[PATCH] ARM: imx6: cpuidle: omit the unnecessary unmask of GINT

Shawn Guo shawnguo at kernel.org
Wed Mar 20 07:28:52 PDT 2019

Hi Peng,

On Wed, Mar 20, 2019 at 07:59:19AM +0000, Peng Fan wrote:
> > So are you saying the workaround of ERR007265 does not need to apply on
> > WAIT_CLOCKED mode?  But WAIT_CLOCKED is one of low-power modes,
> > isn't it?
> If understand correct, I think WAIT_CLOCKED is for run mode, not wait/stop mode.
> According to i.MX6Q RM, "18.6.18 CCM Low Power Control Register (CCM_CLPCR)"
> Setting the low power mode that system will enter on next assertion of dsm_request signal.
> NOTE: Set CCM_CGPR[INT_MEM_CLK_LPM] and CCM_CGPR[1] bits to 1 when setting
> CCM_CLPCR[LPM] bits to 01 (WAIT Mode) or 10 (STOP mode) without power gating.
> CCM_CGPR[INT_MEM_CLK_LPM] and CCM_CGPR[1] bits do not have to be set for STOP
> mode entry.
> 00 Remain in run mode
> 01 Transfer to wait mode
> 10 Transfer to stop mode
> 11 Reserved
> And according to the function:
> int imx6q_set_lpm(enum mxc_cpu_pwr_mode mode)
> {
>         u32 val = readl_relaxed(ccm_base + CLPCR);
>         val &= ~BM_CLPCR_LPM;
>         switch (mode) {
>         case WAIT_CLOCKED:
>                 break;
> WAIT_CLOCKED will clear the LPM bits.
> According to https://www.nxp.com/docs/en/errata/IMX6DQCE.pdf
> "
> ERR007265 CCM: When improper low-power sequence is used, the SoC enters
> low power mode before the ARM core executes WFI
> When software tries to enter Low-Power mode with the following sequence, the SoC enters
> Low-Power mode before the ARM core executes the WFI instruction:
> 1. Set CCM_CLPCR[1:0] to 2’b00
> 2. ARM core enters WFI
> 3. ARM core wakeup from an interrupt event, which is masked by GPC or not visible to GPC,
> such as an interrupt from a local timer
> 4. Set CCM_CLPCR[1:0] to 2’b01 or 2’b10
> 5. ARM core executes WFI
> Before the last step, the SoC enters WAIT mode if CCM_CLPCR[1:0] is set to 2’b01, or STOP
> mode if CCM_CLPCR[1:0] is set to 2’b10.
> "
> It mentions that soc will enter wait/stop mode early before wfi if set CCM_CLPCR to 1/2.
> But WAIT_CLOCKED is not wait mode or stop mode, it will set LPM to run mode.

Thanks much for the detailed explanation.  I messed WAIT_CLOCKED with
WAIT_UNCLOCKED.  Yes, I agree it's an optimization that we skip applying
the workaround to WAIT_CLOCKED.


More information about the linux-arm-kernel mailing list