Re: [Xen-devel] [PATCH for-next 0/6] vpci: initial PCI config space traps

[Roger Pau Monne <roger.pau@xxxxxxxxxx>]

On Tue, Apr 11, 2017 at 11:03:56AM +0100, Roger Pau Monne wrote:
> Hello,
> 
> The following series contain an implementation of handlers for the PCI
> configuration space inside of Xen. This allows Xen to detect accesses to the
> PCI configuration space and react accordingly.
> 
> I'm still not sure that the interface for the handlers is the final one, right
> now only some basic handlers are implemented (header and capability list),
> other maybe more complex handlers (MSI/MSI-X) might need to expand the
> interface.
> 
> Patches #1 and #2 are preparatory, and are not specific to vpci (mostly code
> movement or split). The meat is certainly in patch #3 that implements the
> generic handlers for accesses to the PCI configuration space. Currently a
> per-device red-back tree is used in order to store the list of handlers, and
> they are indexed based on their offset inside of the configuration space. 
> Patch
> #3 also adds the x86 port IO traps and wires them into the newly introduced
> vPCI dispatchers. Patch #4 adds handlers for the ECAM areas (as found on the
> MMCFG ACPI table). Patch #5 adds handlers to the PCI header in order to detect
> BAR related operations (size, position change) and reacts accordingly to them.
> Finally patch #6 allows Xen to mask certain features listed in the
> capabilities list of each PCI device, allowing Xen to mask the MSI/MSI-X
> features, so that the panic in the PVH Dom0 builder can be replaced with a
> printk instead. Note that PVH Dom0 is still missing a lot of functionality, 
> but
> this should hopefully be enough to boot on simple hardware.
> 
> This series is based on top of my previous "x86/dpci: bind legacy PCI
> interrupts to PVHv2 Dom0", which has been rebased on top of the clang fix. The
> full dependency can be seen at:
> 
> http://xenbits.xen.org/gitweb/?p=people/royger/xen.git;a=shortlog;h=refs/heads/vpci_v1
> git://xenbits.xen.org/people/royger/xen.git vpci_v1
> 
> Note that this is only safe to use for the hardware domain (that's trusted),
> any non-trusted domain will need a lot more of traps before it can freely
> access the PCI configuration space.

Hello,

Do not review this version, I have a new one coming up very soon with MSI/MSI-X
and some other fixes for issues that I've found while implementing MSI.

Thanks, Roger.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel