* [PATCH] hw/xen_pt: Don't grant opregion permissions
@ 2020-06-14 23:21 Grzegorz Uriasz
2020-06-15 6:10 ` Philippe Mathieu-Daudé
2020-06-15 12:47 ` Roger Pau Monné
0 siblings, 2 replies; 3+ messages in thread
From: Grzegorz Uriasz @ 2020-06-14 23:21 UTC (permalink / raw)
To: qemu-devel
Cc: Stefano Stabellini, jakub, marmarek, Grzegorz Uriasz, j.nowak26,
Anthony Perard, xen-devel, contact
With the upstreaming of linux based stubdomains to xen, qemu can't
assume it runs inside dom0 - permission assignment must be moved to
libxl running in dom0. This xen patch:
https://lists.xenproject.org/archives/html/xen-devel/2020-06/msg00973.html
implements granting the required permissions to the stubdomain running
qemu. This patch removes granting opregion permissions in qemu - this
should be fine as when qemu is running inside dom0 the memory mapping will
be successfully created without first explicitly granting the permission.
Signed-off-by: Grzegorz Uriasz <gorbak25@gmail.com>
---
hw/xen/xen_pt_graphics.c | 13 -------------
1 file changed, 13 deletions(-)
diff --git a/hw/xen/xen_pt_graphics.c b/hw/xen/xen_pt_graphics.c
index 7d46e9c209..303674365b 100644
--- a/hw/xen/xen_pt_graphics.c
+++ b/hw/xen/xen_pt_graphics.c
@@ -283,19 +283,6 @@ void igd_write_opregion(XenPCIPassthroughState *s, uint32_t val)
igd_guest_opregion = (unsigned long)(val & ~XEN_PCI_INTEL_OPREGION_MASK)
| (igd_host_opregion & XEN_PCI_INTEL_OPREGION_MASK);
- ret = xc_domain_iomem_permission(xen_xc, xen_domid,
- (unsigned long)(igd_host_opregion >> XC_PAGE_SHIFT),
- XEN_PCI_INTEL_OPREGION_PAGES,
- XEN_PCI_INTEL_OPREGION_ENABLE_ACCESSED);
-
- if (ret) {
- XEN_PT_ERR(&s->dev, "[%d]:Can't enable to access IGD host opregion:"
- " 0x%lx.\n", ret,
- (unsigned long)(igd_host_opregion >> XC_PAGE_SHIFT)),
- igd_guest_opregion = 0;
- return;
- }
-
ret = xc_domain_memory_mapping(xen_xc, xen_domid,
(unsigned long)(igd_guest_opregion >> XC_PAGE_SHIFT),
(unsigned long)(igd_host_opregion >> XC_PAGE_SHIFT),
--
2.27.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] hw/xen_pt: Don't grant opregion permissions
2020-06-14 23:21 [PATCH] hw/xen_pt: Don't grant opregion permissions Grzegorz Uriasz
@ 2020-06-15 6:10 ` Philippe Mathieu-Daudé
2020-06-15 12:47 ` Roger Pau Monné
1 sibling, 0 replies; 3+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-06-15 6:10 UTC (permalink / raw)
To: Grzegorz Uriasz, qemu-devel
Cc: Stefano Stabellini, jakub, marmarek, Anthony Perard, j.nowak26,
xen-devel, contact
Hi Grzegorz,
On 6/15/20 1:21 AM, Grzegorz Uriasz wrote:
> With the upstreaming of linux based stubdomains to xen, qemu can't
> assume it runs inside dom0 - permission assignment must be moved to
> libxl running in dom0. This xen patch:
> https://lists.xenproject.org/archives/html/xen-devel/2020-06/msg00973.html
> implements granting the required permissions to the stubdomain running
> qemu. This patch removes granting opregion permissions in qemu - this
> should be fine as when qemu is running inside dom0 the memory mapping will
> be successfully created without first explicitly granting the permission.
>
> Signed-off-by: Grzegorz Uriasz <gorbak25@gmail.com>
> ---
> hw/xen/xen_pt_graphics.c | 13 -------------
> 1 file changed, 13 deletions(-)
>
> diff --git a/hw/xen/xen_pt_graphics.c b/hw/xen/xen_pt_graphics.c
> index 7d46e9c209..303674365b 100644
> --- a/hw/xen/xen_pt_graphics.c
> +++ b/hw/xen/xen_pt_graphics.c
> @@ -283,19 +283,6 @@ void igd_write_opregion(XenPCIPassthroughState *s, uint32_t val)
> igd_guest_opregion = (unsigned long)(val & ~XEN_PCI_INTEL_OPREGION_MASK)
> | (igd_host_opregion & XEN_PCI_INTEL_OPREGION_MASK);
>
> - ret = xc_domain_iomem_permission(xen_xc, xen_domid,
> - (unsigned long)(igd_host_opregion >> XC_PAGE_SHIFT),
> - XEN_PCI_INTEL_OPREGION_PAGES,
> - XEN_PCI_INTEL_OPREGION_ENABLE_ACCESSED);
> -
> - if (ret) {
> - XEN_PT_ERR(&s->dev, "[%d]:Can't enable to access IGD host opregion:"
> - " 0x%lx.\n", ret,
> - (unsigned long)(igd_host_opregion >> XC_PAGE_SHIFT)),
> - igd_guest_opregion = 0;
> - return;
> - }
Shouldn't this be somehow versioned? I.e. if the libxl does not have
the change then keep the current code?
> -
> ret = xc_domain_memory_mapping(xen_xc, xen_domid,
> (unsigned long)(igd_guest_opregion >> XC_PAGE_SHIFT),
> (unsigned long)(igd_host_opregion >> XC_PAGE_SHIFT),
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] hw/xen_pt: Don't grant opregion permissions
2020-06-14 23:21 [PATCH] hw/xen_pt: Don't grant opregion permissions Grzegorz Uriasz
2020-06-15 6:10 ` Philippe Mathieu-Daudé
@ 2020-06-15 12:47 ` Roger Pau Monné
1 sibling, 0 replies; 3+ messages in thread
From: Roger Pau Monné @ 2020-06-15 12:47 UTC (permalink / raw)
To: Grzegorz Uriasz
Cc: Stefano Stabellini, jakub, qemu-devel, marmarek, Anthony Perard,
j.nowak26, xen-devel, contact
On Sun, Jun 14, 2020 at 11:21:09PM +0000, Grzegorz Uriasz wrote:
> With the upstreaming of linux based stubdomains to xen, qemu can't
> assume it runs inside dom0 - permission assignment must be moved to
> libxl running in dom0. This xen patch:
> https://lists.xenproject.org/archives/html/xen-devel/2020-06/msg00973.html
> implements granting the required permissions to the stubdomain running
> qemu. This patch removes granting opregion permissions in qemu - this
> should be fine as when qemu is running inside dom0 the memory mapping will
> be successfully created without first explicitly granting the permission.
In order to avoid breaking certain libxl - QEMU combinations, could
you make the check below non-fatal?
So that the current code can be kept for dom0 while not throwing an
error when used inside of a stub domain?
Thanks, Roger.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2020-06-15 12:48 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-06-14 23:21 [PATCH] hw/xen_pt: Don't grant opregion permissions Grzegorz Uriasz
2020-06-15 6:10 ` Philippe Mathieu-Daudé
2020-06-15 12:47 ` Roger Pau Monné
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).