Re: [PATCH 2/2] drm/xe/hwmon: Fix static analysis tool reported issues

[Jani Nikula <jani.nikula@linux.intel.com>]

On Fri, 22 Mar 2024, "Poosa, Karthik" <karthik.poosa@intel.com> wrote:
> Please find replies inline.
>
> On 21-03-2024 18:30, Jani Nikula wrote:
>> On Thu, 21 Mar 2024, Karthik Poosa <karthik.poosa@intel.com> wrote:
>>> Update xe hwmon with fixes for issues reported by static analysis
>>> tool.
>>> Fix integer overflow with upcasting.
>>> Initialize uninitialized variables.
>>>
>>> Fixes: 4446fcf220ce ("drm/xe/hwmon: Expose power1_max_interval")
>>> Signed-off-by: Karthik Poosa <karthik.poosa@intel.com>
>>> ---
>>>   drivers/gpu/drm/xe/xe_hwmon.c | 18 +++++++++---------
>>>   1 file changed, 9 insertions(+), 9 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/xe/xe_hwmon.c b/drivers/gpu/drm/xe/xe_hwmon.c
>>> index a256af8c2012..6ed9d5c4f6b1 100644
>>> --- a/drivers/gpu/drm/xe/xe_hwmon.c
>>> +++ b/drivers/gpu/drm/xe/xe_hwmon.c
>>> @@ -153,7 +153,7 @@ static void xe_hwmon_process_reg(struct xe_hwmon *hwmon, enum xe_hwmon_reg hwmon
>>>    */
>>>   static void xe_hwmon_power_max_read(struct xe_hwmon *hwmon, long *value)
>>>   {
>>> -	u64 reg_val, min, max;
>>> +	u64 reg_val = 0, min, max;
>> Nah. Just fix xe_hwmon_process_reg() to set *value = 0.
> I think the caller has to ensure value is set to 0, it is not needed in 
> xe_hwmon_process_reg.

If the caller has to set the value to 0, it's a poorly defined
interface. Since xe_hwmon_process_reg() doesn't return a value
indicating an error, it better set the value to 0.

>> Side note, xe_hwmon_get_reg() should return struct xe_reg instead of
>> u32, and xe_hwmon_process_reg() has no business looking into the guts of
>> struct xe_reg.
>>
>>
>> BR,
>> Jani.
>>
> That change is not related to this patch series. It will be a major 
> change which will be handled in another patch series.

Not related to this series, thus a "side note". But the change itself is
trivial.

BR,
Jani.



>
>>>   
>>>   	mutex_lock(&hwmon->hwmon_lock);
>>>   
>>> @@ -182,7 +182,7 @@ static void xe_hwmon_power_max_read(struct xe_hwmon *hwmon, long *value)
>>>   static int xe_hwmon_power_max_write(struct xe_hwmon *hwmon, long value)
>>>   {
>>>   	int ret = 0;
>>> -	u64 reg_val;
>>> +	u64 reg_val = 0;
>>>   
>>>   	mutex_lock(&hwmon->hwmon_lock);
>>>   
>>> @@ -212,7 +212,7 @@ static int xe_hwmon_power_max_write(struct xe_hwmon *hwmon, long value)
>>>   
>>>   static void xe_hwmon_power_rated_max_read(struct xe_hwmon *hwmon, long *value)
>>>   {
>>> -	u64 reg_val;
>>> +	u64 reg_val = 0;
>>>   
>>>   	xe_hwmon_process_reg(hwmon, REG_PKG_POWER_SKU, REG_READ32, &reg_val, 0, 0);
>>>   	reg_val = REG_FIELD_GET(PKG_TDP, reg_val);
>>> @@ -243,7 +243,7 @@ static void
>>>   xe_hwmon_energy_get(struct xe_hwmon *hwmon, long *energy)
>>>   {
>>>   	struct xe_hwmon_energy_info *ei = &hwmon->ei;
>>> -	u64 reg_val;
>>> +	u64 reg_val = 0;
>>>   
>>>   	xe_hwmon_process_reg(hwmon, REG_PKG_ENERGY_STATUS, REG_READ32,
>>>   			     &reg_val, 0, 0);
>>> @@ -264,8 +264,8 @@ xe_hwmon_power1_max_interval_show(struct device *dev, struct device_attribute *a
>>>   				  char *buf)
>>>   {
>>>   	struct xe_hwmon *hwmon = dev_get_drvdata(dev);
>>> -	u32 x, y, x_w = 2; /* 2 bits */
>>> -	u64 r, tau4, out;
>>> +	u32 x = 0, y = 0, x_w = 2; /* 2 bits */
>>> +	u64 r = 0, tau4, out;
>>>   
>>>   	xe_pm_runtime_get(gt_to_xe(hwmon->gt));
>>>   
>>> @@ -291,7 +291,7 @@ xe_hwmon_power1_max_interval_show(struct device *dev, struct device_attribute *a
>>>   	 * As y can be < 2, we compute tau4 = (4 | x) << y
>>>   	 * and then add 2 when doing the final right shift to account for units
>>>   	 */
>>> -	tau4 = ((1 << x_w) | x) << y;
>>> +	tau4 = (u64)((1 << x_w) | x) << y;
>>>   
>>>   	/* val in hwmon interface units (millisec) */
>>>   	out = mul_u64_u32_shr(tau4, SF_TIME, hwmon->scl_shift_time + x_w);
>>> @@ -331,7 +331,7 @@ xe_hwmon_power1_max_interval_store(struct device *dev, struct device_attribute *
>>>   	r = FIELD_PREP(PKG_MAX_WIN, PKG_MAX_WIN_DEFAULT);
>>>   	x = REG_FIELD_GET(PKG_MAX_WIN_X, r);
>>>   	y = REG_FIELD_GET(PKG_MAX_WIN_Y, r);
>>> -	tau4 = ((1 << x_w) | x) << y;
>>> +	tau4 = (u64)((1 << x_w) | x) << y;
>>>   	max_win = mul_u64_u32_shr(tau4, SF_TIME, hwmon->scl_shift_time + x_w);
>>>   
>>>   	if (val > max_win)
>>> @@ -466,7 +466,7 @@ static int xe_hwmon_power_curr_crit_write(struct xe_hwmon *hwmon, long value, u3
>>>   
>>>   static void xe_hwmon_get_voltage(struct xe_hwmon *hwmon, long *value)
>>>   {
>>> -	u64 reg_val;
>>> +	u64 reg_val = 0;
>>>   
>>>   	xe_hwmon_process_reg(hwmon, REG_GT_PERF_STATUS,
>>>   			     REG_READ32, &reg_val, 0, 0);

-- 
Jani Nikula, Intel