Summary: | hdparm init script should flush disk cache on stop. | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Forza <forza> |
Component: | [OLD] Core system | Assignee: | Heinrich Wendel (RETIRED) <lanius> |
Status: | RESOLVED WONTFIX | ||
Severity: | enhancement | CC: | base-system, flash3001 |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- |
Description
Forza
2004-09-29 02:07:34 UTC
Maybe it should not only be in the hdparm script, but also when reboot/halt scripts? For example they could call just before power is turned off: hdparm -W0 -f /dev/hd? sync isn't hdparm -f /dev/hdx enough? why do we also need "sync" and "hdparm -W0" ? Doesn't sync only flush things in the OS/filesystem cache and buffers? I am not sure if both -f and -W0 is needed. I suggested W0 so no more data would enter the cache after flush and sync. maybe anybody here with more knowledge about this topic? I checked on various IRC channels such as #kernel etc and I could not get a clear answer. No one seemed to know how to verify that the harddisk indeed has flushed its caches or not. well, since no other distribution seems to do it, i would go for marking this bug "wontfix", is that ok with you? Yes of course. It was a suggestion/enhancement question, not really a bug. However, I do have sometimes noticed (perhaps it is a filesystem problem and not a cache problem) that somtimes there are errors in the filesystem when shutting down or rebooting the computer. That's how I came to test this solution myself. Ok, maybe you come up with a real explanation/solution one time, then just reopen this one :-) |