linux-spi.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Robin Murphy <robin.murphy@arm.com>
To: Florian Fainelli <f.fainelli@gmail.com>, linux-kernel@vger.kernel.org
Cc: "moderated list:BROADCOM BCM2711/BCM2835 ARM ARCHITECTURE" 
	<linux-arm-kernel@lists.infradead.org>,
	"open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
	<devicetree@vger.kernel.org>,
	Scott Branden <sbranden@broadcom.com>,
	lukas@wunner.de, Ray Jui <rjui@broadcom.com>,
	Rob Herring <robh+dt@kernel.org>,
	"open list:SPI SUBSYSTEM" <linux-spi@vger.kernel.org>,
	Mark Brown <broonie@kernel.org>,
	"maintainer:BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITE..." 
	<bcm-kernel-feedback-list@broadcom.com>,
	"moderated list:BROADCOM BCM2711/BCM2835 ARM ARCHITECTURE" 
	<linux-rpi-kernel@lists.infradead.org>,
	Martin Sperl <kernel@martin.sperl.org>,
	Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Subject: Re: [PATCH v2] spi: bcm2835: Enable shared interrupt support
Date: Fri, 5 Jun 2020 12:34:36 +0100	[thread overview]
Message-ID: <142d48ae-2725-1368-3e11-658449662371@arm.com> (raw)
In-Reply-To: <20200604212819.715-1-f.fainelli@gmail.com>

On 2020-06-04 22:28, Florian Fainelli wrote:
> The 4 SPI controller instances added in BCM2711 and BCM7211 SoCs (SPI3,
> SPI4, SPI5 and SPI6) share the same interrupt line with SPI0.
> 
> For the BCM2835 case which is deemed performance critical, we would like
> to continue using an interrupt handler which does not have the extra
> comparison on BCM2835_SPI_CS_INTR.

FWIW, if I'm reading the patch correctly, then with sensible codegen 
that "overhead" should amount to a bit test on a live register plus a 
not-taken conditional branch - according to the 1176 TRM that should add 
up to a whopping 2 cycles. If that's really significant then I'd have to 
wonder whether you want to be at the mercy of the whole generic IRQ 
stack at all, and should perhaps consider using FIQ instead.

> To support that requirement the common interrupt handling code between
> the shared and non-shared interrupt paths is split into a
> bcm2835_spi_interrupt_common() and both bcm2835_spi_interrupt() as well
> as bcm2835_spi_shared_interrupt() make use of it.
> 
> During probe, we determine if there is at least another instance of this
> SPI controller, and if there is, then we install a shared interrupt
> handler.
> 
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
> Changes in v2:
> 
> - identify other available SPI nodes to determine if we need to set-up
>    interrupt sharing. This needs to happen for the very first instance
>    since we cannot know for the first instance whether interrupt sharing
>    is needed or not.
> 
>   drivers/spi/spi-bcm2835.c | 61 ++++++++++++++++++++++++++++++++-------
>   1 file changed, 50 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/spi/spi-bcm2835.c b/drivers/spi/spi-bcm2835.c
> index 237bd306c268..0288b5b3de1e 100644
> --- a/drivers/spi/spi-bcm2835.c
> +++ b/drivers/spi/spi-bcm2835.c
> @@ -361,11 +361,10 @@ static void bcm2835_spi_reset_hw(struct spi_controller *ctlr)
>   	bcm2835_wr(bs, BCM2835_SPI_DLEN, 0);
>   }
>   
> -static irqreturn_t bcm2835_spi_interrupt(int irq, void *dev_id)
> +static inline irqreturn_t bcm2835_spi_interrupt_common(struct spi_controller *ctlr,
> +						       u32 cs)
>   {
> -	struct spi_controller *ctlr = dev_id;
>   	struct bcm2835_spi *bs = spi_controller_get_devdata(ctlr);
> -	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
>   
>   	/*
>   	 * An interrupt is signaled either if DONE is set (TX FIFO empty)
> @@ -394,6 +393,27 @@ static irqreturn_t bcm2835_spi_interrupt(int irq, void *dev_id)
>   	return IRQ_HANDLED;
>   }
>   
> +static irqreturn_t bcm2835_spi_interrupt(int irq, void *dev_id)
> +{
> +	struct spi_controller *ctlr = dev_id;
> +	struct bcm2835_spi *bs = spi_controller_get_devdata(ctlr);
> +	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
> +
> +	return bcm2835_spi_interrupt_common(ctlr, cs);
> +}
> +
> +static irqreturn_t bcm2835_spi_shared_interrupt(int irq, void *dev_id)
> +{
> +	struct spi_controller *ctlr = dev_id;
> +	struct bcm2835_spi *bs = spi_controller_get_devdata(ctlr);
> +	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
> +
> +	if (!(cs & BCM2835_SPI_CS_INTR))
> +		return IRQ_NONE;
> +
> +	return bcm2835_spi_interrupt_common(ctlr, cs);
> +}
> +
>   static int bcm2835_spi_transfer_one_irq(struct spi_controller *ctlr,
>   					struct spi_device *spi,
>   					struct spi_transfer *tfr,
> @@ -1287,12 +1307,37 @@ static int bcm2835_spi_setup(struct spi_device *spi)
>   	return 0;
>   }
>   
> +static const struct of_device_id bcm2835_spi_match[] = {
> +	{ .compatible = "brcm,bcm2835-spi", },
> +	{}
> +};
> +MODULE_DEVICE_TABLE(of, bcm2835_spi_match);
> +
>   static int bcm2835_spi_probe(struct platform_device *pdev)
>   {
> +	irq_handler_t bcm2835_spi_isr_func = bcm2835_spi_interrupt;
>   	struct spi_controller *ctlr;
> +	unsigned long flags = 0;
> +	struct device_node *dn;
>   	struct bcm2835_spi *bs;
>   	int err;
>   
> +	/* On BCM2711 there can be multiple SPI controllers enabled sharing the
> +	 * same interrupt line, but we also want to minimize the overhead if
> +	 * there is no need to support interrupt sharing. If we find at least
> +	 * another available instane (not counting the one we are probed from),
> +	 * then we assume that interrupt sharing is necessary.
> +	 */
> +	for_each_compatible_node(dn, NULL, bcm2835_spi_match[0].compatible) {
> +		err = of_device_is_available(dn) && dn != pdev->dev.of_node;
> +		of_node_put(dn);

This is in the wrong place - it should only be where you terminate the 
loop early and thus bypass the "of_node_put(from)" call in the iterator 
itself.

> +		if (err) {
> +			flags = IRQF_SHARED;

Is there really any harm to setting IRQF_SHARED even when the interrupt 
isn't shared in hardware? Sure, it means you lose a degree of API-level 
validation and some other driver might also be able to simultaneously 
claim it on bcm283x, but in that case the DT would be wrong and 
*something* isn't going to work correctly anyway, so does this one 
driver really need to care about trying to be the DT police?

Robin.

> +			bcm2835_spi_isr_func = bcm2835_spi_shared_interrupt;
> +			break;
> +		}
> +	}
> +
>   	ctlr = spi_alloc_master(&pdev->dev, ALIGN(sizeof(*bs),
>   						  dma_get_cache_alignment()));
>   	if (!ctlr)
> @@ -1344,8 +1389,8 @@ static int bcm2835_spi_probe(struct platform_device *pdev)
>   	bcm2835_wr(bs, BCM2835_SPI_CS,
>   		   BCM2835_SPI_CS_CLEAR_RX | BCM2835_SPI_CS_CLEAR_TX);
>   
> -	err = devm_request_irq(&pdev->dev, bs->irq, bcm2835_spi_interrupt, 0,
> -			       dev_name(&pdev->dev), ctlr);
> +	err = devm_request_irq(&pdev->dev, bs->irq, bcm2835_spi_isr_func,
> +			       flags, dev_name(&pdev->dev), ctlr);
>   	if (err) {
>   		dev_err(&pdev->dev, "could not request IRQ: %d\n", err);
>   		goto out_dma_release;
> @@ -1400,12 +1445,6 @@ static void bcm2835_spi_shutdown(struct platform_device *pdev)
>   		dev_err(&pdev->dev, "failed to shutdown\n");
>   }
>   
> -static const struct of_device_id bcm2835_spi_match[] = {
> -	{ .compatible = "brcm,bcm2835-spi", },
> -	{}
> -};
> -MODULE_DEVICE_TABLE(of, bcm2835_spi_match);
> -
>   static struct platform_driver bcm2835_spi_driver = {
>   	.driver		= {
>   		.name		= DRV_NAME,
> 

  parent reply	other threads:[~2020-06-05 11:34 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-04 21:28 [PATCH v2] spi: bcm2835: Enable shared interrupt support Florian Fainelli
2020-06-05  8:46 ` Nicolas Saenz Julienne
2020-06-05 10:52   ` Mark Brown
2020-06-05 10:58   ` Nicolas Saenz Julienne
2020-06-05 10:51 ` Lukas Wunner
2020-06-05 11:14 ` Mark Brown
2020-06-05 12:20   ` Mark Brown
2020-06-05 11:34 ` Robin Murphy [this message]
2020-06-05 13:20   ` Mark Brown
2020-06-05 13:46     ` Robin Murphy
2020-06-05 14:41       ` Robin Murphy
2020-06-05 15:27         ` Mark Brown
2020-06-05 22:04         ` Florian Fainelli
2020-06-08 11:11           ` Robin Murphy
2020-06-08 11:28             ` Mark Brown
2020-06-15 16:34               ` Florian Fainelli
2020-06-15 17:00                 ` Mark Brown
2020-06-15 17:04                   ` Florian Fainelli
2020-06-15 17:30                     ` Mark Brown
2020-06-15 17:31                     ` Robin Murphy
2020-06-15 19:26                       ` Mark Brown
2020-06-08 11:41             ` Lukas Wunner
2020-06-15 19:09               ` Robin Murphy
2020-06-15 19:42                 ` Florian Fainelli
2020-06-15 20:48                   ` Mark Brown
2022-07-19 10:53 [PATCH v2] spi: bcm2835: enable " Marc Kleine-Budde
2022-07-26 11:17 ` Mark Brown

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=142d48ae-2725-1368-3e11-658449662371@arm.com \
    --to=robin.murphy@arm.com \
    --cc=bcm-kernel-feedback-list@broadcom.com \
    --cc=broonie@kernel.org \
    --cc=devicetree@vger.kernel.org \
    --cc=f.fainelli@gmail.com \
    --cc=kernel@martin.sperl.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rpi-kernel@lists.infradead.org \
    --cc=linux-spi@vger.kernel.org \
    --cc=lukas@wunner.de \
    --cc=nsaenzjulienne@suse.de \
    --cc=rjui@broadcom.com \
    --cc=robh+dt@kernel.org \
    --cc=sbranden@broadcom.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).