[PATCH 5.15 44/82] x86/entry: Have entry_ibpb() invalidate return predictions

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

5.15-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Johannes Wikner <kwikner@ethz.ch>

commit 50e4b3b94090babe8d4bb85c95f0d3e6b07ea86e upstream.

entry_ibpb() should invalidate all indirect predictions, including return
target predictions. Not all IBPB implementations do this, in which case the
fallback is RSB filling.

Prevent SRSO-style hijacks of return predictions following IBPB, as the return
target predictor can be corrupted before the IBPB completes.

  [ bp: Massage. ]

Signed-off-by: Johannes Wikner <kwikner@ethz.ch>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Cc: <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 arch/x86/entry/entry.S |    5 +++++
 1 file changed, 5 insertions(+)

--- a/arch/x86/entry/entry.S
+++ b/arch/x86/entry/entry.S
@@ -9,6 +9,8 @@
 #include <asm/unwind_hints.h>
 #include <asm/segment.h>
 #include <asm/cache.h>
+#include <asm/cpufeatures.h>
+#include <asm/nospec-branch.h>
 
 .pushsection .noinstr.text, "ax"
 
@@ -17,6 +19,9 @@ SYM_FUNC_START(entry_ibpb)
 	movl	$PRED_CMD_IBPB, %eax
 	xorl	%edx, %edx
 	wrmsr
+
+	/* Make sure IBPB clears return stack preductions too. */
+	FILL_RETURN_BUFFER %rax, RSB_CLEAR_LOOPS, X86_BUG_IBPB_NO_RET
 	RET
 SYM_FUNC_END(entry_ibpb)
 /* For KVM */