[qemu-xen stable-4.7] xen/pt: allow QEMU to request MSI unmasking at bind time

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[qemu-xen stable-4.7] xen/pt: allow QEMU to request MSI unmasking at bind time

patchbot
commit 02659364349730f82977bd20331d4fa374648fa5
Author:     Roger Pau Monne <[hidden email]>
AuthorDate: Thu Aug 24 16:07:03 2017 +0100
Commit:     Anthony PERARD <[hidden email]>
CommitDate: Thu Nov 16 15:40:08 2017 +0000

    xen/pt: allow QEMU to request MSI unmasking at bind time
   
    When a MSI interrupt is bound to a guest using
    xc_domain_update_msi_irq (XEN_DOMCTL_bind_pt_irq) the interrupt is
    left masked by default.
   
    This causes problems with guests that first configure interrupts and
    clean the per-entry MSIX table mask bit and afterwards enable MSIX
    globally. In such scenario the Xen internal msixtbl handlers would not
    detect the unmasking of MSIX entries because vectors are not yet
    registered since MSIX is not enabled, and vectors would be left
    masked.
   
    Introduce a new flag in the gflags field to signal Xen whether a MSI
    interrupt should be unmasked after being bound.
   
    This also requires to track the mask register for MSI interrupts, so
    QEMU can also notify to Xen whether the MSI interrupt should be bound
    masked or unmasked
   
    Signed-off-by: Roger Pau Monné <[hidden email]>
    Reviewed-by: Jan Beulich <[hidden email]>
    Reported-by: Andreas Kinzler <[hidden email]>
    Reviewed-by: Stefano Stabellini <[hidden email]>
    Signed-off-by: Stefano Stabellini <[hidden email]>
    (cherry picked from commit a8036336609d2e184fc3543a4c439c0ba7d7f3a2)
---
 hw/xen/xen_pt.h             |  1 +
 hw/xen/xen_pt_config_init.c | 20 ++++++++++++++++++--
 hw/xen/xen_pt_msi.c         | 13 ++++++++++---
 3 files changed, 29 insertions(+), 5 deletions(-)

diff --git a/hw/xen/xen_pt.h b/hw/xen/xen_pt.h
index 3749711..be6f80f 100644
--- a/hw/xen/xen_pt.h
+++ b/hw/xen/xen_pt.h
@@ -180,6 +180,7 @@ typedef struct XenPTMSI {
     uint32_t addr_hi;  /* guest message upper address */
     uint16_t data;     /* guest message data */
     uint32_t ctrl_offset; /* saved control offset */
+    uint32_t mask;     /* guest mask bits */
     int pirq;          /* guest pirq corresponding */
     bool initialized;  /* when guest MSI is initialized */
     bool mapped;       /* when pirq is mapped */
diff --git a/hw/xen/xen_pt_config_init.c b/hw/xen/xen_pt_config_init.c
index a6100b1..ad7664c 100644
--- a/hw/xen/xen_pt_config_init.c
+++ b/hw/xen/xen_pt_config_init.c
@@ -1313,6 +1313,22 @@ static int xen_pt_msgdata_reg_write(XenPCIPassthroughState *s,
     return 0;
 }
 
+static int xen_pt_mask_reg_write(XenPCIPassthroughState *s, XenPTReg *cfg_entry,
+                                 uint32_t *val, uint32_t dev_value,
+                                 uint32_t valid_mask)
+{
+    int rc;
+
+    rc = xen_pt_long_reg_write(s, cfg_entry, val, dev_value, valid_mask);
+    if (rc) {
+        return rc;
+    }
+
+    s->msi->mask = *val;
+
+    return 0;
+}
+
 /* MSI Capability Structure reg static information table */
 static XenPTRegInfo xen_pt_emu_reg_msi[] = {
     /* Next Pointer reg */
@@ -1391,7 +1407,7 @@ static XenPTRegInfo xen_pt_emu_reg_msi[] = {
         .emu_mask   = 0xFFFFFFFF,
         .init       = xen_pt_mask_reg_init,
         .u.dw.read  = xen_pt_long_reg_read,
-        .u.dw.write = xen_pt_long_reg_write,
+        .u.dw.write = xen_pt_mask_reg_write,
     },
     /* Mask reg (if PCI_MSI_FLAGS_MASKBIT set, for 64-bit devices) */
     {
@@ -1402,7 +1418,7 @@ static XenPTRegInfo xen_pt_emu_reg_msi[] = {
         .emu_mask   = 0xFFFFFFFF,
         .init       = xen_pt_mask_reg_init,
         .u.dw.read  = xen_pt_long_reg_read,
-        .u.dw.write = xen_pt_long_reg_write,
+        .u.dw.write = xen_pt_mask_reg_write,
     },
     /* Pending reg (if PCI_MSI_FLAGS_MASKBIT set, for 32-bit devices) */
     {
diff --git a/hw/xen/xen_pt_msi.c b/hw/xen/xen_pt_msi.c
index 97adc20..d02171a 100644
--- a/hw/xen/xen_pt_msi.c
+++ b/hw/xen/xen_pt_msi.c
@@ -24,6 +24,7 @@
 #define XEN_PT_GFLAGS_SHIFT_DM             9
 #define XEN_PT_GFLAGSSHIFT_DELIV_MODE     12
 #define XEN_PT_GFLAGSSHIFT_TRG_MODE       15
+#define XEN_PT_GFLAGSSHIFT_UNMASKED       16
 
 #define latch(fld) latch[PCI_MSIX_ENTRY_##fld / sizeof(uint32_t)]
 
@@ -155,7 +156,8 @@ static int msi_msix_update(XenPCIPassthroughState *s,
                            int pirq,
                            bool is_msix,
                            int msix_entry,
-                           int *old_pirq)
+                           int *old_pirq,
+                           bool masked)
 {
     PCIDevice *d = &s->dev;
     uint8_t gvec = msi_vector(data);
@@ -171,6 +173,8 @@ static int msi_msix_update(XenPCIPassthroughState *s,
         table_addr = s->msix->mmio_base_addr;
     }
 
+    gflags |= masked ? 0 : (1u << XEN_PT_GFLAGSSHIFT_UNMASKED);
+
     rc = xc_domain_update_msi_irq(xen_xc, xen_domid, gvec,
                                   pirq, gflags, table_addr);
 
@@ -273,8 +277,10 @@ int xen_pt_msi_setup(XenPCIPassthroughState *s)
 int xen_pt_msi_update(XenPCIPassthroughState *s)
 {
     XenPTMSI *msi = s->msi;
+
+    /* Current MSI emulation in QEMU only supports 1 vector */
     return msi_msix_update(s, msi_addr64(msi), msi->data, msi->pirq,
-                           false, 0, &msi->pirq);
+                           false, 0, &msi->pirq, msi->mask & 1);
 }
 
 void xen_pt_msi_disable(XenPCIPassthroughState *s)
@@ -355,7 +361,8 @@ static int xen_pt_msix_update_one(XenPCIPassthroughState *s, int entry_nr,
     }
 
     rc = msi_msix_update(s, entry->addr, entry->data, pirq, true,
-                         entry_nr, &entry->pirq);
+                         entry_nr, &entry->pirq,
+                         vec_ctrl & PCI_MSIX_ENTRY_CTRL_MASKBIT);
 
     if (!rc) {
         entry->updated = false;
--
generated by git-patchbot for /home/xen/git/qemu-xen.git#stable-4.7


_______________________________________________
Xen-changelog mailing list
[hidden email]
https://lists.xenproject.org/xen-changelog