[Xen-devel] [PATCH 01/24] arm: initial Xen support
From: Stefano Stabellini <hidden>
Date: 2012-08-06 10:46:38
Also in:
lkml, xen-devel
On Thu, 2 Aug 2012, Konrad Rzeszutek Wilk wrote:
On Thu, Aug 02, 2012 at 08:35:51AM +0100, Ian Campbell wrote:quoted
On Wed, 2012-08-01 at 19:27 +0100, Rob Herring wrote:quoted
On 07/26/2012 10:33 AM, Stefano Stabellini wrote:quoted
- Basic hypervisor.h and interface.h definitions. - Skelethon enlighten.c, set xen_start_info to an empty struct. - Do not limit xen_initial_domain to PV guests. The new code only compiles when CONFIG_XEN is set, that is going to be added to arch/arm/Kconfig in a later patch. Signed-off-by: Stefano Stabellini <redacted> --- arch/arm/Makefile | 1 + arch/arm/include/asm/hypervisor.h | 6 +++ arch/arm/include/asm/xen/hypervisor.h | 19 ++++++++++ arch/arm/include/asm/xen/interface.h | 64 +++++++++++++++++++++++++++++++++These headers don't seem particularly ARM specific. Could they be moved to asm-generic or include/linux?Or perhaps include/xen. A bunch of it also looks like x86 specific stuff which has crept in. e.g. PARAVIRT_LAZY_FOO and paravirt_get_lazy_mode() are arch/x86 specific and shouldn't be called from common code (and aren't, AFAICT).The could be moved out..
they are called from grant-table.c; sigh, I was the one to add them there :-( interface.h is ARM specific, except for the pvclock structs, that in fact are marked "XXX". hypervisor.h is almost empty but I guess I could move out the following two lines: extern struct shared_info *HYPERVISOR_shared_info; extern struct start_info *xen_start_info; Considering that each arch is free to map them (or not) the way it wants, I don't think is a good idea.