Jonas Maebe
|
246caf3526
* never alias/coalesce cpu registers that are not usable by the register
|
11 anni fa |
florian
|
23c724f885
* prevent a temp. register allocated during spilling being immediately spilled again, resolves #25164
|
11 anni fa |
florian
|
829764e96b
+ replace registers to spill in mov instructions and convert them into ldr/str
|
11 anni fa |
Károly Balogh
|
587acf6452
fix infinite spilling on m68k, patch by Florian, additional IE workaround by me, based on a similar hack in the ARM cg...
|
11 anni fa |
Károly Balogh
|
70f9b098e0
no longer define a segment field for m68k
|
11 anni fa |
sergei
|
d2a9308181
+ SPARC: implemented register spill replacement.
|
11 anni fa |
svenbarth
|
32a18199c5
Print aX or aregX for address registers if DEBUG_REGISTERLIFE is defined.
|
12 anni fa |
florian
|
dc567acc9a
* accumulate register weight instead of taking only the maximum
|
12 anni fa |
florian
|
c2f7ccb253
* coalesce more aggressive
|
12 anni fa |
sergei
|
8b8553991a
+ MIPS: prevent coalescing written-to registers with $sp,$fp,$zero and $at.
|
12 anni fa |
florian
|
23c84fd27c
* more fixes for segment register usage
|
12 anni fa |
florian
|
b85addb451
+ color segment register
|
12 anni fa |
svenbarth
|
d15304c25e
Also check for the type of the register when replacing it. In certain circumstances this can still lead to access to invalid memory resulting either in an access violation or invalid registers.
|
12 anni fa |
pierre
|
322b793506
Try to do something for m68k integer/address registers; not working yet :!(
|
12 anni fa |
svenbarth
|
2ada9a528b
Fix a critical bug in the register allocator (at least for CPUs with seperate address
|
12 anni fa |
florian
|
95830d5a4a
+ DEBUG_REGISTERLIFE define
|
13 anni fa |
florian
|
f75e39e660
+ aliasing of registers, allows to split live ranges
|
13 anni fa |
florian
|
f9dd5bf5e6
+ live_start and live_end_properties
|
13 anni fa |
florian
|
c2d067940a
* get rid of hash chars, modern tools and IDEs can use normal comments as documentation if needed
|
13 anni fa |
florian
|
b78f9826fb
* call AllocMem instead of GetMem/FillChar
|
13 anni fa |
florian
|
ecdec2e431
* disable reg. allocator optimization introduced in r21812 for now because it caused some trouble, see #22405
|
13 anni fa |
masta
|
f9cdf3d4ca
Be more specific in spilling debug messages
|
13 anni fa |
masta
|
aef81cad68
Add debug information about register spilling
|
13 anni fa |
florian
|
9833f642fe
* coalesce if either u or v are precolored, tested on arm and i386, it gives for both a small advantage
|
13 anni fa |
Jonas Maebe
|
0fc422f244
* moved definition of maxcpuregister and tcpuregisterset from cgbase to
|
13 anni fa |
Jonas Maebe
|
6f336a5d13
* the register allocator always knows the exact number of the first imaginary
|
13 anni fa |
Jonas Maebe
|
aee5380ae0
* merged trunk up to r20882
|
13 anni fa |
florian
|
6211a83865
* write variable locations of 64 bit integer values on 32 bit targets correctly
|
14 anni fa |
Jonas Maebe
|
00cb8f5725
* made maxregs related information protected instead of private in trgobj,
|
14 anni fa |
florian
|
68dd23b5a2
* write proper register locations when doing ssa, resolves #7552
|
14 anni fa |