<div dir="ltr"><div dir="ltr"><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, May 30, 2023 at 3:02 PM Mike Snitzer <<a href="mailto:snitzer@kernel.org">snitzer@kernel.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
Also Joe, for you proposed dm-thinp design where you distinquish<br>
between "provision" and "reserve": Would it make sense for REQ_META<br>
(e.g. all XFS metadata) with REQ_PROVISION to be treated as an<br>
LBA-specific hard request? Whereas REQ_PROVISION on its own provides<br>
more freedom to just reserve the length of blocks? (e.g. for XFS<br>
delalloc where LBA range is unknown, but dm-thinp can be asked to<br>
reserve space to accomodate it).<br></blockquote><div><br></div><div>My proposal only involves 'reserve'. Provisioning will be done as part of the usual io path.</div><div><br></div><div>- Joe</div><div> </div></div></div>