linux-kernel - Re: [PATCH 1/8] counter/ti-eqep: implement over/underflow events

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <YXpGLMjXImwEVJGY@shinobu>
Date:   Thu, 28 Oct 2021 15:41:48 +0900
From:   William Breathitt Gray <vilhelm.gray@...il.com>
To:     David Lechner <david@...hnology.com>
Cc:     linux-iio@...r.kernel.org, Robert Nelson <robertcnelson@...il.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/8] counter/ti-eqep: implement over/underflow events

On Wed, Oct 27, 2021 at 10:23:13AM -0500, David Lechner wrote:
> On 10/25/21 2:13 AM, William Breathitt Gray wrote:
> > On Sat, Oct 16, 2021 at 08:33:36PM -0500, David Lechner wrote:
> >> @@ -260,11 +333,17 @@ static int ti_eqep_position_ceiling_write(struct counter_device *counter,
> >>   					  u64 ceiling)
> >>   {
> >>   	struct ti_eqep_cnt *priv = counter->priv;
> >> +	u32 qposmax = ceiling;
> >>   
> >> -	if (ceiling != (u32)ceiling)
> >> +	/* ensure that value fits in 32-bit register */
> >> +	if (qposmax != ceiling)
> >>   		return -ERANGE;
> >>   
> >> -	regmap_write(priv->regmap32, QPOSMAX, ceiling);
> >> +	/* protect against infinite overflow interrupts */
> >> +	if (qposmax == 0)
> >> +		return -EINVAL;
> > 
> > Would you be able to explain this scenario a bit further? My expectation
> > would be that an overflow event would only occur if the position
> > increased past the ceiling (i.e. increased to greater than 0). Of
> > course, running the device with a ceiling of 0 effectively guarantees
> > overflow eventss with every movement, but I would expect a stationary
> > device to sit with a position of 0 and thus no overflow events.
> > 
> 
> This is just the way the hardware works. I discovered this the first
> time I enabled interrupts. Even if you clear the interrupt, it is
> triggered again immediately when QPOSMAX == 0.

For this device, does an overflow event occur once the count value
increases to equal the ceiling value, or once the count value increases
past the ceiling value?

The Counter interface defines ceiling as an inclusive upper limit (count
value is capable of reaching ceiling) and defines COUNTER_EVENT_OVERFLOW
as occuring when the count value increases past ceiling. I want to make
sure the ceiling extension and COUNTER_EVENT_OVERFLOW events for this
driver are behaving as expected of the Counter interface.

Let's use a non-zero example to be clear. Suppose we set ceiling equal
to 10. If count is currently at 9 and increases by 1, count should
become 10 and no COUNTER_EVENT_OVERFLOW event is expected to trigger; if
count is 10 and further increases, count should _not_ become 11 (staying
at 10 or starting over at the floor) but a COUNTER_EVENT_OVERFLOW event
does trigger. Does the driver behave like this currently?

William Breathitt Gray

Download attachment "signature.asc" of type "application/pgp-signature" (834 bytes)