From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Lord Subject: Re: [PATCH] scsi: Silence unnecessary warnings about ioctl to partition Date: Fri, 04 May 2012 09:06:10 -0400 Message-ID: <4FA3D442.9050801@teksavvy.com> References: <1335953452-10460-1-git-send-email-jack@suse.cz> <4FA1092E.9090603@redhat.com> <20120502115447.7dcc3a54@pyramind.ukuu.org.uk> <4FA11454.2010103@redhat.com> <20120502121208.3c19a9bc@pyramind.ukuu.org.uk> <4FA11963.3040007@redhat.com> <4FA18D33.3060607@teksavvy.com> <4FA23809.2090905@redhat.com> <4FA27CDB.2080103@teksavvy.com> <4FA27E7C.6070802@redhat.com> <4FA2C205.3030801@teksavvy.com> <4FA3799A.6010908@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <4FA3799A.6010908@redhat.com> Sender: linux-kernel-owner@vger.kernel.org To: Paolo Bonzini Cc: Alan Cox , Jan Kara , Jens Axboe , LKML , James Bottomley , linux-scsi@vger.kernel.org List-Id: linux-scsi@vger.kernel.org On 12-05-04 02:39 AM, Paolo Bonzini wrote: > Il 03/05/2012 19:36, Mark Lord ha scritto: > >>>>>> Dunno which flags, I'm just ignoring them and waiting >>>>>> for the noise message to get reverted. > > I said which ioctls, not which options. I.e. cut-and-paste from dmesg. Here are some: messages.1:Apr 28 14:08:09 zbox5 kernel: [ 268.800536] hdparm: sending ioctl 330 to a partition! messages.1:Apr 28 14:08:09 zbox5 kernel: [ 268.800547] hdparm: sending ioctl 330 to a partition! messages.1:Apr 28 14:08:57 zbox5 kernel: [ 316.413944] hdparm: sending ioctl 330 to a partition! messages.1:Apr 28 14:08:57 zbox5 kernel: [ 316.413951] hdparm: sending ioctl 330 to a partition! messages.1:Apr 28 14:20:10 zbox5 kernel: [ 989.525085] hdparm: sending ioctl 330 to a partition! messages.1:Apr 28 14:20:10 zbox5 kernel: [ 989.525093] hdparm: sending ioctl 330 to a partition! The in say this about 330: ... /* 0x330 is reserved - used to be HDIO_GETGEO_BIG */ ... So it's HDIO_GETGEO_BIG, which doesn't exist in newer kernels. I wonder when that got removed? Minor userspace breakage there. hdparm issues it first as a backward-compatibility thing, before falling back to the even-more obsolete HDIO_GETGEO, which curiously enough is still in modern kernels. Cheers