linux-kernel - Re: [PATCH v5] PCI: dwc: Clean up some unnecessary codes in dw_pcie_suspend

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] [day] [month] [year] [list]

Message-ID: <Z0YcaXU4LiR+kD07@lizhi-Precision-Tower-5810>
Date: Tue, 26 Nov 2024 14:07:21 -0500
From: Frank Li <Frank.li@....com>
To: Richard Zhu <hongxing.zhu@....com>
Cc: jingoohan1@...il.com, bhelgaas@...gle.com, lpieralisi@...nel.org,
	kw@...ux.com, manivannan.sadhasivam@...aro.org, robh@...nel.org,
	quic_krichai@...cinc.com, imx@...ts.linux.dev,
	kernel@...gutronix.de, linux-pci@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v5] PCI: dwc: Clean up some unnecessary codes in
 dw_pcie_suspend_noirq()

On Tue, Nov 26, 2024 at 03:39:09PM +0800, Richard Zhu wrote:
> Before sending PME_TURN_OFF, don't test the LTSSM state. Since it's safe
> to send PME_TURN_OFF message regardless of whether the link is up or
> down. So, there would be no need to test the LTSSM state before sending
> PME_TURN_OFF message.
>
> Only print the message when ltssm_stat is not in DETECT and POLL.
> In the other words, there isn't an error message when no endpoint is
> connected at all.
>
> Signed-off-by: Richard Zhu <hongxing.zhu@....com>

Reviewed-by: Frank Li <Frank.Li@....com>

> ---
> v5 changes:
> - Remove the redundant check "(val > DW_PCIE_LTSSM_DETECT_WAIT)".
> v4 changes:
> - Keep error return when L2 entry is failed and the endpoint is
>   connected refer to Krishna' comments. Thanks.
> v3 changes:
> - Refine the commit message refer to Manivannan's comments.
> - Regarding Frank's comments, avoid 10ms wait when no link up.
> v2 changes:
> - Don't remove L2 poll check.
> - Add one 1us delay after L2 entry.
> - No error return when L2 entry is timeout
> - Don't print message when no link up.
> ---
>  .../pci/controller/dwc/pcie-designware-host.c | 38 +++++++++++--------
>  drivers/pci/controller/dwc/pcie-designware.h  |  1 +
>  2 files changed, 23 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c b/drivers/pci/controller/dwc/pcie-designware-host.c
> index 24e89b66b772..bbd0ee862c12 100644
> --- a/drivers/pci/controller/dwc/pcie-designware-host.c
> +++ b/drivers/pci/controller/dwc/pcie-designware-host.c
> @@ -927,25 +927,31 @@ int dw_pcie_suspend_noirq(struct dw_pcie *pci)
>  	if (dw_pcie_readw_dbi(pci, offset + PCI_EXP_LNKCTL) & PCI_EXP_LNKCTL_ASPM_L1)
>  		return 0;
>
> -	/* Only send out PME_TURN_OFF when PCIE link is up */
> -	if (dw_pcie_get_ltssm(pci) > DW_PCIE_LTSSM_DETECT_ACT) {
> -		if (pci->pp.ops->pme_turn_off)
> -			pci->pp.ops->pme_turn_off(&pci->pp);
> -		else
> -			ret = dw_pcie_pme_turn_off(pci);
> -
> -		if (ret)
> -			return ret;
> +	if (pci->pp.ops->pme_turn_off)
> +		pci->pp.ops->pme_turn_off(&pci->pp);
> +	else
> +		ret = dw_pcie_pme_turn_off(pci);
> +	if (ret)
> +		return ret;
>
> -		ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
> -					PCIE_PME_TO_L2_TIMEOUT_US/10,
> -					PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
> -		if (ret) {
> -			dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
> -			return ret;
> -		}
> +	ret = read_poll_timeout(dw_pcie_get_ltssm, val,
> +				val == DW_PCIE_LTSSM_L2_IDLE ||
> +				val <= DW_PCIE_LTSSM_DETECT_WAIT,
> +				PCIE_PME_TO_L2_TIMEOUT_US/10,
> +				PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
> +	if (ret) {
> +		/* Only dump message when ltssm_stat isn't in DETECT and POLL */
> +		dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
> +		return ret;
>  	}
>
> +	/*
> +	 * Refer to r6.0, sec 5.3.3.2.1, software should wait at least
> +	 * 100ns after L2/L3 Ready before turning off refclock and
> +	 * main power. It's harmless too when no endpoint connected.
> +	 */
> +	udelay(1);
> +
>  	dw_pcie_stop_link(pci);
>  	if (pci->pp.ops->deinit)
>  		pci->pp.ops->deinit(&pci->pp);
> diff --git a/drivers/pci/controller/dwc/pcie-designware.h b/drivers/pci/controller/dwc/pcie-designware.h
> index 347ab74ac35a..bf036e66717e 100644
> --- a/drivers/pci/controller/dwc/pcie-designware.h
> +++ b/drivers/pci/controller/dwc/pcie-designware.h
> @@ -330,6 +330,7 @@ enum dw_pcie_ltssm {
>  	/* Need to align with PCIE_PORT_DEBUG0 bits 0:5 */
>  	DW_PCIE_LTSSM_DETECT_QUIET = 0x0,
>  	DW_PCIE_LTSSM_DETECT_ACT = 0x1,
> +	DW_PCIE_LTSSM_DETECT_WAIT = 0x6,
>  	DW_PCIE_LTSSM_L0 = 0x11,
>  	DW_PCIE_LTSSM_L2_IDLE = 0x15,
>
> --
> 2.37.1
>