netdev - Re: [net 02/14] Revert "mlx5: move affinity hints assignments to generic code"

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:   Thu, 4 Jan 2018 12:19:30 -0800
From:   Saeed Mahameed <saeedm@...lanox.com>
To:     Sagi Grimberg <sagi@...mberg.me>,
        Saeed Mahameed <saeedm@....mellanox.co.il>
Cc:     "David S. Miller" <davem@...emloft.net>,
        Linux Netdev List <netdev@...r.kernel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Jes Sorensen <jsorensen@...com>
Subject: Re: [net 02/14] Revert "mlx5: move affinity hints assignments to
 generic code"



On 12/26/2017 9:14 AM, Sagi Grimberg wrote:
> 
>> Are you sure it won't get populated at all  ? even if you manually set
>> IRQ affinity via sysfs ?
> 
> Yes, the msi_desc affinity is not initialized without the affinity
> descriptor passed (which is what PCI_IRQ_AFFINITY is for).
> 
>> Anyway we can implement this driver helper function to return the IRQ
>> affinity hint stored in the driver:
>>   "cpumask_first(mdev->priv.irq_info[vector].mask);"
> 
> minus the cpumask_first, but yea. Please send a new patch so we can
> test it out.

Actually using mdev->priv.irq_info[vector].mask is wrong since it only 
gives the initial hint and not the current actual affinity mask.

I found a better way to address this and return the actual dynamic 
affinity of an interrupt vector.

I will send a patch to net soon.