Re: [Xen-devel] [PATCH v4.1] libxc: Defer initialization of start_page for HVM guests

[Ian Campbell <ian.campbell@xxxxxxxxxx>]

On Thu, 2016-01-07 at 17:11 -0500, Boris Ostrovsky wrote:
> With commit 8c45adec18e0 ("libxc: create unmapped initrd in domain
> builder if supported") location of ramdisk may not be available to
> HVMlite guests by the time alloc_magic_pages_hvm() is invoked if the
> guest supports unmapped initrd.
> 
> So let's move ramdisk info initialization (along with a few other
> operations that are not directly related to allocating magic/special
> pages) from alloc_magic_pages_hvm() to bootlate_hvm().

The "few other operations" are just the hvm_info init for standard HVM by
this iteration?

> --- a/tools/libxc/include/xc_dom.h
> +++ b/tools/libxc/include/xc_dom.h
> @@ -71,6 +71,7 @@ struct xc_dom_image {
> Â
> ÂÂÂÂÂ/* arguments and parameters */
> ÂÂÂÂÂchar *cmdline;
> +ÂÂÂÂsize_t cmdline_size;

Perhaps /* Including NULL and padding/alignment */ ?

(That's the sort of thing I don't mind doing on commit if there is
agreement and no other changes required)

> @@ -1095,8 +1044,8 @@ static int vcpu_hvm(struct xc_dom_image *dom)
> ÂÂÂÂÂ/* Set the IP. */
> ÂÂÂÂÂbsp_ctx.cpu.rip = dom->parms.phys_entry;
> Â
> -ÂÂÂÂif ( dom->start_info_pfn )
> -ÂÂÂÂÂÂÂÂbsp_ctx.cpu.rbx = dom->start_info_pfn << PAGE_SHIFT;
> +ÂÂÂÂif ( dom->start_info_seg.pfn )
> +ÂÂÂÂÂÂÂÂbsp_ctx.cpu.rbx = dom->start_info_seg.pfn << PAGE_SHIFT;

Just to check:Âdom->start_info_pfn is unused/nil in the regular-hvm case,
correct (i.e. this code is dm-list specific both before and after)?

> +ÂÂÂÂelse
> +ÂÂÂÂ{
> +ÂÂÂÂÂÂÂÂvoid *hvm_info_page;
> +
> +ÂÂÂÂÂÂÂÂif ( (hvm_info_page = xc_map_foreign_range(
> +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂxch, domid, PAGE_SIZE, PROT_READ | PROT_WRITE,
> +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂHVM_INFO_PFN)) == NULL )
> +ÂÂÂÂÂÂÂÂÂÂÂÂreturn -1;
> +ÂÂÂÂÂÂÂÂbuild_hvm_info(hvm_info_page, dom);
> +ÂÂÂÂÂÂÂÂmunmap(hvm_info_page, PAGE_SIZE);
> +ÂÂÂÂ}

Did you test regular HVM guests after this movement?

Ian.


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