netdev - Re: [PATCH] Use a spinlock to guard `fep->ptp_clk

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]

Date:   Wed, 31 Aug 2022 16:04:10 +0200
From:   Csókás Bence <csokas.bence@...lan.hu>
To:     Richard Cochran <richardcochran@...il.com>
CC:     <netdev@...r.kernel.org>, "David S. Miller" <davem@...emloft.net>,
        "Jakub Kicinski" <kuba@...nel.org>, Andrew Lunn <andrew@...n.ch>,
        <kernel@...gutronix.de>, Marc Kleine-Budde <mkl@...gutronix.de>
Subject: Re: [PATCH] Use a spinlock to guard `fep->ptp_clk_on`


On 2022. 08. 31. 15:54, Richard Cochran wrote:
> On Wed, Aug 31, 2022 at 02:56:31PM +0200, Csókás Bence wrote:
> 
>> diff --git a/drivers/net/ethernet/freescale/fec_ptp.c b/drivers/net/ethernet/freescale/fec_ptp.c
>> index c74d04f4b2fd..dc8564a1f2d2 100644
>> --- a/drivers/net/ethernet/freescale/fec_ptp.c
>> +++ b/drivers/net/ethernet/freescale/fec_ptp.c
>> @@ -365,21 +365,21 @@ static int fec_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta)
>>    */
>>   static int fec_ptp_gettime(struct ptp_clock_info *ptp, struct timespec64 *ts)
>>   {
>> -	struct fec_enet_private *adapter =
>> +	struct fec_enet_private *fep =
>>   	    container_of(ptp, struct fec_enet_private, ptp_caps);
>>   	u64 ns;
>> -	unsigned long flags;
>> +	unsigned long flags, flags2;
>>   
>> -	mutex_lock(&adapter->ptp_clk_mutex);
>> +	spin_lock_irqsave(&fep->ptp_clk_lock, flags);
>>   	/* Check the ptp clock */
>> -	if (!adapter->ptp_clk_on) {
>> -		mutex_unlock(&adapter->ptp_clk_mutex);
>> +	if (!fep->ptp_clk_on) {
> 
> BTW This test is silly. If functionality isn't available then the code
> should simply not register the clock in the first place.

As I understand, `ptp_clk_on` is a flag indicating whether `clk_ipg` is 
running or not, and not a capability thing. The driver switches it 
run-time in `fec_enet_clk_enable()`.

> 
>> +		spin_unlock_irqrestore(&fep->ptp_clk_lock, flags);
>>   		return -EINVAL;
>>   	}
>> -	spin_lock_irqsave(&adapter->tmreg_lock, flags);
>> -	ns = timecounter_read(&adapter->tc);
>> -	spin_unlock_irqrestore(&adapter->tmreg_lock, flags);
>> -	mutex_unlock(&adapter->ptp_clk_mutex);
>> +	spin_lock_irqsave(&fep->tmreg_lock, flags2);
>> +	ns = timecounter_read(&fep->tc);
>> +	spin_unlock_irqrestore(&fep->tmreg_lock, flags2);
>> +	spin_unlock_irqrestore(&fep->ptp_clk_lock, flags);
> 
> Two spin locks?  Why not just use one?

One guards the FEC_ATIME_* registers, the other the `ptp_clk_on` flag. 
For instance, `fec_ptp_adjfreq()` takes `tmreg_lock` but not 
`ptp_clk_lock`, and similarly `fec_enet_clk_enable()` takes 
`ptp_clk_lock` but not `tmreg_lock`.

> 
> Thanks,
> Richard