linux-kernel - Re: REGRESSION: "wlcore: sdio: allow pm to handle sdio power" breaks wifi on HiKey960

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: <20180614113459.GD112168@atomide.com>
Date:   Thu, 14 Jun 2018 04:34:59 -0700
From:   Tony Lindgren <tony@...mide.com>
To:     John Stultz <john.stultz@...aro.org>
Cc:     Valentin Schneider <valentin.schneider@....com>,
        Kalle Valo <kvalo@...eaurora.org>, Eyal Reizer <eyalr@...com>,
        lkml <linux-kernel@...r.kernel.org>,
        Ryan Grachek <ryan@...ted.us>
Subject: Re: REGRESSION: "wlcore: sdio: allow pm to handle sdio power" breaks
 wifi on HiKey960

* John Stultz <john.stultz@...aro.org> [180613 17:17]:
> On Wed, Jun 13, 2018 at 7:42 AM, Valentin Schneider
> <valentin.schneider@....com> wrote:
> > On 13/06/18 05:13, Tony Lindgren wrote:
> >> * John Stultz <john.stultz@...aro.org> [180612 22:15]:
> >>> Hey Folks,
> >>>   I noticed with linus/master wifi wasn't coming up on HiKey960. I
> >>> bisected it down and it seems to be due to:
> >>>
> >>> 60f36637bbbd ("wlcore: sdio: allow pm to handle sdio power")  and
> >>> 728a9dc61f13 ("wlcore: sdio: Fix flakey SDIO runtime PM handling")
> >>>
> >>> When wifi fails to load, the only useful error message I see is:
> >>> [    8.466097] wl1271_sdio mmc1:0001:2: wl12xx_sdio_power_on: failed
> >>> to get_sync(-13)
> >>
> >> Sorry to hear about that.
> >>
> >>> Reverting those two changes gets wifi working again for me:
> >>> [    8.754953] wlcore: wl18xx HW: 183x or 180x, PG 2.2 (ROM 0x11)
> >>> [    8.761778] random: crng init done
> >>> [    8.765185] random: 7 urandom warning(s) missed due to ratelimiting
> >>> [    8.779149] wlcore: loaded
> >>> ...
> >>> [   12.945903] wlcore: PHY firmware version: Rev 8.2.0.0.237
> >>> [   13.058077] wlcore: firmware booted (Rev 8.9.0.0.70)
> >>>
> >>>
> >>> Any suggestions how to resolve this w/o a revert?
> >>
> >> Sounds like we need to ignore also -EACCES if runtime PM is
> >> disabled for MMC. Care to try and see if the patch below
> >> helps?
> >>
> >
> > I don't use wifi with my board (I have an USB ethernet adapter), but I do
> > get the same error message as John.
> >
> > Reverting the patches works and I do see wlan0 being brought up. Sadly,
> > applying your patch doesn't seem to fix the issue - and actually it seems
> > to freeze my board on first boot with this error:
> >
> > [   11.169127] wlcore: ERROR timeout waiting for the hardware to complete initialization
> >
> > On second boot it finally comes to life, but issuing ifconfig freezes it again.
> >
> > $ dmesg | grep wl
> > [    5.922661] wl18xx_driver wl18xx.0.auto: Direct firmware load for ti-connectivity/wl18xx-conf.bin failed with error -2
> > [    5.933904] wlcore: ERROR could not get configuration binary ti-connectivity/wl18xx-conf.bin: -2
> > [    5.949158] wlcore: WARNING falling back to default config
> > [    6.199806] wlcore: wl18xx HW: 183x or 180x, PG 2.2 (ROM 0x11)
> > [    6.210738] wlcore: WARNING Detected unconfigured mac address in nvs, derive from fuse instead.
> > [    6.221644] wlcore: WARNING This default nvs file can be removed from the file system
> > [    6.235180] wlcore: loaded
> > [    6.820146] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready
> > [    7.280611] wlcore: PHY firmware version: Rev 8.2.0.0.236
> > [    7.388339] wlcore: firmware booted (Rev 8.9.0.0.69)
> > [    7.409610] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready
> > [    7.417815] wlcore: down
> > [   10.628867] wlcore: ERROR timeout waiting for the hardware to complete initialization
> >
> >
> > Seems like it's not getting powered on, which might be why those mmc_power_*
> > calls were in there originally ?
> 
> Ryan Grachek came up with a different solution. I still need to
> validate it, but it seems promising:
>   https://lkml.org/lkml/2018/6/13/481

Oh OK good to hear. Yeah that makes sense to me now.

Regards,

Tony