lkml.org 
[lkml]   [2006]   [Dec]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[PATCH] reorder struct pipe_buf_operations
Fields of struct pipe_buf_operations have not a precise layout (ie not 
optimized to fit cache lines nor reduce cache line ping pongs)

The bufs[] array is *large* and is placed near the beginning of the structure,
so all following fields have a large offset. This is unfortunate because many
archs have smaller instructions when using small offsets relative to a base
register. On x86 for example, 7 bits offsets have smaller instruction lengths.

Moving bufs[] at the end of pipe_buf_operations permits all fields to have
small offsets, and reduce text size, and icache pressure.

# size vmlinux.pre vmlinux
text data bss dec hex filename
3268989 664356 492196 4425541 438745 vmlinux.pre
3268765 664356 492196 4425317 438665 vmlinux

So this patch reduces text size by 224 bytes on my x86_64 machine. Similar
results on ia32.


Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
--- linux-2.6.19/include/linux/pipe_fs_i.h 2006-12-11 23:06:57.000000000 +0100
+++ linux-2.6.19-ed/include/linux/pipe_fs_i.h 2006-12-11 22:58:42.000000000 +0100
@@ -41,7 +41,6 @@ struct pipe_buf_operations {
struct pipe_inode_info {
wait_queue_head_t wait;
unsigned int nrbufs, curbuf;
- struct pipe_buffer bufs[PIPE_BUFFERS];
struct page *tmp_page;
unsigned int readers;
unsigned int writers;
@@ -51,6 +50,7 @@ struct pipe_inode_info {
struct fasync_struct *fasync_readers;
struct fasync_struct *fasync_writers;
struct inode *inode;
+ struct pipe_buffer bufs[PIPE_BUFFERS];
};

/* Differs from PIPE_BUF in that PIPE_SIZE is the length of the actual
\
 
 \ /
  Last update: 2006-12-11 22:49    [W:0.154 / U:0.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site