net: sock: Add option for memory optimized hints.
Subsystems (4)
PATCH 1/1 net: sock: Add option for memory optimized hints.
From: Peter Enderborg <[email protected]>
When sending data the socket allocates memory for payload on a cache or a page alloc. The page alloc then might trigger compation that takes long time. This can be avoided with smaller chunks. But userspace can not know what is the right size for the smaller sends. For this we add a SIZEHINT getsocketopt where the userspace can get the size for send that will fit into one page (order 0) or the max for a slab cache allocation.
--- include/uapi/asm-generic/socket.h | 2 ++ include/uapi/linux/socket.h | 9 +++++++++ net/core/sock.c | 17 +++++++++++++++++ 3 files changed, 28 insertions(+)
include/uapi/asm-generic/socket.h +2 -0
@@ -92,4 +92,6 @@#define SO_CNX_ADVICE 53 +#define SO_SIZEHINT 54 +#endif /* __ASM_GENERIC_SOCKET_H */
include/uapi/linux/socket.h +9 -0
@@ -1,6 +1,8 @@#ifndef _UAPI_LINUX_SOCKET_H#define _UAPI_LINUX_SOCKET_H +#include <linux/types.h> +/* * Desired design of maximum size and alignment (see RFC2553) */
@@ -18,4 +20,11 @@ struct __kernel_sockaddr_storage {/* _SS_MAXSIZE value minus size of ss_family */ } __attribute__ ((aligned(_K_SS_ALIGNSIZE))); /* force desired alignment */ +struct sock_sizehint { +__u32 order_zero_size; +/* max payload size that can fit into one page in kernel */ +__u32 cache_size; +/* max payload size that can fit in socket slab cache */ +}; +#endif /* _UAPI_LINUX_SOCKET_H */
net/core/sock.c +17 -0
@@ -1254,6 +1254,23 @@ int sock_getsockopt(struct socket *sock, int level, int optname, v.val = sk->sk_incoming_cpu;break ; +case SO_SIZEHINT : + { +struct sock_sizehint hint; + +if (len >sizeof (hint)) + len =sizeof (hint); + + hint.order_zero_size =PAGE_SIZE - +SKB_DATA_ALIGN (sizeof (struct skb_shared_info )); + hint.cache_size =KMALLOC_MAX_CACHE_SIZE - +SKB_DATA_ALIGN (sizeof (struct skb_shared_info )); + +if (copy_to_user (optval, &hint, len)) +return -EFAULT ; + goto lenout; + } +default :/* We implement the SO_SNDLOWAT etc to not be settable * (1003.1g 7).
Signature · 1 line
2.4.2
David Miller
maintainerIt is not appropriate to submit new features at this time as we are in the merge window and we are only bug fixes.
Please resubmit this when the net-next tree opens back up.
Quoted 11 lines · peter enderborg
From: Peter Enderborg <[email protected]>
When sending data the socket allocates memory for payload on a cache or a page alloc. The page alloc then might trigger compation that takes long time. This can be avoided with smaller chunks. But userspace can not know what is the right size for the smaller sends. For this we add a SIZEHINT getsocketopt where the userspace can get the size for send that will fit into one page (order 0) or the max for a slab cache allocation.
Very obscure thing if you ask me. Exposing kernel internals like that seems to work around a more serious issue.
What kind of sockets would ever use KMALLOC_MAX_CACHE_SIZE allocations exactly ?
af_unix and tcp are definitely clean.
28d6427109d13b0f447cba5761f88d3548e83605 net: attempt high order allocations in sock_alloc_send_pskb() e370a7236321773245c5522d8bb299380830d3b2 af_unix: improve STREAM behavior with fragmented memory eb6a24816b247c0be6b2e97e68933072874bbe54 af_unix: reduce high order page allocations