linux-kernel - Re: [PATCH] ath10k: qmi: Skip host capability request for Xiaomi Poco F1

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]

Message-ID: <20200925152245.GD2510@yoga>
Date:   Fri, 25 Sep 2020 10:22:45 -0500
From:   Bjorn Andersson <bjorn.andersson@...aro.org>
To:     Amit Pundir <amit.pundir@...aro.org>
Cc:     Kalle Valo <kvalo@...eaurora.org>,
        David S Miller <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Jeffrey Hugo <jeffrey.l.hugo@...il.com>,
        John Stultz <john.stultz@...aro.org>,
        Sumit Semwal <sumit.semwal@...aro.org>,
        Konrad Dybcio <konradybcio@...il.com>,
        ath10k <ath10k@...ts.infradead.org>,
        linux-wireless@...r.kernel.org, netdev@...r.kernel.org,
        lkml <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] ath10k: qmi: Skip host capability request for Xiaomi
 Poco F1

On Mon 21 Sep 05:38 CDT 2020, Amit Pundir wrote:

> On Thu, 17 Sep 2020 at 21:35, Bjorn Andersson
> <bjorn.andersson@...aro.org> wrote:
> >
> > On Thu 17 Sep 02:41 CDT 2020, Amit Pundir wrote:
> >
> > > Workaround to get WiFi working on Xiaomi Poco F1 (sdm845)
> > > phone. We get a non-fatal QMI_ERR_MALFORMED_MSG_V01 error
> > > message in ath10k_qmi_host_cap_send_sync(), but we can still
> > > bring up WiFi services successfully on AOSP if we ignore it.
> > >
> > > We suspect either the host cap is not implemented or there
> > > may be firmware specific issues. Firmware version is
> > > QC_IMAGE_VERSION_STRING=WLAN.HL.2.0.c3-00257-QCAHLSWMTPLZ-1
> > >
> > > qcom,snoc-host-cap-8bit-quirk didn't help. If I use this
> > > quirk, then the host capability request does get accepted,
> > > but we run into fatal "msa info req rejected" error and
> > > WiFi interface doesn't come up.
> > >
> >
> > What happens if you skip sending the host-cap message? I had one
> > firmware version for which I implemented a
> > "qcom,snoc-host-cap-skip-quirk".
> >
> > But testing showed that the link was pretty unusable - pushing any real
> > amount of data would cause it to silently stop working - and I realized
> > that I could use the linux-firmware wlanmdsp.mbn instead, which works
> > great on all my devices...
> 
> I skipped the ath10k_qmi_host_cap_send_sync block altogether
> (if that is what you meant by qcom,snoc-host-cap-skip-quirk) and
> so far did not run into any issues with youtube auto-playback loop
> (3+ hours and counting). Does that count as a valid use case?
> Otherwise let me know how could I reproduce a reasonable test
> setup?
> 

Iirc I was able to get an IP but browsing the web would be enough
traffic to stop (without any visible faults from the driver).

So your test sounds good I would like to see a host-cap-skip quirk,
rather than a conditional on the machine compatible.

> >
> > > Attempts are being made to debug the failure reasons but no
> > > luck so far. Hence this device specific workaround instead
> > > of checking for QMI_ERR_MALFORMED_MSG_V01 error message.
> > > Tried ath10k/WCN3990/hw1.0/wlanmdsp.mbn from the upstream
> > > linux-firmware project but it didn't help and neither did
> > > building board-2.bin file from stock bdwlan* files.
> > >
> >
> > "Didn't work" as in the wlanmdsp.mbn from linux-firmware failed to load
> > or some laer problem?
> 
> While using the wlanmdsp.mbn from linux-firmware, I run into
> the following crash 4 times before tqftpserv service gets killed
> eventually:
> 
> [   46.504502] qcom-q6v5-mss 4080000.remoteproc: fatal error received:
> dog_virtual_root.c:89:User-PD grace timer expired for wlan_process
> (ASID: 1)

It loaded, but doesn't seem to come up properly. We can try to debug
this further, but I think getting the quirk in will be useful - as there
seems to be a generation of firmware that has this particular behavior.

Regards,
Bjorn