linux-usb.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ladislav Michl <oss-lists@triops.cz>
To: Mathias Nyman <mathias.nyman@linux.intel.com>
Cc: linux-usb@vger.kernel.org, Sneeker Yeh <sneeker.yeh@gmail.com>
Subject: Re: xHCI host dies on device unplug
Date: Wed, 21 Dec 2022 13:05:50 +0100	[thread overview]
Message-ID: <Y6L2nnXpkkAJVLgh@lenoch> (raw)
In-Reply-To: <Y6Lbxhc/98QA6dMU@lenoch>

On Wed, Dec 21, 2022 at 11:11:19AM +0100, Ladislav Michl wrote:
> On Wed, Dec 21, 2022 at 11:58:42AM +0200, Mathias Nyman wrote:
> > Looked at that same series and turned patch 1/5 into a standalone quick hack that applies on 6.1
> > 
> > Untested, does it work for you?
> 
> Applied on the top of you stop_endpoint_fixes, 6.1.0. is a base tree:
> [   24.800835] xhci-hcd xhci-hcd.0.auto: Delay clearing port-1 CSC
> [   24.806788] usb 1-1: USB disconnect, device number 2
> [   28.148451] ieee80211 phy0: rt2x00usb_vendor_request: Error - Vendor Request 0x07 failed for offset 0x101c with error -19
> [   29.828466] xhci-hcd xhci-hcd.0.auto: xHCI host not responding to stop endpoint command
> [   29.856656] xhci-hcd xhci-hcd.0.auto: xHCI host controller not responding, assume dead
> [   29.864804] xhci-hcd xhci-hcd.0.auto: HC died; cleaning up
> [   29.949460] xhci-hcd xhci-hcd.0.auto: Late clearing port-1 CSC, portsc 0x202a0
> 
> What about checking whenever anything is still connected on command timeout
> and considering device autosuspended instead of killing it?

Just completed test run, it is working, but I'd consider it insane at least...
diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c
index b07d3740f554..d7b7faaac647 100644
--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -1580,6 +1580,21 @@ static bool xhci_pending_command_completion(struct xhci_hcd *xhci)
 	return false;
 }
 
+static bool xhci_is_anything_connected(struct xhci_hcd *xhci)
+{
+	struct usb_hcd *hcd = xhci_to_hcd(xhci);
+	struct xhci_hub *rhub = xhci_get_rhub(hcd);
+	struct xhci_port **ports = rhub->ports;
+	int i, max_ports = rhub->num_ports;
+
+	max_ports = rhub->num_ports;
+	for (i = 0; i < max_ports; i++)
+		if (PORT_CONNECT & readl(ports[i]->addr))
+			return true;
+
+	return false;
+}
+
 void xhci_handle_command_timeout(struct work_struct *work)
 {
 	struct xhci_hcd	*xhci;
@@ -1587,7 +1602,6 @@ void xhci_handle_command_timeout(struct work_struct *work)
 	char		str[XHCI_MSG_MAX];
 	u64		hw_ring_state;
 	u32		cmd_field3;
-	u32		usbsts;
 
 	xhci = container_of(to_delayed_work(work), struct xhci_hcd, cmd_timer);
 
@@ -1602,9 +1616,9 @@ void xhci_handle_command_timeout(struct work_struct *work)
 		return;
 	}
 
-	cmd_field3 = le32_to_cpu(xhci->current_cmd->command_trb->generic.field[3]);
-	usbsts = readl(&xhci->op_regs->status);
-	xhci_dbg(xhci, "Command timeout, USBSTS:%s\n", xhci_decode_usbsts(str, usbsts));
+	xhci_dbg(xhci, "Command timeout, USBSTS:%s, USBCMD: %08x\n",
+		 xhci_decode_usbsts(str, readl(&xhci->op_regs->status)),
+		 readl(&xhci->op_regs->command));
 
 	/* Did hw complete the command but event handler was blocked? */
 	if (xhci_pending_interrupt(xhci) > 0 &&
@@ -1616,10 +1630,16 @@ void xhci_handle_command_timeout(struct work_struct *work)
 		return;
 	}
 
+	cmd_field3 = le32_to_cpu(xhci->current_cmd->command_trb->generic.field[3]);
 	/* Bail out and tear down xhci if a stop endpoint command failed */
 	if (TRB_FIELD_TO_TYPE(cmd_field3) == TRB_STOP_RING) {
 		struct xhci_virt_ep	*ep;
 
+		if (!(xhci_is_anything_connected(xhci))) {
+			xhci_info(xhci, "xHCI autosuspended?\n");
+			goto time_out_completed;
+		}
+
 		xhci_warn(xhci, "xHCI host not responding to stop endpoint command\n");
 
 		ep = xhci_get_virt_ep(xhci, TRB_TO_SLOT_ID(cmd_field3),

  reply	other threads:[~2022-12-21 12:06 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-05 21:27 xHCI host dies on device unplug Ladislav Michl
2022-12-06 13:17 ` Ladislav Michl
2022-12-15 16:12   ` Ladislav Michl
2022-12-16 10:13     ` Mathias Nyman
2022-12-16 21:32       ` Ladislav Michl
2022-12-19 12:25         ` Mathias Nyman
2022-12-19 18:31           ` Ladislav Michl
2022-12-19 21:45             ` Ladislav Michl
2022-12-20  7:58               ` Ladislav Michl
2022-12-21  9:46                 ` Mathias Nyman
2022-12-21  7:14               ` Ladislav Michl
2022-12-21  9:58                 ` Mathias Nyman
2022-12-21 10:11                   ` Ladislav Michl
2022-12-21 12:05                     ` Ladislav Michl [this message]
2022-12-21 12:12                     ` Mathias Nyman
2022-12-21 12:21                       ` Ladislav Michl
2022-12-19  7:11       ` Ladislav Michl

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=Y6L2nnXpkkAJVLgh@lenoch \
    --to=oss-lists@triops.cz \
    --cc=linux-usb@vger.kernel.org \
    --cc=mathias.nyman@linux.intel.com \
    --cc=sneeker.yeh@gmail.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;
as well as URLs for NNTP newsgroup(s).