linux system will die if we offline one of the cpu on R40 based board:
eg: echo 0 > /sys/devices/system/cpu/cpu3/online
The reason is that the R40 version of sunxi_cpu_set_power always passes
0 for the CPU number, so we turn off CPU0, regardless of what CPU the
CPU_OFF request came for.
Fix this by passing the proper CPU number, as there are proper power
clamp registers for every of the four cores.
Signed-off-by: qianfan Zhao <qianfanguijin@163.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Since COUNTER_FREQUENCY is obselete, so set cntfrq_el0 if
CONFIG_COUNTER_FREQUENCY is valid
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
Move this header out of the common header. Network support is used in
quite a few places but it still does not warrant blanket inclusion.
Note that this net.h header itself has quite a lot in it. It could be
split into the driver-mode support, functions, structures, checksumming,
etc.
Signed-off-by: Simon Glass <sjg@chromium.org>
Using memcpy() for some MMIO access is generally frowned upon and might
break things on some platforms. Allwinner H3, which fails to boot, being
an example here.
Moreover, fields being accessed are naturally aligned and warnings
produced by GCC have been quiesced for some time already by:
53dc8ae ("gcc-9: silence 'address-of-packed-member' warning")
That said, it should be okay to revert this commit.
This reverts commit 9bd34a69a4.
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Tomasz Duszynski <tduszyns@gmail.com>
Acked-by: Jagan Teki <jagan@amarulasolutions.com>
Compiling with GCC 9.2.1 leads to build errors:
arch/arm/cpu/armv7/sunxi/psci.c: In function ‘sunxi_cpu_set_power’:
arch/arm/cpu/armv7/sunxi/psci.c:144:21: error: taking address of packed
member of ‘struct sunxi_cpucfg_reg’ may result in an unaligned pointer
value [-Werror=address-of-packed-member]
144 | sunxi_power_switch(&cpucfg->cpu1_pwr_clamp, &cpucfg->cpu1_pwroff,
| ^~~~~~~~~~~~~~~~~~~~~~~
arch/arm/cpu/armv7/sunxi/psci.c:144:46: error: taking address of packed
member of ‘struct sunxi_cpucfg_reg’ may result in an unaligned pointer
value [-Werror=address-of-packed-member]
144 | sunxi_power_switch(&cpucfg->cpu1_pwr_clamp, &cpucfg->cpu1_pwroff,
| ^~~~~~~~~~~~~~~~~~~~
Use memcpy() and void* pointers to resolve the problem caused by packing
the struct sunxi_cpucfg_reg.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Acked-by: Jagan Teki <jagan@amarulasolutions.com>
This patch solves the following warnings:
arch/arm/mach-stm32mp/psci.c:
warning: no previous prototype for ‘psci_set_state’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_arch_cpu_entry’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_features’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_version’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_affinity_info’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_migrate_info_type’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_cpu_on’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_cpu_off’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_system_reset’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_system_off’ [-Wmissing-prototypes]
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
Currently we set the entry address in the psci_cpu_on function.
However R40 has a different register for this. This resulted in
an #ifdef / #else block in psci_cpu_on, which we avoided having
in the first place.
Move this part into a separate function, defined differently for
the R40 as opposed to the other single cluster platforms.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Reviewed-by: Jagan Teki <jagan@amarulasolutions.com>
The R40's CPU controls are a combination of sun6i and sun7i.
All controls are in the CPUCFG block, and it seems the R40 does not
have a PRCM block. The core reset, power gating and clamp controls
are grouped like sun6i.
Last, the R40 does not have a secure SRAM block.
This patch adds a PSCI implementation for CPU bring-up and hotplug
for the R40.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Many ARMv8 boards define a constant COUNTER_FREQUENCY to specify the
frequency of the ARM Generic Timer (aka. arch timer).
ARMv7 boards traditionally used CONFIG_TIMER_CLK_FREQ for the same
purpose. It seems useful to unify them.
Since there are less occurences of the latter version, lets convert all
users over to COUNTER_FREQUENCY.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jagan Teki <jagan@openedev.com>
As part of testing booting Linux kernels on Rockchip devices, it was
discovered by Ziyuan Xu and Sandy Patterson that we had multiple and for
some cases incomplete isb definitions. This was causing a failure to
boot of the Linux kernel.
In order to solve this problem as well as cover any corner cases that we
may also have had a number of changes are made in order to consolidate
things. First, <asm/barriers.h> now becomes the source of isb/dsb/dmb
definitions. This however introduces another complexity. Due to
needing to build SPL for 32bit tegra with -march=armv4 we need to borrow
the __LINUX_ARM_ARCH__ logic from the Linux Kernel in a more complete
form. Move this from arch/arm/lib/Makefile to arch/arm/Makefile and add
a comment about it. Now that we can always know what the target CPU is
capable off we can get always do the correct thing for the barrier. The
final part of this is that need to be consistent everywhere and call
isb()/dsb()/dmb() and NOT call ISB/DSB/DMB in some cases and the
function names in others.
Reviewed-by: Stephen Warren <swarren@nvidia.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
Acked-by: Ziyuan Xu <xzy.xu@rock-chips.com>
Acked-by: Sandy Patterson <apatterson@sightlogix.com>
Reported-by: Ziyuan Xu <xzy.xu@rock-chips.com>
Reported-by: Sandy Patterson <apatterson@sightlogix.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Now that we have a secure data section and space to store per-CPU target
PC address, switch to it instead of storing the target PC on the stack.
Also save clobbered r4-r7 registers on the stack and restore them on
return in psci_cpu_on for Tegra, i.MX7, and LS102xA platforms.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
sunxi and i.mx7 both define the __secure modifier to put functions in
the secure section. Move this to a common place.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
psci_text_end was used to calculate the PSCI stack address following the
secure monitor text. Now that we have an explicit secure stack section,
this is no longer used.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
This patch finishes the rewrite of sunxi specific PSCI parts into C
code.
The assembly-only stack setup code has been factored out into a common
function for ARMv7. The GIC setup code can be renamed as psci_arch_init.
And we can use an empty stub function for psci_text_end.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
To make the PSCI backend more maintainable and easier to port to newer
SoCs, rewrite the current PSCI implementation in C.
Some inline assembly bits are required to access coprocessor registers.
PSCI stack setup is the only part left completely in assembly. In theory
this part could be split out of psci_arch_init into a separate common
function, and psci_arch_init could be completely in C.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
