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
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
|
Fri Jul 09 13:58:42 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/Acceptor.java, EventHandler, TimeValue.java
Solved a few javadoc warnings.
Fri Jul 09 11:08:32 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimedWait.java, Stream*.java, Token.java
Fixed a bug in TimedWait from the recent changes. Made
Streams (and Modules) use absolute times for timeouts.
Changed Tokens slightly so that it ignores interrupts --
will probably change all synchronization mechanisms to
do the same, as well as add a common base class.
Thu Jul 08 14:54:36 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/Condition.java, TimedWait.java, EventHandler.java,
MessageQueue.java, ServiceObject.java, SvcHandler.java,
Task.java, Mutex.java, Semaphore.java, Token.java,
TimerQueue.java
Changed the semantics of Condition, TimedWait, MessageQueue,
Task, Mutex, Semaphore, and Token to use absolute times for
their timeouts.
Changed the semantics of EventHandler, ServiceObject, and
SvcHandler such that handleTimeout receives a TimeValue
representing when the event occured.
Changed TimerQueue internally to reflect the above changes.
Also made better checks in Token and Mutex so that non-owners
can call release without adverse effects. I plan to try to
do this with Semaphore and RWMutex as well.
Thu Jul 08 11:20:17 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimeValue.java: Added relativeTimeOfDay(offsets)
methods which return TimeValues. These should help people
since many classes will be changed to use
absolute TimeValues shortly to correspond to ACE's versions.
Wed Jul 07 22:42:27 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/Token.java: Fixed several bugs in tryacquire and renew.
Specifying a null TimeValue in acquire or renew now means to wait
until the lock is obtained. Specifying TimeValue.zero in acquire
means just to poll and don't call sleepHook(). In renew, it is
the same as specifying a requePosition of 0 -- doesn't give up
the token.
Wed Jul 07 21:54:35 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimedWait.java: Fixed the semantics of
timedWait(TimeValue tv). Now when tv represents zero, it has
polling semantics, checking the condition once and returning
on success or throwing a TimeoutException on failure. Before,
it blocked until notified. (Also affects synch mechanisms)
Wed May 20 14:37:48 CDT 1998 Everett Anderson <eea1@cs.wustl.edu>
Reset the ChangeLog file for future users. If anyone wants more complete
file by file changes, they can be provided with the CVS log command.
Updated the JACE page at http://www.cs.wustl.edu/~eea1/JACE.html to reflect
the Version 1.4.4 changes, which included:
* Removed CUP and JLex from JACE. The java_cup directory is still in CVS,
but won't be included in future versions. This means that the option to
load C++ ACE ServiceConfigurator files is gone.
* Changed the SOCKStream system so it uses DataInputStream and DataOutputStream
to maintain compatibility with other languages and remove the deprecation
warnings from using PrintStream. (Can't use Reader/Writers since they
write 16-bit chars not 8-bit bytes)
* Made lots of modifications to comments and code to remove javadoc warnings
and deprecation warnings. They were all straight-forward changes such as
making sure a @throws line was in the comments for all exceptions a method
had, etc.
|