iio: adc: mt6359: fix unchecked return value in mt6358_read_imp
Subsystems (2)
PATCH 1/1 iio: adc: mt6359: fix unchecked return value in mt6358_read_imp
In mt6358_read_imp(), the return value of regmap_read() is currently ignored. This is problematic because if the bus read fails the variable val_v remains uninitialized.
The function subsequently assigns this uninitialized stack value to *vbat, leading to incorrect measurement results being reported to the IIO subsystem without any error indication.
Update the function to check the return value of regmap_read(). Ensure that mt6358_stop_imp_conv() is still called to clean up the hardware state before returning the error code.
--- drivers/iio/adc/mt6359-auxadc.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/iio/adc/mt6359-auxadc.c +4 -1
@@ -497,10 +497,13 @@ static int mt6358_read_imp(struct mt6359_auxadc *adc_dev,return ret;/* Read the params before stopping */ -regmap_read (regmap, reg_adc0 + (cinfo->imp_adc_num <<1 ), &val_v); + ret =regmap_read (regmap, reg_adc0 + (cinfo->imp_adc_num <<1 ), &val_v);mt6358_stop_imp_conv (adc_dev); +if (ret) +return ret; +if (vbat) *vbat = val_v;if (ibat)
Signature · 1 line
2.43.0
Quoted 11 lines · Salah Triki
In
mt6358_read_imp(), the return value ofregmap_read()is currently ignored. This is problematic because if the bus read fails the variableval_vremains uninitialized.The function subsequently assigns this uninitialized stack value to *vbat, leading to incorrect measurement results being reported to the IIO subsystem without any error indication.
Update the function to check the return value of
regmap_read(). Ensure thatmt6358_stop_imp_conv()is still called to clean up the hardware state before returning the error code.
Sounds like this deserves a Fixes tag, but the problem is that the whole driver is written like this. Why does having this fixed make it special?
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 9 lines · Andy Shevchenko
Quoted 8 lines · Salah Triki
In
mt6358_read_imp(), the return value ofregmap_read()is currently ignored. This is problematic because if the bus read fails the variableval_vremains uninitialized.The function subsequently assigns this uninitialized stack value to *vbat, leading to incorrect measurement results being reported to the IIO subsystem without any error indication.
Quoted 12 lines
Quoted 3 lines
Update the function to check the return value of
regmap_read(). Ensure thatmt6358_stop_imp_conv()is still called to clean up the hardware state before returning the error code.Sounds like this deserves a Fixes tag, but the problem is that the whole driver is written like this. Why does having this fixed make it special?
Signature · 2 lines
With Best Regards, Andy Shevchenko
You are right, Andy. I noticed that several other functions in this driver share the same issue. I will send a V2 that addresses all unchecked regmap_read() calls in this file and I will include the Fixes tag. Thanks for the feedback.
Best Regards
Salah Triki
Jonathan Cameron
maintainerQuoted 23 lines · On Mon, 27 Apr 2026 12:31:34 +0100 Salah Triki <[email protected]> wrote:
Quoted 9 lines · Andy Shevchenko
Quoted 8 lines · Salah Triki
In
mt6358_read_imp(), the return value ofregmap_read()is currently ignored. This is problematic because if the bus read fails the variableval_vremains uninitialized.The function subsequently assigns this uninitialized stack value to *vbat, leading to incorrect measurement results being reported to the IIO subsystem without any error indication.
Quoted 12 lines
Quoted 3 lines
Update the function to check the return value of
regmap_read(). Ensure thatmt6358_stop_imp_conv()is still called to clean up the hardware state before returning the error code.Sounds like this deserves a Fixes tag, but the problem is that the whole driver is written like this. Why does having this fixed make it special?
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 4 lines
You are right, Andy. I noticed that several other functions in this driver share the same issue. I will send a V2 that addresses all unchecked
regmap_read()calls in this file and I will include the Fixes tag. Thanks for the feedback.
What's the bus in this case?
Whilst maybe we should be checking them, it is not uncommon to forgo those checks in internal busses. In general they should be more reliable than something going over PCB tracks like spi or i2c.
I'd be tempted to fix the uninitialized variable by setting it at declaration rather than worry too much about checking all the regmap return values.
Quoted 3 lines
Best Regards
Salah Triki