KVM: SEV: Do cache maintenance on the source VM during intra-host migrationgrafted

Manually perform cache maintenance on the source VM during intra-host
migration to ensure no stale data is left in CPU caches after the VM is
destroyed.  Because the source VM is "converted" to a non-SEV VM, KVM's
memory reclaim flows won't trigger cache maintenance, e.g. when all guest
memory is reclaimed in response to detaching from the mmu_notifier.

Note, relying on the destination VM to do cache maintenance isn't an option
as KVM doesn't require identical guest memory configurations, i.e. the
source VM may have access to memory that the destination VM does not.
Enforcing equivalent memory configurations is infeasible, as it would
require a *deep* comparison of memslots, e.g. to verify that not only are
the memslot identical, but what the memslots point at is also identical.

Fixes: b56639318bb2 ("KVM: SEV: Add support for SEV intra host migration")
Cc: stable@vger.kernel.org
Reported-by: Stefan Teodorescu <fane@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20260923163721.1584779-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

1 files changed, 56 insertions, 0 deletions

diff --git a/arch/alpha/lib/ev67-strcat.S b/arch/alpha/lib/ev67-strcat.S
new file mode 100644
index 000000000..f8c7305b1
--- /dev/null
+++ b/arch/alpha/lib/ev67-strcat.S
@@ -0,0 +1,56 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * arch/alpha/lib/ev67-strcat.S
+ * 21264 version contributed by Rick Gorton <rick.gorton@alpha-processor.com>
+ *
+ * Append a null-terminated string from SRC to DST.
+ *
+ * Much of the information about 21264 scheduling/coding comes from:
+ *	Compiler Writer's Guide for the Alpha 21264
+ *	abbreviated as 'CWG' in other comments here
+ *	ftp.digital.com/pub/Digital/info/semiconductor/literature/dsc-library.html
+ * Scheduling notation:
+ *	E	- either cluster
+ *	U	- upper subcluster; U0 - subcluster U0; U1 - subcluster U1
+ *	L	- lower subcluster; L0 - subcluster L0; L1 - subcluster L1
+ * Try not to change the actual algorithm if possible for consistency.
+ * Commentary: It seems bogus to walk the input string twice - once
+ * to determine the length, and then again while doing the copy.
+ * A significant (future) enhancement would be to only read the input
+ * string once.
+ */
+
+#include <linux/export.h>
+	.text
+
+	.align 4
+	.globl strcat
+	.ent strcat
+strcat:
+	.frame $30, 0, $26
+	.prologue 0
+
+	mov	$16, $0		# E : set up return value
+	/* Find the end of the string.  */
+	ldq_u   $1, 0($16)	# L : load first quadword (a0 may be misaligned)
+	lda     $2, -1		# E :
+	insqh   $2, $16, $2	# U :
+
+	andnot  $16, 7, $16	# E :
+	or      $2, $1, $1	# E :
+	cmpbge  $31, $1, $2	# E : bits set iff byte == 0
+	bne     $2, $found	# U :
+
+$loop:	ldq     $1, 8($16)	# L :
+	addq    $16, 8, $16	# E :
+	cmpbge  $31, $1, $2	# E :
+	beq     $2, $loop	# U :
+
+$found:	cttz	$2, $3		# U0 :
+	addq	$16, $3, $16	# E :
+	/* Now do the append.  */
+	mov	$26, $23	# E :
+	br	__stxcpy	# L0 :
+
+	.end strcat
+	EXPORT_SYMBOL(strcat)