postfix experimental snapshot supports DSN!

Yes, finally postfix is supporting DSN. To quote Witese, this is the first postfix snapshot which provides “full support for RFC 3461..3464 Delivery Status Notifications”!

Still ULE+PREEMPTION

I have done several experiments/asserts to ensure that the panic is related to ULE+PREEMPTION case only.

ULE: no panic
4BSD: no panic
4BSD+PREEMPTION: no panic

My latest progress is that I found that sched_switch in ULE for the threaded case, can decide whether subsequent call would panic. Investigating.

Working on ULE

My company has offered me some test equipment (namely, a HP 360 with ILO, which is *great* for debugging at home!), and I has started to trip over the ULE code and find for bugs.

keys of (j),(z), etc. in FreeBSD scheduler related stuff's comments

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

cvs skill to minimize lock period

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[…]

比碟仙更可怕的拍鬼游戏zz

From: http://y.sina.com.cn 2005年06月17日 11:24 榕树下 文/流浪猫

这是我的一个同事给我讲的他上高中时一个真实的故事。

那一年他上高二，因为他家住在外地，所以他从上高中就在学校住校。他们高中大部分都是长春本市的学生，住校的同学相对较少，他们寝室也就只住了四个人。从生日排行，他排老四。他们寝的老大跟他是同乡，他们俩的关系比较好。

那是快要放暑假的时候，他们考完试，等着发表成绩。这时候已经没有什么课了，学习也比较轻松。那天是星期五，晚上吃完饭没有什么事，几个人就在寝室里侃大山。说着说着，就说到一些奇异的事情上面来了。

老三就说：「哎，你们听没听过一个游戏，说是比‘碟仙’更吓人。」

老二问：「啥游戏啊？」

「拍鬼。你们听过吗？」

我们同事就说：「没听过。咋玩呀？」

我对为什么不能workaround唬弄事的解释

有一天，我们发现了一个bug。

然是有人决定把它workaround掉，大家都很开心。

突然有一天，那个bug被修了。

然后workaround变成了bug……所以说坚决抵制workaround。。。

ULE: kseq_nice_rem is called too much!

It seems that I have hit a bug, where kseq_nice_rem is being called too much. Have read the code and added some KASSERT/MPASS’es.

PR sexiness factors

pav@ has put together this which looks quite interesting. Have a read of it :-)

(Got that from arved@’s blog)

其实老康说的都蛮有道理

细想了想，我有印象的他说的几句话都非常的实在和有道理。

我相信时间能够证明这一切。

• ← 上一页
• 下一页 →