Re: error handling in tty_open

Am Montag, 26. März 2007 20:26 schrieb John:
> Hi,
> No, I don't see how it can work.
> 
> When a device is disconnected serial_open will
> fail at the very beginning

You are right. All failures must be handled. I'll make a new patch.
 
> But there may be a more serious problem.
> port is stored in tty->driver_data and 
> is shared among all files on the same device.
> 
> So, I wonder whether the following may happen:
> 
> proc A -> serial_open file 1 - success
> proc B -> serial_open file 2 - failure 
>   - set cleaning_up flag
> proc A -> serial_close file 1 
>   - aborted because of cleaning_up flag
> proc B -> serial_close file 2
>   - proceeded to close file 2 even if it was
>   never opened.

AFAICT both opening and closing a tty is done with BKL held.
As long as we don't sleep after incrementing cleaning_up we should
be safe. I'll check again.

> By the way, did you receive my patch,
> which I sent on Friday for review?
> E-Mail title was called
> "[PATCH 1/1] race condition fixes for usb-serial - new
> attempt - for review"

Yes. A bit of patience please.

	Regards
		Oliver

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel