| 1 |
On Donnerstag, 8. März 2007, Philip Webb wrote: |
| 2 |
> 070306 Mick wrote: |
| 3 |
> > On Tuesday 06 March 2007 21:22, Philip Webb wrote: |
| 4 |
> >> During the past few months, I've been bit by sudden total freezes |
| 5 |
> >> while using Epiphany & Konqueror to browse Internet sites. |
| 6 |
> >> The only way out is to hit the reset button: eg Ctl-Alt-F2 doesn't work. |
| 7 |
> > |
| 8 |
> > The only time I had suffered similar symptoms (but different |
| 9 |
> > applications) was when I had a bad memory problem. Memtest86 did not |
| 10 |
> > show anything and it was only through trial and error that I found the |
| 11 |
> > faulty module. |
| 12 |
> |
| 13 |
> I emerged Memtest86 & ran it -- it takes 1 hr for my 1 GB memory -- , |
| 14 |
> but I can't find anything to explain the output. It found 2 errors : |
| 15 |
> |
| 16 |
> ... 0000c6e37c0 198.1MB ... |
| 17 |
> ... 0000d6e37a0 214.1MB ... |
| 18 |
> |
| 19 |
> I realise that these lines refer to hex locations, but I don't understand |
| 20 |
> the reference to 'MB' (perhaps they're the same info in a different form). |
| 21 |
> I have a note that : |
| 22 |
|
| 23 |
yes it is the same info. MB = Megabyte. It says, that the errors are at 198,1 |
| 24 |
and 214,1MB. |
| 25 |
|
| 26 |
The problem: memtest86 and memtest86+ both find sometimes errors that aren't |
| 27 |
ones. Or don't find them, when there are some. Let it run several times to |
| 28 |
make sure - and then replace the stick. |
| 29 |
|
| 30 |
> |
| 31 |
> To tell kernel not to use bad patch of memory, append via Lilo : |
| 32 |
> 'badram=<m>,<n>', where m n are obtained via 'memtest'. |
| 33 |
> |
| 34 |
> However, the above lines don't give pairs of addresses, only 1 each. |
| 35 |
> |
| 36 |
> Yes, I did look at the official site & the installed README. |
| 37 |
> Does anyone have further advice ? |
| 38 |
|
| 39 |
yes, get new ram. |
| 40 |
Everything else is not safe. |
| 41 |
|
| 42 |
|
| 43 |
|
| 44 |
-- |
| 45 |
gentoo-user@g.o mailing list |
Archives