linux-kernel - Re: [PATCH V2 2/5] tty: serial: fsl_lpuart: clear UARTCTRL_LOOPS in lpuart32

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: <45769b238521d4b41638406863e31ded@walle.cc>
Date:   Wed, 23 Nov 2022 12:42:47 +0100
From:   Michael Walle <michael@...le.cc>
To:     Sherry Sun <sherry.sun@....com>
Cc:     gregkh@...uxfoundation.org, jirislaby@...nel.org,
        jingchang.lu@...escale.com, tomonori.sakita@...d.co.jp,
        atsushi.nemoto@...d.co.jp, linux-serial@...r.kernel.org,
        linux-kernel@...r.kernel.org, dl-linux-imx <linux-imx@....com>
Subject: Re: [PATCH V2 2/5] tty: serial: fsl_lpuart: clear UARTCTRL_LOOPS in
 lpuart32_shutdown()

Hi Sherry,

Am 2022-11-23 12:30, schrieb Sherry Sun:
>> -----Original Message-----
>> From: Michael Walle <michael@...le.cc>
>> Sent: 2022年11月23日 19:09
>> To: Sherry Sun <sherry.sun@....com>
>> Cc: gregkh@...uxfoundation.org; jirislaby@...nel.org;
>> jingchang.lu@...escale.com; tomonori.sakita@...d.co.jp;
>> atsushi.nemoto@...d.co.jp; linux-serial@...r.kernel.org; linux-
>> kernel@...r.kernel.org; dl-linux-imx <linux-imx@....com>
>> Subject: Re: [PATCH V2 2/5] tty: serial: fsl_lpuart: clear 
>> UARTCTRL_LOOPS in
>> lpuart32_shutdown()
>> 
>> Am 2022-11-23 11:58, schrieb Sherry Sun:
>> >> -----Original Message-----
>> >> From: Michael Walle <michael@...le.cc>
>> >> Sent: 2022年11月23日 18:34
>> >> To: Sherry Sun <sherry.sun@....com>
>> >> Cc: gregkh@...uxfoundation.org; jirislaby@...nel.org;
>> >> jingchang.lu@...escale.com; tomonori.sakita@...d.co.jp;
>> >> atsushi.nemoto@...d.co.jp; linux-serial@...r.kernel.org; linux-
>> >> kernel@...r.kernel.org; dl-linux-imx <linux-imx@....com>
>> >> Subject: Re: [PATCH V2 2/5] tty: serial: fsl_lpuart: clear
>> >> UARTCTRL_LOOPS in
>> >> lpuart32_shutdown()
>> >>
>> >> Am 2022-11-10 09:17, schrieb Sherry Sun:
>> >> > UARTCTRL_LOOPS bit is set in lpuart32_set_mctrl() for loopback
>> >> > mode, but nowhere clear this bit, it should be cleared when closing
>> >> > the uart port to avoid the loopback mode been enabled by default
>> >> > when reopening the uart.
>> >>
>> >> It's cleared in set_mctrl(). What is the expectation from the serial
>> >> core here?
>> >>
>> >
>> > Hi Michael,
>> >
>> > If we call .set_mctrl(TIOCM_LOOP), the UARTCTRL_LOOPS will be set.
>> > Then when we call .shutdown(), serial core won't call .set_mctrl() to
>> > clear it, so the UARTCTRL_LOOPS need to be cleared here.
>> > Per my understanding, .shutdown() should clean up all the uart flags,
>> > as the transmitter and receiver will been disabled, we will
>> > re-configure all the settings needed when re-open the port.
>> 
>> Two things,
>> (1) should the loopback be cleared on a newly opened serial device?
>> (2) as mentioned in my other reply, this can also be handled in
>>      the startup. Eg. the startup can clear the loopback flag.
>>      (together with possible hardware events).
>> 
>> I'm not that deep into the serial core, thus my question about the
>> expectations from the serial core. I guess the answer to
>> (1) is yes, but better to ask.
>> 
> 
> Hi Michael,
> 
> For the (1), I have checked the serial core, seems the answer is no, .
> startup() won't clean the status, only when the uart device is probed,
> lpuart will do the global reset to all the registers instead of
> .startup(). So I think the uart running status cleared in .shutdown()
> is reasonable.

That's not what I've meant. Even with this patch as it is right now,
the loopback flag is cleared on a "newly opened serial device". Just
with one difference, you are clearing the flag in shutdown.

My question was rather, should the loopback (or generally any mctrl 
flags)
be persistent across close/open cycles. E.g. looking at omap-serial.c,
this driver doesn't seem to handle the loopback flag at .startup() or
.shutdown(). Same seems to be true for sh-sci.c.

Greg?

-michael