[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5b05f924-9077-4e68-9925-248470154a90@oracle.com>
Date: Mon, 11 Mar 2024 10:59:43 -0500
From: michael.christie@...cle.com
To: lduncan@...e.com, target-devel@...r.kernel.org
Cc: linux-scsi@...r.kernel.org, linux-kernel@...r.kernel.org, dbond@...e.com,
hare@...e.de, cleech@...hat.com
Subject: Re: [PATCH 2/2] scsi: target: iscsi: don't warn of R/W when no data
On 12/7/23 11:42 AM, lduncan@...e.com wrote:
> From: Lee Duncan <lduncan@...e.com>
>
> The LIO target code has a warning about setting the
> read and/or write header bits with a PDU that has zero
> transfer length, even though the code mentions that the
> SPEC (RFC 3720) allows this, and that some initiators
> set these bits. But in practice such initiators end up
> flooding the logs with thousands of warning messages for
> IO that is allowed.
>
I've never seen us hit this. What initiator is doing this and what is
the command they are sending?
Is it also related to the first patch? Is some initiator sending
something like a TUR with the immediate bit set during some sort of
stall/timeout?
Powered by blists - more mailing lists