On Mon, Sep 26, 2022 at 04:39:59PM +0200, Christoph Hellwig wrote: > On Mon, Sep 26, 2022 at 01:23:25AM +0300, Serge Semin wrote: > > IMO both the approaches seem unclear if a reader doesn't know what > > they have been introduced for. Anyway do you insist on using the > > kmalloc-ed buffer here instead? If so I'll resubmit the series with > > this patch updated accordingly. > > I don't like the __aligend version too much, but I can live with it if > you strongly prefer it. What about converting the NVME hwmon driver to using the kmalloc'ed buffer (seeing the rest of the NVME core driver prefer kmallocing the buffers) and adding the ____cacheline_aligned modifier to the opal_dev buffers? -Sergey
