| Message ID | 20180627122919.23926-1-vigneshr@ti.com |
|---|---|
| Headers | show |
| Series | pci-dra7xx: Enable errata i870 workaround for RC mode | expand |
On Wednesday 27 June 2018 05:59 PM, Vignesh R wrote: > Errata i870 is applicable in both EP and RC mode. Therefore rename > function dra7xx_pcie_ep_unaligned_memaccess(), that implements errata > workaround, to dra7xx_pcie_unaligned_memaccess() and call it from a > common place. So, that errata workaround is applied for both modes of > operation. > > Reported-by: Chris Welch <Chris.Welch@viavisolutions.com> > Signed-off-by: Vignesh R <vigneshr@ti.com> Acked-by: Kishon Vijay Abraham I <kishon@ti.com> > --- > drivers/pci/controller/dwc/pci-dra7xx.c | 12 ++++++------ > 1 file changed, 6 insertions(+), 6 deletions(-) > > diff --git a/drivers/pci/controller/dwc/pci-dra7xx.c b/drivers/pci/controller/dwc/pci-dra7xx.c > index 345aab56ce8b..95d9076e3fde 100644 > --- a/drivers/pci/controller/dwc/pci-dra7xx.c > +++ b/drivers/pci/controller/dwc/pci-dra7xx.c > @@ -542,7 +542,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { > }; > > /* > - * dra7xx_pcie_ep_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 > + * dra7xx_pcie_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 > * @dra7xx: the dra7xx device where the workaround should be applied > * > * Access to the PCIe slave port that are not 32-bit aligned will result > @@ -552,7 +552,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { > * > * To avoid this issue set PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE to 1. > */ > -static int dra7xx_pcie_ep_unaligned_memaccess(struct device *dev) > +static int dra7xx_pcie_unaligned_memaccess(struct device *dev) > { > int ret; > struct device_node *np = dev->of_node; > @@ -695,6 +695,10 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) > if (dra7xx->link_gen < 0 || dra7xx->link_gen > 2) > dra7xx->link_gen = 2; > > + ret = dra7xx_pcie_unaligned_memaccess(dev); > + if (ret) > + dev_err(dev, "WA for Errata i870 not appplied. Update DT\n"); > + > switch (mode) { > case DW_PCIE_RC_TYPE: > if (!IS_ENABLED(CONFIG_PCI_DRA7XX_HOST)) { > @@ -717,10 +721,6 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) > dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, > DEVICE_TYPE_EP); > > - ret = dra7xx_pcie_ep_unaligned_memaccess(dev); > - if (ret) > - goto err_gpio; > - > ret = dra7xx_add_pcie_ep(dra7xx, pdev); > if (ret < 0) > goto err_gpio; > -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wednesday 27 June 2018 05:59 PM, Vignesh R wrote: > Bit positions of PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE and > PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE in CTRL_CORE_SMA_SW_7 are > incorrectly documented in the TRM. In fact, the bit positions are > swapped. Update the DT bindings for PCIe EP to reflect the same. > > Signed-off-by: Vignesh R <vigneshr@ti.com> Shouldn't this be sent to stable fixes? Thanks Kishon > --- > arch/arm/boot/dts/dra7.dtsi | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/arch/arm/boot/dts/dra7.dtsi b/arch/arm/boot/dts/dra7.dtsi > index 7bfe7f28e3bd..27ad193e1a87 100644 > --- a/arch/arm/boot/dts/dra7.dtsi > +++ b/arch/arm/boot/dts/dra7.dtsi > @@ -355,7 +355,7 @@ > ti,hwmods = "pcie1"; > phys = <&pcie1_phy>; > phy-names = "pcie-phy0"; > - ti,syscon-unaligned-access = <&scm_conf1 0x14 2>; > + ti,syscon-unaligned-access = <&scm_conf1 0x14 1>; > status = "disabled"; > }; > }; > -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Jun 27, 2018 at 05:59:17PM +0530, Vignesh R wrote: > Errata i870 is applicable in both EP and RC mode. Therefore rename > function dra7xx_pcie_ep_unaligned_memaccess(), that implements errata > workaround, to dra7xx_pcie_unaligned_memaccess() and call it from a > common place. So, that errata workaround is applied for both modes of > operation. > > Reported-by: Chris Welch <Chris.Welch@viavisolutions.com> > Signed-off-by: Vignesh R <vigneshr@ti.com> > --- > drivers/pci/controller/dwc/pci-dra7xx.c | 12 ++++++------ > 1 file changed, 6 insertions(+), 6 deletions(-) > > diff --git a/drivers/pci/controller/dwc/pci-dra7xx.c b/drivers/pci/controller/dwc/pci-dra7xx.c > index 345aab56ce8b..95d9076e3fde 100644 > --- a/drivers/pci/controller/dwc/pci-dra7xx.c > +++ b/drivers/pci/controller/dwc/pci-dra7xx.c > @@ -542,7 +542,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { > }; > > /* > - * dra7xx_pcie_ep_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 > + * dra7xx_pcie_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 > * @dra7xx: the dra7xx device where the workaround should be applied > * > * Access to the PCIe slave port that are not 32-bit aligned will result > @@ -552,7 +552,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { > * > * To avoid this issue set PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE to 1. > */ > -static int dra7xx_pcie_ep_unaligned_memaccess(struct device *dev) > +static int dra7xx_pcie_unaligned_memaccess(struct device *dev) > { > int ret; > struct device_node *np = dev->of_node; > @@ -695,6 +695,10 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) > if (dra7xx->link_gen < 0 || dra7xx->link_gen > 2) > dra7xx->link_gen = 2; > > + ret = dra7xx_pcie_unaligned_memaccess(dev); > + if (ret) > + dev_err(dev, "WA for Errata i870 not appplied. Update DT\n"); Hi Vignesh, Nit: s/appplied/applied Two questions: - Current code applies the unaligned_memaccess() workaround for all compatible variants. This is fine for current controllers (since they are all affected), the code path above will have to be reworked if there is any other compatible IP re-using the driver that does not require the workaround. - How do you want this series to go upstream ? If it goes via arm-soc, which I think it should, here is my ACK on this patch: Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Please let me know if I can drop this series from the PCI patchwork. Thanks, Lorenzo > + > switch (mode) { > case DW_PCIE_RC_TYPE: > if (!IS_ENABLED(CONFIG_PCI_DRA7XX_HOST)) { > @@ -717,10 +721,6 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) > dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, > DEVICE_TYPE_EP); > > - ret = dra7xx_pcie_ep_unaligned_memaccess(dev); > - if (ret) > - goto err_gpio; > - > ret = dra7xx_add_pcie_ep(dra7xx, pdev); > if (ret < 0) > goto err_gpio; > -- > 2.18.0 > -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Lorenzo, On Wednesday 18 July 2018 04:32 PM, Lorenzo Pieralisi wrote: > On Wed, Jun 27, 2018 at 05:59:17PM +0530, Vignesh R wrote: >> Errata i870 is applicable in both EP and RC mode. Therefore rename >> function dra7xx_pcie_ep_unaligned_memaccess(), that implements errata >> workaround, to dra7xx_pcie_unaligned_memaccess() and call it from a >> common place. So, that errata workaround is applied for both modes of >> operation. >> >> Reported-by: Chris Welch <Chris.Welch@viavisolutions.com> >> Signed-off-by: Vignesh R <vigneshr@ti.com> >> --- >> drivers/pci/controller/dwc/pci-dra7xx.c | 12 ++++++------ >> 1 file changed, 6 insertions(+), 6 deletions(-) >> >> diff --git a/drivers/pci/controller/dwc/pci-dra7xx.c b/drivers/pci/controller/dwc/pci-dra7xx.c >> index 345aab56ce8b..95d9076e3fde 100644 >> --- a/drivers/pci/controller/dwc/pci-dra7xx.c >> +++ b/drivers/pci/controller/dwc/pci-dra7xx.c >> @@ -542,7 +542,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { >> }; >> >> /* >> - * dra7xx_pcie_ep_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 >> + * dra7xx_pcie_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 >> * @dra7xx: the dra7xx device where the workaround should be applied >> * >> * Access to the PCIe slave port that are not 32-bit aligned will result >> @@ -552,7 +552,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { >> * >> * To avoid this issue set PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE to 1. >> */ >> -static int dra7xx_pcie_ep_unaligned_memaccess(struct device *dev) >> +static int dra7xx_pcie_unaligned_memaccess(struct device *dev) >> { >> int ret; >> struct device_node *np = dev->of_node; >> @@ -695,6 +695,10 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) >> if (dra7xx->link_gen < 0 || dra7xx->link_gen > 2) >> dra7xx->link_gen = 2; >> >> + ret = dra7xx_pcie_unaligned_memaccess(dev); >> + if (ret) >> + dev_err(dev, "WA for Errata i870 not appplied. Update DT\n"); > > Hi Vignesh, > > Nit: s/appplied/applied > Oops, let me know if you want me to resend with this fixed. > Two questions: > > - Current code applies the unaligned_memaccess() workaround for all > compatible variants. This is fine for current controllers (since > they are all affected), the code path above will have to be > reworked if there is any other compatible IP re-using the driver > that does not require the workaround. There are no compatible IPs that don't require this workaround. Also, I don't see this IP being re-used in future. If you insist, I can add a errata flag > - How do you want this series to go upstream ? If it goes via arm-soc, > which I think it should, here is my ACK on this patch: > Patch 1 and 2(dt bindings update and driver patch) can go in via PCI tree. And DT changes(patch 3 and 4) can be picked up by Tony via omap/arm-soc tree. They are mostly independent and should not cause any problems. Does that sound good? > Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> > Thanks! > Please let me know if I can drop this series from the PCI patchwork. > > Thanks, > Lorenzo > >> + >> switch (mode) { >> case DW_PCIE_RC_TYPE: >> if (!IS_ENABLED(CONFIG_PCI_DRA7XX_HOST)) { >> @@ -717,10 +721,6 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) >> dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, >> DEVICE_TYPE_EP); >> >> - ret = dra7xx_pcie_ep_unaligned_memaccess(dev); >> - if (ret) >> - goto err_gpio; >> - >> ret = dra7xx_add_pcie_ep(dra7xx, pdev); >> if (ret < 0) >> goto err_gpio; >> -- >> 2.18.0 >>
Hi Tony, On Tuesday 17 July 2018 03:25 PM, Kishon Vijay Abraham I wrote: > > > On Wednesday 27 June 2018 05:59 PM, Vignesh R wrote: >> Bit positions of PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE and >> PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE in CTRL_CORE_SMA_SW_7 are >> incorrectly documented in the TRM. In fact, the bit positions are >> swapped. Update the DT bindings for PCIe EP to reflect the same. >> >> Signed-off-by: Vignesh R <vigneshr@ti.com> > > Shouldn't this be sent to stable fixes? This patch fixes: Fixes: d23f3839fe97 ("ARM: dts: DRA7: Add pcie1 dt node for EP mode") Cc: stable@vger.kernel.org let me know if this needs to be resent with Fixes tag. Regards Vignesh > > Thanks > Kishon >> --- >> arch/arm/boot/dts/dra7.dtsi | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/arch/arm/boot/dts/dra7.dtsi b/arch/arm/boot/dts/dra7.dtsi >> index 7bfe7f28e3bd..27ad193e1a87 100644 >> --- a/arch/arm/boot/dts/dra7.dtsi >> +++ b/arch/arm/boot/dts/dra7.dtsi >> @@ -355,7 +355,7 @@ >> ti,hwmods = "pcie1"; >> phys = <&pcie1_phy>; >> phy-names = "pcie-phy0"; >> - ti,syscon-unaligned-access = <&scm_conf1 0x14 2>; >> + ti,syscon-unaligned-access = <&scm_conf1 0x14 1>; >> status = "disabled"; >> }; >> }; >> -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Jul 19, 2018 at 04:04:34PM +0530, Vignesh R wrote: > Hi Lorenzo, > > On Wednesday 18 July 2018 04:32 PM, Lorenzo Pieralisi wrote: > > On Wed, Jun 27, 2018 at 05:59:17PM +0530, Vignesh R wrote: > >> Errata i870 is applicable in both EP and RC mode. Therefore rename > >> function dra7xx_pcie_ep_unaligned_memaccess(), that implements errata > >> workaround, to dra7xx_pcie_unaligned_memaccess() and call it from a > >> common place. So, that errata workaround is applied for both modes of > >> operation. > >> > >> Reported-by: Chris Welch <Chris.Welch@viavisolutions.com> > >> Signed-off-by: Vignesh R <vigneshr@ti.com> > >> --- > >> drivers/pci/controller/dwc/pci-dra7xx.c | 12 ++++++------ > >> 1 file changed, 6 insertions(+), 6 deletions(-) > >> > >> diff --git a/drivers/pci/controller/dwc/pci-dra7xx.c b/drivers/pci/controller/dwc/pci-dra7xx.c > >> index 345aab56ce8b..95d9076e3fde 100644 > >> --- a/drivers/pci/controller/dwc/pci-dra7xx.c > >> +++ b/drivers/pci/controller/dwc/pci-dra7xx.c > >> @@ -542,7 +542,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { > >> }; > >> > >> /* > >> - * dra7xx_pcie_ep_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 > >> + * dra7xx_pcie_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 > >> * @dra7xx: the dra7xx device where the workaround should be applied > >> * > >> * Access to the PCIe slave port that are not 32-bit aligned will result > >> @@ -552,7 +552,7 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { > >> * > >> * To avoid this issue set PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE to 1. > >> */ > >> -static int dra7xx_pcie_ep_unaligned_memaccess(struct device *dev) > >> +static int dra7xx_pcie_unaligned_memaccess(struct device *dev) > >> { > >> int ret; > >> struct device_node *np = dev->of_node; > >> @@ -695,6 +695,10 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) > >> if (dra7xx->link_gen < 0 || dra7xx->link_gen > 2) > >> dra7xx->link_gen = 2; > >> > >> + ret = dra7xx_pcie_unaligned_memaccess(dev); > >> + if (ret) > >> + dev_err(dev, "WA for Errata i870 not appplied. Update DT\n"); > > > > Hi Vignesh, > > > > Nit: s/appplied/applied > > > > Oops, let me know if you want me to resend with this fixed. I can fix it, no problem but see below. > > Two questions: > > > > - Current code applies the unaligned_memaccess() workaround for all > > compatible variants. This is fine for current controllers (since > > they are all affected), the code path above will have to be > > reworked if there is any other compatible IP re-using the driver > > that does not require the workaround. > > There are no compatible IPs that don't require this workaround. Also, I > don't see this IP being re-used in future. If you insist, I can add a > errata flag I do not insist, I just pointed this out ;-) > > - How do you want this series to go upstream ? If it goes via arm-soc, > > which I think it should, here is my ACK on this patch: > > > Patch 1 and 2(dt bindings update and driver patch) can go in via PCI > tree. And DT changes(patch 3 and 4) can be picked up by Tony via > omap/arm-soc tree. They are mostly independent and should not cause any > problems. Does that sound good? It should be fine but technically as soon as patch (2) is applied we would have a regression if patches (3) and (4) are applied separately. You could re-order the series and send everything via arm-soc. It is not such a big deal, your choice, please let me know. Thanks, Lorenzo > > Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> > > > > Thanks! > > > Please let me know if I can drop this series from the PCI patchwork. > > > > Thanks, > > Lorenzo > > > >> + > >> switch (mode) { > >> case DW_PCIE_RC_TYPE: > >> if (!IS_ENABLED(CONFIG_PCI_DRA7XX_HOST)) { > >> @@ -717,10 +721,6 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) > >> dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, > >> DEVICE_TYPE_EP); > >> > >> - ret = dra7xx_pcie_ep_unaligned_memaccess(dev); > >> - if (ret) > >> - goto err_gpio; > >> - > >> ret = dra7xx_add_pcie_ep(dra7xx, pdev); > >> if (ret < 0) > >> goto err_gpio; > >> -- > >> 2.18.0 > >> > > -- > Regards > Vignesh -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thursday 19 July 2018 04:45 PM, Lorenzo Pieralisi wrote: > On Thu, Jul 19, 2018 at 04:04:34PM +0530, Vignesh R wrote: >> Hi Lorenzo, >> [...] >>>> + ret = dra7xx_pcie_unaligned_memaccess(dev); >>>> + if (ret) >>>> + dev_err(dev, "WA for Errata i870 not appplied. Update DT\n"); >>> >>> Hi Vignesh, >>> >>> Nit: s/appplied/applied >>> >> >> Oops, let me know if you want me to resend with this fixed. > > I can fix it, no problem but see below. > >>> Two questions: >>> >>> - Current code applies the unaligned_memaccess() workaround for all >>> compatible variants. This is fine for current controllers (since >>> they are all affected), the code path above will have to be >>> reworked if there is any other compatible IP re-using the driver >>> that does not require the workaround. >> >> There are no compatible IPs that don't require this workaround. Also, I >> don't see this IP being re-used in future. If you insist, I can add a >> errata flag > > I do not insist, I just pointed this out ;-) > >>> - How do you want this series to go upstream ? If it goes via arm-soc, >>> which I think it should, here is my ACK on this patch: >>> >> Patch 1 and 2(dt bindings update and driver patch) can go in via PCI >> tree. And DT changes(patch 3 and 4) can be picked up by Tony via >> omap/arm-soc tree. They are mostly independent and should not cause any >> problems. Does that sound good? > > It should be fine but technically as soon as patch (2) is applied we > would have a regression if patches (3) and (4) are applied separately. > Right, although not a regression, I see your point. It would be good to merge patch 3 and 4 before patch 2. Tony, Let me know if you are okay with taking this series via omap tree. I can resend re-ordering patch 2 to come at the end of the series. Regards Vignesh > >>> Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> >>> >> >> Thanks! >> >>> Please let me know if I can drop this series from the PCI patchwork. >>> >>> Thanks, >>> Lorenzo >>> >>>> + >>>> switch (mode) { >>>> case DW_PCIE_RC_TYPE: >>>> if (!IS_ENABLED(CONFIG_PCI_DRA7XX_HOST)) { >>>> @@ -717,10 +721,6 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) >>>> dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, >>>> DEVICE_TYPE_EP); >>>> >>>> - ret = dra7xx_pcie_ep_unaligned_memaccess(dev); >>>> - if (ret) >>>> - goto err_gpio; >>>> - >>>> ret = dra7xx_add_pcie_ep(dra7xx, pdev); >>>> if (ret < 0) >>>> goto err_gpio; >>>> -- >>>> 2.18.0 >>>> >> >> -- >> Regards >> Vignesh
Make workaround for errata i870 applicable in Host mode as well(previously it was enabled only for EP mode) as per errata documentation: http://www.ti.com/lit/er/sprz450/sprz450.pdf Tested on DRA72 EVM Changes since v1: Drop IRQ handling rework (will be sent out separately) v1: https://lkml.org/lkml/2017/12/1/59 Vignesh R (4): dt-bindings: PCI: dra7xx: Add bindings for unaligned access in host mode pci: dwc: pci-dra7xx: Enable errata i870 for both EP and RC mode ARM: dts: dra7: Enable workaround for errata i870 in PCIe host mode ARM: dts: dra7: Fix up unaligned access setting for PCIe EP Documentation/devicetree/bindings/pci/ti-pci.txt | 5 +++++ arch/arm/boot/dts/dra7.dtsi | 4 +++- drivers/pci/controller/dwc/pci-dra7xx.c | 12 ++++++------ 3 files changed, 14 insertions(+), 7 deletions(-)