| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | * Make trgobj.uses_registers method return True when registers of appropriate... | sergei | 2015-12-07 | 1 | -2/+6 |
| * | * fix wrong comparisons in trgobj.conservative to enable better coalescing, | florian | 2015-08-28 | 1 | -5/+6 |
| * | * check registers of ait_regalloc for a valid color when translating and comp... | florian | 2015-04-05 | 1 | -1/+7 |
| * | * support top_para in trgllvm.get_spill_temp() | jonas | 2015-04-04 | 1 | -2/+2 |
| * | * factored out spilling code that deals with operand types into virtual | jonas | 2015-04-04 | 1 | -122/+139 |
| * | * don't crash if we have to spill a register of the last instruction in an | jonas | 2015-04-01 | 1 | -0/+2 |
| * | * synchronized with privatetrunk till r30095 | jonas | 2015-03-05 | 1 | -0/+3 |
| |\ |
|
| | * | + New subtype of tai_regalloc, "ra_markused". It marks a physical register as... | sergei | 2015-02-25 | 1 | -0/+3 |
| * | | * synchronised with trunk up till r28402 | jonas | 2014-08-13 | 1 | -2/+2 |
| |\ \
| |/ |
|
| | * | + m68k: initial implementation of do_spill_replace (supports only MOVE instru... | sergei | 2014-07-30 | 1 | -2/+2 |
| * | | * synchronized with trunk up to r27758 | jonas | 2014-05-12 | 1 | -3/+18 |
| |\ \
| |/ |
|
| | * | * never alias/coalesce cpu registers that are not usable by the register | jonas | 2014-03-11 | 1 | -3/+18 |
| * | | * synchronised with trunk up till r26975 | jonas | 2014-03-06 | 1 | -14/+27 |
| |\ \
| |/ |
|
| | * | * prevent a temp. register allocated during spilling being immediately spille... | florian | 2014-03-02 | 1 | -4/+14 |
| | * | + replace registers to spill in mov instructions and convert them into ldr/str | florian | 2014-02-05 | 1 | -2/+2 |
| | * | fix infinite spilling on m68k, patch by Florian, additional IE workaround by ... | karoly | 2014-01-15 | 1 | -2/+5 |
| | * | no longer define a segment field for m68k | karoly | 2014-01-15 | 1 | -6/+6 |
| | * | + SPARC: implemented register spill replacement. | sergei | 2014-01-03 | 1 | -2/+2 |
| * | | + llvm support for the register allocator. While llvm works with virtual | jonas | 2013-11-11 | 1 | -52/+86 |
| * | | + added extra "orgsupreg" parameter to do_spill_read/do_spill_written/ | jonas | 2013-11-11 | 1 | -16/+18 |
| * | | * moved the register colouring and spill temp allocation to a virtual method, | jonas | 2013-11-11 | 1 | -20/+38 |
| |/ |
|
| * | Print aX or aregX for address registers if DEBUG_REGISTERLIFE is defined. | svenbarth | 2013-10-05 | 1 | -3/+3 |
| * | * accumulate register weight instead of taking only the maximum | florian | 2013-09-07 | 1 | -2/+2 |
| * | * coalesce more aggressive | florian | 2013-07-26 | 1 | -1/+1 |
| * | + MIPS: prevent coalescing written-to registers with $sp,$fp,$zero and $at. | sergei | 2013-06-15 | 1 | -3/+6 |
| * | * more fixes for segment register usage | florian | 2013-05-24 | 1 | -0/+10 |
| * | + color segment register | florian | 2013-05-24 | 1 | -0/+12 |
| * | Also check for the type of the register when replacing it. In certain circums... | svenbarth | 2012-10-19 | 1 | -3/+6 |
| * | Try to do something for m68k integer/address registers; not working yet :!( | pierre | 2012-10-19 | 1 | -6/+44 |
| * | Fix a critical bug in the register allocator (at least for CPUs with seperate... | svenbarth | 2012-10-18 | 1 | -2/+5 |
| * | + DEBUG_REGISTERLIFE define | florian | 2012-09-15 | 1 | -2/+6 |
| * | + aliasing of registers, allows to split live ranges | florian | 2012-09-15 | 1 | -9/+39 |
| * | + live_start and live_end_properties | florian | 2012-08-17 | 1 | -2/+31 |
| * | * get rid of hash chars, modern tools and IDEs can use normal comments as doc... | florian | 2012-08-17 | 1 | -11/+11 |
| * | * call AllocMem instead of GetMem/FillChar | florian | 2012-08-15 | 1 | -2/+2 |
| * | * disable reg. allocator optimization introduced in r21812 for now because it... | florian | 2012-08-08 | 1 | -1/+2 |
| * | Be more specific in spilling debug messages | masta | 2012-07-12 | 1 | -2/+2 |
| * | Add debug information about register spilling | masta | 2012-07-08 | 1 | -0/+6 |
| * | * coalesce if either u or v are precolored, tested on arm and i386, it gives ... | florian | 2012-07-07 | 1 | -2/+7 |
| * | * moved definition of maxcpuregister and tcpuregisterset from cgbase to | jonas | 2012-06-15 | 1 | -1/+1 |
| * | * the register allocator always knows the exact number of the first imaginary | jonas | 2012-06-15 | 1 | -1/+3 |
| * | * merged trunk up to r20882 | jonas | 2012-04-15 | 1 | -0/+26 |
| |\ |
|
| | * | * write variable locations of 64 bit integer values on 32 bit targets correctly | florian | 2011-09-03 | 1 | -10/+20 |
| | * | * write proper register locations when doing ssa, resolves #7552 | florian | 2011-08-19 | 1 | -0/+16 |
| * | | * made maxregs related information protected instead of private in trgobj, | jonas | 2011-08-20 | 1 | -4/+5 |
| |/ |
|
| * | * Fixed spell error revealed by lintian. | mazen | 2010-10-06 | 1 | -3/+3 |
| * | * getregisterinline() can be used to replace multiple size variants | jonas | 2010-09-08 | 1 | -6/+20 |
| * | * changed backwards_was_first and extended_backwards from supregsets into | jonas | 2010-07-18 | 1 | -12/+19 |
| * | * disabled superfluous zeroing of another field | jonas | 2010-07-05 | 1 | -3/+3 |
| * | * commented out a number of superfluous zero-initialisations of (large) | jonas | 2010-07-03 | 1 | -18/+21 |
