Brand New AggPas Canvas ready for testing

[graemeg]

Hi everyone. 👋

UPDATE: This branch has been merged into the 'develop' branch. The original Agg2Dpas (old Aggpas based canvas) has been disabled, and replaced with this new Hybrid Canvas. It has been tested under Linux, FreeBSD and Windows.

TL;DR:

Compile it with:

  pasbuild clean
  pasbuild compile -p unix,agg         // for Linux or FreeBSD systems
  pasbuild compile -p windows,agg      // for Windows systems, but requires the freetype.dll (see 'extras' directory)
  pasbuild compile -p macos            // for MacOS systems (AggPas is always used)

The Back Story

All I really wanted was to speed up large amounts of horizontal text rendering. Think of an IDE with 60 lines of code, syntax highlighting, different styles (normal, bold, italic) and about 120 characters wide. Well, the existing AggCanvas (compiled with -dAggCanvas) showed that the quality was great, but the performance was terrible. 🙄

To be fair, AggPas is a software renderer, and using glyph outline mode to render high-quality text took its toll. Running the IDE through the Valgrind Profiler showed that a screen of text caused over 1.2mil internal scanline calculation calls. 😬

So, I had an idea... Why not use AggPas for all the 2D rendering (rectangles, lines, fills etc). We get the the beautiful anti-aliased graphics, transparency support etc. And then use native libraries (Xft or GDI) for high speed text rendering. It sounded great and I had a very elegant design in mind (no include files etc). So I implemented it over a few days. But I hit problems with clipping, then problem with z-order, then problems with event queueing the text rendering, the destruction order of popup window and the list went on... I was playing whack-a-mole!🔨

I needed a new plan. How does GTK3/4 and Qt do it so fast? I then stumbled upon an old Haiku article about how they use AGG for the whole OS rendering, including text. They explained it in detail [thank you!], and that was the Eureka moment! Glyph Bitmap Mask Cacheing! Then bypass AGG's scanline calculations and just blend the glyph masks character by character on the screen using AGG (AggPas in my case). No glyph outline calculations for every piece of text I render!

And so I started with implementation number 2. Lessons learnt from my first attempt, made this go much faster than I expected, and I could reuse a lot of code. The results.... It's way better than both the previous and original attempts - no include files - and way way faster than the original.

I present Hybrid Canvas [the name will change 😂]. I did extensive tests with the ide (maximus), docview, uidesigner and every single examples/gui/* demo. I noted two minor bugs only, which I'll fix soon.

If you are willing to test, it's Linux & FreeBSD only for now. If you can report any bugs in this thread, it would be greatly appreciated. 🙏

[andrewd207]

This sounds amazing Graeme, nicely done. I'll give it ago

[graemeg]

Thanks Andrew. Good to see you are still hanging around after all these years. 🙂

[andrewd207]

Yep, I'm still around tinkering 😁 Glad you are too. Always amazed at what you accomplish

[andrewd207]

The text looks fine in maximus, no problems there. :) I had the window set to full screen though and I think there's a noticeable delay in the painting. This video is with hybrid. Note the keys appear at a constant rate. (The text disappears but I think that's a different bug that happens in dev too.) Making the window smaller

Hybrid/Full screen:

Screencast.from.2026-03-21.09-39-20.mp4

Develop/Full screen

Screencast.from.2026-03-21.09-37-09.mp4

Hybrid/small window:

Screencast.from.2026-03-21.09-56-02.mp4

[andrewd207]

here's the gprof output:

Flat profile:

Each sample counts as 0.01 seconds.
  %   cumulative   self              self     total           
 time   seconds   seconds    calls  Ts/call  Ts/call  name    
 84.40     10.66    10.66                             AGG_PIXFMT_RGBA_$$_BGRA32_BLEND_SOLID_HSPAN$hvrNyasU_KLK
  2.14     10.93     0.27                             AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_SWEEP_SCANLINE$SCANLINE_PTR$$BOOLEAN
  1.50     11.12     0.19                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SORT_CELLS
  1.19     11.27     0.15                             AGG_PIXFMT_RGBA_$$_BLEND_PIX_BGRA$UNSIGNED_PTR$LONGINT$LONGINT$LONGINT$LONGINT$LONGWORD
  1.19     11.42     0.15                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_ADD_CUR_CELL
  0.71     11.51     0.09                             AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_CALCULATE_ALPHA$LONGINT$$LONGWORD
  0.55     11.58     0.07                             AGG_MATH_STROKE_$$_STROKE_CALC_CAP$hR5Fka_s1OoB
  0.48     11.64     0.06                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SET_CUR_CELL$LONGINT$LONGINT
  0.48     11.70     0.06                             AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$_BLEND_SOLID_HSPAN$hGHddf3u68nN
  0.48     11.76     0.06                             AGG_RENDERER_SCANLINE$_$RENDERER_SCANLINE_AA_SOLID_$__$$_RENDER$SCANLINE_PTR
  0.40     11.81     0.05                             AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__XMAX$$LONGINT
  0.32     11.85     0.04                             AGG_TRANS_AFFINE_$$_TRANS_AFFINE_TRANSFORM$TRANS_AFFINE_PTR$DOUBLE_PTR$DOUBLE_PTR
  0.24     11.88     0.03                             AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$_SPAN$LONGINT$LONGINT$LONGWORD$$POINTER
  0.24     11.91     0.03                             MATH_$$_ARCTAN2$EXTENDED$EXTENDED$$EXTENDED
  0.24     11.94     0.03                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_RENDER_LINE$LONGINT$LONGINT$LONGINT$LONGINT
  0.24     11.97     0.03                             AGG_RASTERIZER_SCANLINE_AA_$$_QSORT_CELLS$CELL_AA_PTR_PTR$LONGWORD
  0.24     12.00     0.03                             AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_ADD_CELL$LONGINT$LONGWORD
  0.24     12.03     0.03                             AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_ADD_SPAN$LONGINT$LONGWORD$LONGWORD
  0.20     12.06     0.03                             AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__YMIN$$LONGINT
  0.16     12.08     0.02                             AGG_BASICS_$$_IS_STOP$LONGWORD$$BOOLEAN
  0.16     12.10     0.02                             AGG_CLIP_LIANG_BARSKY_$$_CLIPPING_FLAGS_INT$LONGINT$LONGINT$RECT_PTR$$LONGWORD
  0.16     12.12     0.02                             AGG_COLOR$_$RGBA8_$__$$_CONSTRUCT$LONGWORD$LONGWORD$LONGWORD$LONGWORD$$QWORDBOOL
  0.16     12.14     0.02                             AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_ADD_VERTEX$DOUBLE$DOUBLE$LONGWORD
  0.16     12.16     0.02                             AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_REWIND_SCANLINES$$BOOLEAN
  0.16     12.18     0.02                             AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_RESET$LONGINT$LONGINT
  0.16     12.20     0.02                             AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_RESET_SPANS
  0.16     12.22     0.02                             AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_Y$$LONGINT
  0.16     12.24     0.02                             AGG_SPAN_IMAGE_FILTER_RGBA$_$SPAN_IMAGE_FILTER_RGBA_BILINEAR_$__$$_GENERATE$hwTW1KB__ImL
  0.16     12.26     0.02                             FPC_SYSCALL2
  0.12     12.27     0.02                             AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__XMIN$$LONGINT
  0.08     12.28     0.01                             AGG_ARRAY$_$POD_DEQUE_$__$$_CUT_AT$LONGWORD
  0.08     12.29     0.01                             AGG_ARRAY$_$POD_DEQUE_$__$$_ENTRY$$LONGWORD
  0.08     12.30     0.01                             AGG_BASICS$_$SPRINTF$CHAR_PTR$PCHAR$DOUBLE_$$_APPLY
  0.08     12.31     0.01                             AGG_BASICS_$$_IS_END_POLY$LONGWORD$$BOOLEAN
  0.08     12.32     0.01                             AGG_CONV_ADAPTOR_VCGEN$_$CONV_ADAPTOR_VCGEN_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD
  0.08     12.33     0.01                             AGG_CONV_TRANSFORM$_$CONV_TRANSFORM_$__$$_REWIND$LONGWORD
  0.08     12.34     0.01                             AGG_CURVES$_$CURVE3_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD
  0.08     12.35     0.01                             AGG_FONT_FREETYPE$_$FONT_ENGINE_FREETYPE_BASE_$__$$_UPDATE_SIGNATURE
  0.08     12.36     0.01                             AGG_IMAGE_FILTERS$_$IMAGE_FILTER_BILINEAR_$__$$_CALC_WEIGHT$DOUBLE$$DOUBLE
  0.08     12.37     0.01                             AGG_IMAGE_FILTERS$_$IMAGE_FILTER_BILINEAR_$__$$_RADIUS$$DOUBLE
  0.08     12.38     0.01                             AGG_IMAGE_FILTERS$_$IMAGE_FILTER_LUT_$__$$_CALCULATE$IMAGE_FILTER_BASE_PTR$BOOLEAN
  0.08     12.39     0.01                             AGG_PATH_STORAGE$_$PATH_STORAGE_$__$$_STORAGE_PTRS$DOUBLE_PTR_PTR$$INT8U_PTR
  0.08     12.40     0.01                             AGG_PIXFMT_RGBA_$$_BGRA32_BLEND_VLINE$PIXEL_FORMATS_PTR$LONGINT$LONGINT$LONGWORD$AGGCLR_PTR$BYTE
  0.08     12.41     0.01                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_RENDER_HLINE$LONGINT$LONGINT$LONGINT$LONGINT$LONGINT
  0.08     12.42     0.01                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SCANLINE_CELLS$LONGWORD$$CELL_AA_PTR_PTR
  0.08     12.43     0.01                             AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_TOTAL_CELLS$$LONGWORD
  0.08     12.44     0.01                             AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_CLOSE_POLYGON_NO_CLIP
  0.08     12.45     0.01                             AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_GAMMA$VERTEX_SOURCE_PTR
  0.08     12.46     0.01                             AGG_RENDERER_SCANLINE$_$RENDERER_SCANLINE_AA_SOLID_$__$$_PREPARE$LONGWORD
  0.08     12.47     0.01                             AGG_RENDERING_BUFFER$_$RENDERING_BUFFER_$__$$_ROW$LONGWORD$$INT8U_PTR
  0.08     12.48     0.01                             AGG_RENDERING_BUFFER$_$RENDERING_BUFFER_$__$$_ROW_XY$LONGINT$LONGINT$LONGWORD$$INT8U_PTR
  0.08     12.49     0.01                             AGG_SCANLINE$_$SPAN_OBJ_$__$$_INC_OPERATOR
  0.08     12.50     0.01                             AGG_VCGEN_STROKE$_$VCGEN_STROKE_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD
  0.08     12.51     0.01                             CLASSES$_$TFPLIST_$__$$_GET$LONGINT$$POINTER
  0.08     12.52     0.01                             CTHREADS_$$_CRELOCATETHREADVAR$LONGWORD$$POINTER
  0.08     12.53     0.01                             FPG_GLYPH_CACHE$_$TGLYPHCACHE_$__$$_BLITGLYPH$hiUmZoUCxXAF
  0.08     12.54     0.01                             FPG_HYBRID_CANVAS$_$THYBRIDCANVAS_$__$$_DODRAWSTRING$LONGINT$LONGINT$ANSISTRING
  0.08     12.55     0.01                             FPG_HYBRID_CANVAS_$$_FPGCOLORTOAGG$TFPGCOLOR$$RGBA8
  0.08     12.56     0.01                             FPG_IMGUTILS_$$_FPGCALCULATEGRAY$TFPGCOLOR$BOOLEAN$LONGINT$$TFPGCOLOR
  0.08     12.57     0.01                             FPG_MAIN$_$TFPGSTYLE_$__$$_DRAWBUTTONFACE$TFPGCANVAS$LONGINT$LONGINT$LONGINT$LONGINT$TFPGBUTTONFLAGS
  0.08     12.58     0.01                             FPG_MAIN_$$_FPGCOLORTORGB$TFPGCOLOR$$TFPGCOLOR
  0.08     12.59     0.01                             FPG_TREE$_$TFPGTREEVIEW_$__$$_GETCOLUMNLEFT$LONGINT$$LONGINT
  0.08     12.60     0.01                             MATH_$$_FLOOR$EXTENDED$$LONGINT
  0.08     12.61     0.01                             SYSTEM_$$_FILLXXXX_MORETHANTWOXMMS
  0.08     12.62     0.01                             SYSTEM_$$_MOVE$formal$formal$INT64
  0.08     12.63     0.01                             SYSUTILS_$$_COMPARETEXT$ANSISTRING$ANSISTRING$$LONGINT
  0.00     12.63     0.00        1     0.00     0.00  P$MAXIMUS_$$_MAINPROC
  0.00     12.63     0.00        1     0.00     0.00  main

 %         the percentage of the total running time of the
time       program used by this function.

cumulative a running sum of the number of seconds accounted
 seconds   for by this function and those listed above it.

 self      the number of seconds accounted for by this
seconds    function alone.  This is the major sort for this
           listing.

calls      the number of times this function was invoked, if
           this function is profiled, else blank.

 self      the average number of milliseconds spent in this
ms/call    function per call, if this function is profiled,
	   else blank.

 total     the average number of milliseconds spent in this
ms/call    function and its descendents per call, if this
	   function is profiled, else blank.

name       the name of the function.  This is the minor sort
           for this listing. The index shows the location of
	   the function in the gprof listing. If the index is
	   in parenthesis it shows where it would appear in
	   the gprof listing if it were to be printed.


Copyright (C) 2012-2024 Free Software Foundation, Inc.

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.


		     Call graph (explanation follows)


granularity: each sample hit covers 2 byte(s) for 0.08% of 12.63 seconds

index % time    self  children    called     name
                                                 <spontaneous>
[1]     84.4   10.66    0.00                 AGG_PIXFMT_RGBA_$$_BGRA32_BLEND_SOLID_HSPAN$hvrNyasU_KLK [1]
-----------------------------------------------
                                                 <spontaneous>
[2]      2.1    0.27    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_SWEEP_SCANLINE$SCANLINE_PTR$$BOOLEAN [2]
-----------------------------------------------
                                                 <spontaneous>
[3]      1.5    0.19    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SORT_CELLS [3]
-----------------------------------------------
                                                 <spontaneous>
[4]      1.2    0.15    0.00                 AGG_PIXFMT_RGBA_$$_BLEND_PIX_BGRA$UNSIGNED_PTR$LONGINT$LONGINT$LONGINT$LONGINT$LONGWORD [4]
-----------------------------------------------
                                                 <spontaneous>
[5]      1.2    0.15    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_ADD_CUR_CELL [5]
-----------------------------------------------
                                                 <spontaneous>
[6]      0.7    0.09    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_CALCULATE_ALPHA$LONGINT$$LONGWORD [6]
-----------------------------------------------
                                                 <spontaneous>
[7]      0.6    0.07    0.00                 AGG_MATH_STROKE_$$_STROKE_CALC_CAP$hR5Fka_s1OoB [7]
-----------------------------------------------
                                                 <spontaneous>
[8]      0.5    0.06    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SET_CUR_CELL$LONGINT$LONGINT [8]
-----------------------------------------------
                                                 <spontaneous>
[9]      0.5    0.06    0.00                 AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$_BLEND_SOLID_HSPAN$hGHddf3u68nN [9]
-----------------------------------------------
                                                 <spontaneous>
[10]     0.5    0.06    0.00                 AGG_RENDERER_SCANLINE$_$RENDERER_SCANLINE_AA_SOLID_$__$$_RENDER$SCANLINE_PTR [10]
-----------------------------------------------
                                                 <spontaneous>
[11]     0.4    0.05    0.00                 AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__XMAX$$LONGINT [11]
-----------------------------------------------
                                                 <spontaneous>
[12]     0.3    0.04    0.00                 AGG_TRANS_AFFINE_$$_TRANS_AFFINE_TRANSFORM$TRANS_AFFINE_PTR$DOUBLE_PTR$DOUBLE_PTR [12]
-----------------------------------------------
                                                 <spontaneous>
[13]     0.2    0.03    0.00                 AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$_SPAN$LONGINT$LONGINT$LONGWORD$$POINTER [13]
-----------------------------------------------
                                                 <spontaneous>
[14]     0.2    0.03    0.00                 MATH_$$_ARCTAN2$EXTENDED$EXTENDED$$EXTENDED [14]
-----------------------------------------------
                                                 <spontaneous>
[15]     0.2    0.03    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_RENDER_LINE$LONGINT$LONGINT$LONGINT$LONGINT [15]
-----------------------------------------------
                                                 <spontaneous>
[16]     0.2    0.03    0.00                 AGG_RASTERIZER_SCANLINE_AA_$$_QSORT_CELLS$CELL_AA_PTR_PTR$LONGWORD [16]
-----------------------------------------------
                                                 <spontaneous>
[17]     0.2    0.03    0.00                 AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_ADD_CELL$LONGINT$LONGWORD [17]
-----------------------------------------------
                                                 <spontaneous>
[18]     0.2    0.03    0.00                 AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_ADD_SPAN$LONGINT$LONGWORD$LONGWORD [18]
-----------------------------------------------
                                                 <spontaneous>
[19]     0.2    0.03    0.00                 AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__YMIN$$LONGINT [19]
-----------------------------------------------
                                                 <spontaneous>
[20]     0.2    0.02    0.00                 AGG_BASICS_$$_IS_STOP$LONGWORD$$BOOLEAN [20]
-----------------------------------------------
                                                 <spontaneous>
[21]     0.2    0.02    0.00                 AGG_CLIP_LIANG_BARSKY_$$_CLIPPING_FLAGS_INT$LONGINT$LONGINT$RECT_PTR$$LONGWORD [21]
-----------------------------------------------
                                                 <spontaneous>
[22]     0.2    0.02    0.00                 AGG_COLOR$_$RGBA8_$__$$_CONSTRUCT$LONGWORD$LONGWORD$LONGWORD$LONGWORD$$QWORDBOOL [22]
-----------------------------------------------
                                                 <spontaneous>
[23]     0.2    0.02    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_ADD_VERTEX$DOUBLE$DOUBLE$LONGWORD [23]
-----------------------------------------------
                                                 <spontaneous>
[24]     0.2    0.02    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_REWIND_SCANLINES$$BOOLEAN [24]
-----------------------------------------------
                                                 <spontaneous>
[25]     0.2    0.02    0.00                 AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_RESET$LONGINT$LONGINT [25]
-----------------------------------------------
                                                 <spontaneous>
[26]     0.2    0.02    0.00                 AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_RESET_SPANS [26]
-----------------------------------------------
                                                 <spontaneous>
[27]     0.2    0.02    0.00                 AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_Y$$LONGINT [27]
-----------------------------------------------
                                                 <spontaneous>
[28]     0.2    0.02    0.00                 AGG_SPAN_IMAGE_FILTER_RGBA$_$SPAN_IMAGE_FILTER_RGBA_BILINEAR_$__$$_GENERATE$hwTW1KB__ImL [28]
-----------------------------------------------
                                                 <spontaneous>
[29]     0.2    0.02    0.00                 FPC_SYSCALL2 [29]
-----------------------------------------------
                                                 <spontaneous>
[30]     0.1    0.02    0.00                 AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__XMIN$$LONGINT [30]
-----------------------------------------------
                                                 <spontaneous>
[31]     0.1    0.01    0.00                 AGG_ARRAY$_$POD_DEQUE_$__$$_CUT_AT$LONGWORD [31]
-----------------------------------------------
                                                 <spontaneous>
[32]     0.1    0.01    0.00                 AGG_ARRAY$_$POD_DEQUE_$__$$_ENTRY$$LONGWORD [32]
-----------------------------------------------
                                                 <spontaneous>
[33]     0.1    0.01    0.00                 AGG_BASICS$_$SPRINTF$CHAR_PTR$PCHAR$DOUBLE_$$_APPLY [33]
-----------------------------------------------
                                                 <spontaneous>
[34]     0.1    0.01    0.00                 AGG_BASICS_$$_IS_END_POLY$LONGWORD$$BOOLEAN [34]
-----------------------------------------------
                                                 <spontaneous>
[35]     0.1    0.01    0.00                 AGG_CONV_ADAPTOR_VCGEN$_$CONV_ADAPTOR_VCGEN_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD [35]
-----------------------------------------------
                                                 <spontaneous>
[36]     0.1    0.01    0.00                 AGG_CONV_TRANSFORM$_$CONV_TRANSFORM_$__$$_REWIND$LONGWORD [36]
-----------------------------------------------
                                                 <spontaneous>
[37]     0.1    0.01    0.00                 AGG_CURVES$_$CURVE3_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD [37]
-----------------------------------------------
                                                 <spontaneous>
[38]     0.1    0.01    0.00                 AGG_FONT_FREETYPE$_$FONT_ENGINE_FREETYPE_BASE_$__$$_UPDATE_SIGNATURE [38]
-----------------------------------------------
                                                 <spontaneous>
[39]     0.1    0.01    0.00                 AGG_IMAGE_FILTERS$_$IMAGE_FILTER_BILINEAR_$__$$_CALC_WEIGHT$DOUBLE$$DOUBLE [39]
-----------------------------------------------
                                                 <spontaneous>
[40]     0.1    0.01    0.00                 AGG_IMAGE_FILTERS$_$IMAGE_FILTER_BILINEAR_$__$$_RADIUS$$DOUBLE [40]
-----------------------------------------------
                                                 <spontaneous>
[41]     0.1    0.01    0.00                 AGG_IMAGE_FILTERS$_$IMAGE_FILTER_LUT_$__$$_CALCULATE$IMAGE_FILTER_BASE_PTR$BOOLEAN [41]
-----------------------------------------------
                                                 <spontaneous>
[42]     0.1    0.01    0.00                 AGG_PATH_STORAGE$_$PATH_STORAGE_$__$$_STORAGE_PTRS$DOUBLE_PTR_PTR$$INT8U_PTR [42]
-----------------------------------------------
                                                 <spontaneous>
[43]     0.1    0.01    0.00                 AGG_PIXFMT_RGBA_$$_BGRA32_BLEND_VLINE$PIXEL_FORMATS_PTR$LONGINT$LONGINT$LONGWORD$AGGCLR_PTR$BYTE [43]
-----------------------------------------------
                                                 <spontaneous>
[44]     0.1    0.01    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_RENDER_HLINE$LONGINT$LONGINT$LONGINT$LONGINT$LONGINT [44]
-----------------------------------------------
                                                 <spontaneous>
[45]     0.1    0.01    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SCANLINE_CELLS$LONGWORD$$CELL_AA_PTR_PTR [45]
-----------------------------------------------
                                                 <spontaneous>
[46]     0.1    0.01    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_TOTAL_CELLS$$LONGWORD [46]
-----------------------------------------------
                                                 <spontaneous>
[47]     0.1    0.01    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_CLOSE_POLYGON_NO_CLIP [47]
-----------------------------------------------
                                                 <spontaneous>
[48]     0.1    0.01    0.00                 AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_GAMMA$VERTEX_SOURCE_PTR [48]
-----------------------------------------------
                                                 <spontaneous>
[49]     0.1    0.01    0.00                 AGG_RENDERER_SCANLINE$_$RENDERER_SCANLINE_AA_SOLID_$__$$_PREPARE$LONGWORD [49]
-----------------------------------------------
                                                 <spontaneous>
[50]     0.1    0.01    0.00                 AGG_RENDERING_BUFFER$_$RENDERING_BUFFER_$__$$_ROW$LONGWORD$$INT8U_PTR [50]
-----------------------------------------------
                                                 <spontaneous>
[51]     0.1    0.01    0.00                 AGG_RENDERING_BUFFER$_$RENDERING_BUFFER_$__$$_ROW_XY$LONGINT$LONGINT$LONGWORD$$INT8U_PTR [51]
-----------------------------------------------
                                                 <spontaneous>
[52]     0.1    0.01    0.00                 AGG_SCANLINE$_$SPAN_OBJ_$__$$_INC_OPERATOR [52]
-----------------------------------------------
                                                 <spontaneous>
[53]     0.1    0.01    0.00                 AGG_VCGEN_STROKE$_$VCGEN_STROKE_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD [53]
-----------------------------------------------
                                                 <spontaneous>
[54]     0.1    0.01    0.00                 CLASSES$_$TFPLIST_$__$$_GET$LONGINT$$POINTER [54]
-----------------------------------------------
                                                 <spontaneous>
[55]     0.1    0.01    0.00                 CTHREADS_$$_CRELOCATETHREADVAR$LONGWORD$$POINTER [55]
-----------------------------------------------
                                                 <spontaneous>
[56]     0.1    0.01    0.00                 FPG_GLYPH_CACHE$_$TGLYPHCACHE_$__$$_BLITGLYPH$hiUmZoUCxXAF [56]
-----------------------------------------------
                                                 <spontaneous>
[57]     0.1    0.01    0.00                 FPG_HYBRID_CANVAS$_$THYBRIDCANVAS_$__$$_DODRAWSTRING$LONGINT$LONGINT$ANSISTRING [57]
-----------------------------------------------
                                                 <spontaneous>
[58]     0.1    0.01    0.00                 FPG_HYBRID_CANVAS_$$_FPGCOLORTOAGG$TFPGCOLOR$$RGBA8 [58]
-----------------------------------------------
                                                 <spontaneous>
[59]     0.1    0.01    0.00                 FPG_IMGUTILS_$$_FPGCALCULATEGRAY$TFPGCOLOR$BOOLEAN$LONGINT$$TFPGCOLOR [59]
-----------------------------------------------
                                                 <spontaneous>
[60]     0.1    0.01    0.00                 FPG_MAIN$_$TFPGSTYLE_$__$$_DRAWBUTTONFACE$TFPGCANVAS$LONGINT$LONGINT$LONGINT$LONGINT$TFPGBUTTONFLAGS [60]
-----------------------------------------------
                                                 <spontaneous>
[61]     0.1    0.01    0.00                 FPG_MAIN_$$_FPGCOLORTORGB$TFPGCOLOR$$TFPGCOLOR [61]
-----------------------------------------------
                                                 <spontaneous>
[62]     0.1    0.01    0.00                 FPG_TREE$_$TFPGTREEVIEW_$__$$_GETCOLUMNLEFT$LONGINT$$LONGINT [62]
-----------------------------------------------
                                                 <spontaneous>
[63]     0.1    0.01    0.00                 MATH_$$_FLOOR$EXTENDED$$LONGINT [63]
-----------------------------------------------
                                                 <spontaneous>
[64]     0.1    0.01    0.00                 SYSTEM_$$_FILLXXXX_MORETHANTWOXMMS [64]
-----------------------------------------------
                                                 <spontaneous>
[65]     0.1    0.01    0.00                 SYSTEM_$$_MOVE$formal$formal$INT64 [65]
-----------------------------------------------
                                                 <spontaneous>
[66]     0.1    0.01    0.00                 SYSUTILS_$$_COMPARETEXT$ANSISTRING$ANSISTRING$$LONGINT [66]
-----------------------------------------------
                0.00    0.00       1/1           main [68]
[67]     0.0    0.00    0.00       1         P$MAXIMUS_$$_MAINPROC [67]
-----------------------------------------------
                0.00    0.00       1/1           FPC_SysEntry [4400]
[68]     0.0    0.00    0.00       1         main [68]
                0.00    0.00       1/1           P$MAXIMUS_$$_MAINPROC [67]
-----------------------------------------------

 This table describes the call tree of the program, and was sorted by
 the total amount of time spent in each function and its children.

 Each entry in this table consists of several lines.  The line with the
 index number at the left hand margin lists the current function.
 The lines above it list the functions that called this function,
 and the lines below it list the functions this one called.
 This line lists:
     index	A unique number given to each element of the table.
		Index numbers are sorted numerically.
		The index number is printed next to every function name so
		it is easier to look up where the function is in the table.

     % time	This is the percentage of the `total' time that was spent
		in this function and its children.  Note that due to
		different viewpoints, functions excluded by options, etc,
		these numbers will NOT add up to 100%.

     self	This is the total amount of time spent in this function.

     children	This is the total amount of time propagated into this
		function by its children.

     called	This is the number of times the function was called.
		If the function called itself recursively, the number
		only includes non-recursive calls, and is followed by
		a `+' and the number of recursive calls.

     name	The name of the current function.  The index number is
		printed after it.  If the function is a member of a
		cycle, the cycle number is printed between the
		function's name and the index number.


 For the function's parents, the fields have the following meanings:

     self	This is the amount of time that was propagated directly
		from the function into this parent.

     children	This is the amount of time that was propagated from
		the function's children into this parent.

     called	This is the number of times this parent called the
		function `/' the total number of times the function
		was called.  Recursive calls to the function are not
		included in the number after the `/'.

     name	This is the name of the parent.  The parent's index
		number is printed after it.  If the parent is a
		member of a cycle, the cycle number is printed between
		the name and the index number.

 If the parents of the function cannot be determined, the word
 `<spontaneous>' is printed in the `name' field, and all the other
 fields are blank.

 For the function's children, the fields have the following meanings:

     self	This is the amount of time that was propagated directly
		from the child into the function.

     children	This is the amount of time that was propagated from the
		child's children to the function.

     called	This is the number of times the function called
		this child `/' the total number of times the child
		was called.  Recursive calls by the child are not
		listed in the number after the `/'.

     name	This is the name of the child.  The child's index
		number is printed after it.  If the child is a
		member of a cycle, the cycle number is printed
		between the name and the index number.

 If there are any cycles (circles) in the call graph, there is an
 entry for the cycle-as-a-whole.  This entry shows who called the
 cycle (as parents) and the members of the cycle (as children.)
 The `+' recursive calls entry shows the number of function calls that
 were internal to the cycle, and the calls entry for each member shows,
 for that member, how many times it was called from other members of
 the cycle.


Copyright (C) 2012-2024 Free Software Foundation, Inc.

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.


Index by function name

  [31] AGG_ARRAY$_$POD_DEQUE_$__$$_CUT_AT$LONGWORD [8] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SET_CUR_CELL$LONGINT$LONGINT [26] AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_RESET_SPANS
  [32] AGG_ARRAY$_$POD_DEQUE_$__$$_ENTRY$$LONGWORD [3] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SORT_CELLS [27] AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_Y$$LONGINT
  [33] AGG_BASICS$_$SPRINTF$CHAR_PTR$PCHAR$DOUBLE_$$_APPLY [46] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_TOTAL_CELLS$$LONGWORD [28] AGG_SPAN_IMAGE_FILTER_RGBA$_$SPAN_IMAGE_FILTER_RGBA_BILINEAR_$__$$_GENERATE$hwTW1KB__ImL
  [34] AGG_BASICS_$$_IS_END_POLY$LONGWORD$$BOOLEAN [23] AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_ADD_VERTEX$DOUBLE$DOUBLE$LONGWORD [12] AGG_TRANS_AFFINE_$$_TRANS_AFFINE_TRANSFORM$TRANS_AFFINE_PTR$DOUBLE_PTR$DOUBLE_PTR
  [20] AGG_BASICS_$$_IS_STOP$LONGWORD$$BOOLEAN [6] AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_CALCULATE_ALPHA$LONGINT$$LONGWORD [53] AGG_VCGEN_STROKE$_$VCGEN_STROKE_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD
  [21] AGG_CLIP_LIANG_BARSKY_$$_CLIPPING_FLAGS_INT$LONGINT$LONGINT$RECT_PTR$$LONGWORD [47] AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_CLOSE_POLYGON_NO_CLIP [54] CLASSES$_$TFPLIST_$__$$_GET$LONGINT$$POINTER
  [22] AGG_COLOR$_$RGBA8_$__$$_CONSTRUCT$LONGWORD$LONGWORD$LONGWORD$LONGWORD$$QWORDBOOL [48] AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_GAMMA$VERTEX_SOURCE_PTR [55] CTHREADS_$$_CRELOCATETHREADVAR$LONGWORD$$POINTER
  [35] AGG_CONV_ADAPTOR_VCGEN$_$CONV_ADAPTOR_VCGEN_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD [24] AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_REWIND_SCANLINES$$BOOLEAN [29] FPC_SYSCALL2
  [36] AGG_CONV_TRANSFORM$_$CONV_TRANSFORM_$__$$_REWIND$LONGWORD [2] AGG_RASTERIZER_SCANLINE_AA$_$RASTERIZER_SCANLINE_AA_$__$$_SWEEP_SCANLINE$SCANLINE_PTR$$BOOLEAN [56] FPG_GLYPH_CACHE$_$TGLYPHCACHE_$__$$_BLITGLYPH$hiUmZoUCxXAF
  [37] AGG_CURVES$_$CURVE3_$__$$_VERTEX$DOUBLE_PTR$DOUBLE_PTR$$LONGWORD [16] AGG_RASTERIZER_SCANLINE_AA_$$_QSORT_CELLS$CELL_AA_PTR_PTR$LONGWORD [57] FPG_HYBRID_CANVAS$_$THYBRIDCANVAS_$__$$_DODRAWSTRING$LONGINT$LONGINT$ANSISTRING
  [38] AGG_FONT_FREETYPE$_$FONT_ENGINE_FREETYPE_BASE_$__$$_UPDATE_SIGNATURE [9] AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$_BLEND_SOLID_HSPAN$hGHddf3u68nN [58] FPG_HYBRID_CANVAS_$$_FPGCOLORTOAGG$TFPGCOLOR$$RGBA8
  [39] AGG_IMAGE_FILTERS$_$IMAGE_FILTER_BILINEAR_$__$$_CALC_WEIGHT$DOUBLE$$DOUBLE [13] AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$_SPAN$LONGINT$LONGINT$LONGWORD$$POINTER [59] FPG_IMGUTILS_$$_FPGCALCULATEGRAY$TFPGCOLOR$BOOLEAN$LONGINT$$TFPGCOLOR
  [40] AGG_IMAGE_FILTERS$_$IMAGE_FILTER_BILINEAR_$__$$_RADIUS$$DOUBLE [11] AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__XMAX$$LONGINT [60] FPG_MAIN$_$TFPGSTYLE_$__$$_DRAWBUTTONFACE$TFPGCANVAS$LONGINT$LONGINT$LONGINT$LONGINT$TFPGBUTTONFLAGS
  [41] AGG_IMAGE_FILTERS$_$IMAGE_FILTER_LUT_$__$$_CALCULATE$IMAGE_FILTER_BASE_PTR$BOOLEAN [30] AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__XMIN$$LONGINT [61] FPG_MAIN_$$_FPGCOLORTORGB$TFPGCOLOR$$TFPGCOLOR
   [7] AGG_MATH_STROKE_$$_STROKE_CALC_CAP$hR5Fka_s1OoB [19] AGG_RENDERER_BASE$_$RENDERER_BASE_$__$$__YMIN$$LONGINT [62] FPG_TREE$_$TFPGTREEVIEW_$__$$_GETCOLUMNLEFT$LONGINT$$LONGINT
  [42] AGG_PATH_STORAGE$_$PATH_STORAGE_$__$$_STORAGE_PTRS$DOUBLE_PTR_PTR$$INT8U_PTR [49] AGG_RENDERER_SCANLINE$_$RENDERER_SCANLINE_AA_SOLID_$__$$_PREPARE$LONGWORD [14] MATH_$$_ARCTAN2$EXTENDED$EXTENDED$$EXTENDED
   [1] AGG_PIXFMT_RGBA_$$_BGRA32_BLEND_SOLID_HSPAN$hvrNyasU_KLK [10] AGG_RENDERER_SCANLINE$_$RENDERER_SCANLINE_AA_SOLID_$__$$_RENDER$SCANLINE_PTR [63] MATH_$$_FLOOR$EXTENDED$$LONGINT
  [43] AGG_PIXFMT_RGBA_$$_BGRA32_BLEND_VLINE$PIXEL_FORMATS_PTR$LONGINT$LONGINT$LONGWORD$AGGCLR_PTR$BYTE [50] AGG_RENDERING_BUFFER$_$RENDERING_BUFFER_$__$$_ROW$LONGWORD$$INT8U_PTR [67] P$MAXIMUS_$$_MAINPROC
   [4] AGG_PIXFMT_RGBA_$$_BLEND_PIX_BGRA$UNSIGNED_PTR$LONGINT$LONGINT$LONGINT$LONGINT$LONGWORD [51] AGG_RENDERING_BUFFER$_$RENDERING_BUFFER_$__$$_ROW_XY$LONGINT$LONGINT$LONGWORD$$INT8U_PTR [64] SYSTEM_$$_FILLXXXX_MORETHANTWOXMMS
   [5] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_ADD_CUR_CELL [52] AGG_SCANLINE$_$SPAN_OBJ_$__$$_INC_OPERATOR [65] SYSTEM_$$_MOVE$formal$formal$INT64
  [44] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_RENDER_HLINE$LONGINT$LONGINT$LONGINT$LONGINT$LONGINT [17] AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_ADD_CELL$LONGINT$LONGWORD [66] SYSUTILS_$$_COMPARETEXT$ANSISTRING$ANSISTRING$$LONGINT
  [15] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_RENDER_LINE$LONGINT$LONGINT$LONGINT$LONGINT [18] AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_ADD_SPAN$LONGINT$LONGWORD$LONGWORD [68] main
  [45] AGG_RASTERIZER_SCANLINE_AA$_$OUTLINE_AA_$__$$_SCANLINE_CELLS$LONGWORD$$CELL_AA_PTR_PTR [25] AGG_SCANLINE_U$_$SCANLINE_U8_$__$$_RESET$LONGINT$LONGINT

[graemeg]

Yup, that is the expected output. BGRA32_BLEND_SOLID_HSPAN is how it renders the bitmap glyph masks - avoiding the very expensive scanline calculations (which originally was in the millions).

[graemeg]

That profiler output was expected, but also very revealing.... 84.4% of the time spent calling BGRA32_BLEND_SOLID_HSPAN. That's still a massive chunk.

That gave me an idea. Most of that was rendering fully opaque solid colours - the most common case in application UI with solid background like the text editor, buttons, window backgrounds etc. As the hybrid canvas renders to a buffer (not window canvas), I can add a "fast-path" by modify the buffer directly, instead of going through AggPas's full alpha-blend rasteriser.

Commit aa61ac9 did just that. The results, against the IDE speak for themselves 🥳 :

  ┌──────────────────────────┬────────────────┬───────────────┬────────────────────┐
  │          Metric          │     Before     │     After     │       Change       │
  ├──────────────────────────┼────────────────┼───────────────┼────────────────────┤
  │ BGRA32_BLEND_SOLID_HSPAN │ 84.4% (10.66s) │ 2.33% (0.02s) │ 533x fewer seconds │
  ├──────────────────────────┼────────────────┼───────────────┼────────────────────┤
  │ Total runtime            │ 12.63s         │ 0.86s         │ 14.7x faster       │
  └──────────────────────────┴────────────────┴───────────────┴────────────────────┘

BGRA32_BLEND_SOLID_HSPAN dropped from 84.4% to 2.33%, meaning it's now only called for the small number of fills that
genuinely need alpha blending (transparent colours).

[andrewd207]

I tried it again now with your updates and it's super fast with no lag when typing. I even made it two monitor sized and it worked great. Nicely done

[graemeg]

Thank you for taking the time to test and report back. Much appreciated.

[graemeg]

UPDATE: For easier testing, and because I consider the new Hybrid Canvas way better than the original AggPas canvas, I've merged this branch into 'develop'. See updated instructions on how to activate and build it. It has now taken over the original -dAggCanvas compiler define.

The old Aggpas Canvas (Agg2D.pas) is still their, but unused. I'll probably delete it some time in the future.

[andrewd207]

You had asked about gprof, I'm not sure how I did it. I just did the option -pg and it compiled. But I'm using 3.3.1 not 3.2.2. fwiw here's what it's linked to. I'm on Ubuntu nobel. I had to compile it using the ppcx64 -Fu -FU ... -g -gl -pg options. for whatever reason I couldn't use pasbuild with different options.

linux-vdso.so.1 (0x00007ffe8753e000)
	libX11.so.6 => /lib/x86_64-linux-gnu/libX11.so.6 (0x000078e273f0a000)
	libXext.so.6 => /lib/x86_64-linux-gnu/libXext.so.6 (0x000078e273ef6000)
	libXft.so.2 => /lib/x86_64-linux-gnu/libXft.so.2 (0x000078e273edb000)
	libfontconfig.so.1 => /lib/x86_64-linux-gnu/libfontconfig.so.1 (0x000078e273e8a000)
	libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x000078e273c00000)
	libxcb.so.1 => /lib/x86_64-linux-gnu/libxcb.so.1 (0x000078e273e5f000)
	libfreetype.so.6 => /lib/x86_64-linux-gnu/libfreetype.so.6 (0x000078e273b34000)
	libXrender.so.1 => /lib/x86_64-linux-gnu/libXrender.so.1 (0x000078e273e53000)
	libexpat.so.1 => /lib/x86_64-linux-gnu/libexpat.so.1 (0x000078e273e27000)
	/lib64/ld-linux-x86-64.so.2 (0x000078e27407b000)
	libXau.so.6 => /lib/x86_64-linux-gnu/libXau.so.6 (0x000078e273e21000)
	libXdmcp.so.6 => /lib/x86_64-linux-gnu/libXdmcp.so.6 (0x000078e273e17000)
	libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 (0x000078e273b18000)
	libbz2.so.1.0 => /lib/x86_64-linux-gnu/libbz2.so.1.0 (0x000078e273b04000)
	libpng16.so.16 => /lib/x86_64-linux-gnu/libpng16.so.16 (0x000078e273acc000)
	libbrotlidec.so.1 => /lib/x86_64-linux-gnu/libbrotlidec.so.1 (0x000078e273abe000)
	libbsd.so.0 => /lib/x86_64-linux-gnu/libbsd.so.0 (0x000078e273aa8000)
	libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x000078e2739bf000)
	libbrotlicommon.so.1 => /lib/x86_64-linux-gnu/libbrotlicommon.so.1 (0x000078e27399c000)
	libmd.so.0 => /lib/x86_64-linux-gnu/libmd.so.0 (0x000078e27398d000)

[graemeg]

😃 I use both FPC 3.2.2 and 3.3.1 - initially I thought 3.3.1 broke gprof, but then when I tried 3.2.2 the same problem happened - failed linking stage. Just after I posted, I did some more searching. Seems linux distros scatter libraries in all different places now [I miss FreeBSD], and the default library search path that FPC uses, couldn't find it.

My solution for Ubuntu 24.04 was here: https://lists.freepascal.org/pipermail/fpc-pascal/2026-March/063203.html

gprof is so much better than valgrind and it doesn't slow down the application. Valgrind makes my applications crawl - which I think might affect the profiler results too.

[andrewd207]

Ah I already had this in my /etc/fpc.cfg so that explains it

# path to the gcclib
#ifdef cpux86_64
-Fl/usr/lib/gcc/x86_64-linux-gnu/13
#endif