[PATCH] bus: mhi: host: Range check CHDBOFF and ERDBOFF

[PATCH] bus: mhi: host: Range check CHDBOFF and ERDBOFF

[Jeffrey Hugo]

If the value read from the CHDBOFF and ERDBOFF registers is outside the
range of the MHI register space then an invalid address might be computed
which later causes a kernel panic.  Range check the read value to prevent
a crash due to bad data from the device.

Fixes: 6cd330ae76ff ("bus: mhi: core: Add support for ringing channel/event ring doorbells")
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
---
 drivers/bus/mhi/host/init.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/drivers/bus/mhi/host/init.c b/drivers/bus/mhi/host/init.c
index 3d779ee..ae8ebcfb 100644
--- a/drivers/bus/mhi/host/init.c
+++ b/drivers/bus/mhi/host/init.c
@@ -516,6 +516,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
 		return -EIO;
 	}
 
+	if (val >= mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB)) {
+		dev_err(dev, "CHDB offset: 0x%x is out of range: 0x%zx\n",
+			val, mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB));
+		return -EINVAL;
+	}
+
 	/* Setup wake db */
 	mhi_cntrl->wake_db = base + val + (8 * MHI_DEV_WAKE_DB);
 	mhi_cntrl->wake_set = false;
@@ -532,6 +538,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
 		return -EIO;
 	}
 
+	if (val >= mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings)) {
+		dev_err(dev, "ERDB offset: 0x%x is out of range: 0x%zx\n",
+			val, mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings));
+		return -EINVAL;
+	}
+
 	/* Setup event db address for each ev_ring */
 	mhi_event = mhi_cntrl->mhi_event;
 	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, val += 8, mhi_event++) {
-- 
2.7.4

Re: [PATCH] bus: mhi: host: Range check CHDBOFF and ERDBOFF

[Manivannan Sadhasivam]

On Fri, Mar 17, 2023 at 09:14:19AM -0600, Jeffrey Hugo wrote:
> If the value read from the CHDBOFF and ERDBOFF registers is outside the
> range of the MHI register space then an invalid address might be computed
> which later causes a kernel panic.  Range check the read value to prevent
> a crash due to bad data from the device.
> 
> Fixes: 6cd330ae76ff ("bus: mhi: core: Add support for ringing channel/event ring doorbells")

Please CC stable list.

> Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
> Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
> ---
>  drivers/bus/mhi/host/init.c | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/drivers/bus/mhi/host/init.c b/drivers/bus/mhi/host/init.c
> index 3d779ee..ae8ebcfb 100644
> --- a/drivers/bus/mhi/host/init.c
> +++ b/drivers/bus/mhi/host/init.c
> @@ -516,6 +516,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
>  		return -EIO;
>  	}
>  
> +	if (val >= mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB)) {
> +		dev_err(dev, "CHDB offset: 0x%x is out of range: 0x%zx\n",
> +			val, mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB));
> +		return -EINVAL;

Isn't -ERANGE a better errno here and also in other places of the driver?

Thanks,
Mani

> +	}
> +
>  	/* Setup wake db */
>  	mhi_cntrl->wake_db = base + val + (8 * MHI_DEV_WAKE_DB);
>  	mhi_cntrl->wake_set = false;
> @@ -532,6 +538,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
>  		return -EIO;
>  	}
>  
> +	if (val >= mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings)) {
> +		dev_err(dev, "ERDB offset: 0x%x is out of range: 0x%zx\n",
> +			val, mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings));
> +		return -EINVAL;
> +	}
> +
>  	/* Setup event db address for each ev_ring */
>  	mhi_event = mhi_cntrl->mhi_event;
>  	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, val += 8, mhi_event++) {
> -- 
> 2.7.4
> 
> 

-- 
மணிவண்ணன் சதாசிவம்

Re: [PATCH] bus: mhi: host: Range check CHDBOFF and ERDBOFF

[Jeffrey Hugo]

On 3/24/2023 4:34 AM, Manivannan Sadhasivam wrote:
> On Fri, Mar 17, 2023 at 09:14:19AM -0600, Jeffrey Hugo wrote:
>> If the value read from the CHDBOFF and ERDBOFF registers is outside the
>> range of the MHI register space then an invalid address might be computed
>> which later causes a kernel panic.  Range check the read value to prevent
>> a crash due to bad data from the device.
>>
>> Fixes: 6cd330ae76ff ("bus: mhi: core: Add support for ringing channel/event ring doorbells")
> 
> Please CC stable list.

The stable folks have been automatically picking up everything with a 
fixes tag for years.  CCing the stable list does not seem to do anything 
unless the patch needs specific backporting adjustments.

Do you still want an explicit CC?

> 
>> Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
>> Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
>> ---
>>   drivers/bus/mhi/host/init.c | 12 ++++++++++++
>>   1 file changed, 12 insertions(+)
>>
>> diff --git a/drivers/bus/mhi/host/init.c b/drivers/bus/mhi/host/init.c
>> index 3d779ee..ae8ebcfb 100644
>> --- a/drivers/bus/mhi/host/init.c
>> +++ b/drivers/bus/mhi/host/init.c
>> @@ -516,6 +516,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
>>   		return -EIO;
>>   	}
>>   
>> +	if (val >= mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB)) {
>> +		dev_err(dev, "CHDB offset: 0x%x is out of range: 0x%zx\n",
>> +			val, mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB));
>> +		return -EINVAL;
> 
> Isn't -ERANGE a better errno here and also in other places of the driver?

I suppose that could work.  This was modeled after the existing BHI (eg 
BHIOFF) range checks.  You want those updated in the same change or a 
separate one?

> Thanks,
> Mani
> 
>> +	}
>> +
>>   	/* Setup wake db */
>>   	mhi_cntrl->wake_db = base + val + (8 * MHI_DEV_WAKE_DB);
>>   	mhi_cntrl->wake_set = false;
>> @@ -532,6 +538,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
>>   		return -EIO;
>>   	}
>>   
>> +	if (val >= mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings)) {
>> +		dev_err(dev, "ERDB offset: 0x%x is out of range: 0x%zx\n",
>> +			val, mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings));
>> +		return -EINVAL;
>> +	}
>> +
>>   	/* Setup event db address for each ev_ring */
>>   	mhi_event = mhi_cntrl->mhi_event;
>>   	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, val += 8, mhi_event++) {
>> -- 
>> 2.7.4
>>
>>
> 

Re: [PATCH] bus: mhi: host: Range check CHDBOFF and ERDBOFF

[Manivannan Sadhasivam]

On Fri, Mar 24, 2023 at 07:50:01AM -0600, Jeffrey Hugo wrote:
> On 3/24/2023 4:34 AM, Manivannan Sadhasivam wrote:
> > On Fri, Mar 17, 2023 at 09:14:19AM -0600, Jeffrey Hugo wrote:
> > > If the value read from the CHDBOFF and ERDBOFF registers is outside the
> > > range of the MHI register space then an invalid address might be computed
> > > which later causes a kernel panic.  Range check the read value to prevent
> > > a crash due to bad data from the device.
> > > 
> > > Fixes: 6cd330ae76ff ("bus: mhi: core: Add support for ringing channel/event ring doorbells")
> > 
> > Please CC stable list.
> 
> The stable folks have been automatically picking up everything with a fixes
> tag for years.  CCing the stable list does not seem to do anything unless
> the patch needs specific backporting adjustments.
> 
> Do you still want an explicit CC?
> 

That was the impression I had but Greg wanted us to explicitly CC stable list
and add a hint about how deep the patch needs to be applied.

> > 
> > > Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
> > > Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
> > > ---
> > >   drivers/bus/mhi/host/init.c | 12 ++++++++++++
> > >   1 file changed, 12 insertions(+)
> > > 
> > > diff --git a/drivers/bus/mhi/host/init.c b/drivers/bus/mhi/host/init.c
> > > index 3d779ee..ae8ebcfb 100644
> > > --- a/drivers/bus/mhi/host/init.c
> > > +++ b/drivers/bus/mhi/host/init.c
> > > @@ -516,6 +516,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
> > >   		return -EIO;
> > >   	}
> > > +	if (val >= mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB)) {
> > > +		dev_err(dev, "CHDB offset: 0x%x is out of range: 0x%zx\n",
> > > +			val, mhi_cntrl->reg_len - (8 * MHI_DEV_WAKE_DB));
> > > +		return -EINVAL;
> > 
> > Isn't -ERANGE a better errno here and also in other places of the driver?
> 
> I suppose that could work.  This was modeled after the existing BHI (eg
> BHIOFF) range checks.  You want those updated in the same change or a
> separate one?
> 

That can go into a separate change.

Thanks,
Mani

> > Thanks,
> > Mani
> > 
> > > +	}
> > > +
> > >   	/* Setup wake db */
> > >   	mhi_cntrl->wake_db = base + val + (8 * MHI_DEV_WAKE_DB);
> > >   	mhi_cntrl->wake_set = false;
> > > @@ -532,6 +538,12 @@ int mhi_init_mmio(struct mhi_controller *mhi_cntrl)
> > >   		return -EIO;
> > >   	}
> > > +	if (val >= mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings)) {
> > > +		dev_err(dev, "ERDB offset: 0x%x is out of range: 0x%zx\n",
> > > +			val, mhi_cntrl->reg_len - (8 * mhi_cntrl->total_ev_rings));
> > > +		return -EINVAL;
> > > +	}
> > > +
> > >   	/* Setup event db address for each ev_ring */
> > >   	mhi_event = mhi_cntrl->mhi_event;
> > >   	for (i = 0; i < mhi_cntrl->total_ev_rings; i++, val += 8, mhi_event++) {
> > > -- 
> > > 2.7.4
> > > 
> > > 
> > 
> 

-- 
மணிவண்ணன் சதாசிவம்