New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

global: start removing g_ceph_context #12149

Merged
merged 11 commits into from Dec 23, 2016

Conversation

Projects
None yet
2 participants
@adamemerson
Contributor

adamemerson commented Nov 23, 2016

Unskew the clock,
Add a dout_context macro to simplify handling dout without rewriting too much code
Remove uses of g_ceph_context from the OSD, ObjectStore, and object class code.

More to come!

Stay tuned kids for our next exciting commit.
Same squid time
Same squid repo!

馃

@ghost ghost added core feature labels Nov 23, 2016

@liewegas

This comment has been minimized.

Member

liewegas commented Nov 23, 2016

馃憤 Let's rebase this on master and merge it quick before it grows more conflicts

@liewegas liewegas changed the title from DNM: Exterminate g_ceph_context forever! to global: explicitly define dout_context; unskew clock Nov 23, 2016

@liewegas liewegas added this to the kraken milestone Nov 23, 2016

@ghost

There are a number of whitespace changes which may cause needless backport issues. Would you mind loosing them ?

@liewegas

This comment has been minimized.

Member

liewegas commented Nov 29, 2016

Passed rados test, except for the clock skew test (which I didn't remove yet). @adamemerson any update on the replacement test?

Also there are a couple trivial conflicts that a rebase would resolve

@liewegas liewegas removed this from the kraken milestone Nov 30, 2016

@adamemerson

This comment has been minimized.

Contributor

adamemerson commented Nov 30, 2016

Just got back from Oregon so reintegrating myself into everything. I'll rebase this and deal with Loic's whitespace concern and get an updated test in shortly.

@adamemerson adamemerson changed the title from global: explicitly define dout_context; unskew clock to Start removing g_ceph_context Dec 13, 2016

@adamemerson

This comment has been minimized.

Contributor

adamemerson commented Dec 13, 2016

@dachary I got rid of the relevant white space stuff (though I think we should clean some of it up and some point,(

@liewegas liewegas changed the title from Start removing g_ceph_context to global: unskew clock Dec 14, 2016

@liewegas liewegas changed the title from global: unskew clock to Start removing g_ceph_context Dec 14, 2016

@liewegas liewegas changed the title from Start removing g_ceph_context to global: start removing g_ceph_context Dec 14, 2016

@@ -1,4 +1,3 @@
// -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-

This comment has been minimized.

@liewegas

liewegas Dec 14, 2016

Member

keep this!

This comment has been minimized.

@adamemerson

adamemerson Dec 14, 2016

Contributor

Oops, sorry, that was unintentional. I'll fix that.

// decoded. This object can be default constructed in every field
// /except/ the CephContext*. To avoid contagion (that is, to avoid
// having to shove a CephContext* into *every* object to preserve a
// uniform interface, make a specialized template function to help.

This comment has been minimized.

@liewegas

liewegas Dec 14, 2016

Member

I don't see any users for this yet... ?

This comment has been minimized.

@liewegas

liewegas Dec 14, 2016

Member

Should we leave it out until there are users?

This comment has been minimized.

@adamemerson

adamemerson Dec 14, 2016

Contributor

We can leave it out. I just threw it in here to give people a chance to look at it and see if it was wrong-headed. The first user is PGMap inside the monitor. I'll hold it until I get that done. I just wanted to get what I had working up before master skittered away from me.

@liewegas

This comment has been minimized.

Member

liewegas commented Dec 14, 2016

Again, let's minimize scope so we can test and merge before we get new conflicts

@adamemerson

This comment has been minimized.

Contributor

adamemerson commented Dec 14, 2016

Sure, I can pull out the encode change. I want to get the rest in, though, since it keeps moving away from me and then I have to rebase and retest.

@liewegas

This comment has been minimized.

Member

liewegas commented Dec 14, 2016

@liewegas liewegas requested a review from athanatos Dec 14, 2016

@liewegas

This comment has been minimized.

Member

liewegas commented Dec 14, 2016

@athanatos does this look ok to you?

adamemerson added some commits Nov 15, 2016

common: Unskew clock
In preparation to deglobalizing CephContext, remove the CephContext*
parameter to ceph_clock_now() and ceph::real_clock::now() that carries
a configurable offset.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
common: Remove now unused clock_offset option
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
dout: Use dout_context
As a transition for g_ceph_context removal, make all dout macros depends
on a local macro.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
osd: Remove g_conf and g_ceph_context
So begins the ruthless annihilation of g_ceph_context.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

adamemerson added some commits Dec 8, 2016

os: Add CephContext to ObjectStore and ObjectMap bases
Preparatory to removing g_ceph_context and g_conf uses

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
os/bluestore: Remove g_ceph_context and g_conf
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
os: Continuing the removal of g_ceph_context and g_conf
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
os: Complete removal of g_ceph_context and g_conf
And so it continues.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
objclass: Utterly destroy g_ceph_context
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
msg: Remove mention of g_ceph_context
Since it won't exist any more there's no point in saying it can't be
relied on, now is there?

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
os/filestore: Destroy g_ceph_context and g_conf
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

Most (all?) whitepsace shoudl be fixed. Need to merge this quick before it grows conflicts again.

@liewegas liewegas merged commit a03c5be into ceph:master Dec 23, 2016

3 checks passed

Signed-off-by all commits in this PR are signed
Details
Unmodifed Submodules submodules for project are unmodified
Details
default Build finished.
Details

@adamemerson adamemerson deleted the adamemerson:wip-g_ceph_context-exterminate branch Jan 13, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment