[LTP] [PATCH 1/1] ipneigh01: Replace TCONF error message with TINFO
Mon May 7 14:31:36 CEST 2018
On 05/07/2018 02:13 PM, Anton Smorodskyi wrote:
> Hi Alexey
> This is my first attempt to contribute to LTP project. Disadvantage of this is the fact that I can ask some obvious for you question ( hope you will forgive me that) , but advantage that I have a fresh look so I can point out to something that is wrong but you just get used too.
> First of all let's clarify something not related to this PR directly :
>>> And why it is more appropriate? TCONF is not an error message.
> In LTP documentation TCONF defined as - "The test case was not appropriate for the current hardware or software configuration". So I would not call it error too but I would say that it is some flavor of "SKIPPED" state.
> Where you want to say "this test case can't run currently so I can't say if it is passed or failed"
No, I don't.
> Second thing which needs to be taken in consideration I clarified with Petr Vorel in conversation outside this ML - test counter changed only on TPASS or TFAIL. This lead us to really confusing log output :
> ipneigh01 1 TINFO: initialize 'lhost' 'ltp_ns_veth2' interface
> ipneigh01 1 TINFO: set local addr 10.0.0.2/24
> ipneigh01 1 TINFO: set local addr fd00:1:1:1::2/64
> ipneigh01 1 TINFO: initialize 'rhost' 'ltp_ns_veth1' interface
> ipneigh01 1 TINFO: set remote addr 10.0.0.1/24
> ipneigh01 1 TINFO: set remote addr fd00:1:1:1::1/64
> ipneigh01 1 TINFO: Network config (local -- remote):
> ipneigh01 1 TINFO: ltp_ns_veth2 -- ltp_ns_veth1
> ipneigh01 1 TINFO: 10.0.0.2/24 -- 10.0.0.1/24
> ipneigh01 1 TINFO: fd00:1:1:1::2/64 -- fd00:1:1:1::1/64
> ipneigh01 1 TCONF: 'arp cmd doesn't support IPv6, skipping test-case
> ipneigh01 1 TINFO: Stress auto-creation of NDISC cache entry
> ipneigh01 1 TINFO: by pinging 'fd00:1:1:1::1' and deleting entry again
> ipneigh01 1 TINFO: with 'ip neigh del fd00:1:1:1::1 dev ltp_ns_veth2'
> ipneigh01 1 TPASS: verified adding/removing of NDISC cache entry
> Please correct me if I missing something but I read this like "ok we can't run this test case because arp cmd doesn't support IPv6 , ah but wait test case is passed ".
>> After all discussions around that patch and all info which I gain I can agree that changing TCONF with TINFO was bad idea , but if we will just "fix" issue by changing TST_TOTAL value to 1 ( if I correctly understand your suggestion )
> this will not remove this confusion.
I've not suggested to change TST_TOTAL.
> Only valid fix which I see is change behavior of LTP to change test case counter after TCONF which makes total sense for me , but most probably I missing something and you had some reasons to not changing counter on TCONF messages. Can you please elaborate them ?
Yes, this is a flaw in the test.sh library.
And it would be better to convert the test to use tst_test.sh.
More information about the ltp