linux-kernel - Re: [PATCH v6] xhci: re-initialize the HC during resume if HCE was set

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] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 2 Feb 2022 21:12:53 +0200
From:   Mathias Nyman <mathias.nyman@...ux.intel.com>
To:     Puma Hsu <pumahsu@...gle.com>, mathias.nyman@...el.com,
        gregkh@...uxfoundation.org
Cc:     s.shtylyov@....ru, albertccwang@...gle.com,
        linux-usb@...r.kernel.org, linux-kernel@...r.kernel.org,
        stable@...r.kernel.org
Subject: Re: [PATCH v6] xhci: re-initialize the HC during resume if HCE was
 set

On 29.1.2022 11.30, Puma Hsu wrote:
> When HCE(Host Controller Error) is set, it means an internal
> error condition has been detected. Software needs to re-initialize
> the HC, so add this check in xhci resume.
> 
> Cc: stable@...r.kernel.org
> Signed-off-by: Puma Hsu <pumahsu@...gle.com>
> ---
> v2: Follow Sergey Shtylyov <s.shtylyov@....ru>'s comment.
> v3: Add stable@...r.kernel.org for stable release.
> v4: Refine the commit message.
> v5: Add a debug log. Follow Mathias Nyman <mathias.nyman@...ux.intel.com>'s comment.
> v6: Fix the missing declaration for str.
> 
>  drivers/usb/host/xhci.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
> index dc357cabb265..6f1198068004 100644
> --- a/drivers/usb/host/xhci.c
> +++ b/drivers/usb/host/xhci.c
> @@ -1091,6 +1091,7 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
>  	int			retval = 0;
>  	bool			comp_timer_running = false;
>  	bool			pending_portevent = false;
> +	char			str[XHCI_MSG_MAX];
>  
>  	if (!hcd->state)
>  		return 0;
> @@ -1146,8 +1147,10 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
>  		temp = readl(&xhci->op_regs->status);
>  	}
>  
> -	/* If restore operation fails, re-initialize the HC during resume */
> -	if ((temp & STS_SRE) || hibernated) {
> +	/* If restore operation fails or HC error is detected, re-initialize the HC during resume */
> +	if ((temp & (STS_SRE | STS_HCE)) || hibernated) {
> +		xhci_warn(xhci, "re-initialize HC during resume, USBSTS:%s\n",
> +			  xhci_decode_usbsts(str, temp));
>  
>  		if ((xhci->quirks & XHCI_COMP_MODE_QUIRK) &&
>  				!(xhci_all_ports_seen_u0(xhci))) {
> 

Ended up modifying this patch a bit more than I first intended,
- don't print warning in hibernation case, only error.
- maybe using a lot of stack for a debug string isn't really needed.
- make sure we read the usbsts register before checking for the HCE bit.

Does the below work for you? If yes, and you agree I'll apply it instead

---

diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
index dc357cabb265..04ec2de158bf 100644
--- a/drivers/usb/host/xhci.c
+++ b/drivers/usb/host/xhci.c
@@ -1091,6 +1091,7 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
        int                     retval = 0;
        bool                    comp_timer_running = false;
        bool                    pending_portevent = false;
+       bool                    reinit_xhc = false;
 
        if (!hcd->state)
                return 0;
@@ -1107,10 +1108,11 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
        set_bit(HCD_FLAG_HW_ACCESSIBLE, &xhci->shared_hcd->flags);
 
        spin_lock_irq(&xhci->lock);
-       if ((xhci->quirks & XHCI_RESET_ON_RESUME) || xhci->broken_suspend)
-               hibernated = true;
 
-       if (!hibernated) {
+       if (hibernated || xhci->quirks & XHCI_RESET_ON_RESUME || xhci->broken_suspend)
+               reinit_xhc = true;
+
+       if (!reinit_xhc) {
                /*
                 * Some controllers might lose power during suspend, so wait
                 * for controller not ready bit to clear, just as in xHC init.
@@ -1143,12 +1145,17 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
                        spin_unlock_irq(&xhci->lock);
                        return -ETIMEDOUT;
                }
-               temp = readl(&xhci->op_regs->status);
        }
 
-       /* If restore operation fails, re-initialize the HC during resume */
-       if ((temp & STS_SRE) || hibernated) {
+       temp = readl(&xhci->op_regs->status);
+
+       /* re-initialize the HC on Restore Error, or Host Controller Error */
+       if (temp & (STS_SRE | STS_HCE)) {
+               reinit_xhc = true;
+               xhci_warn(xhci, "xHC error in resume, USBSTS 0x%x, Reinit\n", temp);
+       }
 
+       if (reinit_xhc) {
                if ((xhci->quirks & XHCI_COMP_MODE_QUIRK) &&
                                !(xhci_all_ports_seen_u0(xhci))) {
                        del_timer_sync(&xhci->comp_mode_recovery_timer);