[xen staging] x86/HVM: slightly improve CMPXCHG16B emulation

[patchbot@xxxxxxx]

commit 2488ede7a994c8254b8d9304082d70924bbca141
Author:     Jan Beulich <jbeulich@xxxxxxxx>
AuthorDate: Wed Mar 5 15:41:14 2025 +0100
Commit:     Jan Beulich <jbeulich@xxxxxxxx>
CommitDate: Wed Mar 5 15:41:14 2025 +0100

    x86/HVM: slightly improve CMPXCHG16B emulation
    
    Using hvmemul_linear_mmio_write() directly (as fallback when mapping the
    memory operand isn't possible) won't work properly when the access
    crosses a RAM/MMIO boundary. Use linear_write() instead, which splits at
    such boundaries as necessary.
    
    Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
    Acked-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
---
 xen/arch/x86/hvm/emulate.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c
index 3cd7f2e22f..de38142413 100644
--- a/xen/arch/x86/hvm/emulate.c
+++ b/xen/arch/x86/hvm/emulate.c
@@ -1718,10 +1718,8 @@ static int cf_check hvmemul_cmpxchg(
 {
     struct hvm_emulate_ctxt *hvmemul_ctxt =
         container_of(ctxt, struct hvm_emulate_ctxt, ctxt);
-    struct vcpu *curr = current;
     unsigned long addr;
     uint32_t pfec = PFEC_page_present | PFEC_write_access;
-    struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io;
     int rc;
     void *mapping = NULL;
 
@@ -1745,10 +1743,7 @@ static int cf_check hvmemul_cmpxchg(
     if ( !mapping )
     {
         /* Fix this in case the guest is really relying on r-m-w atomicity. */
-        return hvmemul_linear_mmio_write(addr, bytes, p_new, pfec,
-                                         hvmemul_ctxt, addr,
-                                         hvio->mmio_access.write_access &&
-                                         hvio->mmio_gla == (addr & PAGE_MASK));
+        return linear_write(addr, bytes, p_new, pfec, hvmemul_ctxt);
     }
 
     switch ( bytes )
--
generated by git-patchbot for /home/xen/git/xen.git#staging