[PATCH v5 02/10] vpci/header: Emulate legacy capability list for dom0

[Jiqian Chen <Jiqian.Chen@xxxxxxx>]

Current logic of emulating legacy capability list is only for domU.
So, expand it to emulate for dom0 too. Then it will be easy to hide
a capability whose initialization fails in a function.

And restrict adding PCI_STATUS register only for domU since dom0
has no limitation to access that register.

Signed-off-by: Jiqian Chen <Jiqian.Chen@xxxxxxx>
---
cc: "Roger Pau Monné" <roger.pau@xxxxxxxxxx>
---
v4->v5 changes:
* Return early if dom0, so that I didn't need to change the exiting return 
chunk.

v3->v4 changes:
* Also pass supported_caps to pci_find_next_cap_ttl() for dom0 since the n is 
zero when dom0,
  and add a comment to explain it.
* Restrict adding PCI_STATUS register only for domU since dom0 has no 
limitation to access that register.
* For dom0 register handler, set vpci_hw_write8 to it instead of NULL.

v2->v3 changes:
* Not to add handler of PCI_CAP_LIST_ID when domain is dom0.

v1->v2 changes:
new patch.

Best regards,
Jiqian Chen.
---
 xen/drivers/vpci/header.c | 39 ++++++++++++++++++++++++++-------------
 xen/drivers/vpci/vpci.c   |  6 ++++++
 xen/include/xen/vpci.h    |  2 ++
 3 files changed, 34 insertions(+), 13 deletions(-)

diff --git a/xen/drivers/vpci/header.c b/xen/drivers/vpci/header.c
index 0fb3cfa6a376..d26cbba08ee1 100644
--- a/xen/drivers/vpci/header.c
+++ b/xen/drivers/vpci/header.c
@@ -758,9 +758,9 @@ static int vpci_init_capability_list(struct pci_dev *pdev)
 {
     int rc;
     bool mask_cap_list = false;
+    bool is_hwdom = is_hardware_domain(pdev->domain);
 
-    if ( !is_hardware_domain(pdev->domain) &&
-         pci_conf_read16(pdev->sbdf, PCI_STATUS) & PCI_STATUS_CAP_LIST )
+    if ( pci_conf_read16(pdev->sbdf, PCI_STATUS) & PCI_STATUS_CAP_LIST )
     {
         /* Only expose capabilities to the guest that vPCI can handle. */
         unsigned int next, ttl = 48;
@@ -768,12 +768,18 @@ static int vpci_init_capability_list(struct pci_dev *pdev)
             PCI_CAP_ID_MSI,
             PCI_CAP_ID_MSIX,
         };
+        /*
+         * For dom0, we should expose all capabilities instead of a fixed
+         * capabilities array, so setting n to 0 here is to get the next
+         * capability position directly in pci_find_next_cap_ttl.
+         */
+        const unsigned int n = is_hwdom ? 0 : ARRAY_SIZE(supported_caps);
 
         next = pci_find_next_cap_ttl(pdev->sbdf, PCI_CAPABILITY_LIST,
-                                     supported_caps,
-                                     ARRAY_SIZE(supported_caps), &ttl);
+                                     supported_caps, n, &ttl);
 
-        rc = vpci_add_register(pdev->vpci, vpci_read_val, NULL,
+        rc = vpci_add_register(pdev->vpci, vpci_read_val,
+                               is_hwdom ? vpci_hw_write8 : NULL,
                                PCI_CAPABILITY_LIST, 1,
                                (void *)(uintptr_t)next);
         if ( rc )
@@ -781,7 +787,7 @@ static int vpci_init_capability_list(struct pci_dev *pdev)
 
         next &= ~3;
 
-        if ( !next )
+        if ( !next && !is_hwdom )
             /*
              * If we don't have any supported capabilities to expose to the
              * guest, mask the PCI_STATUS_CAP_LIST bit in the status
@@ -795,15 +801,18 @@ static int vpci_init_capability_list(struct pci_dev *pdev)
 
             next = pci_find_next_cap_ttl(pdev->sbdf,
                                          pos + PCI_CAP_LIST_NEXT,
-                                         supported_caps,
-                                         ARRAY_SIZE(supported_caps), &ttl);
+                                         supported_caps, n, &ttl);
 
-            rc = vpci_add_register(pdev->vpci, vpci_hw_read8, NULL,
-                                   pos + PCI_CAP_LIST_ID, 1, NULL);
-            if ( rc )
-                return rc;
+            if ( !is_hwdom )
+            {
+                rc = vpci_add_register(pdev->vpci, vpci_hw_read8, NULL,
+                                       pos + PCI_CAP_LIST_ID, 1, NULL);
+                if ( rc )
+                    return rc;
+            }
 
-            rc = vpci_add_register(pdev->vpci, vpci_read_val, NULL,
+            rc = vpci_add_register(pdev->vpci, vpci_read_val,
+                                   is_hwdom ? vpci_hw_write8 : NULL,
                                    pos + PCI_CAP_LIST_NEXT, 1,
                                    (void *)(uintptr_t)next);
             if ( rc )
@@ -813,6 +822,10 @@ static int vpci_init_capability_list(struct pci_dev *pdev)
         }
     }
 
+    /* Return early for the hw domain, no masking of PCI_STATUS. */
+    if ( is_hwdom )
+        return 0;
+
     /* Utilize rsvdp_mask to hide PCI_STATUS_CAP_LIST from the guest. */
     return vpci_add_register_mask(pdev->vpci, vpci_hw_read16, vpci_hw_write16,
                                   PCI_STATUS, 2, NULL,
diff --git a/xen/drivers/vpci/vpci.c b/xen/drivers/vpci/vpci.c
index d2f0f97e0a04..09988f04c27c 100644
--- a/xen/drivers/vpci/vpci.c
+++ b/xen/drivers/vpci/vpci.c
@@ -255,6 +255,12 @@ uint32_t cf_check vpci_hw_read32(
     return pci_conf_read32(pdev->sbdf, reg);
 }
 
+void cf_check vpci_hw_write8(
+    const struct pci_dev *pdev, unsigned int reg, uint32_t val, void *data)
+{
+    pci_conf_write8(pdev->sbdf, reg, val);
+}
+
 void cf_check vpci_hw_write16(
     const struct pci_dev *pdev, unsigned int reg, uint32_t val, void *data)
 {
diff --git a/xen/include/xen/vpci.h b/xen/include/xen/vpci.h
index 475981cb8155..fc8d5b470b0b 100644
--- a/xen/include/xen/vpci.h
+++ b/xen/include/xen/vpci.h
@@ -76,6 +76,8 @@ uint32_t cf_check vpci_hw_read16(
     const struct pci_dev *pdev, unsigned int reg, void *data);
 uint32_t cf_check vpci_hw_read32(
     const struct pci_dev *pdev, unsigned int reg, void *data);
+void cf_check vpci_hw_write8(
+    const struct pci_dev *pdev, unsigned int reg, uint32_t val, void *data);
 void cf_check vpci_hw_write16(
     const struct pci_dev *pdev, unsigned int reg, uint32_t val, void *data);
 
-- 
2.34.1