diff options
| author | Dominique Martinet <dominique.martinet@cea.fr> | 2018-07-30 08:14:37 +0200 |
|---|---|---|
| committer | Dominique Martinet <dominique.martinet@cea.fr> | 2018-09-07 18:39:47 +0200 |
| commit | 91a76be37ff89795526c452a6799576b03bec501 (patch) | |
| tree | dc45cdba9b9b8e1b5fa3451024968e137fb23d16 /include/net/9p | |
| parent | 9p: embed fcall in req to round down buffer allocs (diff) | |
| download | linux-91a76be37ff89795526c452a6799576b03bec501.tar.xz linux-91a76be37ff89795526c452a6799576b03bec501.zip | |
9p: add a per-client fcall kmem_cache
Having a specific cache for the fcall allocations helps speed up
end-to-end latency.
The caches will automatically be merged if there are multiple caches
of items with the same size so we do not need to try to share a cache
between different clients of the same size.
Since the msize is negotiated with the server, only allocate the cache
after that negotiation has happened - previous allocations or
allocations of different sizes (e.g. zero-copy fcall) are made with
kmalloc directly.
Some figures on two beefy VMs with Connect-IB (sriov) / trans=rdma,
with ior running 32 processes in parallel doing small 32 bytes IOs:
- no alloc (4.18-rc7 request cache): 65.4k req/s
- non-power of two alloc, no patch: 61.6k req/s
- power of two alloc, no patch: 62.2k req/s
- non-power of two alloc, with patch: 64.7k req/s
- power of two alloc, with patch: 65.1k req/s
Link: http://lkml.kernel.org/r/1532943263-24378-2-git-send-email-asmadeus@codewreck.org
Signed-off-by: Dominique Martinet <dominique.martinet@cea.fr>
Acked-by: Jun Piao <piaojun@huawei.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Greg Kurz <groug@kaod.org>
Diffstat (limited to 'include/net/9p')
| -rw-r--r-- | include/net/9p/9p.h | 4 | ||||
| -rw-r--r-- | include/net/9p/client.h | 1 |
2 files changed, 5 insertions, 0 deletions
diff --git a/include/net/9p/9p.h b/include/net/9p/9p.h index e23896116d9a..beede1e1a919 100644 --- a/include/net/9p/9p.h +++ b/include/net/9p/9p.h @@ -336,6 +336,9 @@ enum p9_qid_t { #define P9_NOFID (u32)(~0) #define P9_MAXWELEM 16 +/* Minimal header size: size[4] type[1] tag[2] */ +#define P9_HDRSZ 7 + /* ample room for Twrite/Rread header */ #define P9_IOHDRSZ 24 @@ -558,6 +561,7 @@ struct p9_fcall { size_t offset; size_t capacity; + struct kmem_cache *cache; u8 *sdata; }; diff --git a/include/net/9p/client.h b/include/net/9p/client.h index c2671d40bb6b..735f3979d559 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -123,6 +123,7 @@ struct p9_client { struct p9_trans_module *trans_mod; enum p9_trans_status status; void *trans; + struct kmem_cache *fcall_cache; union { struct { |