code-generator - Copy of https://svn.cs.ru.nl/repos/clean-code-generator/, where I implement Thumb-2 code generation.

Age	Commit message (Collapse)	Author	Files	Lines
2007-02-27	pass {#Int} and {#Real} using centry,	John van Groningen	1	-50/+177
	add {#Int} and {#Real} results for ccall
2007-02-27	add int_array_to_node_label and real_array_to_node_label	John van Groningen	1	-1/+3

2007-01-12	add centry support for strings	John van Groningen	1	-33/+127

2007-01-10	add passing reals using centry for IA32	John van Groningen	1	-32/+148

2006-12-14	store 64 bits in data section (instead of 32) for cvtsi2sdq instruction	John van Groningen	1	-1/+1

2006-12-01	fix bug in code generation for 64 bit version, if a real is loaded from a node	John van Groningen	1	-4/+13
	in a previous basic block and allocated in a general purpose register incorrect code was generated
2006-11-29	use dINT instead of INT for eq_desc_b to work around problem with the	John van Groningen	1	-1/+7
	microsoft assembler
2006-11-07	add makefile for 64 bit linux	John van Groningen	1	-0/+14

2006-09-18	fix error in rex prefix for move register to global variable and move global	John van Groningen	1	-2/+2
	variable to register, for registers r8-r15
2006-07-18	codegenerator makefile for 64 bit windows	John van Groningen	1	-0/+42

2006-06-30	allow strings with >200 characters for buildAC and eqAC_a	John van Groningen	1	-2/+82

2006-06-26	implement IFABS	John van Groningen	2	-2/+128

2006-06-13	port centry to 64 bit windows, save r15 in ccall and remove @n in ccall	John van Groningen	1	-13/+94

2006-06-13	fix cvtsd2si (source and destination register numbers were swapped)	John van Groningen	1	-2/+2

2006-06-01	move test for NO_STRING_ADDRESS_IN_DESCRIPTOR after include of cgport.h	John van Groningen	1	-3/+4

2006-06-01	fix ccall with a real as normal argument and an integer as state argument	John van Groningen	1	-4/+2
	(after second : )
2006-05-31	pass floating point parameters in registers in ccall for AI64	John van Groningen	1	-24/+36

2006-05-31	fix printing of integers of more than 32 bits	John van Groningen	1	-1/+1

2006-05-29	use offset -24 for apply update entry if profiling is used	John van Groningen	2	-13/+18

2006-05-29	use string with 4 byte length for .string directive for 64 bit architectures	John van Groningen	1	-12/+23

2006-05-11	fix ccall on 64 bit platforms with Real result	John van Groningen	1	-1/+1

2006-05-11	allow shifts up to 63 bits instead of 31	John van Groningen	2	-4/+4

2006-05-10	add shift_s instructions, use RELA for ELF, port to linux	John van Groningen	1	-145/+478

2006-05-10	add shift_s instructions, port printf calls to linux	John van Groningen	1	-2/+88

2006-05-01	add ccall for 64 bit linux	John van Groningen	2	-5/+366

2006-05-01	use int_64 instead of __int64, define int_64, include stdint.h on 64 bits linux	John van Groningen	12	-30/+73

2006-04-25	enable NO_STRING_ADDRESS_IN_DESCRIPTOR for sparc	John van Groningen	1	-5/+7

2006-04-25	optimize apply with >1 arguments	John van Groningen	2	-2/+68

2006-04-25	enable NEW_DESCRIPTORS for sparc	John van Groningen	1	-1/+1

2006-04-18	compute register uses for GFROMF and GTOF	John van Groningen	1	-2/+18

2006-04-18	fix previous revision by using IBLTU instead of IBLT for unsigned < when	John van Groningen	1	-5/+5
	generating code for floating point branch
2006-03-28	intel_asm is only defined on IA32 and AMD64	John van Groningen	1	-0/+4

2006-03-24	remove use of scratch register on IA32 for shift instructions (by adding	John van Groningen	11	-432/+1089
	instruction with extra register), IScc, IFScc, IFBcc (by adding IFCcc instructions with extra register), ICMPW (not used anymore) prevent generating FLD and FSTP instructions between FCOMP and FNSTSW instruction
2006-03-24	optimize shift n bitand 31 or 64 for IA32 and AMD64	John van Groningen	1	-0/+37

2006-03-24	fix entier for 64 bit processors	John van Groningen	1	-4/+8

2006-03-14	improve computation of used registers for reading from arrays, prevents	John van Groningen	1	-2/+4
	incrementing counters too soon in many small loops using one array
2006-02-24	add generating instruction for converting real to int	John van Groningen	2	-38/+64

2006-01-20	fix bugs in code generation for floating point computations for amd64	John van Groningen	5	-7/+56

2006-01-20	optimize cmp $0,r , generate test r,r instead	John van Groningen	4	-15/+50

2006-01-20	fix incorrect code generation for return with > 7 floating point parameters	John van Groningen	3	-0/+43
	in registers and < 2 integer parameters in registers for IA32
2006-01-11	generate code for amd64	John van Groningen	4	-0/+9953

2006-01-10	fiix bug introduced in previous revision because of incorrect merging of ↵	John van Groningen	1	-0/+1
	source code
2006-01-10	port to 64 bit windows	John van Groningen	17	-553/+1911

2005-12-06	add pushLc instruction for pushing the value of a label, with	John van Groningen	3	-0/+31
	a '_' or '.' added at the beginning if this is specified in the c ABI for this platform
2005-11-24	add passing of float argument in ccall for the PowerPC	John van Groningen	1	-1/+13

2005-11-21	use new descriptors on the PowerPC	John van Groningen	1	-3/+4

2005-11-21	change descriptors for apply with >1 argument and selectors	John van Groningen	2	-141/+223

2005-11-11	add .descs directive for selectors	John van Groningen	4	-8/+94

2005-11-11	improve printing of strings	John van Groningen	2	-1/+5

2005-10-10	use yet_args_neededed_n labels for n>4 for new descriptors	John van Groningen	1	-1/+5