AP_HAL: add test for snprintf and thus ftoa_engine behavior

Ensures subnormals aren't formatted grossly wrong nor cause hangs.

Reveals two additional bugs:

* The requested number of digits is not always produced.

* The precision isn't enough to distinguish every float.

Outputs from the (nominally) original assembly implementation running on
a real AVR (avr-gcc 7.3.0) are included. That implementation exhibits
none of these flaws so it's likely they have been introduced during the
C port.

Author: tpwrules

libraries/AP_HAL/tests/test_vsnprintf.cpp

@@ -62,4 +62,53 @@ TEST(vsnprintf_Test, Basic)
     }
 }
 
+static const char* do_subnormal_format(uint32_t val_hex) {
+    // format float represented as a hex number long enough to see all digits
+
+    // note that something else is wrong here and all the strings should be the
+    // same width of 99 chars (or whatever the format string means)! when that
+    // is fixed, update the test and add an assert here.
+
+    static char buf[256];
+
+    float val = *(float*)&val_hex;
+
+    hal.util->snprintf(buf, ARRAY_SIZE(buf), "%.99f", val);
+
+    return buf;
+}
+
+TEST(vsnprintf_Test, SubnormalFormat)
+{
+    // arbitrary small number (1e-31)
+    // EXPECT_STREQ("0.000000000000000000000000000000099999998000000000000000000000000000000000000000000000000000000000000",
+    EXPECT_STREQ("0.0000000000000000000000000000001000000",
+        do_subnormal_format(0x0C01CEB3));
+
+    // smallest normal (1.1754944e-38)
+    // EXPECT_STREQ("0.000000000000000000000000000000000000011754944000000000000000000000000000000000000000000000000000000",
+    EXPECT_STREQ("0.00000000000000000000000000000000000001175494",
+        do_subnormal_format(0x00800000));
+
+    // largest subnormal (1.1754942e-38)
+    // EXPECT_STREQ("0.000000000000000000000000000000000000011754942000000000000000000000000000000000000000000000000000000",
+    EXPECT_STREQ("0.00000000000000000000000000000000000001175494", // !! same as above
+        do_subnormal_format(0x007FFFFF));
+
+    // moderate subnormal (1.1478e-41)
+    // EXPECT_STREQ("0.000000000000000000000000000000000000000011478036000000000000000000000000000000000000000000000000000",
+    EXPECT_STREQ("0.00000000000000000000000000000000000000001147804",
+        do_subnormal_format(0x00001FFF));
+
+    // smallest subnormal (1e-45)
+    // EXPECT_STREQ("0.000000000000000000000000000000000000000000001401290000000000000000000000000000000000000000000000000",
+    EXPECT_STREQ("0.00000000000000000000000000000000000000000000140130",
+        do_subnormal_format(0x00000001));
+
+    // zero
+    EXPECT_STREQ("0.000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000",
+        do_subnormal_format(0x00000000));
+}
+
+
 AP_GTEST_MAIN()