| Message ID | 1433291212-23367-2-git-send-email-toshi.kani@hp.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On Tue, 2015-06-02 at 18:26 -0600, Toshi Kani wrote: > When numa_off is set, NUMA is turned off and node 0 is the only > valid node on the system. The kernel skips parsing ACPI SRAT > table in this case. > > Change acpi_map_pxm_to_node() to always return 0 when numa_off > is set. Also move the range check of a proximity ID from > acpi_get_node() to acpi_map_pxm_to_node() after the numa_off > check. This keeps the interfaces to return 0 regardless of > proximity ID values. > > Signed-off-by: Toshi Kani <toshi.kani@hp.com> > --- > drivers/acpi/numa.c | 12 +++++++++--- > 1 file changed, 9 insertions(+), 3 deletions(-) > > diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c > index 1333cbdc..4898082 100644 > --- a/drivers/acpi/numa.c > +++ b/drivers/acpi/numa.c > @@ -70,7 +70,15 @@ static void __acpi_map_pxm_to_node(int pxm, int node) > > int acpi_map_pxm_to_node(int pxm) > { > - int node = pxm_to_node_map[pxm]; > + int node; > + > + if (numa_off) > + return 0; I found an issue in this patch. numa_off is only defined in x86, but this numa.c (ACPI_NUMA) can be enabled on x86 and IA64. I will fix or drop it in the next version. Thanks, -Toshi
diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c index 1333cbdc..4898082 100644 --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -70,7 +70,15 @@ static void __acpi_map_pxm_to_node(int pxm, int node) int acpi_map_pxm_to_node(int pxm) { - int node = pxm_to_node_map[pxm]; + int node; + + if (numa_off) + return 0; + + if (pxm < 0 || pxm >= MAX_PXM_DOMAINS) + return NUMA_NO_NODE; + + node = pxm_to_node_map[pxm]; if (node == NUMA_NO_NODE) { if (nodes_weight(nodes_found_map) >= MAX_NUMNODES) @@ -328,8 +336,6 @@ int acpi_get_node(acpi_handle handle) int pxm; pxm = acpi_get_pxm(handle); - if (pxm < 0 || pxm >= MAX_PXM_DOMAINS) - return NUMA_NO_NODE; return acpi_map_pxm_to_node(pxm); }
When numa_off is set, NUMA is turned off and node 0 is the only valid node on the system. The kernel skips parsing ACPI SRAT table in this case. Change acpi_map_pxm_to_node() to always return 0 when numa_off is set. Also move the range check of a proximity ID from acpi_get_node() to acpi_map_pxm_to_node() after the numa_off check. This keeps the interfaces to return 0 regardless of proximity ID values. Signed-off-by: Toshi Kani <toshi.kani@hp.com> --- drivers/acpi/numa.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-)