linux-kernel - RE: [patch 0/2] tsc/adjust: Cure suspend/resume issues and prevent TSC deadline timer irq storm

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 for Android: free password hash cracker in your pocket

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

Date:   Thu, 26 Jan 2017 23:40:10 +0000
From:   "Stanton, Kevin B" <kevin.b.stanton@...el.com>
To:     Thomas Gleixner <tglx@...utronix.de>,
        Roland Scheidegger <rscheidegger_lists@...peed.ch>
CC:     LKML <linux-kernel@...r.kernel.org>,
        "x86@...nel.org" <x86@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Borislav Petkov <bp@...en8.de>,
        "Schlobohm, Bruce" <bruce.schlobohm@...el.com>,
        Allen Hung <allen_hung@...l.com>,
        "Sharon, Barak" <barak.sharon@...el.com>,
        "Segev, Gil" <gil.segev@...el.com>
Subject: RE: [patch 0/2] tsc/adjust: Cure suspend/resume issues and prevent
 TSC deadline timer irq storm

On Thurs, 15 Dec 2016, Thomas Gleixner wrote:
>See the SDM. It suggests that the deadline is compared to the TSC value.
>I don't care how it is implemented, but I very much care about it being
>documented in the way it is implemented, which is obviously not the case.

The TSC_DEADLINE behavior definitely diverges from the SDM and we've root caused the issue and are finalizing a fix that addresses the problem. Glad that a software workaround was also possible.  We also plan to expand our documentation on what good behavior of a BIOS is as it relates to adjusting TSC at boot (in almost every circumstance you shouldn't do it).  The official errata will be forthcoming. Thank you for helping to identify this.

Kevin