Hi Martin, Thank you for your patch! Is that possible to give a test case to show the problem clearly?
Thanks! 2011/2/27 Martin Storsjö (JIRA) <j...@apache.org> > > [ > https://issues.apache.org/jira/browse/HARMONY-6682?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel] > > Martin Storsjö updated HARMONY-6682: > ------------------------------------ > > Attachment: outputstream-shortwrite.patch > > Patch fixing the issue > > > SocketOutputStream doesn't handle short writes properly > > ------------------------------------------------------- > > > > Key: HARMONY-6682 > > URL: https://issues.apache.org/jira/browse/HARMONY-6682 > > Project: Harmony > > Issue Type: Bug > > Components: Classlib > > Environment: Tested/noticed on Android, same issue identified in > the Harmony code base, too > > Reporter: Martin Storsjö > > Attachments: outputstream-shortwrite.patch > > > > > > Even blocking writes can return before the full amount is written, e.g. > if a signal is received, or for any other arbitrary reason. > > The java.io.OutputStream.write() interface that SocketOutputStream > implements returns void, and cannot report short writes to the caller, and > therefore is required to loop to retry writing until everything is written, > or until writing fails. > > The same issue is reported to Android at > https://review.source.android.com/21459. > > -- > This message is automatically generated by JIRA. > - > For more information on JIRA, see: http://www.atlassian.com/software/jira > > > -- Regards, Ray Chen
