linux-kernel - RE: [PATCH] gpio: dwapb: Add support for 32 interrupts

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:   Thu, 5 Apr 2018 09:42:46 +0000
From:   Phil Edworthy <phil.edworthy@...esas.com>
To:     Andy Shevchenko <andy.shevchenko@...il.com>
CC:     Hoan Tran <hotran@....com>,
        Linus Walleij <linus.walleij@...aro.org>,
        "Rob Herring" <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        "Michel Pollet" <michel.pollet@...renesas.com>,
        "open list:GPIO SUBSYSTEM" <linux-gpio@...r.kernel.org>,
        devicetree <devicetree@...r.kernel.org>,
        Linux-Renesas <linux-renesas-soc@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH] gpio: dwapb: Add support for 32 interrupts

Hi Andy,

On 30 March 2018 22:26 Andy Shevchenko wrote:
> On Wed, Mar 28, 2018 at 5:22 PM, Phil Edworthy wrote:
> > The DesignWare GPIO IP can be configured for either 1 or 32
> > interrupts,
> 
> 1 to 32, or just a choice between two?
Just a choice of 1 or 32.
Note that by 'configured' I am talking about the hardware being configured in
RTL prior to manufacturing a device. Once made, you cannot change it.
This configuration affects the number of output interrupt signals from the GPIO
Controller block that are connected to an interrupt controller.

> > but the driver currently only supports 1 interrupt. See the DesignWare
> > DW_apb_gpio Databook description of the 'GPIO_INTR_IO' parameter.
> 
> Will see after holiday and perhaps make more comments. Here is just a brief
> review.
> 
> > +- interrupts : The interrupts to the parent controller raised when
> > +GPIOs
> > +  generate the interrupts. If the controller provides one combined
> > +interrupt
> > +  for all GPIOs, specify a single interrupt. If the controller
> > +provides one
> > +  interrupt for each GPIO, provide a list of interrupts that
> > +correspond to each
> > +  of the GPIO pins. When specifying multiple interrupts, if any of
> > +the GPIOs are
> > +  not connected to an interrupt, use the interrupt-mask property.
> > +- interrupt-mask : a 32-bit bit mask that specifies which interrupts
> > +in the list
> > +  of interrupts is valid, bit is 1 for a valid irq.
> 
> So, but why one will need that in practice? GPIO driver usually provides a pin
> based IRQ chip which maps each pin to the corresponding offset inside
> specific IRQ domain.
On an ARM device we have this GPIO block connected to the GIC interrupt
controller, i.e. the Synopsys GPIO controller interrupts can* have a 1 to 1
mapping to the GIC interrupts. At the moment, the GPIO driver only allows a
single irq signal to specified.
* this is not strictly accurate on the device I am working on, there is another
block of IP between the two, but that doesn't matter in this case.

> > +                       struct device_node *np = to_of_node(fwnode);
> > +                       u32 irq_mask = 0xFFFFFFFF;
> 
> Why? Shouldn't it be dependent to the amount of actual pins / ports?
> Intel Quark has only 8 AFAIR.
It's just a default which can be overridden via device tree.
For Quark, since you currently only use a single irq, I guess the HW was
configured that way. In which case, you wouldn't use any of this.

> > +                       int j;
> > +
> > +                       /* Optional irq mask */
> > +                       fwnode_property_read_u32(fwnode,
> > + "interrupt-mask", &irq_mask);
> > +
> > +                       /*
> > +                        * The IP has configuration options to allow a single
> > +                        * combined interrupt or one per gpio. If one per gpio,
> > +                        * some might not be used.
> > +                        */
> 
> > +                       for (j = 0; j < pp->ngpio; j++) {
> > +                               if (irq_mask & BIT(j)) {
> 
> for_each_set_bit() is in kernel for ages!
There's lot of stuff in the kernel for ages that I can't remember!
I'll fix this :)

> > +                                       pp->irq[j] = irq_of_parse_and_map(np, j);
> > +                                       if (pp->irq[j])
> > +                                               pp->has_irq = true;
> > +                               }
> > +                       }
> 
> 
> So, on the first glance the patch looks either superfluous or taking wrong
> approach. Please, elaborate more why it's done in this way and what the
> case for all this in practice.
Hopefully I have explained it a bit better above.

Thanks for your comments
Phil