Hello,
About “surprising” the driver: -EINVAL means that firmware load
machinery is absent, while the rtl8192ce driver (as well as many others)
explicitly depend on it (select FW_LOADER), so that’s not possible in
any correct kernel configuration.
Anyway, driver maintainers answered. Well, I expected such answer. And
I’m afraid they’re right in that the device runs some firmware anyway;
that may be the firmware included on production or last loaded (or maybe
EFI loads it on boot? well, don’t know).
Actually, for a long time I wanted to ask: what’s the real point to
remove firmware from the kernel? That’s not security: complex devices
can’t work without a firmware, so either they don’t work at all or have
some sort of built-in firmware, that may pose all the same security
risks; that’s not the freedom, as unchangeable program on disk isn’t any
worse than unchangeable program in ROM; so what?
By the way, I tried the approach with callback calling, it seems it
doesn’t break anything but neither lets this driver (unaltered) to work.
And the modified driver works somewhat, often crashing the program that
tries to perform scan (tested wpa_supplicant and iw, both crash with
SEGFAULT). It seems that depends on the device state on boot/resume,
which can be different for some reason.
--
Lobachevskii Vitalii https://github.com/numberZero
On 18.08.2016 22:14, Alexandre Oliva wrote:
> Hi,
>
> Thanks for your report!
>
> On Aug 13, 2016, number Zero <[email protected]> wrote:
>
>> However, after I made it to ignore the EINVAL error from the
>> reject_firmware_nowait function (in
>> drivers/net/wireless/realtek/rtlwifi/rtl8192ce/sw.c), it works fine.
> Interesting. Can you check that this remains true even after a full
> power cycle, as in, that it's not a blob loaded by a previous blobbed
> kernel that remains it to work in your setting? (if you never had the
> blob around, or never used a blobbed kernel, just say so, and I'll be
> happy enough about the conclusion ;-)
>
>> Really, are there any reasons for reject_firmware_nowait to return
>> -EINVAL?
> Yes. It the error code to indicate to the caller that the firmware
> loading functionality is not avaialble. It indicates the callback
> supplied by the caller will not be called, so the caller itself should
> take care of e.g. returning any temporary memory the callback would have
> released.
>
> If the driver works even if request_firmware_nowait returns such an
> error, then it ought to tolerate this return code.
>
>> So shouldn’t the reject_firmware_nowait function behave as if the
>> requested firmware merely absent?
> Given the multiple cases in which drivers were "surprised" by this
> return code, I guess we could try to rework reject_firmware_nowait so as
> to actually call the callback, signalling the unsuccessful completion of
> the request. Would you like to give that a try?
>
> Or perhaps you'd prefer to report the bug to the rtl8192ce maintainers,
> so that they could fix their driver so as to work (as it should) even
> when the firmware loading configuration option is disabled?
>
> Thanks,
--- Begin Message ---
Larry Finger <[email protected]> writes:
> On 08/25/2016 08:17 AM, Jes Sorensen wrote:
>> Lobachevskii Vitalii <[email protected]> writes:
>> The realtek devices all require firmwere to operate correctly,
>> including the 8192c series. There are a bunch of commands flying back
>> and forth between the driver and the firmware.
>>
>> If your device happens to work without loading the firmware then you
>> have an old firmware blob loaded.
>>
>> At least this is the case for the USB version of the device, and I find
>> it highly unlikely the PCIe version is any different.
>>
>> Trying to remove the firmware loading error is just plain silly.
>
> I have not bothered my Realtek contacts with such a question, but I
> have a plausible explanation. If an RTL8192CE functions without
> loading external firmware, it is because the device has minimal
> function built in the default firmware. Mostly this rudimentary
> firmware is used to boot the device and to download the firmware for
> complex wireless communication. Rudimentary wifi functions would be
> needed for wake-on-lan operations. The fact that Realtek has never
> implemented WOL for the 8192C chips is highly suggestive that they do
> not function very well in that capacity.
>
> If the RTL8192CE actually runs without loading external firmware, then
> I am quite sure that it will be restricted to 802.11g at the most, and
> more likely 802.11b. If it handles any 802.11n capabilities, then that
> firmware will certainly not have any of the bug fixes applied to the
> firmware since the earliest release.
>
> You are certainly allowed to configure your system any way you want,
> but please do not send any such "fixes" to the kernel sources. They
> will NEVER be accepted!
>
> Configuring a kernel without firmware loading capacity is indeed silly.
Makes perfect sense, but as you also correctly point out, it means the
device actually is running some firmware, but we have no idea what state
or version it is. Trying to run a device with this level of firmware is
both risky and makes it hard to rely on correct operation.
Of course it also makes this whole ostrich process even more pointless,
since the device is in fact running firmware - pretending it isn't is
just silly.
One more thing, yes you can apply this patch to your own degraded
kernel, but if you ship it, kindle remove ALL email addresses of any
driver authors in it. None of us wants to waste our time on bug reports
because of this.
If you truly want to run your system without firmware, I hear there are
great bargains for NE2000 and 3c501 cards on ebay :)
Jes
--- End Message ---
--- Begin Message ---
On 08/25/2016 08:17 AM, Jes Sorensen wrote:
Lobachevskii Vitalii <[email protected]> writes:
Hello,
The RTL8192CE device seems to work fine without any firmware, so you may
make it fully optional, removing dependency on FW_LOADER. Of course that
require some patching, but if I understood the driver internals
correctly, simple complete(&rtlpriv->firmware_loading_complete); would
be enough when firmware loading machinery is unavailable, that is, when
request_firmware_nowait returns -EINVAL (currently that may only happen
in improperly configured or patched kernels, like Linux-libre; see
attached messages for more information)
Of course I will try to fix Linux-libre “deblobbing” technique, as it
should never break anything that may work without a firmware. But
anyway, if a device and its driver may work without a certain kernel
feature, that feature should not be selected, I think.
The realtek devices all require firmwere to operate correctly,
including the 8192c series. There are a bunch of commands flying back
and forth between the driver and the firmware.
If your device happens to work without loading the firmware then you
have an old firmware blob loaded.
At least this is the case for the USB version of the device, and I find
it highly unlikely the PCIe version is any different.
Trying to remove the firmware loading error is just plain silly.
I have not bothered my Realtek contacts with such a question, but I have a
plausible explanation. If an RTL8192CE functions without loading external
firmware, it is because the device has minimal function built in the default
firmware. Mostly this rudimentary firmware is used to boot the device and to
download the firmware for complex wireless communication. Rudimentary wifi
functions would be needed for wake-on-lan operations. The fact that Realtek has
never implemented WOL for the 8192C chips is highly suggestive that they do not
function very well in that capacity.
If the RTL8192CE actually runs without loading external firmware, then I am
quite sure that it will be restricted to 802.11g at the most, and more likely
802.11b. If it handles any 802.11n capabilities, then that firmware will
certainly not have any of the bug fixes applied to the firmware since the
earliest release.
You are certainly allowed to configure your system any way you want, but please
do not send any such "fixes" to the kernel sources. They will NEVER be accepted!
Configuring a kernel without firmware loading capacity is indeed silly.
Larry
--- End Message ---
--- Begin Message ---
Lobachevskii Vitalii <[email protected]> writes:
> Hello,
>
> The RTL8192CE device seems to work fine without any firmware, so you may
> make it fully optional, removing dependency on FW_LOADER. Of course that
> require some patching, but if I understood the driver internals
> correctly, simple complete(&rtlpriv->firmware_loading_complete); would
> be enough when firmware loading machinery is unavailable, that is, when
> request_firmware_nowait returns -EINVAL (currently that may only happen
> in improperly configured or patched kernels, like Linux-libre; see
> attached messages for more information)
>
> Of course I will try to fix Linux-libre “deblobbing” technique, as it
> should never break anything that may work without a firmware. But
> anyway, if a device and its driver may work without a certain kernel
> feature, that feature should not be selected, I think.
The realtek devices all require firmwere to operate correctly,
including the 8192c series. There are a bunch of commands flying back
and forth between the driver and the firmware.
If your device happens to work without loading the firmware then you
have an old firmware blob loaded.
At least this is the case for the USB version of the device, and I find
it highly unlikely the PCIe version is any different.
Trying to remove the firmware loading error is just plain silly.
Jes
--- End Message ---
_______________________________________________
linux-libre mailing list
[email protected]
http://www.fsfla.org/cgi-bin/mailman/listinfo/linux-libre
