Cheatsheet

第一个脚本(1.sh)：
#!/bin/sh
echo test
exit 1

测试命令：./1.sh | grep test && echo 1

csh:
./1.sh | grep test && echo e
test

sh:
./1.sh | grep test && echo e
test
e

注：这篇文章属于作弊条性质，本人不负责其内容所依据理论之正确性。

a. 路由器重置方法：断电，用笔按住背面的RESET按钮，并接上电源，直至路由器提示灯都变成正常的样子为止。
b. 路由器默认IP为192.168.1.1，用户名和口令均为admin。

愈发习惯FreeBSD下的生活之后，用Windows反而时常感觉不爽。例如最近微软推出的Internet Explorer 7的Beta 3，竟然不允许从Beta 2直接覆盖升级，实在让人感觉十分不快。

因为有人问到这个问题，解答一下：

对于：
const char **p
p[i]的值可变，而p[i][j]的值不可变。

对于
char * const *p
p[i]的值不可变，而p[i][j]的值可变。

From des@:

groff -Tlatin1 -mdoc strtonum.3 2>&1 | less

A even better version from ru@:

cd /usr/src/lib/libc
make manlint MAN=strtonum.3 # to check stderr
make all-man MAN=strtonum.3 -DMANBUILDCAT # to format

I used to think that ldap_unbind() is the opposite of ldap_bind(), however I was wrong. Pierangelo Masarati (ando at openldap) pointed out that it was actually opposite of ldap_init(). So, do NOT expect init() -> bind() -> search() -> unbind() -> bind() would success, it’s simply wrong thing. You do not need to unbind() before you do another bind().

Just make a note.

pure-ftpd是一款非常好的FTP服务软件。FreeBSD本身提供了两个很好的FTP服务器，一个是传统的BSD ftpd(8)，另一个是来自NetBSD的lukemftpd(8)，尽管性能和安全性都不错(做过全面代码审计)，但配置虚拟用户会比较麻烦。

You can obtain latest list from sys/sys/proc.h:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

 * Below is a key of locks used to protect each member of struct proc.  The
 * lock is indicated by a reference to a specific character in parens in the
 * associated comment.
 *      * - not yet protected
 *      a - only touched by curproc or parent during fork/wait
 *      b - created at fork, never changes
 *              (exception aiods switch vmspaces, but they are also
 *              marked 'P_SYSTEM' so hopefully it will be left alone)
 *      c - locked by proc mtx
 *      d - locked by allproc_lock lock
 *      e - locked by proctree_lock lock
 *      f - session mtx
 *      g - process group mtx       
 *      h - callout_lock mtx
 *      i - by curproc or the master session mtx
 *      j - locked by sched_lock mtx
 *      k - only accessed by curthread
 *      l - the attaching proc or attaching proc parent
 *      m - Giant
 *      n - not locked, lazy
 *      o - ktrace lock
 *      p - select lock (sellock)
 *      q - td_contested lock
 *      r - p_peers lock
 *      x - created at fork, only changes during single threading in exec
 *      z - zombie threads/ksegroup lock

Today Giorgos Keramidas (keramida@) has noted that the following skill that can minimize lock period:

1. save the commit message to a file (say ~/commitlog)
2. cvs ci -F ~/commitlog file[…]

This article has suggested a way of hiding version.bind string. From a security perceive, this is an overkill and can usually be harmful:

• While it’s true that you can hide version.bind, the fact that version.bind is can be queried reveals that you are running BIND. It makes little sense to fake a version.
• The need of hiding information, which is unnecessarily hidden like this, means that the system administrator is neglecting security.
• Therefore, FOREACH(version.bind is queriable and is hidden) HACKEM :-)

Security can NOT be built on what others are not aware of.