https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 https://jira.whamcloud.com/browse/LU-353 Lustre <p>When I was investigating on <a href="https://jira.whamcloud.com/browse/LU-305" title="utime() fails with EINTR : not conform to POSIX standard" class="issue-link" data-issue-key="LU-305"><del>LU-305</del></a>, I found that ll_file_write() (master code) always return error to user application where error happened, even if there is data be written, which is not POSIX complied. We should return written bytes if there is any written data.</p> LU-353

write should return written bytes when it's interrupted

Bug Major Open Unresolved WC Triage Niu Yawei Mon, 23 May 2011 00:20:08 +0000 Mon, 19 Dec 2011 11:24:43 +0000 Lustre 2.0.0 Lustre 2.1.0 0 2 <p>One difficlty withreturning the bytes written is that they may be sent asynchronously to different stripes, leaving a hole before some of the written bytes. It is easier to return an error to the application and have the VFS truncate the file on write error instead of trying to figure out the largest contiguous range of bytes written. </p> <p>Hi, Andreas</p> <p>As far as I can see, current ll_file_write() write to cache (or synchronously) stripe by stripe sequentially, so there shouldn't be hole problem. Xiong, could you confirm this?</p> Development Rank 1|hzw267: Rank (Obsolete) 10433 Severity