Message ID | 20230707232626.374475-1-marex@denx.de |
---|---|
State | Changes Requested, archived |
Headers | show |
Series | [1/2] dt-bindings: remoteproc: imx_rproc: Support i.MX8MN/P MMIO | expand |
Context | Check | Description |
---|---|---|
robh/checkpatch | success | |
robh/patch-applied | success | |
robh/dtbs-check | warning | build log |
robh/dt-meta-schema | success |
> Subject: [PATCH 1/2] dt-bindings: remoteproc: imx_rproc: Support > i.MX8MN/P MMIO > > The MX8M CM7 boot via SMC call is problematic, since not all versions of > ATF support this interface. Document MMIO support used to boot the > CM7 on MX8MN/MP instead and discern MMIO interface using DT > compatible string. Document GPR register syscon phandle which is required > by the MMIO interface too. > How about introducing a Boolean property fsl,rproc-mmio and only valid for i.MX8MN/P? Regards, Peng. > Signed-off-by: Marek Vasut <marex@denx.de> > --- > Note that the MMIO being discerned using compatible string is similar > approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". > --- > Cc: Bjorn Andersson <andersson@kernel.org> > Cc: Conor Dooley <conor+dt@kernel.org> > Cc: Fabio Estevam <festevam@gmail.com> > Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> > Cc: Mathieu Poirier <mathieu.poirier@linaro.org> > Cc: NXP Linux Team <linux-imx@nxp.com> > Cc: Peng Fan <peng.fan@nxp.com> > Cc: Pengutronix Kernel Team <kernel@pengutronix.de> > Cc: Rob Herring <robh+dt@kernel.org> > Cc: Sascha Hauer <s.hauer@pengutronix.de> > Cc: Shawn Guo <shawnguo@kernel.org> > Cc: devicetree@vger.kernel.org > Cc: linux-arm-kernel@lists.infradead.org > Cc: linux-remoteproc@vger.kernel.org > --- > .../bindings/remoteproc/fsl,imx-rproc.yaml | 20 +++++++++++++++++++ > 1 file changed, 20 insertions(+) > > diff --git a/Documentation/devicetree/bindings/remoteproc/fsl,imx- > rproc.yaml b/Documentation/devicetree/bindings/remoteproc/fsl,imx- > rproc.yaml > index c940199ce89df..78bb976f738ee 100644 > --- a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml > +++ b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml > @@ -20,7 +20,9 @@ properties: > - fsl,imx7ulp-cm4 > - fsl,imx8mm-cm4 > - fsl,imx8mn-cm7 > + - fsl,imx8mn-cm7-mmio > - fsl,imx8mp-cm7 > + - fsl,imx8mp-cm7-mmio > - fsl,imx8mq-cm4 > - fsl,imx8qm-cm4 > - fsl,imx8qxp-cm4 > @@ -30,6 +32,11 @@ properties: > clocks: > maxItems: 1 > > + gpr: > + $ref: /schemas/types.yaml#/definitions/phandle > + description: > + Phandle to IOMUXC GPR block which provide access to CM7 CPUWAIT > bit. > + > syscon: > $ref: /schemas/types.yaml#/definitions/phandle > description: > @@ -84,6 +91,19 @@ properties: > required: > - compatible > > +allOf: > + - if: > + properties: > + compatible: > + not: > + contains: > + enum: > + - fsl,imx8mn-cm7-mmio > + - fsl,imx8mp-cm7-mmio > + then: > + properties: > + gpr: false > + > additionalProperties: false > > examples: > -- > 2.40.1
On 08/07/2023 01:26, Marek Vasut wrote: > The MX8M CM7 boot via SMC call is problematic, since not all versions > of ATF support this interface. Document MMIO support used to boot the > CM7 on MX8MN/MP instead and discern MMIO interface using DT compatible > string. Document GPR register syscon phandle which is required by the > MMIO interface too. > > Signed-off-by: Marek Vasut <marex@denx.de> > --- > Note that the MMIO being discerned using compatible string is similar > approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". > --- Same hardware should not have two compatibles, depending on the bus it is. Best regards, Krzysztof
On 7/10/23 10:13, Krzysztof Kozlowski wrote: > On 08/07/2023 01:26, Marek Vasut wrote: >> The MX8M CM7 boot via SMC call is problematic, since not all versions >> of ATF support this interface. Document MMIO support used to boot the >> CM7 on MX8MN/MP instead and discern MMIO interface using DT compatible >> string. Document GPR register syscon phandle which is required by the >> MMIO interface too. >> >> Signed-off-by: Marek Vasut <marex@denx.de> >> --- >> Note that the MMIO being discerned using compatible string is similar >> approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". >> --- > > Same hardware should not have two compatibles, depending on the bus it is. Yes, I know, but for whatever reason the ST does, so I picked the same approach here. What else would you suggest, boolean properly like Peng wrote ? The compatible is also easier to handle on driver level, at least in Linux, see 2/2 .
On Mon, Jul 10, 2023 at 11:10:48AM +0200, Marek Vasut wrote: > On 7/10/23 10:13, Krzysztof Kozlowski wrote: > > On 08/07/2023 01:26, Marek Vasut wrote: > > > The MX8M CM7 boot via SMC call is problematic, since not all versions > > > of ATF support this interface. Document MMIO support used to boot the > > > CM7 on MX8MN/MP instead and discern MMIO interface using DT compatible > > > string. Document GPR register syscon phandle which is required by the > > > MMIO interface too. > > > > > > Signed-off-by: Marek Vasut <marex@denx.de> > > > --- > > > Note that the MMIO being discerned using compatible string is similar > > > approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". > > > --- > > > > Same hardware should not have two compatibles, depending on the bus it is. > > Yes, I know, but for whatever reason the ST does, so I picked the same > approach here. What else would you suggest, boolean properly like Peng wrote > ? > > The compatible is also easier to handle on driver level, at least in Linux, > see 2/2 . IIRC, in the ST case, the programming model presented to the OS is different, so different compatibles. Seems like this case is similar. However, can't you try and SMC call and if it fails, use MMIO? Rob
On Sat, Jul 08, 2023 at 01:26:25AM +0200, Marek Vasut wrote: > The MX8M CM7 boot via SMC call is problematic, since not all versions > of ATF support this interface. Document MMIO support used to boot the > CM7 on MX8MN/MP instead and discern MMIO interface using DT compatible > string. Document GPR register syscon phandle which is required by the > MMIO interface too. > > Signed-off-by: Marek Vasut <marex@denx.de> > --- > Note that the MMIO being discerned using compatible string is similar > approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". > --- > Cc: Bjorn Andersson <andersson@kernel.org> > Cc: Conor Dooley <conor+dt@kernel.org> > Cc: Fabio Estevam <festevam@gmail.com> > Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> > Cc: Mathieu Poirier <mathieu.poirier@linaro.org> > Cc: NXP Linux Team <linux-imx@nxp.com> > Cc: Peng Fan <peng.fan@nxp.com> > Cc: Pengutronix Kernel Team <kernel@pengutronix.de> > Cc: Rob Herring <robh+dt@kernel.org> > Cc: Sascha Hauer <s.hauer@pengutronix.de> > Cc: Shawn Guo <shawnguo@kernel.org> > Cc: devicetree@vger.kernel.org > Cc: linux-arm-kernel@lists.infradead.org > Cc: linux-remoteproc@vger.kernel.org > --- > .../bindings/remoteproc/fsl,imx-rproc.yaml | 20 +++++++++++++++++++ > 1 file changed, 20 insertions(+) > > diff --git a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml > index c940199ce89df..78bb976f738ee 100644 > --- a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml > +++ b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml > @@ -20,7 +20,9 @@ properties: > - fsl,imx7ulp-cm4 > - fsl,imx8mm-cm4 > - fsl,imx8mn-cm7 > + - fsl,imx8mn-cm7-mmio > - fsl,imx8mp-cm7 > + - fsl,imx8mp-cm7-mmio > - fsl,imx8mq-cm4 > - fsl,imx8qm-cm4 > - fsl,imx8qxp-cm4 > @@ -30,6 +32,11 @@ properties: > clocks: > maxItems: 1 > > + gpr: fsl,gpr or fsl,iomuxc-gpr > + $ref: /schemas/types.yaml#/definitions/phandle > + description: > + Phandle to IOMUXC GPR block which provide access to CM7 CPUWAIT bit. > + > syscon: > $ref: /schemas/types.yaml#/definitions/phandle > description: > @@ -84,6 +91,19 @@ properties: > required: > - compatible > > +allOf: > + - if: > + properties: > + compatible: > + not: > + contains: > + enum: > + - fsl,imx8mn-cm7-mmio > + - fsl,imx8mp-cm7-mmio > + then: > + properties: > + gpr: false > + > additionalProperties: false > > examples: > -- > 2.40.1 >
On 7/10/23 17:09, Rob Herring wrote: > On Mon, Jul 10, 2023 at 11:10:48AM +0200, Marek Vasut wrote: >> On 7/10/23 10:13, Krzysztof Kozlowski wrote: >>> On 08/07/2023 01:26, Marek Vasut wrote: >>>> The MX8M CM7 boot via SMC call is problematic, since not all versions >>>> of ATF support this interface. Document MMIO support used to boot the >>>> CM7 on MX8MN/MP instead and discern MMIO interface using DT compatible >>>> string. Document GPR register syscon phandle which is required by the >>>> MMIO interface too. >>>> >>>> Signed-off-by: Marek Vasut <marex@denx.de> >>>> --- >>>> Note that the MMIO being discerned using compatible string is similar >>>> approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". >>>> --- >>> >>> Same hardware should not have two compatibles, depending on the bus it is. >> >> Yes, I know, but for whatever reason the ST does, so I picked the same >> approach here. What else would you suggest, boolean properly like Peng wrote >> ? >> >> The compatible is also easier to handle on driver level, at least in Linux, >> see 2/2 . > > IIRC, in the ST case, the programming model presented to the OS is > different, so different compatibles. Seems like this case is similar. > > However, can't you try and SMC call and if it fails, use MMIO? They already do that to determine the CM4 CPU state, and if the SMC errors out, then they indicate the CPU is detached. So no, not to my knowledge, which indeed would be my preferred solution though.
Good morning, On Sat, Jul 08, 2023 at 01:26:26AM +0200, Marek Vasut wrote: > The MX8M CM7 boot via SMC call is problematic, since not all versions > of ATF support this interface. Extend the MMIO support so it can boot > the CM7 on MX8MN/MP instead and discern the two alternatives using DT > compatible strings. > > Signed-off-by: Marek Vasut <marex@denx.de> > --- > Cc: Bjorn Andersson <andersson@kernel.org> > Cc: Conor Dooley <conor+dt@kernel.org> > Cc: Fabio Estevam <festevam@gmail.com> > Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> > Cc: Mathieu Poirier <mathieu.poirier@linaro.org> > Cc: NXP Linux Team <linux-imx@nxp.com> > Cc: Peng Fan <peng.fan@nxp.com> > Cc: Pengutronix Kernel Team <kernel@pengutronix.de> > Cc: Rob Herring <robh+dt@kernel.org> > Cc: Sascha Hauer <s.hauer@pengutronix.de> > Cc: Shawn Guo <shawnguo@kernel.org> > Cc: devicetree@vger.kernel.org > Cc: linux-arm-kernel@lists.infradead.org > Cc: linux-remoteproc@vger.kernel.org > --- > drivers/remoteproc/imx_rproc.c | 53 ++++++++++++++++++++++++++++++++-- > drivers/remoteproc/imx_rproc.h | 2 ++ > 2 files changed, 53 insertions(+), 2 deletions(-) > > diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c > index d0eb96d6a4fe1..09589f664a2be 100644 > --- a/drivers/remoteproc/imx_rproc.c > +++ b/drivers/remoteproc/imx_rproc.c > @@ -41,6 +41,12 @@ > #define IMX7D_M4_STOP (IMX7D_ENABLE_M4 | IMX7D_SW_M4C_RST | \ > IMX7D_SW_M4C_NON_SCLR_RST) > > +#define IMX8M_M7_STOP (IMX7D_ENABLE_M4 | IMX7D_SW_M4C_RST) > +#define IMX8M_M7_POLL IMX7D_ENABLE_M4 > + > +#define IMX8M_GPR22 0x58 > +#define IMX8M_GPR22_CM7_CPUWAIT BIT(0) > + > /* Address: 0x020D8000 */ > #define IMX6SX_SRC_SCR 0x00 > #define IMX6SX_ENABLE_M4 BIT(22) > @@ -92,6 +98,7 @@ static int imx_rproc_detach_pd(struct rproc *rproc); > struct imx_rproc { > struct device *dev; > struct regmap *regmap; > + struct regmap *gpr; > struct rproc *rproc; > const struct imx_rproc_dcfg *dcfg; > struct imx_rproc_mem mem[IMX_RPROC_MEM_MAX]; > @@ -287,6 +294,18 @@ static const struct imx_rproc_att imx_rproc_att_imx6sx[] = { > { 0x80000000, 0x80000000, 0x60000000, 0 }, > }; > > +static const struct imx_rproc_dcfg imx_rproc_cfg_imx8mn_mmio = { > + .src_reg = IMX7D_SRC_SCR, > + .src_mask = IMX7D_M4_RST_MASK, > + .src_start = IMX7D_M4_START, > + .src_stop = IMX8M_M7_STOP, > + .gpr_reg = IMX8M_GPR22, > + .gpr_wait = IMX8M_GPR22_CM7_CPUWAIT, > + .att = imx_rproc_att_imx8mn, > + .att_size = ARRAY_SIZE(imx_rproc_att_imx8mn), > + .method = IMX_RPROC_MMIO, > +}; > + > static const struct imx_rproc_dcfg imx_rproc_cfg_imx8mn = { > .att = imx_rproc_att_imx8mn, > .att_size = ARRAY_SIZE(imx_rproc_att_imx8mn), > @@ -367,8 +386,14 @@ static int imx_rproc_start(struct rproc *rproc) > > switch (dcfg->method) { > case IMX_RPROC_MMIO: > - ret = regmap_update_bits(priv->regmap, dcfg->src_reg, dcfg->src_mask, > - dcfg->src_start); > + if (priv->gpr) { > + ret = regmap_clear_bits(priv->gpr, dcfg->gpr_reg, > + dcfg->gpr_wait); > + } else { > + ret = regmap_update_bits(priv->regmap, dcfg->src_reg, > + dcfg->src_mask, > + dcfg->src_start); > + } > break; > case IMX_RPROC_SMC: > arm_smccc_smc(IMX_SIP_RPROC, IMX_SIP_RPROC_START, 0, 0, 0, 0, 0, 0, &res); > @@ -400,6 +425,16 @@ static int imx_rproc_stop(struct rproc *rproc) > > switch (dcfg->method) { > case IMX_RPROC_MMIO: > + if (priv->gpr) { > + ret = regmap_set_bits(priv->gpr, dcfg->gpr_reg, > + dcfg->gpr_wait); > + if (ret) { > + dev_err(priv->dev, > + "Failed to quiescence M4 platform!\n"); > + return ret; > + } > + } > + > ret = regmap_update_bits(priv->regmap, dcfg->src_reg, dcfg->src_mask, > dcfg->src_stop); > break; > @@ -988,6 +1023,10 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv) > break; > } > > + priv->gpr = syscon_regmap_lookup_by_phandle(dev->of_node, "gpr"); > + if (IS_ERR(priv->gpr)) > + priv->gpr = NULL; > + > regmap = syscon_regmap_lookup_by_phandle(dev->of_node, "syscon"); > if (IS_ERR(regmap)) { > dev_err(dev, "failed to find syscon\n"); > @@ -997,6 +1036,14 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv) > priv->regmap = regmap; > regmap_attach_dev(dev, regmap, &config); > > + if (priv->gpr) { > + ret = regmap_read(priv->gpr, dcfg->gpr_reg, &val); > + if (val & dcfg->gpr_wait) { > + imx_rproc_stop(priv->rproc); > + return 0; > + } > + } Can you provide a comment to explain what is happening above? In what kind of scenario would a GPR be set with the wait bit on startup? Thanks, Mathieu > + > ret = regmap_read(regmap, dcfg->src_reg, &val); > if (ret) { > dev_err(dev, "Failed to read src\n"); > @@ -1142,6 +1189,8 @@ static const struct of_device_id imx_rproc_of_match[] = { > { .compatible = "fsl,imx8mm-cm4", .data = &imx_rproc_cfg_imx8mq }, > { .compatible = "fsl,imx8mn-cm7", .data = &imx_rproc_cfg_imx8mn }, > { .compatible = "fsl,imx8mp-cm7", .data = &imx_rproc_cfg_imx8mn }, > + { .compatible = "fsl,imx8mn-cm7-mmio", .data = &imx_rproc_cfg_imx8mn_mmio }, > + { .compatible = "fsl,imx8mp-cm7-mmio", .data = &imx_rproc_cfg_imx8mn_mmio }, > { .compatible = "fsl,imx8qxp-cm4", .data = &imx_rproc_cfg_imx8qxp }, > { .compatible = "fsl,imx8qm-cm4", .data = &imx_rproc_cfg_imx8qm }, > { .compatible = "fsl,imx8ulp-cm33", .data = &imx_rproc_cfg_imx8ulp }, > diff --git a/drivers/remoteproc/imx_rproc.h b/drivers/remoteproc/imx_rproc.h > index 1c7e2127c7584..79a1b8956d142 100644 > --- a/drivers/remoteproc/imx_rproc.h > +++ b/drivers/remoteproc/imx_rproc.h > @@ -31,6 +31,8 @@ struct imx_rproc_dcfg { > u32 src_mask; > u32 src_start; > u32 src_stop; > + u32 gpr_reg; > + u32 gpr_wait; > const struct imx_rproc_att *att; > size_t att_size; > enum imx_rproc_method method; > -- > 2.40.1 >
On 7/17/23 19:10, Mathieu Poirier wrote: > Good morning, > > On Sat, Jul 08, 2023 at 01:26:26AM +0200, Marek Vasut wrote: >> The MX8M CM7 boot via SMC call is problematic, since not all versions >> of ATF support this interface. Extend the MMIO support so it can boot >> the CM7 on MX8MN/MP instead and discern the two alternatives using DT >> compatible strings. >> >> Signed-off-by: Marek Vasut <marex@denx.de> >> --- >> Cc: Bjorn Andersson <andersson@kernel.org> >> Cc: Conor Dooley <conor+dt@kernel.org> >> Cc: Fabio Estevam <festevam@gmail.com> >> Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> >> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> >> Cc: NXP Linux Team <linux-imx@nxp.com> >> Cc: Peng Fan <peng.fan@nxp.com> >> Cc: Pengutronix Kernel Team <kernel@pengutronix.de> >> Cc: Rob Herring <robh+dt@kernel.org> >> Cc: Sascha Hauer <s.hauer@pengutronix.de> >> Cc: Shawn Guo <shawnguo@kernel.org> >> Cc: devicetree@vger.kernel.org >> Cc: linux-arm-kernel@lists.infradead.org >> Cc: linux-remoteproc@vger.kernel.org >> --- >> drivers/remoteproc/imx_rproc.c | 53 ++++++++++++++++++++++++++++++++-- >> drivers/remoteproc/imx_rproc.h | 2 ++ >> 2 files changed, 53 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c >> index d0eb96d6a4fe1..09589f664a2be 100644 >> --- a/drivers/remoteproc/imx_rproc.c >> +++ b/drivers/remoteproc/imx_rproc.c >> @@ -41,6 +41,12 @@ >> #define IMX7D_M4_STOP (IMX7D_ENABLE_M4 | IMX7D_SW_M4C_RST | \ >> IMX7D_SW_M4C_NON_SCLR_RST) >> >> +#define IMX8M_M7_STOP (IMX7D_ENABLE_M4 | IMX7D_SW_M4C_RST) >> +#define IMX8M_M7_POLL IMX7D_ENABLE_M4 >> + >> +#define IMX8M_GPR22 0x58 >> +#define IMX8M_GPR22_CM7_CPUWAIT BIT(0) >> + >> /* Address: 0x020D8000 */ >> #define IMX6SX_SRC_SCR 0x00 >> #define IMX6SX_ENABLE_M4 BIT(22) >> @@ -92,6 +98,7 @@ static int imx_rproc_detach_pd(struct rproc *rproc); >> struct imx_rproc { >> struct device *dev; >> struct regmap *regmap; >> + struct regmap *gpr; >> struct rproc *rproc; >> const struct imx_rproc_dcfg *dcfg; >> struct imx_rproc_mem mem[IMX_RPROC_MEM_MAX]; >> @@ -287,6 +294,18 @@ static const struct imx_rproc_att imx_rproc_att_imx6sx[] = { >> { 0x80000000, 0x80000000, 0x60000000, 0 }, >> }; >> >> +static const struct imx_rproc_dcfg imx_rproc_cfg_imx8mn_mmio = { >> + .src_reg = IMX7D_SRC_SCR, >> + .src_mask = IMX7D_M4_RST_MASK, >> + .src_start = IMX7D_M4_START, >> + .src_stop = IMX8M_M7_STOP, >> + .gpr_reg = IMX8M_GPR22, >> + .gpr_wait = IMX8M_GPR22_CM7_CPUWAIT, >> + .att = imx_rproc_att_imx8mn, >> + .att_size = ARRAY_SIZE(imx_rproc_att_imx8mn), >> + .method = IMX_RPROC_MMIO, >> +}; >> + >> static const struct imx_rproc_dcfg imx_rproc_cfg_imx8mn = { >> .att = imx_rproc_att_imx8mn, >> .att_size = ARRAY_SIZE(imx_rproc_att_imx8mn), >> @@ -367,8 +386,14 @@ static int imx_rproc_start(struct rproc *rproc) >> >> switch (dcfg->method) { >> case IMX_RPROC_MMIO: >> - ret = regmap_update_bits(priv->regmap, dcfg->src_reg, dcfg->src_mask, >> - dcfg->src_start); >> + if (priv->gpr) { >> + ret = regmap_clear_bits(priv->gpr, dcfg->gpr_reg, >> + dcfg->gpr_wait); >> + } else { >> + ret = regmap_update_bits(priv->regmap, dcfg->src_reg, >> + dcfg->src_mask, >> + dcfg->src_start); >> + } >> break; >> case IMX_RPROC_SMC: >> arm_smccc_smc(IMX_SIP_RPROC, IMX_SIP_RPROC_START, 0, 0, 0, 0, 0, 0, &res); >> @@ -400,6 +425,16 @@ static int imx_rproc_stop(struct rproc *rproc) >> >> switch (dcfg->method) { >> case IMX_RPROC_MMIO: >> + if (priv->gpr) { >> + ret = regmap_set_bits(priv->gpr, dcfg->gpr_reg, >> + dcfg->gpr_wait); >> + if (ret) { >> + dev_err(priv->dev, >> + "Failed to quiescence M4 platform!\n"); >> + return ret; >> + } >> + } >> + >> ret = regmap_update_bits(priv->regmap, dcfg->src_reg, dcfg->src_mask, >> dcfg->src_stop); >> break; >> @@ -988,6 +1023,10 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv) >> break; >> } >> >> + priv->gpr = syscon_regmap_lookup_by_phandle(dev->of_node, "gpr"); >> + if (IS_ERR(priv->gpr)) >> + priv->gpr = NULL; >> + >> regmap = syscon_regmap_lookup_by_phandle(dev->of_node, "syscon"); >> if (IS_ERR(regmap)) { >> dev_err(dev, "failed to find syscon\n"); >> @@ -997,6 +1036,14 @@ static int imx_rproc_detect_mode(struct imx_rproc *priv) >> priv->regmap = regmap; >> regmap_attach_dev(dev, regmap, &config); >> >> + if (priv->gpr) { >> + ret = regmap_read(priv->gpr, dcfg->gpr_reg, &val); >> + if (val & dcfg->gpr_wait) { >> + imx_rproc_stop(priv->rproc); >> + return 0; >> + } >> + } > > Can you provide a comment to explain what is happening above? In what kind of > scenario would a GPR be set with the wait bit on startup? This happens right after power-on reset, the core indicates its in wait state in the GPR, but its not fully stopped, so we stop it and then the firmware can be loaded into it.
diff --git a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml index c940199ce89df..78bb976f738ee 100644 --- a/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml +++ b/Documentation/devicetree/bindings/remoteproc/fsl,imx-rproc.yaml @@ -20,7 +20,9 @@ properties: - fsl,imx7ulp-cm4 - fsl,imx8mm-cm4 - fsl,imx8mn-cm7 + - fsl,imx8mn-cm7-mmio - fsl,imx8mp-cm7 + - fsl,imx8mp-cm7-mmio - fsl,imx8mq-cm4 - fsl,imx8qm-cm4 - fsl,imx8qxp-cm4 @@ -30,6 +32,11 @@ properties: clocks: maxItems: 1 + gpr: + $ref: /schemas/types.yaml#/definitions/phandle + description: + Phandle to IOMUXC GPR block which provide access to CM7 CPUWAIT bit. + syscon: $ref: /schemas/types.yaml#/definitions/phandle description: @@ -84,6 +91,19 @@ properties: required: - compatible +allOf: + - if: + properties: + compatible: + not: + contains: + enum: + - fsl,imx8mn-cm7-mmio + - fsl,imx8mp-cm7-mmio + then: + properties: + gpr: false + additionalProperties: false examples:
The MX8M CM7 boot via SMC call is problematic, since not all versions of ATF support this interface. Document MMIO support used to boot the CM7 on MX8MN/MP instead and discern MMIO interface using DT compatible string. Document GPR register syscon phandle which is required by the MMIO interface too. Signed-off-by: Marek Vasut <marex@denx.de> --- Note that the MMIO being discerned using compatible string is similar approach to "st,stm32mp1-rcc" vs "st,stm32mp1-rcc-secure". --- Cc: Bjorn Andersson <andersson@kernel.org> Cc: Conor Dooley <conor+dt@kernel.org> Cc: Fabio Estevam <festevam@gmail.com> Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: NXP Linux Team <linux-imx@nxp.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Pengutronix Kernel Team <kernel@pengutronix.de> Cc: Rob Herring <robh+dt@kernel.org> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Shawn Guo <shawnguo@kernel.org> Cc: devicetree@vger.kernel.org Cc: linux-arm-kernel@lists.infradead.org Cc: linux-remoteproc@vger.kernel.org --- .../bindings/remoteproc/fsl,imx-rproc.yaml | 20 +++++++++++++++++++ 1 file changed, 20 insertions(+)