| 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
| ||
|
Message-Id: <20250108120906.92cda1fba339b072aceec039@hugovil.com>
Date: Wed, 8 Jan 2025 12:09:06 -0500
From: Hugo Villeneuve <hugo@...ovil.com>
To: Andre Werner <andre.werner@...tec-electronic.com>
Cc: gregkh@...uxfoundation.org, jirislaby@...nel.org,
hvilleneuve@...onoff.com, andy@...nel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-serial@...r.kernel.org,
lech.perczak@...lingroup.com, krzk+dt@...nel.org, conor+dt@...nel.org,
robh@...nel.org
Subject: Re: [PATCH v3 2/2] serial: sc16is7xx: Add polling feature if no IRQ
usage possible
On Tue, 7 Jan 2025 15:29:47 +0100
Andre Werner <andre.werner@...tec-electronic.com> wrote:
Hi Andre,
> Fall back to polling mode if no interrupt is configured because not
> possible. If "interrupts" property is missing in devicetree the driver
> uses a delayed worker to pull state of interrupt status registers.
"not possible" is confusing. Maybe:
"Fall back to polling mode if interrupt pin is not connected
or available."
You also could change the commit title in a similar way (Add polling
feature if no IRQ pin is available).
I would also replace "feature" with "mode" for consistency with your
other comments below.
>
> Signed-off-by: Andre Werner <andre.werner@...tec-electronic.com>
> ---
> V2:
> - Change warning for polling mode to debug log entry
> - Correct typo: Resuse -> Reuse
> - Format define with missing tabs for SC16IS7XX_POLL_PERIOD
> - Format struct declaration sc16is7xx_one_config with missing tabs for polling and shutdown
> - Adapt dtbinding with new polling feature
> V3:
> - Use suffix with units and drop a comment SC16IS7XX_POLL_PERIOD_MS. Sorry for that miss.
> - Make Kernel lowercase.
> ---
> ---
> drivers/tty/serial/sc16is7xx.c | 40 ++++++++++++++++++++++++++++++++++
> 1 file changed, 40 insertions(+)
>
> diff --git a/drivers/tty/serial/sc16is7xx.c b/drivers/tty/serial/sc16is7xx.c
> index a3093e09309f..57be5d55205a 100644
> --- a/drivers/tty/serial/sc16is7xx.c
> +++ b/drivers/tty/serial/sc16is7xx.c
> @@ -314,6 +314,7 @@
> #define SC16IS7XX_FIFO_SIZE (64)
> #define SC16IS7XX_GPIOS_PER_BANK 4
>
> +#define SC16IS7XX_POLL_PERIOD_MS 10
> #define SC16IS7XX_RECONF_MD BIT(0)
> #define SC16IS7XX_RECONF_IER BIT(1)
> #define SC16IS7XX_RECONF_RS485 BIT(2)
> @@ -348,6 +349,9 @@ struct sc16is7xx_port {
> u8 mctrl_mask;
> struct kthread_worker kworker;
> struct task_struct *kworker_task;
> + struct kthread_delayed_work poll_work;
> + bool polling;
> + bool shutdown;
> struct sc16is7xx_one p[];
> };
>
> @@ -861,6 +865,19 @@ static irqreturn_t sc16is7xx_irq(int irq, void *dev_id)
> return IRQ_HANDLED;
> }
>
> +static void sc16is7xx_transmission_poll(struct kthread_work *work)
Maybe drop "transmission", as this is used for tx/rx?
And other kthread work functions use _proc suffix, so rename to follow
the same convention -> sc16is7xx_poll_proc()
They also use "wk" as argument name, so also rename it for consistency.
> +{
> + struct sc16is7xx_port *s = container_of(work, struct sc16is7xx_port, poll_work.work);
> +
> + /* Reuse standard IRQ handler. Interrupt ID is unused in this context. */
> + sc16is7xx_irq(0, s);
> +
> + /* Setup delay based on SC16IS7XX_POLL_PERIOD_MS */
> + if (!s->shutdown)
> + kthread_queue_delayed_work(&s->kworker, &s->poll_work,
> + msecs_to_jiffies(SC16IS7XX_POLL_PERIOD_MS));
> +}
> +
> static void sc16is7xx_tx_proc(struct kthread_work *ws)
> {
> struct uart_port *port = &(to_sc16is7xx_one(ws, tx_work)->port);
> @@ -1149,6 +1166,7 @@ static int sc16is7xx_config_rs485(struct uart_port *port, struct ktermios *termi
> static int sc16is7xx_startup(struct uart_port *port)
> {
> struct sc16is7xx_one *one = to_sc16is7xx_one(port, port);
> + struct sc16is7xx_port *s = dev_get_drvdata(port->dev);
> unsigned int val;
> unsigned long flags;
>
> @@ -1210,6 +1228,11 @@ static int sc16is7xx_startup(struct uart_port *port)
> uart_port_lock_irqsave(port, &flags);
> sc16is7xx_enable_ms(port);
> uart_port_unlock_irqrestore(port, flags);
Insert blank line
> + if (s->polling) {
> + s->shutdown = false;
> + kthread_queue_delayed_work(&s->kworker, &s->poll_work,
> + msecs_to_jiffies(SC16IS7XX_POLL_PERIOD_MS));
> + }
>
> return 0;
> }
> @@ -1232,6 +1255,10 @@ static void sc16is7xx_shutdown(struct uart_port *port)
>
> sc16is7xx_power(port, 0);
>
> + if (s->polling) {
> + s->shutdown = true;
Can we avoid using this shutdown variable?
>From what I see, kthread_cancel_delayed_work_sync()
itself will call kthread_cancel_work_sync(), and doc for it states that:
"This function can be used even if the work re-queues itself"
> + kthread_cancel_delayed_work_sync(&s->poll_work);
> + }
Blank line
> kthread_flush_worker(&s->kworker);
> }
>
> @@ -1537,7 +1564,13 @@ int sc16is7xx_probe(struct device *dev, const struct sc16is7xx_devtype *devtype,
>
> /* Always ask for fixed clock rate from a property. */
> device_property_read_u32(dev, "clock-frequency", &uartclk);
Blank line
> + s->polling = !device_property_present(dev, "interrupts");
>
Remove blank line
> + if (s->polling) {
> + dev_dbg(dev,
> + "No interrupt definition found. Falling back to polling mode.\n");
Maybe: "No interrupt pin definition, falling back to polling mode\n"
(add pin and gets rid of trailing point)
> + irq = 0;
> + }
Blank line
> s->clk = devm_clk_get_optional(dev, NULL);
> if (IS_ERR(s->clk))
> return PTR_ERR(s->clk);
> @@ -1664,6 +1697,11 @@ int sc16is7xx_probe(struct device *dev, const struct sc16is7xx_devtype *devtype,
> if (ret)
> goto out_ports;
> #endif
Blank line
> + if (s->polling) {
> + /* Initialize kernel thread for polling */
> + kthread_init_delayed_work(&s->poll_work, sc16is7xx_transmission_poll);
> + return 0;
> + }
>
> /*
> * Setup interrupt. We first try to acquire the IRQ line as level IRQ.
> @@ -1724,6 +1762,8 @@ void sc16is7xx_remove(struct device *dev)
> sc16is7xx_power(&s->p[i].port, 0);
> }
>
> + if (s->polling)
> + kthread_cancel_delayed_work_sync(&s->poll_work);
Blank line
> kthread_flush_worker(&s->kworker);
> kthread_stop(s->kworker_task);
>
> --
> 2.47.1
>
>
>
--
Hugo Villeneuve
Powered by blists - more mailing lists