Explaining the new suspend quirks functionality in F7

Richard Hughes hughsient at gmail.com
Thu May 17 09:34:31 UTC 2007


On Thu, 2007-05-17 at 11:15 +0200, Thorsten Leemhuis wrote:
> 
> On 16.05.2007 16:33, Richard Hughes wrote:
> > New in Fedora 7 we have the new pm-utils and hal-info dmi based matching
> > of suspend quirks. We are doing finer matching to the laptop make and
> > model, to make suspend (and more importantly resume) work for more
> > people.
> > 
> > What this means:
> > 
> > * Some machines that suspended in FC6 might not work in F7
> > * Lots of machines that did not suspend in FC6 might work in F7
> > 
> > So, if you have to edit a file or add stuff to grub to get suspend
> > working in F7, that's a bug. This stuff should just work on the majority
> > or laptops.
> > 
> > To help, and to try and make clearer all this new stuff, I've written a
> > few pages here:
> > http://people.freedesktop.org/~hughsient/temp/quirk/quirk-intro.html
> 
> Some comments:
> 
> The page
> http://people.freedesktop.org/~hughsient/temp/quirk/quirk-debug.html
> only asks "Are you using the proprietary NVIDIA driver?" -- what about
> the proprietary ATI driver?

Good point, I'll add that too. What's the name of the proprietary driver
in xorg.conf?

> You're not asking if people use a plain vga consile (vga=0 or vga=normal
> on the kernel command line -- default in Fedora) or a framebuffer
> console (vga=792 or something like that -- default in Ubuntu, OpenSuse,
> no idea about Mandriva).

Sure, I need to debug this myself. I'll add this to the document.

> Further: Both details mentioned above according to my tests on several
> laptops heavily influence if laptops suspends or what quicks they need.
> So those informations thus IMHO should be tracked in the fdi files.
> Otherwise we soon might end in situations like "my laptops suspends fine
> under Fedora with the fdi file, but does not under OpenSuse (or vice versa).

Sure. Which would be bad. Thanks for your mail.

Richard.





More information about the fedora-devel-list mailing list