Received-SPF: pass (google.com: domain of linux-wireless-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 27 Sep 2022 19:35:35 +0200
From:   Greg KH <gregkh@linuxfoundation.org>
To:     Larry Finger <Larry.Finger@lwfinger.net>
Cc:     LKML <linux-kernel@vger.kernel.org>,
        linux-wireless <linux-wireless@vger.kernel.org>
Subject: Re: Problem with modprobe
Message-ID: <YzM0Z4gXsjZTDkX0@kroah.com>
References: <6c1ca180-8df7-c4d0-bd9f-d6b70f6c5142@lwfinger.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <6c1ca180-8df7-c4d0-bd9f-d6b70f6c5142@lwfinger.net>
Precedence: bulk

On Tue, Sep 27, 2022 at 12:11:24PM -0500, Larry Finger wrote:
> Hi,
> 
> The driver for the Realtek RTL8821CE network card show some asymmetrical
> behavior between loading and unloading.
> 
> Starting from scratch, a load command provides the expected results:
> 
> finger@localhost:~/rtw88>sudo modprobe -v rtw88_8821ce
> insmod /lib/modules/6.1.0-rc3-wireless-next-00623-g2bc09edc7fa2-dirty/kernel/drivers/net/wireless/realtek/rtw88/rtw88_core.ko
> 
> insmod /lib/modules/6.1.0-rc3-wireless-next-00623-g2bc09edc7fa2-dirty/kernel/drivers/net/wireless/realtek/rtw88/rtw88_pci.ko
> 
> insmod /lib/modules/6.1.0-rc3-wireless-next-00623-g2bc09edc7fa2-dirty/kernel/drivers/net/wireless/realtek/rtw88/rtw88_8821c.ko
> 
> insmod /lib/modules/6.1.0-rc3-wireless-next-00623-g2bc09edc7fa2-dirty/kernel/drivers/net/wireless/realtek/rtw88/rtw88_8821ce.ko
> 
> Unloading produces a different result:
> 
> finger@localhost:~/rtw88>sudo modprobe -rv rtw88_8821ce
> rmmod rtw88_8821ce
> rmmod rtw88_pci
> rmmod rtw88_8821c
> 
> Note that rtw88_core is still loaded and needs to be unloaded separately:
> 
> finger@localhost:~/rtw88>sudo modprobe -rv rtw88_core
> rmmod rtw88_core
> 
> The appropriate lines from modules.dep reformatted for clarity are as follows:
> 
> kernel/drivers/net/wireless/realtek/rtw88/rtw88_core.ko:
>    kernel/net/mac80211/mac80211.ko
>    kernel/lib/crypto/libarc4.ko
>    kernel/net/wireless/cfg80211.ko
> kernel/drivers/net/wireless/realtek/rtw88/rtw88_8821c.ko:
>    kernel/drivers/net/wireless/realtek/rtw88/rtw88_core.ko
>    kernel/net/mac80211/mac80211.ko
>    kernel/lib/crypto/libarc4.ko
>    kernel/net/wireless/cfg80211.ko
> kernel/drivers/net/wireless/realtek/rtw88/rtw88_8821ce.ko:
>    kernel/drivers/net/wireless/realtek/rtw88/rtw88_8821c.ko
>    kernel/drivers/net/wireless/realtek/rtw88/rtw88_pci.ko
>    kernel/drivers/net/wireless/realtek/rtw88/rtw88_core.ko
>    kernel/net/mac80211/mac80211.ko
>    kernel/lib/crypto/libarc4.ko
>    kernel/net/wireless/cfg80211.ko
> kernel/drivers/net/wireless/realtek/rtw88/rtw88_pci.ko:
>    kernel/drivers/net/wireless/realtek/rtw88/rtw88_core.ko
>    kernel/net/mac80211/mac80211.ko
>    kernel/lib/crypto/libarc4.ko
>    kernel/net/wireless/cfg80211.ko
> 
> What is the explanation for the failure to auto-unload rtw88_core?

I don't think we ever auto-unload dependant modules as we don't know if
they are still in use or not (hint, module references is not a
representation if the module is in use or not.)

So this is working as intended, right?  And how it's always worked?

thanks,

greg k-h