This patchset uses the infrastructure added in the first set to allow cifs to vastly increase the rsize. It also is more efficient and does less copying as the read of the data is all done directly into the pagecache. Additionally, readpages is now done asynchronously with this set. cifs_readpages just puts the pages in the cache, and fires off the reads and returns. Eventually, when the reads complete, the pages are marked Uptodate, unlocked, etc. With this set in place, I see a dramatic increase in performance for buffered read calls. On a simple dd test of 100M or so: prepatch, with 16k rsize: 104857600 bytes (105 MB) copied, 5.08086 s, 20.6 MB/s postpatch, with 1M rsize: 104857600 bytes (105 MB) copied, 1.2141 s, 86.4 MB/s Other tests show similar results. These results are from my craptacular KVM-based test rig. I expect that the performance boost will be even more dramatic on real hardware or on higher-latency connections. Steve, I'd like to see both of these sets go into 3.2 if at all possible. Both of these sets should be bisectable, so committing them in order is recommended. Jeff Layton (5): cifs: fix protocol definition for READ_RSP cifs: add cifs_async_readv cifs: convert cifs_readpages to use async reads cifs: allow for larger rsize= options and set default to 1M cifs: don't discard readahead pages that are beyond EOF fs/cifs/cifspdu.h | 4 +- fs/cifs/cifsproto.h | 23 ++++ fs/cifs/cifssmb.c | 313 +++++++++++++++++++++++++++++++++++++++++++++++++++ fs/cifs/connect.c | 127 ++++++++++++--------- fs/cifs/file.c | 272 ++++++++++++++++---------------------------- 5 files changed, 509 insertions(+), 230 deletions(-) -- 1.7.6 -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
