[U-Boot] [PATCH v3 0/2] armv8: Support loading 32-bit OS in AArch32 execution state

[U-Boot] [PATCH v3 0/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Alison Wang]

This series is to support loading a 32-bit OS, the execution state will
change from AArch64 to AArch32 when jumping to kernel. The architecture
information will be got through checking FIT image, then U-Boot will
load 32-bit OS or 64-bit OS automatically.

Spin-table method is used for secondary cores to load 32-bit OS. The
architecture information will be got through checking FIT image and
saved in the os_arch element of spin-table, then the secondary cores
will check os_arch and jump to 32-bit OS or 64-bit OS automatically.

---------------------------------------------------------------
Changes in v3:
- Comments the functions and the arguments.
- Rename the real parameters.
- Use the macros instead of the magic values.
- Remove the redundant codes.
- Clean up all of the mess in boot_jump_linux().
- Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system doesn't support AArch32 state.
- Adjust the arguments for armv8_switch_to_el2_m and armv8_switch_to_el1_m.

Changes in v2:
- armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is used
  to switch to AArch64 EL2 or AArch32 Hyp.
- armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is used
  to switch to AArch64 EL1 or AArch32 SVC.
- Support to call armv8_switch_to_el2_m and armv8_switch_to_el1_m.

----------------------------------------------------------------
Alison Wang (2):
      armv8: Support loading 32-bit OS in AArch32 execution state
      armv8: fsl-layerscape: SMP support for loading 32-bit OS

 arch/arm/Kconfig                              |   6 ++
 arch/arm/cpu/armv8/fsl-layerscape/lowlevel.S  |  35 ++++++++--
 arch/arm/cpu/armv8/fsl-layerscape/mp.c        |  10 +++
 arch/arm/cpu/armv8/transition.S               |   8 +--
 arch/arm/include/asm/arch-fsl-layerscape/mp.h |   6 ++
 arch/arm/include/asm/macro.h                  | 174 ++++++++++++++++++++++++++++++++++++-----------
 arch/arm/include/asm/system.h                 | 118 +++++++++++++++++++++++++++++++-
 arch/arm/lib/bootm.c                          |  24 +++++--
 common/image-fit.c                            |  14 +++-
 9 files changed, 339 insertions(+), 56 deletions(-)

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Alison Wang]

To support loading a 32-bit OS, the execution state will change from
AArch64 to AArch32 when jumping to kernel.

The architecture information will be got through checking FIT
image, then U-Boot will load 32-bit OS or 64-bit OS automatically.

Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
Signed-off-by: Alison Wang <alison.wang@nxp.com>
Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
---
Changes in v3:
- Comments the functions and the arguments.
- Rename the real parameters.
- Use the macros instead of the magic values.
- Remove the redundant codes.
- Clean up all of the mess in boot_jump_linux().
- Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system doesn't support AArch32 state.

Changes in v2:
- armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is used
  to switch to AArch64 EL2 or AArch32 Hyp.
- armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is used
  to switch to AArch64 EL1 or AArch32 SVC.

 arch/arm/Kconfig                |   6 ++
 arch/arm/cpu/armv8/transition.S |   8 +-
 arch/arm/include/asm/macro.h    | 174 ++++++++++++++++++++++++++++++----------
 arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
 arch/arm/lib/bootm.c            |  19 ++++-
 common/image-fit.c              |  14 +++-
 6 files changed, 287 insertions(+), 52 deletions(-)

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 729b181..794cb4f 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
 	  If SoC does not support L2CACHE or one do not want to enable
 	  L2CACHE, choose this option.
 
+config ARM64_SUPPORT_AARCH32
+	bool "ARM64 system support AArch32 execution state"
+	default y if ARM64 && !CONFIG_THUNDERX
+	help
+	  This ARM64 system supports AArch32 execution state.
+
 choice
 	prompt "Target select"
 	default TARGET_HIKEY
diff --git a/arch/arm/cpu/armv8/transition.S b/arch/arm/cpu/armv8/transition.S
index 253a39b..417e8b4 100644
--- a/arch/arm/cpu/armv8/transition.S
+++ b/arch/arm/cpu/armv8/transition.S
@@ -11,13 +11,13 @@
 #include <asm/macro.h>
 
 ENTRY(armv8_switch_to_el2)
-	switch_el x0, 1f, 0f, 0f
+	switch_el x4, 1f, 0f, 0f
 0:	ret
-1:	armv8_switch_to_el2_m x0
+1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
 ENDPROC(armv8_switch_to_el2)
 
 ENTRY(armv8_switch_to_el1)
-	switch_el x0, 0f, 1f, 0f
+	switch_el x4, 0f, 1f, 0f
 0:	ret
-1:	armv8_switch_to_el1_m x0, x1
+1:	armv8_switch_to_el1_m x0, x1, x2, x3, x4
 ENDPROC(armv8_switch_to_el1)
diff --git a/arch/arm/include/asm/macro.h b/arch/arm/include/asm/macro.h
index 9bb0efa..dd2c510 100644
--- a/arch/arm/include/asm/macro.h
+++ b/arch/arm/include/asm/macro.h
@@ -8,6 +8,9 @@
 
 #ifndef __ASM_ARM_MACRO_H__
 #define __ASM_ARM_MACRO_H__
+
+#include <asm/system.h>
+
 #ifdef __ASSEMBLY__
 
 /*
@@ -135,13 +138,18 @@ lr	.req	x30
 #endif
 .endm
 
-.macro armv8_switch_to_el2_m, xreg1
-	/* 64bit EL2 | HCE | SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1 */
-	mov	\xreg1, #0x5b1
-	msr	scr_el3, \xreg1
+/*
+ * Switch from EL3 to EL2 for ARMv8
+ * @ep:     kernel entry point
+ * @arch:   machine nr
+ * @ftaddr: fdt address
+ * @flag:   The execution state flag for lower exception
+ *          level, ES_TO_AARCH64 or ES_TO_AARCH32
+ */
+.macro armv8_switch_to_el2_m, ep, arch, ftaddr, flag, xreg5
 	msr	cptr_el3, xzr		/* Disable coprocessor traps to EL3 */
-	mov	\xreg1, #0x33ff
-	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
+	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
+	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
 
 	/* Initialize Generic Timers */
 	msr	cntvoff_el2, xzr
@@ -152,45 +160,92 @@ lr	.req	x30
 	 * and RES0 bits (31,30,27,26,24,21,20,17,15-13,10-6) +
 	 * EE,WXN,I,SA,C,A,M to 0
 	 */
-	mov	\xreg1, #0x0830
-	movk	\xreg1, #0x30C5, lsl #16
-	msr	sctlr_el2, \xreg1
+	ldr	\xreg5, =(SCTLR_EL2_BIT29_28_RES1 | SCTLR_EL2_EE_LE |\
+			  SCTLR_EL2_BIT23_22_RES1 | SCTLR_EL2_WXN_DIS |\
+			  SCTLR_EL2_BIT18_RES1 | SCTLR_EL2_BIT16_RES1 |\
+			  SCTLR_EL2_BIT11_RES1 | SCTLR_EL2_BIT5_4_RES1 |\
+			  SCTLR_EL2_SA_DIS | SCTLR_EL2_DCACHE_DIS |\
+			  SCTLR_EL2_ALIGN_DIS | SCTLR_EL2_MMU_DIS)
+	msr	sctlr_el2, \xreg5
+
+	mov	\xreg5, sp
+	msr	sp_el2, \xreg5		/* Migrate SP */
+	mrs	\xreg5, vbar_el3
+	msr	vbar_el2, \xreg5	/* Migrate VBAR */
+
+	/* Check switch to AArch64 EL2 or AArch32 Hypervisor mode */
+	ldr	\xreg5, =ES_TO_AARCH32
+	cmp	\flag, \xreg5
+	b.eq	1f
+
+	/*
+	 * The next lower exception level is AArch64, 64bit EL2 | HCE |
+	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
+	 */
+	ldr	\xreg5, =(SCR_EL3_RW_AARCH64 | SCR_EL3_HCE_EN |\
+			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
+			  SCR_EL3_NS_EN)
+	msr	scr_el3, \xreg5
 
 	/* Return to the EL2_SP2 mode from EL3 */
-	mov	\xreg1, sp
-	msr	sp_el2, \xreg1		/* Migrate SP */
-	mrs	\xreg1, vbar_el3
-	msr	vbar_el2, \xreg1	/* Migrate VBAR */
-	mov	\xreg1, #0x3c9
-	msr	spsr_el3, \xreg1	/* EL2_SP2 | D | A | I | F */
+	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
+			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
+			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL2H)
+	msr	spsr_el3, \xreg5
 	msr	elr_el3, lr
 	eret
+
+1:
+	/*
+	 * The next lower exception level is AArch32, 32bit EL2 | HCE |
+	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
+	 */
+	ldr	\xreg5, =(SCR_EL3_RW_AARCH32 | SCR_EL3_HCE_EN |\
+			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
+			  SCR_EL3_NS_EN)
+	msr	scr_el3, \xreg5
+
+	/* Return to AArch32 Hypervisor mode */
+	ldr     \xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
+			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
+			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
+			  SPSR_EL_M_HYP)
+	msr	spsr_el3, \xreg5
+	msr     elr_el3, \ep
+
+	mov	\ep, #0
+	eret
 .endm
 
-.macro armv8_switch_to_el1_m, xreg1, xreg2
+/*
+ * Switch from EL2 to EL1 for ARMv8
+ * @ep:     kernel entry point
+ * @arch:   machine nr
+ * @ftaddr: fdt address
+ * @flag:   The execution state flag for lower exception
+ *          level, ES_TO_AARCH64 or ES_TO_AARCH32
+ */
+.macro armv8_switch_to_el1_m, ep, arch, ftaddr, flag, xreg5
 	/* Initialize Generic Timers */
-	mrs	\xreg1, cnthctl_el2
-	orr	\xreg1, \xreg1, #0x3	/* Enable EL1 access to timers */
-	msr	cnthctl_el2, \xreg1
+	mrs	\xreg5, cnthctl_el2
+	/* Enable EL1 access to timers */
+	orr	\xreg5, \xreg5, #(CNTHCTL_EL2_EL1PCEN_EN |\
+		CNTHCTL_EL2_EL1PCTEN_EN)
+	msr	cnthctl_el2, \xreg5
 	msr	cntvoff_el2, xzr
 
 	/* Initilize MPID/MPIDR registers */
-	mrs	\xreg1, midr_el1
-	mrs	\xreg2, mpidr_el1
-	msr	vpidr_el2, \xreg1
-	msr	vmpidr_el2, \xreg2
+	mrs	\xreg5, midr_el1
+	msr	vpidr_el2, \xreg5
+	mrs	\xreg5, mpidr_el1
+	msr	vmpidr_el2, \xreg5
 
 	/* Disable coprocessor traps */
-	mov	\xreg1, #0x33ff
-	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
+	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
+	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
 	msr	hstr_el2, xzr		/* Disable coprocessor traps to EL2 */
-	mov	\xreg1, #3 << 20
-	msr	cpacr_el1, \xreg1	/* Enable FP/SIMD at EL1 */
-
-	/* Initialize HCR_EL2 */
-	mov	\xreg1, #(1 << 31)		/* 64bit EL1 */
-	orr	\xreg1, \xreg1, #(1 << 29)	/* Disable HVC */
-	msr	hcr_el2, \xreg1
+	mov	\xreg5, #(CPACR_EL1_FPEN_EN)
+	msr	cpacr_el1, \xreg5	/* Enable FP/SIMD@EL1 */
 
 	/* SCTLR_EL1 initialization
 	 *
@@ -199,19 +254,56 @@ lr	.req	x30
 	 * UCI,EE,EOE,WXN,nTWE,nTWI,UCT,DZE,I,UMA,SED,ITD,
 	 * CP15BEN,SA0,SA,C,A,M to 0
 	 */
-	mov	\xreg1, #0x0800
-	movk	\xreg1, #0x30d0, lsl #16
-	msr	sctlr_el1, \xreg1
+	ldr	\xreg5, =(SCTLR_EL1_BIT29_28_RES1 | SCTLR_EL1_UCI_DIS |\
+			  SCTLR_EL1_EE_LE | SCTLR_EL1_BIT23_22_RES1 |\
+			  SCTLR_EL1_BIT20_RES1 | SCTLR_EL1_WXN_DIS |\
+			  SCTLR_EL1_NTWE_DIS | SCTLR_EL1_NTWI_DIS |\
+			  SCTLR_EL1_UCT_DIS | SCTLR_EL1_DZE_DIS |\
+			  SCTLR_EL1_ICACHE_DIS | SCTLR_EL1_BIT11_RES1 |\
+			  SCTLR_EL1_UMA_DIS | SCTLR_EL1_SED_EN |\
+			  SCTLR_EL1_ITD_EN | SCTLR_EL1_CP15BEN_DIS |\
+			  SCTLR_EL1_SA0_DIS | SCTLR_EL1_SA_DIS |\
+			  SCTLR_EL1_DCACHE_DIS | SCTLR_EL1_ALIGN_DIS |\
+			  SCTLR_EL1_MMU_DIS)
+	msr	sctlr_el1, \xreg5
+
+	mov	\xreg5, sp
+	msr	sp_el1, \xreg5		/* Migrate SP */
+	mrs	\xreg5, vbar_el2
+	msr	vbar_el1, \xreg5	/* Migrate VBAR */
+
+	/* Check switch to AArch64 EL1 or AArch32 Supervisor mode */
+	ldr	\xreg5, =ES_TO_AARCH32
+	cmp	\flag, \xreg5
+	b.eq	1f
+
+	/* Initialize HCR_EL2 */
+	ldr	\xreg5, =(HCR_EL2_RW_AARCH64 | HCR_EL2_HCD_DIS)
+	msr	hcr_el2, \xreg5
 
 	/* Return to the EL1_SP1 mode from EL2 */
-	mov	\xreg1, sp
-	msr	sp_el1, \xreg1		/* Migrate SP */
-	mrs	\xreg1, vbar_el2
-	msr	vbar_el1, \xreg1	/* Migrate VBAR */
-	mov	\xreg1, #0x3c5
-	msr	spsr_el2, \xreg1	/* EL1_SP1 | D | A | I | F */
+	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
+			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
+			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL1H)
+	msr	spsr_el2, \xreg5
 	msr	elr_el2, lr
 	eret
+
+1:
+	/* Initialize HCR_EL2 */
+	ldr	\xreg5, =(HCR_EL2_RW_AARCH32 | HCR_EL2_HCD_DIS)
+	msr	hcr_el2, \xreg5
+
+	/* Return to AArch32 Supervisor mode from EL2 */
+	ldr	\xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
+			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
+			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
+			  SPSR_EL_M_SVC)
+	msr     spsr_el2, \xreg5
+	msr     elr_el2, \ep
+
+	mov	\ep, #0
+	eret
 .endm
 
 #if defined(CONFIG_GICV3)
diff --git a/arch/arm/include/asm/system.h b/arch/arm/include/asm/system.h
index 9ae890a..0271ddb 100644
--- a/arch/arm/include/asm/system.h
+++ b/arch/arm/include/asm/system.h
@@ -17,6 +17,102 @@
 #define CR_WXN		(1 << 19)	/* Write Permision Imply XN	*/
 #define CR_EE		(1 << 25)	/* Exception (Big) Endian	*/
 
+#define ES_TO_AARCH64		1
+#define ES_TO_AARCH32		0
+
+/*
+ * SCR_EL3 bits definitions
+ */
+#define SCR_EL3_RW_AARCH64	(1 << 10) /* Next lower level is AArch64     */
+#define SCR_EL3_RW_AARCH32	(0 << 10) /* Lower lowers level are AArch32  */
+#define SCR_EL3_HCE_EN		(1 << 8)  /* Hypervisor Call enable          */
+#define SCR_EL3_SMD_DIS		(1 << 7)  /* Secure Monitor Call disable     */
+#define SCR_EL3_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1                  */
+#define SCR_EL3_NS_EN		(1 << 0)  /* EL0 and EL1 in Non-scure state  */
+
+/*
+ * SPSR_EL3/SPSR_EL2 bits definitions
+ */
+#define SPSR_EL_END_LE		(0 << 9)  /* Exception Little-endian          */
+#define SPSR_EL_DEBUG_MASK	(1 << 9)  /* Debug exception masked           */
+#define SPSR_EL_ASYN_MASK	(1 << 8)  /* Asynchronous data abort masked   */
+#define SPSR_EL_SERR_MASK	(1 << 8)  /* System Error exception masked    */
+#define SPSR_EL_IRQ_MASK	(1 << 7)  /* IRQ exception masked             */
+#define SPSR_EL_FIQ_MASK	(1 << 6)  /* FIQ exception masked             */
+#define SPSR_EL_T_A32		(0 << 5)  /* AArch32 instruction set A32      */
+#define SPSR_EL_M_AARCH64	(0 << 4)  /* Exception taken from AArch64     */
+#define SPSR_EL_M_AARCH32	(1 << 4)  /* Exception taken from AArch32     */
+#define SPSR_EL_M_SVC		(0x3)     /* Exception taken from SVC mode    */
+#define SPSR_EL_M_HYP		(0xa)     /* Exception taken from HYP mode    */
+#define SPSR_EL_M_EL1H		(5)       /* Exception taken from EL1h mode   */
+#define SPSR_EL_M_EL2H		(9)       /* Exception taken from EL2h mode   */
+
+/*
+ * CPTR_EL2 bits definitions
+ */
+#define CPTR_EL2_BIT13_12_RES1	(3 << 12) /* Reserved, RES1                   */
+#define CPTR_EL2_BIT9_0_RES1	(0x3ff)   /* Reserved, RES1                   */
+
+/*
+ * SCTLR_EL2 bits definitions
+ */
+#define SCTLR_EL2_BIT29_28_RES1	(3 << 28) /* Reserved, RES1                   */
+#define SCTLR_EL2_EE_LE		(0 << 25) /* Exception Little-endian          */
+#define SCTLR_EL2_BIT23_22_RES1	(3 << 22) /* Reserved, RES1                   */
+#define SCTLR_EL2_WXN_DIS	(0 << 19) /* Write permission is not XN       */
+#define SCTLR_EL2_BIT18_RES1	(1 << 18) /* Reserved, RES1                   */
+#define SCTLR_EL2_BIT16_RES1	(1 << 16) /* Reserved, RES1                   */
+#define SCTLR_EL2_ICACHE_DIS	(0 << 12) /* Instruction cache disabled       */
+#define SCTLR_EL2_BIT11_RES1	(1 << 11) /* Reserved, RES1                   */
+#define SCTLR_EL2_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1                   */
+#define SCTLR_EL2_SA_DIS	(0 << 3)  /* Stack Alignment Check disabled   */
+#define SCTLR_EL2_DCACHE_DIS	(0 << 2)  /* Data cache disabled              */
+#define SCTLR_EL2_ALIGN_DIS	(0 << 1)  /* Alignment check disabled         */
+#define SCTLR_EL2_MMU_DIS	(0)       /* MMU disabled                     */
+
+/*
+ * CNTHCTL_EL2 bits definitions
+ */
+#define CNTHCTL_EL2_EL1PCEN_EN	(1 << 1)  /* Physical timer regs accessible   */
+#define CNTHCTL_EL2_EL1PCTEN_EN	(1 << 0)  /* Physical counter accessible      */
+
+/*
+ * HCR_EL2 bits definitions
+ */
+#define HCR_EL2_RW_AARCH64	(1 << 31) /* EL1 is AArch64                   */
+#define HCR_EL2_RW_AARCH32	(0 << 31) /* Lower levels are AArch32         */
+#define HCR_EL2_HCD_DIS		(1 << 29) /* Hypervisor Call disabled         */
+
+/*
+ * CPACR_EL1 bits definitions
+ */
+#define CPACR_EL1_FPEN_EN	(3 << 20) /* SIMD and FP instruction enabled  */
+
+/*
+ * SCTLR_EL1 bits definitions
+ */
+#define SCTLR_EL1_BIT29_28_RES1	(3 << 28) /* Reserved, RES1                   */
+#define SCTLR_EL1_UCI_DIS	(0 << 26) /* Cache instruction disabled       */
+#define SCTLR_EL1_EE_LE		(0 << 25) /* Exception Little-endian          */
+#define SCTLR_EL1_BIT23_22_RES1	(3 << 22) /* Reserved, RES1                   */
+#define SCTLR_EL1_BIT20_RES1	(1 << 20) /* Reserved, RES1                   */
+#define SCTLR_EL1_WXN_DIS	(0 << 19) /* Write permission is not XN       */
+#define SCTLR_EL1_NTWE_DIS	(0 << 18) /* WFE instruction disabled         */
+#define SCTLR_EL1_NTWI_DIS	(0 << 16) /* WFI instruction disabled         */
+#define SCTLR_EL1_UCT_DIS	(0 << 15) /* CTR_EL0 access disabled          */
+#define SCTLR_EL1_DZE_DIS	(0 << 14) /* DC ZVA instruction disabled      */
+#define SCTLR_EL1_ICACHE_DIS	(0 << 12) /* Instruction cache disabled       */
+#define SCTLR_EL1_BIT11_RES1	(1 << 11) /* Reserved, RES1                   */
+#define SCTLR_EL1_UMA_DIS	(0 << 9)  /* User Mask Access disabled        */
+#define SCTLR_EL1_SED_EN	(0 << 8)  /* SETEND instruction enabled       */
+#define SCTLR_EL1_ITD_EN	(0 << 7)  /* IT instruction enabled           */
+#define SCTLR_EL1_CP15BEN_DIS	(0 << 5)  /* CP15 barrier operation disabled  */
+#define SCTLR_EL1_SA0_DIS	(0 << 4)  /* Stack Alignment EL0 disabled     */
+#define SCTLR_EL1_SA_DIS	(0 << 3)  /* Stack Alignment EL1 disabled     */
+#define SCTLR_EL1_DCACHE_DIS	(0 << 2)  /* Data cache disabled              */
+#define SCTLR_EL1_ALIGN_DIS	(0 << 1)  /* Alignment check disabled         */
+#define SCTLR_EL1_MMU_DIS	(0)       /* MMU disabled                     */
+
 #ifndef __ASSEMBLY__
 
 u64 get_page_table_size(void);
@@ -100,8 +196,26 @@ void __asm_invalidate_icache_all(void);
 int __asm_flush_l3_cache(void);
 void __asm_switch_ttbr(u64 new_ttbr);
 
-void armv8_switch_to_el2(void);
-void armv8_switch_to_el1(void);
+/*
+ * Switch from EL3 to EL2 for ARMv8
+ *
+ * @entry_point: kernel entry point
+ * @mach_nr:     machine nr
+ * @fdt_addr:    fdt address
+ * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
+ */
+void armv8_switch_to_el2(u64 entry_point, u64 mach_nr, u64 fdt_addr,
+			 u64 es_flag);
+/*
+ * Switch from EL2 to EL1 for ARMv8
+ *
+ * @entry_point: kernel entry point
+ * @mach_nr:     machine nr
+ * @fdt_addr:    fdt address
+ * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
+ */
+void armv8_switch_to_el1(u64 entry_point, u64 mach_nr, u64 fdt_addr,
+			 u64 es_flag);
 void gic_init(void);
 void gic_send_sgi(unsigned long sgino);
 void wait_for_wakeup(void);
diff --git a/arch/arm/lib/bootm.c b/arch/arm/lib/bootm.c
index 0838d89..1bbf85c 100644
--- a/arch/arm/lib/bootm.c
+++ b/arch/arm/lib/bootm.c
@@ -189,13 +189,19 @@ static void setup_end_tag(bd_t *bd)
 __weak void setup_board_tags(struct tag **in_params) {}
 
 #ifdef CONFIG_ARM64
-static void do_nonsec_virt_switch(void)
+static void do_nonsec_virt_switch(bootm_headers_t *images, int flag)
 {
 	smp_kick_all_cpus();
 	dcache_disable();	/* flush cache before swtiching to EL2 */
-	armv8_switch_to_el2();
+
 #ifdef CONFIG_ARMV8_SWITCH_TO_EL1
-	armv8_switch_to_el1();
+	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
+			    (u64)images->ft_addr, ES_TO_AARCH64);
+	armv8_switch_to_el1((u64)images->ep, (u64)gd->bd->bi_arch_number,
+			    (u64)images->ft_addr, flag);
+#else
+	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
+			    (u64)images->ft_addr, flag);
 #endif
 }
 #endif
@@ -275,6 +281,7 @@ static void boot_jump_linux(bootm_headers_t *images, int flag)
 	void (*kernel_entry)(void *fdt_addr, void *res0, void *res1,
 			void *res2);
 	int fake = (flag & BOOTM_STATE_OS_FAKE_GO);
+	int es_flag = ES_TO_AARCH64;
 
 	kernel_entry = (void (*)(void *fdt_addr, void *res0, void *res1,
 				void *res2))images->ep;
@@ -286,7 +293,11 @@ static void boot_jump_linux(bootm_headers_t *images, int flag)
 	announce_and_cleanup(fake);
 
 	if (!fake) {
-		do_nonsec_virt_switch();
+		if ((IH_ARCH_DEFAULT == IH_ARCH_ARM64) &&
+		    (images->os.arch == IH_ARCH_ARM))
+			es_flag = ES_TO_AARCH32;
+
+		do_nonsec_virt_switch(images, es_flag);
 		kernel_entry(images->ft_addr, NULL, NULL, NULL);
 	}
 #else
diff --git a/common/image-fit.c b/common/image-fit.c
index 9873957..982021b 100644
--- a/common/image-fit.c
+++ b/common/image-fit.c
@@ -26,6 +26,7 @@ DECLARE_GLOBAL_DATA_PTR;
 #include <u-boot/md5.h>
 #include <u-boot/sha1.h>
 #include <u-boot/sha256.h>
+#include <generated/autoconf.h>
 
 /*****************************************************************************/
 /* New uImage format routines */
@@ -1164,7 +1165,9 @@ int fit_image_check_arch(const void *fit, int noffset, uint8_t arch)
 	if (fit_image_get_arch(fit, noffset, &image_arch))
 		return 0;
 	return (arch == image_arch) ||
-		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64);
+		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64) ||
+		(arch == IH_ARCH_ARM64 && image_arch == IH_ARCH_ARM &&
+		 CONFIG_ARM64_SUPPORT_AARCH32);
 }
 
 /**
@@ -1593,6 +1596,9 @@ int fit_image_load(bootm_headers_t *images, ulong addr,
 	int type_ok, os_ok;
 	ulong load, data, len;
 	uint8_t os;
+#ifndef USE_HOSTCC
+	uint8_t os_arch;
+#endif
 	const char *prop_name;
 	int ret;
 
@@ -1676,6 +1682,12 @@ int fit_image_load(bootm_headers_t *images, ulong addr,
 		return -ENOEXEC;
 	}
 #endif
+
+#ifndef USE_HOSTCC
+	fit_image_get_arch(fit, noffset, &os_arch);
+	images->os.arch = os_arch;
+#endif
+
 	if (image_type == IH_TYPE_FLATDT &&
 	    !fit_image_check_comp(fit, noffset, IH_COMP_NONE)) {
 		puts("FDT image is compressed");
-- 
2.1.0.27.g96db324

[U-Boot] [PATCH v3 2/2] armv8: fsl-layerscape: SMP support for loading 32-bit OS

[Alison Wang]

Spin-table method is used for secondary cores to load 32-bit OS. The
architecture information will be got through checking FIT image and
saved in the os_arch element of spin-table, then the secondary cores
will check os_arch and jump to 32-bit OS or 64-bit OS automatically.

Signed-off-by: Alison Wang <alison.wang@nxp.com>
Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
---
Changes in v3:
- Adjust the arguments for armv8_switch_to_el2_m and armv8_switch_to_el1_m. 

Changes in v2:
- Support to call armv8_switch_to_el2_m and armv8_switch_to_el1_m. 

 arch/arm/cpu/armv8/fsl-layerscape/lowlevel.S  | 35 ++++++++++++++++++++++++---
 arch/arm/cpu/armv8/fsl-layerscape/mp.c        | 10 ++++++++
 arch/arm/include/asm/arch-fsl-layerscape/mp.h |  6 +++++
 arch/arm/lib/bootm.c                          |  5 ++++
 4 files changed, 52 insertions(+), 4 deletions(-)

diff --git a/arch/arm/cpu/armv8/fsl-layerscape/lowlevel.S b/arch/arm/cpu/armv8/fsl-layerscape/lowlevel.S
index 04831ca..7363595 100644
--- a/arch/arm/cpu/armv8/fsl-layerscape/lowlevel.S
+++ b/arch/arm/cpu/armv8/fsl-layerscape/lowlevel.S
@@ -13,6 +13,8 @@
 #ifdef CONFIG_MP
 #include <asm/arch/mp.h>
 #endif
+#include <asm/u-boot.h>
+#include <asm/system.h>
 
 ENTRY(lowlevel_init)
 	mov	x29, lr			/* Save LR */
@@ -320,6 +322,12 @@ ENTRY(secondary_boot_func)
         gic_wait_for_interrupt_m x0, w1
 #endif
 
+	ldr	x5, [x11, #24]
+	ldr	x6, =IH_ARCH_DEFAULT
+	cmp	x6, x5
+	b.ne	slave_cpu
+
+	ldr	x3, =ES_TO_AARCH64
 	bl secondary_switch_to_el2
 #ifdef CONFIG_ARMV8_SWITCH_TO_EL1
 	bl secondary_switch_to_el1
@@ -337,19 +345,38 @@ slave_cpu:
 	tbz     x1, #25, cpu_is_le
 	rev     x0, x0                  /* BE to LE conversion */
 cpu_is_le:
+
+	ldr	x5, [x11, #24]
+	ldr	x6, =IH_ARCH_DEFAULT
+	cmp	x6, x5
+	b.eq	1f
+
+#ifdef CONFIG_ARMV8_SWITCH_TO_EL1
+	ldr	x3, =ES_TO_AARCH64
+	bl	secondary_switch_to_el2
+	ldr	x0, [x11]
+	ldr	x3, =ES_TO_AARCH32
+	bl	secondary_switch_to_el1
+#else
+	ldr	x0, [x11]
+	ldr	x3, =ES_TO_AARCH32
+	bl	secondary_switch_to_el2
+#endif
+
+1:
 	br	x0			/* branch to the given address */
 ENDPROC(secondary_boot_func)
 
 ENTRY(secondary_switch_to_el2)
-	switch_el x0, 1f, 0f, 0f
+	switch_el x4, 1f, 0f, 0f
 0:	ret
-1:	armv8_switch_to_el2_m x0
+1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
 ENDPROC(secondary_switch_to_el2)
 
 ENTRY(secondary_switch_to_el1)
-	switch_el x0, 0f, 1f, 0f
+	switch_el x4, 0f, 1f, 0f
 0:	ret
-1:	armv8_switch_to_el1_m x0, x1
+1:	armv8_switch_to_el1_m x0, x1, x2, x3, x4
 ENDPROC(secondary_switch_to_el1)
 
 	/* Ensure that the literals used by the secondary boot code are
diff --git a/arch/arm/cpu/armv8/fsl-layerscape/mp.c b/arch/arm/cpu/armv8/fsl-layerscape/mp.c
index df7ffb8..dd91550 100644
--- a/arch/arm/cpu/armv8/fsl-layerscape/mp.c
+++ b/arch/arm/cpu/armv8/fsl-layerscape/mp.c
@@ -22,6 +22,16 @@ phys_addr_t determine_mp_bootpg(void)
 	return (phys_addr_t)&secondary_boot_code;
 }
 
+void update_os_arch_secondary_cores(uint8_t os_arch)
+{
+	u64 *table = get_spin_tbl_addr();
+	int i;
+
+	for (i = 1; i < CONFIG_MAX_CPUS; i++)
+		table[i * WORDS_PER_SPIN_TABLE_ENTRY +
+			SPIN_TABLE_ELEM_OS_ARCH_IDX] = os_arch;
+}
+
 int fsl_layerscape_wake_seconday_cores(void)
 {
 	struct ccsr_gur __iomem *gur = (void *)(CONFIG_SYS_FSL_GUTS_ADDR);
diff --git a/arch/arm/include/asm/arch-fsl-layerscape/mp.h b/arch/arm/include/asm/arch-fsl-layerscape/mp.h
index e46e076..55f0e0c 100644
--- a/arch/arm/include/asm/arch-fsl-layerscape/mp.h
+++ b/arch/arm/include/asm/arch-fsl-layerscape/mp.h
@@ -13,6 +13,7 @@
 *      uint64_t entry_addr;
 *      uint64_t status;
 *      uint64_t lpid;
+*      uint64_t os_arch;
 * };
 * we pad this struct to 64 bytes so each entry is in its own cacheline
 * the actual spin table is an array of these structures
@@ -20,6 +21,7 @@
 #define SPIN_TABLE_ELEM_ENTRY_ADDR_IDX	0
 #define SPIN_TABLE_ELEM_STATUS_IDX	1
 #define SPIN_TABLE_ELEM_LPID_IDX	2
+#define SPIN_TABLE_ELEM_OS_ARCH_IDX	3
 #define WORDS_PER_SPIN_TABLE_ENTRY	8	/* pad to 64 bytes */
 #define SPIN_TABLE_ELEM_SIZE		64
 
@@ -35,4 +37,8 @@ phys_addr_t determine_mp_bootpg(void);
 void secondary_boot_func(void);
 int is_core_online(u64 cpu_id);
 #endif
+
+#define IH_ARCH_ARM		2	/* ARM */
+#define IH_ARCH_ARM64		22	/* ARM64 */
+
 #endif /* _FSL_LAYERSCAPE_MP_H */
diff --git a/arch/arm/lib/bootm.c b/arch/arm/lib/bootm.c
index 1bbf85c..4dca065 100644
--- a/arch/arm/lib/bootm.c
+++ b/arch/arm/lib/bootm.c
@@ -274,6 +274,10 @@ bool armv7_boot_nonsec(void)
 }
 #endif
 
+__weak void update_os_arch_secondary_cores(uint8_t os_arch)
+{
+}
+
 /* Subcommand: GO */
 static void boot_jump_linux(bootm_headers_t *images, int flag)
 {
@@ -293,6 +297,7 @@ static void boot_jump_linux(bootm_headers_t *images, int flag)
 	announce_and_cleanup(fake);
 
 	if (!fake) {
+		update_os_arch_secondary_cores(images->os.arch);
 		if ((IH_ARCH_DEFAULT == IH_ARCH_ARM64) &&
 		    (images->os.arch == IH_ARCH_ARM))
 			es_flag = ES_TO_AARCH32;
-- 
2.1.0.27.g96db324

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[York Sun]

On 05/26/2016 01:51 AM, Alison Wang wrote:
> To support loading a 32-bit OS, the execution state will change from
> AArch64 to AArch32 when jumping to kernel.
> 
> The architecture information will be got through checking FIT
> image, then U-Boot will load 32-bit OS or 64-bit OS automatically.
> 
> Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
> Signed-off-by: Alison Wang <alison.wang@nxp.com>
> Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
> ---
> Changes in v3:
> - Comments the functions and the arguments.
> - Rename the real parameters.
> - Use the macros instead of the magic values.
> - Remove the redundant codes.
> - Clean up all of the mess in boot_jump_linux().
> - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system doesn't support AArch32 state.
> 
> Changes in v2:
> - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is used
>   to switch to AArch64 EL2 or AArch32 Hyp.
> - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is used
>   to switch to AArch64 EL1 or AArch32 SVC.
> 
>  arch/arm/Kconfig                |   6 ++
>  arch/arm/cpu/armv8/transition.S |   8 +-
>  arch/arm/include/asm/macro.h    | 174 ++++++++++++++++++++++++++++++----------
>  arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
>  arch/arm/lib/bootm.c            |  19 ++++-
>  common/image-fit.c              |  14 +++-
>  6 files changed, 287 insertions(+), 52 deletions(-)
> 
> diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
> index 729b181..794cb4f 100644
> --- a/arch/arm/Kconfig
> +++ b/arch/arm/Kconfig
> @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
>  	  If SoC does not support L2CACHE or one do not want to enable
>  	  L2CACHE, choose this option.
>  
> +config ARM64_SUPPORT_AARCH32
> +	bool "ARM64 system support AArch32 execution state"
> +	default y if ARM64 && !CONFIG_THUNDERX
> +	help
> +	  This ARM64 system supports AArch32 execution state.
> +
>  choice
>  	prompt "Target select"
>  	default TARGET_HIKEY
> diff --git a/arch/arm/cpu/armv8/transition.S b/arch/arm/cpu/armv8/transition.S
> index 253a39b..417e8b4 100644
> --- a/arch/arm/cpu/armv8/transition.S
> +++ b/arch/arm/cpu/armv8/transition.S
> @@ -11,13 +11,13 @@
>  #include <asm/macro.h>
>  
>  ENTRY(armv8_switch_to_el2)
> -	switch_el x0, 1f, 0f, 0f
> +	switch_el x4, 1f, 0f, 0f
>  0:	ret
> -1:	armv8_switch_to_el2_m x0
> +1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
>  ENDPROC(armv8_switch_to_el2)

Alison,

I see you are chaning macro armv8_switch_to_el2. You may have missed one in
start.S. This is the code you can't test with our ARMv8 chip.

York

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Alexander Graf]

On 26.05.16 10:41, Alison Wang wrote:
> To support loading a 32-bit OS, the execution state will change from
> AArch64 to AArch32 when jumping to kernel.
> 
> The architecture information will be got through checking FIT
> image, then U-Boot will load 32-bit OS or 64-bit OS automatically.
> 
> Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
> Signed-off-by: Alison Wang <alison.wang@nxp.com>
> Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
> ---
> Changes in v3:
> - Comments the functions and the arguments.
> - Rename the real parameters.
> - Use the macros instead of the magic values.
> - Remove the redundant codes.
> - Clean up all of the mess in boot_jump_linux().
> - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system doesn't support AArch32 state.
> 
> Changes in v2:
> - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is used
>   to switch to AArch64 EL2 or AArch32 Hyp.
> - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is used
>   to switch to AArch64 EL1 or AArch32 SVC.
> 
>  arch/arm/Kconfig                |   6 ++
>  arch/arm/cpu/armv8/transition.S |   8 +-
>  arch/arm/include/asm/macro.h    | 174 ++++++++++++++++++++++++++++++----------
>  arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
>  arch/arm/lib/bootm.c            |  19 ++++-
>  common/image-fit.c              |  14 +++-
>  6 files changed, 287 insertions(+), 52 deletions(-)
> 
> diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
> index 729b181..794cb4f 100644
> --- a/arch/arm/Kconfig
> +++ b/arch/arm/Kconfig
> @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
>  	  If SoC does not support L2CACHE or one do not want to enable
>  	  L2CACHE, choose this option.
>  
> +config ARM64_SUPPORT_AARCH32
> +	bool "ARM64 system support AArch32 execution state"
> +	default y if ARM64 && !CONFIG_THUNDERX

I don't think that works. In Kconfig you need to omit the CONFIG_ parts.
It's "ARM64" also, not "CONFIG_ARM64".

Please just try to run a local make defconfig on the thunderx reference
system and see whether it includes the option or not.

> +	help
> +	  This ARM64 system supports AArch32 execution state.
> +
>  choice
>  	prompt "Target select"
>  	default TARGET_HIKEY
> diff --git a/arch/arm/cpu/armv8/transition.S b/arch/arm/cpu/armv8/transition.S
> index 253a39b..417e8b4 100644
> --- a/arch/arm/cpu/armv8/transition.S
> +++ b/arch/arm/cpu/armv8/transition.S
> @@ -11,13 +11,13 @@
>  #include <asm/macro.h>
>  
>  ENTRY(armv8_switch_to_el2)
> -	switch_el x0, 1f, 0f, 0f
> +	switch_el x4, 1f, 0f, 0f
>  0:	ret
> -1:	armv8_switch_to_el2_m x0
> +1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
>  ENDPROC(armv8_switch_to_el2)
>  
>  ENTRY(armv8_switch_to_el1)
> -	switch_el x0, 0f, 1f, 0f
> +	switch_el x4, 0f, 1f, 0f
>  0:	ret
> -1:	armv8_switch_to_el1_m x0, x1
> +1:	armv8_switch_to_el1_m x0, x1, x2, x3, x4
>  ENDPROC(armv8_switch_to_el1)
> diff --git a/arch/arm/include/asm/macro.h b/arch/arm/include/asm/macro.h
> index 9bb0efa..dd2c510 100644
> --- a/arch/arm/include/asm/macro.h
> +++ b/arch/arm/include/asm/macro.h
> @@ -8,6 +8,9 @@
>  
>  #ifndef __ASM_ARM_MACRO_H__
>  #define __ASM_ARM_MACRO_H__
> +
> +#include <asm/system.h>
> +
>  #ifdef __ASSEMBLY__
>  
>  /*
> @@ -135,13 +138,18 @@ lr	.req	x30
>  #endif
>  .endm
>  
> -.macro armv8_switch_to_el2_m, xreg1
> -	/* 64bit EL2 | HCE | SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1 */
> -	mov	\xreg1, #0x5b1
> -	msr	scr_el3, \xreg1
> +/*
> + * Switch from EL3 to EL2 for ARMv8
> + * @ep:     kernel entry point
> + * @arch:   machine nr
> + * @ftaddr: fdt address
> + * @flag:   The execution state flag for lower exception
> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> + */
> +.macro armv8_switch_to_el2_m, ep, arch, ftaddr, flag, xreg5

You're never really using the arch and ftaddr arguments. Just describe
in the macro header that they need to be in x1 and x2 respectively and
leave them out of the argument list for the macro.

Also please rename "xreg5" to something like "scratchreg" or "tmp" or
"tmpreg". Then it's more obvious what this argument is about.

>  	msr	cptr_el3, xzr		/* Disable coprocessor traps to EL3 */
> -	mov	\xreg1, #0x33ff
> -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
> +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
> +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
>  
>  	/* Initialize Generic Timers */
>  	msr	cntvoff_el2, xzr
> @@ -152,45 +160,92 @@ lr	.req	x30
>  	 * and RES0 bits (31,30,27,26,24,21,20,17,15-13,10-6) +
>  	 * EE,WXN,I,SA,C,A,M to 0
>  	 */
> -	mov	\xreg1, #0x0830
> -	movk	\xreg1, #0x30C5, lsl #16
> -	msr	sctlr_el2, \xreg1
> +	ldr	\xreg5, =(SCTLR_EL2_BIT29_28_RES1 | SCTLR_EL2_EE_LE |\
> +			  SCTLR_EL2_BIT23_22_RES1 | SCTLR_EL2_WXN_DIS |\
> +			  SCTLR_EL2_BIT18_RES1 | SCTLR_EL2_BIT16_RES1 |\
> +			  SCTLR_EL2_BIT11_RES1 | SCTLR_EL2_BIT5_4_RES1 |\
> +			  SCTLR_EL2_SA_DIS | SCTLR_EL2_DCACHE_DIS |\
> +			  SCTLR_EL2_ALIGN_DIS | SCTLR_EL2_MMU_DIS)

Do you think you could just define something like SCTLR_EL2_RES1 in the
header file that combines all RES1 fields already?

> +	msr	sctlr_el2, \xreg5
> +
> +	mov	\xreg5, sp
> +	msr	sp_el2, \xreg5		/* Migrate SP */
> +	mrs	\xreg5, vbar_el3
> +	msr	vbar_el2, \xreg5	/* Migrate VBAR */
> +
> +	/* Check switch to AArch64 EL2 or AArch32 Hypervisor mode */
> +	ldr	\xreg5, =ES_TO_AARCH32
> +	cmp	\flag, \xreg5

This can be an immediate cmp, no?

  cmp \flag, #ES_TO_AARCH32

> +	b.eq	1f
> +
> +	/*
> +	 * The next lower exception level is AArch64, 64bit EL2 | HCE |
> +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
> +	 */
> +	ldr	\xreg5, =(SCR_EL3_RW_AARCH64 | SCR_EL3_HCE_EN |\
> +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
> +			  SCR_EL3_NS_EN)

Same here - just define a single mask for SCR_EL3_RES1.

> +	msr	scr_el3, \xreg5
>  
>  	/* Return to the EL2_SP2 mode from EL3 */
> -	mov	\xreg1, sp
> -	msr	sp_el2, \xreg1		/* Migrate SP */
> -	mrs	\xreg1, vbar_el3
> -	msr	vbar_el2, \xreg1	/* Migrate VBAR */
> -	mov	\xreg1, #0x3c9
> -	msr	spsr_el3, \xreg1	/* EL2_SP2 | D | A | I | F */
> +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL2H)
> +	msr	spsr_el3, \xreg5
>  	msr	elr_el3, lr
>  	eret
> +
> +1:
> +	/*
> +	 * The next lower exception level is AArch32, 32bit EL2 | HCE |
> +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
> +	 */
> +	ldr	\xreg5, =(SCR_EL3_RW_AARCH32 | SCR_EL3_HCE_EN |\
> +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
> +			  SCR_EL3_NS_EN)
> +	msr	scr_el3, \xreg5
> +
> +	/* Return to AArch32 Hypervisor mode */
> +	ldr     \xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
> +			  SPSR_EL_M_HYP)
> +	msr	spsr_el3, \xreg5
> +	msr     elr_el3, \ep
> +
> +	mov	\ep, #0
> +	eret
>  .endm
>  
> -.macro armv8_switch_to_el1_m, xreg1, xreg2
> +/*
> + * Switch from EL2 to EL1 for ARMv8
> + * @ep:     kernel entry point
> + * @arch:   machine nr
> + * @ftaddr: fdt address
> + * @flag:   The execution state flag for lower exception
> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> + */
> +.macro armv8_switch_to_el1_m, ep, arch, ftaddr, flag, xreg5

All the comments to the function above apply here too :).

>  	/* Initialize Generic Timers */
> -	mrs	\xreg1, cnthctl_el2
> -	orr	\xreg1, \xreg1, #0x3	/* Enable EL1 access to timers */
> -	msr	cnthctl_el2, \xreg1
> +	mrs	\xreg5, cnthctl_el2
> +	/* Enable EL1 access to timers */
> +	orr	\xreg5, \xreg5, #(CNTHCTL_EL2_EL1PCEN_EN |\
> +		CNTHCTL_EL2_EL1PCTEN_EN)
> +	msr	cnthctl_el2, \xreg5
>  	msr	cntvoff_el2, xzr
>  
>  	/* Initilize MPID/MPIDR registers */
> -	mrs	\xreg1, midr_el1
> -	mrs	\xreg2, mpidr_el1
> -	msr	vpidr_el2, \xreg1
> -	msr	vmpidr_el2, \xreg2
> +	mrs	\xreg5, midr_el1
> +	msr	vpidr_el2, \xreg5
> +	mrs	\xreg5, mpidr_el1
> +	msr	vmpidr_el2, \xreg5
>  
>  	/* Disable coprocessor traps */
> -	mov	\xreg1, #0x33ff
> -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
> +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
> +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
>  	msr	hstr_el2, xzr		/* Disable coprocessor traps to EL2 */
> -	mov	\xreg1, #3 << 20
> -	msr	cpacr_el1, \xreg1	/* Enable FP/SIMD at EL1 */
> -
> -	/* Initialize HCR_EL2 */
> -	mov	\xreg1, #(1 << 31)		/* 64bit EL1 */
> -	orr	\xreg1, \xreg1, #(1 << 29)	/* Disable HVC */
> -	msr	hcr_el2, \xreg1
> +	mov	\xreg5, #(CPACR_EL1_FPEN_EN)
> +	msr	cpacr_el1, \xreg5	/* Enable FP/SIMD at EL1 */
>  
>  	/* SCTLR_EL1 initialization
>  	 *
> @@ -199,19 +254,56 @@ lr	.req	x30
>  	 * UCI,EE,EOE,WXN,nTWE,nTWI,UCT,DZE,I,UMA,SED,ITD,
>  	 * CP15BEN,SA0,SA,C,A,M to 0
>  	 */
> -	mov	\xreg1, #0x0800
> -	movk	\xreg1, #0x30d0, lsl #16
> -	msr	sctlr_el1, \xreg1
> +	ldr	\xreg5, =(SCTLR_EL1_BIT29_28_RES1 | SCTLR_EL1_UCI_DIS |\
> +			  SCTLR_EL1_EE_LE | SCTLR_EL1_BIT23_22_RES1 |\
> +			  SCTLR_EL1_BIT20_RES1 | SCTLR_EL1_WXN_DIS |\
> +			  SCTLR_EL1_NTWE_DIS | SCTLR_EL1_NTWI_DIS |\
> +			  SCTLR_EL1_UCT_DIS | SCTLR_EL1_DZE_DIS |\
> +			  SCTLR_EL1_ICACHE_DIS | SCTLR_EL1_BIT11_RES1 |\
> +			  SCTLR_EL1_UMA_DIS | SCTLR_EL1_SED_EN |\
> +			  SCTLR_EL1_ITD_EN | SCTLR_EL1_CP15BEN_DIS |\
> +			  SCTLR_EL1_SA0_DIS | SCTLR_EL1_SA_DIS |\
> +			  SCTLR_EL1_DCACHE_DIS | SCTLR_EL1_ALIGN_DIS |\
> +			  SCTLR_EL1_MMU_DIS)

Why are the caches getting disabled? Or did you just convert the old
value to the define list?

> +	msr	sctlr_el1, \xreg5
> +
> +	mov	\xreg5, sp
> +	msr	sp_el1, \xreg5		/* Migrate SP */
> +	mrs	\xreg5, vbar_el2
> +	msr	vbar_el1, \xreg5	/* Migrate VBAR */
> +
> +	/* Check switch to AArch64 EL1 or AArch32 Supervisor mode */
> +	ldr	\xreg5, =ES_TO_AARCH32
> +	cmp	\flag, \xreg5
> +	b.eq	1f
> +
> +	/* Initialize HCR_EL2 */
> +	ldr	\xreg5, =(HCR_EL2_RW_AARCH64 | HCR_EL2_HCD_DIS)
> +	msr	hcr_el2, \xreg5
>  
>  	/* Return to the EL1_SP1 mode from EL2 */
> -	mov	\xreg1, sp
> -	msr	sp_el1, \xreg1		/* Migrate SP */
> -	mrs	\xreg1, vbar_el2
> -	msr	vbar_el1, \xreg1	/* Migrate VBAR */
> -	mov	\xreg1, #0x3c5
> -	msr	spsr_el2, \xreg1	/* EL1_SP1 | D | A | I | F */
> +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL1H)
> +	msr	spsr_el2, \xreg5
>  	msr	elr_el2, lr
>  	eret
> +
> +1:
> +	/* Initialize HCR_EL2 */
> +	ldr	\xreg5, =(HCR_EL2_RW_AARCH32 | HCR_EL2_HCD_DIS)
> +	msr	hcr_el2, \xreg5
> +
> +	/* Return to AArch32 Supervisor mode from EL2 */
> +	ldr	\xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
> +			  SPSR_EL_M_SVC)
> +	msr     spsr_el2, \xreg5
> +	msr     elr_el2, \ep
> +
> +	mov	\ep, #0
> +	eret
>  .endm
>  
>  #if defined(CONFIG_GICV3)
> diff --git a/arch/arm/include/asm/system.h b/arch/arm/include/asm/system.h
> index 9ae890a..0271ddb 100644
> --- a/arch/arm/include/asm/system.h
> +++ b/arch/arm/include/asm/system.h
> @@ -17,6 +17,102 @@
>  #define CR_WXN		(1 << 19)	/* Write Permision Imply XN	*/
>  #define CR_EE		(1 << 25)	/* Exception (Big) Endian	*/
>  
> +#define ES_TO_AARCH64		1
> +#define ES_TO_AARCH32		0
> +
> +/*
> + * SCR_EL3 bits definitions
> + */
> +#define SCR_EL3_RW_AARCH64	(1 << 10) /* Next lower level is AArch64     */
> +#define SCR_EL3_RW_AARCH32	(0 << 10) /* Lower lowers level are AArch32  */
> +#define SCR_EL3_HCE_EN		(1 << 8)  /* Hypervisor Call enable          */
> +#define SCR_EL3_SMD_DIS		(1 << 7)  /* Secure Monitor Call disable     */
> +#define SCR_EL3_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1                  */
> +#define SCR_EL3_NS_EN		(1 << 0)  /* EL0 and EL1 in Non-scure state  */
> +
> +/*
> + * SPSR_EL3/SPSR_EL2 bits definitions
> + */
> +#define SPSR_EL_END_LE		(0 << 9)  /* Exception Little-endian          */
> +#define SPSR_EL_DEBUG_MASK	(1 << 9)  /* Debug exception masked           */
> +#define SPSR_EL_ASYN_MASK	(1 << 8)  /* Asynchronous data abort masked   */
> +#define SPSR_EL_SERR_MASK	(1 << 8)  /* System Error exception masked    */
> +#define SPSR_EL_IRQ_MASK	(1 << 7)  /* IRQ exception masked             */
> +#define SPSR_EL_FIQ_MASK	(1 << 6)  /* FIQ exception masked             */
> +#define SPSR_EL_T_A32		(0 << 5)  /* AArch32 instruction set A32      */
> +#define SPSR_EL_M_AARCH64	(0 << 4)  /* Exception taken from AArch64     */
> +#define SPSR_EL_M_AARCH32	(1 << 4)  /* Exception taken from AArch32     */
> +#define SPSR_EL_M_SVC		(0x3)     /* Exception taken from SVC mode    */
> +#define SPSR_EL_M_HYP		(0xa)     /* Exception taken from HYP mode    */
> +#define SPSR_EL_M_EL1H		(5)       /* Exception taken from EL1h mode   */
> +#define SPSR_EL_M_EL2H		(9)       /* Exception taken from EL2h mode   */
> +
> +/*
> + * CPTR_EL2 bits definitions
> + */
> +#define CPTR_EL2_BIT13_12_RES1	(3 << 12) /* Reserved, RES1                   */
> +#define CPTR_EL2_BIT9_0_RES1	(0x3ff)   /* Reserved, RES1                   */
> +
> +/*
> + * SCTLR_EL2 bits definitions
> + */
> +#define SCTLR_EL2_BIT29_28_RES1	(3 << 28) /* Reserved, RES1                   */
> +#define SCTLR_EL2_EE_LE		(0 << 25) /* Exception Little-endian          */
> +#define SCTLR_EL2_BIT23_22_RES1	(3 << 22) /* Reserved, RES1                   */
> +#define SCTLR_EL2_WXN_DIS	(0 << 19) /* Write permission is not XN       */
> +#define SCTLR_EL2_BIT18_RES1	(1 << 18) /* Reserved, RES1                   */
> +#define SCTLR_EL2_BIT16_RES1	(1 << 16) /* Reserved, RES1                   */
> +#define SCTLR_EL2_ICACHE_DIS	(0 << 12) /* Instruction cache disabled       */
> +#define SCTLR_EL2_BIT11_RES1	(1 << 11) /* Reserved, RES1                   */
> +#define SCTLR_EL2_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1                   */
> +#define SCTLR_EL2_SA_DIS	(0 << 3)  /* Stack Alignment Check disabled   */
> +#define SCTLR_EL2_DCACHE_DIS	(0 << 2)  /* Data cache disabled              */
> +#define SCTLR_EL2_ALIGN_DIS	(0 << 1)  /* Alignment check disabled         */
> +#define SCTLR_EL2_MMU_DIS	(0)       /* MMU disabled                     */
> +
> +/*
> + * CNTHCTL_EL2 bits definitions
> + */
> +#define CNTHCTL_EL2_EL1PCEN_EN	(1 << 1)  /* Physical timer regs accessible   */
> +#define CNTHCTL_EL2_EL1PCTEN_EN	(1 << 0)  /* Physical counter accessible      */
> +
> +/*
> + * HCR_EL2 bits definitions
> + */
> +#define HCR_EL2_RW_AARCH64	(1 << 31) /* EL1 is AArch64                   */
> +#define HCR_EL2_RW_AARCH32	(0 << 31) /* Lower levels are AArch32         */
> +#define HCR_EL2_HCD_DIS		(1 << 29) /* Hypervisor Call disabled         */
> +
> +/*
> + * CPACR_EL1 bits definitions
> + */
> +#define CPACR_EL1_FPEN_EN	(3 << 20) /* SIMD and FP instruction enabled  */
> +
> +/*
> + * SCTLR_EL1 bits definitions
> + */
> +#define SCTLR_EL1_BIT29_28_RES1	(3 << 28) /* Reserved, RES1                   */
> +#define SCTLR_EL1_UCI_DIS	(0 << 26) /* Cache instruction disabled       */
> +#define SCTLR_EL1_EE_LE		(0 << 25) /* Exception Little-endian          */
> +#define SCTLR_EL1_BIT23_22_RES1	(3 << 22) /* Reserved, RES1                   */
> +#define SCTLR_EL1_BIT20_RES1	(1 << 20) /* Reserved, RES1                   */
> +#define SCTLR_EL1_WXN_DIS	(0 << 19) /* Write permission is not XN       */
> +#define SCTLR_EL1_NTWE_DIS	(0 << 18) /* WFE instruction disabled         */
> +#define SCTLR_EL1_NTWI_DIS	(0 << 16) /* WFI instruction disabled         */
> +#define SCTLR_EL1_UCT_DIS	(0 << 15) /* CTR_EL0 access disabled          */
> +#define SCTLR_EL1_DZE_DIS	(0 << 14) /* DC ZVA instruction disabled      */
> +#define SCTLR_EL1_ICACHE_DIS	(0 << 12) /* Instruction cache disabled       */
> +#define SCTLR_EL1_BIT11_RES1	(1 << 11) /* Reserved, RES1                   */
> +#define SCTLR_EL1_UMA_DIS	(0 << 9)  /* User Mask Access disabled        */
> +#define SCTLR_EL1_SED_EN	(0 << 8)  /* SETEND instruction enabled       */
> +#define SCTLR_EL1_ITD_EN	(0 << 7)  /* IT instruction enabled           */
> +#define SCTLR_EL1_CP15BEN_DIS	(0 << 5)  /* CP15 barrier operation disabled  */
> +#define SCTLR_EL1_SA0_DIS	(0 << 4)  /* Stack Alignment EL0 disabled     */
> +#define SCTLR_EL1_SA_DIS	(0 << 3)  /* Stack Alignment EL1 disabled     */
> +#define SCTLR_EL1_DCACHE_DIS	(0 << 2)  /* Data cache disabled              */
> +#define SCTLR_EL1_ALIGN_DIS	(0 << 1)  /* Alignment check disabled         */
> +#define SCTLR_EL1_MMU_DIS	(0)       /* MMU disabled                     */
> +
>  #ifndef __ASSEMBLY__
>  
>  u64 get_page_table_size(void);
> @@ -100,8 +196,26 @@ void __asm_invalidate_icache_all(void);
>  int __asm_flush_l3_cache(void);
>  void __asm_switch_ttbr(u64 new_ttbr);
>  
> -void armv8_switch_to_el2(void);
> -void armv8_switch_to_el1(void);
> +/*
> + * Switch from EL3 to EL2 for ARMv8
> + *
> + * @entry_point: kernel entry point
> + * @mach_nr:     machine nr
> + * @fdt_addr:    fdt address
> + * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
> + */
> +void armv8_switch_to_el2(u64 entry_point, u64 mach_nr, u64 fdt_addr,
> +			 u64 es_flag);
> +/*
> + * Switch from EL2 to EL1 for ARMv8
> + *
> + * @entry_point: kernel entry point
> + * @mach_nr:     machine nr
> + * @fdt_addr:    fdt address
> + * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
> + */
> +void armv8_switch_to_el1(u64 entry_point, u64 mach_nr, u64 fdt_addr,
> +			 u64 es_flag);
>  void gic_init(void);
>  void gic_send_sgi(unsigned long sgino);
>  void wait_for_wakeup(void);
> diff --git a/arch/arm/lib/bootm.c b/arch/arm/lib/bootm.c
> index 0838d89..1bbf85c 100644
> --- a/arch/arm/lib/bootm.c
> +++ b/arch/arm/lib/bootm.c
> @@ -189,13 +189,19 @@ static void setup_end_tag(bd_t *bd)
>  __weak void setup_board_tags(struct tag **in_params) {}
>  
>  #ifdef CONFIG_ARM64
> -static void do_nonsec_virt_switch(void)
> +static void do_nonsec_virt_switch(bootm_headers_t *images, int flag)
>  {
>  	smp_kick_all_cpus();
>  	dcache_disable();	/* flush cache before swtiching to EL2 */
> -	armv8_switch_to_el2();
> +
>  #ifdef CONFIG_ARMV8_SWITCH_TO_EL1
> -	armv8_switch_to_el1();
> +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
> +			    (u64)images->ft_addr, ES_TO_AARCH64);
> +	armv8_switch_to_el1((u64)images->ep, (u64)gd->bd->bi_arch_number,
> +			    (u64)images->ft_addr, flag);
> +#else
> +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
> +			    (u64)images->ft_addr, flag);
>  #endif
>  }
>  #endif
> @@ -275,6 +281,7 @@ static void boot_jump_linux(bootm_headers_t *images, int flag)
>  	void (*kernel_entry)(void *fdt_addr, void *res0, void *res1,
>  			void *res2);
>  	int fake = (flag & BOOTM_STATE_OS_FAKE_GO);
> +	int es_flag = ES_TO_AARCH64;
>  
>  	kernel_entry = (void (*)(void *fdt_addr, void *res0, void *res1,
>  				void *res2))images->ep;
> @@ -286,7 +293,11 @@ static void boot_jump_linux(bootm_headers_t *images, int flag)
>  	announce_and_cleanup(fake);
>  
>  	if (!fake) {
> -		do_nonsec_virt_switch();
> +		if ((IH_ARCH_DEFAULT == IH_ARCH_ARM64) &&
> +		    (images->os.arch == IH_ARCH_ARM))
> +			es_flag = ES_TO_AARCH32;
> +
> +		do_nonsec_virt_switch(images, es_flag);
>  		kernel_entry(images->ft_addr, NULL, NULL, NULL);
>  	}
>  #else
> diff --git a/common/image-fit.c b/common/image-fit.c
> index 9873957..982021b 100644
> --- a/common/image-fit.c
> +++ b/common/image-fit.c
> @@ -26,6 +26,7 @@ DECLARE_GLOBAL_DATA_PTR;
>  #include <u-boot/md5.h>
>  #include <u-boot/sha1.h>
>  #include <u-boot/sha256.h>
> +#include <generated/autoconf.h>
>  
>  /*****************************************************************************/
>  /* New uImage format routines */
> @@ -1164,7 +1165,9 @@ int fit_image_check_arch(const void *fit, int noffset, uint8_t arch)
>  	if (fit_image_get_arch(fit, noffset, &image_arch))
>  		return 0;
>  	return (arch == image_arch) ||
> -		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64);
> +		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64) ||
> +		(arch == IH_ARCH_ARM64 && image_arch == IH_ARCH_ARM &&
> +		 CONFIG_ARM64_SUPPORT_AARCH32);

Uh, I think this define is simply not defined for non-compatible systems
(including non-arm64). This really needs to go into an #ifdef.

Thanks a lot for all the refactoring work :). This patch already reads
much nicer than the original code. I guess with next round we'll get
this into really good shape.


Alex

>  }
>  
>  /**
> @@ -1593,6 +1596,9 @@ int fit_image_load(bootm_headers_t *images, ulong addr,
>  	int type_ok, os_ok;
>  	ulong load, data, len;
>  	uint8_t os;
> +#ifndef USE_HOSTCC
> +	uint8_t os_arch;
> +#endif
>  	const char *prop_name;
>  	int ret;
>  
> @@ -1676,6 +1682,12 @@ int fit_image_load(bootm_headers_t *images, ulong addr,
>  		return -ENOEXEC;
>  	}
>  #endif
> +
> +#ifndef USE_HOSTCC
> +	fit_image_get_arch(fit, noffset, &os_arch);
> +	images->os.arch = os_arch;
> +#endif
> +
>  	if (image_type == IH_TYPE_FLATDT &&
>  	    !fit_image_check_comp(fit, noffset, IH_COMP_NONE)) {
>  		puts("FDT image is compressed");
> 

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Huan Wang]

Hi, York,

> On 05/26/2016 01:51 AM, Alison Wang wrote:
> > To support loading a 32-bit OS, the execution state will change from
> > AArch64 to AArch32 when jumping to kernel.
> >
> > The architecture information will be got through checking FIT image,
> > then U-Boot will load 32-bit OS or 64-bit OS automatically.
> >
> > Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
> > Signed-off-by: Alison Wang <alison.wang@nxp.com>
> > Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
> > ---
> > Changes in v3:
> > - Comments the functions and the arguments.
> > - Rename the real parameters.
> > - Use the macros instead of the magic values.
> > - Remove the redundant codes.
> > - Clean up all of the mess in boot_jump_linux().
> > - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system
> doesn't support AArch32 state.
> >
> > Changes in v2:
> > - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is
> used
> >   to switch to AArch64 EL2 or AArch32 Hyp.
> > - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is
> used
> >   to switch to AArch64 EL1 or AArch32 SVC.
> >
> >  arch/arm/Kconfig                |   6 ++
> >  arch/arm/cpu/armv8/transition.S |   8 +-
> >  arch/arm/include/asm/macro.h    | 174 ++++++++++++++++++++++++++++++-
> ---------
> >  arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
> >  arch/arm/lib/bootm.c            |  19 ++++-
> >  common/image-fit.c              |  14 +++-
> >  6 files changed, 287 insertions(+), 52 deletions(-)
> >
> > diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index
> > 729b181..794cb4f 100644
> > --- a/arch/arm/Kconfig
> > +++ b/arch/arm/Kconfig
> > @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
> >  	  If SoC does not support L2CACHE or one do not want to enable
> >  	  L2CACHE, choose this option.
> >
> > +config ARM64_SUPPORT_AARCH32
> > +	bool "ARM64 system support AArch32 execution state"
> > +	default y if ARM64 && !CONFIG_THUNDERX
> > +	help
> > +	  This ARM64 system supports AArch32 execution state.
> > +
> >  choice
> >  	prompt "Target select"
> >  	default TARGET_HIKEY
> > diff --git a/arch/arm/cpu/armv8/transition.S
> > b/arch/arm/cpu/armv8/transition.S index 253a39b..417e8b4 100644
> > --- a/arch/arm/cpu/armv8/transition.S
> > +++ b/arch/arm/cpu/armv8/transition.S
> > @@ -11,13 +11,13 @@
> >  #include <asm/macro.h>
> >
> >  ENTRY(armv8_switch_to_el2)
> > -	switch_el x0, 1f, 0f, 0f
> > +	switch_el x4, 1f, 0f, 0f
> >  0:	ret
> > -1:	armv8_switch_to_el2_m x0
> > +1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
> >  ENDPROC(armv8_switch_to_el2)
> 
> Alison,
> 
> I see you are chaning macro armv8_switch_to_el2. You may have missed one
> in start.S. This is the code you can't test with our ARMv8 chip.
> 
[Alison Wang] Yes, I missed that one. In start.S, armv8_switch_to_el2 will
be called for secondary cores. I will set argument x3 as ES_TO_AARCH64 for
it. As the way to notice secondary cores about the architecture of OS may
be different for other ARMv8 chips, I will not add ES_TO_AARCH32 in start.S.
Do you agree?

Thanks for your reminder.

Best Regards,
Alison Wang

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Huan Wang]

> On 26.05.16 10:41, Alison Wang wrote:
> > To support loading a 32-bit OS, the execution state will change from
> > AArch64 to AArch32 when jumping to kernel.
> >
> > The architecture information will be got through checking FIT image,
> > then U-Boot will load 32-bit OS or 64-bit OS automatically.
> >
> > Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
> > Signed-off-by: Alison Wang <alison.wang@nxp.com>
> > Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
> > ---
> > Changes in v3:
> > - Comments the functions and the arguments.
> > - Rename the real parameters.
> > - Use the macros instead of the magic values.
> > - Remove the redundant codes.
> > - Clean up all of the mess in boot_jump_linux().
> > - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system
> doesn't support AArch32 state.
> >
> > Changes in v2:
> > - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is
> used
> >   to switch to AArch64 EL2 or AArch32 Hyp.
> > - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is
> used
> >   to switch to AArch64 EL1 or AArch32 SVC.
> >
> >  arch/arm/Kconfig                |   6 ++
> >  arch/arm/cpu/armv8/transition.S |   8 +-
> >  arch/arm/include/asm/macro.h    | 174 ++++++++++++++++++++++++++++++-
> ---------
> >  arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
> >  arch/arm/lib/bootm.c            |  19 ++++-
> >  common/image-fit.c              |  14 +++-
> >  6 files changed, 287 insertions(+), 52 deletions(-)
> >
> > diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index
> > 729b181..794cb4f 100644
> > --- a/arch/arm/Kconfig
> > +++ b/arch/arm/Kconfig
> > @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
> >  	  If SoC does not support L2CACHE or one do not want to enable
> >  	  L2CACHE, choose this option.
> >
> > +config ARM64_SUPPORT_AARCH32
> > +	bool "ARM64 system support AArch32 execution state"
> > +	default y if ARM64 && !CONFIG_THUNDERX
> 
> I don't think that works. In Kconfig you need to omit the CONFIG_ parts.
> It's "ARM64" also, not "CONFIG_ARM64".
> 
> Please just try to run a local make defconfig on the thunderx reference
> system and see whether it includes the option or not.
[Alison Wang] Yes, I missed it. Will change in the next version.
> 
> > +	help
> > +	  This ARM64 system supports AArch32 execution state.
> > +
> >  choice
> >  	prompt "Target select"
> >  	default TARGET_HIKEY
> > diff --git a/arch/arm/cpu/armv8/transition.S
> > b/arch/arm/cpu/armv8/transition.S index 253a39b..417e8b4 100644
> > --- a/arch/arm/cpu/armv8/transition.S
> > +++ b/arch/arm/cpu/armv8/transition.S
> > @@ -11,13 +11,13 @@
> >  #include <asm/macro.h>
> >
> >  ENTRY(armv8_switch_to_el2)
> > -	switch_el x0, 1f, 0f, 0f
> > +	switch_el x4, 1f, 0f, 0f
> >  0:	ret
> > -1:	armv8_switch_to_el2_m x0
> > +1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
> >  ENDPROC(armv8_switch_to_el2)
> >
> >  ENTRY(armv8_switch_to_el1)
> > -	switch_el x0, 0f, 1f, 0f
> > +	switch_el x4, 0f, 1f, 0f
> >  0:	ret
> > -1:	armv8_switch_to_el1_m x0, x1
> > +1:	armv8_switch_to_el1_m x0, x1, x2, x3, x4
> >  ENDPROC(armv8_switch_to_el1)
> > diff --git a/arch/arm/include/asm/macro.h
> > b/arch/arm/include/asm/macro.h index 9bb0efa..dd2c510 100644
> > --- a/arch/arm/include/asm/macro.h
> > +++ b/arch/arm/include/asm/macro.h
> > @@ -8,6 +8,9 @@
> >
> >  #ifndef __ASM_ARM_MACRO_H__
> >  #define __ASM_ARM_MACRO_H__
> > +
> > +#include <asm/system.h>
> > +
> >  #ifdef __ASSEMBLY__
> >
> >  /*
> > @@ -135,13 +138,18 @@ lr	.req	x30
> >  #endif
> >  .endm
> >
> > -.macro armv8_switch_to_el2_m, xreg1
> > -	/* 64bit EL2 | HCE | SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1
> */
> > -	mov	\xreg1, #0x5b1
> > -	msr	scr_el3, \xreg1
> > +/*
> > + * Switch from EL3 to EL2 for ARMv8
> > + * @ep:     kernel entry point
> > + * @arch:   machine nr
> > + * @ftaddr: fdt address
> > + * @flag:   The execution state flag for lower exception
> > + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> > + */
> > +.macro armv8_switch_to_el2_m, ep, arch, ftaddr, flag, xreg5
> 
> You're never really using the arch and ftaddr arguments. Just describe
> in the macro header that they need to be in x1 and x2 respectively and
> leave them out of the argument list for the macro.
[Alison Wang] Do you mean rename arch and ftaddr to x1 and x2 here?
> 
> Also please rename "xreg5" to something like "scratchreg" or "tmp" or
> "tmpreg". Then it's more obvious what this argument is about.
[Alison Wang] Ok.
> 
> >  	msr	cptr_el3, xzr		/* Disable coprocessor traps to EL3
> */
> > -	mov	\xreg1, #0x33ff
> > -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
> > +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
> > +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
> >
> >  	/* Initialize Generic Timers */
> >  	msr	cntvoff_el2, xzr
> > @@ -152,45 +160,92 @@ lr	.req	x30
> >  	 * and RES0 bits (31,30,27,26,24,21,20,17,15-13,10-6) +
> >  	 * EE,WXN,I,SA,C,A,M to 0
> >  	 */
> > -	mov	\xreg1, #0x0830
> > -	movk	\xreg1, #0x30C5, lsl #16
> > -	msr	sctlr_el2, \xreg1
> > +	ldr	\xreg5, =(SCTLR_EL2_BIT29_28_RES1 | SCTLR_EL2_EE_LE |\
> > +			  SCTLR_EL2_BIT23_22_RES1 | SCTLR_EL2_WXN_DIS |\
> > +			  SCTLR_EL2_BIT18_RES1 | SCTLR_EL2_BIT16_RES1 |\
> > +			  SCTLR_EL2_BIT11_RES1 | SCTLR_EL2_BIT5_4_RES1 |\
> > +			  SCTLR_EL2_SA_DIS | SCTLR_EL2_DCACHE_DIS |\
> > +			  SCTLR_EL2_ALIGN_DIS | SCTLR_EL2_MMU_DIS)
> 
> Do you think you could just define something like SCTLR_EL2_RES1 in the
> header file that combines all RES1 fields already?
[Alison Wang] Ok, I think so.
> 
> > +	msr	sctlr_el2, \xreg5
> > +
> > +	mov	\xreg5, sp
> > +	msr	sp_el2, \xreg5		/* Migrate SP */
> > +	mrs	\xreg5, vbar_el3
> > +	msr	vbar_el2, \xreg5	/* Migrate VBAR */
> > +
> > +	/* Check switch to AArch64 EL2 or AArch32 Hypervisor mode */
> > +	ldr	\xreg5, =ES_TO_AARCH32
> > +	cmp	\flag, \xreg5
> 
> This can be an immediate cmp, no?
> 
>   cmp \flag, #ES_TO_AARCH32
[Alison Wang] Yes, it can. Will change in the next version.
> 
> > +	b.eq	1f
> > +
> > +	/*
> > +	 * The next lower exception level is AArch64, 64bit EL2 | HCE |
> > +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
> > +	 */
> > +	ldr	\xreg5, =(SCR_EL3_RW_AARCH64 | SCR_EL3_HCE_EN |\
> > +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
> > +			  SCR_EL3_NS_EN)
> 
> Same here - just define a single mask for SCR_EL3_RES1.
[Alison Wang] Ok.
> 
> > +	msr	scr_el3, \xreg5
> >
> >  	/* Return to the EL2_SP2 mode from EL3 */
> > -	mov	\xreg1, sp
> > -	msr	sp_el2, \xreg1		/* Migrate SP */
> > -	mrs	\xreg1, vbar_el3
> > -	msr	vbar_el2, \xreg1	/* Migrate VBAR */
> > -	mov	\xreg1, #0x3c9
> > -	msr	spsr_el3, \xreg1	/* EL2_SP2 | D | A | I | F */
> > +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
> > +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> > +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL2H)
> > +	msr	spsr_el3, \xreg5
> >  	msr	elr_el3, lr
> >  	eret
> > +
> > +1:
> > +	/*
> > +	 * The next lower exception level is AArch32, 32bit EL2 | HCE |
> > +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
> > +	 */
> > +	ldr	\xreg5, =(SCR_EL3_RW_AARCH32 | SCR_EL3_HCE_EN |\
> > +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
> > +			  SCR_EL3_NS_EN)
> > +	msr	scr_el3, \xreg5
> > +
> > +	/* Return to AArch32 Hypervisor mode */
> > +	ldr     \xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
> > +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> > +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
> > +			  SPSR_EL_M_HYP)
> > +	msr	spsr_el3, \xreg5
> > +	msr     elr_el3, \ep
> > +
> > +	mov	\ep, #0
> > +	eret
> >  .endm
> >
> > -.macro armv8_switch_to_el1_m, xreg1, xreg2
> > +/*
> > + * Switch from EL2 to EL1 for ARMv8
> > + * @ep:     kernel entry point
> > + * @arch:   machine nr
> > + * @ftaddr: fdt address
> > + * @flag:   The execution state flag for lower exception
> > + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> > + */
> > +.macro armv8_switch_to_el1_m, ep, arch, ftaddr, flag, xreg5
> 
> All the comments to the function above apply here too :).
[Alison Wang] Ok.
> 
> >  	/* Initialize Generic Timers */
> > -	mrs	\xreg1, cnthctl_el2
> > -	orr	\xreg1, \xreg1, #0x3	/* Enable EL1 access to timers */
> > -	msr	cnthctl_el2, \xreg1
> > +	mrs	\xreg5, cnthctl_el2
> > +	/* Enable EL1 access to timers */
> > +	orr	\xreg5, \xreg5, #(CNTHCTL_EL2_EL1PCEN_EN |\
> > +		CNTHCTL_EL2_EL1PCTEN_EN)
> > +	msr	cnthctl_el2, \xreg5
> >  	msr	cntvoff_el2, xzr
> >
> >  	/* Initilize MPID/MPIDR registers */
> > -	mrs	\xreg1, midr_el1
> > -	mrs	\xreg2, mpidr_el1
> > -	msr	vpidr_el2, \xreg1
> > -	msr	vmpidr_el2, \xreg2
> > +	mrs	\xreg5, midr_el1
> > +	msr	vpidr_el2, \xreg5
> > +	mrs	\xreg5, mpidr_el1
> > +	msr	vmpidr_el2, \xreg5
> >
> >  	/* Disable coprocessor traps */
> > -	mov	\xreg1, #0x33ff
> > -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
> > +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
> > +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
> >  	msr	hstr_el2, xzr		/* Disable coprocessor traps to EL2
> */
> > -	mov	\xreg1, #3 << 20
> > -	msr	cpacr_el1, \xreg1	/* Enable FP/SIMD at EL1 */
> > -
> > -	/* Initialize HCR_EL2 */
> > -	mov	\xreg1, #(1 << 31)		/* 64bit EL1 */
> > -	orr	\xreg1, \xreg1, #(1 << 29)	/* Disable HVC */
> > -	msr	hcr_el2, \xreg1
> > +	mov	\xreg5, #(CPACR_EL1_FPEN_EN)
> > +	msr	cpacr_el1, \xreg5	/* Enable FP/SIMD at EL1 */
> >
> >  	/* SCTLR_EL1 initialization
> >  	 *
> > @@ -199,19 +254,56 @@ lr	.req	x30
> >  	 * UCI,EE,EOE,WXN,nTWE,nTWI,UCT,DZE,I,UMA,SED,ITD,
> >  	 * CP15BEN,SA0,SA,C,A,M to 0
> >  	 */
> > -	mov	\xreg1, #0x0800
> > -	movk	\xreg1, #0x30d0, lsl #16
> > -	msr	sctlr_el1, \xreg1
> > +	ldr	\xreg5, =(SCTLR_EL1_BIT29_28_RES1 | SCTLR_EL1_UCI_DIS |\
> > +			  SCTLR_EL1_EE_LE | SCTLR_EL1_BIT23_22_RES1 |\
> > +			  SCTLR_EL1_BIT20_RES1 | SCTLR_EL1_WXN_DIS |\
> > +			  SCTLR_EL1_NTWE_DIS | SCTLR_EL1_NTWI_DIS |\
> > +			  SCTLR_EL1_UCT_DIS | SCTLR_EL1_DZE_DIS |\
> > +			  SCTLR_EL1_ICACHE_DIS | SCTLR_EL1_BIT11_RES1 |\
> > +			  SCTLR_EL1_UMA_DIS | SCTLR_EL1_SED_EN |\
> > +			  SCTLR_EL1_ITD_EN | SCTLR_EL1_CP15BEN_DIS |\
> > +			  SCTLR_EL1_SA0_DIS | SCTLR_EL1_SA_DIS |\
> > +			  SCTLR_EL1_DCACHE_DIS | SCTLR_EL1_ALIGN_DIS |\
> > +			  SCTLR_EL1_MMU_DIS)
> 
> Why are the caches getting disabled? Or did you just convert the old
> value to the define list?
[Alison Wang] Yes, I just convert the old value to the define list.
> 
> > +	msr	sctlr_el1, \xreg5
> > +
> > +	mov	\xreg5, sp
> > +	msr	sp_el1, \xreg5		/* Migrate SP */
> > +	mrs	\xreg5, vbar_el2
> > +	msr	vbar_el1, \xreg5	/* Migrate VBAR */
> > +
> > +	/* Check switch to AArch64 EL1 or AArch32 Supervisor mode */
> > +	ldr	\xreg5, =ES_TO_AARCH32
> > +	cmp	\flag, \xreg5
> > +	b.eq	1f
> > +
> > +	/* Initialize HCR_EL2 */
> > +	ldr	\xreg5, =(HCR_EL2_RW_AARCH64 | HCR_EL2_HCD_DIS)
> > +	msr	hcr_el2, \xreg5
> >
> >  	/* Return to the EL1_SP1 mode from EL2 */
> > -	mov	\xreg1, sp
> > -	msr	sp_el1, \xreg1		/* Migrate SP */
> > -	mrs	\xreg1, vbar_el2
> > -	msr	vbar_el1, \xreg1	/* Migrate VBAR */
> > -	mov	\xreg1, #0x3c5
> > -	msr	spsr_el2, \xreg1	/* EL1_SP1 | D | A | I | F */
> > +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
> > +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> > +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL1H)
> > +	msr	spsr_el2, \xreg5
> >  	msr	elr_el2, lr
> >  	eret
> > +
> > +1:
> > +	/* Initialize HCR_EL2 */
> > +	ldr	\xreg5, =(HCR_EL2_RW_AARCH32 | HCR_EL2_HCD_DIS)
> > +	msr	hcr_el2, \xreg5
> > +
> > +	/* Return to AArch32 Supervisor mode from EL2 */
> > +	ldr	\xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
> > +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> > +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
> > +			  SPSR_EL_M_SVC)
> > +	msr     spsr_el2, \xreg5
> > +	msr     elr_el2, \ep
> > +
> > +	mov	\ep, #0
> > +	eret
> >  .endm
> >
> >  #if defined(CONFIG_GICV3)
> > diff --git a/arch/arm/include/asm/system.h
> > b/arch/arm/include/asm/system.h index 9ae890a..0271ddb 100644
> > --- a/arch/arm/include/asm/system.h
> > +++ b/arch/arm/include/asm/system.h
> > @@ -17,6 +17,102 @@
> >  #define CR_WXN		(1 << 19)	/* Write Permision Imply XN	*/
> >  #define CR_EE		(1 << 25)	/* Exception (Big) Endian	*/
> >
> > +#define ES_TO_AARCH64		1
> > +#define ES_TO_AARCH32		0
> > +
> > +/*
> > + * SCR_EL3 bits definitions
> > + */
> > +#define SCR_EL3_RW_AARCH64	(1 << 10) /* Next lower level is AArch64
> */
> > +#define SCR_EL3_RW_AARCH32	(0 << 10) /* Lower lowers level are
> AArch32  */
> > +#define SCR_EL3_HCE_EN		(1 << 8)  /* Hypervisor Call enable
> */
> > +#define SCR_EL3_SMD_DIS		(1 << 7)  /* Secure Monitor Call
> disable     */
> > +#define SCR_EL3_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1
> */
> > +#define SCR_EL3_NS_EN		(1 << 0)  /* EL0 and EL1 in Non-scure
> state  */
> > +
> > +/*
> > + * SPSR_EL3/SPSR_EL2 bits definitions  */
> > +#define SPSR_EL_END_LE		(0 << 9)  /* Exception Little-endian
> */
> > +#define SPSR_EL_DEBUG_MASK	(1 << 9)  /* Debug exception masked
> */
> > +#define SPSR_EL_ASYN_MASK	(1 << 8)  /* Asynchronous data abort
> masked   */
> > +#define SPSR_EL_SERR_MASK	(1 << 8)  /* System Error exception masked
> */
> > +#define SPSR_EL_IRQ_MASK	(1 << 7)  /* IRQ exception masked
> */
> > +#define SPSR_EL_FIQ_MASK	(1 << 6)  /* FIQ exception masked
> */
> > +#define SPSR_EL_T_A32		(0 << 5)  /* AArch32 instruction set A32
> */
> > +#define SPSR_EL_M_AARCH64	(0 << 4)  /* Exception taken from AArch64
> */
> > +#define SPSR_EL_M_AARCH32	(1 << 4)  /* Exception taken from AArch32
> */
> > +#define SPSR_EL_M_SVC		(0x3)     /* Exception taken from SVC mode
> */
> > +#define SPSR_EL_M_HYP		(0xa)     /* Exception taken from HYP mode
> */
> > +#define SPSR_EL_M_EL1H		(5)       /* Exception taken from
> EL1h mode   */
> > +#define SPSR_EL_M_EL2H		(9)       /* Exception taken from
> EL2h mode   */
> > +
> > +/*
> > + * CPTR_EL2 bits definitions
> > + */
> > +#define CPTR_EL2_BIT13_12_RES1	(3 << 12) /* Reserved, RES1
> */
> > +#define CPTR_EL2_BIT9_0_RES1	(0x3ff)   /* Reserved, RES1
> */
> > +
> > +/*
> > + * SCTLR_EL2 bits definitions
> > + */
> > +#define SCTLR_EL2_BIT29_28_RES1	(3 << 28) /* Reserved, RES1
> */
> > +#define SCTLR_EL2_EE_LE		(0 << 25) /* Exception Little-endian
> */
> > +#define SCTLR_EL2_BIT23_22_RES1	(3 << 22) /* Reserved, RES1
> */
> > +#define SCTLR_EL2_WXN_DIS	(0 << 19) /* Write permission is not XN
> */
> > +#define SCTLR_EL2_BIT18_RES1	(1 << 18) /* Reserved, RES1
> */
> > +#define SCTLR_EL2_BIT16_RES1	(1 << 16) /* Reserved, RES1
> */
> > +#define SCTLR_EL2_ICACHE_DIS	(0 << 12) /* Instruction cache
> disabled       */
> > +#define SCTLR_EL2_BIT11_RES1	(1 << 11) /* Reserved, RES1
> */
> > +#define SCTLR_EL2_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1
> */
> > +#define SCTLR_EL2_SA_DIS	(0 << 3)  /* Stack Alignment Check
> disabled   */
> > +#define SCTLR_EL2_DCACHE_DIS	(0 << 2)  /* Data cache disabled
> */
> > +#define SCTLR_EL2_ALIGN_DIS	(0 << 1)  /* Alignment check disabled
> */
> > +#define SCTLR_EL2_MMU_DIS	(0)       /* MMU disabled
> */
> > +
> > +/*
> > + * CNTHCTL_EL2 bits definitions
> > + */
> > +#define CNTHCTL_EL2_EL1PCEN_EN	(1 << 1)  /* Physical timer regs
> accessible   */
> > +#define CNTHCTL_EL2_EL1PCTEN_EN	(1 << 0)  /* Physical counter
> accessible      */
> > +
> > +/*
> > + * HCR_EL2 bits definitions
> > + */
> > +#define HCR_EL2_RW_AARCH64	(1 << 31) /* EL1 is AArch64
> */
> > +#define HCR_EL2_RW_AARCH32	(0 << 31) /* Lower levels are AArch32
> */
> > +#define HCR_EL2_HCD_DIS		(1 << 29) /* Hypervisor Call
> disabled         */
> > +
> > +/*
> > + * CPACR_EL1 bits definitions
> > + */
> > +#define CPACR_EL1_FPEN_EN	(3 << 20) /* SIMD and FP instruction
> enabled  */
> > +
> > +/*
> > + * SCTLR_EL1 bits definitions
> > + */
> > +#define SCTLR_EL1_BIT29_28_RES1	(3 << 28) /* Reserved, RES1
> */
> > +#define SCTLR_EL1_UCI_DIS	(0 << 26) /* Cache instruction disabled
> */
> > +#define SCTLR_EL1_EE_LE		(0 << 25) /* Exception Little-endian
> */
> > +#define SCTLR_EL1_BIT23_22_RES1	(3 << 22) /* Reserved, RES1
> */
> > +#define SCTLR_EL1_BIT20_RES1	(1 << 20) /* Reserved, RES1
> */
> > +#define SCTLR_EL1_WXN_DIS	(0 << 19) /* Write permission is not XN
> */
> > +#define SCTLR_EL1_NTWE_DIS	(0 << 18) /* WFE instruction disabled
> */
> > +#define SCTLR_EL1_NTWI_DIS	(0 << 16) /* WFI instruction disabled
> */
> > +#define SCTLR_EL1_UCT_DIS	(0 << 15) /* CTR_EL0 access disabled
> */
> > +#define SCTLR_EL1_DZE_DIS	(0 << 14) /* DC ZVA instruction disabled
> */
> > +#define SCTLR_EL1_ICACHE_DIS	(0 << 12) /* Instruction cache
> disabled       */
> > +#define SCTLR_EL1_BIT11_RES1	(1 << 11) /* Reserved, RES1
> */
> > +#define SCTLR_EL1_UMA_DIS	(0 << 9)  /* User Mask Access disabled
> */
> > +#define SCTLR_EL1_SED_EN	(0 << 8)  /* SETEND instruction enabled
> */
> > +#define SCTLR_EL1_ITD_EN	(0 << 7)  /* IT instruction enabled
> */
> > +#define SCTLR_EL1_CP15BEN_DIS	(0 << 5)  /* CP15 barrier operation
> disabled  */
> > +#define SCTLR_EL1_SA0_DIS	(0 << 4)  /* Stack Alignment EL0 disabled
> */
> > +#define SCTLR_EL1_SA_DIS	(0 << 3)  /* Stack Alignment EL1 disabled
> */
> > +#define SCTLR_EL1_DCACHE_DIS	(0 << 2)  /* Data cache disabled
> */
> > +#define SCTLR_EL1_ALIGN_DIS	(0 << 1)  /* Alignment check disabled
> */
> > +#define SCTLR_EL1_MMU_DIS	(0)       /* MMU disabled
> */
> > +
> >  #ifndef __ASSEMBLY__
> >
> >  u64 get_page_table_size(void);
> > @@ -100,8 +196,26 @@ void __asm_invalidate_icache_all(void);
> >  int __asm_flush_l3_cache(void);
> >  void __asm_switch_ttbr(u64 new_ttbr);
> >
> > -void armv8_switch_to_el2(void);
> > -void armv8_switch_to_el1(void);
> > +/*
> > + * Switch from EL3 to EL2 for ARMv8
> > + *
> > + * @entry_point: kernel entry point
> > + * @mach_nr:     machine nr
> > + * @fdt_addr:    fdt address
> > + * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
> > + */
> > +void armv8_switch_to_el2(u64 entry_point, u64 mach_nr, u64 fdt_addr,
> > +			 u64 es_flag);
> > +/*
> > + * Switch from EL2 to EL1 for ARMv8
> > + *
> > + * @entry_point: kernel entry point
> > + * @mach_nr:     machine nr
> > + * @fdt_addr:    fdt address
> > + * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
> > + */
> > +void armv8_switch_to_el1(u64 entry_point, u64 mach_nr, u64 fdt_addr,
> > +			 u64 es_flag);
> >  void gic_init(void);
> >  void gic_send_sgi(unsigned long sgino);  void wait_for_wakeup(void);
> > diff --git a/arch/arm/lib/bootm.c b/arch/arm/lib/bootm.c index
> > 0838d89..1bbf85c 100644
> > --- a/arch/arm/lib/bootm.c
> > +++ b/arch/arm/lib/bootm.c
> > @@ -189,13 +189,19 @@ static void setup_end_tag(bd_t *bd)  __weak void
> > setup_board_tags(struct tag **in_params) {}
> >
> >  #ifdef CONFIG_ARM64
> > -static void do_nonsec_virt_switch(void)
> > +static void do_nonsec_virt_switch(bootm_headers_t *images, int flag)
> >  {
> >  	smp_kick_all_cpus();
> >  	dcache_disable();	/* flush cache before swtiching to EL2 */
> > -	armv8_switch_to_el2();
> > +
> >  #ifdef CONFIG_ARMV8_SWITCH_TO_EL1
> > -	armv8_switch_to_el1();
> > +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
> > +			    (u64)images->ft_addr, ES_TO_AARCH64);
> > +	armv8_switch_to_el1((u64)images->ep, (u64)gd->bd->bi_arch_number,
> > +			    (u64)images->ft_addr, flag);
> > +#else
> > +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
> > +			    (u64)images->ft_addr, flag);
> >  #endif
> >  }
> >  #endif
> > @@ -275,6 +281,7 @@ static void boot_jump_linux(bootm_headers_t
> *images, int flag)
> >  	void (*kernel_entry)(void *fdt_addr, void *res0, void *res1,
> >  			void *res2);
> >  	int fake = (flag & BOOTM_STATE_OS_FAKE_GO);
> > +	int es_flag = ES_TO_AARCH64;
> >
> >  	kernel_entry = (void (*)(void *fdt_addr, void *res0, void *res1,
> >  				void *res2))images->ep;
> > @@ -286,7 +293,11 @@ static void boot_jump_linux(bootm_headers_t
> *images, int flag)
> >  	announce_and_cleanup(fake);
> >
> >  	if (!fake) {
> > -		do_nonsec_virt_switch();
> > +		if ((IH_ARCH_DEFAULT == IH_ARCH_ARM64) &&
> > +		    (images->os.arch == IH_ARCH_ARM))
> > +			es_flag = ES_TO_AARCH32;
> > +
> > +		do_nonsec_virt_switch(images, es_flag);
> >  		kernel_entry(images->ft_addr, NULL, NULL, NULL);
> >  	}
> >  #else
> > diff --git a/common/image-fit.c b/common/image-fit.c index
> > 9873957..982021b 100644
> > --- a/common/image-fit.c
> > +++ b/common/image-fit.c
> > @@ -26,6 +26,7 @@ DECLARE_GLOBAL_DATA_PTR;  #include <u-boot/md5.h>
> > #include <u-boot/sha1.h>  #include <u-boot/sha256.h>
> > +#include <generated/autoconf.h>
> >
> >
> > /*********************************************************************
> > ********/
> >  /* New uImage format routines */
> > @@ -1164,7 +1165,9 @@ int fit_image_check_arch(const void *fit, int
> noffset, uint8_t arch)
> >  	if (fit_image_get_arch(fit, noffset, &image_arch))
> >  		return 0;
> >  	return (arch == image_arch) ||
> > -		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64);
> > +		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64) ||
> > +		(arch == IH_ARCH_ARM64 && image_arch == IH_ARCH_ARM &&
> > +		 CONFIG_ARM64_SUPPORT_AARCH32);
> 
> Uh, I think this define is simply not defined for non-compatible systems
> (including non-arm64). This really needs to go into an #ifdef.
[Alison Wang] I think arch == IH_ARCH_ARM64 will check for non-arm64 system.
I can't understand your concern. Please explain more.
> 
> Thanks a lot for all the refactoring work :). This patch already reads
> much nicer than the original code. I guess with next round we'll get
> this into really good shape.
[Alison Wang] I hope so. Thank you very much for your suggestions. :)
> 
> 
> Alex
> 
> >  }
> >
> >  /**
> > @@ -1593,6 +1596,9 @@ int fit_image_load(bootm_headers_t *images,
> ulong addr,
> >  	int type_ok, os_ok;
> >  	ulong load, data, len;
> >  	uint8_t os;
> > +#ifndef USE_HOSTCC
> > +	uint8_t os_arch;
> > +#endif
> >  	const char *prop_name;
> >  	int ret;
> >
> > @@ -1676,6 +1682,12 @@ int fit_image_load(bootm_headers_t *images,
> ulong addr,
> >  		return -ENOEXEC;
> >  	}
> >  #endif
> > +
> > +#ifndef USE_HOSTCC
> > +	fit_image_get_arch(fit, noffset, &os_arch);
> > +	images->os.arch = os_arch;
> > +#endif
> > +
> >  	if (image_type == IH_TYPE_FLATDT &&
> >  	    !fit_image_check_comp(fit, noffset, IH_COMP_NONE)) {
> >  		puts("FDT image is compressed");
> >

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Alexander Graf]

On 03.06.16 05:11, Huan Wang wrote:
>> On 26.05.16 10:41, Alison Wang wrote:
>>> To support loading a 32-bit OS, the execution state will change from
>>> AArch64 to AArch32 when jumping to kernel.
>>>
>>> The architecture information will be got through checking FIT image,
>>> then U-Boot will load 32-bit OS or 64-bit OS automatically.
>>>
>>> Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
>>> Signed-off-by: Alison Wang <alison.wang@nxp.com>
>>> Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
>>> ---
>>> Changes in v3:
>>> - Comments the functions and the arguments.
>>> - Rename the real parameters.
>>> - Use the macros instead of the magic values.
>>> - Remove the redundant codes.
>>> - Clean up all of the mess in boot_jump_linux().
>>> - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system
>> doesn't support AArch32 state.
>>>
>>> Changes in v2:
>>> - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is
>> used
>>>   to switch to AArch64 EL2 or AArch32 Hyp.
>>> - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is
>> used
>>>   to switch to AArch64 EL1 or AArch32 SVC.
>>>
>>>  arch/arm/Kconfig                |   6 ++
>>>  arch/arm/cpu/armv8/transition.S |   8 +-
>>>  arch/arm/include/asm/macro.h    | 174 ++++++++++++++++++++++++++++++-
>> ---------
>>>  arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
>>>  arch/arm/lib/bootm.c            |  19 ++++-
>>>  common/image-fit.c              |  14 +++-
>>>  6 files changed, 287 insertions(+), 52 deletions(-)
>>>
>>> diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index
>>> 729b181..794cb4f 100644
>>> --- a/arch/arm/Kconfig
>>> +++ b/arch/arm/Kconfig
>>> @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
>>>  	  If SoC does not support L2CACHE or one do not want to enable
>>>  	  L2CACHE, choose this option.
>>>
>>> +config ARM64_SUPPORT_AARCH32
>>> +	bool "ARM64 system support AArch32 execution state"
>>> +	default y if ARM64 && !CONFIG_THUNDERX
>>
>> I don't think that works. In Kconfig you need to omit the CONFIG_ parts.
>> It's "ARM64" also, not "CONFIG_ARM64".
>>
>> Please just try to run a local make defconfig on the thunderx reference
>> system and see whether it includes the option or not.
> [Alison Wang] Yes, I missed it. Will change in the next version.
>>
>>> +	help
>>> +	  This ARM64 system supports AArch32 execution state.
>>> +
>>>  choice
>>>  	prompt "Target select"
>>>  	default TARGET_HIKEY
>>> diff --git a/arch/arm/cpu/armv8/transition.S
>>> b/arch/arm/cpu/armv8/transition.S index 253a39b..417e8b4 100644
>>> --- a/arch/arm/cpu/armv8/transition.S
>>> +++ b/arch/arm/cpu/armv8/transition.S
>>> @@ -11,13 +11,13 @@
>>>  #include <asm/macro.h>
>>>
>>>  ENTRY(armv8_switch_to_el2)
>>> -	switch_el x0, 1f, 0f, 0f
>>> +	switch_el x4, 1f, 0f, 0f
>>>  0:	ret
>>> -1:	armv8_switch_to_el2_m x0
>>> +1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
>>>  ENDPROC(armv8_switch_to_el2)
>>>
>>>  ENTRY(armv8_switch_to_el1)
>>> -	switch_el x0, 0f, 1f, 0f
>>> +	switch_el x4, 0f, 1f, 0f
>>>  0:	ret
>>> -1:	armv8_switch_to_el1_m x0, x1
>>> +1:	armv8_switch_to_el1_m x0, x1, x2, x3, x4
>>>  ENDPROC(armv8_switch_to_el1)
>>> diff --git a/arch/arm/include/asm/macro.h
>>> b/arch/arm/include/asm/macro.h index 9bb0efa..dd2c510 100644
>>> --- a/arch/arm/include/asm/macro.h
>>> +++ b/arch/arm/include/asm/macro.h
>>> @@ -8,6 +8,9 @@
>>>
>>>  #ifndef __ASM_ARM_MACRO_H__
>>>  #define __ASM_ARM_MACRO_H__
>>> +
>>> +#include <asm/system.h>
>>> +
>>>  #ifdef __ASSEMBLY__
>>>
>>>  /*
>>> @@ -135,13 +138,18 @@ lr	.req	x30
>>>  #endif
>>>  .endm
>>>
>>> -.macro armv8_switch_to_el2_m, xreg1
>>> -	/* 64bit EL2 | HCE | SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1
>> */
>>> -	mov	\xreg1, #0x5b1
>>> -	msr	scr_el3, \xreg1
>>> +/*
>>> + * Switch from EL3 to EL2 for ARMv8
>>> + * @ep:     kernel entry point
>>> + * @arch:   machine nr
>>> + * @ftaddr: fdt address
>>> + * @flag:   The execution state flag for lower exception
>>> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
>>> + */
>>> +.macro armv8_switch_to_el2_m, ep, arch, ftaddr, flag, xreg5
>>
>> You're never really using the arch and ftaddr arguments. Just describe
>> in the macro header that they need to be in x1 and x2 respectively and
>> leave them out of the argument list for the macro.
> [Alison Wang] Do you mean rename arch and ftaddr to x1 and x2 here?

I mean omit arch and ftaddr from the macro arguments and instead mention
that x1 and x2 get passed as x1 and x2 to the guest and are usually
"ftaddr" and "arch".

Also it would make the code more consistent to always use ep.


+/*
+ * Switch from EL3 to EL2 for ARMv8 and jump to @ep.
+ *
+ * @flag:   The execution state flag for lower exception
+ *          level, ES_TO_AARCH64 or ES_TO_AARCH32
+ *
+ * When flag == ES_TO_AARCH32:
+ * @ep:     location to jump to in AArch32 EL2 mode
+ * x1:      "Machine ID" for Linux payloads
+ * x2:      "DT Address" for Linux payloads
+ *
+ * When flag == ES_TO_AARCH64:
+ * @ep:     location to jump to in AArch64 EL2 mode
+ * x1:      "DT Address" for Linux payloads
+ */
+.macro armv8_switch_to_el2_m, ep, flag, tmpreg


>>
>> Also please rename "xreg5" to something like "scratchreg" or "tmp" or
>> "tmpreg". Then it's more obvious what this argument is about.
> [Alison Wang] Ok.
>>
>>>  	msr	cptr_el3, xzr		/* Disable coprocessor traps to EL3
>> */
>>> -	mov	\xreg1, #0x33ff
>>> -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
>>> +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
>>> +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
>>>
>>>  	/* Initialize Generic Timers */
>>>  	msr	cntvoff_el2, xzr
>>> @@ -152,45 +160,92 @@ lr	.req	x30
>>>  	 * and RES0 bits (31,30,27,26,24,21,20,17,15-13,10-6) +
>>>  	 * EE,WXN,I,SA,C,A,M to 0
>>>  	 */
>>> -	mov	\xreg1, #0x0830
>>> -	movk	\xreg1, #0x30C5, lsl #16
>>> -	msr	sctlr_el2, \xreg1
>>> +	ldr	\xreg5, =(SCTLR_EL2_BIT29_28_RES1 | SCTLR_EL2_EE_LE |\
>>> +			  SCTLR_EL2_BIT23_22_RES1 | SCTLR_EL2_WXN_DIS |\
>>> +			  SCTLR_EL2_BIT18_RES1 | SCTLR_EL2_BIT16_RES1 |\
>>> +			  SCTLR_EL2_BIT11_RES1 | SCTLR_EL2_BIT5_4_RES1 |\
>>> +			  SCTLR_EL2_SA_DIS | SCTLR_EL2_DCACHE_DIS |\
>>> +			  SCTLR_EL2_ALIGN_DIS | SCTLR_EL2_MMU_DIS)
>>
>> Do you think you could just define something like SCTLR_EL2_RES1 in the
>> header file that combines all RES1 fields already?
> [Alison Wang] Ok, I think so.
>>
>>> +	msr	sctlr_el2, \xreg5
>>> +
>>> +	mov	\xreg5, sp
>>> +	msr	sp_el2, \xreg5		/* Migrate SP */
>>> +	mrs	\xreg5, vbar_el3
>>> +	msr	vbar_el2, \xreg5	/* Migrate VBAR */
>>> +
>>> +	/* Check switch to AArch64 EL2 or AArch32 Hypervisor mode */
>>> +	ldr	\xreg5, =ES_TO_AARCH32
>>> +	cmp	\flag, \xreg5
>>
>> This can be an immediate cmp, no?
>>
>>   cmp \flag, #ES_TO_AARCH32
> [Alison Wang] Yes, it can. Will change in the next version.
>>
>>> +	b.eq	1f
>>> +
>>> +	/*
>>> +	 * The next lower exception level is AArch64, 64bit EL2 | HCE |
>>> +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
>>> +	 */
>>> +	ldr	\xreg5, =(SCR_EL3_RW_AARCH64 | SCR_EL3_HCE_EN |\
>>> +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
>>> +			  SCR_EL3_NS_EN)
>>
>> Same here - just define a single mask for SCR_EL3_RES1.
> [Alison Wang] Ok.
>>
>>> +	msr	scr_el3, \xreg5
>>>
>>>  	/* Return to the EL2_SP2 mode from EL3 */
>>> -	mov	\xreg1, sp
>>> -	msr	sp_el2, \xreg1		/* Migrate SP */
>>> -	mrs	\xreg1, vbar_el3
>>> -	msr	vbar_el2, \xreg1	/* Migrate VBAR */
>>> -	mov	\xreg1, #0x3c9
>>> -	msr	spsr_el3, \xreg1	/* EL2_SP2 | D | A | I | F */
>>> +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
>>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
>>> +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL2H)
>>> +	msr	spsr_el3, \xreg5
>>>  	msr	elr_el3, lr
>>>  	eret
>>> +
>>> +1:
>>> +	/*
>>> +	 * The next lower exception level is AArch32, 32bit EL2 | HCE |
>>> +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
>>> +	 */
>>> +	ldr	\xreg5, =(SCR_EL3_RW_AARCH32 | SCR_EL3_HCE_EN |\
>>> +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
>>> +			  SCR_EL3_NS_EN)
>>> +	msr	scr_el3, \xreg5
>>> +
>>> +	/* Return to AArch32 Hypervisor mode */
>>> +	ldr     \xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
>>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
>>> +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
>>> +			  SPSR_EL_M_HYP)
>>> +	msr	spsr_el3, \xreg5
>>> +	msr     elr_el3, \ep
>>> +
>>> +	mov	\ep, #0
>>> +	eret
>>>  .endm
>>>
>>> -.macro armv8_switch_to_el1_m, xreg1, xreg2
>>> +/*
>>> + * Switch from EL2 to EL1 for ARMv8
>>> + * @ep:     kernel entry point
>>> + * @arch:   machine nr
>>> + * @ftaddr: fdt address
>>> + * @flag:   The execution state flag for lower exception
>>> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
>>> + */
>>> +.macro armv8_switch_to_el1_m, ep, arch, ftaddr, flag, xreg5
>>
>> All the comments to the function above apply here too :).
> [Alison Wang] Ok.
>>
>>>  	/* Initialize Generic Timers */
>>> -	mrs	\xreg1, cnthctl_el2
>>> -	orr	\xreg1, \xreg1, #0x3	/* Enable EL1 access to timers */
>>> -	msr	cnthctl_el2, \xreg1
>>> +	mrs	\xreg5, cnthctl_el2
>>> +	/* Enable EL1 access to timers */
>>> +	orr	\xreg5, \xreg5, #(CNTHCTL_EL2_EL1PCEN_EN |\
>>> +		CNTHCTL_EL2_EL1PCTEN_EN)
>>> +	msr	cnthctl_el2, \xreg5
>>>  	msr	cntvoff_el2, xzr
>>>
>>>  	/* Initilize MPID/MPIDR registers */
>>> -	mrs	\xreg1, midr_el1
>>> -	mrs	\xreg2, mpidr_el1
>>> -	msr	vpidr_el2, \xreg1
>>> -	msr	vmpidr_el2, \xreg2
>>> +	mrs	\xreg5, midr_el1
>>> +	msr	vpidr_el2, \xreg5
>>> +	mrs	\xreg5, mpidr_el1
>>> +	msr	vmpidr_el2, \xreg5
>>>
>>>  	/* Disable coprocessor traps */
>>> -	mov	\xreg1, #0x33ff
>>> -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
>>> +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
>>> +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
>>>  	msr	hstr_el2, xzr		/* Disable coprocessor traps to EL2
>> */
>>> -	mov	\xreg1, #3 << 20
>>> -	msr	cpacr_el1, \xreg1	/* Enable FP/SIMD at EL1 */
>>> -
>>> -	/* Initialize HCR_EL2 */
>>> -	mov	\xreg1, #(1 << 31)		/* 64bit EL1 */
>>> -	orr	\xreg1, \xreg1, #(1 << 29)	/* Disable HVC */
>>> -	msr	hcr_el2, \xreg1
>>> +	mov	\xreg5, #(CPACR_EL1_FPEN_EN)
>>> +	msr	cpacr_el1, \xreg5	/* Enable FP/SIMD at EL1 */
>>>
>>>  	/* SCTLR_EL1 initialization
>>>  	 *
>>> @@ -199,19 +254,56 @@ lr	.req	x30
>>>  	 * UCI,EE,EOE,WXN,nTWE,nTWI,UCT,DZE,I,UMA,SED,ITD,
>>>  	 * CP15BEN,SA0,SA,C,A,M to 0
>>>  	 */
>>> -	mov	\xreg1, #0x0800
>>> -	movk	\xreg1, #0x30d0, lsl #16
>>> -	msr	sctlr_el1, \xreg1
>>> +	ldr	\xreg5, =(SCTLR_EL1_BIT29_28_RES1 | SCTLR_EL1_UCI_DIS |\
>>> +			  SCTLR_EL1_EE_LE | SCTLR_EL1_BIT23_22_RES1 |\
>>> +			  SCTLR_EL1_BIT20_RES1 | SCTLR_EL1_WXN_DIS |\
>>> +			  SCTLR_EL1_NTWE_DIS | SCTLR_EL1_NTWI_DIS |\
>>> +			  SCTLR_EL1_UCT_DIS | SCTLR_EL1_DZE_DIS |\
>>> +			  SCTLR_EL1_ICACHE_DIS | SCTLR_EL1_BIT11_RES1 |\
>>> +			  SCTLR_EL1_UMA_DIS | SCTLR_EL1_SED_EN |\
>>> +			  SCTLR_EL1_ITD_EN | SCTLR_EL1_CP15BEN_DIS |\
>>> +			  SCTLR_EL1_SA0_DIS | SCTLR_EL1_SA_DIS |\
>>> +			  SCTLR_EL1_DCACHE_DIS | SCTLR_EL1_ALIGN_DIS |\
>>> +			  SCTLR_EL1_MMU_DIS)
>>
>> Why are the caches getting disabled? Or did you just convert the old
>> value to the define list?
> [Alison Wang] Yes, I just convert the old value to the define list.
>>
>>> +	msr	sctlr_el1, \xreg5
>>> +
>>> +	mov	\xreg5, sp
>>> +	msr	sp_el1, \xreg5		/* Migrate SP */
>>> +	mrs	\xreg5, vbar_el2
>>> +	msr	vbar_el1, \xreg5	/* Migrate VBAR */
>>> +
>>> +	/* Check switch to AArch64 EL1 or AArch32 Supervisor mode */
>>> +	ldr	\xreg5, =ES_TO_AARCH32
>>> +	cmp	\flag, \xreg5
>>> +	b.eq	1f
>>> +
>>> +	/* Initialize HCR_EL2 */
>>> +	ldr	\xreg5, =(HCR_EL2_RW_AARCH64 | HCR_EL2_HCD_DIS)
>>> +	msr	hcr_el2, \xreg5
>>>
>>>  	/* Return to the EL1_SP1 mode from EL2 */
>>> -	mov	\xreg1, sp
>>> -	msr	sp_el1, \xreg1		/* Migrate SP */
>>> -	mrs	\xreg1, vbar_el2
>>> -	msr	vbar_el1, \xreg1	/* Migrate VBAR */
>>> -	mov	\xreg1, #0x3c5
>>> -	msr	spsr_el2, \xreg1	/* EL1_SP1 | D | A | I | F */
>>> +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
>>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
>>> +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL1H)
>>> +	msr	spsr_el2, \xreg5
>>>  	msr	elr_el2, lr
>>>  	eret
>>> +
>>> +1:
>>> +	/* Initialize HCR_EL2 */
>>> +	ldr	\xreg5, =(HCR_EL2_RW_AARCH32 | HCR_EL2_HCD_DIS)
>>> +	msr	hcr_el2, \xreg5
>>> +
>>> +	/* Return to AArch32 Supervisor mode from EL2 */
>>> +	ldr	\xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
>>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
>>> +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
>>> +			  SPSR_EL_M_SVC)
>>> +	msr     spsr_el2, \xreg5
>>> +	msr     elr_el2, \ep
>>> +
>>> +	mov	\ep, #0
>>> +	eret
>>>  .endm
>>>
>>>  #if defined(CONFIG_GICV3)
>>> diff --git a/arch/arm/include/asm/system.h
>>> b/arch/arm/include/asm/system.h index 9ae890a..0271ddb 100644
>>> --- a/arch/arm/include/asm/system.h
>>> +++ b/arch/arm/include/asm/system.h
>>> @@ -17,6 +17,102 @@
>>>  #define CR_WXN		(1 << 19)	/* Write Permision Imply XN	*/
>>>  #define CR_EE		(1 << 25)	/* Exception (Big) Endian	*/
>>>
>>> +#define ES_TO_AARCH64		1
>>> +#define ES_TO_AARCH32		0
>>> +
>>> +/*
>>> + * SCR_EL3 bits definitions
>>> + */
>>> +#define SCR_EL3_RW_AARCH64	(1 << 10) /* Next lower level is AArch64
>> */
>>> +#define SCR_EL3_RW_AARCH32	(0 << 10) /* Lower lowers level are
>> AArch32  */
>>> +#define SCR_EL3_HCE_EN		(1 << 8)  /* Hypervisor Call enable
>> */
>>> +#define SCR_EL3_SMD_DIS		(1 << 7)  /* Secure Monitor Call
>> disable     */
>>> +#define SCR_EL3_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1
>> */
>>> +#define SCR_EL3_NS_EN		(1 << 0)  /* EL0 and EL1 in Non-scure
>> state  */
>>> +
>>> +/*
>>> + * SPSR_EL3/SPSR_EL2 bits definitions  */
>>> +#define SPSR_EL_END_LE		(0 << 9)  /* Exception Little-endian
>> */
>>> +#define SPSR_EL_DEBUG_MASK	(1 << 9)  /* Debug exception masked
>> */
>>> +#define SPSR_EL_ASYN_MASK	(1 << 8)  /* Asynchronous data abort
>> masked   */
>>> +#define SPSR_EL_SERR_MASK	(1 << 8)  /* System Error exception masked
>> */
>>> +#define SPSR_EL_IRQ_MASK	(1 << 7)  /* IRQ exception masked
>> */
>>> +#define SPSR_EL_FIQ_MASK	(1 << 6)  /* FIQ exception masked
>> */
>>> +#define SPSR_EL_T_A32		(0 << 5)  /* AArch32 instruction set A32
>> */
>>> +#define SPSR_EL_M_AARCH64	(0 << 4)  /* Exception taken from AArch64
>> */
>>> +#define SPSR_EL_M_AARCH32	(1 << 4)  /* Exception taken from AArch32
>> */
>>> +#define SPSR_EL_M_SVC		(0x3)     /* Exception taken from SVC mode
>> */
>>> +#define SPSR_EL_M_HYP		(0xa)     /* Exception taken from HYP mode
>> */
>>> +#define SPSR_EL_M_EL1H		(5)       /* Exception taken from
>> EL1h mode   */
>>> +#define SPSR_EL_M_EL2H		(9)       /* Exception taken from
>> EL2h mode   */
>>> +
>>> +/*
>>> + * CPTR_EL2 bits definitions
>>> + */
>>> +#define CPTR_EL2_BIT13_12_RES1	(3 << 12) /* Reserved, RES1
>> */
>>> +#define CPTR_EL2_BIT9_0_RES1	(0x3ff)   /* Reserved, RES1
>> */
>>> +
>>> +/*
>>> + * SCTLR_EL2 bits definitions
>>> + */
>>> +#define SCTLR_EL2_BIT29_28_RES1	(3 << 28) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL2_EE_LE		(0 << 25) /* Exception Little-endian
>> */
>>> +#define SCTLR_EL2_BIT23_22_RES1	(3 << 22) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL2_WXN_DIS	(0 << 19) /* Write permission is not XN
>> */
>>> +#define SCTLR_EL2_BIT18_RES1	(1 << 18) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL2_BIT16_RES1	(1 << 16) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL2_ICACHE_DIS	(0 << 12) /* Instruction cache
>> disabled       */
>>> +#define SCTLR_EL2_BIT11_RES1	(1 << 11) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL2_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1
>> */
>>> +#define SCTLR_EL2_SA_DIS	(0 << 3)  /* Stack Alignment Check
>> disabled   */
>>> +#define SCTLR_EL2_DCACHE_DIS	(0 << 2)  /* Data cache disabled
>> */
>>> +#define SCTLR_EL2_ALIGN_DIS	(0 << 1)  /* Alignment check disabled
>> */
>>> +#define SCTLR_EL2_MMU_DIS	(0)       /* MMU disabled
>> */
>>> +
>>> +/*
>>> + * CNTHCTL_EL2 bits definitions
>>> + */
>>> +#define CNTHCTL_EL2_EL1PCEN_EN	(1 << 1)  /* Physical timer regs
>> accessible   */
>>> +#define CNTHCTL_EL2_EL1PCTEN_EN	(1 << 0)  /* Physical counter
>> accessible      */
>>> +
>>> +/*
>>> + * HCR_EL2 bits definitions
>>> + */
>>> +#define HCR_EL2_RW_AARCH64	(1 << 31) /* EL1 is AArch64
>> */
>>> +#define HCR_EL2_RW_AARCH32	(0 << 31) /* Lower levels are AArch32
>> */
>>> +#define HCR_EL2_HCD_DIS		(1 << 29) /* Hypervisor Call
>> disabled         */
>>> +
>>> +/*
>>> + * CPACR_EL1 bits definitions
>>> + */
>>> +#define CPACR_EL1_FPEN_EN	(3 << 20) /* SIMD and FP instruction
>> enabled  */
>>> +
>>> +/*
>>> + * SCTLR_EL1 bits definitions
>>> + */
>>> +#define SCTLR_EL1_BIT29_28_RES1	(3 << 28) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL1_UCI_DIS	(0 << 26) /* Cache instruction disabled
>> */
>>> +#define SCTLR_EL1_EE_LE		(0 << 25) /* Exception Little-endian
>> */
>>> +#define SCTLR_EL1_BIT23_22_RES1	(3 << 22) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL1_BIT20_RES1	(1 << 20) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL1_WXN_DIS	(0 << 19) /* Write permission is not XN
>> */
>>> +#define SCTLR_EL1_NTWE_DIS	(0 << 18) /* WFE instruction disabled
>> */
>>> +#define SCTLR_EL1_NTWI_DIS	(0 << 16) /* WFI instruction disabled
>> */
>>> +#define SCTLR_EL1_UCT_DIS	(0 << 15) /* CTR_EL0 access disabled
>> */
>>> +#define SCTLR_EL1_DZE_DIS	(0 << 14) /* DC ZVA instruction disabled
>> */
>>> +#define SCTLR_EL1_ICACHE_DIS	(0 << 12) /* Instruction cache
>> disabled       */
>>> +#define SCTLR_EL1_BIT11_RES1	(1 << 11) /* Reserved, RES1
>> */
>>> +#define SCTLR_EL1_UMA_DIS	(0 << 9)  /* User Mask Access disabled
>> */
>>> +#define SCTLR_EL1_SED_EN	(0 << 8)  /* SETEND instruction enabled
>> */
>>> +#define SCTLR_EL1_ITD_EN	(0 << 7)  /* IT instruction enabled
>> */
>>> +#define SCTLR_EL1_CP15BEN_DIS	(0 << 5)  /* CP15 barrier operation
>> disabled  */
>>> +#define SCTLR_EL1_SA0_DIS	(0 << 4)  /* Stack Alignment EL0 disabled
>> */
>>> +#define SCTLR_EL1_SA_DIS	(0 << 3)  /* Stack Alignment EL1 disabled
>> */
>>> +#define SCTLR_EL1_DCACHE_DIS	(0 << 2)  /* Data cache disabled
>> */
>>> +#define SCTLR_EL1_ALIGN_DIS	(0 << 1)  /* Alignment check disabled
>> */
>>> +#define SCTLR_EL1_MMU_DIS	(0)       /* MMU disabled
>> */
>>> +
>>>  #ifndef __ASSEMBLY__
>>>
>>>  u64 get_page_table_size(void);
>>> @@ -100,8 +196,26 @@ void __asm_invalidate_icache_all(void);
>>>  int __asm_flush_l3_cache(void);
>>>  void __asm_switch_ttbr(u64 new_ttbr);
>>>
>>> -void armv8_switch_to_el2(void);
>>> -void armv8_switch_to_el1(void);
>>> +/*
>>> + * Switch from EL3 to EL2 for ARMv8
>>> + *
>>> + * @entry_point: kernel entry point
>>> + * @mach_nr:     machine nr
>>> + * @fdt_addr:    fdt address
>>> + * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
>>> + */
>>> +void armv8_switch_to_el2(u64 entry_point, u64 mach_nr, u64 fdt_addr,
>>> +			 u64 es_flag);
>>> +/*
>>> + * Switch from EL2 to EL1 for ARMv8
>>> + *
>>> + * @entry_point: kernel entry point
>>> + * @mach_nr:     machine nr
>>> + * @fdt_addr:    fdt address
>>> + * @es_flag:     execution state flag, ES_TO_AARCH64 or ES_TO_AARCH32
>>> + */
>>> +void armv8_switch_to_el1(u64 entry_point, u64 mach_nr, u64 fdt_addr,
>>> +			 u64 es_flag);
>>>  void gic_init(void);
>>>  void gic_send_sgi(unsigned long sgino);  void wait_for_wakeup(void);
>>> diff --git a/arch/arm/lib/bootm.c b/arch/arm/lib/bootm.c index
>>> 0838d89..1bbf85c 100644
>>> --- a/arch/arm/lib/bootm.c
>>> +++ b/arch/arm/lib/bootm.c
>>> @@ -189,13 +189,19 @@ static void setup_end_tag(bd_t *bd)  __weak void
>>> setup_board_tags(struct tag **in_params) {}
>>>
>>>  #ifdef CONFIG_ARM64
>>> -static void do_nonsec_virt_switch(void)
>>> +static void do_nonsec_virt_switch(bootm_headers_t *images, int flag)
>>>  {
>>>  	smp_kick_all_cpus();
>>>  	dcache_disable();	/* flush cache before swtiching to EL2 */
>>> -	armv8_switch_to_el2();
>>> +
>>>  #ifdef CONFIG_ARMV8_SWITCH_TO_EL1
>>> -	armv8_switch_to_el1();
>>> +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
>>> +			    (u64)images->ft_addr, ES_TO_AARCH64);
>>> +	armv8_switch_to_el1((u64)images->ep, (u64)gd->bd->bi_arch_number,
>>> +			    (u64)images->ft_addr, flag);
>>> +#else
>>> +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
>>> +			    (u64)images->ft_addr, flag);
>>>  #endif
>>>  }
>>>  #endif
>>> @@ -275,6 +281,7 @@ static void boot_jump_linux(bootm_headers_t
>> *images, int flag)
>>>  	void (*kernel_entry)(void *fdt_addr, void *res0, void *res1,
>>>  			void *res2);
>>>  	int fake = (flag & BOOTM_STATE_OS_FAKE_GO);
>>> +	int es_flag = ES_TO_AARCH64;
>>>
>>>  	kernel_entry = (void (*)(void *fdt_addr, void *res0, void *res1,
>>>  				void *res2))images->ep;
>>> @@ -286,7 +293,11 @@ static void boot_jump_linux(bootm_headers_t
>> *images, int flag)
>>>  	announce_and_cleanup(fake);
>>>
>>>  	if (!fake) {
>>> -		do_nonsec_virt_switch();
>>> +		if ((IH_ARCH_DEFAULT == IH_ARCH_ARM64) &&
>>> +		    (images->os.arch == IH_ARCH_ARM))
>>> +			es_flag = ES_TO_AARCH32;
>>> +
>>> +		do_nonsec_virt_switch(images, es_flag);
>>>  		kernel_entry(images->ft_addr, NULL, NULL, NULL);
>>>  	}
>>>  #else
>>> diff --git a/common/image-fit.c b/common/image-fit.c index
>>> 9873957..982021b 100644
>>> --- a/common/image-fit.c
>>> +++ b/common/image-fit.c
>>> @@ -26,6 +26,7 @@ DECLARE_GLOBAL_DATA_PTR;  #include <u-boot/md5.h>
>>> #include <u-boot/sha1.h>  #include <u-boot/sha256.h>
>>> +#include <generated/autoconf.h>
>>>
>>>
>>> /*********************************************************************
>>> ********/
>>>  /* New uImage format routines */
>>> @@ -1164,7 +1165,9 @@ int fit_image_check_arch(const void *fit, int
>> noffset, uint8_t arch)
>>>  	if (fit_image_get_arch(fit, noffset, &image_arch))
>>>  		return 0;
>>>  	return (arch == image_arch) ||
>>> -		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64);
>>> +		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64) ||
>>> +		(arch == IH_ARCH_ARM64 && image_arch == IH_ARCH_ARM &&
>>> +		 CONFIG_ARM64_SUPPORT_AARCH32);
>>
>> Uh, I think this define is simply not defined for non-compatible systems
>> (including non-arm64). This really needs to go into an #ifdef.
> [Alison Wang] I think arch == IH_ARCH_ARM64 will check for non-arm64 system.
> I can't understand your concern. Please explain more.

You either get a

#define CONFIG_ARM64_SUPPORT_AARCH32 1

in the config header if the option is set or simply nothing at all. If
there's nothing, compiling this code will result in an unresolved symbol.


Alex

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Huan Wang]

> On 03.06.16 05:11, Huan Wang wrote:
> >> On 26.05.16 10:41, Alison Wang wrote:
> >>> To support loading a 32-bit OS, the execution state will change from
> >>> AArch64 to AArch32 when jumping to kernel.
> >>>
> >>> The architecture information will be got through checking FIT image,
> >>> then U-Boot will load 32-bit OS or 64-bit OS automatically.
> >>>
> >>> Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
> >>> Signed-off-by: Alison Wang <alison.wang@nxp.com>
> >>> Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
> >>> ---
> >>> Changes in v3:
> >>> - Comments the functions and the arguments.
> >>> - Rename the real parameters.
> >>> - Use the macros instead of the magic values.
> >>> - Remove the redundant codes.
> >>> - Clean up all of the mess in boot_jump_linux().
> >>> - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system
> >> doesn't support AArch32 state.
> >>>
> >>> Changes in v2:
> >>> - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is
> >> used
> >>>   to switch to AArch64 EL2 or AArch32 Hyp.
> >>> - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is
> >> used
> >>>   to switch to AArch64 EL1 or AArch32 SVC.
> >>>
> >>>  arch/arm/Kconfig                |   6 ++
> >>>  arch/arm/cpu/armv8/transition.S |   8 +-
> >>>  arch/arm/include/asm/macro.h    | 174
> ++++++++++++++++++++++++++++++-
> >> ---------
> >>>  arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
> >>>  arch/arm/lib/bootm.c            |  19 ++++-
> >>>  common/image-fit.c              |  14 +++-
> >>>  6 files changed, 287 insertions(+), 52 deletions(-)
> >>>
> >>> diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index
> >>> 729b181..794cb4f 100644
> >>> --- a/arch/arm/Kconfig
> >>> +++ b/arch/arm/Kconfig
> >>> @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
> >>>  	  If SoC does not support L2CACHE or one do not want to enable
> >>>  	  L2CACHE, choose this option.
> >>>
> >>> +config ARM64_SUPPORT_AARCH32
> >>> +	bool "ARM64 system support AArch32 execution state"
> >>> +	default y if ARM64 && !CONFIG_THUNDERX
> >>
> >> I don't think that works. In Kconfig you need to omit the CONFIG_
> parts.
> >> It's "ARM64" also, not "CONFIG_ARM64".
> >>
> >> Please just try to run a local make defconfig on the thunderx
> >> reference system and see whether it includes the option or not.
> > [Alison Wang] Yes, I missed it. Will change in the next version.
> >>
> >>> +	help
> >>> +	  This ARM64 system supports AArch32 execution state.
> >>> +
> >>>  choice
> >>>  	prompt "Target select"
> >>>  	default TARGET_HIKEY
> >>> diff --git a/arch/arm/cpu/armv8/transition.S
> >>> b/arch/arm/cpu/armv8/transition.S index 253a39b..417e8b4 100644
> >>> --- a/arch/arm/cpu/armv8/transition.S
> >>> +++ b/arch/arm/cpu/armv8/transition.S
> >>> @@ -11,13 +11,13 @@
> >>>  #include <asm/macro.h>
> >>>
> >>>  ENTRY(armv8_switch_to_el2)
> >>> -	switch_el x0, 1f, 0f, 0f
> >>> +	switch_el x4, 1f, 0f, 0f
> >>>  0:	ret
> >>> -1:	armv8_switch_to_el2_m x0
> >>> +1:	armv8_switch_to_el2_m x0, x1, x2, x3, x4
> >>>  ENDPROC(armv8_switch_to_el2)
> >>>
> >>>  ENTRY(armv8_switch_to_el1)
> >>> -	switch_el x0, 0f, 1f, 0f
> >>> +	switch_el x4, 0f, 1f, 0f
> >>>  0:	ret
> >>> -1:	armv8_switch_to_el1_m x0, x1
> >>> +1:	armv8_switch_to_el1_m x0, x1, x2, x3, x4
> >>>  ENDPROC(armv8_switch_to_el1)
> >>> diff --git a/arch/arm/include/asm/macro.h
> >>> b/arch/arm/include/asm/macro.h index 9bb0efa..dd2c510 100644
> >>> --- a/arch/arm/include/asm/macro.h
> >>> +++ b/arch/arm/include/asm/macro.h
> >>> @@ -8,6 +8,9 @@
> >>>
> >>>  #ifndef __ASM_ARM_MACRO_H__
> >>>  #define __ASM_ARM_MACRO_H__
> >>> +
> >>> +#include <asm/system.h>
> >>> +
> >>>  #ifdef __ASSEMBLY__
> >>>
> >>>  /*
> >>> @@ -135,13 +138,18 @@ lr	.req	x30
> >>>  #endif
> >>>  .endm
> >>>
> >>> -.macro armv8_switch_to_el2_m, xreg1
> >>> -	/* 64bit EL2 | HCE | SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1
> >> */
> >>> -	mov	\xreg1, #0x5b1
> >>> -	msr	scr_el3, \xreg1
> >>> +/*
> >>> + * Switch from EL3 to EL2 for ARMv8
> >>> + * @ep:     kernel entry point
> >>> + * @arch:   machine nr
> >>> + * @ftaddr: fdt address
> >>> + * @flag:   The execution state flag for lower exception
> >>> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> >>> + */
> >>> +.macro armv8_switch_to_el2_m, ep, arch, ftaddr, flag, xreg5
> >>
> >> You're never really using the arch and ftaddr arguments. Just
> >> describe in the macro header that they need to be in x1 and x2
> >> respectively and leave them out of the argument list for the macro.
> > [Alison Wang] Do you mean rename arch and ftaddr to x1 and x2 here?
> 
> I mean omit arch and ftaddr from the macro arguments and instead mention
> that x1 and x2 get passed as x1 and x2 to the guest and are usually
> "ftaddr" and "arch".
> 
> Also it would make the code more consistent to always use ep.
[Alison Wang] I don't think arch and ftaddr can be removed from the
macro arguments in the current code. For 32-bit kernel startup entry
point, the requirements are: r0 = 0, r1 = machine nr, r2 = atags or
dtb pointer.
> 
> 
> +/*
> + * Switch from EL3 to EL2 for ARMv8 and jump to @ep.
> + *
> + * @flag:   The execution state flag for lower exception
> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> + *
> + * When flag == ES_TO_AARCH32:
> + * @ep:     location to jump to in AArch32 EL2 mode
> + * x1:      "Machine ID" for Linux payloads
> + * x2:      "DT Address" for Linux payloads
> + *
> + * When flag == ES_TO_AARCH64:
> + * @ep:     location to jump to in AArch64 EL2 mode
> + * x1:      "DT Address" for Linux payloads
> + */
> +.macro armv8_switch_to_el2_m, ep, flag, tmpreg
> 
> 
> >>
> >> Also please rename "xreg5" to something like "scratchreg" or "tmp" or
> >> "tmpreg". Then it's more obvious what this argument is about.
> > [Alison Wang] Ok.
> >>
> >>>  	msr	cptr_el3, xzr		/* Disable coprocessor traps to EL3
> >> */
> >>> -	mov	\xreg1, #0x33ff
> >>> -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
> >>> +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
> >>> +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
> >>>
> >>>  	/* Initialize Generic Timers */
> >>>  	msr	cntvoff_el2, xzr
> >>> @@ -152,45 +160,92 @@ lr	.req	x30
> >>>  	 * and RES0 bits (31,30,27,26,24,21,20,17,15-13,10-6) +
> >>>  	 * EE,WXN,I,SA,C,A,M to 0
> >>>  	 */
> >>> -	mov	\xreg1, #0x0830
> >>> -	movk	\xreg1, #0x30C5, lsl #16
> >>> -	msr	sctlr_el2, \xreg1
> >>> +	ldr	\xreg5, =(SCTLR_EL2_BIT29_28_RES1 | SCTLR_EL2_EE_LE |\
> >>> +			  SCTLR_EL2_BIT23_22_RES1 | SCTLR_EL2_WXN_DIS |\
> >>> +			  SCTLR_EL2_BIT18_RES1 | SCTLR_EL2_BIT16_RES1 |\
> >>> +			  SCTLR_EL2_BIT11_RES1 | SCTLR_EL2_BIT5_4_RES1 |\
> >>> +			  SCTLR_EL2_SA_DIS | SCTLR_EL2_DCACHE_DIS |\
> >>> +			  SCTLR_EL2_ALIGN_DIS | SCTLR_EL2_MMU_DIS)
> >>
> >> Do you think you could just define something like SCTLR_EL2_RES1 in
> >> the header file that combines all RES1 fields already?
> > [Alison Wang] Ok, I think so.
> >>
> >>> +	msr	sctlr_el2, \xreg5
> >>> +
> >>> +	mov	\xreg5, sp
> >>> +	msr	sp_el2, \xreg5		/* Migrate SP */
> >>> +	mrs	\xreg5, vbar_el3
> >>> +	msr	vbar_el2, \xreg5	/* Migrate VBAR */
> >>> +
> >>> +	/* Check switch to AArch64 EL2 or AArch32 Hypervisor mode */
> >>> +	ldr	\xreg5, =ES_TO_AARCH32
> >>> +	cmp	\flag, \xreg5
> >>
> >> This can be an immediate cmp, no?
> >>
> >>   cmp \flag, #ES_TO_AARCH32
> > [Alison Wang] Yes, it can. Will change in the next version.
> >>
> >>> +	b.eq	1f
> >>> +
> >>> +	/*
> >>> +	 * The next lower exception level is AArch64, 64bit EL2 | HCE |
> >>> +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
> >>> +	 */
> >>> +	ldr	\xreg5, =(SCR_EL3_RW_AARCH64 | SCR_EL3_HCE_EN |\
> >>> +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
> >>> +			  SCR_EL3_NS_EN)
> >>
> >> Same here - just define a single mask for SCR_EL3_RES1.
> > [Alison Wang] Ok.
> >>
> >>> +	msr	scr_el3, \xreg5
> >>>
> >>>  	/* Return to the EL2_SP2 mode from EL3 */
> >>> -	mov	\xreg1, sp
> >>> -	msr	sp_el2, \xreg1		/* Migrate SP */
> >>> -	mrs	\xreg1, vbar_el3
> >>> -	msr	vbar_el2, \xreg1	/* Migrate VBAR */
> >>> -	mov	\xreg1, #0x3c9
> >>> -	msr	spsr_el3, \xreg1	/* EL2_SP2 | D | A | I | F */
> >>> +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
> >>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> >>> +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL2H)
> >>> +	msr	spsr_el3, \xreg5
> >>>  	msr	elr_el3, lr
> >>>  	eret
> >>> +
> >>> +1:
> >>> +	/*
> >>> +	 * The next lower exception level is AArch32, 32bit EL2 | HCE |
> >>> +	 * SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1.
> >>> +	 */
> >>> +	ldr	\xreg5, =(SCR_EL3_RW_AARCH32 | SCR_EL3_HCE_EN |\
> >>> +			  SCR_EL3_SMD_DIS | SCR_EL3_BIT5_4_RES1 |\
> >>> +			  SCR_EL3_NS_EN)
> >>> +	msr	scr_el3, \xreg5
> >>> +
> >>> +	/* Return to AArch32 Hypervisor mode */
> >>> +	ldr     \xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
> >>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> >>> +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
> >>> +			  SPSR_EL_M_HYP)
> >>> +	msr	spsr_el3, \xreg5
> >>> +	msr     elr_el3, \ep
> >>> +
> >>> +	mov	\ep, #0
> >>> +	eret
> >>>  .endm
> >>>
> >>> -.macro armv8_switch_to_el1_m, xreg1, xreg2
> >>> +/*
> >>> + * Switch from EL2 to EL1 for ARMv8
> >>> + * @ep:     kernel entry point
> >>> + * @arch:   machine nr
> >>> + * @ftaddr: fdt address
> >>> + * @flag:   The execution state flag for lower exception
> >>> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
> >>> + */
> >>> +.macro armv8_switch_to_el1_m, ep, arch, ftaddr, flag, xreg5
> >>
> >> All the comments to the function above apply here too :).
> > [Alison Wang] Ok.
> >>
> >>>  	/* Initialize Generic Timers */
> >>> -	mrs	\xreg1, cnthctl_el2
> >>> -	orr	\xreg1, \xreg1, #0x3	/* Enable EL1 access to timers */
> >>> -	msr	cnthctl_el2, \xreg1
> >>> +	mrs	\xreg5, cnthctl_el2
> >>> +	/* Enable EL1 access to timers */
> >>> +	orr	\xreg5, \xreg5, #(CNTHCTL_EL2_EL1PCEN_EN |\
> >>> +		CNTHCTL_EL2_EL1PCTEN_EN)
> >>> +	msr	cnthctl_el2, \xreg5
> >>>  	msr	cntvoff_el2, xzr
> >>>
> >>>  	/* Initilize MPID/MPIDR registers */
> >>> -	mrs	\xreg1, midr_el1
> >>> -	mrs	\xreg2, mpidr_el1
> >>> -	msr	vpidr_el2, \xreg1
> >>> -	msr	vmpidr_el2, \xreg2
> >>> +	mrs	\xreg5, midr_el1
> >>> +	msr	vpidr_el2, \xreg5
> >>> +	mrs	\xreg5, mpidr_el1
> >>> +	msr	vmpidr_el2, \xreg5
> >>>
> >>>  	/* Disable coprocessor traps */
> >>> -	mov	\xreg1, #0x33ff
> >>> -	msr	cptr_el2, \xreg1	/* Disable coprocessor traps to EL2 */
> >>> +	mov	\xreg5, #(CPTR_EL2_BIT13_12_RES1 | CPTR_EL2_BIT9_0_RES1)
> >>> +	msr	cptr_el2, \xreg5	/* Disable coprocessor traps to EL2 */
> >>>  	msr	hstr_el2, xzr		/* Disable coprocessor traps to EL2
> >> */
> >>> -	mov	\xreg1, #3 << 20
> >>> -	msr	cpacr_el1, \xreg1	/* Enable FP/SIMD at EL1 */
> >>> -
> >>> -	/* Initialize HCR_EL2 */
> >>> -	mov	\xreg1, #(1 << 31)		/* 64bit EL1 */
> >>> -	orr	\xreg1, \xreg1, #(1 << 29)	/* Disable HVC */
> >>> -	msr	hcr_el2, \xreg1
> >>> +	mov	\xreg5, #(CPACR_EL1_FPEN_EN)
> >>> +	msr	cpacr_el1, \xreg5	/* Enable FP/SIMD at EL1 */
> >>>
> >>>  	/* SCTLR_EL1 initialization
> >>>  	 *
> >>> @@ -199,19 +254,56 @@ lr	.req	x30
> >>>  	 * UCI,EE,EOE,WXN,nTWE,nTWI,UCT,DZE,I,UMA,SED,ITD,
> >>>  	 * CP15BEN,SA0,SA,C,A,M to 0
> >>>  	 */
> >>> -	mov	\xreg1, #0x0800
> >>> -	movk	\xreg1, #0x30d0, lsl #16
> >>> -	msr	sctlr_el1, \xreg1
> >>> +	ldr	\xreg5, =(SCTLR_EL1_BIT29_28_RES1 | SCTLR_EL1_UCI_DIS |\
> >>> +			  SCTLR_EL1_EE_LE | SCTLR_EL1_BIT23_22_RES1 |\
> >>> +			  SCTLR_EL1_BIT20_RES1 | SCTLR_EL1_WXN_DIS |\
> >>> +			  SCTLR_EL1_NTWE_DIS | SCTLR_EL1_NTWI_DIS |\
> >>> +			  SCTLR_EL1_UCT_DIS | SCTLR_EL1_DZE_DIS |\
> >>> +			  SCTLR_EL1_ICACHE_DIS | SCTLR_EL1_BIT11_RES1 |\
> >>> +			  SCTLR_EL1_UMA_DIS | SCTLR_EL1_SED_EN |\
> >>> +			  SCTLR_EL1_ITD_EN | SCTLR_EL1_CP15BEN_DIS |\
> >>> +			  SCTLR_EL1_SA0_DIS | SCTLR_EL1_SA_DIS |\
> >>> +			  SCTLR_EL1_DCACHE_DIS | SCTLR_EL1_ALIGN_DIS |\
> >>> +			  SCTLR_EL1_MMU_DIS)
> >>
> >> Why are the caches getting disabled? Or did you just convert the old
> >> value to the define list?
> > [Alison Wang] Yes, I just convert the old value to the define list.
> >>
> >>> +	msr	sctlr_el1, \xreg5
> >>> +
> >>> +	mov	\xreg5, sp
> >>> +	msr	sp_el1, \xreg5		/* Migrate SP */
> >>> +	mrs	\xreg5, vbar_el2
> >>> +	msr	vbar_el1, \xreg5	/* Migrate VBAR */
> >>> +
> >>> +	/* Check switch to AArch64 EL1 or AArch32 Supervisor mode */
> >>> +	ldr	\xreg5, =ES_TO_AARCH32
> >>> +	cmp	\flag, \xreg5
> >>> +	b.eq	1f
> >>> +
> >>> +	/* Initialize HCR_EL2 */
> >>> +	ldr	\xreg5, =(HCR_EL2_RW_AARCH64 | HCR_EL2_HCD_DIS)
> >>> +	msr	hcr_el2, \xreg5
> >>>
> >>>  	/* Return to the EL1_SP1 mode from EL2 */
> >>> -	mov	\xreg1, sp
> >>> -	msr	sp_el1, \xreg1		/* Migrate SP */
> >>> -	mrs	\xreg1, vbar_el2
> >>> -	msr	vbar_el1, \xreg1	/* Migrate VBAR */
> >>> -	mov	\xreg1, #0x3c5
> >>> -	msr	spsr_el2, \xreg1	/* EL1_SP1 | D | A | I | F */
> >>> +	ldr	\xreg5, =(SPSR_EL_DEBUG_MASK | SPSR_EL_SERR_MASK |\
> >>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> >>> +			  SPSR_EL_M_AARCH64 | SPSR_EL_M_EL1H)
> >>> +	msr	spsr_el2, \xreg5
> >>>  	msr	elr_el2, lr
> >>>  	eret
> >>> +
> >>> +1:
> >>> +	/* Initialize HCR_EL2 */
> >>> +	ldr	\xreg5, =(HCR_EL2_RW_AARCH32 | HCR_EL2_HCD_DIS)
> >>> +	msr	hcr_el2, \xreg5
> >>> +
> >>> +	/* Return to AArch32 Supervisor mode from EL2 */
> >>> +	ldr	\xreg5, =(SPSR_EL_END_LE | SPSR_EL_ASYN_MASK |\
> >>> +			  SPSR_EL_IRQ_MASK | SPSR_EL_FIQ_MASK |\
> >>> +			  SPSR_EL_T_A32 | SPSR_EL_M_AARCH32 |\
> >>> +			  SPSR_EL_M_SVC)
> >>> +	msr     spsr_el2, \xreg5
> >>> +	msr     elr_el2, \ep
> >>> +
> >>> +	mov	\ep, #0
> >>> +	eret
> >>>  .endm
> >>>
> >>>  #if defined(CONFIG_GICV3)
> >>> diff --git a/arch/arm/include/asm/system.h
> >>> b/arch/arm/include/asm/system.h index 9ae890a..0271ddb 100644
> >>> --- a/arch/arm/include/asm/system.h
> >>> +++ b/arch/arm/include/asm/system.h
> >>> @@ -17,6 +17,102 @@
> >>>  #define CR_WXN		(1 << 19)	/* Write Permision Imply XN
> 	*/
> >>>  #define CR_EE		(1 << 25)	/* Exception (Big) Endian	*/
> >>>
> >>> +#define ES_TO_AARCH64		1
> >>> +#define ES_TO_AARCH32		0
> >>> +
> >>> +/*
> >>> + * SCR_EL3 bits definitions
> >>> + */
> >>> +#define SCR_EL3_RW_AARCH64	(1 << 10) /* Next lower level is
> AArch64
> >> */
> >>> +#define SCR_EL3_RW_AARCH32	(0 << 10) /* Lower lowers level are
> >> AArch32  */
> >>> +#define SCR_EL3_HCE_EN		(1 << 8)  /* Hypervisor Call enable
> >> */
> >>> +#define SCR_EL3_SMD_DIS		(1 << 7)  /* Secure Monitor Call
> >> disable     */
> >>> +#define SCR_EL3_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1
> >> */
> >>> +#define SCR_EL3_NS_EN		(1 << 0)  /* EL0 and EL1 in Non-
> scure
> >> state  */
> >>> +
> >>> +/*
> >>> + * SPSR_EL3/SPSR_EL2 bits definitions  */
> >>> +#define SPSR_EL_END_LE		(0 << 9)  /* Exception Little-endian
> >> */
> >>> +#define SPSR_EL_DEBUG_MASK	(1 << 9)  /* Debug exception masked
> >> */
> >>> +#define SPSR_EL_ASYN_MASK	(1 << 8)  /* Asynchronous data abort
> >> masked   */
> >>> +#define SPSR_EL_SERR_MASK	(1 << 8)  /* System Error exception masked
> >> */
> >>> +#define SPSR_EL_IRQ_MASK	(1 << 7)  /* IRQ exception masked
> >> */
> >>> +#define SPSR_EL_FIQ_MASK	(1 << 6)  /* FIQ exception masked
> >> */
> >>> +#define SPSR_EL_T_A32		(0 << 5)  /* AArch32 instruction set
> A32
> >> */
> >>> +#define SPSR_EL_M_AARCH64	(0 << 4)  /* Exception taken from AArch64
> >> */
> >>> +#define SPSR_EL_M_AARCH32	(1 << 4)  /* Exception taken from AArch32
> >> */
> >>> +#define SPSR_EL_M_SVC		(0x3)     /* Exception taken from
> SVC mode
> >> */
> >>> +#define SPSR_EL_M_HYP		(0xa)     /* Exception taken from
> HYP mode
> >> */
> >>> +#define SPSR_EL_M_EL1H		(5)       /* Exception taken from
> >> EL1h mode   */
> >>> +#define SPSR_EL_M_EL2H		(9)       /* Exception taken from
> >> EL2h mode   */
> >>> +
> >>> +/*
> >>> + * CPTR_EL2 bits definitions
> >>> + */
> >>> +#define CPTR_EL2_BIT13_12_RES1	(3 << 12) /* Reserved, RES1
> >> */
> >>> +#define CPTR_EL2_BIT9_0_RES1	(0x3ff)   /* Reserved, RES1
> >> */
> >>> +
> >>> +/*
> >>> + * SCTLR_EL2 bits definitions
> >>> + */
> >>> +#define SCTLR_EL2_BIT29_28_RES1	(3 << 28) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL2_EE_LE		(0 << 25) /* Exception Little-endian
> >> */
> >>> +#define SCTLR_EL2_BIT23_22_RES1	(3 << 22) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL2_WXN_DIS	(0 << 19) /* Write permission is not XN
> >> */
> >>> +#define SCTLR_EL2_BIT18_RES1	(1 << 18) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL2_BIT16_RES1	(1 << 16) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL2_ICACHE_DIS	(0 << 12) /* Instruction cache
> >> disabled       */
> >>> +#define SCTLR_EL2_BIT11_RES1	(1 << 11) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL2_BIT5_4_RES1	(3 << 4)  /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL2_SA_DIS	(0 << 3)  /* Stack Alignment Check
> >> disabled   */
> >>> +#define SCTLR_EL2_DCACHE_DIS	(0 << 2)  /* Data cache disabled
> >> */
> >>> +#define SCTLR_EL2_ALIGN_DIS	(0 << 1)  /* Alignment check
> disabled
> >> */
> >>> +#define SCTLR_EL2_MMU_DIS	(0)       /* MMU disabled
> >> */
> >>> +
> >>> +/*
> >>> + * CNTHCTL_EL2 bits definitions
> >>> + */
> >>> +#define CNTHCTL_EL2_EL1PCEN_EN	(1 << 1)  /* Physical timer regs
> >> accessible   */
> >>> +#define CNTHCTL_EL2_EL1PCTEN_EN	(1 << 0)  /* Physical counter
> >> accessible      */
> >>> +
> >>> +/*
> >>> + * HCR_EL2 bits definitions
> >>> + */
> >>> +#define HCR_EL2_RW_AARCH64	(1 << 31) /* EL1 is AArch64
> >> */
> >>> +#define HCR_EL2_RW_AARCH32	(0 << 31) /* Lower levels are
> AArch32
> >> */
> >>> +#define HCR_EL2_HCD_DIS		(1 << 29) /* Hypervisor Call
> >> disabled         */
> >>> +
> >>> +/*
> >>> + * CPACR_EL1 bits definitions
> >>> + */
> >>> +#define CPACR_EL1_FPEN_EN	(3 << 20) /* SIMD and FP instruction
> >> enabled  */
> >>> +
> >>> +/*
> >>> + * SCTLR_EL1 bits definitions
> >>> + */
> >>> +#define SCTLR_EL1_BIT29_28_RES1	(3 << 28) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL1_UCI_DIS	(0 << 26) /* Cache instruction disabled
> >> */
> >>> +#define SCTLR_EL1_EE_LE		(0 << 25) /* Exception Little-endian
> >> */
> >>> +#define SCTLR_EL1_BIT23_22_RES1	(3 << 22) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL1_BIT20_RES1	(1 << 20) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL1_WXN_DIS	(0 << 19) /* Write permission is not XN
> >> */
> >>> +#define SCTLR_EL1_NTWE_DIS	(0 << 18) /* WFE instruction
> disabled
> >> */
> >>> +#define SCTLR_EL1_NTWI_DIS	(0 << 16) /* WFI instruction
> disabled
> >> */
> >>> +#define SCTLR_EL1_UCT_DIS	(0 << 15) /* CTR_EL0 access disabled
> >> */
> >>> +#define SCTLR_EL1_DZE_DIS	(0 << 14) /* DC ZVA instruction disabled
> >> */
> >>> +#define SCTLR_EL1_ICACHE_DIS	(0 << 12) /* Instruction cache
> >> disabled       */
> >>> +#define SCTLR_EL1_BIT11_RES1	(1 << 11) /* Reserved, RES1
> >> */
> >>> +#define SCTLR_EL1_UMA_DIS	(0 << 9)  /* User Mask Access disabled
> >> */
> >>> +#define SCTLR_EL1_SED_EN	(0 << 8)  /* SETEND instruction enabled
> >> */
> >>> +#define SCTLR_EL1_ITD_EN	(0 << 7)  /* IT instruction enabled
> >> */
> >>> +#define SCTLR_EL1_CP15BEN_DIS	(0 << 5)  /* CP15 barrier operation
> >> disabled  */
> >>> +#define SCTLR_EL1_SA0_DIS	(0 << 4)  /* Stack Alignment EL0 disabled
> >> */
> >>> +#define SCTLR_EL1_SA_DIS	(0 << 3)  /* Stack Alignment EL1 disabled
> >> */
> >>> +#define SCTLR_EL1_DCACHE_DIS	(0 << 2)  /* Data cache disabled
> >> */
> >>> +#define SCTLR_EL1_ALIGN_DIS	(0 << 1)  /* Alignment check
> disabled
> >> */
> >>> +#define SCTLR_EL1_MMU_DIS	(0)       /* MMU disabled
> >> */
> >>> +
> >>>  #ifndef __ASSEMBLY__
> >>>
> >>>  u64 get_page_table_size(void);
> >>> @@ -100,8 +196,26 @@ void __asm_invalidate_icache_all(void);
> >>>  int __asm_flush_l3_cache(void);
> >>>  void __asm_switch_ttbr(u64 new_ttbr);
> >>>
> >>> -void armv8_switch_to_el2(void);
> >>> -void armv8_switch_to_el1(void);
> >>> +/*
> >>> + * Switch from EL3 to EL2 for ARMv8
> >>> + *
> >>> + * @entry_point: kernel entry point
> >>> + * @mach_nr:     machine nr
> >>> + * @fdt_addr:    fdt address
> >>> + * @es_flag:     execution state flag, ES_TO_AARCH64 or
> ES_TO_AARCH32
> >>> + */
> >>> +void armv8_switch_to_el2(u64 entry_point, u64 mach_nr, u64 fdt_addr,
> >>> +			 u64 es_flag);
> >>> +/*
> >>> + * Switch from EL2 to EL1 for ARMv8
> >>> + *
> >>> + * @entry_point: kernel entry point
> >>> + * @mach_nr:     machine nr
> >>> + * @fdt_addr:    fdt address
> >>> + * @es_flag:     execution state flag, ES_TO_AARCH64 or
> ES_TO_AARCH32
> >>> + */
> >>> +void armv8_switch_to_el1(u64 entry_point, u64 mach_nr, u64 fdt_addr,
> >>> +			 u64 es_flag);
> >>>  void gic_init(void);
> >>>  void gic_send_sgi(unsigned long sgino);  void
> >>> wait_for_wakeup(void); diff --git a/arch/arm/lib/bootm.c
> >>> b/arch/arm/lib/bootm.c index 0838d89..1bbf85c 100644
> >>> --- a/arch/arm/lib/bootm.c
> >>> +++ b/arch/arm/lib/bootm.c
> >>> @@ -189,13 +189,19 @@ static void setup_end_tag(bd_t *bd)  __weak
> >>> void setup_board_tags(struct tag **in_params) {}
> >>>
> >>>  #ifdef CONFIG_ARM64
> >>> -static void do_nonsec_virt_switch(void)
> >>> +static void do_nonsec_virt_switch(bootm_headers_t *images, int
> >>> +flag)
> >>>  {
> >>>  	smp_kick_all_cpus();
> >>>  	dcache_disable();	/* flush cache before swtiching to EL2 */
> >>> -	armv8_switch_to_el2();
> >>> +
> >>>  #ifdef CONFIG_ARMV8_SWITCH_TO_EL1
> >>> -	armv8_switch_to_el1();
> >>> +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
> >>> +			    (u64)images->ft_addr, ES_TO_AARCH64);
> >>> +	armv8_switch_to_el1((u64)images->ep, (u64)gd->bd->bi_arch_number,
> >>> +			    (u64)images->ft_addr, flag); #else
> >>> +	armv8_switch_to_el2((u64)images->ep, (u64)gd->bd->bi_arch_number,
> >>> +			    (u64)images->ft_addr, flag);
> >>>  #endif
> >>>  }
> >>>  #endif
> >>> @@ -275,6 +281,7 @@ static void boot_jump_linux(bootm_headers_t
> >> *images, int flag)
> >>>  	void (*kernel_entry)(void *fdt_addr, void *res0, void *res1,
> >>>  			void *res2);
> >>>  	int fake = (flag & BOOTM_STATE_OS_FAKE_GO);
> >>> +	int es_flag = ES_TO_AARCH64;
> >>>
> >>>  	kernel_entry = (void (*)(void *fdt_addr, void *res0, void *res1,
> >>>  				void *res2))images->ep;
> >>> @@ -286,7 +293,11 @@ static void boot_jump_linux(bootm_headers_t
> >> *images, int flag)
> >>>  	announce_and_cleanup(fake);
> >>>
> >>>  	if (!fake) {
> >>> -		do_nonsec_virt_switch();
> >>> +		if ((IH_ARCH_DEFAULT == IH_ARCH_ARM64) &&
> >>> +		    (images->os.arch == IH_ARCH_ARM))
> >>> +			es_flag = ES_TO_AARCH32;
> >>> +
> >>> +		do_nonsec_virt_switch(images, es_flag);
> >>>  		kernel_entry(images->ft_addr, NULL, NULL, NULL);
> >>>  	}
> >>>  #else
> >>> diff --git a/common/image-fit.c b/common/image-fit.c index
> >>> 9873957..982021b 100644
> >>> --- a/common/image-fit.c
> >>> +++ b/common/image-fit.c
> >>> @@ -26,6 +26,7 @@ DECLARE_GLOBAL_DATA_PTR;  #include <u-boot/md5.h>
> >>> #include <u-boot/sha1.h>  #include <u-boot/sha256.h>
> >>> +#include <generated/autoconf.h>
> >>>
> >>>
> >>> /*******************************************************************
> >>> **
> >>> ********/
> >>>  /* New uImage format routines */
> >>> @@ -1164,7 +1165,9 @@ int fit_image_check_arch(const void *fit, int
> >> noffset, uint8_t arch)
> >>>  	if (fit_image_get_arch(fit, noffset, &image_arch))
> >>>  		return 0;
> >>>  	return (arch == image_arch) ||
> >>> -		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64);
> >>> +		(arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64) ||
> >>> +		(arch == IH_ARCH_ARM64 && image_arch == IH_ARCH_ARM &&
> >>> +		 CONFIG_ARM64_SUPPORT_AARCH32);
> >>
> >> Uh, I think this define is simply not defined for non-compatible
> >> systems (including non-arm64). This really needs to go into an #ifdef.
> > [Alison Wang] I think arch == IH_ARCH_ARM64 will check for non-arm64
> system.
> > I can't understand your concern. Please explain more.
> 
> You either get a
> 
> #define CONFIG_ARM64_SUPPORT_AARCH32 1
> 
> in the config header if the option is set or simply nothing at all. If
> there's nothing, compiling this code will result in an unresolved symbol.
> 
[Alison Wang] Yes, I will change as follows in the next version.

int fit_image_check_arch(const void *fit, int noffset, uint8_t arch)
 {
        uint8_t image_arch;
+       int aarch32_support = 0;
+
+#ifdef CONFIG_ARM64_SUPPORT_AARCH32
+       aarch32_support = 1;
+#endif

        if (fit_image_get_arch(fit, noffset, &image_arch))
                return 0;
        return (arch == image_arch) ||
                (arch == IH_ARCH_I386 && image_arch == IH_ARCH_X86_64) ||
                (arch == IH_ARCH_ARM64 && image_arch == IH_ARCH_ARM &&
-                CONFIG_ARM64_SUPPORT_AARCH32);
+                aarch32_support);
 }


Best Regards,
Alison Wang

[U-Boot] [PATCH v3 1/2] armv8: Support loading 32-bit OS in AArch32 execution state

[Alexander Graf]

Am 03.06.2016 um 11:31 schrieb Huan Wang <alison.wang@nxp.com>:

>> On 03.06.16 05:11, Huan Wang wrote:
>>>> On 26.05.16 10:41, Alison Wang wrote:
>>>>> To support loading a 32-bit OS, the execution state will change from
>>>>> AArch64 to AArch32 when jumping to kernel.
>>>>> 
>>>>> The architecture information will be got through checking FIT image,
>>>>> then U-Boot will load 32-bit OS or 64-bit OS automatically.
>>>>> 
>>>>> Signed-off-by: Ebony Zhu <ebony.zhu@nxp.com>
>>>>> Signed-off-by: Alison Wang <alison.wang@nxp.com>
>>>>> Signed-off-by: Chenhui Zhao <chenhui.zhao@nxp.com>
>>>>> ---
>>>>> Changes in v3:
>>>>> - Comments the functions and the arguments.
>>>>> - Rename the real parameters.
>>>>> - Use the macros instead of the magic values.
>>>>> - Remove the redundant codes.
>>>>> - Clean up all of the mess in boot_jump_linux().
>>>>> - Add CONFIG_ARM64_SUPPORT_AARCH32 to detect for some ARM64 system
>>>> doesn't support AArch32 state.
>>>>> 
>>>>> Changes in v2:
>>>>> - armv8_switch_to_el2_aarch32() is removed. armv8_switch_to_el2_m is
>>>> used
>>>>>  to switch to AArch64 EL2 or AArch32 Hyp.
>>>>> - armv8_switch_to_el1_aarch32() is removed. armv8_switch_to_el1_m is
>>>> used
>>>>>  to switch to AArch64 EL1 or AArch32 SVC.
>>>>> 
>>>>> arch/arm/Kconfig                |   6 ++
>>>>> arch/arm/cpu/armv8/transition.S |   8 +-
>>>>> arch/arm/include/asm/macro.h    | 174
>> ++++++++++++++++++++++++++++++-
>>>> ---------
>>>>> arch/arm/include/asm/system.h   | 118 ++++++++++++++++++++++++++-
>>>>> arch/arm/lib/bootm.c            |  19 ++++-
>>>>> common/image-fit.c              |  14 +++-
>>>>> 6 files changed, 287 insertions(+), 52 deletions(-)
>>>>> 
>>>>> diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index
>>>>> 729b181..794cb4f 100644
>>>>> --- a/arch/arm/Kconfig
>>>>> +++ b/arch/arm/Kconfig
>>>>> @@ -77,6 +77,12 @@ config SYS_L2CACHE_OFF
>>>>>      If SoC does not support L2CACHE or one do not want to enable
>>>>>      L2CACHE, choose this option.
>>>>> 
>>>>> +config ARM64_SUPPORT_AARCH32
>>>>> +    bool "ARM64 system support AArch32 execution state"
>>>>> +    default y if ARM64 && !CONFIG_THUNDERX
>>>> 
>>>> I don't think that works. In Kconfig you need to omit the CONFIG_
>> parts.
>>>> It's "ARM64" also, not "CONFIG_ARM64".
>>>> 
>>>> Please just try to run a local make defconfig on the thunderx
>>>> reference system and see whether it includes the option or not.
>>> [Alison Wang] Yes, I missed it. Will change in the next version.
>>>> 
>>>>> +    help
>>>>> +      This ARM64 system supports AArch32 execution state.
>>>>> +
>>>>> choice
>>>>>    prompt "Target select"
>>>>>    default TARGET_HIKEY
>>>>> diff --git a/arch/arm/cpu/armv8/transition.S
>>>>> b/arch/arm/cpu/armv8/transition.S index 253a39b..417e8b4 100644
>>>>> --- a/arch/arm/cpu/armv8/transition.S
>>>>> +++ b/arch/arm/cpu/armv8/transition.S
>>>>> @@ -11,13 +11,13 @@
>>>>> #include <asm/macro.h>
>>>>> 
>>>>> ENTRY(armv8_switch_to_el2)
>>>>> -    switch_el x0, 1f, 0f, 0f
>>>>> +    switch_el x4, 1f, 0f, 0f
>>>>> 0:    ret
>>>>> -1:    armv8_switch_to_el2_m x0
>>>>> +1:    armv8_switch_to_el2_m x0, x1, x2, x3, x4
>>>>> ENDPROC(armv8_switch_to_el2)
>>>>> 
>>>>> ENTRY(armv8_switch_to_el1)
>>>>> -    switch_el x0, 0f, 1f, 0f
>>>>> +    switch_el x4, 0f, 1f, 0f
>>>>> 0:    ret
>>>>> -1:    armv8_switch_to_el1_m x0, x1
>>>>> +1:    armv8_switch_to_el1_m x0, x1, x2, x3, x4
>>>>> ENDPROC(armv8_switch_to_el1)
>>>>> diff --git a/arch/arm/include/asm/macro.h
>>>>> b/arch/arm/include/asm/macro.h index 9bb0efa..dd2c510 100644
>>>>> --- a/arch/arm/include/asm/macro.h
>>>>> +++ b/arch/arm/include/asm/macro.h
>>>>> @@ -8,6 +8,9 @@
>>>>> 
>>>>> #ifndef __ASM_ARM_MACRO_H__
>>>>> #define __ASM_ARM_MACRO_H__
>>>>> +
>>>>> +#include <asm/system.h>
>>>>> +
>>>>> #ifdef __ASSEMBLY__
>>>>> 
>>>>> /*
>>>>> @@ -135,13 +138,18 @@ lr    .req    x30
>>>>> #endif
>>>>> .endm
>>>>> 
>>>>> -.macro armv8_switch_to_el2_m, xreg1
>>>>> -    /* 64bit EL2 | HCE | SMD | RES1 (Bits[5:4]) | Non-secure EL0/EL1
>>>> */
>>>>> -    mov    \xreg1, #0x5b1
>>>>> -    msr    scr_el3, \xreg1
>>>>> +/*
>>>>> + * Switch from EL3 to EL2 for ARMv8
>>>>> + * @ep:     kernel entry point
>>>>> + * @arch:   machine nr
>>>>> + * @ftaddr: fdt address
>>>>> + * @flag:   The execution state flag for lower exception
>>>>> + *          level, ES_TO_AARCH64 or ES_TO_AARCH32
>>>>> + */
>>>>> +.macro armv8_switch_to_el2_m, ep, arch, ftaddr, flag, xreg5
>>>> 
>>>> You're never really using the arch and ftaddr arguments. Just
>>>> describe in the macro header that they need to be in x1 and x2
>>>> respectively and leave them out of the argument list for the macro.
>>> [Alison Wang] Do you mean rename arch and ftaddr to x1 and x2 here?
>> 
>> I mean omit arch and ftaddr from the macro arguments and instead mention
>> that x1 and x2 get passed as x1 and x2 to the guest and are usually
>> "ftaddr" and "arch".
>> 
>> Also it would make the code more consistent to always use ep.
> [Alison Wang] I don't think arch and ftaddr can be removed from the
> macro arguments in the current code. For 32-bit kernel startup entry
> point, the requirements are: r0 = 0, r1 = machine nr, r2 = atags or
> dtb pointer.

You don't refer to them, so they can be removed. Don't remove them from the function prototype - only from the macro.

Or if you find it too confusing to read then, pass a struct that contains ep, x0, x1, x2 and load the values from there.

Alex