On Wed, Jan 30, 2019 at 08:57:47AM +0100, Ulf Hansson wrote: > On Mon, 14 Jan 2019 at 10:58, Christoph Hellwig <hch@xxxxxx> wrote: > > > > Instead of setting up a kernel pointer to track the current PIO address, > > track the offset in the current page, and do an atomic kmap for the page > > while doing the actual PIO operations. > > > > Signed-off-by: Christoph Hellwig <hch@xxxxxx> > > Nitpick: This one have some trailing whitespace errors, which git > complains about when I apply it. > > No need to re-spin due to this, I can fix it up. I'll need to respin to limit the segment size to a page anyway, so I will take this into account as well for v2.
