Re: Random x86-64 seg-fault finally fixed

[Tim Darby <t+dfbsd@xxxxxxxxxxxx>]

Here's hoping everyone has an MPSAFE New Year!  And I'm hoping one of these 48-core monsters shows up under my Christmas tree.

Tim

On Wed, Dec 22, 2010 at 1:19 PM, Matthew Dillon <dillon@apollo.backplane.com> wrote:

  The random utility seg-fault (usually cc1) on x86-64 appears to have
  finally been fixed.  It turned out to be a vm_page race in the pageout
  demon when it cycles pages onto the free page list.

  This clears the way for running the system with the remaining global
  tokens set to MPSAFE mode.  Test boxes are now running that way with no
  errors so I expect we will be changing the tokens to run MPSAFE after
  christmas sometime.  We do still have to audit the code paths.

  I also expect to have a fine-grained VM solution at least for standard
  VM faults by the next release, hopefully sooner.  It appears to be the
  biggest bottleneck on the monster 48-core test box now.

  Sascha has brought in the mps disk driver which is now undergoing
  testing.  This is the last big chipset support piece needed for 48-core
  opteron supermicro support.  The Gigabit ethernet uses the igb driver
  which seems to work pretty well in polling mode.

                                       -Matt
                                       Matthew Dillon
                                       <dillon@backplane.com>