- 05 Aug, 2009 1 commit
-
-
Matthias Braun authored
[r26317]
-
- 05 Jul, 2009 2 commits
-
-
Michael Beck authored
[r26239]
-
Michael Beck authored
- SymConst's are now ALWAYS placed in the start block [r26236]
-
- 19 Jun, 2009 1 commit
-
-
Matthias Braun authored
verify that Projs are in the same Block as their predecessors (and fix bugs in firm where this wasn't the case) [r26163]
-
- 15 May, 2009 1 commit
-
-
Matthias Braun authored
[r25967]
-
- 07 May, 2009 1 commit
-
-
Matthias Braun authored
- No exceptions for argument order in the Builtin constructor - add missing functions for Borrow node [r25888]
-
- 02 Mar, 2009 1 commit
-
-
Matthias Braun authored
[r25583]
-
- 27 Feb, 2009 1 commit
-
-
Michael Beck authored
and seems generally to be a good idea [r25572]
-
- 21 Feb, 2009 1 commit
-
-
Moritz Kroll authored
Renamed get_Cond_defaultProj() to get_Cond_default_proj() for consistency. Added set_Cond_default_proj() and get_cond_kind_name(). Made get_cond_jmp_predicate_name() consistent with other enum name functions. [r25545]
-
- 19 Jan, 2009 2 commits
-
-
Michael Beck authored
[r25264]
-
Michael Beck authored
[r25256]
-
- 09 Jan, 2009 1 commit
-
-
Michael Beck authored
[r25128]
-
- 29 Dec, 2008 1 commit
-
-
Michael Beck authored
[r24938]
-
- 26 Dec, 2008 1 commit
-
-
Michael Beck authored
The old code was fundamental broken, because it - changes the cls of a node WITHOUT changing a following be_Keep - produced the xmm0 reload BEFORE the Keep -> both nodes define xmm0 The new code handles some simple cases even more gracefully. [r24897]
-
- 24 Dec, 2008 4 commits
-
-
Christoph Mallon authored
[r24890]
-
Christoph Mallon authored
[r24889]
-
Michael Beck authored
this should safe a huge amount of useless calls [r24888]
-
Michael Beck authored
- implemented ir_bk_debugbreak [r24885]
-
- 23 Dec, 2008 6 commits
-
-
Michael Beck authored
[r24882]
-
Christoph Mallon authored
[r24881]
-
Christoph Mallon authored
[r24880]
-
Christoph Mallon authored
- BSF/BSR cannot do 16 bit AM in our model, because they only set the lower 16 bits of the destination register then. - Rename gen_unop_dest() to gen_unop_AM(). [r24876]
-
Michael Beck authored
[r24870]
-
Michael Beck authored
[r24867]
-
- 21 Dec, 2008 3 commits
-
-
Michael Beck authored
[r24832]
-
Michael Beck authored
[r24830]
-
Michael Beck authored
- add athlon_old for athlons WITHOUT SSE [r24829]
-
- 19 Dec, 2008 1 commit
-
-
Michael Beck authored
- fully implemented __builtin_(frame|return)_address in the backend [r24812]
-
- 18 Dec, 2008 2 commits
-
-
Michael Beck authored
[r24793]
-
Michael Beck authored
[r24785]
-
- 13 Dec, 2008 3 commits
-
-
Michael Beck authored
- fixed an FIXME [r24629]
-
Matthias Braun authored
some adjustment to float convs (skip some strictconvs when no_fold_const is not set, but always strict conv in int->float converstion) [r24617]
-
Matthias Braun authored
[r24598]
-
- 05 Dec, 2008 1 commit
-
-
Michael Beck authored
[r24329]
-
- 04 Dec, 2008 2 commits
-
-
Michael Beck authored
[r24298]
-
Michael Beck authored
- add support for Psi(Cmp(!mode_float), Const(mode_float), Const(mode_float)) [r24295]
-
- 03 Dec, 2008 1 commit
-
-
Michael Beck authored
[r24259]
-
- 01 Dec, 2008 1 commit
-
-
Christoph Mallon authored
Perform dark rituals and resurrect ProduceVal. We are not ready to dispose it, yet, because the Spiller is too clever for the Keep-trick. [r24182]
-
- 28 Nov, 2008 2 commits
-
-
Christoph Mallon authored
Remove ProduceVal. Replace it by Xor0 (to produce 0), Sbb0 (to produce 0/-1) and by Keep to make Cltd register pressure faithful. [r24128]
-
Matthias Braun authored
[r24112]
-