mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Jon Hunter <jonathanh@nvidia.com>
To: Vishwaroop A <va@nvidia.com>,
	Thierry Reding <thierry.reding@kernel.org>,
	Mark Brown <broonie@kernel.org>
Cc: Laxman Dewangan <ldewangan@nvidia.com>,
	Sowjanya Komatineni <skomatineni@nvidia.com>,
	Breno Leitao <leitao@debian.org>,
	Suresh Mangipudi <smangipudi@nvidia.com>,
	Krishna Yarlagadda <kyarlagadda@nvidia.com>,
	linux-tegra@vger.kernel.org, linux-spi@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/3] spi: tegra210-quad: Convert to hard IRQ with high-priority workqueue
Date: Tue, 19 May 2026 10:15:24 +0100	[thread overview]
Message-ID: <d058322b-1239-4429-97eb-ab49002881d7@nvidia.com> (raw)
In-Reply-To: <20260518160739.3286438-2-va@nvidia.com>



On 18/05/2026 17:07, Vishwaroop A wrote:
> Threaded IRQ handlers suffer from scheduler latency on heavily loaded
> systems, causing false transfer timeouts. Convert to hard IRQ handler
> that schedules work on a high-priority unbound workqueue.
> 
> The hard IRQ handler verifies the interrupt, caches FIFO status,
> clears and masks interrupts, then schedules bottom-half processing.
> The workqueue handler runs in process context (can sleep for DMA)
> and can execute on any CPU, avoiding CPU0 bottlenecks.
> 
> Signed-off-by: Vishwaroop A <va@nvidia.com>
> ---
>   drivers/spi/spi-tegra210-quad.c | 120 ++++++++++++++++++++++----------
>   1 file changed, 83 insertions(+), 37 deletions(-)

...

>   static struct tegra_qspi_soc_data tegra210_qspi_soc_data = {
> @@ -1793,12 +1826,21 @@ static int tegra_qspi_probe(struct platform_device *pdev)
>   
>   	pm_runtime_put_autosuspend(&pdev->dev);
>   
> -	ret = request_threaded_irq(tqspi->irq, NULL,
> -				   tegra_qspi_isr_thread, IRQF_ONESHOT,
> -				   dev_name(&pdev->dev), tqspi);
> +	tqspi->wq = alloc_workqueue("%s", WQ_HIGHPRI | WQ_UNBOUND, 0,
> +				    dev_name(&pdev->dev));

Why not used devm_alloc_workqueue() here?

> +	if (!tqspi->wq) {
> +		dev_err(&pdev->dev, "failed to allocate workqueue\n");
> +		ret = -ENOMEM;
> +		goto exit_pm_disable;
> +	}
> +
> +	INIT_WORK(&tqspi->irq_work, tegra_qspi_work_handler);
> +
> +	ret = request_irq(tqspi->irq, tegra_qspi_isr, IRQF_SHARED,
> +			  dev_name(&pdev->dev), tqspi);

And use devm_request_irq() here?

>   	if (ret < 0) {
>   		dev_err(&pdev->dev, "failed to request IRQ#%u: %d\n", tqspi->irq, ret);
> -		goto exit_pm_disable;
> +		goto exit_destroy_wq;
>   	}
>   
>   	ret = spi_register_controller(host);
> @@ -1810,7 +1852,9 @@ static int tegra_qspi_probe(struct platform_device *pdev)
>   	return 0;
>   
>   exit_free_irq:
> -	free_irq(qspi_irq, tqspi);
> +	free_irq(tqspi->irq, tqspi);
> +exit_destroy_wq:
> +	destroy_workqueue(tqspi->wq);
>   exit_pm_disable:
>   	pm_runtime_dont_use_autosuspend(&pdev->dev);
>   	pm_runtime_force_suspend(&pdev->dev);
> @@ -1825,6 +1869,8 @@ static void tegra_qspi_remove(struct platform_device *pdev)
>   
>   	spi_unregister_controller(host);
>   	free_irq(tqspi->irq, tqspi);
> +	flush_workqueue(tqspi->wq);
> +	destroy_workqueue(tqspi->wq);
>   	pm_runtime_dont_use_autosuspend(&pdev->dev);
>   	pm_runtime_force_suspend(&pdev->dev);
>   	tegra_qspi_deinit_dma(tqspi);

Jon

-- 
nvpublic


  reply	other threads:[~2026-05-19  9:15 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-18 16:07 [PATCH 0/3] spi: tegra210-quad: Improve interrupt handling for loaded systems Vishwaroop A
2026-05-18 16:07 ` [PATCH 1/3] spi: tegra210-quad: Convert to hard IRQ with high-priority workqueue Vishwaroop A
2026-05-19  9:15   ` Jon Hunter [this message]
2026-05-18 16:07 ` [PATCH 2/3] spi: tegra210-quad: Cache TRANS_STATUS in ISR for timeout handler Vishwaroop A
2026-05-19  9:27   ` Jon Hunter
2026-05-18 16:07 ` [PATCH 3/3] spi: tegra210-quad: Process small PIO transfers in hard IRQ context Vishwaroop A

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=d058322b-1239-4429-97eb-ab49002881d7@nvidia.com \
    --to=jonathanh@nvidia.com \
    --cc=broonie@kernel.org \
    --cc=kyarlagadda@nvidia.com \
    --cc=ldewangan@nvidia.com \
    --cc=leitao@debian.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-spi@vger.kernel.org \
    --cc=linux-tegra@vger.kernel.org \
    --cc=skomatineni@nvidia.com \
    --cc=smangipudi@nvidia.com \
    --cc=thierry.reding@kernel.org \
    --cc=va@nvidia.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

all inboxes | Powered by JetHome®