From 9b3d423707c3b1f6633be1be7e959623e10c596b Mon Sep 17 00:00:00 2001 From: Jiada Wang Date: Wed, 30 Oct 2013 04:25:51 -0700 Subject: [PATCH 1/9] ARM: i.MX6q: fix the wrong parent of can_root clock instead of pll3_usb_otg the parent of can_root clock should be pll3_60m. Signed-off-by: Jiada Wang Signed-off-by: Shawn Guo --- arch/arm/mach-imx/clk-imx6q.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mach-imx/clk-imx6q.c b/arch/arm/mach-imx/clk-imx6q.c index d756d91fd741..edd522e8c4af 100644 --- a/arch/arm/mach-imx/clk-imx6q.c +++ b/arch/arm/mach-imx/clk-imx6q.c @@ -298,7 +298,7 @@ static void __init imx6q_clocks_init(struct device_node *ccm_node) clk[asrc_podf] = imx_clk_divider("asrc_podf", "asrc_pred", base + 0x30, 9, 3); clk[spdif_pred] = imx_clk_divider("spdif_pred", "spdif_sel", base + 0x30, 25, 3); clk[spdif_podf] = imx_clk_divider("spdif_podf", "spdif_pred", base + 0x30, 22, 3); - clk[can_root] = imx_clk_divider("can_root", "pll3_usb_otg", base + 0x20, 2, 6); + clk[can_root] = imx_clk_divider("can_root", "pll3_60m", base + 0x20, 2, 6); clk[ecspi_root] = imx_clk_divider("ecspi_root", "pll3_60m", base + 0x38, 19, 6); clk[gpu2d_core_podf] = imx_clk_divider("gpu2d_core_podf", "gpu2d_core_sel", base + 0x18, 23, 3); clk[gpu3d_core_podf] = imx_clk_divider("gpu3d_core_podf", "gpu3d_core_sel", base + 0x18, 26, 3); From b6e23bb63f28f0a8ffa7cf9824fa48000c08f9b2 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Mon, 28 Oct 2013 16:37:12 +0800 Subject: [PATCH 2/9] ARM: imx: remove imx_src_prepare_restart() call There is ~10% possibility that the following emergency restart command fails to reboot imx6q. $ echo b > /proc/sysrq-trigger The IMX restart routine mxc_restart() assumes that it will always run on primary core, and will call imx_src_prepare_restart() to disable secondary cores in order to get them come to online in the following boot. However, the assumption is only true for normal kernel_restart() case where migrate_to_reboot_cpu() will be called to migrate to primary core, but not necessarily true for emergency_restart() case. So when emergency_restart() calls into mxc_restart() on any secondary core, system will hang immediately once imx_src_prepare_restart() is called to disabled secondary cores. Since emergency_restart() is defined as a function that is safe to call in interrupt context, we cannot just call migrate_to_reboot_cpu() to fix the issue. Fortunately, we just found that the issue can be fixed at imx6q platform level. We used to call imx_src_prepare_restart() to disable all secondary cores before resetting hardware. Otherwise, the secondary will fail come to online in the reboot. However, we recently found that after commit 6050d18 (ARM: imx: reset core along with enable/disable operation) comes to play, we do not need to reset the secondary cores any more. That said, mxc_restart() now can run on any core to reboot the system, as long as we remove the imx_src_prepare_restart() call from mxc_restart(). So let's simply remove imx_src_prepare_restart() call to fix the above emergency restart failure. Reported-by: Jiada Wang Signed-off-by: Shawn Guo --- arch/arm/mach-imx/common.h | 5 ----- arch/arm/mach-imx/src.c | 15 --------------- arch/arm/mach-imx/system.c | 3 --- 3 files changed, 23 deletions(-) diff --git a/arch/arm/mach-imx/common.h b/arch/arm/mach-imx/common.h index 7cbe22d0c6e9..24a7899e36a8 100644 --- a/arch/arm/mach-imx/common.h +++ b/arch/arm/mach-imx/common.h @@ -127,11 +127,6 @@ static inline void imx_smp_prepare(void) {} static inline void imx_scu_standby_enable(void) {} #endif void imx_src_init(void); -#ifdef CONFIG_HAVE_IMX_SRC -void imx_src_prepare_restart(void); -#else -static inline void imx_src_prepare_restart(void) {} -#endif void imx_gpc_init(void); void imx_gpc_pre_suspend(void); void imx_gpc_post_resume(void); diff --git a/arch/arm/mach-imx/src.c b/arch/arm/mach-imx/src.c index 4754373e7e7d..45f7f4e0a447 100644 --- a/arch/arm/mach-imx/src.c +++ b/arch/arm/mach-imx/src.c @@ -115,21 +115,6 @@ void imx_set_cpu_arg(int cpu, u32 arg) writel_relaxed(arg, src_base + SRC_GPR1 + cpu * 8 + 4); } -void imx_src_prepare_restart(void) -{ - u32 val; - - /* clear enable bits of secondary cores */ - spin_lock(&scr_lock); - val = readl_relaxed(src_base + SRC_SCR); - val &= ~(0x7 << BP_SRC_SCR_CORE1_ENABLE); - writel_relaxed(val, src_base + SRC_SCR); - spin_unlock(&scr_lock); - - /* clear persistent entry register of primary core */ - writel_relaxed(0, src_base + SRC_GPR1); -} - void __init imx_src_init(void) { struct device_node *np; diff --git a/arch/arm/mach-imx/system.c b/arch/arm/mach-imx/system.c index e6edcd38b282..826b72ba08d9 100644 --- a/arch/arm/mach-imx/system.c +++ b/arch/arm/mach-imx/system.c @@ -42,9 +42,6 @@ void mxc_restart(enum reboot_mode mode, const char *cmd) { unsigned int wcr_enable; - if (cpu_is_imx6q() || cpu_is_imx6dl()) - imx_src_prepare_restart(); - if (wdog_clk) clk_enable(wdog_clk); From 2c11b57a8a8d83ffa91aebb12c90488c8802e6f3 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Thu, 31 Oct 2013 10:35:40 +0800 Subject: [PATCH 3/9] ARM: imx: improve mxc_restart() on the SRC bit writes The current comment in the code does not make it clear why the double writes on SRC bit is needed. Let's quote the errata to get it clear. Also, to ensure there are at least 2 writes happen in the same one 32kHz period, we actually need 3 writes. Let's add the third one. Signed-off-by: Shawn Guo --- arch/arm/mach-imx/system.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/arch/arm/mach-imx/system.c b/arch/arm/mach-imx/system.c index 826b72ba08d9..5e3027d3692f 100644 --- a/arch/arm/mach-imx/system.c +++ b/arch/arm/mach-imx/system.c @@ -52,7 +52,14 @@ void mxc_restart(enum reboot_mode mode, const char *cmd) /* Assert SRS signal */ __raw_writew(wcr_enable, wdog_base); - /* write twice to ensure the request will not get ignored */ + /* + * Due to imx6q errata ERR004346 (WDOG: WDOG SRS bit requires to be + * written twice), we add another two writes to ensure there must be at + * least two writes happen in the same one 32kHz clock period. We save + * the target check here, since the writes shouldn't be a huge burden + * for other platforms. + */ + __raw_writew(wcr_enable, wdog_base); __raw_writew(wcr_enable, wdog_base); /* wait for reset to assert... */ From 5d48217210a0ae4a08953b7e3db95ed5a992673d Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Thu, 31 Oct 2013 16:37:12 +0800 Subject: [PATCH 4/9] ARM: imx: v7_cpu_resume() is needed by imx6sl build Building a kernel with the following options, CONFIG_SMP=n CONFIG_PM=y CONFIG_SOC_IMX6SL=y CONFIG_SOC_IMX6Q=n we will see the build error below. arch/arm/mach-imx/built-in.o: In function `imx6q_pm_enter': platform-spi_imx.c:(.text+0x2648): undefined reference to `v7_cpu_resume' make[1]: *** [vmlinux] Error 1 This is because that v7_cpu_resume() implemented in headsmp.S is also needed by imx6sl build. Let's build headsmp.S for CONFIG_SOC_IMX6SL as well. Reported-by: Russell King Signed-off-by: Shawn Guo --- arch/arm/mach-imx/Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/mach-imx/Makefile b/arch/arm/mach-imx/Makefile index bbe1f5bb799c..1789e2b31903 100644 --- a/arch/arm/mach-imx/Makefile +++ b/arch/arm/mach-imx/Makefile @@ -102,8 +102,8 @@ obj-$(CONFIG_SOC_IMX6SL) += clk-imx6sl.o mach-imx6sl.o ifeq ($(CONFIG_PM),y) obj-$(CONFIG_SOC_IMX6Q) += pm-imx6q.o headsmp.o -# i.MX6SL reuses pm-imx6q.c -obj-$(CONFIG_SOC_IMX6SL) += pm-imx6q.o +# i.MX6SL reuses i.MX6Q code +obj-$(CONFIG_SOC_IMX6SL) += pm-imx6q.o headsmp.o endif # i.MX5 based machines From ec9de6cd95a1ee326269699777301d5e1d5ddb75 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Lothar=20Wa=C3=9Fmann?= Date: Thu, 31 Oct 2013 12:55:48 +0100 Subject: [PATCH 5/9] ARM: imx6q: add missing sentinel to divider table MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The clk_enet_ref_table[] is missing a final empty entry as end of list marker. Also make the existing markers more obvious. Signed-off-by: Lothar Waßmann Signed-off-by: Shawn Guo --- arch/arm/mach-imx/clk-imx6q.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/arm/mach-imx/clk-imx6q.c b/arch/arm/mach-imx/clk-imx6q.c index edd522e8c4af..04cfd0fcb0e5 100644 --- a/arch/arm/mach-imx/clk-imx6q.c +++ b/arch/arm/mach-imx/clk-imx6q.c @@ -122,13 +122,14 @@ static struct clk_div_table clk_enet_ref_table[] = { { .val = 1, .div = 10, }, { .val = 2, .div = 5, }, { .val = 3, .div = 4, }, + { /* sentinel */ } }; static struct clk_div_table post_div_table[] = { { .val = 2, .div = 1, }, { .val = 1, .div = 2, }, { .val = 0, .div = 4, }, - { } + { /* sentinel */ } }; static struct clk_div_table video_div_table[] = { @@ -136,7 +137,7 @@ static struct clk_div_table video_div_table[] = { { .val = 1, .div = 2, }, { .val = 2, .div = 1, }, { .val = 3, .div = 4, }, - { } + { /* sentinel */ } }; static void __init imx6q_clocks_init(struct device_node *ccm_node) From 322503a15740bd9383bb4ed452e5dd5a40598170 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Wed, 30 Oct 2013 15:12:55 +0800 Subject: [PATCH 6/9] ARM: imx: add sleep for pllv3 relock The pllv3 relock time varies in the range of 50us ~ 500us, depending on the specific PLL type, e.g. 50us for ARM PLL and 450us for Audio/Video PLL. Let's add a usleep_range() call instead of doing busy wait during relock. Signed-off-by: Shawn Guo --- arch/arm/mach-imx/clk-pllv3.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/mach-imx/clk-pllv3.c b/arch/arm/mach-imx/clk-pllv3.c index f6640b6a7b31..c9ca19184420 100644 --- a/arch/arm/mach-imx/clk-pllv3.c +++ b/arch/arm/mach-imx/clk-pllv3.c @@ -12,6 +12,7 @@ #include #include +#include #include #include #include @@ -66,6 +67,7 @@ static int clk_pllv3_prepare(struct clk_hw *hw) break; if (time_after(jiffies, timeout)) break; + usleep_range(50, 500); } while (1); if (readl_relaxed(pll->base) & BM_PLL_LOCK) From bc3b84da8a55752d8c54005e558d59ac10fe9953 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Wed, 30 Oct 2013 15:56:22 +0800 Subject: [PATCH 7/9] ARM: imx: pllv3 needs relock in .set_rate() call The pllv3 nees relock not only when powering up but also when rate changes. The patch creates a helper function clk_pllv3_wait_lock() and moves the relock code from clk_pllv3_prepare() into there, so that both .prepare() and .set_rate() hooks of pllv3 can call into the helper for relocking. Since relock is only needed when PLL is powered up while clk_set_rate() could be called before clk is prepared, we need to add a check in clk_pllv3_wait_lock() to skip the relock if PLL is not powered. Signed-off-by: Shawn Guo --- arch/arm/mach-imx/clk-pllv3.c | 43 ++++++++++++++++++++--------------- 1 file changed, 25 insertions(+), 18 deletions(-) diff --git a/arch/arm/mach-imx/clk-pllv3.c b/arch/arm/mach-imx/clk-pllv3.c index c9ca19184420..df1736232961 100644 --- a/arch/arm/mach-imx/clk-pllv3.c +++ b/arch/arm/mach-imx/clk-pllv3.c @@ -46,10 +46,30 @@ struct clk_pllv3 { #define to_clk_pllv3(_hw) container_of(_hw, struct clk_pllv3, hw) +static int clk_pllv3_wait_lock(struct clk_pllv3 *pll) +{ + unsigned long timeout = jiffies + msecs_to_jiffies(10); + u32 val = readl_relaxed(pll->base) & BM_PLL_POWER; + + /* No need to wait for lock when pll is not powered up */ + if ((pll->powerup_set && !val) || (!pll->powerup_set && val)) + return 0; + + /* Wait for PLL to lock */ + do { + if (readl_relaxed(pll->base) & BM_PLL_LOCK) + break; + if (time_after(jiffies, timeout)) + break; + usleep_range(50, 500); + } while (1); + + return readl_relaxed(pll->base) & BM_PLL_LOCK ? 0 : -ETIMEDOUT; +} + static int clk_pllv3_prepare(struct clk_hw *hw) { struct clk_pllv3 *pll = to_clk_pllv3(hw); - unsigned long timeout; u32 val; val = readl_relaxed(pll->base); @@ -60,20 +80,7 @@ static int clk_pllv3_prepare(struct clk_hw *hw) val &= ~BM_PLL_POWER; writel_relaxed(val, pll->base); - timeout = jiffies + msecs_to_jiffies(10); - /* Wait for PLL to lock */ - do { - if (readl_relaxed(pll->base) & BM_PLL_LOCK) - break; - if (time_after(jiffies, timeout)) - break; - usleep_range(50, 500); - } while (1); - - if (readl_relaxed(pll->base) & BM_PLL_LOCK) - return 0; - else - return -ETIMEDOUT; + return clk_pllv3_wait_lock(pll); } static void clk_pllv3_unprepare(struct clk_hw *hw) @@ -148,7 +155,7 @@ static int clk_pllv3_set_rate(struct clk_hw *hw, unsigned long rate, val |= div; writel_relaxed(val, pll->base); - return 0; + return clk_pllv3_wait_lock(pll); } static const struct clk_ops clk_pllv3_ops = { @@ -204,7 +211,7 @@ static int clk_pllv3_sys_set_rate(struct clk_hw *hw, unsigned long rate, val |= div; writel_relaxed(val, pll->base); - return 0; + return clk_pllv3_wait_lock(pll); } static const struct clk_ops clk_pllv3_sys_ops = { @@ -278,7 +285,7 @@ static int clk_pllv3_av_set_rate(struct clk_hw *hw, unsigned long rate, writel_relaxed(mfn, pll->base + PLL_NUM_OFFSET); writel_relaxed(mfd, pll->base + PLL_DENOM_OFFSET); - return 0; + return clk_pllv3_wait_lock(pll); } static const struct clk_ops clk_pllv3_av_ops = { From 43c9b9e8a4c64b1dd3026ab233703a4321ac6d7c Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Thu, 31 Oct 2013 09:46:17 +0800 Subject: [PATCH 8/9] ARM: imx: set up pllv3 POWER and BYPASS sequentially Currently, POWER and BYPASS bits are set up in a single write to pllv3 register. This causes problem occasionally from the IPU/HDMI testing. Let's follow FSL BSP code to set up POWER bit, relock, and then BYPASS sequentially. Signed-off-by: Shawn Guo --- arch/arm/mach-imx/clk-pllv3.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/arch/arm/mach-imx/clk-pllv3.c b/arch/arm/mach-imx/clk-pllv3.c index df1736232961..61364050fccd 100644 --- a/arch/arm/mach-imx/clk-pllv3.c +++ b/arch/arm/mach-imx/clk-pllv3.c @@ -71,16 +71,24 @@ static int clk_pllv3_prepare(struct clk_hw *hw) { struct clk_pllv3 *pll = to_clk_pllv3(hw); u32 val; + int ret; val = readl_relaxed(pll->base); - val &= ~BM_PLL_BYPASS; if (pll->powerup_set) val |= BM_PLL_POWER; else val &= ~BM_PLL_POWER; writel_relaxed(val, pll->base); - return clk_pllv3_wait_lock(pll); + ret = clk_pllv3_wait_lock(pll); + if (ret) + return ret; + + val = readl_relaxed(pll->base); + val &= ~BM_PLL_BYPASS; + writel_relaxed(val, pll->base); + + return 0; } static void clk_pllv3_unprepare(struct clk_hw *hw) From 036e29918078e38b0b127b89deb4c595028c1bc0 Mon Sep 17 00:00:00 2001 From: Alexander Shiyan Date: Tue, 5 Nov 2013 18:00:18 +0400 Subject: [PATCH 9/9] ARM: dts: i.MX51: Fix OTG PHY clock Proper clock ID for USB OTG PHY is "usb_phy_gate". The patch changes this mismatch. Signed-off-by: Alexander Shiyan Signed-off-by: Shawn Guo --- arch/arm/boot/dts/imx51.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/boot/dts/imx51.dtsi b/arch/arm/boot/dts/imx51.dtsi index f4dcff3a9969..4bcdd3ad15e5 100644 --- a/arch/arm/boot/dts/imx51.dtsi +++ b/arch/arm/boot/dts/imx51.dtsi @@ -190,7 +190,7 @@ usbphy0: usbphy@0 { compatible = "usb-nop-xceiv"; - clocks = <&clks 124>; + clocks = <&clks 75>; clock-names = "main_clk"; status = "okay"; };