re: atl1c: Atheros L1C Gigabit Ethernet driver

re: atl1c: Atheros L1C Gigabit Ethernet driver

[Dan Carpenter]

Hello Jie Yang,

The patch 43250ddd75a3: "atl1c: Atheros L1C Gigabit Ethernet driver"
from Feb 18, 2009, leads to the following Smatch warning:
"drivers/net/ethernet/atheros/atl1c/atl1c_hw.c:472 atl1c_phy_setup_adv()
	 warn: odd binop '0x300 & 0xfffffffffffffcff'"

[ This check as too many false positives so I haven't released it ]

drivers/net/ethernet/atheros/atl1c/atl1c_hw.c
   472          u16 mii_giga_ctrl_data = GIGA_CR_1000T_DEFAULT_CAP &
   473                                  ~GIGA_CR_1000T_SPEED_MASK;

The defines are set up like this:

#define GIGA_CR_1000T_SPEED_MASK        0x0300
#define GIGA_CR_1000T_DEFAULT_CAP       0x0300

So we're just setting mii_giga_ctrl_data to zero.  It seems odd.  Did
you intend to do the bitwise negate of GIGA_CR_1000T_SPEED_MASK?
They're only used one time so I can't tell from the context what was
intented.

regards,
dan carpenter

Re: atl1c: Atheros L1C Gigabit Ethernet driver

[Dan Carpenter]

Jie Yang's email is bouncing.  2009 was a long time ago.  Jay,
Chris, any ideas?

regards,
dan carpenter

On Fri, Feb 15, 2013 at 01:34:28PM +0300, Dan Carpenter wrote:
> Hello Jie Yang,
> 
> The patch 43250ddd75a3: "atl1c: Atheros L1C Gigabit Ethernet driver"
> from Feb 18, 2009, leads to the following Smatch warning:
> "drivers/net/ethernet/atheros/atl1c/atl1c_hw.c:472 atl1c_phy_setup_adv()
> 	 warn: odd binop '0x300 & 0xfffffffffffffcff'"
> 
> [ This check as too many false positives so I haven't released it ]
> 
> drivers/net/ethernet/atheros/atl1c/atl1c_hw.c
>    472          u16 mii_giga_ctrl_data = GIGA_CR_1000T_DEFAULT_CAP &
>    473                                  ~GIGA_CR_1000T_SPEED_MASK;
> 
> The defines are set up like this:
> 
> #define GIGA_CR_1000T_SPEED_MASK        0x0300
> #define GIGA_CR_1000T_DEFAULT_CAP       0x0300
> 
> So we're just setting mii_giga_ctrl_data to zero.  It seems odd.  Did
> you intend to do the bitwise negate of GIGA_CR_1000T_SPEED_MASK?
> They're only used one time so I can't tell from the context what was
> intented.
> 
> regards,
> dan carpenter

Re: atl1c: Atheros L1C Gigabit Ethernet driver

[J. K. Cliburn]

Xiong,

Is Jie still an active maintainer of atl1c?

Thanks,
Jay

On 02/15/2013 05:12 AM, Dan Carpenter wrote:
> Jie Yang's email is bouncing.  2009 was a long time ago.  Jay,
> Chris, any ideas?
>
> regards,
> dan carpenter
>
> On Fri, Feb 15, 2013 at 01:34:28PM +0300, Dan Carpenter wrote:
>> Hello Jie Yang,
>>
>> The patch 43250ddd75a3: "atl1c: Atheros L1C Gigabit Ethernet driver"
>> from Feb 18, 2009, leads to the following Smatch warning:
>> "drivers/net/ethernet/atheros/atl1c/atl1c_hw.c:472 atl1c_phy_setup_adv()
>> 	 warn: odd binop '0x300 & 0xfffffffffffffcff'"
>>
>> [ This check as too many false positives so I haven't released it ]
>>
>> drivers/net/ethernet/atheros/atl1c/atl1c_hw.c
>>     472          u16 mii_giga_ctrl_data = GIGA_CR_1000T_DEFAULT_CAP &
>>     473                                  ~GIGA_CR_1000T_SPEED_MASK;
>>
>> The defines are set up like this:
>>
>> #define GIGA_CR_1000T_SPEED_MASK        0x0300
>> #define GIGA_CR_1000T_DEFAULT_CAP       0x0300
>>
>> So we're just setting mii_giga_ctrl_data to zero.  It seems odd.  Did
>> you intend to do the bitwise negate of GIGA_CR_1000T_SPEED_MASK?
>> They're only used one time so I can't tell from the context what was
>> intented.
>>
>> regards,
>> dan carpenter

RE: atl1c: Atheros L1C Gigabit Ethernet driver

[Huang, Xiong]

No, he has left this company.  Can you please replace it with nic-devel@qca.qualcomm.com ? thanks a lot !

-Xiong

> -----Original Message-----
> From: J. K. Cliburn [mailto:jcliburn@gmail.com]
> Sent: Tuesday, February 19, 2013 5:54 AM
> To: Dan Carpenter
> Cc: Chris Snook; kernel-janitors@vger.kernel.org; Huang, Xiong
> Subject: Re: atl1c: Atheros L1C Gigabit Ethernet driver
> 
> Xiong,
> 
> Is Jie still an active maintainer of atl1c?
> 
> Thanks,
> Jay
> 
> On 02/15/2013 05:12 AM, Dan Carpenter wrote:
> > Jie Yang's email is bouncing.  2009 was a long time ago.  Jay, Chris,
> > any ideas?
> >
> > regards,
> > dan carpenter
> >
> > On Fri, Feb 15, 2013 at 01:34:28PM +0300, Dan Carpenter wrote:
> >> Hello Jie Yang,
> >>
> >> The patch 43250ddd75a3: "atl1c: Atheros L1C Gigabit Ethernet driver"
> >> from Feb 18, 2009, leads to the following Smatch warning:
> >> "drivers/net/ethernet/atheros/atl1c/atl1c_hw.c:472
> atl1c_phy_setup_adv()
> >> 	 warn: odd binop '0x300 & 0xfffffffffffffcff'"
> >>
> >> [ This check as too many false positives so I haven't released it ]
> >>
> >> drivers/net/ethernet/atheros/atl1c/atl1c_hw.c
> >>     472          u16 mii_giga_ctrl_data = GIGA_CR_1000T_DEFAULT_CAP &
> >>     473                                  ~GIGA_CR_1000T_SPEED_MASK;
> >>
> >> The defines are set up like this:
> >>
> >> #define GIGA_CR_1000T_SPEED_MASK        0x0300
> >> #define GIGA_CR_1000T_DEFAULT_CAP       0x0300
> >>
> >> So we're just setting mii_giga_ctrl_data to zero.  It seems odd.  Did
> >> you intend to do the bitwise negate of GIGA_CR_1000T_SPEED_MASK?
> >> They're only used one time so I can't tell from the context what was
> >> intented.
> >>
> >> regards,
> >> dan carpenter