Message ID | 20240520-am62p-fdt-fix-v1-1-49845dcb3667@ti.com |
---|---|
State | Accepted |
Commit | 44b1d2210f0aa062c45e39860ebdcdbb22c84933 |
Delegated to: | Tom Rini |
Headers | show |
Series | arm: mach-k3: am62p: Fixup TF-A/OP-TEE reserved-memory node in FDT | expand |
On 5/23/24 11:43 AM, Bryan Brattlof wrote: > The address we load TFA and OPTEE is configurable by the > CONFIG_K3_{ATF,OPTEE)_LOAD_ADDR, but the DT node reservations remain > static which can cause some confusion about where exactly these firmware > are exactly. Fix this by updating the reserved-memory{} nodes when the > loaded address does not match the address in DT. > > Reported-by: Andrew Davis <afd@ti.com> > Signed-off-by: Bryan Brattlof <bb@ti.com> > --- > Hello everyone, > > This is a little fixup to avoid any confusion once we're in the kernel. > Because TF-A can be configured in U-Boot to be anywhere we want, we need > up update the reserved-memory{} node with this change. > > Thanks for reviewing > ~Bryan > --- > arch/arm/mach-k3/Makefile | 1 + > arch/arm/mach-k3/am62p5_fdt.c | 16 ++++++++++++++++ You'll want to rebase this on -next, these _fdt.c files all got moved into directories for each SoC. Andrew > arch/arm/mach-k3/am62px/Kconfig | 1 + > 3 files changed, 18 insertions(+) > > diff --git a/arch/arm/mach-k3/Makefile b/arch/arm/mach-k3/Makefile > index 1bd523329a4f8..4e9d0925f13f5 100644 > --- a/arch/arm/mach-k3/Makefile > +++ b/arch/arm/mach-k3/Makefile > @@ -13,6 +13,7 @@ obj-$(CONFIG_SOC_K3_J721S2) += j721s2_fdt.o > obj-$(CONFIG_SOC_K3_AM625) += am625_fdt.o > obj-$(CONFIG_SOC_K3_AM62A7) += am62a7_fdt.o > obj-$(CONFIG_SOC_K3_J784S4) += j784s4_fdt.o > +obj-$(CONFIG_SOC_K3_AM62P5) += am62p5_fdt.o > endif > ifeq ($(CONFIG_SPL_BUILD),y) > obj-$(CONFIG_SOC_K3_AM654) += am654_init.o > diff --git a/arch/arm/mach-k3/am62p5_fdt.c b/arch/arm/mach-k3/am62p5_fdt.c > new file mode 100644 > index 0000000000000..d67f012a5dcc4 > --- /dev/null > +++ b/arch/arm/mach-k3/am62p5_fdt.c > @@ -0,0 +1,16 @@ > +// SPDX-License-Identifier: GPL-2.0-or-later > +/* > + * Copyright (C) 2024 Texas Instruments Incorporated - https://www.ti.com/ > + */ > + > +#include <asm/hardware.h> > +#include "common_fdt.h" > +#include <fdt_support.h> > + > +int ft_system_setup(void *blob, struct bd_info *bd) > +{ > + fdt_fixup_reserved(blob, "tfa", CONFIG_K3_ATF_LOAD_ADDR, 0x80000); > + fdt_fixup_reserved(blob, "optee", CONFIG_K3_OPTEE_LOAD_ADDR, 0x1800000); > + > + return 0; > +} > diff --git a/arch/arm/mach-k3/am62px/Kconfig b/arch/arm/mach-k3/am62px/Kconfig > index 38a9e6811b119..76ae86b66222f 100644 > --- a/arch/arm/mach-k3/am62px/Kconfig > +++ b/arch/arm/mach-k3/am62px/Kconfig > @@ -13,6 +13,7 @@ config TARGET_AM62P5_A53_EVM > bool "TI K3 based AM62P5 EVM running on A53" > select ARM64 > select BINMAN > + select OF_SYSTEM_SETUP > > config TARGET_AM62P5_R5_EVM > bool "TI K3 based AM62P5 EVM running on R5" > > --- > base-commit: a7f0154c412859323396111dd0c09dbafbc153cb > change-id: 20240520-am62p-fdt-fix-7c51e1a1cd54 > > Best regards,
On May 23, 2024 thus sayeth Andrew Davis: > On 5/23/24 11:43 AM, Bryan Brattlof wrote: > > The address we load TFA and OPTEE is configurable by the > > CONFIG_K3_{ATF,OPTEE)_LOAD_ADDR, but the DT node reservations remain > > static which can cause some confusion about where exactly these firmware > > are exactly. Fix this by updating the reserved-memory{} nodes when the > > loaded address does not match the address in DT. > > > > Reported-by: Andrew Davis <afd@ti.com> > > Signed-off-by: Bryan Brattlof <bb@ti.com> > > --- > > Hello everyone, > > > > This is a little fixup to avoid any confusion once we're in the kernel. > > Because TF-A can be configured in U-Boot to be anywhere we want, we need > > up update the reserved-memory{} node with this change. > > > > Thanks for reviewing > > ~Bryan > > --- > > arch/arm/mach-k3/Makefile | 1 + > > arch/arm/mach-k3/am62p5_fdt.c | 16 ++++++++++++++++ > > You'll want to rebase this on -next, these _fdt.c files all got moved > into directories for each SoC. > Ah! my bad, you're right this belongs in -next. v2 incoming momentarily Thanks Andrew ~Bryan
On May 23, 2024 thus sayeth Bryan Brattlof: > On May 23, 2024 thus sayeth Andrew Davis: > > On 5/23/24 11:43 AM, Bryan Brattlof wrote: > > > The address we load TFA and OPTEE is configurable by the > > > CONFIG_K3_{ATF,OPTEE)_LOAD_ADDR, but the DT node reservations remain > > > static which can cause some confusion about where exactly these firmware > > > are exactly. Fix this by updating the reserved-memory{} nodes when the > > > loaded address does not match the address in DT. > > > > > > Reported-by: Andrew Davis <afd@ti.com> > > > Signed-off-by: Bryan Brattlof <bb@ti.com> > > > --- > > > Hello everyone, > > > > > > This is a little fixup to avoid any confusion once we're in the kernel. > > > Because TF-A can be configured in U-Boot to be anywhere we want, we need > > > up update the reserved-memory{} node with this change. > > > > > > Thanks for reviewing > > > ~Bryan > > > --- > > > arch/arm/mach-k3/Makefile | 1 + > > > arch/arm/mach-k3/am62p5_fdt.c | 16 ++++++++++++++++ > > > > You'll want to rebase this on -next, these _fdt.c files all got moved > > into directories for each SoC. > > > > Ah! my bad, you're right this belongs in -next. v2 incoming momentarily > Wait no on second thought, this needs to be in v2024.07 also. I'll leave this for Tom for -master and maybe send out another version for -next to preempt the merge conflict for v2024.10? ~Bryan
On Thu, May 23, 2024 at 12:22:56PM -0500, Bryan Brattlof wrote: > On May 23, 2024 thus sayeth Bryan Brattlof: > > On May 23, 2024 thus sayeth Andrew Davis: > > > On 5/23/24 11:43 AM, Bryan Brattlof wrote: > > > > The address we load TFA and OPTEE is configurable by the > > > > CONFIG_K3_{ATF,OPTEE)_LOAD_ADDR, but the DT node reservations remain > > > > static which can cause some confusion about where exactly these firmware > > > > are exactly. Fix this by updating the reserved-memory{} nodes when the > > > > loaded address does not match the address in DT. > > > > > > > > Reported-by: Andrew Davis <afd@ti.com> > > > > Signed-off-by: Bryan Brattlof <bb@ti.com> > > > > --- > > > > Hello everyone, > > > > > > > > This is a little fixup to avoid any confusion once we're in the kernel. > > > > Because TF-A can be configured in U-Boot to be anywhere we want, we need > > > > up update the reserved-memory{} node with this change. > > > > > > > > Thanks for reviewing > > > > ~Bryan > > > > --- > > > > arch/arm/mach-k3/Makefile | 1 + > > > > arch/arm/mach-k3/am62p5_fdt.c | 16 ++++++++++++++++ > > > > > > You'll want to rebase this on -next, these _fdt.c files all got moved > > > into directories for each SoC. > > > > > > > Ah! my bad, you're right this belongs in -next. v2 incoming momentarily > > > > Wait no on second thought, this needs to be in v2024.07 also. I'll leave > this for Tom for -master and maybe send out another version for -next to > preempt the merge conflict for v2024.10? If this needs to go in master then git should hopefully do the right thing when I merge the next rc tag in.
On Thu, May 23, 2024 at 11:43:20AM -0500, Bryan Brattlof wrote: > The address we load TFA and OPTEE is configurable by the > CONFIG_K3_{ATF,OPTEE)_LOAD_ADDR, but the DT node reservations remain > static which can cause some confusion about where exactly these firmware > are exactly. Fix this by updating the reserved-memory{} nodes when the > loaded address does not match the address in DT. > > Reported-by: Andrew Davis <afd@ti.com> > Signed-off-by: Bryan Brattlof <bb@ti.com> Applied to u-boot/master, thanks!
diff --git a/arch/arm/mach-k3/Makefile b/arch/arm/mach-k3/Makefile index 1bd523329a4f8..4e9d0925f13f5 100644 --- a/arch/arm/mach-k3/Makefile +++ b/arch/arm/mach-k3/Makefile @@ -13,6 +13,7 @@ obj-$(CONFIG_SOC_K3_J721S2) += j721s2_fdt.o obj-$(CONFIG_SOC_K3_AM625) += am625_fdt.o obj-$(CONFIG_SOC_K3_AM62A7) += am62a7_fdt.o obj-$(CONFIG_SOC_K3_J784S4) += j784s4_fdt.o +obj-$(CONFIG_SOC_K3_AM62P5) += am62p5_fdt.o endif ifeq ($(CONFIG_SPL_BUILD),y) obj-$(CONFIG_SOC_K3_AM654) += am654_init.o diff --git a/arch/arm/mach-k3/am62p5_fdt.c b/arch/arm/mach-k3/am62p5_fdt.c new file mode 100644 index 0000000000000..d67f012a5dcc4 --- /dev/null +++ b/arch/arm/mach-k3/am62p5_fdt.c @@ -0,0 +1,16 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (C) 2024 Texas Instruments Incorporated - https://www.ti.com/ + */ + +#include <asm/hardware.h> +#include "common_fdt.h" +#include <fdt_support.h> + +int ft_system_setup(void *blob, struct bd_info *bd) +{ + fdt_fixup_reserved(blob, "tfa", CONFIG_K3_ATF_LOAD_ADDR, 0x80000); + fdt_fixup_reserved(blob, "optee", CONFIG_K3_OPTEE_LOAD_ADDR, 0x1800000); + + return 0; +} diff --git a/arch/arm/mach-k3/am62px/Kconfig b/arch/arm/mach-k3/am62px/Kconfig index 38a9e6811b119..76ae86b66222f 100644 --- a/arch/arm/mach-k3/am62px/Kconfig +++ b/arch/arm/mach-k3/am62px/Kconfig @@ -13,6 +13,7 @@ config TARGET_AM62P5_A53_EVM bool "TI K3 based AM62P5 EVM running on A53" select ARM64 select BINMAN + select OF_SYSTEM_SETUP config TARGET_AM62P5_R5_EVM bool "TI K3 based AM62P5 EVM running on R5"
The address we load TFA and OPTEE is configurable by the CONFIG_K3_{ATF,OPTEE)_LOAD_ADDR, but the DT node reservations remain static which can cause some confusion about where exactly these firmware are exactly. Fix this by updating the reserved-memory{} nodes when the loaded address does not match the address in DT. Reported-by: Andrew Davis <afd@ti.com> Signed-off-by: Bryan Brattlof <bb@ti.com> --- Hello everyone, This is a little fixup to avoid any confusion once we're in the kernel. Because TF-A can be configured in U-Boot to be anywhere we want, we need up update the reserved-memory{} node with this change. Thanks for reviewing ~Bryan --- arch/arm/mach-k3/Makefile | 1 + arch/arm/mach-k3/am62p5_fdt.c | 16 ++++++++++++++++ arch/arm/mach-k3/am62px/Kconfig | 1 + 3 files changed, 18 insertions(+) --- base-commit: a7f0154c412859323396111dd0c09dbafbc153cb change-id: 20240520-am62p-fdt-fix-7c51e1a1cd54 Best regards,