[kvm-unit-tests PATCH] powerpc: Add emulator test for the lswi instruction

[Thomas Huth <thuth@redhat.com>]

This test checks some special cases of the lswi instruction. Test
works fine on real hardware, but in QEMU, this reveals a bug with
the final "don't overwrite RA" test (RA gets destroyed since the
check in QEMU is still wrong).
The code is based on the lswx test by Laurent Vivier.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 powerpc/emulator.c | 128 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)

diff --git a/powerpc/emulator.c b/powerpc/emulator.c
index 3696d83..f6ffe21 100644
--- a/powerpc/emulator.c
+++ b/powerpc/emulator.c
@@ -71,6 +71,133 @@ static void test_64bit(void)
 	report_prefix_pop();
 }
 
+/**
+ * Test 'Load String Word Immediate' instruction
+ */
+static void test_lswi(void)
+{
+	int i;
+	char addr[128];
+	uint64_t regs[32];
+
+	report_prefix_push("lswi");
+
+	/* fill memory with sequence */
+	for (i = 0; i < 128; i++)
+		addr[i] = 1 + i;
+
+	/* check incomplete register filling */
+	alignment = 0;
+	asm volatile ("li r12,-1;"
+		      "mr r11, r12;"
+		      "lswi r11, %[addr], %[len];"
+		      "std r11, 0*8(%[regs]);"
+		      "std r12, 1*8(%[regs]);"
+		      ::
+		      [len] "i" (3),
+		      [addr] "b" (addr),
+		      [regs] "r" (regs)
+		      :
+		      "r11", "r12", "memory");
+
+#if  __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
+	/*
+	 * lswi is supposed to cause an alignment exception in little endian
+	 * mode, but QEMU does not support it. So in case we do not get an
+	 * exception, this is an expected failure and we run the other tests
+	 */
+	report_xfail("alignment", !alignment, alignment);
+	if (alignment) {
+		report_prefix_pop();
+		return;
+	}
+#endif
+	report("partial", regs[0] == 0x01020300 && regs[1] == (uint64_t)-1);
+
+	/* check NB = 0 ==> 32 bytes. */
+	asm volatile ("li r19,-1;"
+		      "mr r11, r19; mr r12, r19; mr r13, r19;"
+		      "mr r14, r19; mr r15, r19; mr r16, r19;"
+		      "mr r17, r19; mr r18, r19;"
+		      "lswi r11, %[addr], %[len];"
+		      "std r11, 0*8(%[regs]);"
+		      "std r12, 1*8(%[regs]);"
+		      "std r13, 2*8(%[regs]);"
+		      "std r14, 3*8(%[regs]);"
+		      "std r15, 4*8(%[regs]);"
+		      "std r16, 5*8(%[regs]);"
+		      "std r17, 6*8(%[regs]);"
+		      "std r18, 7*8(%[regs]);"
+		      "std r19, 8*8(%[regs]);"
+		      ::
+		      [len] "i" (0),
+		      [addr] "b" (addr),
+		      [regs] "r" (regs)
+		      :
+		      /* as 32 is the number of bytes,
+		       * we should modify 32/4 = 8 regs, from r11 to r18
+		       * We check r19 is unmodified by filling it with 1s
+		       * before the instruction.
+		       */
+		      "r11", "r12", "r13", "r14", "r15", "r16", "r17",
+		      "r18", "r19", "memory");
+
+	report("length", regs[0] == 0x01020304 && regs[1] == 0x05060708 &&
+			 regs[2] == 0x090a0b0c && regs[3] == 0x0d0e0f10 &&
+			 regs[4] == 0x11121314 && regs[5] == 0x15161718 &&
+			 regs[6] == 0x191a1b1c && regs[7] == 0x1d1e1f20 &&
+			 regs[8] == (uint64_t)-1);
+
+	/* check wrap around to r0 */
+	asm volatile ("li r31,-1;"
+		      "mr r0, r31;"
+		      "lswi r31, %[addr], %[len];"
+		      "std r31, 0*8(%[regs]);"
+		      "std r0, 1*8(%[regs]);"
+		      ::
+		      [len] "i" (8),
+		      [addr] "b" (addr),
+		      [regs] "r" (regs)
+		      :
+		      /* modify two registers from r31, wrap around to r0 */
+		      "r31", "r0", "memory");
+
+	report("wrap around to r0", regs[0] == 0x01020304 &&
+			            regs[1] == 0x05060708);
+
+	/* check wrap around doesn't break RA */
+	asm volatile ("mr r28,r1\n"
+		      "mr r29,r2\n"
+		      "li r31,-1\n"
+		      "mr r2,r31\n"
+		      "mr r0,r31\n"
+		      "mr r1, %[addr]\n"
+		      ".long 0x7fe184aa\n"       /* lswi r31, r1, 16 */
+		      "std r31, 0*8(%[regs])\n"
+		      "std r0, 1*8(%[regs])\n"
+		      "std r1, 2*8(%[regs])\n"
+		      "std r2, 3*8(%[regs])\n"
+		      "mr r1,r28\n"
+		      "mr r2,r29\n"
+		      ::
+		      [addr] "r" (addr),
+		      [regs] "r" (regs)
+		      :
+		      /* loading four registers from r31 wraps around to r2,
+		       * r1 is saved to r29, as adding it to the clobber
+		       * list doesn't protect it
+		       */
+		      "r31", "r0", "r28", "r29", "memory");
+
+	/* doc says it is invalid, real proc stops when it comes to
+	 * overwrite the register.
+	 * In all the cases, the register must stay untouched
+	 */
+	report("Don't overwrite RA", regs[2] == (uint64_t)addr);
+
+	report_prefix_pop();
+}
+
 /*
  * lswx: Load String Word Indexed X-form
  *
@@ -234,6 +361,7 @@ int main(int argc, char **argv)
 
 	test_64bit();
 	test_illegal();
+	test_lswi();
 	test_lswx();
 
 	report_prefix_pop();
-- 
1.8.3.1