| Message ID | 20210827205718.536008-1-aford173@gmail.com |
|---|---|
| State | Deferred |
| Delegated to: | Tom Rini |
| Headers | show |
| Series | [1/2] arm: mach-rmobile: Move cpu_reset to rzg2-reset.c | expand |
On 8/27/21 10:57 PM, Adam Ford wrote: > Various RZ/G2[MNH] boards have copies of cpu_reset which > creates duplicate code, so move this to a common file shared > among all users of the RZG2 and declare it weak so it can be > overwritten if necessary. RCar3 has the same reset code too, so this could be unified further. However, I wonder, can't you simply call the PSCI reset ? There is a PSCI reset driver in drivers/reset/ too I think.
On Sun, Aug 29, 2021 at 12:53 PM Marek Vasut <marex@denx.de> wrote: > > On 8/27/21 10:57 PM, Adam Ford wrote: > > Various RZ/G2[MNH] boards have copies of cpu_reset which > > creates duplicate code, so move this to a common file shared > > among all users of the RZG2 and declare it weak so it can be > > overwritten if necessary. > > RCar3 has the same reset code too, so this could be unified further. > However, I wonder, can't you simply call the PSCI reset ? There is a > PSCI reset driver in drivers/reset/ too I think. Thanks for the suggestion. I removed the custom cpu_reset code from the board file, and enabled sysreset and the psci, and that appeared to still reset when needed. I'll send a different patch shortly. adam
On 8/30/21 7:52 PM, Adam Ford wrote: > On Sun, Aug 29, 2021 at 12:53 PM Marek Vasut <marex@denx.de> wrote: >> >> On 8/27/21 10:57 PM, Adam Ford wrote: >>> Various RZ/G2[MNH] boards have copies of cpu_reset which >>> creates duplicate code, so move this to a common file shared >>> among all users of the RZG2 and declare it weak so it can be >>> overwritten if necessary. >> >> RCar3 has the same reset code too, so this could be unified further. >> However, I wonder, can't you simply call the PSCI reset ? There is a >> PSCI reset driver in drivers/reset/ too I think. > > Thanks for the suggestion. I removed the custom cpu_reset code from > the board file, and enabled sysreset and the psci, and that appeared > to still reset when needed. > > I'll send a different patch shortly. Thanks!
diff --git a/arch/arm/mach-rmobile/Makefile b/arch/arm/mach-rmobile/Makefile index 195bbeb5c8..2bd173ec3a 100644 --- a/arch/arm/mach-rmobile/Makefile +++ b/arch/arm/mach-rmobile/Makefile @@ -13,7 +13,7 @@ obj-$(CONFIG_SH73A0) += lowlevel_init.o cpu_info-sh73a0.o pfc-sh73a0.o obj-$(CONFIG_R8A7740) += lowlevel_init.o cpu_info-r8a7740.o pfc-r8a7740.o obj-$(CONFIG_RCAR_GEN2) += lowlevel_init_ca15.o cpu_info-rcar.o obj-$(CONFIG_RCAR_GEN3) += lowlevel_init_gen3.o cpu_info-rcar.o memmap-gen3.o -obj-$(CONFIG_RZ_G2) += cpu_info-rzg.o +obj-$(CONFIG_RZ_G2) += cpu_info-rzg.o rzg2-reset.o ifneq ($(CONFIG_R8A779A0),) obj-$(CONFIG_ARMV8_PSCI) += psci-r8a779a0.o diff --git a/arch/arm/mach-rmobile/rzg2-reset.c b/arch/arm/mach-rmobile/rzg2-reset.c new file mode 100644 index 0000000000..8c3608f34a --- /dev/null +++ b/arch/arm/mach-rmobile/rzg2-reset.c @@ -0,0 +1,26 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Copyright (C) 2021 Renesas Electronics Corporation + */ + +#include <asm/io.h> + +#define RST_BASE 0xE6160000 +#define RST_CA57RESCNT (RST_BASE + 0x40) +#define RST_CA53RESCNT (RST_BASE + 0x44) +#define RST_CA57_CODE 0xA5A5000F +#define RST_CA53_CODE 0x5A5A000F + +__weak void reset_cpu(void) +{ + unsigned long midr, cputype; + + asm volatile("mrs %0, midr_el1" : "=r" (midr)); + cputype = (midr >> 4) & 0xfff; + + if (cputype == 0xd03) + writel(RST_CA53_CODE, RST_CA53RESCNT); + else + writel(RST_CA57_CODE, RST_CA57RESCNT); +} +
Various RZ/G2[MNH] boards have copies of cpu_reset which creates duplicate code, so move this to a common file shared among all users of the RZG2 and declare it weak so it can be overwritten if necessary. Signed-off-by: Adam Ford <aford173@gmail.com>