1 |
On Wed, Mar 20, 2013 at 11:14 AM, Grant Edwards |
2 |
<grant.b.edwards@×××××.com> wrote: |
3 |
> |
4 |
> On 2013-03-20, Carlos Hendson <skyclan@×××.net> wrote: |
5 |
> |
6 |
> > That's by no means conclusive, however, I've also run a complete pass of |
7 |
> > memcheck for over an hour without any issues reported. |
8 |
> |
9 |
> FWIW. I've had flakey memory that ran memcheck fine for several hours |
10 |
> and multiple passes -- but if I let it run long enough, it would fail. |
11 |
> I wouldn't be confident unless memtest ran for at least 12 hours (24 |
12 |
> would be even better). |
13 |
|
14 |
|
15 |
When I've had memory problems, it seemed like it was always shown in |
16 |
tests 5 and 8 from memtest86+. So, now, to expedite the tests, I set |
17 |
it to only run tests 5 and 8. A few hours of those can find problems |
18 |
faster than a couple days of running the full battery of tests. I |
19 |
always run the full set at least once, but my experience and those of |
20 |
people I've seen on Google seems to indicate that on modern systems 5 |
21 |
and 8 are where errors are exposed. YMMV of course. :) |