> -----Original Message----- > From: John Hubbard <jhubbard@xxxxxxxxxx> > Sent: Tuesday, October 20, 2020 6:42 PM > To: Xiong, Jianxin <jianxin.xiong@xxxxxxxxx>; linux-rdma@xxxxxxxxxxxxxxx; dri-devel@xxxxxxxxxxxxxxxxxxxxx > Cc: Doug Ledford <dledford@xxxxxxxxxx>; Jason Gunthorpe <jgg@xxxxxxxx>; Leon Romanovsky <leon@xxxxxxxxxx>; Sumit Semwal > <sumit.semwal@xxxxxxxxxx>; Christian Koenig <christian.koenig@xxxxxxx>; Vetter, Daniel <daniel.vetter@xxxxxxxxx> > Subject: Re: [PATCH v5 0/5] RDMA: Add dma-buf support > > On 10/15/20 3:02 PM, Jianxin Xiong wrote: > > This is the fifth version of the patch set. Changelog: > > > > Hi, > > A minor point, but if you can tweak your email sending setup, it would be nice. > Specifically, make follow-up patches a reply to the first item. That's a list convention, and git format-patch + git send-email *.patch is > normally sufficient to make that happen, unless you override it by doing something like sending each patch separately...which is my first > suspicion as to how this happened. > > These patches are difficult to link to, because they don't follow the convention of patches 1-5 being in-reply-to patch 0. So if we want to > ask people outside of this list to take a peek (I was about to), we have to go collect 5 or 6 different lore.kernel.org URLs, one for each > patch... > > Take a look on lore and you can see the problem. Here's patch 0, and there is no way from there to find the remaining patches: > > https://lore.kernel.org/dri-devel/1602799340-138152-1-git-send-email-jianxin.xiong@xxxxxxxxx/ > > Hi John, Thanks for pointing this out. I didn't realize sending out patches individually would cause the difference compared to sending with a single command. Only version 4 and 5 have this issue and I will switch back to my old script for the next version. Thanks, Jianxin > thanks, > -- > John Hubbard > NVIDIA >
