[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-devel] [PATCH for-4.12 v2 3/7] x86/pvh: reorder PVH dom0 iommu initialization



So that the iommu is initialized before populating the p2m, and
entries added get the corresponding iommu page table entries if
required. This requires splitting the current pvh_setup_p2m into two
different functions. One that crafts dom0 physmap and sets the paging
allocation, and another one that actually populates the p2m with RAM
regions.

Note that this allows to remove the special casing done for the low
1MB in hwdom_iommu_map.

Signed-off-by: Roger Pau Monné <roger.pau@xxxxxxxxxx>
---
Cc: Jan Beulich <jbeulich@xxxxxxxx>
Cc: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
Cc: Wei Liu <wei.liu2@xxxxxxxxxx>
---
Changes since v1:
 - New in this version.
---
The previous flow, where the iommu was enabled after having populated
the p2m was used to workaround an issue on an old pre-Haswell Intel
box. I no longer have that box, and when I tested PVH dom0 on it was
before the iommu map-reserved fixes.

If hardware needing the iommu page tables to contain RAM regions is
found I'm happy to work on other solutions, like performing the setup
of the iommu at the start of dom0_construct_pvh but only enabling it
when dom0 p2m is fully populated.
---
 xen/arch/x86/hvm/dom0_build.c       | 35 +++++++++++++++++++----------
 xen/drivers/passthrough/x86/iommu.c |  7 +-----
 2 files changed, 24 insertions(+), 18 deletions(-)

diff --git a/xen/arch/x86/hvm/dom0_build.c b/xen/arch/x86/hvm/dom0_build.c
index a571d15c13..aa599f09ef 100644
--- a/xen/arch/x86/hvm/dom0_build.c
+++ b/xen/arch/x86/hvm/dom0_build.c
@@ -409,14 +409,10 @@ static __init void pvh_setup_e820(struct domain *d, 
unsigned long nr_pages)
     ASSERT(cur_pages == nr_pages);
 }
 
-static int __init pvh_setup_p2m(struct domain *d)
+static void __init pvh_init_p2m(struct domain *d)
 {
-    struct vcpu *v = d->vcpu[0];
     unsigned long nr_pages = dom0_compute_nr_pages(d, NULL, 0);
-    unsigned int i;
-    int rc;
     bool preempted;
-#define MB1_PAGES PFN_DOWN(MB(1))
 
     pvh_setup_e820(d, nr_pages);
     do {
@@ -425,6 +421,14 @@ static int __init pvh_setup_p2m(struct domain *d)
                               &preempted);
         process_pending_softirqs();
     } while ( preempted );
+}
+
+static int __init pvh_populate_p2m(struct domain *d)
+{
+    struct vcpu *v = d->vcpu[0];
+    unsigned int i;
+    int rc;
+#define MB1_PAGES PFN_DOWN(MB(1))
 
     /*
      * Memory below 1MB is identity mapped initially. RAM regions are
@@ -1134,13 +1138,6 @@ int __init dom0_construct_pvh(struct domain *d, const 
module_t *image,
 
     printk(XENLOG_INFO "*** Building a PVH Dom%d ***\n", d->domain_id);
 
-    rc = pvh_setup_p2m(d);
-    if ( rc )
-    {
-        printk("Failed to setup Dom0 physical memory map\n");
-        return rc;
-    }
-
     /*
      * NB: MMCFG initialization needs to be performed before iommu
      * initialization so the iommu code can fetch the MMCFG regions used by the
@@ -1148,8 +1145,22 @@ int __init dom0_construct_pvh(struct domain *d, const 
module_t *image,
      */
     pvh_setup_mmcfg(d);
 
+    /*
+     * Craft dom0 physical memory map and set the paging allocation. This must
+     * be done before the iommu initializion, since iommu initialization code
+     * will likely add mappings required by devices to the p2m (ie: RMRRs).
+     */
+    pvh_init_p2m(d);
+
     iommu_hwdom_init(d);
 
+    rc = pvh_populate_p2m(d);
+    if ( rc )
+    {
+        printk("Failed to setup Dom0 physical memory map\n");
+        return rc;
+    }
+
     rc = pvh_load_kernel(d, image, image_headroom, initrd, 
bootstrap_map(image),
                          cmdline, &entry, &start_info);
     if ( rc )
diff --git a/xen/drivers/passthrough/x86/iommu.c 
b/xen/drivers/passthrough/x86/iommu.c
index a88ef9b189..42b1a1bbc3 100644
--- a/xen/drivers/passthrough/x86/iommu.c
+++ b/xen/drivers/passthrough/x86/iommu.c
@@ -151,12 +151,7 @@ static bool __hwdom_init hwdom_iommu_map(const struct 
domain *d,
      * inclusive mapping additionally maps in every pfn up to 4GB except those
      * that fall in unusable ranges for PV Dom0.
      */
-    if ( (pfn > max_pfn && !mfn_valid(mfn)) || xen_in_range(pfn) ||
-         /*
-          * Ignore any address below 1MB, that's already identity mapped by the
-          * Dom0 builder for HVM.
-          */
-         (!d->domain_id && is_hvm_domain(d) && pfn < PFN_DOWN(MB(1))) )
+    if ( (pfn > max_pfn && !mfn_valid(mfn)) || xen_in_range(pfn) )
         return false;
 
     switch ( type = page_get_ram_type(mfn) )
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.