On Mon, May 13, 2019 at 8:29 AM Christoph Hellwig <h...@infradead.org> wrote: > > On Mon, May 13, 2019 at 08:20:03AM -0400, tedheadster wrote: > > On Mon, May 13, 2019 at 3:02 AM Christoph Hellwig <h...@infradead.org> > > wrote: > > > > > > On Sun, May 12, 2019 at 02:55:48PM -0400, tedheadster wrote: > > > > Christoph, > > > > On the same hardware (reboot with different kernel) I am getting > > > > _horrible_ disk I/O performance on the 5.1.1. kernel compiled on a > > > > 32-bit platform using HIGHMEM64G (PAE) to access 32GiB of physical > > > > memory. > > > > > > > > The numbers are truly terrible to copy a 16GiB file from one disk to a > > > > different one: > > >
Christoph, I believe I found the problem, and it does not relate to anything I considered before. I forgot that I had chosen the SLOB memory allocator for a previous test and it was still enabled. There was a huge amount of locking slowing the system down while SLOB was allocating new memory with its simple algorithm. Switching to SLUB has improved it immensely. I am sorry I missed this rather important item. - Matthew
