From de6ef7ef5e944ec8e2d0a9b6ab7628b7919b0f5c Mon Sep 17 00:00:00 2001 From: Irmen de Jong Date: Fri, 11 Feb 2022 00:16:39 +0100 Subject: [PATCH] doc --- CompilerDevelopment.md | 3 +++ docs/source/todo.rst | 56 ++++++++++++++++++++++++------------------ 2 files changed, 35 insertions(+), 24 deletions(-) diff --git a/CompilerDevelopment.md b/CompilerDevelopment.md index 57ffcace9..14b3f7b48 100644 --- a/CompilerDevelopment.md +++ b/CompilerDevelopment.md @@ -12,6 +12,9 @@ * The same argument applies to `IMemSizer`, and - not entirely sure about that - `IBuiltinFunctions`. #### Steps to take, in conceptual (!) order: + +(note: all these steps have been implemented, rejected or otherwise solved now.) + 1. introduce an abstraction `SourceCode` that encapsulates the origin and actual loading of Prog8 source code - from the local file system (use case: user programs) - from resources (prog8lib) diff --git a/docs/source/todo.rst b/docs/source/todo.rst index 73d1958f2..f62d3358d 100644 --- a/docs/source/todo.rst +++ b/docs/source/todo.rst @@ -3,7 +3,9 @@ TODO For next release ^^^^^^^^^^^^^^^^ -... +- answer the question in zeropage.kt makeAllocation(): numerical variables in zeropage never have an initial value here TODO why not? +- (for 7.9:) remove support for old @"screencodes" string encoding syntax, parser+code+docs +- ... Need help with @@ -21,47 +23,53 @@ Blocked by an official Commander-x16 r39 release Future Things and Ideas ^^^^^^^^^^^^^^^^^^^^^^^ -- remove support for old @"screencodes" string encoding syntax (parser+code+docs) +Compiler: + +- writeAssembly(): make it possible to actually get rid of the VarDecl nodes by fixing the rest of the code mentioned there. - allow "xxx" * constexpr (where constexpr is not a number literal), now gives expression error not same type - unify FunctioncallExpression + FunctioncallStatement and PipeExpression + Pipe statement classes, may require moving Expression/Statement into interfaces instead of abstract base classes - for the pipe operator: recognise a placeholder (``?`` or ``%`` or ``_``) in a non-unary function call to allow non-unary functions in the chain; ``4 |> mkword(?, $44) |> print_uw`` - for the pipe operator: make it 100% syntactic sugar so there's no need for asm codegen like translatePipeExpression - make it possible to inline non-asmsub routines that just contain a single statement (return, functioncall, assignment) - but this requires all identifiers in the inlined expression to be changed to fully scoped names -- simplifyConditionalExpression() should not split expression if it still results in stack-based evaluation -- simplifyConditionalExpression() sometimes introduces needless assignment to r9 tempvar + but this requires all identifiers in the inlined expression to be changed to fully scoped names. + If we can do that why not perhaps also able to inline multi-line subroutines? Why would it be limited to just 1 line? Maybe to protect against code bloat. +- simplifyConditionalExpression() should not split expression if it still results in stack-based evaluation, but how does it know? +- simplifyConditionalExpression() sometimes introduces needless assignment to r9 tempvar (scenario sought) - consider adding McCarthy evaluation to shortcircuit and and or expressions. First do ifs by splitting them up? Then do expressions that compute a value? - use more of Result<> and Either<> to handle errors/ nulls better? -- rethink the whole "isAugmentable" business. Because the way this is determined, should always also be exactly mirrorred in the AugmentableAssignmentAsmGen or you'll get a crash at code gen time. -- can we get rid of pieces of asmgen.AssignmentAsmGen by just reusing the AugmentableAssignment ? generated code should not suffer - make it possible to use cpu opcodes such as 'nop' as variable names by prefixing all asm vars with something such as ``p8v_``? Or not worth it (most 3 letter opcodes as variables are nonsensical anyway) then we can get rid of the instruction lists in the machinedefinitions as well? -- c64: make the graphics.BITMAP_ADDRESS configurable (VIC banking) -- optimize several inner loops in gfx2 even further? -- add modes 2 and 3 to gfx2 (lowres 4 color and 16 color)? -- add a flood fill routine to gfx2? -- add a diskio.f_seek() routine for the Cx16 that uses its seek dos api? -- make it possible for diskio to read and write from more than one file at the same time (= use multiple io channels)? -- fix problems in c128 target -- add (u)word array type (or modifier?) that puts the array into memory as 2 separate byte-arrays 1 for LSB 1 for MSB -> allows for word arrays of length 256 - [problematic due to 64tass:] add a compiler option to not remove unused subroutines. this allows for building library programs. But this won't work with 64tass's .proc ... Perhaps replace all uses of .proc/.pend by .block/.bend will fix that? (but we lose the optimizing aspect of the assembler where it strips out unused code. There's not really a dynamic switch possible as all assembly lib code is static and uses one or the other) -- zig try-based error handling where the V flag could indicate error condition? and/or BRK to jump into monitor on failure? (has to set BRK vector for this) -- get rid of all TODO's in the code ;) +- Zig-like try-based error handling where the V flag could indicate error condition? and/or BRK to jump into monitor on failure? (has to set BRK vector for this) +- add special (u)word array type (or modifier?) that puts the array into memory as 2 separate byte-arrays 1 for LSB 1 for MSB -> allows for word arrays of length 256 +Libraries: -More optimization ideas -^^^^^^^^^^^^^^^^^^^^^^^ -- VariableAllocator: can we think of a smarter strategy for allocating variables into zeropage, rather than first-come-first-served -- translateFunctioncall() in BuiltinFunctionsAsmGen: should be able to assign parameters to a builtin function directly from register(s), this will make the use of a builtin function in a pipe expression more efficient without using a temporary variable -- translateNormalAssignment() -> better code gen for assigning boolean comparison expressions -- when a for loop's loopvariable isn't referenced in the body, and the iterations are known, replace the loop by a repeatloop -- automatically convert if statements that test for multiple values (if X==1 or X==2..) to if X in [1,2,..] statements, instead of just a warning. +- fix the problems in c128 target, and flesh out its libraries. +- c64: make the graphics.BITMAP_ADDRESS configurable (VIC banking) +- optimize several inner loops in gfx2 even further? +- add modes 2 and 3 to gfx2 (lowres 4 color and 16 color)? +- add a flood fill routine to gfx2? +- add a diskio.f_seek() routine for the Cx16 that uses its seek dos api? (only if that's stable) + +Expressions: + +- rethink the whole "isAugmentable" business. Because the way this is determined, should always also be exactly mirrorred in the AugmentableAssignmentAsmGen or you'll get a crash at code gen time. +- can we get rid of pieces of asmgen.AssignmentAsmGen by just reusing the AugmentableAssignment ? generated code should not suffer - rewrite expression tree evaluation such that it doesn't use an eval stack but flatten the tree into linear code that uses a fixed number of predetermined value 'variables'? "Three address code" was mentioned. https://en.wikipedia.org/wiki/Three-address_code these variables have to be unique for each subroutine because they could otherwise be interfered with from irq routines etc. - this removes the need for the BinExprSplitter? (which is problematic and very limited now) and perhaps as well the assignment splitting in BeforeAsmAstChanger too + +Optimizations: + +- VariableAllocator: can we think of a smarter strategy for allocating variables into zeropage, rather than first-come-first-served +- translateFunctioncall() in BuiltinFunctionsAsmGen: should be able to assign parameters to a builtin function directly from register(s), this will make the use of a builtin function in a pipe expression more efficient without using a temporary variable +- translateNormalAssignment() -> better code gen for assigning boolean comparison expressions +- when a for loop's loopvariable isn't referenced in the body, and the iterations are known, replace the loop by a repeatloop +- automatically convert if statements that test for multiple values (if X==1 or X==2..) to if X in [1,2,..] statements, instead of just a warning. - introduce byte-index operator to avoid index multiplications in loops over arrays? see github issue #4