Messages in this thread | | | Date | Fri, 26 Jun 1998 09:22:04 -0700 (PDT) | From | David Lang <> | Subject | Re: uniform input device packets? |
| |
-----BEGIN PGP SIGNED MESSAGE-----
but what about multiple events? (move the mouse diagnally while clicking a button) 3 2/4 byte reads+mangle could be combined to 1 cache read while 3 16 byte reads would be 3 cache reads.
David Lang
On Fri, 26 Jun 1998, Mike Jagdis wrote:
> Date: Fri, 26 Jun 1998 12:13:34 +0100 (GMT/BST) > From: Mike Jagdis <mike@roan.co.uk> > To: Allanah Myles <dossy@panoptic.com> > Cc: linux-kernel@vger.rutgers.edu > Subject: Re: uniform input device packets? > > On Fri, 26 Jun 1998, Allanah Myles wrote: > > > I'd rather read 2 bytes and mangle their bits around rather than > > having to, say, read 16 bytes to obtain the same amount of data. > > Unless, you can tell me that it's faster to just read the 16 > > bytes. > > Well, lets think about it for a minute... Cache loads happen > in terms of groups of bytes so reading a few more bytes is > not very low cost. Write gathering and write back means > that writing a few more bytes is also low cost. Instructions > to move data to and from memory can typically execute in > parallel on pipelined processors so moving a few more bytes > is also low cost again. > > Now on the 6x86 at least bit shift operations are only > performed in the X pipeline, cause a pipeline stall, and > are relatively slow compared to memory loads and stores. > You can do a lot of memory loads and stores in the time > it takes to do a bit shift. > > I think you might find that a 16 byte read is actually > faster than a 2 byte read and mangle... > > Mike > > -- > .----------------------------------------------------------------------. > | Mike Jagdis | Internet: mailto:mike@roan.co.uk | > | Roan Technology Ltd. | | > | 54A Peach Street, Wokingham | Telephone: +44 118 989 0403 | > | RG40 1XG, ENGLAND | Fax: +44 118 989 1195 | > `----------------------------------------------------------------------' > > > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.rutgers.edu >
-----BEGIN PGP SIGNATURE----- Version: PGP for Personal Privacy 5.0 Charset: noconv
iQEVAwUBNZPKsT7msCGEppcbAQECiAgAsAkL5gKW0rtEqr5PqaLebGvBZQOTfkfx gjPLJiR9Xh0PQ3RHPnGrRk4v9vQy6gj/Glc61niTj8e2TAXfOSvJqU0w4Tm2SSbu PqVcOs40McO1tYPjWX1lZ8cnzFBZEIui05o3Sepg47RvNgjkZfU3LsJcvKBb3UfC o4AOKwZ73bhjUVhJTwmijmLQsw7eYKIUHU3cc1o/lTAc9qjX9320dXGMoEkugYCg MT998c+lIevQ2TKsXad4oomV4TTulS56kqLyr4IA9VRVxk6Uj59vvRQESjgiXLmb +9RN48kiarGHFK6TN5Zt9caw2b/Q643XfiixgT5pdceHbqmudQ8qFg== =Z9WR -----END PGP SIGNATURE-----
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
| |