Re: [Qemu-devel] [RFC v3 00/27] QMP: out-of-band (OOB) execution support

[Fam Zheng <famz@redhat.com>]

On Thu, 11/16 13:32, Peter Xu wrote:
> On Wed, Nov 15, 2017 at 09:42:46AM +0000, Stefan Hajnoczi wrote:
> > On Mon, Nov 06, 2017 at 09:08:00PM +0800, Peter Xu wrote:
> > > On Mon, Nov 06, 2017 at 02:12:17AM -0800, no-reply@patchew.org wrote:
> > > > Hi,
> > > > 
> > > > This series seems to have some coding style problems. See output below for
> > > > more information:
> > > > 
> > > > Subject: [Qemu-devel] [RFC v3 00/27] QMP: out-of-band (OOB) execution support
> > > > Type: series
> > > > Message-id: 20171106094643.14881-1-peterx@redhat.com
> > > > 
> > > > === TEST SCRIPT BEGIN ===
> > > > #!/bin/bash
> > > > 
> > > > BASE=base
> > > > n=1
> > > > total=$(git log --oneline $BASE.. | wc -l)
> > > > failed=0
> > > > 
> > > > git config --local diff.renamelimit 0
> > > > git config --local diff.renames True
> > > > 
> > > > commits="$(git log --format=%H --reverse $BASE..)"
> > > > for c in $commits; do
> > > >     echo "Checking PATCH $n/$total: $(git log -n 1 --format=%s $c)..."
> > > >     if ! git show $c --format=email | ./scripts/checkpatch.pl --mailback -; then
> > > >         failed=1
> > > >         echo
> > > >     fi
> > > >     n=$((n+1))
> > > > done
> > > > 
> > > > exit $failed
> > > > === TEST SCRIPT END ===
> > > > 
> > > > Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
> > > > From https://github.com/patchew-project/qemu
> > > >  * [new tag]               patchew/20171106094643.14881-1-peterx@redhat.com -> patchew/20171106094643.14881-1-peterx@redhat.com
> > > > Switched to a new branch 'test'
> > > > 5525c4e791 tests: qmp-test: add oob test
> > > > ccc9e4c399 tests: qmp-test: verify command batching
> > > > 7f45b4c6c0 docs: update QMP documents for OOB commands
> > > > 58cfe877d2 monitor: enable IO thread for (qmp & !mux) typed
> > > > 5e1d56ce74 qmp: isolate responses into io thread
> > > > aef4275536 qmp: let migrate-incoming allow out-of-band
> > > > 5e68beacf3 qmp: support out-of-band (oob) execution
> > > > 43c7215a30 qapi: introduce new cmd option "allow-oob"
> > > > e11127ba4b monitor: send event when request queue full
> > > > 45cef4f7e4 qmp: add new event "request-dropped"
> > > > 485da28be1 monitor: separate QMP parser and dispatcher
> > > > 4892fe9ca2 monitor: let monitor_{suspend|resume} thread safe
> > > > 1b86166d9c monitor: introduce monitor_qmp_respond()
> > > > 0f48093add qmp: introduce some capability helpers
> > > > 8d3f33043d qmp: negociate QMP capabilities
> > > > 023b386d0e qmp: introduce QMPCapability
> > > > 2bde5ca8ce monitor: allow to use IO thread for parsing
> > > > f4cc112f80 monitor: create monitor dedicate iothread
> > > > 3590fdc1d4 monitor: let mon_list be tail queue
> > > > 11c818a9ac monitor: unify global init
> > > > 36d3efb87d monitor: move the cur_mon hack deeper for QMP
> > > > bf3e493a86 qjson: add "opaque" field to JSONMessageParser
> > > > 17367fe7a1 monitor: move skip_flush into monitor_data_init
> > > > 0c98d4baa4 qobject: let object_property_get_str() use new API
> > > > aa4b973dd5 qobject: introduce qobject_get_try_str()
> > > > 981ccebc1e qobject: introduce qstring_get_try_str()
> > > > d40ba38085 char-io: fix possible race on IOWatchPoll
> > > > 
> > > > === OUTPUT BEGIN ===
> > > > Checking PATCH 1/27: char-io: fix possible race on IOWatchPoll...
> > > > Checking PATCH 2/27: qobject: introduce qstring_get_try_str()...
> > > > Checking PATCH 3/27: qobject: introduce qobject_get_try_str()...
> > > > Checking PATCH 4/27: qobject: let object_property_get_str() use new API...
> > > > Checking PATCH 5/27: monitor: move skip_flush into monitor_data_init...
> > > > Checking PATCH 6/27: qjson: add "opaque" field to JSONMessageParser...
> > > > Checking PATCH 7/27: monitor: move the cur_mon hack deeper for QMP...
> > > > Checking PATCH 8/27: monitor: unify global init...
> > > > Checking PATCH 9/27: monitor: let mon_list be tail queue...
> > > > Checking PATCH 10/27: monitor: create monitor dedicate iothread...
> > > > Checking PATCH 11/27: monitor: allow to use IO thread for parsing...
> > > > Checking PATCH 12/27: qmp: introduce QMPCapability...
> > > > Checking PATCH 13/27: qmp: negociate QMP capabilities...
> > > > Checking PATCH 14/27: qmp: introduce some capability helpers...
> > > > Checking PATCH 15/27: monitor: introduce monitor_qmp_respond()...
> > > > Checking PATCH 16/27: monitor: let monitor_{suspend|resume} thread safe...
> > > > ERROR: braces {} are necessary for all arms of this statement
> > > > #28: FILE: monitor.c:4014:
> > > > +    if (atomic_dec_fetch(&mon->suspend_cnt) == 0)
> > > > [...]
> > > > 
> > > > ERROR: Missing Signed-off-by: line(s)
> > > 
> > > Will add it in next post.  I still haven't found a good way to let
> > > magit add this line for me every time automatically.
> > 
> > If you don't want to type "git commit -s" all the time you could use the
> > "format.signOff = on" git-config(1) variable.
> 
> I tried it, but it seems to be for when formatting patches rather than
> putting that s-o-b line into commit messages.  I think what I wanted
> was something like commit.signOff but after some searches I found that
> idea was proposed but rejected somehow:
> 
> http://git.661346.n2.nabble.com/PATCH-Add-a-commit-signoff-configuration-variable-to-always-use-signoff-td1886619.html#a2093996
> 
> So I think I'll just live with my current approach.  Anway, thanks for
> the hint!

I think what you could better use here is a checkpatch.pl hook, either before
committing or before posting. I personally prefer the checkpatch.pl check in the
pre-publish-tag hook of git-publish so I can freely do dirty commits that
doesn't need s-o-b lines when I'm working on the branch, but still remember to
add it before posting.

Fam