Add "BASICVFP", a VFP version of BASIC64

Detail:
  See Doc/BASICVFP for overall design/specification
  File changes:
  - !Mk,fd7, !MkClean,fd7, !MkRom,fd7, Makefile - update build scripts to support BASICVFP target
  - Resources/UK/Messages - Add a few new messages
  - hdr/Definitions - Define BASICVFP register usage and a few utility macros
  - hdr/WorkSpace - Define BASICVFP workspace adjustments. Add a few asserts to check some workspace constraints are maintained.
  - s/* - Lots of changes to add support for BASICVFP variant of the module
Admin:
  Tested on Raspberry Pi 1/2/3, BB-xM


Version 1.67. Tagged as 'BASIC-1_67'

!Mk,fd7

@@ -15,3 +15,4 @@
 Dir <Obey$Dir>
 amu_machine standalone COMPONENT=BASIC105 THROWBACK=-throwback
 amu_machine standalone COMPONENT=BASIC64 THROWBACK=-throwback
+amu_machine standalone COMPONENT=BASICVFP THROWBACK=-throwback

!MkClean,fd7

@@ -15,4 +15,5 @@
 Dir <Obey$Dir>
 amu_machine clean COMPONENT=BASIC105
 amu_machine clean COMPONENT=BASIC64
+amu_machine clean COMPONENT=BASICVFP
 stripdepnd

!MkRom,fd7

@@ -15,3 +15,4 @@
 Dir <Obey$Dir>
 amu_machine rom COMPONENT=BASIC105 THROWBACK=-throwback
 amu_machine rom COMPONENT=BASIC64 THROWBACK=-throwback
+amu_machine rom COMPONENT=BASICVFP THROWBACK=-throwback

Doc/BASICVFP

+BASICVFP considerations
+=======================
+
+"BASICVFP" aims to produce a release of BASIC which uses the VFP instruction set for floating point math, in place of the obsolete FPA instruction set used by BASIC64.
+
+FPA supports single, double, and extended double precision IEEE 754 floats. VFP only supports single and double precision. Lukcily BASIC64 only uses double precision floats, so on the surface it would appear that a VFP version of BASIC64 would be trivial to produce, without requiring any loss of precision. However there are some other differences between FPA and VFP which make things a bit more complicated.
+
+
+FPA vs. VFP issues
+==================
+
+Word order
+----------
+
+FPA stored double-precision floats using a big-endian word order, i.e. when using STFD to store a double-precision float in memory the most significant 32 bits of the 64 bit value were stored in the lowest word of the pair.
+
+On the other hand, VFP uses a little-endian word order (or more correctly, the endianness of the entire 64-bit value will match the configured data endian mode of the CPU).
+
+This has the potential to cause compatibility issues with any interfaces where the memory representation of BASIC floats are exposed to the wider world.
+
+No support for trig/pow instructions
+------------------------------------
+
+The FPA instruction set has a wide range of trigonometric and power functions available, and BASIC64 made use of them for implementing its trig and power operations.
+
+The only power function VFP implements is square root. Therefore additional effort will be required to implement the full range of trig/power operations in BASICVFP.
+
+Since double-precision floats offer higher precision than BASIC 5-byte floats, we cannot simply re-use the 5-byte float code. However as a stopgap solution we could conceivably perform the operations using the FPA instruction set (which would in turn rely on FPEmulator to implement the operation in software, at 80 bit precision)
+
+Exception handling
+------------------
+
+FPA had full support for hardware trapping of floating point exceptions (division by zero, NaN, etc.). BASIC64 relied on this for generating most of the floating point errors.
+
+On the other hand, hardware trapping of exceptions is optional in VFP. The hardware can still detect when an error has occured and set the relevant bit(s) in the FPSCR, but in most VFP implementations software must manually poll the register to detect any errors. This will likely result in the VFP code sequences being longer than the FPA sequences, and some care may be needed to try and avoid pipeline stalls caused by reading the FPSCR.
+
+VFP context creation
+--------------------
+
+FPEmulator provides the system with a default FPA context which programs can make use of. VFPSupport, on the other hand, requires that each program makes its own VFP context. Therefore some extra logic will be needed on startup/shutdown to create/destroy a VFP context for BASICVFP. Care will also be needed on any entry points (e.g. error environment handler) to make sure that BASIC restores its VFP context before calling any FP code.
+
+
+VFP advantages
+==============
+
+Some VFP implementations can perform operations on vectors of numbers, which could be used to speed up array/matrix operations.
+
+There's also the potential to use the NEON instruction set for integer operations (again, most likely for array/matrix operations)
+
+
+BASIC64 specification review
+============================
+
+The following are key points from the BBC BASIC reference manual (ISBN 1 85250 103 0), with regards to BASIC64
+
+"Exchanging data between BBC BASIC and other languages, like C, is now easier" (p6)
+
+  Although the Norcroft C compiler is currently restricted to FPA, GCC has supported VFP under RISC OS for some time now. So it's worth considering which word order (VFP or FPA) will make data exchange easier in the future.
+
+"The '|' indirection operator" (p161)
+
+  Depending on whether BASIC V or BASIC VI are in use, this will access either a 5-byte float or a (FPA) double-precision float. Consideration is necessary for what word order to use with VFP.
+
+"When an error occurs ... the values of all the variables and so on will still be intact" (p164)
+
+  I.e. when handling VFP errors we need to make sure we check for any error before we write back potentially erroneous values to program variables.
+
+"Format of the CALL parameter block" (p224)
+
+  CALL and USR allow 8-byte floats to be passed to assembler routines. This type will only ever be used in BASIC64, but some consideration is necessary as to whether FPA or VFP word order should be used in BASICVFP.
+
+  If we are to aim for BASICVFP to be a drop-in replacement for BASIC64 then it is only natural for FPA word order to be used. However since FPA is obsolete, it would be more convenient if there was a version of BASIC available which used VFP word ordering, and allowed assembler code to make use of the VFP context which BASIC had created. This would then allow BASICVFP programs to easily be augmented with VFP or NEON assembler routines.
+
+"VARIND" (p229), "STOREA" (p229), "EXPR" (p231)
+
+  The BASIC64 version of these routines use the F0 register to pass and return floating-point values. For BASICVFP this raises two possibilities:
+
+  * Maintain use of the F0 register. For optimal performance BASIC could have two versions of the routines; an internal routine which uses the VFP D0 register, and the external form (as exposed to assembler routines) which uses the FPA F0 register. However this may be difficult to achieve as some routines are specified as not using the stack.
+
+  * Only use VFP D0. This will result in optimal performance but will break compatibility with BASIC64 assembler.
+
+"INPUT#" (p294)
+"PRINT#" (p347)
+
+  An exact specification of the BASIC64 float format is given here, describing the fact that the words are stored in big-endian order. Additionally, both versions of BASIC are capable of reading both 5-byte and 8-byte floats. Therefore for continued interoperability of data files the only sensible choice is to have BASICVFP use FPA word ordering for floats when reading and writing files.
+
+"OSCLI" (p334)
+
+  This makes mention of how some of the interpreter state (e.g. CALL environment information pointer) is exposed to the executed command. Notably, there does not appear to be any way for the executed command to determine whether BASIC or BASIC64 is in use. Potentially a program could probe the end of the CALL environment block to look for the 5-byte float routines. However this does not help us if we require executed commands to be able to differentiate calls made from BASIC64 and calls made from BASICVFP.
+
+"Numeric types" (p411)
+
+  The diagram on p411 represents the storage format of 8-byte floats in BASIC64 (i.e. FPA double precision floats)
+
+
+BASIC64 code review
+===================
+
+FPSR
+----
+
+On entry to the interpreter, BASIC64 initialises the FPSR to &70000, i.e. the Invalid Operation, Division by Zero and Overflow exception traps will be enabled. At no other point during execution is the FPSR read or written, creating an implicit contract between BASIC64 and the user program that correct operation is only guaranteed if the program does not program the FPSR with conflicting settings.
+
+Also of note is that the FPSR is not reset when an error occurs - any code which manipulates the FPSR (e.g. a SWI) and then generates an error without restoring the correct value may break BASIC.
+
+FPOINT
+------
+
+The FPOINT assembly constant is used to select between BASIC (FPOINT=0) and BASIC64 (FPOINT=1), allowing all relevant code to easily be located. An obvious extension of this would be to use FPOINT=2 for BASICVFP.
+
+Workspace layout
+----------------
+
+Although the program and its variables can be relocated (by manipulating the PAGE, TOP, LOMEM and HIMEM variables), BASIC is also reliant on a block of non-relocatable workspace which is anchored to the start of application space (&8000). This static workspace would make a suitable place for storing the VFP context.
+
+
+BASICVFP proposal
+=================
+
+Two of the primary advantages of BASIC64 over standard BASIC are the increased accuracy of floating point calculations, and speed (when FPA hardware is available).
+
+BASICVFP can't improve on precision (FPA supported IEEE extended double precision floats but VFP only supports single and double precision), but due to the prevalence of VFP hardware it can provide significant speed improvements over BASIC64. Therefore the main focus on this proposal is on gaining as much speed as possible, without being too concerned about compatibility with programs that made advanced use of BASIC64 features. I.e. "Generic" BASIC programs which don't make any assumptions about the floating point format (and thus run fine under BASIC or BASIC64) will continue to run under BASICVFP, but programs which assume FPA word ordering or FPA register usage will fail.
+
+In detail:
+
+* INPUT# and PRINT# will use FPA word ordering for double-precision floats, to maintain data file compatibility between different BASIC versions
+
+* All other interfaces will use VFP word ordering and VFP registers (e.g. the "|" operator will use VFP word ordering, and VARIND will return any floating-point value in D0)
+
+* On startup BASIC will create a "full" VFP context, i.e. supporting both VFP+NEON and with all data registers available. This will allow assembler code to easily reuse the context for its own VFP/NEON calculations (although it is still the responsibility of the program to ensure the appropriate instructions/registers are available before it attempts to use them)
+
+* The VFP context will be stored in the non-relocatable workspace located at &8000. However since the context size is variable, the amount of memory to reserve for the context will have to be determined at runtime.
+
+* The FPSCR will be initialised to zero. It will be reset to zero by the error handler (ensuring correct recovery from code which uses VFP short vectors), but other than that it's expected that code will not return to BASIC with it set to a value that will conflict with BASIC's VFP usage.
+
+* FP exceptions will be detected by polling the cumulative exception bits in the FPSCR. For consistency with BASIC64 only the following exceptions will generate errors:
+  * Invalid Operation
+  * Division by Zero
+  * Overflow
+
+* The initial version of BASICVFP is expected to rely on FPA/FPEmulator for implementing the trig/power operations that are not supported by VFP. Future versions may provide more optimal routines, e.g. based on the routines used by Steve Drain's "Float" module.
+
+* The initial version of BASICVFP is not expected to use VFP short vectors or NEON vectors to accelerate array/matrix operations. However these are viable future improvements.
+
+* The initial version of BASICVFP is not expected to tackle the issue of allowing commands executed via OSCLI to detect that BASICVFP is in use.
+
+* A BASICVFP build of the module will be selected by setting FPOINT to 2.
+
+* Since BASICVFP will be incompatible with some BASIC/BASIC64 programs, it will have to respond to a separate command (i.e. '*BASICVFP'), and will most likely be a separate module. A future goal might be to produce a version of BASICVFP which provides "full" BASIC64 compatibility, at the expense of some speed.
+
+* DCFD behaviour will be left alone, for now at least (see discussion below)
+
+
+Other considerations
+====================
+
+Trig/power operations
+---------------------
+
+Since VFP doesn't provide trig/power operations, it may be desirable to expose BASICVFP's trig/power functions to assembler code in a similar manner to how the 5-byte float operations are exposed to BASIC V assembler routines. The addition of these extra routines could also serve as a way for commands executed via OSCLI to detect that BASICVFP is in use. However, there is no standard defined for how this list of routines should be extended, so some care may be needed in order to avoid compatibility issues with code which detects BASIC/BASIC64 via checking for the presence of the 5-byte float ops.
+
+ROM space
+---------
+
+If BASICVFP is to be included in ROM, it is worth revisiting the idea to split the assembler out into a separate module, to allow it to be shared between all three implementations: https://www.riscosopen.org/forum/forums/3/topics/903
+
+Alternatively some ROM space could be saved by dropping BASIC64 from ROM for machines where VFP is available.
+
+DCFD
+----
+
+The DCFD assembler directive has three forms:
+
+* DCFD <number> - produces a FPA-format double precision float
+* DCFD.fpa <number> - produces a FPA-format double precision float
+* DCFD.vfp <number> - produces a VFP-format double precision float
+
+Arguments for whether the plain "DCFD" form should be changed to produce VFP-format double precision floats go both ways.
+
+Arguments for leaving DCFD the same:
+
+* It would result in consistent behaviour across all BASIC versions
+* Since any float-using assembler code is likely to require a rewrite to work with BASICVFP (e.g. to switch from using FPA instructions to VFP), it's not a big deal to require the author to also update any DCFD directives
+
+Arguments for changing:
+
+* Apart from INPUT# and PRINT#, all other code which interacts with double-precision floats expects them to be in VFP format. Therefore it makes sense for DCFD to match the native float type.
+* Making features related to obsolete instruction sets easier to use than features relating to current instruction sets is counter-productive. Therefore "DCFD" (which is shorter and therefore easier to type than "DCFD.vfp") should default to VFP format in BASICVFP.
+* Programmers might forget to add the suffix (DCFS, DCFE and DCFH require no suffixes)
+
+A potential solution might be to introduce a new OPT bit or some other configuration variable which controls the default behaviour. However this may run into issues if the program is run on versions of BASIC which do not recognise the option.

Makefile

@@ -42,4 +42,8 @@ endif
 include StdTools
 include AAsmModule
 
+ifeq ($(COMPONENT),BASICVFP)
+AS            = ObjAsmVFP
+endif
+
 # Dynamic dependencies:

Resources/UK/Messages

@@ -27,6 +27,8 @@ M23:Failed with %0 on line %1
 M24:%1 at line %0
 M25:The program size is %0 bytes, the variables use %1 bytes. There are %2 bytes of memory remaining.
 M26:%0 assembled on %1.
+M27:BASICVFP -help activated (use HELP at the > prompt for more help):
+M28:BASICVFP [-chain] <filename> to run a file (text/tokenised).
BASICVFP -quit <filename> to run a file (text/tokenised) and quit when done.
BASICVFP -load <filename> to start with a file (text/tokenised).
BASICVFP @xxxxxxxx,xxxxxxxx to start with in-core text/tokenised program.
BASICVFP -chain @xxxxxxxx,xxxxxxxx to run in-core text/tokenised program.
 # The errors
 E0:Unknown setting of exception control.
 E1:Silly!
@@ -152,3 +154,4 @@ E120:VFP scalar offset out of range
 E121:VFP scalar must not change in register list
 E122:Too many registers in list
 E123:Label/offset must be word aligned
+E124:Invalid arithmetic operation

VersionASM

@@ -11,13 +11,13 @@
                         GBLS    Module_HelpVersion
                         GBLS    Module_ComponentName
                         GBLS    Module_ComponentPath
-Module_MajorVersion     SETS    "1.66"
-Module_Version          SETA    166
+Module_MajorVersion     SETS    "1.67"
+Module_Version          SETA    167
 Module_MinorVersion     SETS    ""
-Module_Date             SETS    "13 Apr 2017"
-Module_ApplicationDate  SETS    "13-Apr-17"
+Module_Date             SETS    "03 Jun 2017"
+Module_ApplicationDate  SETS    "03-Jun-17"
 Module_ComponentName    SETS    "BASIC"
 Module_ComponentPath    SETS    "castle/RiscOS/Sources/Programmer/BASIC"
-Module_FullVersion      SETS    "1.66"
-Module_HelpVersion      SETS    "1.66 (13 Apr 2017)"
+Module_FullVersion      SETS    "1.67"
+Module_HelpVersion      SETS    "1.67 (03 Jun 2017)"
                         END

VersionNum

-/* (1.66)
+/* (1.67)
  *
  * This file is automatically maintained by srccommit, do not edit manually.
  * Last processed by srccommit version: 1.1.
  *
  */
-#define Module_MajorVersion_CMHG        1.66
+#define Module_MajorVersion_CMHG        1.67
 #define Module_MinorVersion_CMHG        
-#define Module_Date_CMHG                13 Apr 2017
+#define Module_Date_CMHG                03 Jun 2017
 
-#define Module_MajorVersion             "1.66"
-#define Module_Version                  166
+#define Module_MajorVersion             "1.67"
+#define Module_Version                  167
 #define Module_MinorVersion             ""
-#define Module_Date                     "13 Apr 2017"
+#define Module_Date                     "03 Jun 2017"
 
-#define Module_ApplicationDate          "13-Apr-17"
+#define Module_ApplicationDate          "03-Jun-17"
 
 #define Module_ComponentName            "BASIC"
 #define Module_ComponentPath            "castle/RiscOS/Sources/Programmer/BASIC"
 
-#define Module_FullVersion              "1.66"
-#define Module_HelpVersion              "1.66 (13 Apr 2017)"
-#define Module_LibraryVersionInfo       "1:66"
+#define Module_FullVersion              "1.67"
+#define Module_HelpVersion              "1.67 (03 Jun 2017)"
+#define Module_LibraryVersionInfo       "1:67"

hdr/Definitions

@@ -46,11 +46,15 @@ FWACC   RN      R4
 FWACCX  RN      R5
 FWGRD   RN      R6
 FWSIGN  RN      R7
- |
+ ELIF FPOINT=1
 FACC    FN      0
 F1      FN      1
 F2      FN      2
 F7      FN      7
+ ELIF FPOINT=2
+FACC    DN      0
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 IACC    RN      R0                     ;pointer to some data (i.e. a variable's value)
 CLEN    RN      R2                     ;actually contains STRACC+CLEN
@@ -143,11 +147,53 @@ $L      LDMIA   $ADDR,{FACC,FACCX}
         AND     FSIGN,FACCX,#&80000000
         AND     FACCX,FACCX,#255
         MEND
- |
+ ELIF FPOINT=1
         MACRO
 $L      FPUSH
 $L      STFD    FACC,[SP,#-8]!
         MEND
+ ELIF FPOINT=2
+        MACRO
+$L      FPUSH
+$L      FSTD    FACC,[SP,#-8]!
+        MEND
+
+        ; Check for VFP exceptions
+        ; Corrupts NZ, $temp
+        MACRO
+        FPSCRCheck $temp
+        FMRX    $temp,FPSCR
+        TST     $temp,#FPSCR_IOC+FPSCR_DZC+FPSCR_OFC
+        BNE     VFPException
+        MEND
+
+        ;VFP FACC -> FPA F0
+        MACRO
+        FACCtoFPA$cc
+        ASSERT  FACC=0
+        FSTS$cc S0,[SP,#-4]!
+        FSTS$cc S1,[SP,#-4]!
+        LDF$cc.D F0,[SP],#8
+        MEND
+
+        ;FPA F0 -> VFP FACC
+        MACRO
+        FPAtoFACC
+        ASSERT  FACC=0
+        STFD    F0,[SP,#-8]!
+        FLDS    S1,[SP],#4
+        FLDS    S0,[SP],#4
+        MEND
+
+        MACRO
+        FPAop $op
+        ! 0, "TODO native $op"
+        FACCtoFPA
+        $op F0,F0
+        FPAtoFACC
+        MEND
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 
         ;do an expression

hdr/WorkSpace

@@ -68,11 +68,19 @@ VFPTABLES                      #       4                      ;Pointer to assemb
  ]
 R12STORE                       #       4                      ;safe place for r12
 ERRXLATE                       #       4
+ [ FPOINT=2
+FREEPTR                        #       4                      ;dynamic FREE value
+VFPFLAGS                       #       4
+VFPFLAG_Vectors                *       1                      ;VFP short vectors available
+VFPFLAG_NEON                   *       2                      ;Integer NEON available
+ ]
 ; <- Insert new bits here
+        ASSERT @+VARS < VARS
 CALLEDNAME                     #       0-(@)             ;needs to be /at least/ one byte for -quit flag
 ; Note that apart from the first byte (which is used as a Quit flag)
 ; CALLEDNAME is only really used by TWIN & TWINO commands (now removed).
 ; It could (should?) probably be changed to 1 byte and renamed QUIT.
+        ! 0, "CALLEDNAME size " :CC::STR: ?CALLEDNAME
 
  [ FPOINT=0
 CACHESIZE                      *       128
@@ -94,6 +102,17 @@ OPT_extended_instrs            *       16                     ;Enable 'extended
 OPT_mask                       *       15                     ;Mask of our supported flags (not used in any significant manner?)
 
 VCACHE                         #       CACHESIZE*16
+ [ FPOINT=2
+VFPCONTEXT                     #       0                      ;VFP context sits here
+ |
 FREE                           #       0                      ;start of everything else
+ ]
+
+; The BASIC reference manual states that the following must be true
+; (TRACEFILE exposed via CALL environment information pointer)
+        ASSERT TRACEFILE+4 = LOCALARLIST
+        ASSERT TRACEFILE+8 = INSTALLLIST
+        ASSERT TRACEFILE+12 = LIBRARYLIST
+        ASSERT TRACEFILE+16 = OVERPTR
 
                                END

s/Assembler

@@ -2095,9 +2095,15 @@ CASMFPIMM1
         LDMDB   R4!,{R6,R7}
         TEQ     R6,FACC
         TEQEQ   R7,FGRD
- |
+ ELIF FPOINT=1
         LDFS    F1,[R4,#-4]!
         CMF     FACC,F1
+ ELIF FPOINT=2
+        FLDD    D1,[R4,#-8]!
+        FCMPD   FACC,D1
+        FMRX    PC,FPSCR
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         BEQ     CASMFPGOTIMM
         SUBS    R5,R5,#1
@@ -2130,7 +2136,7 @@ CASMFPIMMTB
         =       &80,0,0,0              ; 0.5
         &       &A0000000
         =       &84,0,0,0              ; 10
- |
+ ELIF FPOINT=1
         ;       Aasm doesn't seem to like comma-separated FP constants
         DCFS    0
         DCFS    1
@@ -2140,6 +2146,18 @@ CASMFPIMMTB
         DCFS    5
         DCFS    0.5
         DCFS    10
+ ELIF FPOINT=2
+        ; No way of directly comparing float & double values, so just store as doubles
+        DCFD    0
+        DCFD    1
+        DCFD    2
+        DCFD    3
+        DCFD    4
+        DCFD    5
+        DCFD    0.5
+        DCFD    10
+ |
+        ! 1, "Unknown FPPOINT setting"
  ]
 CASMFPIMMTBEND
 
@@ -2696,10 +2714,17 @@ CASMDCFSSMALL
         RSB     FACCX,FACCX,#9
         ORR     R1,FSIGN,FACC,LSR FACCX
         B       CASMICHK
- |
+ ELIF FPOINT=1
         STFS    FACC,[SP,#-4]!
         LDR     R1,[SP],#4
         B       CASMICHK
+ ELIF FPOINT=2
+        FCVTSD  S0,FACC ; Assume FACC corruptible
+        FPSCRCheck R1
+        FMRS    R1,S0
+        B       CASMICHK
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 CASMDCFD
 
@@ -2717,8 +2742,14 @@ CASMDCFD
         STR     R7,[R2],#4
         MOV     R7,FACC,LSL #21
         STR     R7,[R2],#4
- |
+ ELIF FPOINT=1
         STFD    FACC,[R2],#8
+ ELIF FPOINT=2
+        ASSERT  FACC = 0
+        FSTS    S1,[R2],#4
+        FSTS    S0,[R2],#4
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       CASMXCHK
 
@@ -2738,11 +2769,15 @@ CASMDCFv
         STR     R7,[R2,#4]
         MOV     R7,FACC,LSL #21
         STR     R7,[R2],#8
-  |
+  ELIF FPOINT=1
         STFD    FACC,[R2]
         LDMIA   R2,{R0,R1}
         STR     R0,[R2,#4]
         STR     R1,[R2],#8
+  ELIF FPOINT=2
+        FSTD    FACC,[R2],#8
+  |
+        ! 1, "Unknown FPOINT setting"
   ]
         B       CASMXCHK
 
@@ -2768,8 +2803,15 @@ CASMDCFHSMALL
         ORR     R1,FSIGN,FACC,LSR FACCX
         B       CASMXCHK
   |
+    [ FPOINT=1
         STFS    FACC,[SP,#-4]!
         LDR     R1,[SP],#4
+    |
+        ; We could try and use the half-precision extension here, but it's probably not worth the effort
+        FCVTSD  S0,FACC ; Assume FACC corruptible
+        FPSCRCheck R1
+        FMRS    R1,S0
+    ]
         MOV     R2,R1,LSR #23 ; exponent
         MOV     R0,R1,LSL #8 ; mantissa
         AND     R2,R2,#&FF
@@ -2807,8 +2849,33 @@ CASMDCFE
         STR     FACC,[R2],#4
         MOV     FACC,#0
         STR     FACC,[R2],#4
- |
+ ELIF FPOINT=1
         STFE    FACC,[R2],#12
+ ELIF FPOINT=2
+        FMRRD   R14,R1,FACC ; Transfer to FPA word order
+        ; Only deal with three types of numbers:
+        ; * Zero
+        ; * Denormalised
+        ; * Normalised
+        EORS    R0,R14,R1,LSL #1 ; EQ if (+/-) zero
+        ; Move the sign bit to R0
+        AND     R0,R1,#&80000000
+        BIC     R1,R1,#&80000000
+        BEQ     CASMDCFE0
+        ; Detect denormalised numbers
+        CMP     R1,#1<<20
+        ; Transfer exponent to R0
+        ORRHS   R0,R0,R1,LSR #20
+        ORRHS   R0,R0,#16383-1023 ; Adjust bias
+        ; Shift over the fraction
+        MOV     R1,R1,LSL #31-20
+        ORR     R1,R1,R14,LSR #21
+        MOV     R14,R14,LSL #32-21
+        ; Set J
+        ORRHS   R1,R1,#&80000000
+        BICLO   R1,R1,#&80000000
+CASMDCFE0
+        STMIA   R2!,{R0,R1,R14}
  ]
         B       CASMXCHK
 

s/BASICVFP

+; Copyright 2017 Castle Technology Ltd
+;
+; Licensed under the Apache License, Version 2.0 (the "License");
+; you may not use this file except in compliance with the License.
+; You may obtain a copy of the License at
+;
+;     http://www.apache.org/licenses/LICENSE-2.0
+;
+; Unless required by applicable law or agreed to in writing, software
+; distributed under the License is distributed on an "AS IS" BASIS,
+; WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+; See the License for the specific language governing permissions and
+; limitations under the License.
+;
+;> BASICVFP
+
+FPOINT                          *       2                      ;2 for ARM VFP instruction set
+
+        LNK     ModHead.s

s/Basic

@@ -23,8 +23,23 @@ VFPDataResourceFile
 MODULEMAIN
         WritePSRc USR_mode,R1           ; IRQs+FIQs on, USR26/32 mode
         SWI     OS_GetEnv
+      [ FPOINT=2
+        MOV     R2,R1
+        MOV     R0,#VFPSupport_Context_UserMode+VFPSupport_Context_AppSpace
+        MOV     R1,#32
+        SWI     XVFPSupport_CheckContext
+        MOVVS   R0,#VFPSupport_Context_UserMode+VFPSupport_Context_AppSpace
+        MOVVS   R1,#16                  ; 32 registers not available, try 16
+        SWIVS   VFPSupport_CheckContext
+        MOV     R6,R0
+        MOV     R7,R1
+        SUB     R1,R2,R0
+        SUB     R1,R1,#VARS
+        SUBS    R1,R1,#VFPCONTEXT+256
+      |
         SUB     R1,R1,#VARS
         SUBS    R1,R1,#FREE+256
+      ]
         BPL     MAIN
         ADR     R0,SEVEREERROR
         MOV     R1,#0                   ;global messages
@@ -61,6 +76,12 @@ PUTBACKHAND
         MOV     R0,#0
         STR     R0,[R12,#TRACEFILE-OLDERR] ;kill handle!
         SWINE   XOS_Find
+      [ FPOINT=2
+        LDR     R0,=VFPCONTEXT-OLDERR
+        ADD     R0,R12,R0
+        MOV     R1,#0
+        SWI     XVFPSupport_DestroyContext
+      ]
         MOV     R0,#6
         LDMIA   R12!,{R1,R2,R3}
         SWI     XOS_ChangeEnvironment
@@ -194,6 +215,12 @@ OSQUITR LDR     R1,[R12,#TRACEFILE-OLDERR] ;tracefile handle
         MOV     R0,#0
         STR     R0,[R12,#TRACEFILE-OLDERR] ;kill handle!
         SWINE   XOS_Find                   ;close
+      [ FPOINT=2
+        LDR     R0,=VFPCONTEXT-OLDERR
+        ADD     R0,R12,R0
+        MOV     R1,#0
+        SWI     XVFPSupport_DestroyContext
+      ]
         MOV     R0,#6
         LDMIA   R12!,{R1,R2,R3}
         SWI     XOS_ChangeEnvironment
@@ -214,6 +241,7 @@ MYNAME  =       "ARW!"
 EXITERR DCD     ErrorBase_BASIC + &FE
         DCB     "BASIC program exceeded return code limit",0
         ALIGN
+        LTORG
 MAIN    MOV     ARGP,#VARS
         ;Create the SWI 0              ;MOV PC, R14 in the SWICODE data area
         ;and    then do a SYNCHRONISECODEAREAS, ranged.
@@ -224,7 +252,7 @@ MAIN    MOV     ARGP,#VARS
         ADD     R2, R1, #8             ;Size of SWICODE area
         MOV     R0, #1
         SWI     XOS_SynchroniseCodeAreas
- [ FPOINT=1
+ [ FPOINT=1 :LOR: FPOINT=2             ;only necessary for FPOINT=2 since we still rely on FPA for some ops
         MOV     R0,#&70000
         WFS     R0
  ]
@@ -242,6 +270,25 @@ MAIN    MOV     ARGP,#VARS
         MOV     R0,#0
         MOV     R1,R3
         SWI     OS_Find
+ ]
+ [ FPOINT=2
+        ; Set up VFP context
+        MOV     R0,#VFPSupport_Context_UserMode+VFPSupport_Context_AppSpace+VFPSupport_CreateContext_Activate
+        MOV     R1,R7
+        ADD     R2,ARGP,#VFPCONTEXT
+        MOV     R3,#0
+        SWI     VFPSupport_CreateContext
+        ; Check for interesting features
+        MOV     R0,#VFPSupport_Features_Misc
+        SWI     XVFPSupport_Features
+        ASSERT  VFPSupport_MiscFeature_VFPVectors_HW = VFPFLAG_Vectors
+        ASSERT  VFPFLAG_Vectors < 4 ; if we got back an error pointer, we want the flag to be zero
+        AND     R7,R0,#VFPFLAG_Vectors
+        MOV     R0,#VFPSupport_Features_SystemRegs
+        SWI     VFPSupport_Features ; No X bit, this reason code has been around forever
+        TST     R2,#&F000
+        ORRNE   R7,R7,#VFPFLAG_NEON
+        STR     R7,[ARGP,#VFPFLAGS]
  ]
         ADRL    R0,MSGATLINE
         STR     R0,[ARGP,#ERRXLATE]
@@ -267,7 +314,13 @@ MAIN    MOV     ARGP,#VARS
         ADD     R2,ARGP,#OLDERR
         SWI     XOS_ChangeEnvironment
         STMIA   R9!,{R1,R2}
+ [ FPOINT=2
+        ADD     R0,ARGP,#VFPCONTEXT
+        ADD     R0,R0,R6
+        STR     R0,[ARGP,#FREEPTR]
+ |
         ADD     R0,ARGP,#FREE          ;lomem
+ ]
         STR     R0,[ARGP,#PAGE]
         SWI     OS_GetEnv
         MOV     SP,R1                  ;get himem limit
@@ -382,6 +435,8 @@ TITLE   STMFD   SP!,{R0-R12,R14}
 REPSTR  =       "ARM BBC BASIC V"
  [ FPOINT=1
         =       "I"
+ ELIF FPOINT=2
+        =       "I (VFP)"
  ]
         =       " (C) Acorn 1989",10,13,0
  [ RELEASEVER=0
@@ -493,7 +548,11 @@ ENTRYKEYW2
         SWI     OS_WriteS
         =       "$Name. -help activated (use HELP at the > prompt for more help):",10,13,0
  |
+      [ FPOINT=2
+        MOV     R0,#27
+      |
         MOV     R0,#17+FPOINT
+      ]
         BL      MSGPRNXXX
  ]
         B       ENTRYHELP
@@ -516,7 +575,11 @@ ENTRYHELP
         =       "$Name. @xxxxxxxx,xxxxxxxx to start with in-core text/tokenised program.",10,13
         =       "$Name. -chain @xxxxxxxx,xxxxxxxx to run in-core text/tokenised program.",10,13,0
  |
+      [ FPOINT=2
+        MOV     R0,#28
+      |
         MOV     R0,#20+FPOINT
+      ]
         BL      MSGPRNXXX
  ]
         B       FSASET
@@ -693,8 +756,12 @@ MINUSBC BL      EXPR
         BPL     PLUSBC
         TEQ     FACC,#0
         EORNE   FSIGN,FSIGN,#&80000000
- |
+ ELIF FPOINT=1
         RSFMID  FACC,FACC,#0
+ ELIF FPOINT=2
+        FNEGD   FACC,FACC
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       PLUSBC
 GOTLT2  CMP     R10,#"-"
@@ -769,10 +836,17 @@ PLUSBCFP
         BL      F1LDA
         BL      FADDW
         BL      F1STA
- |
+ ELIF FPOINT=1
         LDFD    F1,[R4]
         ADFD    FACC,F1,FACC
         STFD    FACC,[R4]
+ ELIF FPOINT=2
+        FLDD    D1,[R4]
+        FADDD   FACC,D1,FACC
+        FPSCRCheck R14
+        FSTD    FACC,[R4]
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       NXT
 LETSTNOTCACHE
@@ -1259,6 +1333,13 @@ MSGERR  MOV     ARGP,#VARS             ;cardinal error from outside world
         MOV     R14,#0
         STRB    R14,[ARGP,#MEMM]       ;clear MEMM (and r12) flags
         BL      FLUSHCACHE
+      [ FPOINT=2
+        ; Reset FP context
+        ADD     R0,ARGP,#VFPCONTEXT
+        MOV     R1,#0
+        SWI     XVFPSupport_ChangeContext
+        FMXR    FPSCR,R1
+      ]
         MOV     R0,#&DA
         MOV     R1,#0
         MOV     R2,#0

s/Command

@@ -42,11 +42,17 @@ APPENDNO
         BL      FTENFQ
         BL      FTENFQ
         BL      SFIX
- |
+ ELIF FPOINT=1
         FLTD    FACC,R7
         DVFD    FACC,FACC,#10
         DVFD    FACC,FACC,#10
         FIXZ    IACC,FACC
+ ELIF FPOINT=2
+        ; If we have VFP, we definitely have SMULL, so just use that
+        LDR     IACC,=&28F5C28+1 ; (1<<32)/100 plus a small correction
+        SMULL   R14,IACC,R7,IACC
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         ADD     IACC,IACC,#1
         RSB     R4,IACC,IACC,LSL #2    ;5
@@ -161,6 +167,8 @@ HELP    BL      SPACES
 HELPTTL =       "ARM BBC BASIC V"
  [ FPOINT=1
         =       "I (64b FPA)"
+ ELIF FPOINT=2
+        =       "I (64b VFP)"
  ]
         =       0
 HELPDATE

s/ErrorMsgs

@@ -538,7 +538,7 @@ ERFEXP  BL      MSG
         =       "Exponent range",0
  ]
         ALIGN
- ]
+ ] ; FPOINT=0
 FACERR  BL      MSG
         =       26,72
  [ OWNERRORS=1
@@ -865,7 +865,20 @@ EVFPRIA BL      MSG
   ]
         ALIGN
  ]
-
+ [ FPOINT=2
+VFPException
+        FMRX    R0,FPSCR
+        TST     R0,#FPSCR_DZC
+        BNE     ZDIVOR
+        TST     R0,#FPSCR_OFC
+        BNE     FOVR
+EINVOP  BL      MSG
+        =       0,124
+  [ OWNERRORS=1
+        =       "Invalid operation", 0
+  ]
+        ALIGN
+ ]
  [ standalone
 Basic_ResFiles
   [ :DEF: VFPAssembler

s/Expr

@@ -98,7 +98,7 @@ CREAND  STR     TYPE,[SP,#-4]!
         MOVCS   TYPE,#4                ;ref arrays are integers
         ADDCS   AELINE,AELINE,#1       ;and need an extra character to be ignored
         TEQ     TYPE,#4                ;word align if integer
- [ FPOINT=1
+ [ FPOINT=1 :LOR: FPOINT=2
         TEQNE   TYPE,#8                ;word align if floating point
  ]
         ADDEQ   R2,R2,#3
@@ -371,7 +371,7 @@ LOOKA2  ADD     R0,R0,#1               ;adjust by -2
 LOOKA3  TST     R5,#&FF                ;Check list entry has ended
         BNE     LOOKU1                 ;Otherwise loop
         TEQ     TYPE,#4                ;found
- [ FPOINT=1
+ [ FPOINT=1 :LOR: FPOINT=2
         TEQNE   TYPE,#8
  ]
         SUBNE   R0,R0,#3               ;adjust by -3 but add 3 if type=4, 8
@@ -616,8 +616,12 @@ FPULL
         LDMFD   SP!,{FACC,FACCX}
         AND     FSIGN,FACCX,#&80000000
         AND     FACCX,FACCX,#255
- |
+ ELIF FPOINT=1
         LDFD    FACC,[SP],#8
+ ELIF FPOINT=2
+        FLDD    FACC,[SP],#8
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         MOV     PC,R14
 DIVOP   STR     R14,[SP,#-4]!
@@ -710,7 +714,7 @@ FCMPNE  CMP     FACCX,FWACCX
 FCOMPZZ CMPEQ   FACC,FWACC
 FCOMPRX MOV     IACC,#0
         LDMFD   SP!,{R7,PC}
- |
+ ELIF FPOINT=1
 FCOMPS  LDR     R4,[SP],#4
         FLTD    F1,R4                  ;float first number
         B       FCOMPT
@@ -725,6 +729,26 @@ FCOMPT  MOV     TYPE,#TINTEGER
         CMF     F1,FACC
         MOV     IACC,#0
         LDMFD   SP!,{R7,PC}
+ ELIF FPOINT=2
+FCOMPS  FLDS    S2,[SP],#4
+        FSITOD  D1,S2             ;float first number
+        B       FCOMPT
+FCOMPR  STMFD   SP!,{R14,R7}
+        FPUSH
+        BL      EXPRRECUR              ;R5
+        TEQ     TYPE,#0
+        BEQ     ERTYPEINT
+        ASSERT  FACC = 0
+        FMSRPL  S0,IACC                ;float if rqd
+        FSITODPL FACC,S0
+        FLDD    D1,[SP],#8             ;get first
+FCOMPT  MOV     TYPE,#TINTEGER
+        FCMPD   D1,FACC
+        MOV     IACC,#0
+        FMRX    PC,FPSCR
+        LDMFD   SP!,{R7,PC}
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 STNCMP  STMFD   SP!,{R7,R14}
         BL      SPUSH                  ;preserves R5
@@ -896,7 +920,7 @@ FPLUS   STMFD   SP!,{FACC,FACCX,FSIGN,R7}
 FPLUSS  LDMFD   SP!,{FWACC,FWACCX,FWSIGN}
         TEQ     FWACC,#0
         BLNE    FADDW
- |
+ ELIF FPOINT=1
 FPLUS   STR     R7,[SP,#-4]!
         FPUSH
         BL      EXPRRECUR
@@ -906,6 +930,21 @@ FPLUS   STR     R7,[SP,#-4]!
         LDFD    F1,[SP],#8
         MOV     TYPE,#TFP
 FPLUSS  ADFD    FACC,FACC,F1
+ ELIF FPOINT=2
+FPLUS   STR     R7,[SP,#-4]!
+        FPUSH
+        BL      EXPRRECUR
+        TEQ     TYPE,#0
+        BEQ     ERTYPEINT
+        ASSERT  FACC = 0
+        FMSRPL  S0,IACC
+        FSITODPL FACC,S0
+        FLDD    D1,[SP],#8
+        MOV     TYPE,#TFP
+FPLUSS  FADDD   FACC,FACC,D1
+        FPSCRCheck R14
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 EXPRPULLNEXT
         LDR     R7,[SP],#4
@@ -927,7 +966,7 @@ FMINUS  STMFD   SP!,{FACC,FACCX,FSIGN,R7}
         TEQ     FACC,#0
         EORNE   FSIGN,FSIGN,#&80000000
         B       FPLUSS
- |
+ ELIF FPOINT=1
 FPLUST  LDR     IACC,[SP],#4
         FLTD    F1,IACC                ;no need to change type
         B       FPLUSS
@@ -945,6 +984,30 @@ FMINUS  STR     R7,[SP,#-4]!
         SUFD    FACC,F1,FACC
         MOV     TYPE,#TFP
         B       EXPRPULLNEXT
+ ELIF FPOINT=2
+FPLUST  FLDS    S2,[SP],#4
+        FSITOD  D1,S2
+        B       FPLUSS
+FMINUT  FLDS    S2,[SP],#4
+        FSITOD  D1,S2
+        FSUBD   FACC,D1,FACC
+        FPSCRCheck R14
+        B       EXPRPULLNEXT
+FMINUS  STR     R7,[SP,#-4]!
+        FPUSH
+        BL      EXPRRECUR
+        TEQ     TYPE,#0
+        BEQ     ERTYPEINT
+        ASSERT  FACC = 0
+        FMSRPL  S0,IACC
+        FSITODPL FACC,S0
+        FLDD    D1,[SP],#8
+        FSUBD   FACC,D1,FACC
+        MOV     TYPE,#TFP
+        FPSCRCheck R14
+        B       EXPRPULLNEXT
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 EXPRSUB TEQ     TYPE,#0
         BEQ     ERTYPEINT
@@ -993,7 +1056,7 @@ FTIME   STMFD   SP!,{FACC,FACCX,FSIGN,R7}
 FTIMES  LDMFD   SP!,{FWACC,FWACCX,FWSIGN}
         TEQ     FACC,#0
         BLNE    FMUL0
- |
+ ELIF FPOINT=1
 FTIMEL  FLTD    FACC,IACC
 FTIMET  LDR     R4,[SP],#4
         FLTD    F1,R4
@@ -1008,6 +1071,31 @@ FTIME   STR     R7,[SP,#-4]!
         LDFD    F1,[SP],#8
 FTIMES  MUFD    FACC,F1,FACC
         MOV     TYPE,#TFP
+ ELIF FPOINT=2
+FTIMEL
+        ASSERT  FACC = 0
+        FMSR    S0,IACC
+        FSITOD  FACC,S0
+FTIMET  FLDS    S2,[SP],#4
+        FSITOD  D1,S2
+        B       FTIMES
+FTIMF
+        ASSERT  FACC = 0
+        FMSR    S0,IACC
+        FSITOD  FACC,S0
+FTIME   STR     R7,[SP,#-4]!
+        FPUSH
+        BL      EXPRRECUR              ;R10
+        TEQ     TYPE,#0
+        BEQ     ERTYPEINT
+        FMSRPL  S0,IACC
+        FSITODPL FACC,S0
+        FLDD    D1,[SP],#8
+FTIMES  FMULD   FACC,D1,FACC
+        MOV     TYPE,#TFP
+        FPSCRCheck R14
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         LDR     R7,[SP],#4
         CMP     R7,R10,LSR #28
@@ -1023,7 +1111,7 @@ EXPRDIV BL      FLOATY
         BEQ     ZDIVOR
         LDMFD   SP!,{FWACC,FWACCX,FWSIGN}
         BL      FXDIV0
- |
+ ELIF FPOINT=1
 EXPRDIV TEQ     TYPE,#0
         BLPL    FLOATQ
         STR     R7,[SP,#-4]!
@@ -1035,6 +1123,22 @@ EXPRDIV TEQ     TYPE,#0
         BEQ     ZDIVOR
         LDFD    F1,[SP],#8
         DVFD    FACC,F1,FACC
+ ELIF FPOINT=2
+EXPRDIV TEQ     TYPE,#0
+        BLPL    FLOATQ
+        STR     R7,[SP,#-4]!
+        FPUSH
+        BL      EXPRRECUR              ;R10
+        TEQ     TYPE,#0
+        BLPL    FLOATQ
+        FCMPZD  FACC
+        FMRX    PC,FPSCR
+        BEQ     ZDIVOR
+        FLDD    D1,[SP],#8
+        FDIVD   FACC,D1,FACC
+        FPSCRCheck R14
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         LDR     R7,[SP],#4
         CMP     R7,R10,LSR #28
@@ -1156,7 +1260,7 @@ FPOWC   FPUSH                          ;multiplier
         BL      FMUL
         PULLJ   6
         B       POWEREND
- |
+ ELIF FPOINT=1
         BL      FLOATY
         FPUSH
         BL      FACTOR                 ;shortcut!
@@ -1164,6 +1268,27 @@ FPOWC   FPUSH                          ;multiplier
         FLTPLD  FACC,IACC
         LDFD    F1,[SP],#8
         POWD    FACC,F1,FACC
+POWEREND
+        LDRB    R10,[AELINE],#1
+        CMP     R10,#" "
+        BEQ     POWEREND
+        MOV     TYPE,#TFP
+        LDR     R7,[SP],#4
+        B       EXPRNEXT
+ ELIF FPOINT=2
+        BL      FLOATY
+        FPUSH
+        BL      FACTOR                 ;shortcut!
+        BEQ     ERTYPEINT
+        ! 0, "TODO native POWD"
+        FLTPLD  F0,IACC
+        ASSERT  FACC=0
+        FACCtoFPAMI
+        LDR     R10,[SP,#4]            ;SP VFP -> FPA
+        STR     R10,[SP,#-4]!
+        LDFD    F1,[SP],#12
+        POWD    F0,F1,F0
+        FPAtoFACC
 POWEREND
         LDRB    R10,[AELINE],#1
         CMP     R10,#" "

s/Factor

@@ -408,8 +408,12 @@ VALCMP  RSBPL   IACC,IACC,#0           ;negate integer
         TEQ     FACC,#0
         EORNE   FSIGN,FSIGN,#&80000000 ;negate floating point
         TEQ     TYPE,#0
- |
+ ELIF FPOINT=1
         RSFD    FACC,FACC,#0
+ ELIF FPOINT=2
+        FNEGD   FACC,FACC
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         LDR     PC,[SP],#4
 DATAST  LDRB    R10,[AELINE],#1
@@ -475,8 +479,12 @@ VARFP
         LOAD5   FACC,FACCX,IACC,FGRD,FSIGN
         AND     FSIGN,FACC,#&80000000
         ORRNE   FACC,FACC,#&80000000
- |
+ ELIF FPOINT=1
         LDFD    FACC,[IACC]
+ ELIF FPOINT=2
+        FLDD    FACC,[IACC]
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         MOVS    TYPE,#TFP
         MOV     PC,R14
@@ -527,9 +535,13 @@ TSTN    AND     R1,AELINE,#CACHEMASK
  [ FPOINT=0
         AND     FACCX,TYPE,#255
         AND     FSIGN,TYPE,#TFP
- |
+ ELIF FPOINT=1
         STMFD   SP!,{IACC,TYPE}
         LDFD    FACC,[SP],#8
+ ELIF FPOINT=2
+        FMDRR   FACC,IACC,TYPE
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         MOVS    TYPE,#TFP
         MOV     PC,R14
@@ -541,9 +553,13 @@ TSTNNOTCACHE
         TEQ     TYPE,#0
  [ FPOINT=0
         ORRMI   TYPE,FACCX,FSIGN
- |
+ ELIF FPOINT=1
         STFMID  FACC,[SP,#-8]!
         LDMMIFD SP!,{IACC,TYPE}
+ ELIF FPOINT=2
+        FMRRDMI IACC,TYPE,FACC
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         AND     R5,R6,#CACHEMASK
         ADD     R5,ARGP,R5,LSL #CACHESHIFT
@@ -637,8 +653,12 @@ ABS     STR     R14,[SP,#-4]!
         BEQ     ERTYPEINT
  [ FPOINT=0
         MOVMI   FSIGN,#0
- |
+ ELIF FPOINT=1
         ABSMID  FACC,FACC
+ ELIF FPOINT=2
+        FABSDMI FACC,FACC
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         LDRMI   PC,[SP],#4               ;do fp abs (easy)
         TEQ     IACC,#0
@@ -672,10 +692,16 @@ ACS     STR     R14,[SP,#-4]!
  [ FPOINT=0
         MOV     R10,#1
         B       ASN1
- |
+ ELIF FPOINT=1
         ACSD    FACC,FACC
         MOVS    TYPE,#TFP
         LDR     PC,[SP],#4
+ ELIF FPOINT=2
+        FPAop   ACSD
+        MOVS    TYPE,#TFP
+        LDR     PC,[SP],#4
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 BBGET   STR     R14,[SP,#-4]!
         BL      CHAN
@@ -711,8 +737,12 @@ EXP     STR     R14,[SP,#-4]!
         BLPL    FLOATQ
  [ FPOINT=0
         BL      FEXP
- |
+ ELIF FPOINT=1
         EXPD    FACC,FACC
+ ELIF FPOINT=2
+        FPAop   EXPD
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 FSINSTK MOVS    TYPE,#TFP
         LDR     PC,[SP],#4
@@ -818,10 +848,23 @@ INTF    BL      SFIX
         B       PSINSTK
 INTS    MVN     IACC,#0
         B       PSINSTK
- |
+ ELIF FPOINT=1
         FIXM    IACC,FACC
         MOVS    TYPE,#TINTEGER
         LDR     PC,[SP],#4
+ ELIF FPOINT=2
+        MOV     R14,#FPSCR_RMODE_DOWN
+        FMXR    FPSCR,R14
+        ASSERT  FACC = 0
+        FTOSID  S0,FACC                ; Double -> int, rounding to minus infinity
+        FPSCRCheck R14
+        FMRS    IACC,S0
+        MOV     R14,#0
+        FMXR    FPSCR,R14              ; Restore default rounding mode
+        MOVS    TYPE,#TINTEGER
+        LDR     PC,[SP],#4
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 LEN     STR     R14,[SP,#-4]!
         BL      FACTOR
@@ -835,8 +878,12 @@ LN      STR     R14,[SP,#-4]!
         BLPL    FLOATQ
  [ FPOINT=0
         BL      FLOG
- |
+ ELIF FPOINT=1
         LGND    FACC,FACC
+ ELIF FPOINT=2
+        FPAop   LGND
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       FSINSTK
 LOG     STR     R14,[SP,#-4]!
@@ -846,10 +893,16 @@ LOG     STR     R14,[SP,#-4]!
         BL      FLOG
         ADR     TYPE,RPLN10
         B       FMULFSINSTK
- |
+ ELIF FPOINT=1
         LOGD    FACC,FACC
         MOVS    TYPE,#TFP
         LDR     PC,[SP],#4
+ ELIF FPOINT=2
+        FPAop   LOGD
+        MOVS    TYPE,#TFP
+        LDR     PC,[SP],#4
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 DEG     STR     R14,[SP,#-4]!
         BL      FACTOR
@@ -871,7 +924,7 @@ F180DP  =       &D3,&E0,&2E,&E5
         =       &86,0,0,0              ;5.729577951E1
 FPID180 =       &12,&35,&FA,&8E
         =       &7B,0,0,0              ;1.745329252E-2
- |
+ ELIF FPOINT=1
 FMULFSINSTK
         LDFD    F1,[TYPE]
         MUFD    FACC,FACC,F1
@@ -879,6 +932,17 @@ FMULFSINSTK
         LDR     PC,[SP],#4
 F180DP  DCFD    57.2957795130823208767981548141
 FPID180 DCFD    0.0174532925199432957692369076849
+ ELIF FPOINT=2
+FMULFSINSTK
+        FLDD    D1,[TYPE]
+        FMULD   FACC,FACC,D1
+        FPSCRCheck R14
+        MOVS    TYPE,#TFP
+        LDR     PC,[SP],#4
+F180DP  DCFD    57.2957795130823208767981548141
+FPID180 DCFD    0.0174532925199432957692369076849
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 NOT     STR     R14,[SP,#-4]!
         BL      FACTOR
@@ -952,13 +1016,26 @@ RND     STR     R14,[SP,#-4]!
         BL      SFIX
         ADD     IACC,IACC,#1
         PULLJ   2
- |
+ ELIF FPOINT=1
         FLTD    F7,IACC
         BL      DORANDOM
         BL      FRNDAA
         MUFDZ   FACC,FACC,F7
         FIXZ    IACC,FACC
         ADD     IACC,IACC,#1
+ ELIF FPOINT=2
+        FMSR    S14,IACC
+        FSITOD  D7,S14
+        BL      DORANDOM
+        BL      FRNDAA
+        FMULD   FACC,FACC,D7
+        ASSERT  FACC = 0
+        FTOSIZD S0,FACC
+        FPSCRCheck IACC
+        FMRS    IACC,S0
+        ADD     IACC,IACC,#1
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       PSINSTK
 FRND1   BL      DORANDOM
@@ -1004,8 +1081,12 @@ FCONVERT2
         ORR     R2,R2,IACC,LSR #12     ;build 1st word
         MOV     R3,IACC,LSL #20
         ORR     R3,R3,R1,LSR #12       ;build second word
+     [ FPOINT=1
         STMFD   SP!,{R2,R3}            ;save...
         LDFD    FACC,[SP],#8           ;...and restore in FACC
+     |
+        FMDRR   FACC,R3,R2
+     ]
         MOV     PC,R14
 FCONVERTB
         MOVS    IACC,R1                ;if msword of mantissa zero then 32 bit move
@@ -1036,10 +1117,17 @@ SGN     STR     R14,[SP,#-4]!
         BEQ     SINSTK
         TEQ     FSIGN,#0
         BMI     TRUE
- |
+ ELIF FPOINT=1
         CMF     FACC,#0
         BEQ     FALSE
         BMI     TRUE
+ ELIF FPOINT=2
+        FCMPZD  FACC
+        FMRX    PC, FPSCR
+        BEQ     FALSE
+        BMI     TRUE
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
 INTONE  MOV     IACC,#1
         B       SINSTK
@@ -1203,9 +1291,14 @@ TAN2A   TST     R10,#1
         EORNE   FSIGN,FSIGN,#&80000000
         BLNE    FRECIP
         ADD     SP,SP,#8
- |
+ ELIF FPOINT=1
         BLPL    FLOATQ
         TAND    FACC,FACC
+ ELIF FPOINT=2
+        BLPL    FLOATQ
+        FPAop   TAND
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       FSINSTK
 SQR     STR     R14,[SP,#-4]!
@@ -1213,8 +1306,13 @@ SQR     STR     R14,[SP,#-4]!
         BLPL    FLOATQ
  [ FPOINT=0
         BL      FSQRT
- |
+ ELIF FPOINT=1
         SQTD    FACC,FACC
+ ELIF FPOINT=2
+        FSQRTD  FACC,FACC
+        FPSCRCheck R14
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         B       FSINSTK
 TO      LDRB    R10,[AELINE],#1

s/Funct

@@ -634,8 +634,12 @@ GTARGRETRESTRVFP
         LDMFD   AELINE!,{FACC,FACCX}
         AND     FSIGN,FACCX,#&80000000
         AND     FACCX,FACCX,#255
- |
+ ELIF FPOINT=1
         LDFD    FACC,[AELINE],#8
+ ELIF FPOINT=2
+        FLDD    FACC,[AELINE],#8
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         MOV     TYPE,#TFP
 GTARGRETRESTRV1
@@ -691,7 +695,7 @@ RETWORD STRB    R7,[R6],#1
         MOV     R7,R7,ROR #8
 RETBYTE STRB    R7,[R6]
         MOV     PC,R14
- [ FPOINT=1
+ [ FPOINT=1 :LOR: FPOINT=2
 RETFP   LDMFD   SP!,{R7,R8}
         STMIA   R6,{R7,R8}
         MOV     ARGP,#VARS

s/ModHead

@@ -43,6 +43,9 @@ international_help SETL {TRUE}          ; Default to RISC OS 3.60+ international
         GET     PublicWS
         GET     Sound
         GET     OsWords
+      [ FPOINT=2
+        GET     VFPSupport
+      ]
 
 RELEASEVER                     *       1              ;1 for release version: no Roger
 OWNERRORS                      *       0              ;1 for english error messages in module
@@ -53,8 +56,12 @@ SAFECRUNCH                     *       15             ;15 is a nice safe looking
                                GBLS    NameMsgs
  [ FPOINT=0
 Name                           SETS    "BASIC"
- |
+ ELIF FPOINT=1
 Name                           SETS    "BASIC64"
+ ELIF FPOINT=2
+Name                           SETS    "BASICVFP"
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
  [ standalone
 NameMsgs                       SETS    "$Name.Msgs"
@@ -65,6 +72,7 @@ NameMsgs                       SETS    "Messages"     ;share when not standalone
         AREA    |BASIC$$Code|, CODE, READONLY, PIC
         ENTRY
 
+
 Module_BaseAddr                *       .
 
 Basic_ModuleStart
@@ -126,8 +134,12 @@ Basic_HelpStr
 
  [ FPOINT=0
         DCB     "BBC BASIC V",9,"$Module_HelpVersion", 0
- |
+ ELIF FPOINT=1
         DCB     "BBC BASIC VI",9,"$Module_HelpVersion", 0
+ ELIF FPOINT=2
+        DCB     "BBC BASIC VI (VFP)",9,"$Module_HelpVersion", 0
+ |
+        ! 1, "Unknown FPOINT setting"
  ]
         ALIGN