Messages in this thread | | | Date | Tue, 26 Oct 2010 09:57:45 -0400 | From | "John Stoffel" <> | Subject | Re: [PATCH 06/11] IMA: use i_writecount rather than a private counter |
| |
>>>>> "Eric" == Eric Paris <eparis@redhat.com> writes:
Eric> On Mon, 2010-10-25 at 15:25 -0700, H. Peter Anvin wrote: >> On 10/25/2010 02:52 PM, Eric Paris wrote: >> > On Mon, 2010-10-25 at 15:27 -0400, John Stoffel wrote: >> > >> >> The problems with kernel.org is a perfect exmaple of how an annocuous >> >> feature like this, can kill a system's performance. >> > >> > You admit that you don't know what you are talking about and then state >> > that this kills systems performance. Interesting conclusion. >> > >> > I'm not going to try to refute you point by point but will instead paint >> > a broad picture. I see 3 possible states: >> > 1) Configured out - 0 overhead. period. >> > 2) Configured in but default disabled >> > 3) Configured in and enabled by admin intervention >> > >> > I have (I think) pretty clearly discussed the overhead and the changes >> > made in case #2. We expand struct inode by 4 bytes, we increment and >> > decrement those 4 bytes on open/close() and we use a new inode->i_flags. >> > >> >> Case #2 is the bad one, as long as distros are likely to compile it in.
Eric> Agreed. And that's the case this whole patch series is addressing. It Eric> makes it (literally not figuratively) hundreds of times better than it Eric> is today :)
And just to chime in, I really appreciate your hard work on this cleanup!
John
| |