linux-kernel - Re: [PATCH] usb: host: ehci: workaround PME bug on AMD EHCI controller

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

Date:   Tue, 20 Jun 2017 10:36:59 +0800
From:   Kai-Heng Feng <kai.heng.feng@...onical.com>
To:     Alan Stern <stern@...land.harvard.edu>
Cc:     Bjorn Helgaas <helgaas@...nel.org>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        gregkh@...uxfoundation.org, USB list <linux-usb@...r.kernel.org>,
        linux-pci@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>,
        "Rafael J. Wysocki" <rafael.j.wysocki@...el.com>,
        linux-pm@...r.kernel.org
Subject: Re: [PATCH] usb: host: ehci: workaround PME bug on AMD EHCI controller

On Tue, Jun 20, 2017 at 2:32 AM, Alan Stern <stern@...land.harvard.edu> wrote:
>
> It's possible that the test was invalid.  Kai-Heng did not say whether
> /sys/.../power/wakeup was set to "enabled" for both the EHCI controller
> and the USB root hub beneath it, before the test was started.  If
> either of them was set to "disabled" then we would not expect a plug or
> unplug event to wake up the system.

You are right, it's "disabled" on USB root hub.
Changed it to "enabled", the test result remains the same.

>
> In any case, the controller should be set to the lowest power setting
> that is consistent with the desired wakeup behavior.  If wakeup is set
> to "enabled" then the state should be D2 -- if possible.  That's the
> theory, anyway.  If the system supports putting devices only into D3
> during S3 sleep then there's no choice, but if we do have a choice then
> we should take it.
>
> BTW, I just noticed that pci_target_state() uses device_may_wakeup() to
> get the desired wakeup behavior.  That is correct for system sleep, but
> it is wrong for runtime PM.  For runtime PM, wakeup should be enabled
> whenever the hardware allows it, so the test should be
> device_can_wakeup().
>
> This means that pci_target_state() should behave differently depending
> on whether it is called from pci_prepare_to_sleep() or from
> pci_finish_runtime_suspend().  Probably nobody noticed this before
> because it usually doesn't make any difference.
>
> Alan Stern
>