[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ocfs2-devel
Subject:    Re: [Ocfs2-devel] ocfs2 patches reviewing request
From:       Changwei Ge <chge () linux ! alibaba ! com>
Date:       2019-09-16 3:29:08
Message-ID: 18616125-bbd1-469a-eb9c-4b62a03a911b () linux ! alibaba ! com
[Download RAW message or body]

Hi Joseph,

On 2019/9/15 9:44 上午, Joseph Qi wrote:
> Hi Changwei,
> 
> On 19/9/11 11:03, Changwei Ge wrote:
> > Hi list,
> > 
> > 
> > I have two ocfs2 patches[1] queued in -mm tree for about 5 months.
> > 
> > Does anyone can help review them?
> > 
> > Maybe, @Joseph do you have some time?
> > 
> > Any comments, advise and questions will be welcomed.
> > 
> > Hopefully, we can merge them into mainline during next merge window to close the \
> > two issues. 
> > 
> > Thanks,
> > 
> > Changwei
> > 
> > 
> > [1]:
> > 
> > ocfs2: wait for recovering done after direct unlock request
> > 
> > https://urldefense.proofpoint.com/v2/url?u=http-3A__git.cmpxchg.org_cgit.cgi_linux \
> > -2Dmmotm.git_commit_-3Fh-3Dv5.3-2Drc6-2Dmmotm-2D2019-2D08-2D27-2D20-2D39-26id-3D2a \
> > aebca745e4cc7c49401374ba1d430ec1fbfe29&d=DwIDaQ&c=RoP1YumCXCgaWHvlZYR8PZh8Bv7qIrMU \
> > B65eapI_JnE&r=C7gAd4uDxlAvTdc0vmU6X8CMk6L2iDY8-HD0qT6Fo7Y&m=_t_Of2eXInVn3n8f73nWj6pDN3XgYy0cqf2-rhjXSIo&s=JApCD55Np6-m_ICRyaHBbie44eldDkeLAEBaoAgS0o0&e= \
> >  
> I've gone through the thread, it seems that you agreed on a better
> solution after discussed with Jun?

I tried the way as Jun suggested, but after that, IMO, adding a point of 
calculating dlm lock resource usage in 
`dlm_finish_local_lockres_recovery()` might make related code subtle. 
Although Jun's way can run _dlm unlock_ and _dlm recovery_ 
asynchronously, but the issue I reported was a corner case during 
unmounting ocfs2 from different nodes together. So performance should 
not be our biggest concern. A explicit fix would be my preference. 
Practically, most dlm code's behavior is decided after RECOVERING DONE 
if there is recovering process running.



> 
> 
> > ocfs2: checkpoint appending truncate log transaction before flushing
> > 
> > https://urldefense.proofpoint.com/v2/url?u=http-3A__git.cmpxchg.org_cgit.cgi_linux \
> > -2Dmmotm.git_commit_-3Fh-3Dv5.3-2Drc6-2Dmmotm-2D2019-2D08-2D27-2D20-2D39-26id-3Dab \
> > d1a359d33331ae21cf880d579215dc793d925c&d=DwIDaQ&c=RoP1YumCXCgaWHvlZYR8PZh8Bv7qIrMU \
> > B65eapI_JnE&r=C7gAd4uDxlAvTdc0vmU6X8CMk6L2iDY8-HD0qT6Fo7Y&m=_t_Of2eXInVn3n8f73nWj6pDN3XgYy0cqf2-rhjXSIo&s=UmlDvPwCLbXPnqDcbXZ_C8ihvRTVwRf3_f_KU5tAksA&e= \
> > 
> This looks good to me.
> So should I give my reviewed-by on the original thread, or you'll start a
> new one?

Adding your feedback at my original thread would be fine. :-)

Thanks,
Changwei

> 
> Thanks,
> Joseph
> 

_______________________________________________
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
https://oss.oracle.com/mailman/listinfo/ocfs2-devel


[prev in list] [next in list] [prev in thread] [next in thread]