On 2/21/17 12:30 PM, GerritHub wrote: > william.allen.simp...@gmail.com uploaded this change for *review*. > > View Change <https://review.gerrithub.io/349831> > > pull up ntirpc #41 > > Remove unused SVC_VC_CREATE_BOTHWAYS > This is now: pull up ntirpc #41 & #42
Many thanks to Daniel G for his review! No idea why CEA failed for #41, but not for #41 + #42. Didn't touch anything related in either case. On the Ganesha side, still only removes one unused symbol (never used internally to ntirpc). This patch series removes a lot of alloc/free activity for every new transport connection, and makes a fair number of changes to headers externally visible by Ganesha (and much more internally). Most of the client and server create calls are now inlines with const parameters, perhaps generating smarter code. "Napalm svc_ioq output threads" patch some time ago sped up the TCP server output -- but had a side effect on the client call side of handling the xids out of order. (Only noticed affecting the NLM_async code, but could have affected v4.x back-channels too.) That's because the first call ran the output queue for other senders, then handled its own result. This now uses a worker task queue instead, restoring the xid ordering. There's not enough testing of UDP and version 3. Basically, just mount, umount, cp in new file, then rm the new file. Surely somebody who uses v3 regularly has a better battery of tests. ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot _______________________________________________ Nfs-ganesha-devel mailing list Nfs-ganesha-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel
