hw/display/ati.c | 137 ++++++++++++- hw/display/ati_2d.c | 462 +++++++++++++++++++++++++++++------------- hw/display/ati_dbg.c | 9 + hw/display/ati_int.h | 26 ++- hw/display/ati_regs.h | 20 ++ 5 files changed, 509 insertions(+), 145 deletions(-)
This series implements HOST_DATA as a blit source enabling text rendering in
xterm under X.org with 2D acceleration.
The series builds up functionality incrementally:
* Patches 1-6: Bug fixes and register implementations
* Patches 7-14: Refactor of ati_2d_blt to decouple from ATIVGAState
* Patch 15: Scissor clipping implementation
* Patches 16-17: HOST_DATA register writes, color expansion, and
accumulator flush
Tested with xterm on X.org using the r128 driver built with
--disable-dri (MMIO mode).
Hardware tests can be found at:
https://codeberg.org/cjab/ati-tests/src/commit/32ed42625cb3f094d4a3d3ce3ad26cfa5f872ac2/tests/common/host_data.c
Changes from v10:
- Rename ati_{flush,finish}_host_data to ati_host_data_{flush,finish}
- Use sizeof accumulator when flushing SRC_COLOR datatype
- Simplify ati_host_data_finish by flushing and immediately ending the blit
- Combine HOST_DATA* and HOST_DATA_LAST handlers
- Reset host_data.next within ati_host_data_flush
Changes from v9:
- Remove unneeded FIXME comment in ati_2d_do_blt
- ati_set_dirty is no longer called when the blit exits early
- Drop patch 15 moving src bounds validation to ati_2d_blt
- Check src vram bounds using clipped values, only when host_data is inactive
- Return to 128-bit accumulator for simplicity
- Small tweak to color expansion to avoid a multiply
Changes from v8:
- Rename tmp_stride to tmp_stride_words
- Small formatting tweaks
- Add clipping debug logging
- Replace extract8 with BIT masking in expansion hot path
- Add SRC_COLOR support
- Fix SRC_COLOR value (was 0x20000, should have been 0x30000)
- Add partial SRC_MONO_FRGD support
- Src bound validation now only done on ROP3_SRCCOPY blits
- Other improved log messages
- Host data accumulator expanded from 128-bits to 256-bits based on hardware
testing. Real hardware flushes after 128-bits but early HOST_DATA_LAST writes
make use of the entire 256-bits to finish the blit.
- Added ati_finish_host_data to match hardware in the early HOST_DATA_LAST case
- Added active flag for host_data blits ignoring HOST_DATA writes
when no blit is active.
- Removed ati_host_data_reset. This is now handled either when flushing or a new
blit is started.
Changes from v7:
- Avoid potential issue with dst_offset calculation (found by Zoltan)
- Remove unnecessary newlines and other small formatting fixes
- Squash patch 15 into patches 13 and 14
- Note: Zoltan's suggestion to embed ATI2DCtx in ATIVGAState will be
revisited in a follow-up series
Changes from v6:
- Reduce churn by adding ctx_ struct and pointer
- Moved extraction of ati_set_dirty helper earlier in the series
- Moved register logging into setup_2d_blt_ctx
- Remove unnecessary comments
Changes from v5:
- Added previously standalone aperture size patch to series
- Introduced ATI2DCtx to hold all blit state (suggested by Zoltan)
- Made refactor patches more granular
- Simplified DP_GUI_MASTER_CNTL bit shifting
- Sorted clipping regs by offsets
- Removed dst_x/y update after blit (confirmed by hardware testing)
Changes from v4:
- Refactored ati_2d_blt to accept src and dst state
- Implemented host_data blits on top of the refactored ati_2d_blt
(suggested by Zoltan)
- Dropped separate field storage for DP_DATATYPE/DP_MIX (per Zoltan)
Changes from v3:
- New patch 1 fixing DP_DATATYPE/MIX register aliasing, this supersedes old patch 5
- Fix MSB/LSB HOST_DATA bit ordering, it is per-byte rather than per-word
- Fixed a missing break in register write handler
- Squashed and reorganized the dst/scissor helper patches (per Zoltan)
- Simplified register field constants and increment logic (per Zoltan)
- Tested with MorphOS and TVPaint (opaque bitmap fonts now work)
Changes from v2:
- Verified with hardware that clipping default bit latches
- Verified with hardware that pitch/offset default bits latch
(supersedes Zoltan's "ati-vga: Separate default control bit for
source")
- A new approach to HOST_DATA using a 128-bit accumulator instead of a
large buffer
- Fixed a longstanding bug in (DST/SRC)_PITCH reads that zeroed pitch value.
- Removed mask from the DP_GUI_MASTER_CNTL write, all fields are
important
- Created helpers for code shared between ati_2d_blt and
ati_host_data_flush
Changes from v1:
- Split monolithic patch into 7 logical patches as requested
- Integrate HOST_DATA blit into existing ati_2d_blt()
- Add fallback implementation for non-pixman builds
- Remove unnecessary initialization in realize (kept in reset only)
- Fixed DP_GUI_MASTER_CNTL mask to preserve GMC_SRC_SOURCE field
- Implement scissor clipping
Chad Jablonski (17):
ati-vga: Fix framebuffer mapping by using hardware-correct aperture
sizes
ati-vga: Fix DST_PITCH and SRC_PITCH reads
ati-vga: Read aliased values from DP_GUI_MASTER_CNTL
ati-vga: Latch src and dst pitch and offset on master_cntl default
ati-vga: Implement foreground and background color register writes
ati-vga: Add scissor clipping register support
ati-vga: Remove dst_x/y updates after blit
ati-vga: Consolidate dirty region tracking in ati_2d_blt
ati-vga: Remove src and dst stride mutation in ati_2d_blt
ati-vga: Use local variables for register values in ati_2d_blt
ati-vga: Introduce ATI2DCtx struct for 2D blit context
ati-vga: Extract setup_2d_blt_ctx from ati_2d_blt
ati-vga: Split ati_2d_do_blt from ati_2d_blt
ati-vga: Remove ATIVGAState param from ati_2d_do_blt
ati-vga: Implement scissor rectangle clipping for 2D operations
ati-vga: Implement HOST_DATA register writes
ati-vga: Implement HOST_DATA flush to VRAM
hw/display/ati.c | 137 ++++++++++++-
hw/display/ati_2d.c | 462 +++++++++++++++++++++++++++++-------------
hw/display/ati_dbg.c | 9 +
hw/display/ati_int.h | 26 ++-
hw/display/ati_regs.h | 20 ++
5 files changed, 509 insertions(+), 145 deletions(-)
--
2.52.0
Hi Chad,
On 3/3/26 03:47, Chad Jablonski wrote:
> This series implements HOST_DATA as a blit source enabling text rendering in
> xterm under X.org with 2D acceleration.
>
> The series builds up functionality incrementally:
>
> * Patches 1-6: Bug fixes and register implementations
> * Patches 7-14: Refactor of ati_2d_blt to decouple from ATIVGAState
> * Patch 15: Scissor clipping implementation
> * Patches 16-17: HOST_DATA register writes, color expansion, and
> accumulator flush
> Chad Jablonski (17):
> ati-vga: Fix framebuffer mapping by using hardware-correct aperture
> sizes
> ati-vga: Fix DST_PITCH and SRC_PITCH reads
> ati-vga: Read aliased values from DP_GUI_MASTER_CNTL
> ati-vga: Latch src and dst pitch and offset on master_cntl default
> ati-vga: Implement foreground and background color register writes
> ati-vga: Add scissor clipping register support
> ati-vga: Remove dst_x/y updates after blit
> ati-vga: Consolidate dirty region tracking in ati_2d_blt
> ati-vga: Remove src and dst stride mutation in ati_2d_blt
> ati-vga: Use local variables for register values in ati_2d_blt
> ati-vga: Introduce ATI2DCtx struct for 2D blit context
> ati-vga: Extract setup_2d_blt_ctx from ati_2d_blt
> ati-vga: Split ati_2d_do_blt from ati_2d_blt
> ati-vga: Remove ATIVGAState param from ati_2d_do_blt
> ati-vga: Implement scissor rectangle clipping for 2D operations
> ati-vga: Implement HOST_DATA register writes
> ati-vga: Implement HOST_DATA flush to VRAM
Zoltan asked me to look at your series [*] but unfortunately I'm
getting a build failure when building without PIXMAN starting
with patch #9:
../hw/display/ati_2d.c: In function ‘ati_2d_do_blt’:
../hw/display/ati_2d.c:185:13: error: unused variable ‘src_stride_words’
[-Werror=unused-variable]
185 | int src_stride_words = ctx->src_stride / sizeof(uint32_t);
| ^~~~~~~~~~~~~~~~
../hw/display/ati_2d.c:147:9: error: unused variable ‘dst_stride_words’
[-Werror=unused-variable]
147 | int dst_stride_words = ctx->dst_stride / sizeof(uint32_t);
| ^~~~~~~~~~~~~~~~
../hw/display/ati_2d.c:137:10: error: unused variable ‘use_pixman_blt’
[-Werror=unused-variable]
137 | bool use_pixman_blt = use_pixman & BIT(1);
| ^~~~~~~~~~~~~~
../hw/display/ati_2d.c:136:10: error: unused variable ‘use_pixman_fill’
[-Werror=unused-variable]
136 | bool use_pixman_fill = use_pixman & BIT(0);
| ^~~~~~~~~~~~~~~
I'm queueing the first 8 patches hoping it helps.
Also, please fix the checkpatch.pl coding style errors
before re-posting.
Regards,
Phil.
[*]
https://lore.kernel.org/qemu-devel/7f1df452-05c2-efa3-fdce-a5d987cfdcc8@eik.bme.hu/
On Sun, 8 Mar 2026, Philippe Mathieu-Daudé wrote: > Hi Chad, > > On 3/3/26 03:47, Chad Jablonski wrote: >> This series implements HOST_DATA as a blit source enabling text rendering >> in >> xterm under X.org with 2D acceleration. >> >> The series builds up functionality incrementally: >> >> * Patches 1-6: Bug fixes and register implementations >> * Patches 7-14: Refactor of ati_2d_blt to decouple from ATIVGAState >> * Patch 15: Scissor clipping implementation >> * Patches 16-17: HOST_DATA register writes, color expansion, and >> accumulator flush > > >> Chad Jablonski (17): >> ati-vga: Fix framebuffer mapping by using hardware-correct aperture >> sizes >> ati-vga: Fix DST_PITCH and SRC_PITCH reads >> ati-vga: Read aliased values from DP_GUI_MASTER_CNTL >> ati-vga: Latch src and dst pitch and offset on master_cntl default >> ati-vga: Implement foreground and background color register writes >> ati-vga: Add scissor clipping register support >> ati-vga: Remove dst_x/y updates after blit >> ati-vga: Consolidate dirty region tracking in ati_2d_blt >> ati-vga: Remove src and dst stride mutation in ati_2d_blt >> ati-vga: Use local variables for register values in ati_2d_blt >> ati-vga: Introduce ATI2DCtx struct for 2D blit context >> ati-vga: Extract setup_2d_blt_ctx from ati_2d_blt >> ati-vga: Split ati_2d_do_blt from ati_2d_blt >> ati-vga: Remove ATIVGAState param from ati_2d_do_blt >> ati-vga: Implement scissor rectangle clipping for 2D operations >> ati-vga: Implement HOST_DATA register writes >> ati-vga: Implement HOST_DATA flush to VRAM > > Zoltan asked me to look at your series [*] but unfortunately I'm > getting a build failure when building without PIXMAN starting > with patch #9: > > ../hw/display/ati_2d.c: In function ‘ati_2d_do_blt’: > ../hw/display/ati_2d.c:185:13: error: unused variable ‘src_stride_words’ > [-Werror=unused-variable] > 185 | int src_stride_words = ctx->src_stride / sizeof(uint32_t); > | ^~~~~~~~~~~~~~~~ > ../hw/display/ati_2d.c:147:9: error: unused variable ‘dst_stride_words’ > [-Werror=unused-variable] > 147 | int dst_stride_words = ctx->dst_stride / sizeof(uint32_t); > | ^~~~~~~~~~~~~~~~ > ../hw/display/ati_2d.c:137:10: error: unused variable ‘use_pixman_blt’ > [-Werror=unused-variable] > 137 | bool use_pixman_blt = use_pixman & BIT(1); > | ^~~~~~~~~~~~~~ > ../hw/display/ati_2d.c:136:10: error: unused variable ‘use_pixman_fill’ > [-Werror=unused-variable] > 136 | bool use_pixman_fill = use_pixman & BIT(0); > | ^~~~~~~~~~~~~~~ > > I'm queueing the first 8 patches hoping it helps. > > Also, please fix the checkpatch.pl coding style errors > before re-posting. Sorry, my mistake I did not notice this during review and testing. I'm fixing it up and will send updated version of the last part of the series not yet queued. Thanks a lot for your help. Regards, BALATON Zoltan
On Mon, 9 Mar 2026, BALATON Zoltan wrote: > On Sun, 8 Mar 2026, Philippe Mathieu-Daudé wrote: >> Hi Chad, >> >> On 3/3/26 03:47, Chad Jablonski wrote: >>> This series implements HOST_DATA as a blit source enabling text rendering >>> in >>> xterm under X.org with 2D acceleration. >>> >>> The series builds up functionality incrementally: >>> >>> * Patches 1-6: Bug fixes and register implementations >>> * Patches 7-14: Refactor of ati_2d_blt to decouple from ATIVGAState >>> * Patch 15: Scissor clipping implementation >>> * Patches 16-17: HOST_DATA register writes, color expansion, and >>> accumulator flush >> >> >>> Chad Jablonski (17): >>> ati-vga: Fix framebuffer mapping by using hardware-correct aperture >>> sizes >>> ati-vga: Fix DST_PITCH and SRC_PITCH reads >>> ati-vga: Read aliased values from DP_GUI_MASTER_CNTL >>> ati-vga: Latch src and dst pitch and offset on master_cntl default >>> ati-vga: Implement foreground and background color register writes >>> ati-vga: Add scissor clipping register support >>> ati-vga: Remove dst_x/y updates after blit >>> ati-vga: Consolidate dirty region tracking in ati_2d_blt >>> ati-vga: Remove src and dst stride mutation in ati_2d_blt >>> ati-vga: Use local variables for register values in ati_2d_blt >>> ati-vga: Introduce ATI2DCtx struct for 2D blit context >>> ati-vga: Extract setup_2d_blt_ctx from ati_2d_blt >>> ati-vga: Split ati_2d_do_blt from ati_2d_blt >>> ati-vga: Remove ATIVGAState param from ati_2d_do_blt >>> ati-vga: Implement scissor rectangle clipping for 2D operations >>> ati-vga: Implement HOST_DATA register writes >>> ati-vga: Implement HOST_DATA flush to VRAM >> >> Zoltan asked me to look at your series [*] but unfortunately I'm >> getting a build failure when building without PIXMAN starting >> with patch #9: >> >> ../hw/display/ati_2d.c: In function ‘ati_2d_do_blt’: >> ../hw/display/ati_2d.c:185:13: error: unused variable ‘src_stride_words’ >> [-Werror=unused-variable] >> 185 | int src_stride_words = ctx->src_stride / sizeof(uint32_t); >> | ^~~~~~~~~~~~~~~~ >> ../hw/display/ati_2d.c:147:9: error: unused variable ‘dst_stride_words’ >> [-Werror=unused-variable] >> 147 | int dst_stride_words = ctx->dst_stride / sizeof(uint32_t); >> | ^~~~~~~~~~~~~~~~ >> ../hw/display/ati_2d.c:137:10: error: unused variable ‘use_pixman_blt’ >> [-Werror=unused-variable] >> 137 | bool use_pixman_blt = use_pixman & BIT(1); >> | ^~~~~~~~~~~~~~ >> ../hw/display/ati_2d.c:136:10: error: unused variable ‘use_pixman_fill’ >> [-Werror=unused-variable] >> 136 | bool use_pixman_fill = use_pixman & BIT(0); >> | ^~~~~~~~~~~~~~~ >> >> I'm queueing the first 8 patches hoping it helps. >> >> Also, please fix the checkpatch.pl coding style errors >> before re-posting. > > Sorry, my mistake I did not notice this during review and testing. I'm fixing > it up and will send updated version of the last part of the series not yet > queued. Thanks a lot for your help. I sent updated v12 fixing the above. Before that I've also sent this small fix independent of this series: https://patchew.org/QEMU/cover.1773009887.git.balaton@eik.bme.hu/ Thank you, BALATON Zoltan
© 2016 - 2026 Red Hat, Inc.