| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
modify values if text and/or font differs, skipping markShapeDirty() saves performance.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
undesired (Graph VBAA + MSAA); Add NonFSAAGLCapabilitiesChooser
Notable: On RaspiPi4b w/ Mesa3D's Broadcom/VC driver,
the chosen capabilities is a multisamnple one even though not requested.
This causes
- extra performance overhead
- doubled AA: 1st our VBAA, then the FSAA (multisample) -> loss of sharpness
Simply dropping the undersired FSAA helps and ups performance
on the Raspi board (22 -> 35 fps).
|
| |
|
|
|
|
| |
selected commits)
|
|
|
|
| |
TextRegionUtil.countStringRegion() allowing to use Region.setBufferCapacity()
|
|
|
|
| |
than enough
|
| |
|
|
|
|
| |
early if not needed (track capacity); Align all VBORegion* buffer init/set/grow impl.
|
|
|
|
| |
addOutlineShape1() (slow perf+debug), rename growBufferSize() -> growBuffer()
|
|
|
|
| |
buffer data-type to directly put[34][sif](..) skipping GLArrayDataClient/Buffers buffer-growth and validations
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
used for getElemCount() instead of 0==position, ... (API change)
API Change
- sealed() moved up from GLArrayDataEditable -> GLArrayData
- GLArrayDataWrapper is sealed by default
- getSizeInBytes() -> getByteCount()
- Semantics of getElemCount() and getByteCount()
- Correctly use sealed() to switch from position to limit - instead of 0==position
Aligned method names:
- getElemCount()
- elemPosition()
- remainingElems()
- getElemCapacity()
to corresponding byte counts:
- getByteCount()
- bytePosition()
- remainingBytes()
- getByteCapacity()
|
|
|
|
| |
but recommended)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
performance-hit measuring performance.
This was mostly notable on a Raspberry-Pi 4 arm64, where perfromance degragated around 3x using high-freq counter.
Using our well determined Clock.currentNanos() removes this overhead,
back to 'easy measuring' and having a well defined 'currentNanos()' since module start.
TestTextRendererNEWT00 can enable Region and Font perf-counter w/ '-perf',
w/o it only uses its own counter and hence reduce the high-freq burden (64% perf win on raspi4).
+++
Below numbers show that Region.addOutlineShape() perhaps needs a little performance work
to allow long text to be processed in 'real time' on embedded platform.
Hower, usually we cache the Region for long text and can have at least one liner
to be renderer within 60fps fast, i.e. Region produced in ~26ms for a 81 char line
instead of ~130ms for 664 chars.
+++
Raspberry Pi 4b, OpenJDK17, Debian 11:
Using current medium sized text_1 w/ 664 chars, w/o '-perf'
and after having passed 40 frames, we have following durations:
- process the OutlineShape -> Region: 129ms (text)
- Render the Region: 53ms
Startup Times:
- loading GlueGen - loading test 0 [ms]
- loading GlueGen - start test 1,910 [ms]
- loading test - start test 1,910 [ms]
- loading test - gl 2,631 [ms]
- loading test - graph 2,636 [ms]
- loading test - txt 2,844 [ms]
- loading test - draw 3,062 [ms]
Perf ..
1 / 1: Perf Launch: Total: graph 5, txt 207, draw 218, txt+draw 425 [ms]
1 / 1: Perf Launch: PerLoop: graph 5,505,740, txt 207,530,736, draw 218,393,680, txt+draw 425,924,416 [ns]
20 / 20: Perf Frame20: Total: graph 16, txt 376, draw 281, txt+draw 657 [ms]
20 / 20: Perf Frame20: PerLoop: graph 807,055, txt 18,820,824, draw 14,075,146, txt+draw 32,895,970 [ns]
20 / 40: Perf Frame40: Total: graph 3, txt 129, draw 53, txt+draw 182 [ms]
20 / 40: Perf Frame40: PerLoop: graph 176,670, txt 6,451,330, draw 2,658,217, txt+draw 9,109,547 [ns]
+++
On a modern desktop (~2y old), GNU/Linux Debian 11, AMD GPU on Mesa3D:
Using current medium sized text_1 w/ 664 chars, w/o '-perf'
and after having passed 40 frames, we have following durations:
- process the OutlineShape -> Region: 42ms (text)
- Render the Region: 5ms
Startup Times:
- loading GlueGen - loading test 0 [ms]
- loading GlueGen - start test 310 [ms]
- loading test - start test 309 [ms]
- loading test - gl 459 [ms]
- loading test - graph 460 [ms]
- loading test - txt 490 [ms]
- loading test - draw 506 [ms]
Perf ..
1 / 1: Perf Launch: Total: graph 1, txt 29, draw 15, txt+draw 45 [ms]
1 / 1: Perf Launch: PerLoop: graph 1,191,096, txt 29,868,436, draw 15,519,445, txt+draw 45,387,881 [ns]
20 / 20: Perf Frame20: Total: graph 240, txt 68, draw 21, txt+draw 89 [ms]
20 / 20: Perf Frame20: PerLoop: graph 12,045,651, txt 3,415,402, draw 1,069,348, txt+draw 4,484,750 [ns]
20 / 40: Perf Frame40: Total: graph 283, txt 42, draw 5, txt+draw 47 [ms]
20 / 40: Perf Frame40: PerLoop: graph 14,152,395, txt 2,116,114, draw 265,292, txt+draw 2,381,406 [ns]
|
|
|
|
| |
Clock.getMonotonicTime() ...
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
indices growBufferSize(); Add GLRegion.create(..) w/ initial vertices/indices count; Up default[VI]Count;
Following heuristcs were found, hence we might want to calculate these for each font (TODO):
/**
* Heuristics with TestTextRendererNEWT00 text_1 + text_2 = 1334 chars
* - FreeSans ~ vertices 64/char, indices 33/char
* - Ubuntu Light ~ vertices 100/char, indices 50/char
* - FreeSerif ~ vertices 115/char, indices 61/char
*
* Now let's assume a minimum of 10 chars will be rendered
*/
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
variants for flexibility/performance
Notable: The array-put is slower than small range single-puts, e.g. put3i(..).
Uses GlueGen's Buffers change commit 69b748925038b7d44fa6318536642b426e3d3e38
|
|
|
|
| |
createDummyDrawable(..)
|
|
|
|
| |
is not orig-owner
|
|
|
|
|
|
| |
dropping its usage (GLArrayDataWrapper validation)
Skip GLProfile based index, comps, type validation, might not be future proof.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Ease GLArrayData* buffer growth.
Using integer indices, i.e. GL_UNSIGNED_INT, requires us to pass a GLProfile 'hint' to the GLRegion ctor.
Region.max_indices is computed in this regard and used in Region.addOutlineShape().
TODO: If exceeding max_indices, the code path needs some work.
Buffer growth is eased via GLArrayData using its golden growth ratio
and manually triggering growth before processing all triangles in Region.addOutlineShape().
+++
TextRegionUtil static drawText() won't clear passed Region anymore, caller has to do this if so intended.
|
|
|
|
| |
additional components ...
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
finalize immutables, add growthFactor (default golden ratio 1.618), add getCapacity*() and printStats(..)
The growthFactor becomes essential for better growth behavior and can be set via setGrowthFactor().
The other changes were merely to clean up the GLArrayData interface and its 4 implementations.
Not great to change its API, but one name was misleading ['getComponentCount' -> 'getCompsPerEleme'],
so overall .. readability is enhanced.
Motivation for this change was the performance analysis and improvement of our Graph Curve Renderer.
|
|
|
|
|
|
| |
(fontName, text) for equals
Otherwise we would need to use a mostly collision free secure hash algo, Black2b-512 or sha256/512
|
|
|
|
| |
(redundant), add getLeftSideBearings*() from htmx table; Flatten TypecastGlyph impl by merging its Metrics, add equal() complementing hashCode()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
mapping, rely on no-shape for 'space' non-contour.
Drop erroneous Glyph ID_SPACE, ID_CR to rely on no-shape for 'space' non-contour
resolves different cmap-mappings of fonts, not following 'some std'.
Hence getGlyph(glyph_id) no more uses the `font.getGlyph(Glyph.ID_UNKNOWN)` shape,
but a null-shape as intended and using the hmtx and hhea table values for asvance and bounds.
This fixes 'space' spacing in general and specifically FreeSerif-Regular and the like.
This path also simplifies processing/layout of glyphs in process(..) and get*BoundsFU(..).
|
|
|
|
| |
lineGap, document them properly and fix Font.getLineHeightFU()
|
| |
|
| |
|
|\
| |
| | |
Add missing case in getDbgSeverityString()
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
rules for OutlineShape and add get/setWinding in Outline
Loop.initFromPolyline()'s Winding determination used a 3-point triangle-area method,
which is insufficent for complex shapes like serif 'g' or 'æ'.
Solved by using the whole area over the Outline shape.
Note: Loop.initFromPolyline()'s Winding determination is used to convert
the inner shape or holes to CW only.
Therefor the outter bondary shapes must be CCW.
This details has been documented within OutlineShape, anchor 'windingrules'.
Since the conversion of 'CCW -> CW' for inner shapes or holes is covered,
a safe user path would be to completely create CCW shapes.
However, this has not been hardcoded and is left to the user.
Impact: Fixes rendering serif 'g' or 'æ'.
The enhanced unit test TestTextRendererNEWT01 produces snapshots for all fonts within FontSet01.
While it shows proper rendering of the single Glyphs it exposes another Region/Curve Renderer bug,
i.e. sort-of a Region overflow crossing over from the box-end to the start.
|
| |
| |
| |
| | |
dropping redundant getMetricWidth*(); Fix getMetricBoundsFU()
|
| |
| |
| |
| | |
optional AffineTransform (see Label0)
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
processString(..) to Font, cleaning up ..
Further having Font.processString() return the AABBox of the whole string's 'laidout' OutlineShapes,
which is used for (debug) Font.getPointsBounds2(..) just to validate the coordinated with
the Glyph based AABBox of Font.getPointsBounds(..).
Static TextRegionUtil.drawString(..) no more require to pass the temp AffineTransform instances (ugly).
|
| | |
|
| |
| |
| |
| | |
getPointsBoundsFU(); Glyph: Drop getSymbol()
|
|/
|
|
| |
plane.
|
|
|
|
|
|
|
|
|
| |
system.
- All pixelSize metrics methods are dropped in Font*
- TypecastGlyph.Advance dropped, i.e. dropping prescales glyph advance based on pixelSize
- TextRegionUtil produces OutlineShape in font em-size [0..1] added to GLRegion
- Adjusted demos
|
|
|
|
|
|
|
|
|
| |
Use TestTextRendererNEWT01 to produce validation snaps
- Move kerning handling from Font to Font.Glyph using binary-search for right-glyph-id on kerning subset from this instance (left)
- TextRegionUtil: Kerning must be added before translation as it applies before the current right-glyph.
- TestTextRendererNEWT01 produces validation snapshots against LibreOffice print-preview snapshots
- GPUTextRendererListenerBase01 added another text for kerning validation, show more font-size details (pt, px, mm)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
original font-units (FU) to have them scaled later ( fu * pixelScale / unitsPerEM )
Scaling from font-units (funits, or FU) is now performed by the renderer itself,
i.e. applying the scale-factor 'fontPixelSize / font.getMetrics().getUnitsPerEM()'
to the PMV matrix to render the whole graph GLRegion.
This finally provides proper device and resolution independent font utilization.
Further, preliminary kerning has been added.
+++
Also ...
TypecastFont:
- getGlyphID(..) getGlyph(..) enforce symbol mapping to Glyph.ID_SPACE Glyph.ID_CR,
as some fonts ave an erroneous cmap (FreeSerif-Regular)
- add getKerning(..)
TODO: Add binary search
- Set TypecastFont.USE_PRESCALED_ADVANCE := false,
i.e. dropping all prescaled pixel-sized advance values mapped to font pixel-size
as we utilize font-units only for later uniform scaling.
- Drop virtual getPixelSize() and add static FontScale.toPixels(..)
- Add fullString() for debugging purposed, including some font tables
|
|
|
|
|
|
| |
Add Path2F addPath(..), emphasize required Winding.CW
GPURegionGLListener01 used by TestRegionRendererNEWT01 covers Path2F CCW and CW (reverse add) methods.
|
|
|
|
| |
addVertex methods
|
| |
|
| |
|