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] [day] [month] [year] [list]
Date:   Wed, 9 Nov 2016 09:36:13 +0000
From:   Marc Zyngier <marc.zyngier@....com>
To:     Vineet Gupta <Vineet.Gupta1@...opsys.com>
Cc:     Yuriy Kolerov <yuriy.kolerov@...opsys.com>,
        "linux-snps-arc@...ts.infradead.org" 
        <linux-snps-arc@...ts.infradead.org>,
        "Alexey.Brodkin@...opsys.com" <Alexey.Brodkin@...opsys.com>,
        "tglx@...utronix.de" <tglx@...utronix.de>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [RFC] ARCv2: MCIP: Deprecate setting of affinity in Device Tree

Hi Vineet,

On 08/11/16 20:57, Vineet Gupta wrote:
> +CC MarcZ
> 
> Hi Marc,
> 
> I have a question below
> 

[...]

> So as discussed in a prior thread, we no longer support setting the default
> affinity via DT (so removed from xlate function). However we are now using the
> domain map function to set "seed" affinity based on cpu_online_mask(). Will that
> be the wrong place again to do this too.
> Also as Yuriy notes above, it is not guaranteed that all cpus will be online when
> the map function is called. So then better to set to boot cpu only : wondering
> what approach you guys take for ARM ?

For global interrupts (that can target any CPU), we make them affine to
the boot CPU, and let userspace move them around if required. Note that
we do not support round-robin on any of our recent interrupt
controllers, but maybe that's an acceptable configuration for you.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

Powered by blists - more mailing lists