will send a v5 shortly without this patch: question below.
about this patch: this is in prevision of a world where INET can be
compiled out. So it is not something that matters today with current
kernels.
Now, as you mentioned in another patch, the only socket that survives
various reasonable CONFIG_* gymnastics is netlink. So even though
non-IPv4 kernels with IP support is not feasible today, I believe
there is some logic to using netlink sockets for ethtool purposes,
instead of IPv4 or IPv6 sockets. Shall I propose a patch to add
ethtool support on AF_NETLINK sockets, and update the tool to try
AF_INET first for backward compatibility reasons, then fallback to
AF_NETLINK?
On Sun, Mar 13, 2016 at 10:24 AM, Ben Hutchings <b...@decadent.org.uk> wrote:
> On Fri, 2016-03-11 at 09:58 -0800, David Decotigny wrote:
>> From: David Decotigny <de...@googlers.com>
>>
>>
>> Signed-off-by: David Decotigny <de...@googlers.com>
>> ---
>> ethtool.c | 3 +++
>> 1 file changed, 3 insertions(+)
>>
>> diff --git a/ethtool.c b/ethtool.c
>> index 761252f..f9336e3 100644
>> --- a/ethtool.c
>> +++ b/ethtool.c
>> @@ -4615,6 +4615,9 @@ opt_found:
>> /* Open control socket. */
>> ctx.fd = socket(AF_INET, SOCK_DGRAM, 0);
>> if (ctx.fd < 0) {
>> + ctx.fd = socket(AF_UNIX, SOCK_DGRAM, 0);
>> + }
>
> You still haven't answered whether this is a real problem on Linux.
>
> Ben.
>
>> + if (ctx.fd < 0) {
>> perror("Cannot get control socket");
>> return 70;
>> }
> --
> Ben Hutchings
> If at first you don't succeed, you're doing about average.
