linux-kernel - Re: [PATCH] Bluetooth: btusb: avoid unused function warning

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: <A91B8FA0-50F3-44FC-9466-F8569EA89C99@holtmann.org>
Date:   Thu, 19 Sep 2019 14:36:18 +0200
From:   Marcel Holtmann <marcel@...tmann.org>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     Johan Hedberg <johan.hedberg@...il.com>,
        Alex Lu <alex_lu@...lsil.com.cn>,
        Matthias Kaehlcke <mka@...omium.org>,
        Rajat Jain <rajatja@...gle.com>,
        Raghuram Hegde <raghuram.hegde@...el.com>,
        Bluez mailing list <linux-bluetooth@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Bluetooth: btusb: avoid unused function warning

Hi Arnd,

>>> The btusb_rtl_cmd_timeout() function is used inside of an
>>> ifdef, leading to a warning when this part is hidden
>>> from the compiler:
>>> 
>>> drivers/bluetooth/btusb.c:530:13: error: unused function 'btusb_rtl_cmd_timeout' [-Werror,-Wunused-function]
>>> 
>>> Use an IS_ENABLED() check instead so the compiler can see
>>> the code and then discard it silently.
>>> 
>>> Fixes: d7ef0d1e3968 ("Bluetooth: btusb: Use cmd_timeout to reset Realtek device")
>>> Signed-off-by: Arnd Bergmann <arnd@...db.de>
>>> ---
>>> drivers/bluetooth/btusb.c | 5 ++---
>>> 1 file changed, 2 insertions(+), 3 deletions(-)
>>> 
>>> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
>>> index a9c35ebb30f8..23e606aaaea4 100644
>>> --- a/drivers/bluetooth/btusb.c
>>> +++ b/drivers/bluetooth/btusb.c
>>> @@ -3807,8 +3807,8 @@ static int btusb_probe(struct usb_interface *intf,
>>>              btusb_check_needs_reset_resume(intf);
>>>      }
>>> 
>>> -#ifdef CONFIG_BT_HCIBTUSB_RTL
>>> -     if (id->driver_info & BTUSB_REALTEK) {
>>> +     if (IS_ENABLED(CONFIG_BT_HCIBTUSB_RTL) &&
>>> +         (id->driver_info & BTUSB_REALTEK)) {
>>>              hdev->setup = btrtl_setup_realtek;
>>>              hdev->shutdown = btrtl_shutdown_realtek;
>>>              hdev->cmd_timeout = btusb_rtl_cmd_timeout;
>>> @@ -3819,7 +3819,6 @@ static int btusb_probe(struct usb_interface *intf,
>>>               */
>>>              set_bit(BTUSB_WAKEUP_DISABLE, &data->flags);
>>>      }
>>> -#endif
>> 
>> I prefer that we stick another ifdef around the btusb_rtl_cmd_timeout function since that
>> is how we did it for the other vendors as well.
> 
> Ok. Can you just commit that with 'Reported-by: Arnd Bergmann <arnd@...db.de>'?
> 
>> However I start to wonder if we need all these vendor ifdef anyway. The vendor specific
>> functions should turn into empty stubs if their support is not selected.
> 
> It just wastes a little bit of object code space, which my approach
> above avoids.
> 
> I guess one could also be clever and redefine those macros like
> 
> #define BTUSB_REALTEK (IS_ENABLED(CONFIG_BT_HCIBTUSB_RTL) ? 0x20000 : 0)
> 
> so the if() section gets silently dropped, in addition to treating
> zero like BTUSB_IGNORE.

we could do that as well. I just want it to be consistent in the code.

Regards

Marcel