linux-kernel - Re: Suspend/resume issues with Mediatek MT7961 wifi card

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>] [day] [month] [year] [list]

Message-ID: <CACAwPwaYxRa09Qm3N=L7PVBkAabJ5tWO_ZryoF9FZNLxg2v+tA@mail.gmail.com>
Date:   Sat, 16 Oct 2021 19:26:59 +0300
From:   Maxim Levitsky <maximlevitsky@...il.com>
To:     Felix Fietkau <nbd@....name>
Cc:     Lorenzo Bianconi <lorenzo.bianconi83@...il.com>,
        Ryder Lee <ryder.lee@...iatek.com>,
        Sean Wang <sean.wang@...iatek.com>,
        Soul Huang <Soul.Huang@...iatek.com>,
        Deren Wu <deren.wu@...iatek.com>,
        linux-wireless@...r.kernel.org, netdev@...r.kernel.org,
        LKML <linux-kernel@...r.kernel.org>, mlevitsk@...hat.com
Subject: Re: Suspend/resume issues with Mediatek MT7961 wifi card

On Sat, Oct 16, 2021 at 7:23 PM Maxim Levitsky <maximlevitsky@...il.com> wrote:
>
> I recently bought an Aya Neo tablet which now has this wifi chipset in its latest 2021 hardware revision.
>
> I am using kernel 5.15.0-rc4 (compiled from source), and I also merged in it and tested the master branch of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-testing.git
>
> I tested with and without wireless running before suspend (using gnome's wireless disable option in the gnome's menu).
>
> In all cases, the wireless card dies after the resume, which itself takes a lot of time most likely due to timeouts in the resume handler of this driver, and then the system is also partially hung.
>
> Looks like the driver's resume (or suspend) function isn't working correctly, and dmesg also indicates several failures.
>
> I had attached the full dmesg after a single suspend/resume cycle showing those errors.
>
> If I unload the mt7921e prior to suspend and load it after, there are no errors, no timeouts, and wifi connects and works fine after resume.
>
> My distro is Fedora 35 beta. Bluetooth also used to fail sometimes after resume when I used older Fedora versions, but now it seems very robust, and doesn't need any disabling, reloading of btusb/etc.
>
> I am a kernel developer (I work on KVM these days), and I can test any patches/workarounds/firmware/etc when I find some time for that.
>
> Best regards,
>     Maxim Levitsky
>

(resend using plain text - reminds me to never use Gmail's web interface.)

>

Download attachment "after_suspend.dmesg" of type "application/octet-stream" (151490 bytes)