lighttpd 1.4.52 发布,带来性能改进

栏目: 软件资讯 · 发布时间: 5年前

内容简介:lighttpd 1.4.52 发布了,该版本带来了性能改进和 Bug 修复。 server.http-parseopts 的推荐设置如下   server.http-parseopts = (         "header-strict"            => "enable",...

lighttpd 1.4.52 发布了,该版本带来了性能改进和 Bug 修复。

server.http-parseopts 的推荐设置如下

  server.http-parseopts = (
        "header-strict"            => "enable",
        "host-strict"              => "enable",
        "host-normalize"           => "enable",
        "url-normalize"            => "enable",
        "url-normalize-unreserved" => "enable",
        "url-normalize-required"   => "enable",
        "url-ctrls-reject"         => "enable",
        "url-path-2f-decode"       => "enable",
        "url-path-dotseg-remove"   => "enable",
        "url-query-20-plus"        => "enable"
      )

自 1.4.51 以来的更改

  • [mysql] MySQL 8 deprecates my_bool

  • [core] typo in trace

  • [build] Fix unportable test(1) operator

  • [core] perf: call connection_reset() fewer times

  • [core] perf: array_reset_data_strings()

  • [core] perf: buffer_free_ptr() attribute cold

  • [core] perf: one-element cache for host normalize

  • [core] perf: buffer_copy_string_len()

  • [core] perf: skip redundant prepare copy calls

  • [core] perf: buffer_align_size() identity if align

  • [core] perf: size write buffers for reuse

  • [core] perf: prepend headers directly into write q

  • [core] perf: copy small strings; better buf reuse

  • [core] perf: copy small strings; extend last chunk

  • [core] perf: specialized func for array sorting

  • [core] perf: append response directly into write q

  • [core] perf: better buf reuse reading from backend

  • [core] chunk.c code reuse

  • [multiple] perf: write headers to backend write cq

  • [multiple] perf: power-2 alloc large headers

  • [multiple] perf: use larger initial backend buffer

  • [core] permit env vars to be set with blank value

  • [mod_fastcgi] perf: reduce data copies

  • [mod_fastcgi] perf: reduce data copies

  • [core] perf: chunk.c chunk pool

  • [multiple] perf: reuse large buffers w/ backend

  • [multiple] better packing of struct chunk

  • [core] perf: inline buffer_append_string_buffer()

  • [core] slightly simpler flag append to string

  • [mod_cgi] perf: reuse buffers for creating CGI env

  • [mod_fastcgi,mod_scgi] perf: env accumulation

  • [core] Don’t call RAND_cleanup with OpenSSL 1.1.x

  • [mod_openssl] move SSL_shutdown() to separate func

  • [mod_openssl] SSL_read before second SSL_shutdown

  • [mod_cgi] perf: use stat_cache for cgi handler

  • [mod_openssl] prefer using TLS_server_method()

  • [mod_webdav] return 403 if file should exist

  • [core] perf: chunkqueue buffers already sized up

  • [core] perf: simpler buffer_string_space()

  • [multiple] dynamic handlers hint backend header sz

  • [core] use chunk_buf_sz instead of hard-coded num

  • [multiple] perf: simplify chunkqueue_get_memory()

  • [mod_wstunnel] perf: reuse large buffers

  • [mod_cgi] perf: cache getenv() results at start up

  • [core] fix 301 → 302 overwrite with Location (fixes #2918)

  • [core] fix setting of headers previously reset (fixes #2919)

  • [mod_webdav] quiet coverity false positive

  • [core] server.compat-module-load = “disable”

  • [core] server.chunkqueue-chunk-sz = 4096

  • [core] perf: simpler buffer_string_space() (fixed)

  • [core] perf: faster HTTP pipelined requests

  • [core] perf: simpler buffer_string_space() (tests)

  • [mod_cgi] reset reused buffer on internal redir

  • [core] clear chunk buffer upon release

  • [mod_fastcgi] minor: copy packet without padding

  • [mod_redirect,mod_rewrite] use server_name

  • [mod_fastcgi] transfer chunks minus packet padding

  • [core] separate func to reset FILE_CHUNK

  • [core] perf: simple, quick buffer_clear()

  • [core] perf: small improvement to encoding CGI var

  • [core] perf: small improvement buffer_string_space

  • [core] simpler physical path concatenation

  • [mod_webdav] fix LOCK on incorrect URI path

  • [mod_webdav] one fewer buffer copy for COPY,MOVE

  • [core] perf: simplify buffer_move()

  • [mod_cml] parse query string without modifying it

  • [core] perf: buffer optimizations

  • [mod_wstunnel] use buffer_string_length()

  • [core] perf: inline buffer_copy_buffer()

  • [core] cygwin helper func for getcwd

  • [core] cygwin sample to run lighttpd under NSSM

  • [core] limit con&rarr;uri.authority < 1024 octets

  • [mod_webdav] separate func for each request method

  • [core] reject decoded url-path without leading &lsquo;/&rsquo;

  • [multiple] validate UTF-8 in url-decoded paths

  • [mod_proxy] silence coverity false positive

  • [core] fix typo

  • [core] buffer_append_path_len()

  • [core] quiet indexfile warning if mod not loaded

  • </ul


【声明】文章转载自:开源中国社区 [http://www.oschina.net]


以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持 码农网

查看所有标签

猜你喜欢:

本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们

R语言编程艺术

R语言编程艺术

(美)Norman Matloff / 陈堰平、邱怡轩、潘岚锋 等 / 机械工业出版社 / 2013-5 / 69.00

【编辑推荐】 这本书涵盖了R语言编程的诸多方面,尤其在面向对象编程、程序调试、提升程序运行速度以及并行计算等方面,填补了同类图书的空白。关于程序调试的章节更是作者多年经验的总结。不管是初学者还是有一定编程经验的读者,阅读这本书都会有所收获。 ——统计之都 【内容简介】 R语言是世界上最流行的用于数据处理和统计分析的脚本语言。考古学家用它来跟踪古代文明的传播,医药公司用它来探......一起来看看 《R语言编程艺术》 这本书的介绍吧!

MD5 加密
MD5 加密

MD5 加密工具

RGB CMYK 转换工具
RGB CMYK 转换工具

RGB CMYK 互转工具

HEX HSV 转换工具
HEX HSV 转换工具

HEX HSV 互换工具