On 03/28/2011 06:32 PM, Takuya Yoshikawa wrote:
From: Takuya Yoshikawa<yoshikawa.takuya@xxxxxxxxxxxxx>
This stops "CMP r/m, reg" to write back the data into memory.
Pointed out by Avi.
Signed-off-by: Takuya Yoshikawa<yoshikawa.takuya@xxxxxxxxxxxxx>
---
arch/x86/kvm/emulate.c | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
index 14c5ad5..8a73805 100644
--- a/arch/x86/kvm/emulate.c
+++ b/arch/x86/kvm/emulate.c
@@ -3084,6 +3084,7 @@ special_insn:
emulate_2op_SrcV("xor", c->src, c->dst, ctxt->eflags);
break;
case 0x38 ... 0x3d:
+ c->dst.type = OP_NONE; /* Disable writeback. */
cmp: /* cmp */
Why not disable writeback here? As a prelude to having em_cmp() which
does everything?
I see SCAS also does a 'goto cmp', but it also benefits from disabling
writeback.
emulate_2op_SrcV("cmp", c->src, c->dst, ctxt->eflags);
break;
@@ -3138,6 +3139,7 @@ special_insn:
case 6:
goto xor;
case 7:
+ c->dst.type = OP_NONE; /* Disable writeback. */
goto cmp;
}
break;
--
error compiling committee.c: too many arguments to function
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
