| Message ID | 8fce2c60a0095fffdab12f2204eaa71b387882e3.1418231681.git.joseph.salisbury@canonical.com |
|---|---|
| State | New |
| Headers | show |
On Wed, Dec 10, 2014 at 01:13:16PM -0500, Joseph Salisbury wrote: > From: Dave Hansen <dave.hansen@linux.intel.com> > > BugLink: http://bugs.launchpad.net/bugs/1386473 > > Commit: > > cebf15eb09a2 ("x86, sched: Add new topology for multi-NUMA-node CPUs") > > some code to try to detect the situation where we have a NUMA node > inside of the "DIE" sched domain. > > It detected this by looking for cpus which match_die() but do not match > NUMA nodes via topology_same_node(). > > I wrote it up as: > > if (match_die(c, o) == !topology_same_node(c, o)) > > which actually seemed to work some of the time, albiet > accidentally. > > It should have been doing an &&, not an ==. > > This code essentially chopped off the "DIE" domain on one of > Andrew Morton's systems. He reported that this patch fixed his > issue. > > Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> > Reported-by: Andrew Morton <akpm@linux-foundation.org> > Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> > Cc: Dave Hansen <dave@sr71.net> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: David Rientjes <rientjes@google.com> > Cc: Igor Mammedov <imammedo@redhat.com> > Cc: Jan Kiszka <jan.kiszka@siemens.com> > Cc: Lan Tianyu <tianyu.lan@intel.com> > Cc: Linus Torvalds <torvalds@linux-foundation.org> > Cc: Prarit Bhargava <prarit@redhat.com> > Cc: Toshi Kani <toshi.kani@hp.com> > Link: http://lkml.kernel.org/r/20140930214546.FD481CFF@viggo.jf.intel.com > Signed-off-by: Ingo Molnar <mingo@kernel.org> > (cherry picked from commit 728e5653e6fdb2a0892e94a600aef8c9a036c7eb) > > Signed-off-by: Joseph Salisbury <joseph.salisbury@canonical.com> > --- > arch/x86/kernel/smpboot.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c > index cf12440..c8748a1 100644 > --- a/arch/x86/kernel/smpboot.c > +++ b/arch/x86/kernel/smpboot.c > @@ -449,7 +449,7 @@ void set_cpu_sibling_map(int cpu) > } else if (i != cpu && !c->booted_cores) > c->booted_cores = cpu_data(i).booted_cores; > } > - if (match_die(c, o) == !topology_same_node(c, o)) > + if (match_die(c, o) && !topology_same_node(c, o)) > primarily_use_numa_for_topology(); > } > } > -- > 1.7.9.5 > > > -- > kernel-team mailing list > kernel-team@lists.ubuntu.com > https://lists.ubuntu.com/mailman/listinfo/kernel-team Looks good.
On Wed, Dec 10, 2014 at 01:13:16PM -0500, Joseph Salisbury wrote: > From: Dave Hansen <dave.hansen@linux.intel.com> > > BugLink: http://bugs.launchpad.net/bugs/1386473 > > Commit: > > cebf15eb09a2 ("x86, sched: Add new topology for multi-NUMA-node CPUs") > > some code to try to detect the situation where we have a NUMA node > inside of the "DIE" sched domain. > > It detected this by looking for cpus which match_die() but do not match > NUMA nodes via topology_same_node(). > > I wrote it up as: > > if (match_die(c, o) == !topology_same_node(c, o)) > > which actually seemed to work some of the time, albiet > accidentally. > > It should have been doing an &&, not an ==. > > This code essentially chopped off the "DIE" domain on one of > Andrew Morton's systems. He reported that this patch fixed his > issue. > > Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> > Reported-by: Andrew Morton <akpm@linux-foundation.org> > Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> > Cc: Dave Hansen <dave@sr71.net> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: David Rientjes <rientjes@google.com> > Cc: Igor Mammedov <imammedo@redhat.com> > Cc: Jan Kiszka <jan.kiszka@siemens.com> > Cc: Lan Tianyu <tianyu.lan@intel.com> > Cc: Linus Torvalds <torvalds@linux-foundation.org> > Cc: Prarit Bhargava <prarit@redhat.com> > Cc: Toshi Kani <toshi.kani@hp.com> > Link: http://lkml.kernel.org/r/20140930214546.FD481CFF@viggo.jf.intel.com > Signed-off-by: Ingo Molnar <mingo@kernel.org> > (cherry picked from commit 728e5653e6fdb2a0892e94a600aef8c9a036c7eb) > > Signed-off-by: Joseph Salisbury <joseph.salisbury@canonical.com> > --- > arch/x86/kernel/smpboot.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c > index cf12440..c8748a1 100644 > --- a/arch/x86/kernel/smpboot.c > +++ b/arch/x86/kernel/smpboot.c > @@ -449,7 +449,7 @@ void set_cpu_sibling_map(int cpu) > } else if (i != cpu && !c->booted_cores) > c->booted_cores = cpu_data(i).booted_cores; > } > - if (match_die(c, o) == !topology_same_node(c, o)) > + if (match_die(c, o) && !topology_same_node(c, o)) > primarily_use_numa_for_topology(); > } > } > -- > 1.7.9.5 > > > -- > kernel-team mailing list > kernel-team@lists.ubuntu.com > https://lists.ubuntu.com/mailman/listinfo/kernel-team Somehow the first line of the commit text got dropped. I've added that back and applied the commit to Utopic master-next
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index cf12440..c8748a1 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -449,7 +449,7 @@ void set_cpu_sibling_map(int cpu) } else if (i != cpu && !c->booted_cores) c->booted_cores = cpu_data(i).booted_cores; } - if (match_die(c, o) == !topology_same_node(c, o)) + if (match_die(c, o) && !topology_same_node(c, o)) primarily_use_numa_for_topology(); } }