| 1 |
Stephan Feder wrote: |
| 2 |
> MAL wrote: |
| 3 |
> |
| 4 |
>>Jesse Jacobs wrote: |
| 5 |
>> |
| 6 |
>>>I've read about a deadlock condition. |
| 7 |
>>>What would constitute a oversive tmpfs? |
| 8 |
>>>I thought anything unable to be allocated would be paged out to swap. |
| 9 |
>> |
| 10 |
>>No, if you run out of tmpfs, that's it.. you're out of disk space.. it |
| 11 |
>>wont use swap to emulate disk space. |
| 12 |
> |
| 13 |
> ... |
| 14 |
> |
| 15 |
> Not true. tmpfs lives in virtual memory, not physical memory only. |
| 16 |
|
| 17 |
Sorry, badly researched... either way, that's not going to aid |
| 18 |
performance at all.. actual file access is going to be faster than |
| 19 |
file->tmpfs->swap. |
| 20 |
|
| 21 |
MAL |
| 22 |
|
| 23 |
|
| 24 |
-- |
| 25 |
gentoo-performance@g.o mailing list |
Archives