From: "Marek Marczykowski-Górecki" <marmarek@invisiblethingslab.com>
To: Suwan Kim <suwan.kim027@gmail.com>
Cc: linux-usb@vger.kernel.org, Shuah Khan <skhan@linuxfoundation.org>,
Valentina Manea <valentina.manea.m@gmail.com>
Subject: Re: "usbip: Implement SG support to vhci-hcd and stub driver" causes a deadlock
Date: Wed, 11 Dec 2019 12:01:30 +0100 [thread overview]
Message-ID: <20191211110130.GN11116@mail-itl> (raw)
In-Reply-To: <20191211062705.GA11367@localhost.localdomain>
[-- Attachment #1: Type: text/plain, Size: 2595 bytes --]
On Wed, Dec 11, 2019 at 03:27:05PM +0900, Suwan Kim wrote:
> On Wed, Dec 11, 2019 at 04:20:32AM +0100, Marek Marczykowski-Górecki wrote:
> > On Wed, Dec 11, 2019 at 12:07:54PM +0900, Suwan Kim wrote:
> > > HCD should giveback URB to driver calling usb_hcd_giveback_urb().
> > > But in the case of transaction error, vhci_recv_ret_submit()
> > > terminates without giveback URB. So, I think the error path in
> > > usbip_recv_xbuff and usbip_recv_iso should unlink URB from ep and
> > > insert proper error code in urb->status that indicates error
> > > status to driver and handle giveback URB to driver.
> > >
> > > Then hub_irq doesn't need to flush error URB.
> > > That will be helpful to graceful connection shutdown.
> > >
> > >
> > > static void vhci_recv_ret_submit(struct vhci_device *vdev,
> > > struct usbip_header *pdu)
> > > ...
> > > ...
> > > if (usbip_recv_xbuff(ud, urb) < 0) {
> > > urb->status = -EPIPE;
> > > goto error; // goto error path
> > > }
> > >
> > > /* recv iso_packet_descriptor */
> > > if (usbip_recv_iso(ud, urb) < 0) {
> > > urb->status = -EPIPE;
> > > goto error; // goto error path
> > > }
> > > ...
> > > ...
> > > error://error path
> > > spin_lock_irqsave(&vhci->lock, flags);
> > > usb_hcd_unlink_urb_from_ep(vhci_hcd_to_hcd(vhci_hcd), urb);
> > > spin_unlock_irqrestore(&vhci->lock, flags);
> > >
> > > usb_hcd_giveback_urb(vhci_hcd_to_hcd(vhci_hcd), urb, urb->status);
> > >
> > > usbip_dbg_vhci_rx("Leave\n");
> > > }
> >
> > Yup, that works! Now the error is handled gracefully instead of hanging.
>
> Marek, Could you test if it works too?
Yes, this seems to work too.
> Regards,
> Suwan Kim
>
>
> diff --git a/drivers/usb/usbip/vhci_rx.c b/drivers/usb/usbip/vhci_rx.c
> index 33f8972ba842..19a807e398eb 100644
> --- a/drivers/usb/usbip/vhci_rx.c
> +++ b/drivers/usb/usbip/vhci_rx.c
> @@ -78,11 +78,11 @@ static void vhci_recv_ret_submit(struct vhci_device *vdev,
>
> /* recv transfer buffer */
> if (usbip_recv_xbuff(ud, urb) < 0)
> - return;
> + urb->status = -EPIPE;
>
> /* recv iso_packet_descriptor */
> if (usbip_recv_iso(ud, urb) < 0)
> - return;
> + urb->status = -EPIPE;
>
> /* restore the padding in iso packets */
> usbip_pad_iso(ud, urb);
--
Best Regards,
Marek Marczykowski-Górecki
Invisible Things Lab
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
prev parent reply other threads:[~2019-12-11 11:01 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-06 3:24 "usbip: Implement SG support to vhci-hcd and stub driver" causes a deadlock Marek Marczykowski-Górecki
2019-12-06 6:50 ` Suwan Kim
2019-12-06 20:57 ` Marek Marczykowski-Górecki
2019-12-06 21:12 ` Shuah Khan
2019-12-07 0:58 ` Marek Marczykowski-Górecki
2019-12-07 18:45 ` Marek Marczykowski-Górecki
2019-12-09 2:01 ` Suwan Kim
2019-12-09 3:37 ` Marek Marczykowski-Górecki
2019-12-09 6:35 ` Suwan Kim
2019-12-09 14:19 ` Marek Marczykowski-Górecki
2019-12-10 14:25 ` Suwan Kim
2019-12-10 15:32 ` Marek Marczykowski-Górecki
2019-12-11 3:07 ` Suwan Kim
2019-12-11 3:20 ` Marek Marczykowski-Górecki
2019-12-11 4:53 ` Suwan Kim
2019-12-11 6:27 ` Suwan Kim
2019-12-11 11:01 ` Marek Marczykowski-Górecki [this message]
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=20191211110130.GN11116@mail-itl \
--to=marmarek@invisiblethingslab.com \
--cc=linux-usb@vger.kernel.org \
--cc=skhan@linuxfoundation.org \
--cc=suwan.kim027@gmail.com \
--cc=valentina.manea.m@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).