Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754434AbaLHTwS (ORCPT ); Mon, 8 Dec 2014 14:52:18 -0500 Received: from mail-wg0-f50.google.com ([74.125.82.50]:58984 "EHLO mail-wg0-f50.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752066AbaLHTwP (ORCPT ); Mon, 8 Dec 2014 14:52:15 -0500 From: Pali =?utf-8?q?Roh=C3=A1r?= To: Marcel Holtmann Subject: Re: wl1251: NVS firmware data Date: Mon, 8 Dec 2014 20:52:11 +0100 User-Agent: KMail/1.13.7 (Linux/3.18.0-031800rc5-generic; KDE/4.14.1; x86_64; ; ) Cc: "Greg Kroah-Hartman" , Ming Lei , Pavel Machek , "John W. Linville" , Grazvydas Ignotas , "linux-wireless@vger.kernel.org" , Network Development , Linux Kernel Mailing List , Ivaylo Dimitrov , Aaro Koskinen , Kalle Valo , Sebastian Reichel , David Gnedt References: <201411271506.20457@pali> <201412082015.18501@pali> <9E6DD65F-8AF4-49C6-85EE-AA2BB4C6C39D@holtmann.org> In-Reply-To: <9E6DD65F-8AF4-49C6-85EE-AA2BB4C6C39D@holtmann.org> MIME-Version: 1.0 Content-Type: multipart/signed; boundary="nextPart8923115.Zzh4vTgOgY"; protocol="application/pgp-signature"; micalg=pgp-sha1 Content-Transfer-Encoding: 7bit Message-Id: <201412082052.12049@pali> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org --nextPart8923115.Zzh4vTgOgY Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable On Monday 08 December 2014 20:41:13 Marcel Holtmann wrote: > Hi Pali, >=20 > >>>>>>>> On Saturday 06 December 2014 13:49:54 Pavel Machek > >>>>>>>> wrote: /** > >>>>>>>>=20 > >>>>>>>> + * request_firmware_prefer_user: - prefer usermode > >>>>>>>> helper for loading firmware + * @firmware_p: pointer > >>>>>>>> to firmware image > >>>>>>>> + * @name: name of firmware file > >>>>>>>> + * @device: device for which firmware is being > >>>>>>>> loaded + * > >>>>>>>> + * This function works pretty much like > >>>>>>>> request_firmware(), but it prefer + * usermode > >>>>>>>> helper. If usermode helper fails then it fallback to > >>>>>>>> direct access. + * Usefull for dynamic or model > >>>>>>>> specific firmware data. + **/ > >>>>>>>> +int request_firmware_prefer_user(const struct > >>>>>>>> firmware **firmware_p, + =20 > >>>>>>>> const char *name, struct device *device) +{ > >>>>>>>> + int ret; > >>>>>>>> + __module_get(THIS_MODULE); > >>>>>>>> + ret =3D _request_firmware(firmware_p, name, > >>>>>>>> device, + FW_OPT_UEVENT > >>>>>>>>=20 > >>>>>>>> | FW_OPT_PREFER_USER); + > >>>>>>>>=20 > >>>>>>>> module_put(THIS_MODULE); + return ret; > >>>>>>>> +} > >>>>>>>> +EXPORT_SYMBOL_GPL(request_firmware_prefer_user); > >>>>>>>=20 > >>>>>>> I'd like to introduce request_firmware_user() which > >>>>>>> only requests firmware from user space, and this way > >>>>>>> is simpler and more flexible since we have > >>>>>>> request_firmware_direct() already. > >>>>>>=20 > >>>>>> Why would a driver care about what program provides the > >>>>>> firmware? It shouldn't at all, and we want to get rid > >>>>>> of the userspace firmware loader, not encourage > >>>>>> drivers to use it "exclusively" at all. > >>>>>=20 > >>>>> Do not remove it! Without userspace firmware loader it > >>>>> is impossible to load dynamic firmware files. > >>>>=20 > >>>> why is this dynamic in the first place. It does not sound > >>>> like dynamic data to me at all. This is like the WiFi MAC > >>>> address(es) or Bluetooth BD_ADDR. They are all static > >>>> information. The only difference is that they are on the > >>>> host accessibly filesystem or storage and not on the > >>>> device itself. > >>>>=20 > >>>> To be honest, for Bluetooth we solved this now. If the > >>>> device is missing key information like the calibration > >>>> data or BD_ADDR, then it comes up unconfigured. A > >>>> userspace process can then go and load the right data > >>>> into it and then the device becomes available as > >>>> Bluetooth device. > >>>>=20 > >>>> Trying to use request_firmware to load some random data > >>>> and insist on going through userspace helper for that > >>>> sounds crazy to me. Especially since we are trying hard > >>>> to get away from the userspace loader. Forcing to keep > >>>> it for new stuff sounds backwards to me. > >>>>=20 > >>>> With the special Nokia partition in mind, why hasn't this > >>>> been turned into a mountable filesystem or into a > >>>> driver/subsystem that can access the data direct from the > >>>> kernel. I advocated for this some time ago. Maybe there > >>>> should be a special subsystem for access to these factory > >>>> persistent information that drivers then just can access. > >>>> I seem to remember that some systems provide these via > >>>> ACPI. Why does the ARM platform has to be special here? > >>>>=20 > >>>> And the problem of getting Ethernet and WiFi MAC address > >>>> and Bluetooth BD_ADDR comes up many many times. Why not > >>>> have something generic here. And don't tell me > >>>> request_firmware is that generic solution ;) > >>>>=20 > >>>> Regards > >>>>=20 > >>>> Marcel > >>>=20 > >>> Hi Marcel. I think you did not understand this problem. > >>> This discussion is not about mac address. Please read > >>> email thread again and if there are some unclear pars, > >>> then ask. Thanks! > >>=20 > >> I think that I pretty clearly understand the problem. > >> Calibration data, MAC address, what is the difference? For > >> me this is all the same. It is data that is specific to a > >> device or type of devices and it is stored somewhere else. > >> In most cases in some immutable memory/flash area. > >=20 > > Those calibration data (in form of binary NVS firmware file) > > needs to be sent to wl1251 chip. Mac address is not needed > > at this step (and kernel generate some random if is not > > provided). >=20 > the MAC address is just an example or similar data. And to be > clear the kernel generating some random address is not a good > idea either. If you get a new random address on every boot > that is total disaster. Because it sort of works does not > mean it is the right way to do it. That is why I am including > MAC address in the list here. It is same kind of data that is > needed before a device can be declared fully functional. >=20 > > (Just to note wl1271 driver loads both MAC address and NVS > > data via one firmware file which is prepared by userspace, > > but this discussion is about wl1251...) >=20 > There is no difference between any drivers here. I do not know > why are you trying to tie this to a specific driver. Why does > it matter what kind of information these are. The point is > they are not static, they are device specific and come from > different sources. And the kernel driver needs them. >=20 Right. > >> What you want is access to this data since the kernel > >> driver needs it. Do I get this so far ;) > >=20 > > Yes, we need to provide NVS data to kernel when kernel ask > > for them. > >=20 > >> So my take is that request_firmware is not the right way to > >> get this data. Or more precisely make sure that this data > >> is available to kernel drivers. And what I am seeing here > >> is that instead of actually solving the bigger problem, we > >> just hack around it with request_firmware. Now > >> surprisingly the request_firmware loads files directly > >> from the kernel and all the hacks do not work anymore. > >>=20 > >> Regards > >>=20 > >> Marcel > >=20 > > Just read emails again... > >=20 > > Our problem is: > >=20 > > linux-firmware.git tree provides two binary firmware files: > >=20 > > ti-connectivity/wl1251-fw.bin > > ti-connectivity/wl1251-nvs.bin > >=20 > > First is firmware file, second NVS file with generic > > calibration data. Kernel driver wl1251 now loads both > > firmware files via request_firmware. Generic calibration > > data are enough for wl1251 chip (it should work). But > > devices have own calibration data stored somewhere else. >=20 > Loading generic data that is static and stored on the > filesystem via request_firmware is totally fine. If you have > the NVS data in that file, then great. If you have specific > data, then overwrite the file, link it to the real file or do > something with it. As long as it is a file on the filesystem, > you will be just fine. >=20 I think that I should be able to rsync disk image of system and=20 then copy it into another device (of same type). So I do not want=20 to store device specific files into /lib/firmware/... > If you however want to hook into some magic userspace helper > to build the content of the file and somehow load it, then > that sounds like the wrong approach to me. >=20 If data depends on something from userspace (location of device=20 or type of sim card, etc, ...) then there is needed some=20 userspace program which prepare data. It cannot be automatically=20 done from kernel. > > On Nokia N900 NVS data are generated on-the-fly from some > > bytes from CAL (/dev/mtd1), from state of cellular network > > and from some other regulation settings. > >=20 > > So I think that files stored in linux-firmware.git tree > > (which are also installed into /lib/firmware/) should be > > loaded with request_firmware function. Or not? Do you think > > something else? What other developers think? > >=20 > > I'm against kernel driver for CAL (/dev/mtd1) for more > > reasons: > >=20 > > 1) we have userspace open source code, but licensed under > > GPLv3. And until kernel change license, we cannot include > > it. > >=20 > > 2) NVS data are (probably) not in one place, plus they > > depends on something other. > >=20 > > 3) If manufacture XYZ create new device with its own storage > > format of calibration data this means that correct solution > > for XYZ is also to implement new kernel fs driver for its > > own format. Do you really want to have in kernel all those > > drivers for all different (proprietary) storage formats? > >=20 > > 4) It does not help us with existence of generic file > > /lib/firmware/ti-connectivity/wl1251-nvs.bin which comes > > from linux-firmware.git tree. >=20 > As I said before, I think that a driver should not register > with its subsystem until it has all data that it needs. Or if > it can tell the subsystem that it is missing data and the > subsystem knows how to provide hooks for getting this data. >=20 > We all know that many embedded devices need extra data to > operation properly. This data is normally programmed onto the > device in the factory. So if someone would now build a > subsystem that can retrieve magic blobs of data from magic > places like ACPI, devicetree, userspace or whatever that > would help. I can see that request_firmware looks a lot like > this. However the reality is that you have a race condition > here. request_firmware relies on the fact that it is file in > userspace. That is what it was designed for in the first > place. Firmware files that are place on the hosts filesystem. > It does not have the option to start a notifier when blob xyz > becomes available. And that is what you essentially need for > the drivers. The driver finds the hardware and goes, now I > need blob xyz to function and then it sits and waits until it > gets told that blob is now available. Then it initializes the > hardware and registers it to the subsystem. >=20 > I fully realize that request_firmware is pretty close in this > regard, but the semantics and timing that many of these NVS > data like addresses, our calibration information are > different. It is more than just this specific drivers > problem. There are many devices out there that have certain > settings stored somewhere and it needs these based on how the > device is build or how it is provisioning in the factory. >=20 > What I would actually prefer to see that the driver just > requests this blob of information and then a separate > subsystem deals with getting it. As I said, in some cases the > information might be in ACPI or devicetree or accessible by a > special driver. In that case no userspace interaction would > be needed at all. However the driver has to deal with the > fact that the data blob might not be available for a certain > period of time. If you mention cellular modem, then that one > has to boot up first and get its data. More reason to > actually design this cleanly so that there are no race > conditions. >=20 > Regards >=20 > Marcel Yes, this sounds good. Some subsystem which reads needed data (or=20 generate them or wait until something other provide them) sounds=20 good. =2D-=20 Pali Roh=C3=A1r pali.rohar@gmail.com --nextPart8923115.Zzh4vTgOgY Content-Type: application/pgp-signature; name=signature.asc Content-Description: This is a digitally signed message part. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) iEYEABECAAYFAlSGAWwACgkQi/DJPQPkQ1KovgCguCcFSwrKfa23hW7ruKfgIQSl phUAnRaqNPmqKfGnbuRi3rblxXWmmWX5 =ZSx5 -----END PGP SIGNATURE----- --nextPart8923115.Zzh4vTgOgY-- -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/