 Sven/Sarah Barth
|
bc46d8f33a
* x86_64 Win64 callback implementation: fix building the argument array if the result is passed as a parameter
|
3 年之前 |
 Karoly Balogh
|
5a84342348
* rtl-console: fix ; in the previous commit
|
3 年之前 |
|
Karoly Balogh
|
d3de4b5951
* rtl-console: generalized the partial screen update feature which was added to Amiga. the function was renamed to UpdateScreenArea(). it's still an empty call on other platforms, that falls back to UpdateScreen() but the infrastructure is now there to support it.
|
3 年之前 |
 J. Gareth "Curious Kit" Moreton
|
f289f2694a
x86: Additions to OptPass2Movx to better synergise with new CMP optimisation under -O2
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
683a92bcc8
i386: Correction to GetIntRegisterBetween to ensure we only get 8-bit registers that we can actually encode
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
1da7ce46de
x86: New double CMP optimisation to remove a branch
|
3 年之前 |
 florian
|
781b2d0a80
* patch by Rika: second part of #39496, resolves #30496
|
3 年之前 |
|
florian
|
abc4a0a4e6
* patch by Rika: another improvement of cutils.newalignment, part of #39496
|
3 年之前 |
|
florian
|
c0979bbc0c
* TObject.InitInstance got very big so inlining has no advantage anymore, resolves #39494
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
cafd708b6d
Refactoring of OptPass2Movx to remove goto
|
3 年之前 |
|
Karoly Balogh
|
9894fe8082
* rtl-console: on Amiga added a custom screen update call, to do partial screen updates. doing a full update can be very slow on low-end systems (checking the entire screen buffer every time), so software tuned for these can now request partial updates, when the updated already is already known
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
22cd8d5d62
Fixed bug in MovxMovx2Movx optimisation that would specify a 64-bit destination instead of 32-bit one
|
3 年之前 |
|
Karoly Balogh
|
4fe5e53fcc
* rtl-console: on Amiga, remove the initial testing of the buffer for changes. it just slowed things down with frequent partial updates, especially on low-end processors with limited or no data cache
|
3 年之前 |
|
Karoly Balogh
|
cc7c328a7c
* rtl-console: slight refactor of how Amiga SysUpdateScreen() addresses the videobuf. this results in slightly more optimal code with less complex addressing, which helps low-end processors
|
3 年之前 |
|
Karoly Balogh
|
5264e48d90
* rtl-console: always compile Amiga KVM units in OBJFPC mode. This makes it easier to compile these units outside of this package
|
3 年之前 |
|
florian
|
ab969e0a9b
* patch by Rika: improve ctuils.newalign/align, part of #39496
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
b4c8c1da12
Overflow bug fixes to MovZX/SX optimisations when CMP instructions are encountered.
|
3 年之前 |
|
florian
|
6dbe71cd30
* TX86AsmOptimizer.OptPass1MOVXX should search only over other instructions if it works with registers only
|
3 年之前 |
|
florian
|
6147d6d8a0
* compilation with i386 fixed
|
3 年之前 |
 Michaël Van Canneyt
|
ebd7bb7cb9
* Create API description as Pascal
|
3 年之前 |
|
Michael Van Canneyt
|
bd77787712
* Fix from Дмитрий Ворошин, fix issue #39495
|
3 年之前 |
|
Michaël Van Canneyt
|
e6fe075a42
* Patch from Henrique Werlan, correctly fetch RTTI info for external classes (pas2js issue 38943)
|
3 年之前 |
|
Michaël Van Canneyt
|
a16449c54e
* Fix from Henrique Werlang, use safename when writing pascal code (pas2js issue 39196)
|
3 年之前 |
 Nikolay Nikolov
|
b00fdb7205
* ptcpas keyboard handling improvements for macOS and X11
|
3 年之前 |
|
Michaël Van Canneyt
|
8ee49ef930
* Patch from Werner Pamler to fix issue #39493
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
d083cc7247
New MovxAndTest2Test optimisation to mirror the regular MovAndTest2Test optimisation
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
5b4c104aaf
Massive overhaul to OptPass2Movx to favour operand shrinkage
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
d255ffba8b
Improved handling of signed sequences in OptPass2Movx
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
01e5f4855a
MovZX->MovSX optimisation
|
3 年之前 |
|
J. Gareth "Curious Kit" Moreton
|
4825d2d16c
New Movz ###,%ecx, shift/rotate %cl,... optimisation
|
3 年之前 |
