linux-kernel - Expected behaviour for device driver aio_read(), aio

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <c384c5ea0906151348w3835632aob0034a4b877163bc@mail.gmail.com>
Date:	Mon, 15 Jun 2009 22:48:33 +0200
From:	Leon Woestenberg <leon.woestenberg@...il.com>
To:	linux-kernel@...r.kernel.org,
	kernelnewbies <kernelnewbies@...linux.org>
Subject: Expected behaviour for device driver aio_read(), aio_write() and 
	(girl)friends?

Hello,

as I cannot find any existing drivers using aio_read()/write(), and
the interface has changed since LDD3 now supporting vectored I/O,
where can I get more information on it?

Some questions I came up with:

For a multi vector I/O, can aio_read() still return the number of bytes?
Must it do so for synchronous I/O (either requested or possible) only,
or is returning the queued amount of I/O bytes also valid?
Or, must it still return -EIOCBQUEUED if everything was queued?

For multi-segment vectored IO, what must aio_complete return in res2?
The total of bytes across all completed I/O vectors?
And what if the I/O fails half-way the segments?

LDD3 says the first result of aio_complete() is the "usual result
status" for the operation.
I'm not sure what that is for multi-vectored I/O; I guess the total
amount of bytes transferred?

Regards,
-- 
Leon
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/