From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Wilcox Date: Tue, 4 Feb 2020 09:16:31 -0800 Subject: [Cluster-devel] [PATCH v4 00/12] Change readahead API In-Reply-To: <20200204153227.GF2654@twin.jikos.cz> References: <20200201151240.24082-1-willy@infradead.org> <20200204153227.GF2654@twin.jikos.cz> Message-ID: <20200204171631.GM8731@bombadil.infradead.org> List-Id: To: cluster-devel.redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit On Tue, Feb 04, 2020 at 04:32:27PM +0100, David Sterba wrote: > On Sat, Feb 01, 2020 at 07:12:28AM -0800, Matthew Wilcox wrote: > > From: "Matthew Wilcox (Oracle)" > > > > I would particularly value feedback on this from the gfs2 and ocfs2 > > maintainers. They have non-trivial changes, and a review on patch 5 > > would be greatly appreciated. > > > > This series adds a readahead address_space operation to eventually > > replace the readpages operation. The key difference is that > > pages are added to the page cache as they are allocated (and > > then looked up by the filesystem) instead of passing them on a > > list to the readpages operation and having the filesystem add > > them to the page cache. It's a net reduction in code for each > > implementation, more efficient than walking a list, and solves > > the direct-write vs buffered-read problem reported by yu kuai at > > https://lore.kernel.org/linux-fsdevel/20200116063601.39201-1-yukuai3 at huawei.com/ > > > > v4: > > - Rebase on current Linus (a62aa6f7f50a ("Merge tag 'gfs2-for-5.6'")) > > I've tried to test the patchset but haven't got very far, it crashes at boot > ritht after VFS mounts the root. The patches are from mailinglist, applied on > current master, bug I saw the same crash with the git branch in your > repo (probably v1). Yeah, I wasn't able to test at the time due to what turned out to be the hpet bug in Linus' tree. Now that's fixed, I've found & fixed a couple more bugs. There'll be a v5 once I fix the remaining problem (looks like a missing page unlock somewhere).