 J. Gareth "Curious Kit" Moreton
|
b804fc5164
* Flags specific to TAddNode have been moved to their own field
|
il y a 1 an |
|
J. Gareth "Curious Kit" Moreton
|
2a44ffe51b
* x86: BZHI optimisation now sets the subtract and shift nodes to 'do not execute' as they are skipped over.
|
il y a 2 ans |
|
J. Gareth "Curious Kit" Moreton
|
16bd996e74
* x86: 'and not' optimisation now sets the NOT node to 'do not execute' as it is skipped over
|
il y a 2 ans |
|
J. Gareth "Curious Kit" Moreton
|
7da9b4a988
* x86: The code generator will now attempt to manipulate "x and ((1 shl y) - 1)" to use BZHI
|
il y a 2 ans |
|
J. Gareth "Curious Kit" Moreton
|
a1ea1dd98e
* x86: Changed BMI2 check to BMI1 for ordinal ANDN optimisation
|
il y a 2 ans |
|
J. Gareth "Curious Kit" Moreton
|
626e23d89f
* x86: The code generator will now attempt manipulate "x and (not y)" (where x and y are ordinals) to use ANDN.
|
il y a 2 ans |
 florian
|
f097d1ea35
* use the bts instruction on x86 if possible, the reg,reg variant is better than a mov/shl/or sequence
|
il y a 2 ans |
|
J. Gareth "Curious Kit" Moreton
|
6f24c8b4ef
* x86: Code generation fixes where FLAGS
|
il y a 3 ans |
|
J. Gareth "Curious Kit" Moreton
|
9b4033fa80
Fast mod
|
il y a 3 ans |
|
J. Gareth "Curious Kit" Moreton
|
2dc0995067
- Bug fix to new ADD/SUB optimisation where conditions are concerned
|
il y a 3 ans |
|
florian
|
01d6358f93
* some fixes for expectloc
|
il y a 3 ans |
 Florian Klämpfl
|
e121a753e1
* mark temps as deallocated
|
il y a 3 ans |
 Yuriy Sydorov
|
b8befe3d1f
* Fixed a typo in the condition, though it seems this code path is not used now.
|
il y a 3 ans |
|
florian
|
c02b43fb93
* change fma cpu flag into an fpu flag
|
il y a 4 ans |
|
florian
|
e46b426ca1
+ tx86addnode.second_opvector makes use of AVX instructions
|
il y a 4 ans |
|
florian
|
637976e83f
* patch by Marģers to unify internal error numbers, resolves #37888
|
il y a 4 ans |
|
florian
|
043c3df292
* properly handle boolean64 xor boolean64 on i386
|
il y a 5 ans |
|
florian
|
b2ed667a86
* swap fpu stack correctly when moving data into xmm registers
|
il y a 5 ans |
|
florian
|
5f67d830e7
* better rules if a 3 op mul operation is tried
|
il y a 5 ans |
|
florian
|
b4de0b2e9c
* insert fwait after fcompp, resolves #37136
|
il y a 5 ans |
|
florian
|
97efb03ae6
* call x86 addordinal for non short boolean evaluated boolean nodes
|
il y a 5 ans |
|
florian
|
fca4d3315b
* let the operation do the type conversion for x87 floating point operations
|
il y a 5 ans |
|
florian
|
3c16324f80
* x86: avoid to put variables into registers if they are handled by the x87-FPU
|
il y a 5 ans |
|
florian
|
91f3f022fe
* MM registers which might not be changed can be still used in 3 op mathematical operations as source
|
il y a 5 ans |
|
florian
|
5f3050961f
* load left properly into a register in tx86addnode.second_addordinal, resolves #36287
|
il y a 5 ans |
|
florian
|
c8f746b881
* do not allocate an extra register for some integer operations if not needed
|
il y a 5 ans |
 Jonas Maebe
|
ce598c15ec
* factored out the conditions under which add nodes need to perform
|
il y a 6 ans |
|
Jonas Maebe
|
281b3ad276
* fix case completeness and unreachable code warnings in compiler that would
|
il y a 6 ans |
 pierre
|
9c90f593ab
Add global range check disable for i8086 cgcpu and x86 nx86add units
|
il y a 6 ans |
|
Jonas Maebe
|
4686f61002
* keep track of the temp position separately from the offset in references,
|
il y a 7 ans |
