i.MX31 suspend and VSTBY question

[Nguyen Dinh-R00091]

Hi,


>-----Original Message-----
>From: linux-arm-kernel-bounces at lists.infradead.org [mailto:linux-arm-kernel-
>bounces at lists.infradead.org] On Behalf Of Helmut Raiger
>Sent: Monday, March 21, 2011 7:54 AM
>To: linux-arm-kernel at lists.infradead.org
>Subject: i.MX31 suspend and VSTBY question
>
>Hi,
>     I'm currently testing 'suspend to memory' on our i.MX31 platform,
>therefore I created a file named pm-imx31.c in arch/arm/plat-mxc. It
>contains some debugging code for the PMIC registers, that was hacked
>into mc13xxx_core.c.
>
>------------------- snip: pm-imx31.c --------------------------
>
>#include <linux/kernel.h>
>#include <linux/err.h>
>#include <linux/suspend.h>
>#include <linux/regulator/machine.h>
>
>#include <mach/system.h>
>// HACK
>#include "../mach-mx3/crm_regs.h"
>
>extern void mc13xxx_dump_registers(void);
>
>static void mxc_modify_reg(void *reg_offset, unsigned int mask,
>                         unsigned int data)
>{
>     u32 reg;
>
>     reg = __raw_readl(reg_offset);
>     reg = (reg & (~mask)) | data;
>     __raw_writel(reg, reg_offset);
>}
>
>static int mx31_suspend_prepare(void)
>{
>#ifdef CONFIG_PM_DEBUG
>     mc13xxx_dump_registers();
>#endif
>     return 0;
>}
>
>static int mx31_suspend_enter(suspend_state_t state)
>{
>     unsigned long *reg;
>
>     switch (state) {
>     case PM_SUSPEND_MEM:
>         /* TODO: Enable Well Bias, disable core clock in standby */
>         mxc_modify_reg(MXC_CCM_CCMR,
>                 MXC_CCM_CCMR_WBEN | MXC_CCM_CCMR_SBYCS,
>                 MXC_CCM_CCMR_SBYCS);

For our release that has LPM working:

		/* Enable Well Bias and set VSTBY
		  * VSTBY pin will be asserted during SR mode. This asks the
		  * PM IC to set the core voltage to the standby voltage
		  * Must clear the MXC_CCM_CCMR_SBYCS bit as well  */
		mxc_ccm_modify_reg(MXC_CCM_CCMR,
				   MXC_CCM_CCMR_WBEN | MXC_CCM_CCMR_VSTBY |
					MXC_CCM_CCMR_SBYCS,
				   MXC_CCM_CCMR_WBEN | MXC_CCM_CCMR_VSTBY);

Dinh

>
>         mxc_modify_reg(MXC_CCM_CCMR,
>                 MXC_CCM_CCMR_LPM_MASK,
>                 2 << MXC_CCM_CCMR_LPM_OFFSET);
>
>         /* enable VSTBY signalling */
>         mxc_modify_reg(MXC_CCM_CCMR, MXC_CCM_CCMR_VSTBY,
>MXC_CCM_CCMR_VSTBY);
>
>         printk(KERN_INFO "Now supsending! CCMR=0x%x\n",
>__raw_readl(MXC_CCM_CCMR));
>
>         for (reg = MXC_CCM_CCMR; reg <= MXC_CCM_PDR2; reg++)
>             printk(KERN_INFO "0x%p: 0x%08lx\n", reg, *reg);
>
>         arch_idle();
>
>         printk(KERN_INFO "Now resuming\n");
>         break;
>     default:
>         return -EINVAL;
>     }
>     return 0;
>}
>
>static void mx31_resume_done(void)
>{
>#ifdef CONFIG_PM_DEBUG
>     mc13xxx_dump_registers();
>#endif
>
>}
>
>struct platform_suspend_ops mx31_suspend_ops = {
>       .valid = suspend_valid_only_mem,
>       .prepare = mx31_suspend_prepare,
>       .enter = mx31_suspend_enter,
>       .end = mx31_resume_done,
>};
>
>static int __init mx31_pm_init(void)
>{
>       printk(KERN_INFO "Power Management for Freescale MX31\n");
>       suspend_set_ops(&mx31_suspend_ops);
>
>       return 0;
>}
>
>------------------- snip --------------------------
>
>'echo mem >/sys/power/state' does what is expected, however the VSTBY
>line is not asserted.
>
>We had this platform running with a 2.6.21 LTIB based kernel. When I use
>this image (hardly comparable)  the signal appears as expected. I'm
>running out of ideas, so any pointers would be very much appreciated.
>
>Helmut
>
>
>
>--
>Scanned by MailScanner.
>
>
>_______________________________________________
>linux-arm-kernel mailing list
>linux-arm-kernel at lists.infradead.org
>http://lists.infradead.org/mailman/listinfo/linux-arm-kernel