imx.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
From: Frank Li <Frank.li@nxp.com>
To: Richard Zhu <hongxing.zhu@nxp.com>
Cc: imx@lists.linux.dev
Subject: Re: [PATCH v2] PCI: dwc: Fix suspend hang when e1000e network card is connected
Date: Tue, 14 May 2024 10:28:42 -0400	[thread overview]
Message-ID: <ZkN1GgczTvzAwYDQ@lizhi-Precision-Tower-5810> (raw)
In-Reply-To: <1715666198-9140-1-git-send-email-hongxing.zhu@nxp.com>

On Tue, May 14, 2024 at 01:56:38PM +0800, Richard Zhu wrote:
> In dw_pcie_suspend_noirq() and after the PME_TURN_OFF messgage is sent out,
> one PMC version2.0 endpoint device E1000E network card hangs in LTSSM stat
> polling and L2 stat check.
> 
> To avoid this hang issue after the PME_TURN_OFF message is sent out.
> Only poll and check LTSSM L2 stat if the EP's PMC version is 3 or later,
> 
> Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
> ---
> - Add a new help function dwc_get_devices_pmc_version() to get devices PMC
>   versions.
> 
>  .../pci/controller/dwc/pcie-designware-host.c | 46 ++++++++++++++++++-
>  1 file changed, 45 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c b/drivers/pci/controller/dwc/pcie-designware-host.c
> index cb8c3c2bcc790..0c2794e7bcf0a 100644
> --- a/drivers/pci/controller/dwc/pcie-designware-host.c
> +++ b/drivers/pci/controller/dwc/pcie-designware-host.c
> @@ -884,6 +884,42 @@ int dw_pcie_setup_rc(struct dw_pcie_rp *pp)
>  }
>  EXPORT_SYMBOL_GPL(dw_pcie_setup_rc);
>  
> +static int dwc_get_devices_pmc_version(struct dw_pcie *pci)
> +{
> +	int ret;
> +	u16 pmc, ver = 0;
> +	struct dw_pcie_rp *pp = &pci->pp;
> +	struct pci_bus *child, *root_bus = NULL;
> +	struct pci_dev *pdev;
> +
> +	/* Check the remote EP's PMC version */
> +	list_for_each_entry(child, &pp->bridge->bus->children, node) {
> +		if (child->parent == pp->bridge->bus) {
> +			root_bus = child;
> +			break;
> +		}
> +	}
> +
> +	if (!root_bus) {
> +		dev_err(pci->dev, "Failed to find downstream devices\n");
> +		return -ENODEV;
> +	}
> +
> +	list_for_each_entry(pdev, &root_bus->devices, bus_list) {
> +		if (PCI_SLOT(pdev->devfn) == 0) {
> +			/* find PCI PM capability in list */
> +			ret = pci_find_capability(pdev, PCI_CAP_ID_PM);
> +			if (!ret)
> +				return ret;
> +			/* Check device's PM ability version */
> +			pci_read_config_word(pdev, ret + PCI_PM_PMC, &pmc);
> +			ver &= pmc;

ver is 0.  so ver will be always 0 by '&='.

By the ways, in 7.5.2 PCI Power Management Capability Structure
I have not found version info.

Frank 

> +		}
> +	}
> +
> +	return ver;
> +}
> +
>  static int dw_pcie_pme_turn_off(struct dw_pcie *pci)
>  {
>  	struct dw_pcie_ob_atu_cfg atu = { 0 };
> @@ -924,7 +960,7 @@ int dw_pcie_suspend_noirq(struct dw_pcie *pci)
>  {
>  	u8 offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
>  	u32 val;
> -	int ret = 0;
> +	int pmc_ver, ret = 0;
>  
>  	/*
>  	 * If L1SS is supported, then do not put the link into L2 as some
> @@ -933,6 +969,8 @@ 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;
>  
> +	pmc_ver = dwc_get_devices_pmc_version(pci);
> +
>  	if (dw_pcie_get_ltssm(pci) > DW_PCIE_LTSSM_DETECT_ACT) {
>  		/* Only send out PME_TURN_OFF when PCIE link is up */
>  		if (pci->pp.ops->pme_turn_off)
> @@ -942,7 +980,13 @@ int dw_pcie_suspend_noirq(struct dw_pcie *pci)
>  
>  		if (ret)
>  			return ret;
> +	}
>  
> +	/*
> +	 * PMC Ver2.0 device e1000e NIC hangs in LTSSM polling and L2 check.
> +	 * Only do the L2 check when PMC version is Ver3.0 or later.
> +	 */
> +	if (pmc_ver >= 3) {
>  		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);
> -- 
> 2.37.1
> 

  parent reply	other threads:[~2024-05-14 14:28 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-14  5:56 [PATCH v2] PCI: dwc: Fix suspend hang when e1000e network card is connected Richard Zhu
2024-05-14 11:03 ` Fabio Estevam
2024-05-15  1:06   ` Hongxing Zhu
2024-05-14 14:28 ` Frank Li [this message]
2024-05-15  1:05   ` Hongxing Zhu

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=ZkN1GgczTvzAwYDQ@lizhi-Precision-Tower-5810 \
    --to=frank.li@nxp.com \
    --cc=hongxing.zhu@nxp.com \
    --cc=imx@lists.linux.dev \
    /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 read-only IMAP folder(s) and NNTP newsgroup(s).