diff mbox series

[1/5] x86/HVM: reduce recursion in linear_{read,write}()

Message ID f1147a7a-e232-48e8-b9ea-664c7f5192a4@suse.com (mailing list archive)
State Superseded
Headers show
Series x86/HVM: emulation (MMIO) improvements | expand

Commit Message

Jan Beulich Sept. 4, 2024, 1:27 p.m. UTC 
Let's make explicit what the compiler may or may not do on our behalf:
The 2nd of the recursive invocations each can fall through rather than
re-invoking the function. This will save us from adding yet another
parameter (or more) to the function, just for the recursive invocations.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

Comments

Andrew Cooper Sept. 6, 2024, 5:51 p.m. UTC | #1 
On 04/09/2024 2:27 pm, Jan Beulich wrote:
> Let's make explicit what the compiler may or may not do on our behalf:
> The 2nd of the recursive invocations each can fall through rather than
> re-invoking the function. This will save us from adding yet another
> parameter (or more) to the function, just for the recursive invocations.
>
> Signed-off-by: Jan Beulich <jbeulich@suse.com>

Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
diff mbox series

Patch

--- a/xen/arch/x86/hvm/emulate.c
+++ b/xen/arch/x86/hvm/emulate.c
@@ -1146,7 +1146,7 @@  static int linear_read(unsigned long add
     pagefault_info_t pfinfo;
     struct hvm_vcpu_io *hvio = &current->arch.hvm.hvm_io;
     unsigned int offset = addr & ~PAGE_MASK;
-    int rc = HVMTRANS_bad_gfn_to_mfn;
+    int rc;
 
     if ( offset + bytes > PAGE_SIZE )
     {
@@ -1154,12 +1154,16 @@  static int linear_read(unsigned long add
 
         /* Split the access at the page boundary. */
         rc = linear_read(addr, part1, p_data, pfec, hvmemul_ctxt);
-        if ( rc == X86EMUL_OKAY )
-            rc = linear_read(addr + part1, bytes - part1, p_data + part1,
-                             pfec, hvmemul_ctxt);
-        return rc;
+        if ( rc != X86EMUL_OKAY )
+            return rc;
+
+        addr += part1;
+        bytes -= part1;
+        p_data += part1;
     }
 
+    rc = HVMTRANS_bad_gfn_to_mfn;
+
     /*
      * If there is an MMIO cache entry for the access then we must be re-issuing
      * an access that was previously handled as MMIO. Thus it is imperative that
@@ -1201,7 +1205,7 @@  static int linear_write(unsigned long ad
     pagefault_info_t pfinfo;
     struct hvm_vcpu_io *hvio = &current->arch.hvm.hvm_io;
     unsigned int offset = addr & ~PAGE_MASK;
-    int rc = HVMTRANS_bad_gfn_to_mfn;
+    int rc;
 
     if ( offset + bytes > PAGE_SIZE )
     {
@@ -1209,12 +1213,16 @@  static int linear_write(unsigned long ad
 
         /* Split the access at the page boundary. */
         rc = linear_write(addr, part1, p_data, pfec, hvmemul_ctxt);
-        if ( rc == X86EMUL_OKAY )
-            rc = linear_write(addr + part1, bytes - part1, p_data + part1,
-                              pfec, hvmemul_ctxt);
-        return rc;
+        if ( rc != X86EMUL_OKAY )
+            return rc;
+
+        addr += part1;
+        bytes -= part1;
+        p_data += part1;
     }
 
+    rc = HVMTRANS_bad_gfn_to_mfn;
+
     /*
      * If there is an MMIO cache entry for the access then we must be re-issuing
      * an access that was previously handled as MMIO. Thus it is imperative that