RE: [GSoC] Status of font-rs port

[Anurag Thakur]

Hello Alexei,

> Can you confirm that straight line drawing still takes most of the time with the dense rastrerizer?

Running perf on ftbench reports:

 

10ppem:

 

  22.20%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_render_line

  15.37%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_render_quadratic

   6.01%  lt-ftbench  libfreetype.so.6.18.3  [.] TT_Load_Simple_Glyph

   5.76%  lt-ftbench  libfreetype.so.6.18.3  [.] FT_Outline_Get_CBox

   3.79%  lt-ftbench  libfreetype.so.6.18.3  [.] load_truetype_glyph

   3.22%  ld          libbfd-2.39.so         [.] bfd_link_hash_traverse

   2.46%  lt-ftbench  libfreetype.so.6.18.3  [.] FT_Outline_Decompose

   1.23%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_raster_render

 

100ppem:

 

  48.12%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_render_line

  18.94%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_raster_render

   9.13%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_render_quadratic

   2.33%  lt-ftbench  libfreetype.so.6.18.3  [.] TT_Load_Simple_Glyph

   2.11%  lt-ftbench  libfreetype.so.6.18.3  [.] FT_Outline_Get_CBox

   1.75%  lt-ftbench  libm.so.6              [.] fminf32x

   1.40%  lt-ftbench  libfreetype.so.6.18.3  [.] load_truetype_glyph

   1.36%  lt-ftbench  libm.so.6              [.] fmaxf32x

   1.10%  lt-ftbench  libfreetype.so.6.18.3  [.] FT_Outline_Decompose

 

1000ppem:

 

  70.29%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_raster_render

  10.60%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_render_line

   6.88%  lt-ftbench  libc.so.6              [.] 0x00000000001541c4

   6.29%  lt-ftbench  libc.so.6              [.] 0x00000000001541ce

   1.25%  lt-ftbench  libc.so.6              [.] 0x00000000001541d3

   1.22%  lt-ftbench  libc.so.6              [.] 0x00000000001541c9

   0.68%  lt-ftbench  libfreetype.so.6.18.3  [.] dense_render_quadratic

   0.38%  lt-ftbench  libm.so.6              [.] fminf32x

   0.32%  lt-ftbench  libm.so.6              [.] fmaxf32x

 

I believe something changes at around 200ppem that causes the rendering time to increase dramatically.

Will investigate and report.

 

Also, if there is some way of viewing the libc function name instead of address, that would be helpful.

 

Regards

Anurag

 

From: Alexei Podtelezhnikov <apodtele@gmail.com>
Sent: Friday, September 16, 2022 8:04 AM
To: Anurag Thakur <anurag105csec21@bpitindia.edu.in>
Cc: freetype-devel@nongnu.org
Subject: Re: [GSoC] Status of font-rs port

 

 

Hi Anurag,

 

The rasterizer code (including SIMD) has been integrated into the FreeType and seems to be working fine,

Comparison images: dense and smooth renderer (outlines and dots disabled)

 

This is a really nice progress indeed. The images look almost identical. Can also look at some more challenging fonts with very fine curves, eg, Windows' KUNSTLER.TTF. We need to make sure that both rasterizers are tuned to similar quality.

 

The performance is comparable (sometimes better) to the smooth renderer for small font sizes, however it is several times slower for larger sizes

 

This is really quite stunning. I look forward to further progress. Can you confirm that straight line drawing still takes most of the time with the dense rastrerizer?

 

I am thinking of working on the “Direct rendering mode” next. Any pointers regarding that would be appreciated.

 

I am not sure about this one. The whole point of direct mode is to leave the canvas allocation to the client. The dense rasterizer defeats this purpose.

 

And just for reference

List of bugs I have encountered:
1. Ftgrid produces weird output and crashes at large sizes(~170ppem)

2. The inverted bitmap pitch causes rendering differences (can be seen in the “Q” letter of the ftview output) 

 

Hmm. Is this only when Q is on the last row? That would be an ftview bug then. I do not see any problems with parentheses? Is this 2.12.1 demos?
 

3. Ftgrid crashes when selecting monochrome output

 

Do you think this is also a pitch sign bug?

 

4. [TODO] lcd rendering needs to be implemented

 

This is either three interlaced bitmaps after shifts, or 3x resolution with filtering. You should be able to copy this from smooth.

 

Best,

Alexei