Let's print the result to the screen

I have found and vendored a piece of code that allows me to convert a
16-bit number so it can be easily printed on screen. With this, let's be
a bit more flashy and print the result to the screen.

Signed-off-by: Miquel Sabaté Solà <msabate@suse.com>

4 files changed, 467 insertions, 19 deletions

diff --git a/assets/alphanum.chr b/assets/alphanum.chr
new file mode 100644
index 0000000..622eb8e
--- /dev/null
+++ b/assets/alphanum.chr
diff --git a/assets/session.nss b/assets/session.nss
new file mode 100644
index 0000000..42f6b13
--- /dev/null
+++ b/assets/session.nss
@@ -0,0 +1,186 @@
+NSTssTXT

+

+BtnTiles=1

+BtnChecker=0

+BtnSelTiles=0

+BtnChrBank1=0

+BtnChrBank2=1

+BtnGridAll=0

+BtnGridTile=0

+BtnGridAtr=1

+BtnGridBlock=0

+BtnGridScr=0

+BtnPal=1

+BtnTypeIn=0

+BtnFrameAll=1

+BtnFrameSelected=0

+BtnFrameNone=0

+BtnSpriteSnap=0

+BtnSprite8x16=0

+BtnForce8x16=1

+MenuBestOffsets=0

+MenuLossy=0

+MenuThreshold=0

+MenuNoColorData=0

+MenuMetaSprAutoInc=0

+MenuMetaSprSkipZero=0

+MenuMetaSprMerge=0

+MenuSaveIncName=1

+MenuSaveIncAttr=1

+MenuSaveRLE=0

+VarTileViewTagBuf=0

+VarBgPalCur=1

+VarPalActive=0

+VarTileActive=54

+VarBankActive=4096

+VarPPUMask=0

+VarPPUMaskSet0=0

+VarPPUMaskSet1=0

+VarPPUMaskSet2=0

+VarPPUMaskSet3=0

+VarPalBank=0

+VarMetaSpriteActive=0

+VarSpriteActive=0

+VarSpriteGridX=64

+VarSpriteGridY=64

+VarNameW=32

+VarNameH=30

+VarNameViewX=0

+VarNameViewY=-1

+VarNameSelectionL=-1

+VarNameSelectionR=0

+VarNameSelectionT=-1

+VarNameSelectionB=0

+VarNameCopyW=-1

+VarNameCopyH=-1

+VarCHRSelectionL=6

+VarCHRSelectionR=7

+VarCHRSelectionT=3

+VarCHRSelectionB=4

+VarCHRCopyW=-1

+VarCHRCopyH=-1

+VarCHRCopyRect=1

+VarCHRSelectRect=1

+VarCHRBanks=4

+VarCHRBanksUndo=4

+VarCheckpoint_CHRBanks=4

+VarCheckpoint_CHRBanksUndo=4

+

+

+VarCHRSelected=00[36]0100[c9]

+

+

+Palette=0f3010000f0121310f0616260f0919290f1121310f1222320f1323330f1424340f1525350f1626360f1727370f1828380f1929390f1a2a3a0f1b2b3b0f1c2c3c

+

+

+PalUndo=0f3010000f0121310f0616260f0919290f1121310f1222320f1323330f1424340f1525350f1626360f1727370f1828380f1929390f1a2a3a0f1b2b3b0f1c2c3c

+

+

+CHRMain=00[1101]3c464a52623c00[a]182808[3]3e00[a]3c42023c407e00[a]3c420c02423c00[a]081828487e0800[a]7e407c02423c00[a]3e407c42[2]3c00[a]3e02040810[2]00[a]3c423c42[2]3c00[a]3c42[2]3e023c00[9]3c4299a1[2]99423c00[9]3c42[2]7e42[2]00[a]7c427c42[2]3c00[a]3c4240[2]423c00[a]7c42[4]7c00[a]7c407040[2]7c00[a]3c40[2]7040[2]00[a]3c42404e423c00[a]42[3]7e42[2]00[a]3e08[4]3e00[a]02[3]42[2]3c00[a]44487048444200[a]40[5]7e00[a]42665a42[3]00[a]4262524a464200[a]3c42[4]3c00[a]7c42[2]7c40[2]00[a]3c42[3]3c0600[a]3c42[2]7c444200[a]3e403c02[2]7c00[a]3e08[5]00[a]42[5]3c00[a]22[4]140800[a]42[3]5a664200[a]422418[2]244200[a]41221408[3]00[a]7e040810207e00[f]102000[e]18[2]00[c98]

+

+

+CHRCopy=00[1000]

+

+

+CHRUndo=00[1101]3c464a52623c00[a]182808[3]3e00[a]3c42023c407e00[a]3c420c02423c00[a]081828487e0800[a]7e407c02423c00[a]3e407c42[2]3c00[a]3e02040810[2]00[a]3c423c42[2]3c00[a]3c42[2]3e023c00[9]3c4299a1[2]99423c00[9]3c42[2]7e42[2]00[a]7c427c42[2]3c00[a]3c4240[2]423c00[a]7c42[4]7c00[a]7c407040[2]7c00[a]3c40[2]7040[2]00[a]3c42404e423c00[a]42[3]7e42[2]00[a]3e08[4]3e00[a]02[3]42[2]3c00[a]44487048444200[a]40[5]7e00[a]42665a42[3]00[a]4262524a464200[a]3c42[4]3c00[a]7c42[2]7c40[2]00[a]3c42[3]3c0600[a]3c42[2]7c444200[a]3e403c02[2]7c00[a]3e08[5]00[a]42[5]3c00[a]22[4]140800[a]42[3]5a664200[a]422418[2]244200[a]41221408[3]00[a]7e040810207e00[f]102000[e]18[2]00[c98]

+

+

+NameTable=00[12a]2a261f1b2d1f00311b232e36[3]00[288]

+

+

+NameCopy=00[3c0]

+

+

+NameUndo=00[12a]2a261f1b2d1f00311b232e36[2]00[289]

+

+

+AttrTable=00[40]

+

+

+AttrCopy=00[40]

+

+

+AttrUndo=00[40]

+

+

+MetaSprites=ff[10000]

+

+

+32_TileCollisionProperties=00000000[200]

+

+

+32_TileCollisionPropertiesUndo=00000000[200]

+

+

+PropertyConditions=1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]

+

+

+PropertyConditionsUndo=1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]1[2]0[2]

+

+

+TilePropConditionType=00[7]02

+

+

+TilePropConditionTypeUndo=00[7]02

+

+

+Checkpoint_Palette=00[40]

+

+

+Checkpoint_PalUndo=00[40]

+

+

+Checkpoint_CHRMain=00[2000]

+

+

+Checkpoint_CHRUndo=00[2000]

+

+

+Checkpoint_NameTable=00[3c0]

+

+

+Checkpoint_NameUndo=00[3c0]

+

+

+Checkpoint_AttrTable=00[40]

+

+

+Checkpoint_AttrUndo=00[40]

+

+

+Checkpoint_MetaSprites=ff[10000]

+

+

+Checkpoint_32_TileCollisionProperties=00000000[200]

+

+

+Checkpoint_32_TileCollisionPropertiesUndo=00000000[200]

+

+

+Checkpoint_PropertyConditions=[20]

+

+

+Checkpoint_PropertyConditionsUndo=[20]

+

+

+Checkpoint_TilePropConditionType=00[7]02

+

+

+Checkpoint_TilePropConditionTypeUndo=00[7]02

+propertyCHRlabel0=solid

+propertyCHRlabel1=dangerous

+propertyCHRlabel2=unlabeled2

+propertyCHRlabel3=unlabeled3

+propertyCHRlabel4=unlabeled4

+propertyCHRlabel5=unlabeled5

+propertyCHRlabel6=escape bit

+propertyCHRlabel7=global condition

+MetaSpriteBankName=session

+

+FilterCHR=1

+FilterName=1

+FileNameCHR=C:\users\mssola\Desktop\src\github.com\mssola\aoc2023.nes\assets\alphanum.chr

+FileNameName=

+FileNamePal=

+FileNameMetaSpriteBank=

+FileNameTileProps=

diff --git a/src/1.s b/src/1.s
index 2980951..202c3ee 100644
--- a/src/1.s
+++ b/src/1.s
@@ -2,13 +2,8 @@
 ;; Day 1 https://adventofcode.com/2023/day/1
 ;;
 ;; This program then iterates over the array on `data` and computes the value on
-;; each row, then accumulating it into a sum variable. This variable is stored
-;; as a 16-bit number on $08-$09. In order to see the actual result, in FCEUX,
-;; for example, you need to go to `Tools -> RAM Watch`, and set a watch on
-;; address $08 which is unsigned and 2-bytes long. This will show you the end
-;; result after a couple of seconds (yes, it's a big ass computation for poor
-;; old NES, and it's quite fast considering that we are not rendering anything
-;; on screen!). The result must be: `54953`.
+;; each row, then accumulating it into a sum variable. The end result will be
+;; printed to the screen, which must be `54953`.
 
 ;; Luckily for us, good ol' NROM can fit the data set for this exercise :D
 .segment "HEADER"
@@ -24,6 +19,8 @@
     .addr nmi, reset, irq
 
 .segment "CHARS"
+    .incbin "../assets/alphanum.chr"
+
 .segment "STARTUP"
 .segment "CODE"
 
@@ -32,6 +29,7 @@
 .include "../include/ppu.s"
 .include "../include/reset.s"
 .include "../include/globals.s"
+.include "../vendor/bcd16.s"
 
 ;; "Variables" used by this program.
 .scope Vars
@@ -80,23 +78,24 @@
 ;; The main function will be called at the end of the `reset` vector as defined
 ;; in `include/reset.s`.
 .proc main
-    cli
-
-    lda #%10010000
-    sta PPU::CONTROL
-    lda #%00011110
-    sta PPU::MASK
-
-    ;; Initialize variables being used for this program.
+    ;; Initialize palettes and variables being used for this program.
+    jsr init_palettes
     jsr Vars::init
 
     ;; This program uses two flags:
-    ;;   - 7 (`render`): whether NMI-code can render stuff on screen. Unused but
-    ;;                   reserved out of habit.
+    ;;   - 7 (`render`): whether NMI-code can render stuff on screen.
     ;;   - 6 (`done`): whether the computation has been done.
-    lda #0
+    ;; Note that we are setting the render flag so the initial message is
+    ;; shown.
+    lda #%10000000
     sta Globals::m_flags
 
+    cli
+
+    lda #%10010000
+    sta PPU::CONTROL
+    lda #%00011110
+    sta PPU::MASK
 @loop:
     jsr compute_next
     jmp @loop
@@ -187,6 +186,17 @@
     lda #%01000000
     ora Globals::m_flags
     sta Globals::m_flags
+
+    ;; Now that we are done, convert the 16-bit number into separate bytes so it
+    ;; can be rendered on NMI-code.
+    lda Vars::m_sum + 1
+    sta bcdNum
+    lda Vars::m_sum
+    sta bcdNum + 1
+    jsr bcdConvert
+
+    ;; Because of the previous, we can set the `render` flag again.
+    SET_RENDER_FLAG
 @end:
     rts
 .endproc
@@ -221,8 +231,136 @@
     rts
 .endproc
 
-;; Unused.
+;; init_palettes copies all the palettes into the proper PPU address.
+.proc init_palettes
+    PPU_ADDR $3F00
+
+    ldx #0
+@load_palettes_loop:
+    lda palettes, x
+    sta PPU::DATA
+    inx
+    cpx #$20
+    bne @load_palettes_loop
+    rts
+palettes:
+    ;; Background (only first one used)
+    .byte $0F, $30, $10, $00
+    .byte $0F, $00, $00, $00
+    .byte $0F, $00, $00, $00
+    .byte $0F, $00, $00, $00
+
+    ;; Foreground (unused)
+    .byte $0F, $00, $00, $00
+    .byte $0F, $00, $00, $00
+    .byte $0F, $00, $00, $00
+    .byte $0F, $00, $00, $00
+.endproc
+
+;; Non-Maskable Interrupts handler.
 nmi:
+    bit PPU::STATUS
+
+    ;; Skip rendering if the `render` flag is not set.
+    bit Globals::m_flags
+    bpl @nmi_next
+
+    ;; Backup registers.
+    pha
+    txa
+    pha
+    tya
+    pha
+
+    ;; If the result is not there yet, then we just print an initial message.
+    ;; Otherwise print the given result.
+    bit Globals::m_flags
+    bvs :+
+    jsr print_message
+    jmp :++
+:
+    jsr print_result
+:
+
+    ;; Reset the scroll.
+    bit PPU::STATUS
+    lda #$00
+    sta PPU::SCROLL
+    sta PPU::SCROLL
+
+    ;; And unset the render flag so the `main` code is unblocked.
+    UNSET_RENDER_FLAG
+
+    ;; Restore registers.
+    pla
+    tay
+    pla
+    tax
+    pla
+@nmi_next:
+    rti
+
+;; Simply print "Please wait..." to the screen.
+.proc print_message
+    WRITE_PPU_DATA $2129, $2A
+    WRITE_PPU_DATA $212A, $26
+    WRITE_PPU_DATA $212B, $1F
+    WRITE_PPU_DATA $212C, $1B
+    WRITE_PPU_DATA $212D, $2D
+    WRITE_PPU_DATA $212E, $1F
+
+    WRITE_PPU_DATA $2130, $31
+    WRITE_PPU_DATA $2131, $1B
+    WRITE_PPU_DATA $2132, $23
+    WRITE_PPU_DATA $2133, $2E
+    WRITE_PPU_DATA $2134, $36
+    WRITE_PPU_DATA $2135, $36
+    WRITE_PPU_DATA $2136, $36
+
+    rts
+.endproc
+
+;; Print the result on screen. This assumes that `bcdResult` already contains
+;; the proper data.
+.proc print_result
+    ;; Clear tiles that were written by `print_message` and that are not re-used
+    ;; here.
+    WRITE_PPU_DATA $2129, $00
+    WRITE_PPU_DATA $212A, $00
+    WRITE_PPU_DATA $212B, $00
+    WRITE_PPU_DATA $212C, $00
+    WRITE_PPU_DATA $212D, $00
+    WRITE_PPU_DATA $2133, $00
+    WRITE_PPU_DATA $2134, $00
+    WRITE_PPU_DATA $2135, $00
+    WRITE_PPU_DATA $2136, $00
+
+    ;; And loop so PPU::ADDRESS $2E21-$2E32 has the data as stored on bcdResult,
+    ;; which is the binary to decimal conversion result when the final
+    ;; computation was done.
+    ldx #4
+    ldy #$2E
+@loop:
+    ;; PPU address.
+    bit PPU::STATUS
+    lda #$21
+    sta PPU::ADDRESS
+    sty PPU::ADDRESS
+
+    ;; PPU data.
+    lda bcdResult, x
+    clc
+    adc #$10
+    sta PPU::DATA
+    dex
+    iny
+    cpy #$33
+    bne @loop
+
+    rts
+.endproc
+
+;; Unused.
 irq:
     rti
 
diff --git a/vendor/bcd16.s b/vendor/bcd16.s
new file mode 100644
index 0000000..452b294
--- /dev/null
+++ b/vendor/bcd16.s
@@ -0,0 +1,124 @@
+; bcd16.s
+; version 20060201
+;
+; Copyright (C) 2006 Damian Yerrick
+;
+; Copying and distribution of this file, with or without
+; modification, are permitted in any medium without royalty provided
+; the copyright notice and this notice are preserved in any source
+; code copies.  This file is offered as-is, without any warranty.
+;
+; @mssola: The memory addresses being used here have been modified so they don't
+; clash with my programs (from $0x to $8x).
+
+.p02
+
+.exportzp bcdNum, bcdResult
+.export bcdConvert
+
+; bcdConvert
+;
+; Given a number in bcdNum (16-bit), converts it to 5 decimal digits
+; in bcdResult.  Unlike most 6502 binary-to-decimal converters, this
+; subroutine doesn't use the decimal mode that was removed from the
+; 2A03 variant of the 6502 processor.
+;
+; For each value of n from 4 to 1, it compares the number to 8*10^n,
+; then 4*10^n, then 2*10^n, then 1*10^n, each time subtracting if
+; possible. After finishing all the comparisons and subtractions in
+; each decimal place value, it writes the digit to the output array
+; as a byte value in the range [0, 9].  Finally, it writes the
+; remainder to element 0.
+;
+; Extension to 24-bit and larger numbers is straightforward:
+; Add a third bcdTable, increase BCD_BITS, and extend the
+; trial subtraction.
+
+; Constants _________________________________________________________
+; BCD_BITS
+;   The highest possible number of bits in the BCD output. Should
+;   roughly equal 4 * log10(2) * x, where x is the width in bits
+;   of the largest binary number to be put in bcdNum.
+; bcdTableLo[y], bcdTableHi[y]
+;   Contains (1 << y) converted from BCD to binary.
+BCD_BITS = 19
+
+; Variables _________________________________________________________
+; bcdNum (input)
+;   Number to be converted to decimal (16-bit little endian).
+;   Overwritten.
+; bcdResult (output)
+;   Decimal digits of result (5-digit little endian).
+; X
+;   Offset of current digit being worked on.
+; Y
+;   Offset into bcdTable*.
+; curDigit
+;   The lower holds the digit being constructed.
+;   The upper nibble contains a sentinel value; when a 1 is shifted
+;   out, the byte is complete and should be copied to result.
+;   (This behavior is called a "ring counter".)
+;   Overwritten.
+; b
+;   Low byte of the result of trial subtraction.
+;   Overwritten.
+bcdNum = $80
+bcdResult = $82
+curDigit = $87
+b = $82
+
+;
+; Completes within 670 cycles.
+;
+
+bcdConvert:
+  lda #$80 >> ((BCD_BITS - 1) & 3)
+  sta curDigit
+  ldx #(BCD_BITS - 1) >> 2
+  ldy #BCD_BITS - 5
+
+@loop:
+  ; Trial subtract this bit to A:b
+  sec
+  lda bcdNum
+  sbc bcdTableLo,y
+  sta b
+  lda bcdNum+1
+  sbc bcdTableHi,y
+
+  ; If A:b > bcdNum then bcdNum = A:b
+  bcc @trial_lower
+  sta bcdNum+1
+  lda b
+  sta bcdNum
+@trial_lower:
+
+  ; Copy bit from carry into digit and pick up
+  ; end-of-digit sentinel into carry
+  rol curDigit
+  dey
+  bcc @loop
+
+  ; Copy digit into result
+  lda curDigit
+  sta bcdResult,x
+  lda #$10  ; Empty digit; sentinel at 4 bits
+  sta curDigit
+  ; If there are digits left, do those
+  dex
+  bne @loop
+  lda bcdNum
+  sta bcdResult
+  rts
+
+bcdTableLo:
+  .byt <10, <20, <40, <80
+  .byt <100, <200, <400, <800
+  .byt <1000, <2000, <4000, <8000
+  .byt <10000, <20000, <40000
+
+bcdTableHi:
+  .byt >10, >20, >40, >80
+  .byt >100, >200, >400, >800
+  .byt >1000, >2000, >4000, >8000
+  .byt >10000, >20000, >40000