netdev - RE: [net-next 05/14] i40e/i40evf: always set the CLEARPBA flag when re-enabling interrupts

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]

Message-ID: <063D6719AE5E284EB5DD2968C1650D6DD008ECFE@AcuExch.aculab.com>
Date:   Tue, 10 Oct 2017 13:07:31 +0000
From:   David Laight <David.Laight@...LAB.COM>
To:     'Jeff Kirsher' <jeffrey.t.kirsher@...el.com>,
        "davem@...emloft.net" <davem@...emloft.net>
CC:     Jacob Keller <jacob.e.keller@...el.com>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "nhorman@...hat.com" <nhorman@...hat.com>,
        "sassmann@...hat.com" <sassmann@...hat.com>,
        "jogreene@...hat.com" <jogreene@...hat.com>
Subject: RE: [net-next 05/14] i40e/i40evf: always set the CLEARPBA flag when
 re-enabling interrupts

From: Jeff Kirsher
> Sent: 09 October 2017 23:39
> In the past we changed driver behavior to not clear the PBA when
> re-enabling interrupts. This change was motivated by the flawed belief
> that clearing the PBA would cause a lost interrupt if a receive
> interrupt occurred while interrupts were disabled.
> 
> According to empirical testing this isn't the case. Additionally, the
> data sheet specifically says that we should set the CLEARPBA bit when
> re-enabling interrupts in a polling setup.

I presume this if the MSI-X Pending Bit Array?
Normally this bit is cleared when the interrupt is actioned.

If request the device clear the PBA then it (probably) won't
raise an interrupt when it is unmasked (by clearing the 'masked' bit).

If you've just checked all the rings (with the interrupt masked)
and you clear the PBA bit when you unmask interrupts then you will
need to do another scan of the rings to pick up any packets that
arrived (or tried to signal an interrupt) in that small gap.

'Empirical testing' probably won't hit the timing window.

	David