HI-TECH C Compiler for PIC18 MCUs Version 9.80 Release Notes

Transcription

1 HI-TECH C Compiler for PIC18 MCUs Version 9.80 Release Notes Copyright (C) 2011 Microchip Technology Inc. All Rights Reserved. Printed in Australia. Produced on: September 27, 2011 Australian Design Centre 45 Colebard Street West Acacia Ridge QLD 4110 Australia web: THIS FILE CONTAINS IMPORTANT INFORMATION RELATING TO THIS COMPILER. PLEASE READ IT BEFORE RUNNING THIS SOFTWARE.

2 Chapter 1 Introduction 1.1 Description This release is a major update of the HI-TECH C Compiler for PIC18 TM MCUs. It brings the compiler up-to-date with the latest OCG framework. This will resolve a number of long-standing issues and feature requests since the first OCG compiler was released. A large number of limitations have been lifted as a result of this update. This compiler can run in Lite or PRO mode, but there is no Standard mode. The HI-TECH C Compiler for PIC18 TM MCUs (Standard) still fills this need. 1.2 Further Help The compiler s user s guide covers all aspects of the compiler s operation, as well as other useful information. Check the well populated index for your search term. Common problems are explained in the FAQ list. You can also ask questions of other users of this product in the forums. Please direct any bug reports or feature requests via to support. 1.3 Updates and Feedback Microchip welcomes bug reports, suggestions or comments regarding this compiler version. Please direct any feedback via to support. Updates to this product will be made available on Microchip s web site or on Microchip Direct. 2

3 Introduction Previous Versions At times, advisory message 1395 may be issued by the compiler. This message is part of a new testing process. The compiler will display this message if it encounters a specific code sequence that results in internal compiler templates being used in a unique way. It does not mean there is a bug in the generated code, but that the code sequence encounted could be used to further improve the compiler s performance. If you wish to participate by contributing the code that generated this message, you are welcomed to send the project to support; otherwise, you may ignore this message. 1.4 Previous Versions The previous version of HI-TECH C Compiler for PIC18 MCUs was 9.66 released in April

4 Chapter 2 New Features The following are new features the compiler now supports. The version number in brackets indicates the first compiler version to support the feature. See Section 6.1 for a full list of all supported devices. 2.1 General C18 compatibility (9.80) This release offers beta support of C code compatibility with MPLAB C Compiler for PIC18 MCUs projects. A replacement C18 driver (mcc18) is also supplied so this compatibility also extends to the command-line options used to build C18 projects. Any MPLAB IDE projects that are configured for MPLAB C Compiler for PIC18 MCUs need to be associated with the replacement driver (mcc18) in the HI-TECH compiler s bin directory and rebuilt. The installer for version 9.80 allows you to automatically convert existing C18 projects over to the replacement driver. See the C18 Compatibility section in the User s Guide for more information. Html-format debug information (9.80) The compiler can now output a HTML summary of the build process. This currently contains limited information, but does provide some details not available elsewhere. It also allows easy access to intermediate files, particularly those that are not normally in a human-readable format. See --HTML in the user s manual for more information. New --fill specification (9.80) The --fill option, which fills unused addresses in a HEX file, has been updated to be compatible with MPLAB C3X compilers. It adds the ability to use multiple instances of --fill on different address ranges, as well as adding the ability to do increment-filling over an address range. For more details consult the manual. 4

5 New Features C Code New Chips Supported (9.65) The following devices are now supported: 18F23K22, 18F24K22, 18F25K22, 18F25K80, 18F26J13, 18F26J53, 18F26K22, 18F26K80, 18F27J13, 18F27J53, 18F43K22, 18F44K22, 18F45K22, 18F45K80, 18F46J13, 18F46J53, 18F46K22, 18F46K80, 18F47J13, 18F47J53, 18F65K22, 18F65K80, 18F65K90, 18F66K22, 18F66K80, 18F66K90, 18F67K22, 18F67K90, 18F85K22, 18F85K90, 18F86J72, 18F86K22, 18F86K90, 18F87J72, 18LF23K22, 18LF24K22, 18LF25K22, 18LF25K80, 18LF26J13, 18LF26J53, 18LF26K22, 18LF26K80, 18LF27J13, 18LF27J53, 18LF43K22, 18LF44K22, 18LF45K22, 18LF45K80, 18LF46J13, 18LF46J53, 18LF46K22, 18LF46K80, 18LF47J13, 18LF47J53 and 18LF65K C Code Multi-bank stack allocation (9.80) The compiler still places auto and parameter variables in a compiled stack; however, this stack can not be built up in more than one memory bank. This will allow more objects to be allocated to the stack and reduce stack-related memory errors. Absolute const objects (9.80) Objects that are qualified const can now be made absolute and, hence, placed at a user-defined address. These objects can be initialized or uninitialized. #pragma psect reinstated (9.80) This directive can now be used as it was on older STD compilers. It only affects code in the module in which the pragma is contained and allows you to move blocks of code or data into user-defined psects, which can then be placed at user-defined addresses. 2.3 Libraries and Header files Peripheral library support (9.80) The HI-TECH peripheral library has been updated with the latest functions, corrections and changes, made to the C18 peripheral library. Alternative delay routines (9.80) There are now two in-built delay routines: _delaywdt, which may use a CLRWDT instruction in the delay loop, depending on the delay value; and _delay, which is guaranteed not to use this instruction. There are also delaywdt_us and delaywdt_ms macros that can be used in addition to the non-clrwdt versions. setjmp/longjmp library functions (9.80) These functions have been implemented in the standard library. Used with caution, they allow jumps across functions. New typedef types (9.80) The intmax_t, uintmax_t, intptr_t, uintptr_t, int24 and uint24 types have been added to stdint.h. 5

6 Chapter 3 Changes The following are features that are now handled differently by the compiler. These changes may require modification to your source code if porting code to this compiler version. The version number in brackets indicates the first compiler version to implement the change. 3.1 General USB memory (9.80) Previously, those devices that had the USB peripheral did not by default include the USB buffer memory in general purpose memory usable by your programs. This has now been changed. The assumption is now that the USB module is not being used and that this memory is available for general purpose use. If you are using this module and want to reserve the USB memory, use the --RAM option to limit the memory available. 3.2 Command-Line Driver --callgraph option removed (9.80) This option is no longer needed and the call graph (along with call tables) now appears in the assembly list file, not the map file as was previously the case. --fill specification radix (9.80) The default radix of values used by the --fill option has changed from hexadecimal to decimal. In order to be compatible with the new --fill specification, prefix all values with 0x, as per C hexadecimal constants. 6

7 Changes C Code 3.3 C Code Symbol length (9.80) The maximum C and preprocessor macro identifier default length has changed from 31 to 255. Compiling with the --STRICT option will set this to the value 31. The -N option no longer has any effect. Fake symbols and external functions (9.80) External functions (those defined outside of the C domain) which are called via a constant cast to be a function pointer use a new symbol for their return value. It has been renamed from?fake to a name that is based on the function s address. See External Functions in the manual for more information. Configuration bit (9.80) The compiler now support the use of #pragma config as the main mechanism of specifying these bit settings. The syntax, setting and values are identical to those used by MPLAB C for PIC18 MCUs. Open the file chipinfo.html, which is located in the DOCS directory of your compiler installation. Click on your target device and it will show you the settings and values that are appropriate with this pragma. The CONFIG and PROG_CONFIG macros are still supported for legacy projects. Structure return types (9.80) Structures (and unions) which are larger than 4 bytes in size are now returned in the function s parameter memory and the address of this memory is returned in the FSR0 register. The limitation concerning returning of structures in program memory has now been lifted. 3.4 Assembler/Assembly Code FN-type directives (9.80) All FN-type directives are no longer supported and should not be used. Such directive include: FNBREAK, FNSIZE, FNROOT etc. The FNCALL and FNROOT directives are still issued by the compiler, but it is recommended that these not be used in handwritten assembly code. 3.5 Libraries and Header Files Warnings with legacy SFRs (9.80) Using legacy SFRs or SFR bit names will generate a warning indicating that the variable is deprecated. Persistent variable check functions (9.80) The functions, which previously validated and checked persistent variables, have been removed. These no longer work with the new memory allocation scheme used by this compiler version. 7

8 Changes Libraries and Header Files EEPROM and flash routines (9.80) The functions to read and write the EEPROM and flash memory are now supplied by the Microchip peripheral library. This library is built from the same code used by the MPLAB C Compiler for PIC18 MCUs. The prototypes for these routines are contained in the manual, in the Section EEPROM and Flash Runtime Access. You can also find information on these in the peripheral library documentation. Note that these routines are less automated than the HI-TECH C routines they replace. When flash memory is written, the entire block that contains the new values must be erased and then written as a whole. If you only wish to write some of the locations in the block, then you must read in the block, modify the copy to include the changes required, then write then modifed copy back to the flash. The HI-TECH C routines allowed you to specify any address range and this would be accommodated by the library routine. Peripheral libraries (9.80) The peripheral libraries are now linked in by default. The --RUNTIME option can be used to prevent these library files being searched, if required. If you are using the MPLAB IDE, new projects will default to linking in these libraries; however, existing projects will use the setting you had previously selected in the Linker tab of the Build Options dialog. If you previously used the EEPROM or flash library routines, you must ensure that the peripheral libraries are linked in to continue to use them. Single character SFR names (9.80) SFRs whose name was a single character have been removed from the header files. These were typically STATUS register bit variables, e.g. Z and C. Standard library names (9.80) The names of the standard libraries have changed. See the user s guide for more information. Optimizer defines (9.80) The macro OPTIMIZE_SIZE has been changed to OPTIMIZE_SPACE. In addition, the OPTIMIZE_SPACE or OPTIMIZE_SPEED macros are only defined if the optimizer is enabled. Previously the space macro was defined even if the optimizer was disabled. _fassert prototype change (9.80) The function called by the assert() macro, _fassert, has had a minor change to its prototype. CAN header file (9.80) The can18.h header file has been removed from the distribution. The contents of this file now appear in the device-specific header files for devices that implement CAN. Floating-point unpack routines (9.80) The routines that used to appear in ftunpack.c and flunpack.c are no longer included. Library routines that previously used these embed them inline to save stack levels and optimize the code. 8

9 Chapter 4 Limitations The following are limitations in the compiler s operation. These may be general coding restrictions, or deviations from information contained in the user s manual. 4.1 General Code size In some instances the code size may be larger than that produced by the PICC-18 STD compiler. File names If you add an assembly source file to an MPLAB project, the base name of the assembly source file must be different to the MPLAB project base name. Errors will be generated if this is not the case and the source file may be overwritten with the project s intermediate assembly file. REAL ICE debugging If you are using the REAL ICE TRACE and LOG debugging macros, these must be added by hand into your source code. You cannot use MPLAB IDE to add these macros to your code. The debugging settings should be configured in MPLAB IDE in the usual way. 4.2 Preprocessor Sizeof Pointers The result of the sizeof preprocessor operator when applied to a pointer type will always return 1. Use of the C sizeof operator will return the correct value. 9

10 Limitations C Code 4.3 C Code Can t generate code message This message may appear for complex expressions that involve multiple assignment operators, e.g. *= and floating-point operands. Splitting the expression into multiple statements will correct this issue. Can t generate code message This message may appear when compiling code that casts floatingpoint values to an integer type. This is more likely to occur on complex expressions. Try splitting the expression into multiple statements. Accessing timers Some timer peripherals need to have multi-byte registers written in a specific order, but this is not guaranteed with all C statements. Macros are provided which should be used to write to timer registers, e.g. WRITETIMER0(x). The multi-byte timer registers, e.g. TMR0, are not supported by this compiler; however, the individual registers that make up the timer count, e.g. TMR0L and TMR0H, are supported. Indirect calls from ISRs Indirect calls made in interrupt service routines to functions that are also called from main-line code may fail and are now flagged as an error. This does not affect direct calls made from an ISR. Far Variables Variables qualified as far may be defined, but will not be assigned initial values. The usual caveats with far variables still apply with regard to the size of the code generated to access them. Extended PIC18 instruction set Presently the extended PIC18 instruction set and indexed addressing mode are not supported by either the HI-TECH C PRO for the PIC18 MCU Family or PICC-18 STD compilers. If using a device which is equipped with the extended instruction set, it can be configured for legacy mode (extended instructions disabled) with the following configuration setting: CONFIG(4, XINSTDIS); Other configuration word 4 parameters are inherently ANDed into this directive as required. Recursion Recursion is not supported by the compiler in any operating mode. For cases where a function is called by both interrupt and main-line code, the function output is duplicated to give the impression of re-entrancy from the C source level. See the user s guide Interrupt section for more information. Absolute initialized variables (9.80) Variables which are absolute cannot be initialized. Missing parenthesis not detected A missing end parenthesis is not detected in variable declarations when the variable is type cast. For example, 10

11 Limitations Libraries and Header Files static char ((unsigned)&porta; This problem does not appear to affect any output code, but may cause problems when using third-party software tools. Qualified arrays generate error The compiler generates an error when compiling complicated arrays involving qualified types, e.g.: const char * const * const listnames[] = {menu0, menu1}; The error is not issued for arrays of more basic types or for less complicated types, e.g. const char * * const listnames[] = {menu0, menu1}; will compile without error. Structure initialization Locally defined structures are not initialized correctly. Indirect Function call Use of indirect function calls when the function parameters include pointers, may result in compiler error. Try to avoid using pointers as parameters to the functions called indirectly. 4.4 Libraries and Header Files Configuration pragma and macro usage On some devices, assigning a literal constant to an entire configuration register (for example, CONFIG1L below) using the #pragma config will fail. #pragma config CONFIG1L=0x37F Use the individual settings to specify and build up the config value. You may safely assign constant values to each setting. Use of the PROG_CONFIG macro may also fail, so it is recommended that you only use the CONFIG macro for legacy projects. Legacy configuration macro syntax If you are using the CONFIG legacy macro, the parsing of the config setting symbols is very primitive and may not use brackets. The following example CONFIG(1,(XINSTDIS & DEBUGDIS & STVRDIS & PLLDIV1 & WDTEN & CPUDIV3 will genereate an error due to the extra parenthesis used. This should be specified as follows. 11

12 Limitations Debug Limitations CONFIG(1, XINSTDIS & DEBUGDIS & STVRDIS & PLLDIV1 & WDTEN & CPUDIV3 Configuration bit name omissions Some configuration bit names are missing from the header files and cannot be used, for example EXTCLK0, BBSIZ4K, BBSIZ512, and OSC_EXT_RC_OSC_CLKOUT Flash Write Routine The flash write routine will fail if the memory being written is in the same block as the functions associated with writing flash. Fast Double Libraries The fast form of the floating-point library routines have not been implemented. The suboption in the --double option has been removed to that effect. Floating point accuracy The floating point math routines, despite DBL_MIN being defined as E- 38, cannot perform operations predictably on numbers below the order of 1e-35. The result is either correct OR zero. 4.5 Debug Limitations ICE2000/4000 Users of the Microchip ICE2000/4000 should be aware of a documented limitation of these tools concerning side-effects when reading particular SFRs. Reading the following SFRs can affect a STATUS bit, but due to a limitation of the ICE some general purpose registers will also be affected. Reading from the following SFRs can affect the GPRs as follows: PORTB - x81 (081, 181, 281, 391 etc.) RCREG1 - xae RCREG2 - x6e PORTD - x83 SSPBUF - xc9 It is advised to reserve the corresponding locations using a --RAM option (while debugging) if your program reads from the above registers. More information can be found in MPLAB via the ICE2000/4000 help topics under Limitations: PIC18XXXX General Limitations. COF debug limitation For users of MPLAB IDE, please be aware that due to a limitation in the COF debugging format, pointers are always assumed to be 16-bits in length by the IDE. The compiler will encode pointers to be either 1, 2 or 3 bytes wide. This limitation will mean that the addresses contained in 1- or 3-byte wide pointers may be incorrectly shown in the Watch window. In addition, if a pointer is a member of a structure, 12

13 Limitations Debug Limitations those members located after the pointer in the structure s memory will also be shown incorrectly in the Watch window if the pointer member is not 2 bytes wide. The actual contents of the pointer and program execution are not affected. 13

14 Chapter 5 Bugfixes The following are corrections that have been made to the compiler. These may fix bugs in the generated code or alter the operation of the compiler to that which was intended or specified by the user s manual. The version number in brackets indicates the first compiler version to implement the fix. 5.1 Driver Runtime startup code for far objects (9.80) The runtime startup code generated by the driver when uninitialized far variables are used attempted to clear these variables in RAM, not in the external memory. This mean that the far variables would not have been cleared properly and other objects may be overwritten. 5.2 C Code Calling functions via a constant (9.80) If a literal constant is cast to a function pointer and used to call a function (presumably defined in assembly code or via some non-c means) the compiler will now attempt to make this call. Previously, an error would result. If the function being called takes arguments or returns a value in memory, the use must define the location of the symbols used to store these values. The names of the symbols will be printed as undefined if you build without defining them. Calling undefined functions (9.80) If an array of function pointers is not initialized and one of the pointers is used to call a function, the compiler will now produce an error to indicate the called 14

15 Bugfixes C Code function is unknown. Error on invalid warning pragma (9.80) If a negative value was specified with the #pragma warning disable the parser may have crashed. This has been corrected and an error will be produced in this situation. Wrong access for known pointers (9.80) If a pointer is assigned only one value, this address is known by the code generator (it is a constant), and the pointer is used to access an object, an optimization may directly access the object without using the pointer. For some program memory addresses, the code produced would access the correct address, but the data space memory. Access is now made to the correct memory space. Error on absolute variable initialization (9.80) Initialization of absolute variables is not allowed. Previously, if such code was encountered, it would silently be ignored and the variable s indicated address may not be honoured. If such code is encountered now, an error will be issued. Initialized function pointer in a structure (9.80) If a structure, which includes a function pointer member, was initialized at its declaration, then accessing this object may have resulted in error. Structures are now correctly initialized regardless of member type. Undefined i1/i2 symbols (9.80) In some instances where a function was called or had its address taken and this function was used or referenced in both interrupt and main-line code, some symbols were not defined by the compiler and a compiler error would result. Crash when sizeof used as array dimension (9.80) If an array was defined that did not have its size specified in the square brackets, and the size of this array was taken using the sizeof operator, and this size was then used as the dimension of another array, the compiler would crash. Such code is now compiled without error. Conditional operator (9.80) in some instances, the ordering of code generated for the conditional operator? : may have been out of sequence. This would not have affected the function aspect of generated code, but the order of side effects may have changed. This has been corrected. Can t generate code with function parameters (9.80) Many instances of code, where a pointer was passed to a function, may have triggered a can t generate code error message. These errors related to an incorrect calculation of the pointer size for the parameter. This pointer size is now correctly calculated. Access of large arrays (9.80) When large arrays were accessed, an int index variable may have been incorrectly reduced to a signed char rather than an unsigned char, where such an optimization was possible. The wrong array element would have been accessed. These optimizations are now correctly performed. 15

16 Bugfixes C Code Access of const parameters (9.80) Parameters that were qualified as const were in some instances being read from program memory rather than data memory. Wrong bank selection in comparing unsigned long with literal (9.66) Bank was not being selected for temporary variables used in computation. Right shift of long by zero (9.66) Right shifting of a long variable by zero was incorrectly handled. Incorrect result for xor_between long and char (9.66) XOR between a long with char was resulting into incorrect results. Right shift of integers (9.66) Logical shift was being performed instead of arithmetic shift. Wrong value assigned to union variable (9.66) Assigning value to union variable was incorrectly done. Banksel missing when assigning bitfield from temporary variable (9.66) Banksel was added in assignment to bitfield from a temporary value. Pointer comparison to NULL (9.65) If a 2-byte pointer was compared to NULL, the comparison may result a true value if the LSB of the pointer s content was 0. A full 2-byte comparison is now performed and this comparison will return false where expected. Incorrect bit-field initialization (9.65) If a structure with bit-fields was initialized, but the number of initial values was less than the number of bit-field members contained in the structure, the compiler was not padding the initial values with extra 0 values. Undefined symbol null error (9.65) In some instances pointer variables may not have acquired the correct size which would have lead to this error. The size of pointers is now always performed correctly. Looping around MACH_ALLOC error (9.65) This error may have been produced in some situations where memory allocation of variables failed. Better sorting has been implemented. Wrong bank selection after pointer access (9.65) One of two banks might have been selected after executing code that performed a dereference of a pointer that has targets to both data and program memory. After executing this code, the compiler would have assumed only one bank selection was possible and this may have caused subsequent code to fail. Wrong bank selection after bit-field assignment (9.65) One of two banks might have been selected after executing code that performed an assignment of a one-bit wide bit-field to another, but the compiler would have assumed only one selection was possible. This may have caused subsequent code to fail. 16

17 Bugfixes Assembler/Assembly Code 5.3 Assembler/Assembly Code Crash with long symbols (9.80) Symbols larger than 100 characters may have crashed the assembler. The assembler has been adjust to allow symbols up to 256 characters in size. 5.4 Chip Configuration and Header Files Updated files (9.80) The device configuration and header files are now built entirely from a central database. Any changes, such as corrections to memory ranges, names or addresses of SFRs or bits within these SFRs, made to this database will show in these files. 5.5 Debugging Watching anonymous structure (9.80) Watching anonymous structure members may not have revealed the correct contents. The debugging information has been extended to allow such members to be viewed correctly. 17

18 Chapter 6 Addendum 6.1 List of Supported Devices The following is the full list of all devices supported by this version of HI-TECH C Compiler for PIC18 MCUs (PRO) sorted alpha-numerically. Check the chipinfo.html file in the DOCS directory, or the Driver tab in the MPLAB IDE Build Options dialog for the supported devices for other compiler versions. 18C242, 18C252, 18C442, 18C452, 18C601, 18C658, 18C801, 18C858, 18F1220, 18F1230, 18F1320, 18F1330, 18F13K20, 18F13K22, 18F13K50, 18F14K20, 18F14K22, 18F14K50, 18F2220, 18F2221, 18F2320, 18F2321, 18F2331, 18F23K20, 18F23K22, 18F2410, 18F242, 18F2420, 18F2423, 18F2431, 18F2439, 18F2450, 18F2455, 18F2458, 18F248, 18F2480, 18F24J10, 18F24J11, 18F24J50, 18F24K20, 18F24K22, 18F2510, 18F2515, 18F252, 18F2520, 18F2523, 18F2525, 18F2539, 18F2550, 18F2553, 18F258, 18F2580, 18F2585, 18F25J10, 18F25J11, 18F25J50, 18F25K20, 18F25K22, 18F25K80, 18F2610, 18F2620, 18F2680, 18F2682, 18F2685, 18F26J11, 18F26J13, 18F26J50, 18F26J53, 18F26K20, 18F26K22, 18F26K80, 18F27J13, 18F27J53, 18F4220, 18F4221, 18F4320, 18F4321, 18F4331, 18F43K20, 18F43K22, 18F4410, 18F442, 18F4420, 18F4423, 18F4431, 18F4439, 18F4450, 18F4455, 18F4458, 18F448, 18F4480, 18F44J10, 18F44J11, 18F44J50, 18F44K20, 18F44K22, 18F4510, 18F4515, 18F452, 18F4520, 18F4523, 18F4525, 18F4539, 18F4550, 18F4553, 18F458, 18F4580, 18F4585, 18F45J10, 18F45J11, 18F45J50, 18F45K20, 18F45K22, 18F45K80, 18F4610, 18F4620, 18F4680, 18F4682, 18F4685, 18F46J11, 18F46J13, 18F46J50, 18F46J53, 18F46K20, 18F46K22, 18F46K80, 18F47J13, 18F47J53, 18F6310, 18F6390, 18F6393, 18F63J11, 18F63J90, 18F6410, 18F6490, 18F6493, 18F64J11, 18F64J90, 18F6520, 18F6525, 18F6527, 18F6585, 18F65J10, 18

19 Addendum List of Supported Devices 18F65J11, 18F65J15, 18F65J50, 18F65J90, 18F65K22, 18F65K80, 18F65K90, 18F6620, 18F6621, 18F6622, 18F6627, 18F6628, 18F6680, 18F66J10, 18F66J11, 18F66J15, 18F66J16, 18F66J50, 18F66J55, 18F66J60, 18F66J65, 18F66J90, 18F66J93, 18F66K22, 18F66K80, 18F66K90, 18F6720, 18F6722, 18F6723, 18F67J10, 18F67J11, 18F67J50, 18F67J60, 18F67J90, 18F67J93, 18F67K22, 18F67K90, 18F8310, 18F8390, 18F8393, 18F83J11, 18F83J90, 18F8410, 18F8490, 18F8493, 18F84J11, 18F84J90, 18F8520, 18F8525, 18F8527, 18F8585, 18F85J10, 18F85J11, 18F85J15, 18F85J50, 18F85J90, 18F85K22, 18F85K90, 18F8620, 18F8621, 18F8622, 18F8627, 18F8628, 18F8680, 18F86J10, 18F86J11, 18F86J15, 18F86J16, 18F86J50, 18F86J55, 18F86J60, 18F86J65, 18F86J72, 18F86J90, 18F86J93, 18F86K22, 18F86K90, 18F8720, 18F8722, 18F8723, 18F87J10, 18F87J11, 18F87J50, 18F87J60, 18F87J72, 18F87J90, 18F87J93, 18F87K22, 18F87K90, 18F96J60, 18F96J65, 18F97J60, 18LF13K22, 18LF13K50, 18LF14K22, 18LF14K50, 18LF23K22, 18LF24J10, 18LF24J11, 18LF24J50, 18LF24K22, 18LF25J10, 18LF25J11, 18LF25J50, 18LF25K22, 18LF25K80, 18LF26J11, 18LF26J13, 18LF26J50, 18LF26J53, 18LF26K22, 18LF26K80, 18LF27J13, 18LF27J53, 18LF43K22, 18LF44J10, 18LF44J11, 18LF44J50, 18LF44K22, 18LF45J10, 18LF45J11, 18LF45J50, 18LF45K22, 18LF45K80, 18LF46J11, 18LF46J13, 18LF46J50, 18LF46J53, 18LF46K22, 18LF46K80, 18LF47J13, 18LF47J53, 18LF65K80, 18LF66K8018C242, 18C252, 18C442, 18C452, 18C601, 18C658, 18C801, 18C858, 18F1220, 18F1230, 18F1320, 18F1330, 18F13K22, 18F13K50, 18F14K22, 18F14K22LIN, 18F14K50, 18F2220, 18F2221, 18F2320, 18F2321, 18F2331, 18F23K20, 18F23K22, 18F2410, 18F242, 18F2420, 18F2423, 18F2431, 18F2439, 18F2450, 18F2455, 18F2458, 18F248, 18F2480, 18F24J10, 18F24J11, 18F24J50, 18F24K20, 18F24K22, 18F2510, 18F2515, 18F252, 18F2520, 18F2523, 18F2525, 18F2539, 18F2550, 18F2553, 18F258, 18F2580, 18F2585, 18F25J10, 18F25J11, 18F25J50, 18F25K20, 18F25K22, 18F25K80, 18F2610, 18F2620, 18F2680, 18F2682, 18F2685, 18F26J11, 18F26J13, 18F26J50, 18F26J53, 18F26K20, 18F26K22, 18F26K80, 18F27J13, 18F27J53, 18F4220, 18F4221, 18F4320, 18F4321, 18F4331, 18F43K20, 18F43K22, 18F4410, 18F442, 18F4420, 18F4423, 18F4431, 18F4439, 18F4450, 18F4455, 18F4458, 18F448, 18F4480, 18F44J10, 18F44J11, 18F44J50, 18F44K20, 18F44K22, 18F4510, 18F4515, 18F452, 18F4520, 18F4523, 18F4525, 18F4539, 18F4550, 18F4553, 18F458, 18F4580, 18F4585, 18F45J10, 18F45J11, 18F45J50, 18F45K20, 18F45K22, 18F45K80, 18F4610, 18F4620, 18F4680, 18F4682, 18F4685, 18F46J11, 18F46J13, 18F46J50, 18F46J53, 18F46K20, 18F46K22, 18F46K80, 18F47J13, 18F47J53, 18F6310, 18F6390, 18F6393, 18F63J11, 18F63J90, 18F6410, 18F6490, 18F6493, 18F64J11, 18F64J90, 18F6520, 18F6525, 18F6527, 18F6585, 18F65J10, 18F65J11, 18F65J15, 18F65J50, 18F65J90, 18F65K22, 18F65K80, 18F65K90, 18F6620, 18F6621, 18F6622, 18F6627, 18F6628, 18F6680, 18F66J10, 18F66J11, 18F66J15, 18F66J16, 18F66J50, 18F66J55, 18F66J60, 18F66J65, 18F66J90, 18F66J93, 18F66K22, 18F66K80, 18F66K90, 18F6720, 18F6722, 18F6723, 18F67J10, 18F67J11, 18F67J50, 18F67J60, 18F67J90, 18F67J93, 18F67K22, 19

20 Addendum Microchip Errata 18F67K90, 18F8310, 18F8390, 18F8393, 18F83J11, 18F83J90, 18F8410, 18F8490, 18F8493, 18F84J11, 18F84J90, 18F8520, 18F8525, 18F8527, 18F8585, 18F85J10, 18F85J11, 18F85J15, 18F85J50, 18F85J90, 18F85K22, 18F85K90, 18F8620, 18F8621, 18F8622, 18F8627, 18F8628, 18F8680, 18F86J10, 18F86J11, 18F86J15, 18F86J16, 18F86J50, 18F86J55, 18F86J60, 18F86J65, 18F86J72, 18F86J90, 18F86J93, 18F86K22, 18F86K90, 18F8720, 18F8722, 18F8723, 18F87J10, 18F87J11, 18F87J50, 18F87J60, 18F87J72, 18F87J90, 18F87J93, 18F87K22, 18F87K90, 18F96J60, 18F96J65, 18F97J60, 18LF13K22, 18LF13K50, 18LF14K22, 18LF14K50, 18LF23K22, 18LF24J10, 18LF24J11, 18LF24J50, 18LF24K22, 18LF25J10, 18LF25J11, 18LF25J50, 18LF25K22, 18LF25K80, 18LF26J11, 18LF26J13, 18LF26J50, 18LF26J53, 18LF26K22, 18LF26K80, 18LF27J13, 18LF27J53, 18LF43K22, 18LF44J10, 18LF44J11, 18LF44J50, 18LF44K22, 18LF45J10, 18LF45J11, 18LF45J50, 18LF45K22, 18LF45K80, 18LF46J11, 18LF46J13, 18LF46J50, 18LF46J53, 18LF46K22, 18LF46K80, 18LF47J13, 18LF47J53, 18LF65K80, 18LF66K Microchip Errata This release of the PICC-18 compiler recognises the published silicon errata issues listed in the table below. Some of these issues have been corrected and no longer apply in recent silicon revisions. Refer to Microchip s device errata documents for details on which issues are still pertinent for your silicon revision. The compiler s chip configuration file records which issues are applicable to each device. Specific errata workarounds can be selectively enabled or disabled via the driver s --ERRATA command line option. PICC-18 Errata Workarounds Name Description Workaround details 4000 Execution of some flow control operations may yeild unexpected results when certain instructions vector code execution A continuous block of program code is not allowed to grow over the 4000h address across the 4000h address boundary. boundary. Additional NOP instructions are inserted at LFSR DAW Using the LFSR instruction to load a value into a specified FSR register may also corrupt a RAM location. The DAW instruction may improperly clear the CARRY bit (STATUS<0>) when executed. Continued... prescribed locations. The compiler will load FSR registers without using the LFSR instruction. The compiler is not affected by this issue. 20

21 Addendum Microchip Errata PICC-18 Errata Workarounds Name Description Workaround details MINUS40 Table read operations above the user Affected library sources program space (>1FFFFFh) may yield config.c, idloc.c and devread.c erroneous results at the extreme low end employ additional of the device s rated temperature range NOP instructions at prescribed (-40 o C). locations. EEDAT EEADR EELVD FLLVD RESET FASTINTS BSR15 When reading EEPROM, the contents of the EEDATA register may become corrupted in the second instruction cycle after setting the RD bit (EECON1<0>). The result returned from an EEPROM read operation can be corrupted if the RD bit is set immediately following the loading of the EEADR register. Writes to EEPROM memory may not succeed if the internal voltage reference is not set. Writes to program memory may not succeed if the internal voltage reference is not set. A GOTO instruction placed at the reset vector may not execute. If a high-priority interrupt occurs during a two-cycle instruction which modifies WREG, BSR or STATUS, the fastinterrupt return mechanism (via shadow registers) will restore the value held by the register before the instruction. Peripheral flags may be erroneously affected if the BSR register holds the value 15, and an instruction is executed that holds the value C9h in its 8 least significant bits. Continued... The EEPROM_READ macro and eeprom_read library function read EEDATA immeadiately. The compiler is not affected by this issue. No workaround applied No workaround applied Additional NOP instruction inserted at reset vector if following instruction is GOTO. Additional code reloads the shadow registers with the correct values of WREG, STATUS and BSR. Compiler avoids generating MOVLB 15 instructions. A warning is issued if this instruction is detected. 21

22 Addendum Microchip Errata PICC-18 Errata Workarounds Name Description Workaround details TBLWTINT If a peripheral interrupt occurs during a TBLWT operation, data can be corrupted. Library routine flash_write() will temporarily disable all applicable interrupt-enable bits before execution of a TBLWT instruction. FW4000 Self write operations initiated from and No workaround applied acting upon a range within the same side of the 4000h boundary may fail based on sequences of instructions executed following the write. RESETRAM Data in a RAM location can become corrupted if an asynchronous reset (eg. WDT, MCLR event) occurs during a write operation to that location. A warning will be issued if the length nvram psect is greater than zero bytes (persistent variables populate this psect). 22