved it by adding a IMX_DEPS var, which holds a list
of main u-boot targets, which must be finished, before
calling imx targets and fixed the build for imx
targets which enabled CONFIG_MULTI_DTB_FIT.
I think it is just luck, that imx targets with
CONFIG_OF_SEPARATE enabled build, because the
u-boot-dtb.imx target depends on u-boot-dtb.bin
which gets build early enough before starting with
u-boot-dtb.imx. May this targets should be fixed too.
Signed-off-by: Heiko Schocher <hs@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
add DM PCI support on the socrates board.
use PCIE_FSL now.
Signed-off-by: Heiko Schocher <hs@denx.de>
Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
add socrates device tree from linux:
commit 71ae5fc87c34 ("Merge tag 'linux-kselftest-5.2-rc1' of
git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest")
and added SPDX license identifier.
Did not fix checkpatch warnings:
arch/powerpc/dts/socrates.dts:235: check: Please don't use multiple blank lines
arch/powerpc/dts/socrates.dts:238: error: code indent should use tabs where possible
Also, add me as board maintainer.
Signed-off-by: Heiko Schocher <hs@denx.de>
Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
add DM based PCI Configuration space access support for
MPC85xx PCI Bridge. This driver is based on
arch/powerpc/cpu/mpc85xx/pci.c
In the old driver there is a fix for a hw issue on the
TARGET_MPC8555CDS and TARGET_MPC8541CDS boards. As I
have no such hardware I did not port this part.
Signed-off-by: Heiko Schocher <hs@denx.de>
Reviewed-by: Priyanka Jain <priyanka.jain@nxp.com>
With python3 we're now tripping over a long-standing problem with how we
add to the buildman file with some toolchains. We cannot have multiple
toolchain-alias sections as that leads to a parse error.
Signed-off-by: Tom Rini <trini@konsulko.com>
To enable HS400(ES) and UHS for imx8m platforms, update the driver data
to share with imx8qm esdhc_soc_data.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
During mmc initialization, there are several calls to mmc_set_clock
and mmc_set_ios. When mmc_power_off, the mmc->clock will be set,
but the imx driver will use 400KHz. So the following calls
to mmc_set_ios will set the clock several times which is redundant
in fsl_esdhc_imx driver. So let's simplify to remove redundant
clock settings.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
The QorIQ eSDHC on all platforms supports checking write protect
state through register bit. So check it always.
Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
The fsl_esdhc_init() was actually to get configuration of mmc_config.
So rename it to fsl_esdhc_get_cfg_common() and make it common for both
DM_MMC and non-DM_MMC.
Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
This patch is to clean up bus width setting code.
- For DM_MMC, remove getting "bus-width" from device tree.
This has been done in mmc_of_parse().
- For non-DM_MMC, move bus width configuration from fsl_esdhc_init()
to fsl_esdhc_initialize() which is non-DM_MMC specific.
And fix up bus width configuration to support only 1-bit, 4-bit,
or 8-bit. Keep using 8-bit if it's not set because many platforms
use driver without providing max bus width.
- Remove bus_width member from fsl_esdhc_priv structure.
Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Voltage validation should be done by CMD8. Current comparison between
mmc_cfg voltages and host voltage capabilities is meaningless.
So drop current comparison and let voltage validation is through CMD8.
Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Controller initialization is not needed in fsl_esdhc_init().
It will be done in esdhc_init() for non-DM_MMC, and in
esdhc_init_common() in probe for DM_MMC.
Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Update this tool to use Python 3 to meet the 2020 deadline.
Unfortunately this introduces a test failure due to a problem in pylibfdt
on Python 3. I will investigate.
Signed-off-by: Simon Glass <sjg@chromium.org>
When preparing to possible expand or contract an entry we reset the size
to the original value from the binman device-tree definition, which is
often None.
This causes binman to forget the original size of the entry. Remember this
so that it can be used when needed.
Signed-off-by: Simon Glass <sjg@chromium.org>
This function should use a void * type, not char *. This causes an error:
TypeError: in method 'fdt_property_stub', argument 3 of type 'char const *'
Fix it.
Signed-off-by: Simon Glass <sjg@chromium.org>
