ARM: imx: add smp support for imx7d

Add SMP support for i.MX7D, including CPU hotplug support, for
systems where TFA is not present.

The motivation for bringing up the second i.MX7D core inside the kernel
is that legacy vendor bootloaders usually do not implement PSCI support.

This is a significant blocker for systems in the field that are running old
bootloader versions to upgrade to a modern mainline kernel version, as only
one CPU of the i.MX7D would be brought up.

Bring up the second i.MX7D core inside the kernel to make the migration
path to mainline kernel easier for the existing iMX7D users.

Signed-off-by: Anson Huang <b20788@freescale.com>
Signed-off-by: Arulpandiyan Vadivel <arulpandiyan_vadivel@mentor.com> # Fix merge conflicts
Signed-off-by: Leonard Crestez <leonard.crestez@nxp.com>
Signed-off-by: Marek Vasut <marex@denx.de> # heavy cleanup
Signed-off-by: Fabio Estevam <festevam@denx.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>

7 files changed, 136 insertions, 11 deletions

diff --git a/arch/arm/mach-imx/Makefile b/arch/arm/mach-imx/Makefile
index 9cebd360d58e..d1506ef7a537 100644
--- a/arch/arm/mach-imx/Makefile
+++ b/arch/arm/mach-imx/Makefile
@@ -35,7 +35,7 @@ obj-$(CONFIG_HAVE_IMX_ANATOP) += anatop.o
 obj-$(CONFIG_HAVE_IMX_GPC) += gpc.o
 obj-$(CONFIG_HAVE_IMX_MMDC) += mmdc.o
 obj-$(CONFIG_HAVE_IMX_SRC) += src.o
-ifneq ($(CONFIG_SOC_IMX6)$(CONFIG_SOC_LS1021A),)
+ifneq ($(CONFIG_SOC_IMX6)$(CONFIG_SOC_IMX7D_CA7)$(CONFIG_SOC_LS1021A),)
 AFLAGS_headsmp.o :=-Wa,-march=armv7-a
 obj-$(CONFIG_SMP) += headsmp.o platsmp.o
 obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o
diff --git a/arch/arm/mach-imx/common.h b/arch/arm/mach-imx/common.h
index 474dedb73bc7..f0a073a71401 100644
--- a/arch/arm/mach-imx/common.h
+++ b/arch/arm/mach-imx/common.h
@@ -68,11 +68,13 @@ void imx_set_cpu_arg(int cpu, u32 arg);
 void v7_secondary_startup(void);
 void imx_scu_map_io(void);
 void imx_smp_prepare(void);
+void imx_gpcv2_set_core1_pdn_pup_by_software(bool pdn);
 #else
 static inline void imx_scu_map_io(void) {}
 static inline void imx_smp_prepare(void) {}
 #endif
 void imx_src_init(void);
+void imx7_src_init(void);
 void imx_gpc_pre_suspend(bool arm_power_off);
 void imx_gpc_post_resume(void);
 void imx_gpc_mask_all(void);
@@ -131,6 +133,7 @@ static inline void imx_init_l2cache(void) {}
 #endif
 
 extern const struct smp_operations imx_smp_ops;
+extern const struct smp_operations imx7_smp_ops;
 extern const struct smp_operations ls1021a_smp_ops;
 
 #endif
diff --git a/arch/arm/mach-imx/headsmp.S b/arch/arm/mach-imx/headsmp.S
index 766dbdb2ae27..fcba58be8e79 100644
--- a/arch/arm/mach-imx/headsmp.S
+++ b/arch/arm/mach-imx/headsmp.S
@@ -21,6 +21,15 @@ diag_reg_offset:
 
 ENTRY(v7_secondary_startup)
 ARM_BE8(setend be)			@ go BE8 if entered LE
+	mrc	p15, 0, r0, c0, c0, 0
+	lsl	r0, r0, #16
+	lsr	r0, r0, #20
+	/* 0xc07 is cortex A7's ID */
+	mov	r1, #0xc00
+	orr	r1, #0x7
+	cmp	r0, r1
+	beq	secondary_startup
+
 	set_diag_reg
 	b	secondary_startup
 ENDPROC(v7_secondary_startup)
diff --git a/arch/arm/mach-imx/hotplug.c b/arch/arm/mach-imx/hotplug.c
index 82e22398d43d..e24a46dc5703 100644
--- a/arch/arm/mach-imx/hotplug.c
+++ b/arch/arm/mach-imx/hotplug.c
@@ -11,6 +11,7 @@
 #include <asm/proc-fns.h>
 
 #include "common.h"
+#include "hardware.h"
 
 /*
  * platform-specific code to shutdown a CPU
@@ -40,5 +41,7 @@ int imx_cpu_kill(unsigned int cpu)
 			return 0;
 	imx_enable_cpu(cpu, false);
 	imx_set_cpu_arg(cpu, 0);
+	if (cpu_is_imx7d())
+		imx_gpcv2_set_core1_pdn_pup_by_software(true);
 	return 1;
 }
diff --git a/arch/arm/mach-imx/mach-imx7d.c b/arch/arm/mach-imx/mach-imx7d.c
index ccf64ddf8b7e..6fdd06bcf988 100644
--- a/arch/arm/mach-imx/mach-imx7d.c
+++ b/arch/arm/mach-imx/mach-imx7d.c
@@ -69,7 +69,7 @@ static void __init imx7d_init_late(void)
 static void __init imx7d_init_irq(void)
 {
 	imx_init_revision_from_anatop();
-	imx_src_init();
+	imx7_src_init();
 	irqchip_init();
 }
 
@@ -80,6 +80,7 @@ static const char *const imx7d_dt_compat[] __initconst = {
 };
 
 DT_MACHINE_START(IMX7D, "Freescale i.MX7 Dual (Device Tree)")
+	.smp            = smp_ops(imx7_smp_ops),
 	.init_irq	= imx7d_init_irq,
 	.init_machine	= imx7d_init_machine,
 	.init_late      = imx7d_init_late,
diff --git a/arch/arm/mach-imx/platsmp.c b/arch/arm/mach-imx/platsmp.c
index cf4e9335831c..972639038be5 100644
--- a/arch/arm/mach-imx/platsmp.c
+++ b/arch/arm/mach-imx/platsmp.c
@@ -92,6 +92,32 @@ const struct smp_operations imx_smp_ops __initconst = {
 #endif
 };
 
+/*
+ * Initialise the CPU possible map early - this describes the CPUs
+ * which may be present or become present in the system.
+ */
+static void __init imx7_smp_init_cpus(void)
+{
+	struct device_node *np;
+	int i, ncores = 0;
+
+	/* The iMX7D SCU does not report core count, get it from DT */
+	for_each_of_cpu_node(np)
+		ncores++;
+
+	for (i = ncores; i < NR_CPUS; i++)
+		set_cpu_possible(i, false);
+}
+
+const struct smp_operations imx7_smp_ops __initconst = {
+	.smp_init_cpus		= imx7_smp_init_cpus,
+	.smp_boot_secondary	= imx_boot_secondary,
+#ifdef CONFIG_HOTPLUG_CPU
+	.cpu_die		= imx_cpu_die,
+	.cpu_kill		= imx_cpu_kill,
+#endif
+};
+
 #define DCFG_CCSR_SCRATCHRW1	0x200
 
 static int ls1021a_boot_secondary(unsigned int cpu, struct task_struct *idle)
diff --git a/arch/arm/mach-imx/src.c b/arch/arm/mach-imx/src.c
index f52f371292ac..95fd1fbb0826 100644
--- a/arch/arm/mach-imx/src.c
+++ b/arch/arm/mach-imx/src.c
@@ -6,15 +6,19 @@
 
 #include <linux/init.h>
 #include <linux/io.h>
+#include <linux/iopoll.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
 #include <linux/reset-controller.h>
 #include <linux/smp.h>
 #include <asm/smp_plat.h>
 #include "common.h"
+#include "hardware.h"
 
 #define SRC_SCR				0x000
-#define SRC_GPR1			0x020
+#define SRC_GPR1_V1			0x020
+#define SRC_GPR1_V2			0x074
+#define SRC_GPR1(gpr_v2)		((gpr_v2) ? SRC_GPR1_V2 : SRC_GPR1_V1)
 #define BP_SRC_SCR_WARM_RESET_ENABLE	0
 #define BP_SRC_SCR_SW_GPU_RST		1
 #define BP_SRC_SCR_SW_VPU_RST		2
@@ -23,9 +27,18 @@
 #define BP_SRC_SCR_SW_IPU2_RST		12
 #define BP_SRC_SCR_CORE1_RST		14
 #define BP_SRC_SCR_CORE1_ENABLE		22
+/* below is for i.MX7D */
+#define SRC_A7RCR1			0x008
+#define BP_SRC_A7RCR1_A7_CORE1_ENABLE	1
+#define GPC_CPU_PGC_SW_PUP_REQ		0xf0
+#define GPC_CPU_PGC_SW_PDN_REQ		0xfc
+#define GPC_PGC_C1			0x840
+#define BM_CPU_PGC_SW_PDN_PUP_REQ_CORE1_A7	0x2
 
 static void __iomem *src_base;
 static DEFINE_SPINLOCK(scr_lock);
+static bool gpr_v2;
+static void __iomem *gpc_base;
 
 static const int sw_reset_bits[5] = {
 	BP_SRC_SCR_SW_GPU_RST,
@@ -73,17 +86,64 @@ static struct reset_controller_dev imx_reset_controller = {
 	.nr_resets = ARRAY_SIZE(sw_reset_bits),
 };
 
+static void imx_gpcv2_set_m_core_pgc(bool enable, u32 offset)
+{
+	writel_relaxed(enable, gpc_base + offset);
+}
+
+/*
+ * The motivation for bringing up the second i.MX7D core inside the kernel
+ * is that legacy vendor bootloaders usually do not implement PSCI support.
+ * This is a significant blocker for systems in the field that are running old
+ * bootloader versions to upgrade to a modern mainline kernel version, as only
+ * one CPU of the i.MX7D would be brought up.
+ * Bring up the second i.MX7D core inside the kernel to make the migration
+ * path to mainline kernel easier for the existing iMX7D users.
+ */
+void imx_gpcv2_set_core1_pdn_pup_by_software(bool pdn)
+{
+	u32 reg = pdn ? GPC_CPU_PGC_SW_PDN_REQ : GPC_CPU_PGC_SW_PUP_REQ;
+	u32 val, pup;
+	int ret;
+
+	imx_gpcv2_set_m_core_pgc(true, GPC_PGC_C1);
+	val = readl_relaxed(gpc_base + reg);
+	val |= BM_CPU_PGC_SW_PDN_PUP_REQ_CORE1_A7;
+	writel_relaxed(val, gpc_base + reg);
+
+	ret = readl_relaxed_poll_timeout_atomic(gpc_base + reg, pup,
+				!(pup & BM_CPU_PGC_SW_PDN_PUP_REQ_CORE1_A7),
+				5, 1000000);
+	if (ret < 0) {
+		pr_err("i.MX7D: CORE1_A7 power up timeout\n");
+		val &= ~BM_CPU_PGC_SW_PDN_PUP_REQ_CORE1_A7;
+		writel_relaxed(val, gpc_base + reg);
+	}
+
+	imx_gpcv2_set_m_core_pgc(false, GPC_PGC_C1);
+}
+
 void imx_enable_cpu(int cpu, bool enable)
 {
 	u32 mask, val;
 
 	cpu = cpu_logical_map(cpu);
-	mask = 1 << (BP_SRC_SCR_CORE1_ENABLE + cpu - 1);
 	spin_lock(&scr_lock);
-	val = readl_relaxed(src_base + SRC_SCR);
-	val = enable ? val | mask : val & ~mask;
-	val |= 1 << (BP_SRC_SCR_CORE1_RST + cpu - 1);
-	writel_relaxed(val, src_base + SRC_SCR);
+	if (gpr_v2) {
+		if (enable)
+			imx_gpcv2_set_core1_pdn_pup_by_software(false);
+
+		mask = 1 << (BP_SRC_A7RCR1_A7_CORE1_ENABLE + cpu - 1);
+		val = readl_relaxed(src_base + SRC_A7RCR1);
+		val = enable ? val | mask : val & ~mask;
+		writel_relaxed(val, src_base + SRC_A7RCR1);
+	} else {
+		mask = 1 << (BP_SRC_SCR_CORE1_ENABLE + cpu - 1);
+		val = readl_relaxed(src_base + SRC_SCR);
+		val = enable ? val | mask : val & ~mask;
+		val |= 1 << (BP_SRC_SCR_CORE1_RST + cpu - 1);
+		writel_relaxed(val, src_base + SRC_SCR);
+	}
 	spin_unlock(&scr_lock);
 }
 
@@ -91,19 +151,19 @@ void imx_set_cpu_jump(int cpu, void *jump_addr)
 {
 	cpu = cpu_logical_map(cpu);
 	writel_relaxed(__pa_symbol(jump_addr),
-		       src_base + SRC_GPR1 + cpu * 8);
+		       src_base + SRC_GPR1(gpr_v2) + cpu * 8);
 }
 
 u32 imx_get_cpu_arg(int cpu)
 {
 	cpu = cpu_logical_map(cpu);
-	return readl_relaxed(src_base + SRC_GPR1 + cpu * 8 + 4);
+	return readl_relaxed(src_base + SRC_GPR1(gpr_v2) + cpu * 8 + 4);
 }
 
 void imx_set_cpu_arg(int cpu, u32 arg)
 {
 	cpu = cpu_logical_map(cpu);
-	writel_relaxed(arg, src_base + SRC_GPR1 + cpu * 8 + 4);
+	writel_relaxed(arg, src_base + SRC_GPR1(gpr_v2) + cpu * 8 + 4);
 }
 
 void __init imx_src_init(void)
@@ -131,3 +191,26 @@ void __init imx_src_init(void)
 	writel_relaxed(val, src_base + SRC_SCR);
 	spin_unlock(&scr_lock);
 }
+
+void __init imx7_src_init(void)
+{
+	struct device_node *np;
+
+	gpr_v2 = true;
+
+	np = of_find_compatible_node(NULL, NULL, "fsl,imx7d-src");
+	if (!np)
+		return;
+
+	src_base = of_iomap(np, 0);
+	if (!src_base)
+		return;
+
+	np = of_find_compatible_node(NULL, NULL, "fsl,imx7d-gpc");
+	if (!np)
+		return;
+
+	gpc_base = of_iomap(np, 0);
+	if (!gpc_base)
+		return;
+}