Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 17 Aug 2022 12:28:09 -0700
From:   Matthias Kaehlcke <mka@chromium.org>
To:     Will McVicker <willmcvicker@google.com>
Cc:     Bjorn Helgaas <bhelgaas@google.com>, kernel-team@android.com,
        Sajid Dalvi <sdalvi@google.com>, linux-pci@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] PCI/PM: Switch D3Hot delay to also use usleep_range
Message-ID: <Yv1BSfkd6WUh2SFp@google.com>
References: <20220817185202.1689955-1-willmcvicker@google.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <20220817185202.1689955-1-willmcvicker@google.com>
Precedence: bulk

Hi Will,

On Wed, Aug 17, 2022 at 06:52:02PM +0000, Will McVicker wrote:
> From: Sajid Dalvi <sdalvi@google.com>
> 
> Since the PCI spec requires a 10ms D3Hot delay (defined by
> PCI_PM_D3HOT_WAIT) and a few of the PCI quirks update the D3Hot delay up
> to 120ms, let's add support for both usleep_range and msleep based on
> the delay time to improve the delay accuracy.
> 
> This patch is based off of a commit from Sajid Dalvi <sdalvi@google.com>
> in the Pixel 6 kernel tree [1]. Testing on a Pixel 6, found that the
> 10ms delay for the Exynos PCIe device was on average delaying for 19ms
> when the spec requires 10ms. Switching from msleep to uslseep_range
> therefore decreases the resume time on a Pixel 6 on average by 9ms.
> 
> [1] https://android.googlesource.com/kernel/gs/+/18a8cad68d8e6d50f339a716a18295e6d987cee3
> 
> Signed-off-by: Sajid Dalvi <sdalvi@google.com>
> Signed-off-by: Will McVicker <willmcvicker@google.com>
> ---
>  drivers/pci/pci.c | 20 +++++++++++++++-----
>  1 file changed, 15 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
> index 95bc329e74c0..97a042ca9032 100644
> --- a/drivers/pci/pci.c
> +++ b/drivers/pci/pci.c
> @@ -63,16 +63,26 @@ struct pci_pme_device {
>  };
>  
>  #define PME_TIMEOUT 1000 /* How long between PME checks */
> +#define MAX(a, b) ((a) >= (b) ? (a) : (b))

no need to define this macro, you can use max() from
include/linux/minmax.h instead.

>  
>  static void pci_dev_d3_sleep(struct pci_dev *dev)
>  {
> -	unsigned int delay = dev->d3hot_delay;
> +	unsigned int delay_ms = dev->d3hot_delay;
>  
> -	if (delay < pci_pm_d3hot_delay)
> -		delay = pci_pm_d3hot_delay;
> +	if (delay_ms < pci_pm_d3hot_delay)
> +		delay_ms = pci_pm_d3hot_delay;

nit: since you are already touching this code you could change it to:

	unsigned int delay_ms = max(dev->d3hot_delay, pci_pm_d3hot_delay);

>  
> -	if (delay)
> -		msleep(delay);
> +	if (delay_ms) {
> +		if (delay_ms <= 20) {
> +			/* Use a 20-25% upper bound with 1ms minimum */
> +			unsigned int upper = MAX((delay_ms >> 3) << 1, 1);

Not sure this optimization of using bit shifts instead of the clearer do_div()
is really needed here. pci_dev_d3_sleep() is not a super hot code path IIUC.

> +
> +			usleep_range(delay_ms * USEC_PER_MSEC,
> +				     (delay_ms + upper) * USEC_PER_MSEC);
> +		} else {
> +			msleep(delay_ms);
> +		}
> +	}
>  }
>  
>  bool pci_reset_supported(struct pci_dev *dev)
> 
> base-commit: 274a2eebf80c60246f9edd6ef8e9a095ad121264
> -- 
> 2.37.1.595.g718a3a8f04-goog
>