The newly introduced 'perf_stats' attribute uses the default access
mode of 0444 letting non-root users access performance stats of an
nvdimm and potentially force the kernel into issuing large number of
expensive HCALLs. Since the information exposed by this attribute
cannot be cached hence its better to ward of access to this attribute
from non-root users.
Hence this patch updates the access-mode of 'perf_stats' sysfs
attribute file to 0400 to make it only readable to root-users.
Reported-by: Aneesh Kumar K.V <redacted>
Signed-off-by: Vaibhav Jain <redacted>
---
arch/powerpc/platforms/pseries/papr_scm.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/powerpc/platforms/pseries/papr_scm.c b/arch/powerpc/platforms/pseries/papr_scm.c
index f439f0dfea7d1..31864d167a2ce 100644
--- a/arch/powerpc/platforms/pseries/papr_scm.c
+++ b/arch/powerpc/platforms/pseries/papr_scm.c
@@ -822,7 +822,7 @@ static ssize_t perf_stats_show(struct device *dev,
kfree(stats);
return rc ? rc : seq_buf_used(&s);
}
-DEVICE_ATTR_RO(perf_stats);
+DEVICE_ATTR(perf_stats, 0400, perf_stats_show, NULL);
static ssize_t flags_show(struct device *dev,
struct device_attribute *attr, char *buf)
--
2.26.2
Vaibhav Jain [off-list ref] writes:
The newly introduced 'perf_stats' attribute uses the default access
mode of 0444 letting non-root users access performance stats of an
nvdimm and potentially force the kernel into issuing large number of
expensive HCALLs. Since the information exposed by this attribute
cannot be cached hence its better to ward of access to this attribute
from non-root users.
Hence this patch updates the access-mode of 'perf_stats' sysfs
attribute file to 0400 to make it only readable to root-users.
Or should we ratelimit it?
Fixes: ??
Reported-by: Aneesh Kumar K.V <redacted>
Signed-off-by: Vaibhav Jain <redacted>
cheers
Hi Mpe,
Thanks for reviewing this patch. My responses below:
Michael Ellerman [off-list ref] writes:
Vaibhav Jain [off-list ref] writes:
quoted
The newly introduced 'perf_stats' attribute uses the default access
mode of 0444 letting non-root users access performance stats of an
nvdimm and potentially force the kernel into issuing large number of
expensive HCALLs. Since the information exposed by this attribute
cannot be cached hence its better to ward of access to this attribute
from non-root users.
Hence this patch updates the access-mode of 'perf_stats' sysfs
attribute file to 0400 to make it only readable to root-users.
Or should we ratelimit it?
Ideal consumers of this data will be users with CAP_PERFMON or
CAP_SYS_ADMIN. Also they need up-to-date values for these performance stats
as these values can be time sensitive.
So rate limiting may not be a complete solution since a user running
'perf' might be throttled by another user who is simply reading the
sysfs file contents.
So instead of setting attribute mode to 0400, will add a check for
'perfmon_capable()' in perf_stats_show() denying read access to users
without CAP_PERFMON or CAP_SYS_ADMIN.
Fixes: ??
Right. I will add this in v2.
quoted
Reported-by: Aneesh Kumar K.V <redacted>
Signed-off-by: Vaibhav Jain <redacted>
cheers
--
Cheers
~ Vaibhav