ARMv7M: Add STM32F4 support

Signed-off-by: Kamil Lulko <rev13@wp.pl>
Reviewed-by: Tom Rini <trini@konsulko.com>
diff --git a/arch/arm/cpu/armv7m/Makefile b/arch/arm/cpu/armv7m/Makefile
index aff60e8..b662e03 100644
--- a/arch/arm/cpu/armv7m/Makefile
+++ b/arch/arm/cpu/armv7m/Makefile
@@ -7,3 +7,5 @@
 
 extra-y := start.o
 obj-y += cpu.o
+
+obj-$(CONFIG_STM32F4) += stm32f4/
diff --git a/arch/arm/cpu/armv7m/stm32f4/Makefile b/arch/arm/cpu/armv7m/stm32f4/Makefile
new file mode 100644
index 0000000..e982830
--- /dev/null
+++ b/arch/arm/cpu/armv7m/stm32f4/Makefile
@@ -0,0 +1,11 @@
+#
+# (C) Copyright 2000-2006
+# Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+#
+# (C) Copyright 2015
+# Kamil Lulko, <rev13@wp.pl>
+#
+# SPDX-License-Identifier:	GPL-2.0+
+#
+
+obj-y += soc.o clock.o timer.o flash.o
diff --git a/arch/arm/cpu/armv7m/stm32f4/clock.c b/arch/arm/cpu/armv7m/stm32f4/clock.c
new file mode 100644
index 0000000..2eded1f
--- /dev/null
+++ b/arch/arm/cpu/armv7m/stm32f4/clock.c
@@ -0,0 +1,209 @@
+/*
+ * (C) Copyright 2015
+ * Kamil Lulko, <rev13@wp.pl>
+ *
+ * (C) Copyright 2014
+ * STMicroelectronics
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include <common.h>
+#include <asm/io.h>
+#include <asm/arch/stm32.h>
+
+#define RCC_CR_HSION		(1 << 0)
+#define RCC_CR_HSEON		(1 << 16)
+#define RCC_CR_HSERDY		(1 << 17)
+#define RCC_CR_HSEBYP		(1 << 18)
+#define RCC_CR_CSSON		(1 << 19)
+#define RCC_CR_PLLON		(1 << 24)
+#define RCC_CR_PLLRDY		(1 << 25)
+
+#define RCC_PLLCFGR_PLLM_MASK	0x3F
+#define RCC_PLLCFGR_PLLN_MASK	0x7FC0
+#define RCC_PLLCFGR_PLLP_MASK	0x30000
+#define RCC_PLLCFGR_PLLQ_MASK	0xF000000
+#define RCC_PLLCFGR_PLLSRC	(1 << 22)
+#define RCC_PLLCFGR_PLLN_SHIFT	6
+#define RCC_PLLCFGR_PLLP_SHIFT	16
+#define RCC_PLLCFGR_PLLQ_SHIFT	24
+
+#define RCC_CFGR_AHB_PSC_MASK	0xF0
+#define RCC_CFGR_APB1_PSC_MASK	0x1C00
+#define RCC_CFGR_APB2_PSC_MASK	0xE000
+#define RCC_CFGR_SW0		(1 << 0)
+#define RCC_CFGR_SW1		(1 << 1)
+#define RCC_CFGR_SW_MASK	0x3
+#define RCC_CFGR_SW_HSI		0
+#define RCC_CFGR_SW_HSE		RCC_CFGR_SW0
+#define RCC_CFGR_SW_PLL		RCC_CFGR_SW1
+#define RCC_CFGR_SWS0		(1 << 2)
+#define RCC_CFGR_SWS1		(1 << 3)
+#define RCC_CFGR_SWS_MASK	0xC
+#define RCC_CFGR_SWS_HSI	0
+#define RCC_CFGR_SWS_HSE	RCC_CFGR_SWS0
+#define RCC_CFGR_SWS_PLL	RCC_CFGR_SWS1
+#define RCC_CFGR_HPRE_SHIFT	4
+#define RCC_CFGR_PPRE1_SHIFT	10
+#define RCC_CFGR_PPRE2_SHIFT	13
+
+#define RCC_APB1ENR_PWREN	(1 << 28)
+
+#define PWR_CR_VOS0		(1 << 14)
+#define PWR_CR_VOS1		(1 << 15)
+#define PWR_CR_VOS_MASK		0xC000
+#define PWR_CR_VOS_SCALE_MODE_1	(PWR_CR_VOS0 | PWR_CR_VOS1)
+#define PWR_CR_VOS_SCALE_MODE_2	(PWR_CR_VOS1)
+#define PWR_CR_VOS_SCALE_MODE_3	(PWR_CR_VOS0)
+
+#define FLASH_ACR_WS(n)		n
+#define FLASH_ACR_PRFTEN	(1 << 8)
+#define FLASH_ACR_ICEN		(1 << 9)
+#define FLASH_ACR_DCEN		(1 << 10)
+
+struct pll_psc {
+	u8	pll_m;
+	u16	pll_n;
+	u8	pll_p;
+	u8	pll_q;
+	u8	ahb_psc;
+	u8	apb1_psc;
+	u8	apb2_psc;
+};
+
+#define AHB_PSC_1		0
+#define AHB_PSC_2		0x8
+#define AHB_PSC_4		0x9
+#define AHB_PSC_8		0xA
+#define AHB_PSC_16		0xB
+#define AHB_PSC_64		0xC
+#define AHB_PSC_128		0xD
+#define AHB_PSC_256		0xE
+#define AHB_PSC_512		0xF
+
+#define APB_PSC_1		0
+#define APB_PSC_2		0x4
+#define APB_PSC_4		0x5
+#define APB_PSC_8		0x6
+#define APB_PSC_16		0x7
+
+#if !defined(CONFIG_STM32_HSE_HZ)
+#error "CONFIG_STM32_HSE_HZ not defined!"
+#else
+#if (CONFIG_STM32_HSE_HZ == 8000000)
+struct pll_psc pll_psc_168 = {
+	.pll_m = 8,
+	.pll_n = 336,
+	.pll_p = 2,
+	.pll_q = 7,
+	.ahb_psc = AHB_PSC_1,
+	.apb1_psc = APB_PSC_4,
+	.apb2_psc = APB_PSC_2
+};
+#else
+#error "No PLL/Prescaler configuration for given CONFIG_STM32_HSE_HZ exists"
+#endif
+#endif
+
+int configure_clocks(void)
+{
+	/* Reset RCC configuration */
+	setbits_le32(&STM32_RCC->cr, RCC_CR_HSION);
+	writel(0, &STM32_RCC->cfgr); /* Reset CFGR */
+	clrbits_le32(&STM32_RCC->cr, (RCC_CR_HSEON | RCC_CR_CSSON
+		| RCC_CR_PLLON));
+	writel(0x24003010, &STM32_RCC->pllcfgr); /* Reset value from RM */
+	clrbits_le32(&STM32_RCC->cr, RCC_CR_HSEBYP);
+	writel(0, &STM32_RCC->cir); /* Disable all interrupts */
+
+	/* Configure for HSE+PLL operation */
+	setbits_le32(&STM32_RCC->cr, RCC_CR_HSEON);
+	while (!(readl(&STM32_RCC->cr) & RCC_CR_HSERDY))
+		;
+
+	/* Enable high performance mode, System frequency up to 168 MHz */
+	setbits_le32(&STM32_RCC->apb1enr, RCC_APB1ENR_PWREN);
+	writel(PWR_CR_VOS_SCALE_MODE_1, &STM32_PWR->cr);
+
+	setbits_le32(&STM32_RCC->cfgr, ((
+		pll_psc_168.ahb_psc << RCC_CFGR_HPRE_SHIFT)
+		| (pll_psc_168.apb1_psc << RCC_CFGR_PPRE1_SHIFT)
+		| (pll_psc_168.apb2_psc << RCC_CFGR_PPRE2_SHIFT)));
+
+	writel(pll_psc_168.pll_m
+		| (pll_psc_168.pll_n << RCC_PLLCFGR_PLLN_SHIFT)
+		| (((pll_psc_168.pll_p >> 1) - 1) << RCC_PLLCFGR_PLLP_SHIFT)
+		| (pll_psc_168.pll_q << RCC_PLLCFGR_PLLQ_SHIFT),
+		&STM32_RCC->pllcfgr);
+	setbits_le32(&STM32_RCC->pllcfgr, RCC_PLLCFGR_PLLSRC);
+
+	setbits_le32(&STM32_RCC->cr, RCC_CR_PLLON);
+
+	while (!(readl(&STM32_RCC->cr) & RCC_CR_PLLRDY))
+		;
+
+	/* 5 wait states, Prefetch enabled, D-Cache enabled, I-Cache enabled */
+	writel(FLASH_ACR_WS(5) | FLASH_ACR_PRFTEN | FLASH_ACR_ICEN
+		| FLASH_ACR_DCEN, &STM32_FLASH->acr);
+
+	clrbits_le32(&STM32_RCC->cfgr, (RCC_CFGR_SW0 | RCC_CFGR_SW1));
+	setbits_le32(&STM32_RCC->cfgr, RCC_CFGR_SW_PLL);
+
+	while ((readl(&STM32_RCC->cfgr) & RCC_CFGR_SWS_MASK) !=
+			RCC_CFGR_SWS_PLL)
+		;
+
+	return 0;
+}
+
+unsigned long clock_get(enum clock clck)
+{
+	u32 sysclk = 0;
+	u32 shift = 0;
+	/* Prescaler table lookups for clock computation */
+	u8 ahb_psc_table[16] = {
+		0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 3, 4, 6, 7, 8, 9
+	};
+	u8 apb_psc_table[8] = {
+		0, 0, 0, 0, 1, 2, 3, 4
+	};
+
+	if ((readl(&STM32_RCC->cfgr) & RCC_CFGR_SWS_MASK) ==
+			RCC_CFGR_SWS_PLL) {
+		u16 pllm, plln, pllp;
+		pllm = (readl(&STM32_RCC->pllcfgr) & RCC_PLLCFGR_PLLM_MASK);
+		plln = ((readl(&STM32_RCC->pllcfgr) & RCC_PLLCFGR_PLLN_MASK)
+			>> RCC_PLLCFGR_PLLN_SHIFT);
+		pllp = ((((readl(&STM32_RCC->pllcfgr) & RCC_PLLCFGR_PLLP_MASK)
+			>> RCC_PLLCFGR_PLLP_SHIFT) + 1) << 1);
+		sysclk = ((CONFIG_STM32_HSE_HZ / pllm) * plln) / pllp;
+	}
+
+	switch (clck) {
+	case CLOCK_CORE:
+		return sysclk;
+		break;
+	case CLOCK_AHB:
+		shift = ahb_psc_table[(
+			(readl(&STM32_RCC->cfgr) & RCC_CFGR_AHB_PSC_MASK)
+			>> RCC_CFGR_HPRE_SHIFT)];
+		return sysclk >>= shift;
+		break;
+	case CLOCK_APB1:
+		shift = apb_psc_table[(
+			(readl(&STM32_RCC->cfgr) & RCC_CFGR_APB1_PSC_MASK)
+			>> RCC_CFGR_PPRE1_SHIFT)];
+		return sysclk >>= shift;
+		break;
+	case CLOCK_APB2:
+		shift = apb_psc_table[(
+			(readl(&STM32_RCC->cfgr) & RCC_CFGR_APB2_PSC_MASK)
+			>> RCC_CFGR_PPRE2_SHIFT)];
+		return sysclk >>= shift;
+		break;
+	default:
+		return 0;
+		break;
+	}
+}
diff --git a/arch/arm/cpu/armv7m/stm32f4/flash.c b/arch/arm/cpu/armv7m/stm32f4/flash.c
new file mode 100644
index 0000000..e5c6111
--- /dev/null
+++ b/arch/arm/cpu/armv7m/stm32f4/flash.c
@@ -0,0 +1,143 @@
+/*
+ * (C) Copyright 2015
+ * Kamil Lulko, <rev13@wp.pl>
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include <common.h>
+#include <asm/io.h>
+#include <asm/arch/stm32.h>
+
+#define STM32_FLASH_KEY1	0x45670123
+#define STM32_FLASH_KEY2	0xCDEF89AB
+
+flash_info_t flash_info[CONFIG_SYS_MAX_FLASH_BANKS];
+
+const u32 sect_sz_kb[CONFIG_SYS_MAX_FLASH_SECT] = {
+	[0 ... 3] =	16 * 1024,
+	[4] =		64 * 1024,
+	[5 ... 11] =	128 * 1024
+};
+
+static void stm32f4_flash_lock(u8 lock)
+{
+	if (lock) {
+		setbits_le32(&STM32_FLASH->cr, STM32_FLASH_CR_LOCK);
+	} else {
+		writel(STM32_FLASH_KEY1, &STM32_FLASH->key);
+		writel(STM32_FLASH_KEY2, &STM32_FLASH->key);
+	}
+}
+
+unsigned long flash_init(void)
+{
+	unsigned long total_size = 0;
+	u8 i, j;
+
+	for (i = 0; i < CONFIG_SYS_MAX_FLASH_BANKS; i++) {
+		flash_info[i].flash_id = FLASH_STM32F4;
+		flash_info[i].sector_count = CONFIG_SYS_MAX_FLASH_SECT;
+		flash_info[i].start[0] = CONFIG_SYS_FLASH_BASE + (i << 20);
+		flash_info[i].size = sect_sz_kb[0];
+		for (j = 1; j < CONFIG_SYS_MAX_FLASH_SECT; j++) {
+			flash_info[i].start[j] = flash_info[i].start[j - 1]
+				+ (sect_sz_kb[j - 1]);
+			flash_info[i].size += sect_sz_kb[j];
+		}
+		total_size += flash_info[i].size;
+	}
+
+	return total_size;
+}
+
+void flash_print_info(flash_info_t *info)
+{
+	int i;
+
+	if (info->flash_id == FLASH_UNKNOWN) {
+		printf("missing or unknown FLASH type\n");
+		return;
+	} else if (info->flash_id == FLASH_STM32F4) {
+		printf("STM32F4 Embedded Flash\n");
+	}
+
+	printf("  Size: %ld MB in %d Sectors\n",
+	       info->size >> 20, info->sector_count);
+
+	printf("  Sector Start Addresses:");
+	for (i = 0; i < info->sector_count; ++i) {
+		if ((i % 5) == 0)
+			printf("\n   ");
+		printf(" %08lX%s",
+		       info->start[i],
+			info->protect[i] ? " (RO)" : "     ");
+	}
+	printf("\n");
+	return;
+}
+
+int flash_erase(flash_info_t *info, int first, int last)
+{
+	u8 bank = 0xFF;
+	int i;
+
+	for (i = 0; i < CONFIG_SYS_MAX_FLASH_BANKS; i++) {
+		if (info == &flash_info[i]) {
+			bank = i;
+			break;
+		}
+	}
+	if (bank == 0xFF)
+		return -1;
+
+	stm32f4_flash_lock(0);
+
+	for (i = first; i <= last; i++) {
+		while (readl(&STM32_FLASH->sr) & STM32_FLASH_SR_BSY)
+			;
+
+		if (bank == 0) {
+			setbits_le32(&STM32_FLASH->cr,
+				     (i << STM32_FLASH_CR_SNB_OFFSET));
+		} else if (bank == 1) {
+			setbits_le32(&STM32_FLASH->cr,
+				     ((0x10 | i) << STM32_FLASH_CR_SNB_OFFSET));
+		} else {
+			stm32f4_flash_lock(1);
+			return -1;
+		}
+		setbits_le32(&STM32_FLASH->cr, STM32_FLASH_CR_SER);
+		setbits_le32(&STM32_FLASH->cr, STM32_FLASH_CR_STRT);
+
+		while (readl(&STM32_FLASH->sr) & STM32_FLASH_SR_BSY)
+			;
+
+		clrbits_le32(&STM32_FLASH->cr, STM32_FLASH_CR_SER);
+		stm32f4_flash_lock(1);
+	}
+
+	return 0;
+}
+
+int write_buff(flash_info_t *info, uchar *src, ulong addr, ulong cnt)
+{
+	ulong i;
+
+	while (readl(&STM32_FLASH->sr) & STM32_FLASH_SR_BSY)
+		;
+
+	stm32f4_flash_lock(0);
+
+	setbits_le32(&STM32_FLASH->cr, STM32_FLASH_CR_PG);
+	/* To make things simple use byte writes only */
+	for (i = 0; i < cnt; i++) {
+		*(uchar *)(addr + i) = src[i];
+		while (readl(&STM32_FLASH->sr) & STM32_FLASH_SR_BSY)
+			;
+	}
+	clrbits_le32(&STM32_FLASH->cr, STM32_FLASH_CR_PG);
+	stm32f4_flash_lock(1);
+
+	return 0;
+}
diff --git a/arch/arm/cpu/armv7m/stm32f4/soc.c b/arch/arm/cpu/armv7m/stm32f4/soc.c
new file mode 100644
index 0000000..202a126
--- /dev/null
+++ b/arch/arm/cpu/armv7m/stm32f4/soc.c
@@ -0,0 +1,37 @@
+/*
+ * (C) Copyright 2015
+ * Kamil Lulko, <rev13@wp.pl>
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include <common.h>
+#include <asm/io.h>
+#include <asm/armv7m.h>
+#include <asm/arch/stm32.h>
+
+u32 get_cpu_rev(void)
+{
+	return 0;
+}
+
+int arch_cpu_init(void)
+{
+	configure_clocks();
+
+	/*
+	 * Configure the memory protection unit (MPU) to allow full access to
+	 * the whole 4GB address space.
+	 */
+	writel(0, &V7M_MPU->rnr);
+	writel(0, &V7M_MPU->rbar);
+	writel((V7M_MPU_RASR_AP_RW_RW | V7M_MPU_RASR_SIZE_4GB
+		| V7M_MPU_RASR_EN), &V7M_MPU->rasr);
+	writel(V7M_MPU_CTRL_ENABLE | V7M_MPU_CTRL_HFNMIENA, &V7M_MPU->ctrl);
+
+	return 0;
+}
+
+void s_init(void)
+{
+}
diff --git a/arch/arm/cpu/armv7m/stm32f4/timer.c b/arch/arm/cpu/armv7m/stm32f4/timer.c
new file mode 100644
index 0000000..102ae6d
--- /dev/null
+++ b/arch/arm/cpu/armv7m/stm32f4/timer.c
@@ -0,0 +1,118 @@
+/*
+ * (C) Copyright 2015
+ * Kamil Lulko, <rev13@wp.pl>
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include <common.h>
+#include <asm/io.h>
+#include <asm/armv7m.h>
+#include <asm/arch/stm32.h>
+
+DECLARE_GLOBAL_DATA_PTR;
+
+#define STM32_TIM2_BASE	(STM32_APB1PERIPH_BASE + 0x0000)
+
+#define RCC_APB1ENR_TIM2EN	(1 << 0)
+
+struct stm32_tim2_5 {
+	u32 cr1;
+	u32 cr2;
+	u32 smcr;
+	u32 dier;
+	u32 sr;
+	u32 egr;
+	u32 ccmr1;
+	u32 ccmr2;
+	u32 ccer;
+	u32 cnt;
+	u32 psc;
+	u32 arr;
+	u32 reserved1;
+	u32 ccr1;
+	u32 ccr2;
+	u32 ccr3;
+	u32 ccr4;
+	u32 reserved2;
+	u32 dcr;
+	u32 dmar;
+	u32 or;
+};
+
+#define TIM_CR1_CEN	(1 << 0)
+
+#define TIM_EGR_UG	(1 << 0)
+
+int timer_init(void)
+{
+	struct stm32_tim2_5 *tim = (struct stm32_tim2_5 *)STM32_TIM2_BASE;
+
+	setbits_le32(&STM32_RCC->apb1enr, RCC_APB1ENR_TIM2EN);
+
+	if (clock_get(CLOCK_AHB) == clock_get(CLOCK_APB1))
+		writel((clock_get(CLOCK_APB1) / CONFIG_SYS_HZ_CLOCK) - 1,
+		       &tim->psc);
+	else
+		writel(((clock_get(CLOCK_APB1) * 2) / CONFIG_SYS_HZ_CLOCK) - 1,
+		       &tim->psc);
+
+	writel(0xFFFFFFFF, &tim->arr);
+	writel(TIM_CR1_CEN, &tim->cr1);
+	setbits_le32(&tim->egr, TIM_EGR_UG);
+
+	gd->arch.tbl = 0;
+	gd->arch.tbu = 0;
+	gd->arch.lastinc = 0;
+
+	return 0;
+}
+
+ulong get_timer(ulong base)
+{
+	return (get_ticks() / (CONFIG_SYS_HZ_CLOCK / CONFIG_SYS_HZ)) - base;
+}
+
+unsigned long long get_ticks(void)
+{
+	struct stm32_tim2_5 *tim = (struct stm32_tim2_5 *)STM32_TIM2_BASE;
+	u32 now;
+
+	now = readl(&tim->cnt);
+
+	if (now >= gd->arch.lastinc)
+		gd->arch.tbl += (now - gd->arch.lastinc);
+	else
+		gd->arch.tbl += (0xFFFFFFFF - gd->arch.lastinc) + now;
+
+	gd->arch.lastinc = now;
+
+	return gd->arch.tbl;
+}
+
+void reset_timer(void)
+{
+	struct stm32_tim2_5 *tim = (struct stm32_tim2_5 *)STM32_TIM2_BASE;
+
+	gd->arch.lastinc = readl(&tim->cnt);
+	gd->arch.tbl = 0;
+}
+
+/* delay x useconds */
+void __udelay(ulong usec)
+{
+	unsigned long long start;
+
+	start = get_ticks();		/* get current timestamp */
+	while ((get_ticks() - start) < usec)
+		;			/* loop till time has passed */
+}
+
+/*
+ * This function is derived from PowerPC code (timebase clock frequency).
+ * On ARM it returns the number of timer ticks per second.
+ */
+ulong get_tbclk(void)
+{
+	return CONFIG_SYS_HZ_CLOCK;
+}