netdev - Re: [PATCH net-next v3 06/10] net: ethtool: Add infrastructure for reporting cable test results

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 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:   Sun, 10 May 2020 11:00:13 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     Andrew Lunn <andrew@...n.ch>
Cc:     Michal Kubecek <mkubecek@...e.cz>, netdev@...r.kernel.org,
        David Miller <davem@...emloft.net>,
        Florian Fainelli <f.fainelli@...il.com>,
        Heiner Kallweit <hkallweit1@...il.com>,
        Chris Healy <cphealy@...il.com>, michael@...le.cc
Subject: Re: [PATCH net-next v3 06/10] net: ethtool: Add infrastructure for
 reporting cable test results

On Sun, 10 May 2020 18:07:58 +0200 Andrew Lunn wrote:
> On Sun, May 10, 2020 at 05:12:26PM +0200, Michal Kubecek wrote:
> > On Sat, May 09, 2020 at 06:28:47PM +0200, Andrew Lunn wrote:  
> > > Provide infrastructure for PHY drivers to report the cable test
> > > results.  A netlink skb is associated to the phydev. Helpers will be
> > > added which can add results to this skb. Once the test has finished
> > > the results are sent to user space.
> > > 
> > > When netlink ethtool is not part of the kernel configuration stubs are
> > > provided. It is also impossible to trigger a cable test, so the error
> > > code returned by the alloc function is of no consequence.
> > > 
> > > v2:
> > > Include the status complete in the netlink notification message
> > > 
> > > Signed-off-by: Andrew Lunn <andrew@...n.ch>  
> > 
> > Reviewed-by: Michal Kubecek <mkubecek@...e.cz>
> > 
> > It seems you applied the changes to ethnl_cable_test_alloc() suggested
> > in v2 review as part of patch 7 rather than here. I don't think it's
> > necessary to fix that unless there is some actual problem that would
> > require a resubmit.  
> 
> Hi Michal
> 
> Yes, squashed it into the wrong patch. But since all it does it change
> one errno for another, it is unlikely to break bisect. As i agree, we
> can live with this.

Sorry Andrew, would you mind doing one more quick spin? :(

Apart from what Michal pointed out there is a new line added after
ETHTOOL_A_TSINFO_MAX in patch 3 and removed in patch 4. 

More importantly we should not use the ENOTSUPP error code, AFAIU it's
for NFS, it's not a standard error code and user space struggles to
translate it with strerror(). Would you mind replacing all ENOTSUPPs
with EOPNOTSUPPs?