Re: [PATCH 0/3] Fix two PEIe hotplug issues

[Kristen Carlson Accardi <kristen.c.accardi@intel.com>]

On Thu, 18 Oct 2007 13:06:21 -0400
Mark Lord <lkml@rtr.ca> wrote:

> Kristen Carlson Accardi wrote:
> > On Wed, 17 Oct 2007 23:09:45 -0400
> > Mark Lord <lkml@rtr.ca> wrote:
> > 
> >> Mark Lord wrote:
> >>> Fix PCIe Hotplug so that it works with ExpressCard slots on Dell notebooks
> >>> (and others?) in conjunction with the modparam of pciehp_force=1.
> >>
> >> To make things simpler for distro people, I'm contemplating another patch
> >> in this series, to allow something like:  pciehp_force=2
> >>
> >> This would then *try* the ACPI BIOS calls, and only fall back to pcie_force=1
> >> if the BIOS support fails.  This would be an ideal default for most desktop/notebook
> >> distros to consider using.
> >>
> >> Without this, they don't have any good choice:  use the defaults, and things fail
> >> on the most popular brand of machines in the marketplace.  Use pciehp_force=1,
> >> and they may break (?) on other brands.
> >>
> >> So a hybrid of the two seems best.  Pity it couldn't be the default behaviour, though.
> >> Or could it?  We're early enough in the 2.6.24 cycle for it..
> >>
> >> Opinions?
> >>
> > 
> > NAK!  Absolutely no way will I take a patch that does this.  
> 
> 
> No big deal.  I personally don't have a distro to maintain,
> so no pain for me here.
> 
> > I've been actually having philosophical issues with
> > even having pciehp_force as a module parameter at all.  As I said before,
> > using pciehp_force violates the PCI spec.
> 
> No, it just provides a way to use the hardware when the vendor
> didn't include BIOS functionality for it.

No, it actually does violate the spec.  Feel free to read it yourself.
We are not supposed to do Native PCIe without first successfully executing
OSC.  Period.  If you do, you are violating the spec.

> 
> These notebooks fully support hotplug in the external ExpressCard slot
> (which is the *whole point* of an external slot), and according to Dell
> they work just fine with that other OS.  I don't use the "other OS" here,
> but the hardware also works just fine with Linux now.

that's because the other OS probably uses ACPI based hotplug.  As anyone
will tell you the arguement "well, it works fine on X" is bogus, because 
often vendors will put workarounds in the drivers or in the case of ACPI
the firmware.


> 
> I'm guessing they just left out the BIOS functionality because
> it was one of the very first machines to market with such slots,
> and the BIOS wasn't mature enough.   So they rely upon more easily
> updated software drivers instead.  The only gotcha is they do specify
> that booting from the slot is not possible (because no BIOS support).
> 
> Cheers
> 

the point is that we don't know why they left it out.  I know specifically
of hardware where they left it out on PURPOSE because they don't want people
to use it.  Maybe your hardware is behaving fine, maybe it will break on
you eventually after you use it, maybe not, but I've already got specific
examples of broken hardware, so we will not do this.