> From: Atari Ace [mailto:atari_ace@verizon.net]
> Sent: Sunday, September 27, 2009 7:58 AM
> To: submit@mamedev.org
> Cc: atariace@hotmail.com
> Subject: [patch] More _NAME macros
>
> Hi mamedev,
>
> MAME's idiom for function/data macros is to first implement
> <name>_NAME, then implement the other macros in terms of the _NAME
> macro. Then in principle only a single line needs editing to change
> the naming convention.
>
> This patchset implements this idiom more completely. The first patch
> adds some missing _NAME macros and fixes cases in source files that
> should be using the macros. The second patch then changes header
> files where the macros should have been used, but weren't. This
> required changing the idiom for removing a machine driver function
> pointer from MDRV_<FUNCTION>(NULL) to MDRV_<FUNCTION>(0), to avoid
> problems with NULL being macro expanded. This actually unifies the
> handling of all such cases, as we already had ipt_0 and driver_init_0.
> It also required reworking the devtempl.h implementation in a way that
> triggered a warning on MSVC about using empty macros, so vconv.c
> needed to be updated. The third patch then renames all the _NAME and
> _0 macros to verify that all the cases have been covered, so it isn't
> intended to be applied.
>
> ~aa
to be working reliably.
Deprecated the ROMREGION_DISPOSE flag, as 98% of the use of it no
longer is applicable with on-the-fly decoding, and the remaining
cases are difficult to identify among the others.
gfx_element. Updated the drivers that did this to use the new function, fixing
random crashes.
Fixed a couple of other minor regressions with recent drawgfx changes.
- Added built-in dirty tile tracking to the gfx_element. This removes
the need for all drivers that had dynamically populated graphics
to do their own dirty tracking. Tiles are marked dirty via the
new function gfx_element_mark_dirty(). Any driver that needs access
to the decoded data must call gfx_element_get_data() in order to
ensure that the referenced tile is clean before proceeding.
- In order to support dirty tracking, the gfx_element was enhanced to
keep track of the original source pointer, so that it can go back
and regenerate tiles on demand. For systems that set NULL for the
region in the gfxdecode, they must use gfx_element_set_source()
to specify a pointer to the raw data before drawing anything.
- Changed allocgfx() to gfx_element_alloc(), and added parameters to
specify the source data pointer, base color index, and total colors.
Many drivers had to whack these values in after the fact, so this
allowed for some minor additional cleanup.
- Added a dirtyseq member to the gfx_element struct. This is
incremented on each tile dirty, and can be used to sniff if
something has changed.
- Added logic in the tilemap engine to track which gfx_elements are
used for a given tilemap, and automatically detect changes to the
tiles so that drivers no longer have to explicitly invalidate the
tilemap when tiles change. In the future, this may grow smarter to
only invalidate the affected tiles, but for now it invalidates the
entire tilemap.
- Updated a number of drivers to remove their own dirty handling and
leverage the new internal dirty marking.
- Because the source data must always be present, updated the atarigen
zwackery and mystwarr graphics handing code to support this.
- Thanks to the dirty tracking, this actually allows all gfx decoding
to happen on the fly instead of all at once up front. Since there
was some concern that this would cause undesirable behavior due to
decoding lots of tiles on the fly, it is controlled with a compile-
time constant in mame.h (PREDECODE_GFX). Set this to 1 to get the
old behavior back.
- Moved decodechar() and decodegfx() to deprecat.h. All drivers in MAME
have been updated to simply mark tiles dirty and let the rendering
system decode them as needed, so these functions may go away in the
future.
- Rewrote entirely the rendering code in drawgfx. This code previously
used extensive recursive #includes and tricks to build, and was
very difficult to understand. The new code is based off of a set of
macros defined in drawgfxm.h. These new macros separate the core
rendering logic from the per-pixel operation, allowing the operation
to be easily "plugged" into any of the renderers. These macros are
also available to any driver that wants custom rendering behavior
that is similar to existing core behavior, without needing to
populate the core with esoteric one-off rendering behaviors.
- Added a set of new functions for [p]drawgfx[zoom], one for each
transparency type. The old [p]drawgfx[zoom] functions are still
present, but now switch off the transparency type and call through
to one of these new transparency-specific functions. The old
functions are also now reduced to only supporting TRANSPARENCY_NONE,
TRANSPARENCY_PEN, and TRANSPARENCY_PENS. All other rendering types
must use the new functions.
- All new rendering functions have extensive asserts to catch improper
clipping rectangles and other common errors.
- All new rendering functions automatically downgrade to optimized
versions where appropriate. For example, calling drawgfx_transpen
with an out-of-range pen automatically falls back to drawgfx_opaque.
And drawgfxzoom_* with xscale=yscale=1.0 automatically falls back
to drawgfx_*. And many other examples. In general, this relieves
drivers from needing to make these sorts of decisions.
- All new rendering functions have a consistent parameter order that
is a bit different from the existing functions. The cliprect
parameter is now specified immediately after the destination bitmap,
to match the convention used throughout the rest of the system.
The core parameters are followed by the scale parameters (for the
zoom functions), and then followed by the priority parameters (for
the pdrawgfx* functions), finally followed by any PIXEL_OP*-specific
parameters (such as transparent pen, alpha, drawing tables, etc.)
- Removed drawgfx_alpha_cache, alpha_set_level(), and the inline
functions alpha_blend16() and alpha_blend32(). To render graphics
with alpha, use the new [p]drawgfx[zoom]_alpha functions, which
take an explicit alpha value. To render tilemaps with alpha, the
TILEMAP_DRAW_ALPHA option now takes an explicit alpha parameter.
And to do you own alpha blending, use the alpha_blend_r16() and
alpha_blend_r32() functions, which take an explicit alpha.
- Updated a number of drivers as a result of removing the implicit
alpha in the drawgfx_alpha_cache.
- Removed drawgfx_pen_table and TRANSPARENCY_PEN_TABLE. To achieve
the same effect, build your own table and pass it to
[p]drawgfx[zoom]_transtable, along with a pointer to the
machine->shadow_table to use for shadows. Eventually
machine->shadow_table is likely to go away, and drivers will need
to fetch the shadow table from the palette directly.
- Updated a number of drivers to remove use of drawgfx_pen_table.
- Removed TRANSPARENCY_ALPHARANGE; it was only used by the psikyosh
driver, so it is now moved locally into that driver and built
using the macros in drawgfxm.h.
- Removed TRANSPARENCY_PEN_RAW; to achieve the same effect, call the
new [p]drawgfx[zoom]_transpen_raw() functions. Updated drivers to
make this change.
- Removed the unused mdrawgfx* functions entirely.
- Added new function gfx_element_set_source_clip() to specify a
source clipping rectangle for any element. This replaces the nasty
hacks that were being used in bnstars, ms32, namcos86, and namcos1
to achieve similar behaviors.
- Simplified the copyrozbitmap() functions to match the copybitmap()
functions in having separate opaque and transparent versions. Also
removed the 'priority' parameter which was only used by one driver,
and moved that logic into a custom renderer built using macros in
drawgfxm.h. Updated copyrozbitmap* to use the destbitmap, cliprect
parameter ordering convention as well.
- Simplified the draw_scanline*() functions to always render opaque.
Only one driver was doing otherwise, and it now does its work
internally (draw_scanline is dead-simple ever since we moved
rotation to the OSD code; I almost just removed it entirely).
Other changes:
- Added a cliprect to the bitmap_t type, which describes the full
bitmap.
- Removed tilemap_set_pen_data_offset; unfortunately, this adds a
random tile offset behind the scenes and goes against the dirty
tile detection and invalidation. Updated the mainsnk, snk, and
snk68 drivers to use old fashioned tile banking. (Sorry Nicola.)
- Changed zac2650 gfxdecode to use scale factors.
- Added function video_assert_out_of_range_pixels() to help find
the source of invalid pixels (generally out-of-range palette
entries due to invalid data or sloppy calculations). Place this
after each step in your rendering in a debug build to discover
which code is generating improper pixels.
Sent: Saturday, December 13, 2008 5:07 PM
To: submit@mamedev.org
Cc: atariace@hotmail.com
Subject: [patch] Add machine to allocgfx
Hi mamedev,
This patch eliminates the #include "deprecat.h" from drawgfx.h. It
does so in a fashion similar to my recent tilemap patch, adding the
machine pointer to gfx_element, changing allocgfx to take a machine,
and then adjusting the internals to use the machine field as needed.
The changes outside of drawgfx.[ch] were done with the attached
script.
~aa
integer value, regions are now referred to by a region class and
a region tag. The class specifies the type of region (one of CPU,
gfx, sound, user, disk, prom, pld) while the tag uniquely specifies
the region. This change required updating all the ROM region
definitions in the project to specify the class/tag instead of
region number.
Updated the core memory_region_* functions to accept a class/tag
pair. Added new memory_region_next() function to allow for iteration
over all memory regions of a given class. Added new function
memory_region_class_name() to return the name for a given CPU
memory region class.
Changed the auto-binding behavior of CPU regions. Previously, the
first CPU would auto-bind to REGION_CPU1 (that is, any ROM references
would automatically assume that they lived in the corresponding
region). Now, each CPU automatically binds to the RGNCLASS_CPU region
with the same tag as the CPU itself. This behavior required ensuring
that all previous REGION_CPU* regions were changed to RGNCLASS_CPU
with the same tag as the CPU.
Introduced a new auto-binding mechanism for sound cores. This works
similarly to the CPU binding. Each sound core that requires a memory
region now auto-binds to the RGNCLASS_SOUND with the same tag as the
sound core. In almost all cases, this allowed for the removal of the
explicit region item in the sound configuration, which in turn
allowed for many sound configurations to removed altogether.
Updated the expression engine's memory reference behavior. A recent
update expanded the scope of memory references to allow for referencing
data in non-active CPU spaces, in memory regions, and in EEPROMs.
However, this previous update required an index, which is no longer
appropriate for regions and will become increasingly less appropriate
for CPUs over time. Instead, a new syntax is supported, of the form:
"[tag.][space]size@addr", where 'tag' is an optional tag for the CPU
or memory region you wish to access, followed by a period as a
separator; 'space' is the memory address space or region class you
wish to access (p/d/i for program/data/I/O spaces; o for opcode space;
r for direct RAM; c/u/g/s for CPU/user/gfx/sound regions; e for
EEPROMs); and 'size' is the usual b/w/d/q for byte/word/dword/qword.
Cleaned up ROM definition flags and removed some ugly hacks that had
existed previously. Expanded to support up to 256 BIOSes. Updated
ROM_COPY to support specifying class/tag for the source region.
Updated the address map AM_REGION macro to support specifying a
class/tag for the region.
Updated debugger windows to display the CPU and region tags where
appropriate.
Updated -listxml to output region class and tag for each ROM entry.
longer used. Source needs to be recompiled because of the changed enum.
- Changed copybitmap and copyscrollbitmap:
There are now 2 versions of each, one without and with transparency:
void copybitmap(mame_bitmap *dest,mame_bitmap *src,int flipx,int flipy, int sx,int sy,const rectangle *clip);
void copybitmap_trans(mame_bitmap *dest,mame_bitmap *src,int flipx,int flipy, int sx,int sy,const rectangle *clip, pen_t transparent_pen);
void copyscrollbitmap(mame_bitmap *dest,mame_bitmap *src, nt rows,const int *rowscroll,int cols,const int *colscroll, const rectangle *clip);
void copyscrollbitmap_trans(mame_bitmap *dest,mame_bitmap *src, int rows,const int *rowscroll,int cols,const int *colscroll, const rectangle *clip, pen_t transparent_pen);
The version without _trans is the equivalent of the old TRANSPARENCY_NONE, The *_trans version is the equivalent
of the old TRANSPARENCY_PEN. The old TRANSPARENCY_COLOR mode is done via calling *_trans version and passing in
the pen that has been looked up via machine->pens[].
So for example, copybitmap(..., TRANSPARENCY_COLOR, 0) becomes
copybitmap_trans(..., machine->pens[0])
- Changed all drivers to the new calls. Suprising how few drivers still use these functions.
Most have still not been converted to tilemaps, or they are still writing to a tmpbitmap
which gets copied over to the real bitmap in VIDEO_UPDATE.
- Changed machine->screen[0].visarea to 'cliprect' where appropriate.
While investigating alternate gfx layout schemes, I stumbled across
the fact that some drivers are allocating graphics with one layout and
then decoding them with another (!). There's no guarantee this will
work, but for the drivers that do so (all Konami games), the layouts
are similar enough that it does. A related potential bug is that many
drivers are decoding using the layout provided to allocgfx, not the
layout attached the element returned from allocgfx. If the element
had scaling applied to it, this would be incorrect, but since scaling
is rare these are also benign. It would also be a problem if the
layout data had a different internal representation (which is
something I'm experimenting with), so to reduce the possibility of
coding errors and allow for future changes, I'd like to remove the
layout parameter from decodechar.
So here's two patches, the first fixes the affected Konami drivers to
allocate and decode using the same layouts. It changes the notion of
plane_order and bpp in the functions somewhat to let the start
routines select the appropriate layout, but acceptably so IMHO (I can
clean this up further if there are loud objections). The second patch
then removes the layout parameter from all the decodechar() calls.
I also reviewed MESS to see if it had similar problems and didn't find
any.
0. This patch does minor cleanup to existing layouts, trimming/padding
entries as appropriate and reformating a few layouts.
1. This patch introduces a GFXLAYOUT_RAW() macro, and uses it
throughout. It codifies the requirements for a raw layout in one
place.
2. This patch adds new validation code to the core for some previously
unchecked assumptions about layouts, and reduces the number of
references to the gfx_layout fields in preparation for a change in the
representation.
3. This patch constifies the remaining non-const gfx_layouts in MAME.
It does this by adjusting the code so that the only modification to a
layout ever needed is for the total field, which is then handled by
modifying a stack-based copy of the layout before invoking allocgfx. I
also spent some time consolidating and simplifying the layout code in
konamiic.c.
- removed years from copyright notices
- removed redundant (c) from copyright notices
- updated "the MAME Team" to be "Nicola Salmoria and the MAME Team"
This patch should complete the addition of static qualifiers to all
MAME symbols that aren't explicitly exported. It primarily handles
generated code (e.g. amspdwy.c), plus a handful of cases I'd
previously missed and some new cases introduced in the last update.
One interesting bit was the discovery that the 32-bit scanline
routines in drawgfx.c are unused. I debated eliminating them but
decided instead to just export them. Various internal drawgfx
functions were conditionally removed by examining a new RAW define,
although one routine (blockmove_8toN_alphaone) was determined to be
dead code.
While investigating constifying MESS, I came across a few core APIs
that were missing const qualifiers which this patch fixes. I also
consted up tx1.c while I was at it.
