* Robert Hailey <robert at freenetproject.org> [2008-04-07 11:00:32]:
>
> On Apr 5, 2008, at 8:18 AM, toad at freenetproject.org wrote:
>
>> Author: toad
>> Date: 2008-04-05 13:18:53 +0000 (Sat, 05 Apr 2008)
>> New Revision: 19014
>>
>> Modified:
>> trunk/freenet/src/freenet/node/SSKInsertSender.java
>> Log:
>> Synchronize on access to status.
>> Inspired by Daniel Cheng's patch.
>>
>> Modified: trunk/freenet/src/freenet/node/SSKInsertSender.java
>> ===================================================================
>> --- trunk/freenet/src/freenet/node/SSKInsertSender.java 2008-04-05
>> 13:16:17 UTC (rev 19013)
>> +++ trunk/freenet/src/freenet/node/SSKInsertSender.java 2008-04-05
>> 13:18:53 UTC (rev 19014)
>> @@ -525,7 +525,7 @@
>> // Nothing to wait for, no downstream transfers, just exit.
>> }
>>
>> - public int getStatus() {
>> + public synchronized int getStatus() {
>> return status;
>> }
>>
>> @@ -536,7 +536,7 @@
>> /**
>> * @return The current status as a string
>> */
>> - public String getStatusString() {
>> + public synchronized String getStatusString() {
>> if(status == SUCCESS)
>> return "SUCCESS";
>> if(status == ROUTE_NOT_FOUND)
>>
>
> I have not really examined nextgens recent volatile patches, but my
> understanding is that this is actually one of the limited cases where a
> volatile field is exactly what is desired. Other threads (like fproxy,
> logging) would have to block to read the status, but the synchronization
> does not help since the moment the status is returned any atomicity is
> lost. The return value of getStatus()/getStatusString() is therefore
> inherently 'dirty'.
>
Agreed, synchronization is not necessary here
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL:
<https://emu.freenetproject.org/pipermail/devl/attachments/20080408/19a6db6d/attachment.pgp>
