Hi Leo Tuesday, September 22, 2020 6:43 AM, Leo Li wrote: > > > > -----Original Message----- > > From: Ran Wang <ran.wan...@nxp.com> > > Sent: Wednesday, September 16, 2020 3:18 AM > > To: Leo Li <leoyang...@nxp.com>; Rob Herring <robh...@kernel.org>; > > Shawn Guo <shawn...@kernel.org> > > Cc: linuxppc-dev@lists.ozlabs.org; > > linux-arm-ker...@lists.infradead.org; > > devicet...@vger.kernel.org; linux-ker...@vger.kernel.org; Biwen Li > > <biwen...@nxp.com>; Ran Wang <ran.wan...@nxp.com> > > Subject: [PATCH 2/5] soc: fsl: handle RCPM errata A-008646 on SoC > > LS1021A > > > > From: Biwen Li <biwen...@nxp.com> > > > > Description: > > - Reading configuration register RCPM_IPPDEXPCR1 > > always return zero > > > > Workaround: > > - Save register RCPM_IPPDEXPCR1's value to > > register SCFG_SPARECR8.(uboot's psci also > > need reading value from the register SCFG_SPARECR8 > > to set register RCPM_IPPDEXPCR1) > > > > Impact: > > - FlexTimer module will cannot wakeup system in > Will not.. > Also it will be better to merge this with the issue description part above to > prevent confusion.
OK > > deep sleep on SoC LS1021A > > > > Signed-off-by: Biwen Li <biwen...@nxp.com> > > Signed-off-by: Ran Wang <ran.wan...@nxp.com> > > --- > > drivers/soc/fsl/rcpm.c | 42 > > +++++++++++++++++++++++++++++++++++++++++- > > 1 file changed, 41 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/soc/fsl/rcpm.c b/drivers/soc/fsl/rcpm.c index > > a093dbe..e6354f5 100644 > > --- a/drivers/soc/fsl/rcpm.c > > +++ b/drivers/soc/fsl/rcpm.c > > @@ -2,7 +2,7 @@ > > // > > // rcpm.c - Freescale QorIQ RCPM driver // -// Copyright 2019 NXP > > +// Copyright 2019-2020 NXP > > // > > // Author: Ran Wang <ran.wan...@nxp.com> > > > > @@ -13,6 +13,9 @@ > > #include <linux/slab.h> > > #include <linux/suspend.h> > > #include <linux/kernel.h> > > +#include <linux/acpi.h> > > +#include <linux/mfd/syscon.h> > > +#include <linux/regmap.h> > > > > #define RCPM_WAKEUP_CELL_MAX_SIZE 7 > > > > @@ -37,6 +40,9 @@ static int rcpm_pm_prepare(struct device *dev) > > struct device_node *np = dev->of_node; > > u32 value[RCPM_WAKEUP_CELL_MAX_SIZE + 1]; > > u32 setting[RCPM_WAKEUP_CELL_MAX_SIZE] = {0}; > > + struct regmap *scfg_addr_regmap = NULL; > > + u32 reg_offset[2]; > > + u32 reg_value = 0; > > > > rcpm = dev_get_drvdata(dev); > > if (!rcpm) > > @@ -90,6 +96,40 @@ static int rcpm_pm_prepare(struct device *dev) > > tmp |= ioread32be(address); > > iowrite32be(tmp, address); > > } > > + /* > > + * Workaround of errata A-008646 on SoC LS1021A: > > + * There is a bug of register ippdexpcr1. > > + * Reading configuration register RCPM_IPPDEXPCR1 > > + * always return zero. So save ippdexpcr1's value > > + * to register SCFG_SPARECR8.And the value of > > + * ippdexpcr1 will be read from SCFG_SPARECR8. > > + */ > > + if (device_property_present(dev, "fsl,ippdexpcr1-alt-addr")) > > { > > + if (dev_of_node(dev)) { > > + scfg_addr_regmap = > > syscon_regmap_lookup_by_phandle(np, > > + > > "fsl,ippdexpcr1-alt-addr"); > > + } else if (is_acpi_node(dev->fwnode)) { > > + continue; > > + } > > + > > + if (scfg_addr_regmap && (i == 1)) { > > + if (device_property_read_u32_array(dev, > > + "fsl,ippdexpcr1-alt-addr", > > + reg_offset, > > + 2)) { > > It is not necessary to read out the whole fsl,ippdexpcr1-alt-addr property if > we > only need the offset. Also you can change to use the > syscon_regmap_lookup_by_phandle_args() API above to simplify the code. Got it, will update it in next version, thanks. Regards, Ran > > + scfg_addr_regmap = NULL; > > + continue; > > + } > > + /* Read value from register SCFG_SPARECR8 > > */ > > + regmap_read(scfg_addr_regmap, > > + reg_offset[1], > > + ®_value); > > + /* Write value to register SCFG_SPARECR8 */ > > + regmap_write(scfg_addr_regmap, > > + reg_offset[1], > > + tmp | reg_value); > > + } > > + } > > } > > > > return 0; > > -- > > 2.7.4