[PATCH] pci.id performance patch

[Danny Kukawka]

On Friday 13 October 2006 12:35, Kay Sievers wrote:
> On Fri, 2006-10-13 at 12:10 +0200, Danny Kukawka wrote:
> > On Thursday 12 October 2006 18:23, Artem Kachitchkine wrote:
> > > FWIW, if we want HAL to run on embedded/limited resource platforms, we
> > > should probably invest time in making it more modular and customizable.
> > > 600KB is really nothing on a regular PC,
> >
> > And on a embedded plattform you don't have/should have such a big pci.id
> > file. Such machines have a limited number of known devices and should
> > also use a pci.id file with limited information.
>
> That's true for pci, but you never know what usb device will show up on
> a small device, and that's the same problem.

The patch only effected the pci.ids part.

> Why not compile it into a simple indexed berkeley db file. That's how
> every mailserver works with huge alias tables. That way you don't need
> to allocate any memory, have fast access and efficient lookup.

Not sure if this is possible with pci.ids because this file have IIRC a 3 
level tab seperated structure.

Danny