If the kernel has CONFIG_NUMA unset namespaces in sysfs will lack a
numa_node attribute.  In such cases ndctl will report a value of 0 for the
namespace numa_node in 'ndctl list'.  Instead of reporting potentially bad
data just hide the numa_node field if it is unsupported.

Signed-off-by: Ross Zwisler <ross.zwis...@linux.intel.com>
Fixes: commit f7d3de80a121 ("ndctl: support machines without numa")
---
 ndctl/lib/libndctl.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/ndctl/lib/libndctl.c b/ndctl/lib/libndctl.c
index ed5a65b..b7180e8 100644
--- a/ndctl/lib/libndctl.c
+++ b/ndctl/lib/libndctl.c
@@ -3008,6 +3008,8 @@ static void *add_namespace(void *parent, int id, const 
char *ndns_base)
        sprintf(path, "%s/numa_node", ndns_base);
        if (sysfs_read_attr(ctx, path, buf) == 0)
                ndns->numa_node = strtol(buf, NULL, 0);
+       else
+               ndns->numa_node = -1;
 
        sprintf(path, "%s/holder_class", ndns_base);
        if (sysfs_read_attr(ctx, path, buf) == 0)
-- 
2.14.3

_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm

Reply via email to