From: "zbr@ioremap.net" <zbr@ioremap.net>
Envelope-From: drustafa@yandex.ru
To: David Fries <david@fries.net>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
In-Reply-To: <1391390104-31154-1-git-send-email-David@Fries.net>
References: <1391390104-31154-1-git-send-email-David@Fries.net>
Subject: Re: [PATCH 0/4] w1: refcnt fix, skip non-error send, docs
MIME-Version: 1.0
Message-Id: <9071391471978@web15j.yandex.ru>
Date: Tue, 04 Feb 2014 03:59:38 +0400
Content-Transfer-Encoding: 8bit
Content-Type: text/plain; charset=koi8-r
Sender: linux-kernel-owner@vger.kernel.org

Hi

03.02.2014, 05:15, "David Fries" <David@Fries.net>:

> ?I could submit these patches as in, which would require the previous
> ?set, or I could merge the documentation into the previous set and
> ?resubmit them all since they haven't made it into the kernel tree yet.
> ?Opinions?
>
> ?Here's a small refcnt fix, skipping sending non-error messages, and
> ?documentation and comment updates.
>
> ?non-error error messages:
> ?Currently every master or slave command is sending a response with
> ?w1_netlink_send_error no matter if there is an error or not. ?This
> ?makes commands like list slaves W1_CMD_LIST_SLAVES or W1_CMD_READ
> ?return two messages, one with data and one without. ?That is a problem
> ?with the list slaves because they are identical except for one having
> ?data and one not, and since there could be no slaves known to the
> ?kernel you can't just discard the no data case, unless the program
> ?were to expect two replies. ?So I propose only sending the error reply
> ?if there is an error, in which case there wouldn't be a normal reply
> ?(such as read). ?This would mean commands like write would no longer
> ?return a response unless there was an error. ?If an application wanted
> ?to verify the kernel received the write message it could follow it by
> ?a read to verify the data or just that read came after write and had a
> ?response so write must have completed without error. ?I think it is
> ?safe to do away with the extra replies. ?If someone sees a big enough
> ?need for this, I could modify it so all commands return one response,
> ?with commands like write always calling send error even if there
> ?wasn't one.

I created this protocol to handle cases like nothing is returned, but yet userspace knows
operations has been completed. Also, you can not really change it at this time - there are
already userspace application which may depend on the last ack to find out its request completed.

Reference counter fix is correct, please submit it in the separate patch.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/