linux-kernel - Re: [PATCH] soundwire: bus: Make IRQ handling conditionally built

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

Message-ID: <20230920085133.GE103419@ediswmail.ad.cirrus.com>
Date:   Wed, 20 Sep 2023 08:51:33 +0000
From:   Charles Keepax <ckeepax@...nsource.cirrus.com>
To:     Vinod Koul <vkoul@...nel.org>
CC:     <yung-chuan.liao@...ux.intel.com>,
        <pierre-louis.bossart@...ux.intel.com>, <sanyog.r.kale@...el.com>,
        <linux-kernel@...r.kernel.org>, <patches@...nsource.cirrus.com>,
        <rdunlap@...radead.org>
Subject: Re: [PATCH] soundwire: bus: Make IRQ handling conditionally built

On Wed, Sep 20, 2023 at 09:05:13AM +0200, Vinod Koul wrote:
> On 18-09-23, 17:10, Charles Keepax wrote:
> > SoundWire has provisions for a simple callback for the IRQ handling so
> > has no hard dependency on IRQ_DOMAIN, but the recent addition of IRQ
> > handling was causing builds without IRQ_DOMAIN to fail. Resolve this by
> > moving the IRQ handling into its own file and only add it to the build
> > when IRQ_DOMAIN is included in the kernel.
> > 
> > Fixes: 12a95123bfe1 ("soundwire: bus: Allow SoundWire peripherals to register IRQ handlers")
> > Reported-by: kernel test robot <lkp@...el.com>
> > Closes: https://lore.kernel.org/oe-kbuild-all/202309150522.MoKeF4jx-lkp@intel.com/
> > Signed-off-by: Charles Keepax <ckeepax@...nsource.cirrus.com>
> > ---
> >  drivers/soundwire/Makefile   |  4 +++
> >  drivers/soundwire/bus.c      | 31 +++----------------
> >  drivers/soundwire/bus_type.c | 11 +++----
> >  drivers/soundwire/irq.c      | 59 ++++++++++++++++++++++++++++++++++++
> >  drivers/soundwire/irq.h      | 43 ++++++++++++++++++++++++++
> >  5 files changed, 115 insertions(+), 33 deletions(-)
> >  create mode 100644 drivers/soundwire/irq.c
> >  create mode 100644 drivers/soundwire/irq.h
> > 
> > diff --git a/drivers/soundwire/Makefile b/drivers/soundwire/Makefile
> > index c3d3ab3262d3a..657f5888a77b0 100644
> > --- a/drivers/soundwire/Makefile
> > +++ b/drivers/soundwire/Makefile
> > @@ -15,6 +15,10 @@ ifdef CONFIG_DEBUG_FS
> >  soundwire-bus-y += debugfs.o
> >  endif
> >  
> > +ifdef CONFIG_IRQ_DOMAIN
> > +soundwire-bus-y += irq.o
> > +endif
> 
> Any reason why we cant use depends for this?
> 

No reason we can't, but my thinking was really that SoundWire doesn't
really have a dependency on IRQ_DOMAIN, as you can use the original
callback mechanism. It seemed a shame to force it as a dependency,
when the whole subsystem can function happily without it.

That said, I am happy to switch to a simple dependency if you prefer?
It would certainly be a much simpler change.

Thanks,
Charles