Comment 80 for bug 2056461

Revision history for this message
In , dhowells (dhowells-linux-kernel-bugs) wrote :

(In reply to Dominique Martinet from comment #58)
> Created attachment 306570 [details]
> tenative fix to allow reads past cached size for direct IO

How does it fix the O_NONBLOCK behaviour? Don't you also need to set NETFS_RREQ_BLOCKED too at some point?