Date: Sat, 30 Dec 2023 16:58:00 +0000 From: bugzilla-noreply@freebsd.org To: fs@FreeBSD.org Subject: [Bug 276002] nfscl: data corruption using both copy_file_range and mmap'd I/O Message-ID: <bug-276002-3630-eUPCNojKmi@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-276002-3630@https.bugs.freebsd.org/bugzilla/> References: <bug-276002-3630@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D276002 --- Comment #21 from Rick Macklem <rmacklem@FreeBSD.org> --- Created attachment 247360 --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=3D247360&action= =3Dedit Don't clear the dirty bits on pages being written through the buffer cache You could try this completely untested patch. The last time we played this game, the problem was that a call to vfs_busy_pages() cleared the dirty bits when it should not have done so. This is the only other place where the NFS client code does the same thing. Note that I think kib@'s copy_file_range patch is needed. I plan to do some testing of it, to see if I can break copy_file_range() without the patch and without using mmap'd I/O. --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-276002-3630-eUPCNojKmi>