I am trying to refocus this thread from a particular issue to more generic needs... Regards, Boris > -----Original Message----- > From: Matthew Wilcox [mailto:willy@xxxxxxxxxxxxxxx] > Sent: Tuesday, September 30, 2014 3:24 PM > To: Zuckerman, Boris > Cc: Matthew Wilcox; Valdis.Kletnieks@xxxxxx; Matthew Wilcox; linux- > fsdevel@xxxxxxxxxxxxxxx; linux-mm@xxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx > Subject: Re: [PATCH v11 00/21] Add support for NV-DIMMs to ext4 > > On Tue, Sep 30, 2014 at 05:10:26PM +0000, Zuckerman, Boris wrote: > > > > > > The more I think about this, the more I think this is a bad idea. > > > When you have a file open with O_DIRECT, your I/O has to be done in > > > 512-byte multiples, and it has to be aligned to 512-byte boundaries > > > in memory. If an unsuspecting application has O_DIRECT forced on > > > it, it isn't going to know to do that, and so all its I/Os will fail. > > > It'll also be horribly inefficient if a program has the file mmaped. > > > > > > What problem are you really trying to solve? Some big files hogging the page > cache? > > > -- > > > > Page cache? As another copy in RAM? > > NV_DIMMs may be viewed as a caching device. This caching can be implemented on > the level of NV block/offset or may have some hints from FS and applications. > Temporary files is one example. They may not need to hit NV domain ever. Some > transactional journals or DB files is another example. They may stay in RAM until power > off. > > Boris, you're confused. Valdis is trying to solve an unrelated problem (and hopes my > DAX patches will do it for him). I'm explaining to him why what he wants to do is a bad > idea. This tangent is unrelated to NV-DIMMs. -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html