On (10/03/16 13:15), Lukas Slebodnik wrote:
>On (10/03/16 13:10), Pavel Březina wrote:
>>On 03/10/2016 01:05 PM, Lukas Slebodnik wrote:
>>>On (10/03/16 12:58), Pavel Březina wrote:
>>>>On 03/09/2016 05:39 PM, Lukas Slebodnik wrote:
>>>>>ehlo,
>>>>>
>>>>>I read log files from latest 1.13 today and it was a small challenge
>>>>>due to missing line feed after some ldb messages.
>>>>>
>>>>>LS
>>>>
>>>>Can we check if new line is present instead of always appending it?
>>>We do not always append it. We append it only in libldb.
>>>Because libldb is inconsistent. I sent patch to samba upstream
>>
>>That's why I ask for it.
>>
>>>but we cannot reply on specific version of libldb.
>>>In future, we can conditionally remove this flag if we detect there is
>>>recent enough libldb.
>>>
>>>We could append it but it would be additional slowdown in debug messages
>>>even in other parts of code which already have line feed in message
>>>(sssd debug messages, libhbac ...)
>>
>>It doesn't need to be on sssd debug, but only in ldb_debug_messages. I
don't
>>think that if (fmt[strlen-1] '= '\n') is such a slow down, but I
don'ẗ
>>insist.
>>
>strlen is time consuming.
>
>But if there will be more developers who like idea with strlen.
>Then we might consider to add it.
>
>>Ack to the patches then.
>>
>Thank you for review.
>
>I will wait few days
>for other opinions before pushing.
>
http://sssd-ci.duckdns.org/logs/job/39/88/summary.html
master:
* 7c30eade4ae794ed809845f2ef70dda849b6e7c9
* 558ec7d717735bb16c210c675c2cc5bee1da4576
and the same issue is in 1.13 branch therefore backported patches:
sssd-1-13:
* f07332dca37952b371936be5005c10319b3e32bd
* a7eba44333da1f711d027ef0f034f1ac6edb8403
LS