netdev - Re: [PATCH net-next 0/8] net: Extend availability of PHY statistics

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: <98af5159-e53d-72a3-c5ec-6c99c347d4ea@gmail.com>
Date:   Fri, 27 Apr 2018 10:27:38 -0700
From:   Florian Fainelli <f.fainelli@...il.com>
To:     David Miller <davem@...emloft.net>
Cc:     netdev@...r.kernel.org, andrew@...n.ch,
        vivien.didelot@...oirfairelinux.com, cphealy@...il.com,
        nikita.yoush@...entembedded.com
Subject: Re: [PATCH net-next 0/8] net: Extend availability of PHY statistics



On 04/27/2018 10:19 AM, Florian Fainelli wrote:
> 
> 
> On 04/27/2018 10:13 AM, David Miller wrote:
>> From: David Miller <davem@...emloft.net>
>> Date: Fri, 27 Apr 2018 12:10:14 -0400 (EDT)
>>
>>> Ugh, this breaks the allmodconfig build.
>>>
>>> You can't move interfaces to drivers/net/phy/phy.c which are
>>> referenced in core networking like net/core/ethtool.c unless you make
>>> PHYLIB 'bool' only.
>>>
>>> net/core/ethtool.o: In function `__ethtool_get_sset_count':
>>> ethtool.c:(.text+0x1cd8): undefined reference to `phy_ethtool_get_sset_count'
>>> net/core/ethtool.o: In function `dev_ethtool':
>>> (.text+0x5407): undefined reference to `phy_ethtool_get_strings'
>>> (.text+0x6869): undefined reference to `phy_ethtool_get_sset_count'
>>> (.text+0x69b3): undefined reference to `phy_ethtool_get_stats'
>>>
>>> Please fix this somehow, thanks Florian.
>>
>> I'm applying this temporary fix to net-next so that I can get work
>> down without having to revert all of your changes.
> 
> I forgot to put a static keyword in front of the functions declared
> within the else part of the #IS_ENABLED(CONFIG_PHYLIB) in
> include/linux/phy.h, will submit a fix for that immediately, *sigh*.

Still not enough coffee, what I am suggesting is actually to make the
body of the functions be static inline functions within
include/net/phy.h that way they get inlined into net/core/ethtool.c and
we can keep PHYLIB modular, does that look acceptable?

> 
>>
>> ====================
>> [PATCH] phy: Temporary build fix after phylib changes.
>>
>> Make PHYLIB boolean, because we reference phylib provided symbols now
>> from net/core/ethtool.c and therefore 'm' doesn't work.
>>
>> Signed-off-by: David S. Miller <davem@...emloft.net>
>> ---
>>  drivers/net/phy/Kconfig | 3 +--
>>  1 file changed, 1 insertion(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/phy/Kconfig b/drivers/net/phy/Kconfig
>> index edb8b9ab827f..7c5e8c1e9370 100644
>> --- a/drivers/net/phy/Kconfig
>> +++ b/drivers/net/phy/Kconfig
>> @@ -9,7 +9,6 @@ menuconfig MDIO_DEVICE
>>  
>>  config MDIO_BUS
>>  	tristate
>> -	default m if PHYLIB=m
>>  	default MDIO_DEVICE
>>  	help
>>  	  This internal symbol is used for link time dependencies and it
>> @@ -171,7 +170,7 @@ config PHYLINK
>>  	  autonegotiation modes.
>>  
>>  menuconfig PHYLIB
>> -	tristate "PHY Device support and infrastructure"
>> +	bool "PHY Device support and infrastructure"
>>  	depends on NETDEVICES
>>  	select MDIO_DEVICE
>>  	help
>>
> 

-- 
Florian