Update cmd/rsp endian handling in exp_inband

Data written/read from the command/response/data
buffers is done so in transaction sizes of 4 or 8
bytes.  The data in those transactions will be
little endian byte order.  The byte order of those
transactions must be corrected to represent the
data structures stored in the buffers correctly
for extraction of field data.  The fields themselves
will be big endian byte ordering.

This patch also adds two attributes:

ATTR_MSS_OCMB_EXP_STRUCT_MMIO_ENDIAN_CTRL -
Controls whether or not bytes read from and
written two the buffer are swapped.  The default
is to swap the byte order.

ATTR_MSS_OCMB_EXP_STRUCT_ENDIAN -
Controls whether or not the bytes of buffer
structure fields are little endian or not.
The default is big endian.

Change-Id: I734d25dea2a3e4a864a2d35df2576676cad27cfe
Reviewed-on: http://rchgit01.rchland.ibm.com/gerrit1/72314
Tested-by: FSP CI Jenkins <fsp-CI-jenkins+hostboot@us.ibm.com>
Tested-by: Jenkins Server <pfd-jenkins+hostboot@us.ibm.com>
Tested-by: Hostboot CI <hostboot-ci+hostboot@us.ibm.com>
Reviewed-by: Benjamin Gass <bgass@us.ibm.com>
Reviewed-by: Louis Stermole <stermole@us.ibm.com>
Reviewed-by: Jennifer A. Stofer <stofer@us.ibm.com>
Reviewed-on: http://rchgit01.rchland.ibm.com/gerrit1/80988
Reviewed-by: RAJA DAS <rajadas2@in.ibm.com>
Tested-by: RAJA DAS <rajadas2@in.ibm.com>

src/import/chips/ocmb/explorer/procedures/hwp/memory/lib/shared/exp_consts.H

@@ -92,6 +92,7 @@ enum ffdc_codes
     EXP_I2C_SET_FIELD = 0x0001,
     READ_HOST_FW_RESPONSE_STRUCT = 0x0003,
     READ_SENSOR_CACHE_STRUCT = 0x0004,
+    READ_CRCT_ENDIAN = 0x0005,
 
     SET_BYTE_ENABLES = 0x1041,
     SET_NIBBLE_ENABLES = 0x1042,

src/import/generic/memory/lib/utils/endian_utils.H

@@ -189,6 +189,66 @@ bool readLEArray(const std::vector<uint8_t>& i_input, const uint32_t i_size, uin
     return l_passing;
 }
 
+///
+/// @brief Converts BE data into native order
+/// @tparam T the data type to output to
+/// @param[in] i_input inputted data to process
+/// @param[in,out] io_idx current index
+/// @param[out] o_data data that has been converted into native endianness
+/// @return bool true if passing false if failing
+/// @note Real FFDC will be handled outside
+///
+template < typename T >
+bool readBE(const std::vector<uint8_t>& i_input, uint32_t& io_idx, T& o_data)
+{
+    const uint32_t l_sz = static_cast<uint32_t>(sizeof(o_data));
+    uint64_t l_idx = io_idx;
+    io_idx = l_sz + io_idx;
+
+    // Checks that our final index is within the data range
+    // Note: we decrement the index prior, so equal to is ok
+    if(io_idx > i_input.size())
+    {
+        return false;
+    }
+
+    o_data = 0;
+
+    for(uint64_t i = 0; i < l_sz; i++)
+    {
+        uint8_t v = i_input[l_idx];
+        o_data <<= BITS_PER_BYTE;
+        o_data |= v;
+        l_idx++;
+    }
+
+    return true;
+}
+
+///
+/// @brief Converts BE data into native order
+/// @tparam T the data type to output to
+/// @param[in] i_input inputted data to process
+/// @param[in] i_size size of the array
+/// @param[in,out] io_idx current index
+/// @param[out] o_data data that has been converted into native endianness
+/// @return bool true if passing false if failing
+/// @note Real FFDC will be handled outside
+///
+template < typename T >
+bool readBEArray(const std::vector<uint8_t>& i_input, const uint32_t i_size, uint32_t& io_idx, T* o_data)
+{
+    // Loop while the readLE is still passing and we haven't looped through the array's boundaries
+    bool l_passing = true;
+
+    for(uint32_t i = 0; i < i_size && l_passing; ++i)
+    {
+        l_passing = readBE(i_input, io_idx, o_data[i]);
+    }
+
+    return l_passing;
+}
+
 }
 
 #endif