Re: 【Ask for upstream branch】

[Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>]

On Thu, 2017-08-24 at 16:34 +0800, oulijun wrote:
> 在 2017/8/24 15:35, Wei Hu (Xavier) 写道:
> > 
> > 
> > On 2017/8/21 20:57, Doug Ledford wrote:
> > > On 8/21/2017 2:16 AM, oulijun wrote:
> > > > Hi,  Doug Ledford
> > > > 
> > > >     We are preparing the hip08 RoCE driver for upstreaming.  I
> > > > am found that
> > > > 
> > > > you have create a new branch named net-next-base and this
> > > > branch has
> > > > included
> > > > 
> > > > hns3 patchset.  If we generate patchset of hip08 RoCE driver,
> > > > can
> > > > directly based
> > > > 
> > > > on this branch?
> > > 
> > > Yes.
> > > 
> > 
> > Hi, Doug
> > 
> >    In the repo https://git.kernel.org/pub/scm/linux/kernel/git/dled
> > ford/rdma.git,
> >    we found for-next-merged branch looks newer than k.o/net-next-
> > base, right ?
> >    please reconfirm which branch we should generate patcheset based
> > on.
> >    Thank you very much!
> > 
> > 
> > 1. k.o/for-next-merged branch
> > 
> > git log --oneline drivers/infiniband/hw/hns/
> > 
> > b0e32e2 Merge branch 'k.o/for-4.13-rc' into k.o/for-next
> > 5b59a39 IB/hns: fix memory leak on ah on error return path
> > 3e5f088 IB/hns: Avoid compile test under non 64bit environments
> > 8ccf098 Revert "RDMA/hns: fix build regression"
> > 5db465f IB/hns: checking for IS_ERR() instead of NULL
> > 967de35 IB/hns: include linux/interrupt.h
> > ecd840f RDMA/hns: fix build regression
> > d1d7149 IB/hns: fix semicolon.cocci warnings
> > 87809f8 IB/hns: fix returnvar.cocci warnings
> > 3756c7f IB/hns: fix boolreturn.cocci warnings
> > e89bf46 IB/hns: Support compile test for hns RoCE driver
> > 5f110ac4 IB/hns: Fix for checkpatch.pl comment style warnings
> > d322f00 IB/hns: Fix the bug with modifying the MAC address without
> > removing the driver
> > 9de61d3 IB/hns: Fix the bug with rdma operation
> > 58c4f0d IB/hns: Fix the bug with wild pointer when destroy rc qp
> > 
> > 2. k.o/net-next-base
> > 
> > git log --oneline drivers/infiniband/hw/hns/
> > 
> > 5db465f IB/hns: checking for IS_ERR() instead of NULL
> > 5f110ac4 IB/hns: Fix for checkpatch.pl comment style warnings
> > d322f00 IB/hns: Fix the bug with modifying the MAC address without
> > removing the driver
> > 9de61d3 IB/hns: Fix the bug with rdma operation
> > 58c4f0d IB/hns: Fix the bug with wild pointer when destroy rc qp
> > 
> > 
> >    Regards
> > Wei Hu
> > 
> 
> Please use net-next-base. Doug have comfirm. Please refer to his
> reply for this email.

The net-next-based branch is not intended to go to the next kernel
merge window, it's destined for the merge window after that.  If you
don't need anything in net-next-based, then use for-next instead. 
That's what will actually be going to Linus.

As for what's newer, the for-next-merged branch is everything merged
together (for-4.13-rc, for-next, and net-next-based), and I really only
have it there for one reason: so I can put the for-next tag on it and
Stephen Rothwell can then pull it into his for-next kernel testing.  I
put stuff in there (the net-next-based branch) that isn't going into
the next merge window so that it gets early for-next testing, but that
doesn't mean it is actually going into for-next.

So, like I said, use the for-next branch unless you absolutely depend
on code in the net-next-based branch, and if you do, it means the
submission will flow through into the second merge window instead of
the next merge window.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
    GPG KeyID: B826A3330E572FDD
    Key fingerprint = AE6B 1BDA 122B 23B4 265B  1274 B826 A333 0E57 2FDD

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html