Hello,
On Thu, 15 Oct 2015, lucien xin wrote:
> yeah, I don't understand why err > 0 is necessary to set IPSKB_DOREDIRECT
> to send redirects.
> FIB_RES_NH(res).nh_scope >= RT_SCOPE_HOST, what's that mean?
It tells us that packet comes from remote address that
we can reach direc
On Wed, Oct 14, 2015 at 3:24 AM, Julian Anastasov wrote:
>
> Hello,
>
> On Wed, 14 Oct 2015, Xin Long wrote:
>
>> commit fe3edf45792a ("ipv4: Remove all RTCF_DIRECTSRC handliing.").
>>
>> Before this commit, we had a tristate:
>> < 0: error
>> 0: ok
>> 1: ok and set RTCF_DIRECTSRC
>>
>> Bu
Hello,
On Wed, 14 Oct 2015, Xin Long wrote:
> commit fe3edf45792a ("ipv4: Remove all RTCF_DIRECTSRC handliing.").
>
> Before this commit, we had a tristate:
> < 0: error
> 0: ok
> 1: ok and set RTCF_DIRECTSRC
>
> But now we only care about 0 or < 0.
Not true, __mkroute_input()
Hi Xin,
[auto build test WARNING on net-next/master -- if it's inappropriate base,
please suggest rules for selecting the more suitable base]
url:
https://github.com/0day-ci/linux/commits/Xin-Long/route-fib_validate_source-remove-the-RT_SCOPE_HOST-test/20151014-004206
config: cris-etrax-100l
commit fe3edf45792a ("ipv4: Remove all RTCF_DIRECTSRC handliing.").
Before this commit, we had a tristate:
< 0: error
0: ok
1: ok and set RTCF_DIRECTSRC
But now we only care about 0 or < 0.
Signed-off-by: Xin Long
---
net/ipv4/fib_frontend.c | 14 +++---
1 file changed, 3 insertions(+)
