mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Alan Cox <alan@lxorguk.ukuu.org.uk>
To: Toby Gray <toby.gray@realvnc.com>
Cc: Oliver Neukum <oliver@neukum.name>,
	Greg Kroah-Hartman <gregkh@suse.de>,
	linux-usb@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] USB: cdc-acm: Prevent data loss when filling tty buffer.
Date: Mon, 21 Mar 2011 16:56:12 +0000	[thread overview]
Message-ID: <20110321165612.0f764046@lxorguk.ukuu.org.uk> (raw)
In-Reply-To: <1300722745-2404-1-git-send-email-toby.gray@realvnc.com>

On Mon, 21 Mar 2011 15:52:25 +0000
Toby Gray <toby.gray@realvnc.com> wrote:

> When sending large quantities of data through a CDC ACM channel it is possible
> for data to be lost when attempting to copy the data to the tty buffer. This
> occurs due to the return value from tty_insert_flip_string not being checked.

For a tty that is normally the right thing to do - no flow control was
asserted and the internal 64K of buffering was overrun so discard.

> This patch adds checking for how many bytes have been inserted into the tty
> buffer and returns any remaining bytes back to the filled read buffer list.

How does ACM handle flow control - is the expectation that it gets flow
controlled in hardware by not having the opportunity to send bits
to the host end ? If so this seems to make sense.


> +	copied = 0;

Surely copied = buf->size is the no tty assumption "we had the lot and
discarded it"

>  	if (tty) {
>  		spin_lock_irqsave(&acm->throttle_lock, flags);
>  		throttled = acm->throttle;
>  		spin_unlock_irqrestore(&acm->throttle_lock, flags);
>  		if (!throttled) {
> -			tty_insert_flip_string(tty, buf->base, buf->size);
> +			copied = tty_insert_flip_string(tty,
> +							buf->base, buf->size);
>  			tty_flip_buffer_push(tty);
>  		} else {
>  			tty_kref_put(tty);
> @@ -440,9 +443,26 @@ next_buffer:
>  		}
>  	}
>  
> -	spin_lock_irqsave(&acm->read_lock, flags);
> -	list_add(&buf->list, &acm->spare_read_bufs);
> -	spin_unlock_irqrestore(&acm->read_lock, flags);
> +	if (copied == buf->size || !tty) {

Which would simplify this lot

> +		spin_lock_irqsave(&acm->read_lock, flags);
> +		list_add(&buf->list, &acm->spare_read_bufs);
> +		spin_unlock_irqrestore(&acm->read_lock, flags);
> +	} else {
> +		tty_kref_put(tty);
> +		dbg("Partial buffer fill");
> +		if (copied > 0) {
> +			memmove(buf->base,
> +				buf->base + copied,
> +				buf->size - copied);
> +			buf->size -= copied;
> +		}

Would it not be cleaner to add a buf->head pointer that could simply be
advanced so the code would become

	buf->head += copied;
	buf->size -= copied;
	if (buf->size != 0)
		list_add(&buf->list, &acm->filled_read_bufs);

instead of all the memmove uglies ?



  reply	other threads:[~2011-03-21 16:55 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-03-21 15:52 Toby Gray
2011-03-21 16:56 ` Alan Cox [this message]
2011-03-21 17:58   ` Toby Gray
2011-03-22  8:07   ` Oliver Neukum
2011-03-22 11:07     ` Alan Cox
2011-03-21 18:04 ` [PATCH v2] " Toby Gray
2011-03-22 10:05   ` Johan Hovold
2011-03-22 10:35     ` Alan Cox
2011-03-22 11:34       ` Johan Hovold
2011-03-22 14:11     ` Toby Gray
2011-03-22 18:05       ` Alan Cox
2011-03-22  7:43 ` [PATCH] " Oliver Neukum
2011-03-23 12:15   ` Toby Gray

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=20110321165612.0f764046@lxorguk.ukuu.org.uk \
    --to=alan@lxorguk.ukuu.org.uk \
    --cc=gregkh@suse.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=oliver@neukum.name \
    --cc=toby.gray@realvnc.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

all inboxes | Powered by JetHome®