| Message ID | 20240506150830.23709-10-johan+linaro@kernel.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | arm64: dts: qcom: sc8280xp-x13s: enable pm8008 camera pmic | expand |
On 06/05/2024 16:08, Johan Hovold wrote: > The SPMI GPIO driver assumes that the parent device is an SPMI device > and accesses random data when backcasting the parent struct device > pointer for non-SPMI devices. > > Fortunately this does not seem to cause any issues currently when the > parent device is an I2C client like the PM8008, but this could change if > the structures are reorganised (e.g. using structure randomisation). > > Notably the interrupt implementation is also broken for non-SPMI devices. > > Also note that the two GPIO pins on PM8008 are used for interrupts and > reset so their practical use should be limited. > > Drop the broken GPIO support for PM8008 for now. > > Fixes: ea119e5a482a ("pinctrl: qcom-pmic-gpio: Add support for pm8008") > Cc: stable@vger.kernel.org # 5.13 > Signed-off-by: Johan Hovold <johan+linaro@kernel.org> > --- > drivers/pinctrl/qcom/pinctrl-spmi-gpio.c | 1 - > 1 file changed, 1 deletion(-) > > diff --git a/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c b/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c > index f4e2c88a7c82..e61be7d05494 100644 > --- a/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c > +++ b/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c > @@ -1206,7 +1206,6 @@ static const struct of_device_id pmic_gpio_of_match[] = { > { .compatible = "qcom,pm7325-gpio", .data = (void *) 10 }, > { .compatible = "qcom,pm7550ba-gpio", .data = (void *) 8}, > { .compatible = "qcom,pm8005-gpio", .data = (void *) 4 }, > - { .compatible = "qcom,pm8008-gpio", .data = (void *) 2 }, > { .compatible = "qcom,pm8019-gpio", .data = (void *) 6 }, > /* pm8150 has 10 GPIOs with holes on 2, 5, 7 and 8 */ > { .compatible = "qcom,pm8150-gpio", .data = (void *) 10 }, Since there are no upstream dtsi users there's no harm in dropping and re-adding when i2c can be taken account of. Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Quoting Johan Hovold (2024-05-06 08:08:26) > The SPMI GPIO driver assumes that the parent device is an SPMI device > and accesses random data when backcasting the parent struct device > pointer for non-SPMI devices. > > Fortunately this does not seem to cause any issues currently when the > parent device is an I2C client like the PM8008, but this could change if > the structures are reorganised (e.g. using structure randomisation). > > Notably the interrupt implementation is also broken for non-SPMI devices. > > Also note that the two GPIO pins on PM8008 are used for interrupts and > reset so their practical use should be limited. > > Drop the broken GPIO support for PM8008 for now. > > Fixes: ea119e5a482a ("pinctrl: qcom-pmic-gpio: Add support for pm8008") > Cc: stable@vger.kernel.org # 5.13 > Signed-off-by: Johan Hovold <johan+linaro@kernel.org> > --- Reviewed-by: Stephen Boyd <swboyd@chromium.org>
On Mon, May 6, 2024 at 5:10 PM Johan Hovold <johan+linaro@kernel.org> wrote: > The SPMI GPIO driver assumes that the parent device is an SPMI device > and accesses random data when backcasting the parent struct device > pointer for non-SPMI devices. > > Fortunately this does not seem to cause any issues currently when the > parent device is an I2C client like the PM8008, but this could change if > the structures are reorganised (e.g. using structure randomisation). > > Notably the interrupt implementation is also broken for non-SPMI devices. > > Also note that the two GPIO pins on PM8008 are used for interrupts and > reset so their practical use should be limited. > > Drop the broken GPIO support for PM8008 for now. > > Fixes: ea119e5a482a ("pinctrl: qcom-pmic-gpio: Add support for pm8008") > Cc: stable@vger.kernel.org # 5.13 > Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Is this something I can just apply, maybe with the DT binding drop patch right (8/13) after it? IIUC it does not need to go into fixes because there are no regressions, right? Yours, Linus Walleij
On Mon, May 27, 2024 at 03:35:41PM +0200, Linus Walleij wrote: > On Mon, May 6, 2024 at 5:10 PM Johan Hovold <johan+linaro@kernel.org> wrote: > > > The SPMI GPIO driver assumes that the parent device is an SPMI device > > and accesses random data when backcasting the parent struct device > > pointer for non-SPMI devices. > > > > Fortunately this does not seem to cause any issues currently when the > > parent device is an I2C client like the PM8008, but this could change if > > the structures are reorganised (e.g. using structure randomisation). > > > > Notably the interrupt implementation is also broken for non-SPMI devices. > > > > Also note that the two GPIO pins on PM8008 are used for interrupts and > > reset so their practical use should be limited. > > > > Drop the broken GPIO support for PM8008 for now. > > > > Fixes: ea119e5a482a ("pinctrl: qcom-pmic-gpio: Add support for pm8008") > > Cc: stable@vger.kernel.org # 5.13 > > Signed-off-by: Johan Hovold <johan+linaro@kernel.org> > > Is this something I can just apply, maybe with the DT binding drop > patch right (8/13) after it? Yes, I guess so, unless it's easier to let everything go through MFD (except possibly the regulator driver). I'll be posting a v2 in a bit and include these two there too. You can either pick them there or ack them as you prefer. > IIUC it does not need to go into fixes because there are no regressions, > right? As I mentioned in the commit message, the driver is backcasting a pointer to an incorrect type, which could lead to all sorts of trouble even if it does not seem to be the case currently (I did not check 5.13 for example). Since it has always been broken, I'd rather err on the safe side and just drop it also from the stable trees. Johan
diff --git a/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c b/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c index f4e2c88a7c82..e61be7d05494 100644 --- a/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c +++ b/drivers/pinctrl/qcom/pinctrl-spmi-gpio.c @@ -1206,7 +1206,6 @@ static const struct of_device_id pmic_gpio_of_match[] = { { .compatible = "qcom,pm7325-gpio", .data = (void *) 10 }, { .compatible = "qcom,pm7550ba-gpio", .data = (void *) 8}, { .compatible = "qcom,pm8005-gpio", .data = (void *) 4 }, - { .compatible = "qcom,pm8008-gpio", .data = (void *) 2 }, { .compatible = "qcom,pm8019-gpio", .data = (void *) 6 }, /* pm8150 has 10 GPIOs with holes on 2, 5, 7 and 8 */ { .compatible = "qcom,pm8150-gpio", .data = (void *) 10 },
The SPMI GPIO driver assumes that the parent device is an SPMI device and accesses random data when backcasting the parent struct device pointer for non-SPMI devices. Fortunately this does not seem to cause any issues currently when the parent device is an I2C client like the PM8008, but this could change if the structures are reorganised (e.g. using structure randomisation). Notably the interrupt implementation is also broken for non-SPMI devices. Also note that the two GPIO pins on PM8008 are used for interrupts and reset so their practical use should be limited. Drop the broken GPIO support for PM8008 for now. Fixes: ea119e5a482a ("pinctrl: qcom-pmic-gpio: Add support for pm8008") Cc: stable@vger.kernel.org # 5.13 Signed-off-by: Johan Hovold <johan+linaro@kernel.org> --- drivers/pinctrl/qcom/pinctrl-spmi-gpio.c | 1 - 1 file changed, 1 deletion(-)