[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [patch 27/30] xen/events: Only force affinity mask for percpu interrupts
 
- To: Thomas Gleixner <tglx@xxxxxxxxxxxxx>, LKML <linux-kernel@xxxxxxxxxxxxxxx>
 
- From: boris.ostrovsky@xxxxxxxxxx
 
- Date: Thu, 10 Dec 2020 18:20:46 -0500
 
- Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>, Marc Zyngier <maz@xxxxxxxxxx>,        Juergen Gross <jgross@xxxxxxxx>,        Stefano Stabellini <sstabellini@xxxxxxxxxx>,        xen-devel@xxxxxxxxxxxxxxxxxxxx,        "James E.J. Bottomley" <James.Bottomley@xxxxxxxxxxxxxxxxxxxxx>,        Helge Deller <deller@xxxxxx>, afzal mohammed <afzal.mohd.ma@xxxxxxxxx>,        linux-parisc@xxxxxxxxxxxxxxx, Russell King <linux@xxxxxxxxxxxxxxx>,        linux-arm-kernel@xxxxxxxxxxxxxxxxxxx,        Mark Rutland <mark.rutland@xxxxxxx>,        Catalin Marinas <catalin.marinas@xxxxxxx>,        Will Deacon <will@xxxxxxxxxx>,        Christian Borntraeger <borntraeger@xxxxxxxxxx>,        Heiko Carstens <hca@xxxxxxxxxxxxx>, linux-s390@xxxxxxxxxxxxxxx,        Jani Nikula <jani.nikula@xxxxxxxxxxxxxxx>,        Joonas Lahtinen <joonas.lahtinen@xxxxxxxxxxxxxxx>,        Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx>, David Airlie <airlied@xxxxxxxx>,        Daniel Vetter <daniel@xxxxxxxx>,        Pankaj Bharadiya <pankaj.laxminarayan.bharadiya@xxxxxxxxx>,        Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>,        Wambui Karuga <wambui.karugax@xxxxxxxxx>,        intel-gfx@xxxxxxxxxxxxxxxxxxxxx, dri-devel@xxxxxxxxxxxxxxxxxxxxx,        Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxxxxxxxx>,        Linus Walleij <linus.walleij@xxxxxxxxxx>, linux-gpio@xxxxxxxxxxxxxxx,        Lee Jones <lee.jones@xxxxxxxxxx>, Jon Mason <jdmason@xxxxxxxx>,        Dave Jiang <dave.jiang@xxxxxxxxx>, Allen Hubbe <allenbh@xxxxxxxxx>,        linux-ntb@xxxxxxxxxxxxxxxx,        Lorenzo Pieralisi <lorenzo.pieralisi@xxxxxxx>,        Rob Herring <robh@xxxxxxxxxx>, Bjorn Helgaas <bhelgaas@xxxxxxxxxx>,        Michal Simek <michal.simek@xxxxxxxxxx>, linux-pci@xxxxxxxxxxxxxxx,        Karthikeyan Mitran <m.karthikeyan@xxxxxxxxxxxxxx>,        Hou Zhiqiang <Zhiqiang.Hou@xxxxxxx>, Tariq Toukan <tariqt@xxxxxxxxxx>,        "David S. Miller" <davem@xxxxxxxxxxxxx>,        Jakub Kicinski <kuba@xxxxxxxxxx>, netdev@xxxxxxxxxxxxxxx,        linux-rdma@xxxxxxxxxxxxxxx, Saeed Mahameed <saeedm@xxxxxxxxxx>,        Leon Romanovsky <leon@xxxxxxxxxx>
 
- Delivery-date: Thu, 10 Dec 2020 23:21:54 +0000
 
- List-id: Xen developer discussion <xen-devel.lists.xenproject.org>
 
 
 
On 12/10/20 2:26 PM, Thomas Gleixner wrote:
> All event channel setups bind the interrupt on CPU0 or the target CPU for
> percpu interrupts and overwrite the affinity mask with the corresponding
> cpumask. That does not make sense.
>
> The XEN implementation of irqchip::irq_set_affinity() already picks a
> single target CPU out of the affinity mask and the actual target is stored
> in the effective CPU mask, so destroying the user chosen affinity mask
> which might contain more than one CPU is wrong.
>
> Change the implementation so that the channel is bound to CPU0 at the XEN
> level and leave the affinity mask alone. At startup of the interrupt
> affinity will be assigned out of the affinity mask and the XEN binding will
> be updated. 
If that's the case then I wonder whether we need this call at all and instead 
bind at startup time.
-boris
> Only keep the enforcement for real percpu interrupts.
 
 
    
     |