Re: [PATCH v6 11/12] arm64: topology: enable ACPI/PPTT based CPU topology
From: Jeremy Linton
Date: Fri Feb 23 2018 - 23:37:45 EST
On 02/23/2018 05:02 AM, Lorenzo Pieralisi wrote:
On Thu, Jan 25, 2018 at 09:56:30AM -0600, Jeremy Linton wrote:
Hi,
On 01/25/2018 06:15 AM, Xiongfeng Wang wrote:
Hi Jeremy,
I have tested the patch with the newest UEFI. It prints the below error:
[ 4.017371] BUG: arch topology borken
[ 4.021069] BUG: arch topology borken
[ 4.024764] BUG: arch topology borken
[ 4.028460] BUG: arch topology borken
[ 4.032153] BUG: arch topology borken
[ 4.035849] BUG: arch topology borken
[ 4.039543] BUG: arch topology borken
[ 4.043239] BUG: arch topology borken
[ 4.046932] BUG: arch topology borken
[ 4.050629] BUG: arch topology borken
[ 4.054322] BUG: arch topology borken
I checked the code and found that the newest UEFI set PPTT physical_package_flag on a physical package node and
the NUMA domain (SRAT domains) starts from the layer of DIE. (The topology of our board is core->cluster->die->package).
I commented about that on the EDK2 mailing list. While the current spec
doesn't explicitly ban having the flag set multiple times between the leaf
and the root I consider it a "bug" and there is an effort to clarify the
spec and the use of that flag.
When the kernel starts to build sched_domain, the multi-core sched_domain contains all the cores within a package,
and the lowest NUMA sched_domain contains all the cores within a die. But the kernel requires that the multi-core
sched_domain should be a subset of the lowest NUMA sched_domain, so the BUG info is printed.
Right. I've mentioned this problem a couple of times.
At at the moment, the spec isn't clear about how the proximity domain is
detected/located within the PPTT topology (a node with a 1:1 correspondence
isn't even required). As you can see from this patch set, we are making the
general assumption that the proximity domains are at the same level as the
physical socket. This isn't ideal for NUMA topologies, like the D05, that
don't align with the physical socket.
There are efforts underway to clarify and expand upon the specification to
deal with this general problem. The simple solution is another flag (say
PPTT_PROXIMITY_DOMAIN which would map to the D05 die) which could be used to
find nodes with 1:1 correspondence. At that point we could add a fairly
trivial patch to correct just the scheduler topology without affecting the
rest of the system topology code.
I think Morten asked already but isn't this the same end result we end
up having if we remove the DIE level if NUMA-within-package is detected
(instead of using the default_topology[]) and we create our own ARM64
domain hierarchy (with DIE level removed) through set_sched_topology()
accordingly ?
I'm not sure what removing the die level does for you, but its not
really the problem AFAIK, the problem is because MC layer is larger than
the NUMA domains.
Put it differently: do we really need to rely on another PPTT flag to
collect this information ?
Strictly no, and I have a partial patch around here i've been meaning to
flush out which uses the early node information to detect if there are
nodes smaller than the package. Initially I've been claiming i was going
to stay away from making scheduler topology changes in this patch set,
but it seems that at least providing a patch which does the minimal bits
is in the cards. The PXN flag was is more of a shortcut to finding the
cache levels at or below the numa domains, rather than any hard
requirement. Similarly, to the request someone else was making for a
leaf node flag (or node ordering) to avoid multiple passes in the table.
That request would simplify the posted code a bit but it works without it.
I can't merge code that breaks a platform with legitimate firmware
bindings.
Breaks in this case is a BUG warning that shows up right before it
"corrects" a scheduler domain.
Basically, as i've mentioned a few times, this patch set corrects the
existing topology problems, in doing so it uncovers issues with the way
we are mapping that topology for the scheduler. That is actually not
difficult thing to fix, my assumption originally is that we would
already be at the point of discussion the finer points of the scheduler
changes but we are still here.
Anyway, I was planning on posting a v7 this week, but time flys... I
will include a further scheduler tweak to work around the inverted numa
domain problem in that set early next week.
Thanks,
Thanks,
Lorenzo
If we modify the UEFI to make NUMA sched_domain start from the layer of package, then all the topology information
within the package will be discarded. I think we need to build the multi-core sched_domain using the cores within
the cluster instead of the cores within the package. I think that's what 'multi-core' means. Multi cores form a cluster. I guess.
If we build the multi-core sched_domain using the cores within a cluster, I think we need to add fields in struct cpu_topology
to record which cores are in each cluster.
The problem is that there isn't a generic way to identify which level of
cache sharing is the "correct" top layer MC domain. For one system cluster
might be appropriate, for another it might be the highest caching level
within a socket, for another is might be a something in between or a group
of clusters or LLCs..
Hence the effort to standardize/guarantee a PPTT node that exactly matches a
SRAT domain. With that, each SOC/system provider has clearly defined method
for communicating where they want the proximity domain information to begin.
Thanks,
Thanks,
Xiongfeng
On 2018/1/13 8:59, Jeremy Linton wrote:
Propagate the topology information from the PPTT tree to the
cpu_topology array. We can get the thread id, core_id and
cluster_id by assuming certain levels of the PPTT tree correspond
to those concepts. The package_id is flagged in the tree and can be
found by calling find_acpi_cpu_topology_package() which terminates
its search when it finds an ACPI node flagged as the physical
package. If the tree doesn't contain enough levels to represent
all of the requested levels then the root node will be returned
for all subsequent levels.
Cc: Juri Lelli <juri.lelli@xxxxxxx>
Signed-off-by: Jeremy Linton <jeremy.linton@xxxxxxx>
---
arch/arm64/kernel/topology.c | 46 +++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 45 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/kernel/topology.c b/arch/arm64/kernel/topology.c
index 7b06e263fdd1..ce8ec7fd6b32 100644
--- a/arch/arm64/kernel/topology.c
+++ b/arch/arm64/kernel/topology.c
@@ -11,6 +11,7 @@
* for more details.
*/
+#include <linux/acpi.h>
#include <linux/arch_topology.h>
#include <linux/cpu.h>
#include <linux/cpumask.h>
@@ -22,6 +23,7 @@
#include <linux/sched.h>
#include <linux/sched/topology.h>
#include <linux/slab.h>
+#include <linux/smp.h>
#include <linux/string.h>
#include <asm/cpu.h>
@@ -300,6 +302,46 @@ static void __init reset_cpu_topology(void)
}
}
+#ifdef CONFIG_ACPI
+/*
+ * Propagate the topology information of the processor_topology_node tree to the
+ * cpu_topology array.
+ */
+static int __init parse_acpi_topology(void)
+{
+ bool is_threaded;
+ int cpu, topology_id;
+
+ is_threaded = read_cpuid_mpidr() & MPIDR_MT_BITMASK;
+
+ for_each_possible_cpu(cpu) {
+ topology_id = find_acpi_cpu_topology(cpu, 0);
+ if (topology_id < 0)
+ return topology_id;
+
+ if (is_threaded) {
+ cpu_topology[cpu].thread_id = topology_id;
+ topology_id = find_acpi_cpu_topology(cpu, 1);
+ cpu_topology[cpu].core_id = topology_id;
+ topology_id = find_acpi_cpu_topology_package(cpu);
+ cpu_topology[cpu].package_id = topology_id;
+ } else {
+ cpu_topology[cpu].thread_id = -1;
+ cpu_topology[cpu].core_id = topology_id;
+ topology_id = find_acpi_cpu_topology_package(cpu);
+ cpu_topology[cpu].package_id = topology_id;
+ }
+ }
+
+ return 0;
+}
+
+#else
+static inline int __init parse_acpi_topology(void)
+{
+ return -EINVAL;
+}
+#endif
void __init init_cpu_topology(void)
{
@@ -309,6 +351,8 @@ void __init init_cpu_topology(void)
* Discard anything that was parsed if we hit an error so we
* don't use partial information.
*/
- if (of_have_populated_dt() && parse_dt_topology())
+ if ((!acpi_disabled) && parse_acpi_topology())
+ reset_cpu_topology();
+ else if (of_have_populated_dt() && parse_dt_topology())
reset_cpu_topology();
}