[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAGXu5jK2kLiOGTbq2TCBp97c2KZsZKPDNLSWDWY4r+RQpAP4fQ@mail.gmail.com>
Date: Wed, 1 Aug 2018 12:50:26 -0700
From: Kees Cook <keescook@...omium.org>
To: Bart Van Assche <Bart.VanAssche@....com>
Cc: "hch@...radead.org" <hch@...radead.org>,
"david.kershner@...sys.com" <david.kershner@...sys.com>,
"linux-block@...r.kernel.org" <linux-block@...r.kernel.org>,
"rdunlap@...radead.org" <rdunlap@...radead.org>,
"u.kleine-koenig@...gutronix.de" <u.kleine-koenig@...gutronix.de>,
"nab@...ux-iscsi.org" <nab@...ux-iscsi.org>,
"tglx@...utronix.de" <tglx@...utronix.de>,
"hch@....de" <hch@....de>,
"martin.petersen@...cle.com" <martin.petersen@...cle.com>,
"manoj@...ux.vnet.ibm.com" <manoj@...ux.vnet.ibm.com>,
"axboe@...nel.dk" <axboe@...nel.dk>,
"jgross@...e.com" <jgross@...e.com>,
"cyrille.pitchen@...e-electrons.com"
<cyrille.pitchen@...e-electrons.com>,
"sdharia@...eaurora.org" <sdharia@...eaurora.org>,
"viresh.kumar@...aro.org" <viresh.kumar@...aro.org>,
"linux-scsi@...r.kernel.org" <linux-scsi@...r.kernel.org>,
"target-devel@...r.kernel.org" <target-devel@...r.kernel.org>,
"pombredanne@...b.com" <pombredanne@...b.com>,
"linux-ide@...r.kernel.org" <linux-ide@...r.kernel.org>,
"bp@...en8.de" <bp@...en8.de>,
"mrochs@...ux.vnet.ibm.com" <mrochs@...ux.vnet.ibm.com>,
"tj@...nel.org" <tj@...nel.org>,
"davem@...emloft.net" <davem@...emloft.net>,
"jejb@...ux.vnet.ibm.com" <jejb@...ux.vnet.ibm.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"sboyd@...eaurora.org" <sboyd@...eaurora.org>,
"ukrishn@...ux.vnet.ibm.com" <ukrishn@...ux.vnet.ibm.com>
Subject: Re: [PATCH v2 3/9] scsi: build scsi_common.o for all scsi passthrough
request users
On Tue, Jul 31, 2018 at 1:18 PM, Bart Van Assche <Bart.VanAssche@....com> wrote:
> On Tue, 2018-07-31 at 13:12 -0700, hch@...radead.org wrote:
>> It shouldn't. All these are built into scsi_mod.o, which is only built
>> when CONFIG_SCSI is set. Under what circumstances do you see them built?
>
> Although I think you are right, I still prefer that the scsi_common.c file
> would be moved to a new directory. That will prevent confusion later on for
> people who want to add additional code. This patch makes it nontrivial to
> figure out which code is built when SCSI target functionality is enabled but
> SCSI initiator functionality is not selected. I think moving scsi_common.c
> into a new directory would make it much easier to figure out which code is
> built depending on the kernel configuration.
While I don't disagree with you, this series is the result of several
back-and-forth discussions where that option was seemingly rejected.
Christoph's approach seemed to satisfy Jens. If you can convince them,
sure! I'll do whatever the consensus is. :)
-Kees
--
Kees Cook
Pixel Security
Powered by blists - more mailing lists