Subject: Re: [PATCH] brcmfmac: run firmware state watchdog on the host machine
To:     =?UTF-8?B?UmFmYcWCIE1pxYJlY2tp?= <zajec5@gmail.com>,
        Kalle Valo <kvalo@codeaurora.org>
Cc:     linux-wireless@vger.kernel.org,
        brcm80211-dev-list.pdl@broadcom.com,
        brcm80211-dev-list@cypress.com,
        =?UTF-8?B?UmFmYcWCIE1pxYJlY2tp?= <rafal@milecki.pl>
References: <20190222170739.14266-1-zajec5@gmail.com>
From:   Arend Van Spriel <arend.vanspriel@broadcom.com>
Message-ID: <751f8cc0-3ec2-9aad-2c21-c52d5098272a@broadcom.com>
Date:   Sat, 23 Feb 2019 00:06:27 +0100
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.5.1
MIME-Version: 1.0
In-Reply-To: <20190222170739.14266-1-zajec5@gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Sender: linux-wireless-owner@vger.kernel.org
Precedence: bulk

On 2/22/2019 6:07 PM, Rafał Miłecki wrote:
> From: Rafał Miłecki <rafal@milecki.pl>
> 
> FullMAC firmware may happen to crash due to some potential bugs exposed
> by e.g. a specific traffic or host-requested setup. It usually results
> in various timeouts & running our of resources (e.g. ring slots).
> 
> Monitoring firmware state allows handling such a situation more
> gracefully. At this point the watchdog:
> 1) Prints a clear error message about a firmware crash
> 2) Tries to dump a crash info data

Hi Rafał,

I like the idea of having firmware crashes detected, but not sure I am a 
big fan of the watchdog mechanism.

> That should be helpful for users & should allow providing a valuable
> reports for the Broadcom developers. It obviously doesn't really fix
> anything.

Agree on the value of it and it is new functionality indeed and not a fix.

> This watchdog is important for USB & SDIO devices which don't have
> anything alike implemented yet. It's also there to complement PCIe with
> its FWHALT signal which does not work in all situations.

Actually, SDIO does already have a watchdog, but it serves different 
purpose (get firmware console, idle detect). There is also a 
brcmf_sdio_checkdied() function to detect a firmware crash. It is being 
called in brcmf_sdio_bus_rxctl() when firwmare does not respond as 
expected [1]. Similar thing can be done for PCIE and I actually started 
working on that last week. USB is a bit different as it will switch to 
bootloader image when firmware crashes.

Regards,
Arend

[1] 
https://elixir.bootlin.com/linux/latest/source/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c#L3144