| Message ID | 1477349722-26635-1-git-send-email-nick@shmanahar.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Mon, Oct 24, 2016 at 11:55:22PM +0100, Nick Dyer wrote: > We have encountered some RMI4 firmwares where there are blank pages in between > PDT pages which contain functions. Add a device tree property which can be set > to force reading the first N pages. Cann we get updated firmware instead? This seems like violation of RMI protocol. Or, if it is allowed by the protocol, can we avoid DT parameter and keep scanning until the end? Thanks. > > Signed-off-by: Nick Dyer <nick@shmanahar.org> > --- > Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt | 2 ++ > drivers/input/rmi4/rmi_driver.c | 16 ++++++++++++++-- > include/linux/rmi.h | 1 + > 3 files changed, 17 insertions(+), 2 deletions(-) > > diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt > index ec908b9..a611374 100644 > --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt > +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt > @@ -25,6 +25,8 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt > - syna,startup-delay-ms: The number of milliseconds to wait after powering on > the device. > > +- syna,pdt-scan-pages: Scan the first N PDT pages, ignoring blank pages. > + > - vdd-supply: VDD power supply. > See ../regulator/regulator.txt > > diff --git a/drivers/input/rmi4/rmi_driver.c b/drivers/input/rmi4/rmi_driver.c > index 4a88312..8f94160 100644 > --- a/drivers/input/rmi4/rmi_driver.c > +++ b/drivers/input/rmi4/rmi_driver.c > @@ -428,6 +428,8 @@ static int rmi_scan_pdt_page(struct rmi_device *rmi_dev, > const struct pdt_entry *entry)) > { > struct rmi_driver_data *data = dev_get_drvdata(&rmi_dev->dev); > + const struct rmi_device_platform_data *pdata = > + rmi_get_platform_data(rmi_dev); > struct pdt_entry pdt_entry; > u16 page_start = RMI4_PAGE_SIZE * page; > u16 pdt_start = page_start + PDT_START_SCAN_LOCATION; > @@ -449,8 +451,13 @@ static int rmi_scan_pdt_page(struct rmi_device *rmi_dev, > return retval; > } > > - return (data->f01_bootloader_mode || addr == pdt_start) ? > - RMI_SCAN_DONE : RMI_SCAN_CONTINUE; > + if (data->f01_bootloader_mode) > + return RMI_SCAN_DONE; > + > + if (page >= pdata->pdt_scan_pages && addr == pdt_start) > + return RMI_SCAN_DONE; > + > + return RMI_SCAN_CONTINUE; > } > > static int rmi_scan_pdt(struct rmi_device *rmi_dev, void *ctx, > @@ -832,6 +839,11 @@ static int rmi_driver_of_probe(struct device *dev, > if (retval) > return retval; > > + retval = rmi_of_property_read_u32(dev, &pdata->pdt_scan_pages, > + "syna,pdt-scan-pages", 1); > + if (retval) > + return retval; > + > return 0; > } > #else > diff --git a/include/linux/rmi.h b/include/linux/rmi.h > index e0aca14..2c4570d 100644 > --- a/include/linux/rmi.h > +++ b/include/linux/rmi.h > @@ -207,6 +207,7 @@ struct rmi_device_platform_data_spi { > */ > struct rmi_device_platform_data { > int reset_delay_ms; > + u32 pdt_scan_pages; > > struct rmi_device_platform_data_spi spi_data; > > -- > 2.7.4 >
On 10/24/2016 04:39 PM, Dmitry Torokhov wrote: > On Mon, Oct 24, 2016 at 11:55:22PM +0100, Nick Dyer wrote: >> We have encountered some RMI4 firmwares where there are blank pages in between >> PDT pages which contain functions. Add a device tree property which can be set >> to force reading the first N pages. > Cann we get updated firmware instead? This seems like violation of RMI > protocol. Or, if it is allowed by the protocol, can we avoid DT > parameter and keep scanning until the end? It is a violation of the RMI4 spec and we found at least one other device with misconfigured firmware, so this problem extends beyond just a single device. We are adding steps to our verification process to make sure we catch these misconfigurations in the future. But, I'm not sure we can guarantee that a device with this issue didn't go into production. I like Guenter's suggestion of requiring two empty pages to stop scanning. If we did that for all devices it would add one extra read to the PDT scan, but we would avoid us having to maintain a list of devices or set additional parameters. Andrew > Thanks. > >> Signed-off-by: Nick Dyer <nick@shmanahar.org> >> --- >> Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt | 2 ++ >> drivers/input/rmi4/rmi_driver.c | 16 ++++++++++++++-- >> include/linux/rmi.h | 1 + >> 3 files changed, 17 insertions(+), 2 deletions(-) >> >> diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt >> index ec908b9..a611374 100644 >> --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt >> +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt >> @@ -25,6 +25,8 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt >> - syna,startup-delay-ms: The number of milliseconds to wait after powering on >> the device. >> >> +- syna,pdt-scan-pages: Scan the first N PDT pages, ignoring blank pages. >> + >> - vdd-supply: VDD power supply. >> See ../regulator/regulator.txt >> >> diff --git a/drivers/input/rmi4/rmi_driver.c b/drivers/input/rmi4/rmi_driver.c >> index 4a88312..8f94160 100644 >> --- a/drivers/input/rmi4/rmi_driver.c >> +++ b/drivers/input/rmi4/rmi_driver.c >> @@ -428,6 +428,8 @@ static int rmi_scan_pdt_page(struct rmi_device *rmi_dev, >> const struct pdt_entry *entry)) >> { >> struct rmi_driver_data *data = dev_get_drvdata(&rmi_dev->dev); >> + const struct rmi_device_platform_data *pdata = >> + rmi_get_platform_data(rmi_dev); >> struct pdt_entry pdt_entry; >> u16 page_start = RMI4_PAGE_SIZE * page; >> u16 pdt_start = page_start + PDT_START_SCAN_LOCATION; >> @@ -449,8 +451,13 @@ static int rmi_scan_pdt_page(struct rmi_device *rmi_dev, >> return retval; >> } >> >> - return (data->f01_bootloader_mode || addr == pdt_start) ? >> - RMI_SCAN_DONE : RMI_SCAN_CONTINUE; >> + if (data->f01_bootloader_mode) >> + return RMI_SCAN_DONE; >> + >> + if (page >= pdata->pdt_scan_pages && addr == pdt_start) >> + return RMI_SCAN_DONE; >> + >> + return RMI_SCAN_CONTINUE; >> } >> >> static int rmi_scan_pdt(struct rmi_device *rmi_dev, void *ctx, >> @@ -832,6 +839,11 @@ static int rmi_driver_of_probe(struct device *dev, >> if (retval) >> return retval; >> >> + retval = rmi_of_property_read_u32(dev, &pdata->pdt_scan_pages, >> + "syna,pdt-scan-pages", 1); >> + if (retval) >> + return retval; >> + >> return 0; >> } >> #else >> diff --git a/include/linux/rmi.h b/include/linux/rmi.h >> index e0aca14..2c4570d 100644 >> --- a/include/linux/rmi.h >> +++ b/include/linux/rmi.h >> @@ -207,6 +207,7 @@ struct rmi_device_platform_data_spi { >> */ >> struct rmi_device_platform_data { >> int reset_delay_ms; >> + u32 pdt_scan_pages; >> >> struct rmi_device_platform_data_spi spi_data; >> >> -- >> 2.7.4 >> -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Oct 24, 2016 at 05:03:30PM -0700, Andrew Duggan wrote: > On 10/24/2016 04:39 PM, Dmitry Torokhov wrote: > >On Mon, Oct 24, 2016 at 11:55:22PM +0100, Nick Dyer wrote: > >>We have encountered some RMI4 firmwares where there are blank pages in between > >>PDT pages which contain functions. Add a device tree property which can be set > >>to force reading the first N pages. > >Cann we get updated firmware instead? This seems like violation of RMI > >protocol. Or, if it is allowed by the protocol, can we avoid DT > >parameter and keep scanning until the end? > > It is a violation of the RMI4 spec and we found at least one other > device with misconfigured firmware, so this problem extends beyond > just a single device. We are adding steps to our verification > process to make sure we catch these misconfigurations in the future. > But, I'm not sure we can guarantee that a device with this issue > didn't go into production. > > I like Guenter's suggestion of requiring two empty pages to stop > scanning. If we did that for all devices it would add one extra read > to the PDT scan, but we would avoid us having to maintain a list of > devices or set additional parameters. OK, if that fixes broken firmware then that's what I would prefer. Thanks.
diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt index ec908b9..a611374 100644 --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt @@ -25,6 +25,8 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt - syna,startup-delay-ms: The number of milliseconds to wait after powering on the device. +- syna,pdt-scan-pages: Scan the first N PDT pages, ignoring blank pages. + - vdd-supply: VDD power supply. See ../regulator/regulator.txt diff --git a/drivers/input/rmi4/rmi_driver.c b/drivers/input/rmi4/rmi_driver.c index 4a88312..8f94160 100644 --- a/drivers/input/rmi4/rmi_driver.c +++ b/drivers/input/rmi4/rmi_driver.c @@ -428,6 +428,8 @@ static int rmi_scan_pdt_page(struct rmi_device *rmi_dev, const struct pdt_entry *entry)) { struct rmi_driver_data *data = dev_get_drvdata(&rmi_dev->dev); + const struct rmi_device_platform_data *pdata = + rmi_get_platform_data(rmi_dev); struct pdt_entry pdt_entry; u16 page_start = RMI4_PAGE_SIZE * page; u16 pdt_start = page_start + PDT_START_SCAN_LOCATION; @@ -449,8 +451,13 @@ static int rmi_scan_pdt_page(struct rmi_device *rmi_dev, return retval; } - return (data->f01_bootloader_mode || addr == pdt_start) ? - RMI_SCAN_DONE : RMI_SCAN_CONTINUE; + if (data->f01_bootloader_mode) + return RMI_SCAN_DONE; + + if (page >= pdata->pdt_scan_pages && addr == pdt_start) + return RMI_SCAN_DONE; + + return RMI_SCAN_CONTINUE; } static int rmi_scan_pdt(struct rmi_device *rmi_dev, void *ctx, @@ -832,6 +839,11 @@ static int rmi_driver_of_probe(struct device *dev, if (retval) return retval; + retval = rmi_of_property_read_u32(dev, &pdata->pdt_scan_pages, + "syna,pdt-scan-pages", 1); + if (retval) + return retval; + return 0; } #else diff --git a/include/linux/rmi.h b/include/linux/rmi.h index e0aca14..2c4570d 100644 --- a/include/linux/rmi.h +++ b/include/linux/rmi.h @@ -207,6 +207,7 @@ struct rmi_device_platform_data_spi { */ struct rmi_device_platform_data { int reset_delay_ms; + u32 pdt_scan_pages; struct rmi_device_platform_data_spi spi_data;
We have encountered some RMI4 firmwares where there are blank pages in between PDT pages which contain functions. Add a device tree property which can be set to force reading the first N pages. Signed-off-by: Nick Dyer <nick@shmanahar.org> --- Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt | 2 ++ drivers/input/rmi4/rmi_driver.c | 16 ++++++++++++++-- include/linux/rmi.h | 1 + 3 files changed, 17 insertions(+), 2 deletions(-)