summary |
shortlog |
log |
commit | commitdiff |
tree
raw |
patch |
inline | side by side (from parent 1:
2d9f98b)
Add extra parenthesis around terms in the common MIN() and MAX() macros to prevent issues with non-trivial macro inputs (thanks to David Lyons).
* \return The larger of the two input parameters
*/
#if !defined(MAX) || defined(__DOXYGEN__)
* \return The larger of the two input parameters
*/
#if !defined(MAX) || defined(__DOXYGEN__)
- #define MAX(x, y) ((x > y) ? x : y)
+ #define MAX(x, y) (((x) > (y)) ? (x) : (y))
#endif
/** Convenience macro to determine the smaller of two values.
#endif
/** Convenience macro to determine the smaller of two values.
* \return The smaller of the two input parameters
*/
#if !defined(MIN) || defined(__DOXYGEN__)
* \return The smaller of the two input parameters
*/
#if !defined(MIN) || defined(__DOXYGEN__)
- #define MIN(x, y) ((x < y) ? x : y)
+ #define MIN(x, y) (((x) < (y)) ? (x) : (y))
#endif
#if !defined(STRINGIFY) || defined(__DOXYGEN__)
#endif
#if !defined(STRINGIFY) || defined(__DOXYGEN__)
return ENDPOINT_RWSTREAM_NoError;\r
}\r
\r
return ENDPOINT_RWSTREAM_NoError;\r
}\r
\r
+/* The following abuses the C preprocessor in order to copy-paste common code with slight alterations,\r
+ * so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
+\r
#define TEMPLATE_FUNC_NAME Endpoint_Write_Stream_LE\r
#define TEMPLATE_BUFFER_TYPE const void*\r
#define TEMPLATE_CLEAR_ENDPOINT() Endpoint_ClearIN()\r
#define TEMPLATE_FUNC_NAME Endpoint_Write_Stream_LE\r
#define TEMPLATE_BUFFER_TYPE const void*\r
#define TEMPLATE_CLEAR_ENDPOINT() Endpoint_ClearIN()\r
return PIPE_RWSTREAM_NoError;\r
}\r
\r
return PIPE_RWSTREAM_NoError;\r
}\r
\r
-/* The following abuses the C preprocessor in order to copy-past common code with slight alterations,\r
+/* The following abuses the C preprocessor in order to copy-paste common code with slight alterations,\r
* so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
\r
#define TEMPLATE_FUNC_NAME Pipe_Write_Stream_LE\r
* so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
\r
#define TEMPLATE_FUNC_NAME Pipe_Write_Stream_LE\r
return ENDPOINT_RWSTREAM_NoError;\r
}\r
\r
return ENDPOINT_RWSTREAM_NoError;\r
}\r
\r
+/* The following abuses the C preprocessor in order to copy-paste common code with slight alterations,\r
+ * so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
+\r
#define TEMPLATE_FUNC_NAME Endpoint_Write_Stream_LE\r
#define TEMPLATE_BUFFER_TYPE const void*\r
#define TEMPLATE_CLEAR_ENDPOINT() Endpoint_ClearIN()\r
#define TEMPLATE_FUNC_NAME Endpoint_Write_Stream_LE\r
#define TEMPLATE_BUFFER_TYPE const void*\r
#define TEMPLATE_CLEAR_ENDPOINT() Endpoint_ClearIN()\r
return PIPE_RWSTREAM_NoError;\r
}\r
\r
return PIPE_RWSTREAM_NoError;\r
}\r
\r
-/* The following abuses the C preprocessor in order to copy-past common code with slight alterations,\r
+/* The following abuses the C preprocessor in order to copy-paste common code with slight alterations,\r
* so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
\r
#define TEMPLATE_FUNC_NAME Pipe_Write_Stream_LE\r
* so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
\r
#define TEMPLATE_FUNC_NAME Pipe_Write_Stream_LE\r
return ENDPOINT_RWSTREAM_NoError;\r
}\r
\r
return ENDPOINT_RWSTREAM_NoError;\r
}\r
\r
+/* The following abuses the C preprocessor in order to copy-paste common code with slight alterations,\r
+ * so that the code needs to be written once. It is a crude form of templating to reduce code maintenance. */\r
+\r
#define TEMPLATE_FUNC_NAME Endpoint_Write_Stream_LE\r
#define TEMPLATE_BUFFER_TYPE const void*\r
#define TEMPLATE_CLEAR_ENDPOINT() Endpoint_ClearIN()\r
#define TEMPLATE_FUNC_NAME Endpoint_Write_Stream_LE\r
#define TEMPLATE_BUFFER_TYPE const void*\r
#define TEMPLATE_CLEAR_ENDPOINT() Endpoint_ClearIN()\r
* reconfiguring all Endpoints/Pipes in order each time a new Endpoint/Pipe is created. To minimize the compiled program
* size, the new \c ORDERED_EP_CONFIG compile time option may be defined in the project makefile to restrict the ordering
* in exchange for a smaller compiled binary size.
* reconfiguring all Endpoints/Pipes in order each time a new Endpoint/Pipe is created. To minimize the compiled program
* size, the new \c ORDERED_EP_CONFIG compile time option may be defined in the project makefile to restrict the ordering
* in exchange for a smaller compiled binary size.
- * - The previous F_CLOCK symbol, required in the project makefile, has been renamed to F_USB. This is due to the previous name
+ * - The previous \c F_CLOCK symbol, required in the project makefile, has been renamed to \c F_USB. This is due to the previous name
* being far too generic for use in future architecture ports, where multiple clock domains are used.
*
* <b>Device Mode</b>
* being far too generic for use in future architecture ports, where multiple clock domains are used.
*
* <b>Device Mode</b>
* error code to allow the user application to determine when to send the next chunk of data.
* - The \ref CDC_Device_SendString() function now expects a null terminated string instead of an explicit length. Existing code
* should use the new \ref CDC_Device_SendData() function, or remove the length parameter from the function call.
* error code to allow the user application to determine when to send the next chunk of data.
* - The \ref CDC_Device_SendString() function now expects a null terminated string instead of an explicit length. Existing code
* should use the new \ref CDC_Device_SendData() function, or remove the length parameter from the function call.
- * - The Endpoint_ResetFIFO() function has been renamed to \ref Endpoint_ResetEndpoint(), to make the API function names more
+ * - The \c Endpoint_ResetFIFO() function has been renamed to \ref Endpoint_ResetEndpoint(), to make the API function names more
* consistent. Existing applications using the old function name should simply replace it with a call to the new function name.
* consistent. Existing applications using the old function name should simply replace it with a call to the new function name.
- * - The Endpoint_*_Byte() functions have been renamed Endpoint_*_8() to ensure they are correct across all architectures. Existing
+ * - The \c Endpoint_*_Byte() functions have been renamed Endpoint_*_8() to ensure they are correct across all architectures. Existing
* code using these functions should replace the previous function names with the new function names.
* code using these functions should replace the previous function names with the new function names.
- * - The Endpoint_*_Word() functions have been renamed Endpoint_*_16() to ensure they are correct across all architectures. Existing
+ * - The \c Endpoint_*_Word() functions have been renamed Endpoint_*_16() to ensure they are correct across all architectures. Existing
* code using these functions should replace the previous function names with the new function names.
* code using these functions should replace the previous function names with the new function names.
- * - The Endpoint_*_DWord() functions have been renamed Endpoint_*_32() to ensure they are correct across all architectures. Existing
+ * - The \c Endpoint_*_DWord() functions have been renamed Endpoint_*_32() to ensure they are correct across all architectures. Existing
* code using these functions should replace the previous function names with the new function names.
* - The Device mode RNDIS class driver no longer stores the incoming and outgoing packets in the class driver instance; the user is
* now expected to manually define a storage location for the packet data. Packets must now be sent and received manually via a call
* code using these functions should replace the previous function names with the new function names.
* - The Device mode RNDIS class driver no longer stores the incoming and outgoing packets in the class driver instance; the user is
* now expected to manually define a storage location for the packet data. Packets must now be sent and received manually via a call
* - The \ref PRNT_Host_SendString() and \ref CDC_Host_SendString() functions now expect a null terminated string instead of an explicit
* length. Existing code should use the new \ref PRNT_Host_SendData() and \ref CDC_Host_SendData() functions, or remove the
* length parameter from the function call.
* - The \ref PRNT_Host_SendString() and \ref CDC_Host_SendString() functions now expect a null terminated string instead of an explicit
* length. Existing code should use the new \ref PRNT_Host_SendData() and \ref CDC_Host_SendData() functions, or remove the
* length parameter from the function call.
- * - The Pipe_ClearErrorFlags() function has been removed, as the pipe error flags are now automatically cleared when the
+ * - The \c Pipe_ClearErrorFlags() function has been removed, as the pipe error flags are now automatically cleared when the
* \ref Pipe_ClearError() function is called.
* \ref Pipe_ClearError() function is called.
- * - The Pipe_*_Byte() functions have been renamed Pipe_*_8() to ensure they are correct across all architectures. Existing code using
+ * - The \c Pipe_*_Byte() functions have been renamed Pipe_*_8() to ensure they are correct across all architectures. Existing code using
* these functions should replace the previous function names with the new function names.
* these functions should replace the previous function names with the new function names.
- * - The Pipe_*_Word() functions have been renamed Pipe_*_16() to ensure they are correct across all architectures. Existing code using
+ * - The \c Pipe_*_Word() functions have been renamed Pipe_*_16() to ensure they are correct across all architectures. Existing code using
* these functions should replace the previous function names with the new function names.
* these functions should replace the previous function names with the new function names.
- * - The Pipe_*_DWord() functions have been renamed Pipe_*_32() to ensure they are correct across all architectures. Existing code using
+ * - The \c Pipe_*_DWord() functions have been renamed Pipe_*_32() to ensure they are correct across all architectures. Existing code using
* these functions should replace the previous function names with the new function names.
* these functions should replace the previous function names with the new function names.
+ * - The \c USB_Host_ClearPipeStall() function has been renamed to USB_Host_ClearEndpointStall(), as it operates on a full endpoing address
+ * within the attached device and not a pipe within the host. Existing code using the old function name should update the function calls and
+ * check for correct usage.
*
* \section Sec_Migration101122 Migrating from 100807 to 101122
* <b>USB Core</b>
*
* \section Sec_Migration101122 Migrating from 100807 to 101122
* <b>USB Core</b>
projects / pub / lufa.git / commitdiff