[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: COPY_DELAY could perform worse than COPY_NONE

From: Sergio López
Subject: Re: COPY_DELAY could perform worse than COPY_NONE
Date: Wed, 5 Oct 2011 20:41:08 +0200

2011/10/4 Samuel Thibault <samuel.thibault@gnu.org>:
> Sergio López, le Tue 04 Oct 2011 09:54:55 +0200, a écrit :
>> In objects where copy attribute is MEMORY_OBJECT_COPY_DELAY, for every
>> read operation (usually done with vm_copy) all pages in the source
>> object are write protected to be able to use copy-on-write
>> optimizations. The cost for this operation seems to be higher than
>> actually copying the pages for objects larger than just a bunch of
>> pages. Moreover, this CoW optimization is lost for requests larger
>> than 2k, since MiG stubs copy (with a conventional memcpy) the data
>> from the buffer received in the message, to the actual address
>> expected by the client.
> I see. I guess it'll be the same for other FSes, and maybe
> storeio/pager.c too?  Could you send a patch so that you show up as
> author of the change?

Sure, but I think applying this changes on all translators could be
dangerous. I've just tested ext2fs, I don't know if there could be
some side effects for other servers.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]