From: William Roche <william.ro...@oracle.com>
The Corrected Error events collected by the cec_add_elem() have to be
consistently filtered out.
We fix the case where the value of find_elem() to find the slot of a pfn
was mistakenly used as the return value of the function.
Now the MCE notifiers chain relying on MCE_HANDLED_CEC would only report
filtered corrected errors that reached the action threshold.
Signed-off-by: William Roche <william.ro...@oracle.com>
---
Notes:
This is the new patch version using an additional 'err' variable.
Unit tested it on a VM instance and a "Bare Metal" machine.
No reporting is done by the MCE_HANDLED_CEC aware notifiers until
the action threshold is reached.
drivers/ras/cec.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/ras/cec.c b/drivers/ras/cec.c
index ddecf25..b926c67 100644
--- a/drivers/ras/cec.c
+++ b/drivers/ras/cec.c
@@ -312,8 +312,8 @@ static bool sanity_check(struct ce_array *ca)
static int cec_add_elem(u64 pfn)
{
struct ce_array *ca = &ce_arr;
+ int count, err, ret = 0;
unsigned int to = 0;
- int count, ret = 0;
/*
* We can be called very early on the identify_cpu() path where we are
@@ -330,8 +330,8 @@ static int cec_add_elem(u64 pfn)
if (ca->n == MAX_ELEMS)
WARN_ON(!del_lru_elem_unlocked(ca));
- ret = find_elem(ca, pfn, &to);
- if (ret < 0) {
+ err = find_elem(ca, pfn, &to);
+ if (err < 0) {
/*
* Shift range [to-end] to make room for one more element.
*/
--
1.8.3.1
