Diff of /sml/trunk/HISTORY

Parent Directory | Revision Log | Patch

	revision 1142, Wed Mar 13 22:25:37 2002 UTC	revision 1159, Fri Mar 22 03:18:24 2002 UTC
#	Line 13	Line 13
13	Description:	Description:
14
15	----------------------------------------------------------------------	----------------------------------------------------------------------
16		Name: Matthias Blume
17		Date: 2002/03/21 22:20:00 EST
18		Tag: blume-20020321-kmp-bugfix
19		Description:
20
21		Changed the interface to the KMP routine in PreString and fixed
22		a minor bug in one place where it was used.
23
24		----------------------------------------------------------------------
25		Name: Allen Leung
26		Date: 2002/03/21 20:30:00 EST
27		Tag: leunga-20020321-cfg
28		Description:
29
30		Fixed a potential problem in cfg edge splitting.
31
32		----------------------------------------------------------------------
33		Name: Allen Leung
34		Date: 2002/03/21 17:15:00 EST
35		Tag: leunga-20020321-x86-fp-cfg
36		Description:
37
38		1. Recoded the buggy parts of x86-fp.
39
40		a. All the block reordering code has been removed.
41		We now depend on the block placement phases to do this work.
42
43		b. Critical edge splitting code has been simplified and moved into the
44		CFG modules, as where they belong.
45
46		Both of these were quite buggy and complex.  The code is now much, much
47		simpler.
48
49		2. X86 backend.
50
51		a. Added instructions for 64-bit support.  Instruction selection for
52		64-bit has not been committed, however, since that
53		requires changes to MLTREE which haven't been approved by
54		Lal and John.
55
56		b. Added support for FUCOMI and FUCOMIP when generating code for
57		PentiumPro and above.  We only generate these instructions in
58		the fast-fp mode.
59
60		c. Added cases for JP and JNP in X86FreqProps.
61
62		3. CFG
63
64		CFG now has a bunch of methods for edge splitting and merging.
65
66		4. Machine description.
67
68		John's simplification of MLTREE_BASIS.fcond broke a few machine
69		description things:
70
71		rtl-build.{sig,sml} and hppa.mdl fixed.
72
73		NOTE: the machine description stuff in the repository is still broken.
74		Again, I can't put my fixes in because that involves
75		changes to MLTREE.
76
77		----------------------------------------------------------------------
78		Name: Matthias Blume
79		Date: 2002/03/20 15:55:00 EST
80		Tag: blume-20020320-kmp
81		Description:
82
83		Implemented Knuth-Morris-Pratt string matching in PreString and used
84		it for String.isSubstring, Substring.isSubstring, and
85		Substring.position.
86
87		(Might need some stress-testing.  Simple examples worked fine.)
88
89		----------------------------------------------------------------------
90		Name: Matthias Blume
91		Date: 2002/03/19 16:37:00 EST
92		Tag: blume-20020319-witnesses
93		Description:
94
95		Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
96
97		This implements a generic mechanism for changing constness qualifiers
98		anywhere within big C types without resorting to outright "casts".
99		(So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
100		modify the constness at the outermost level.)
101		The implementation of "convert" is based on the idea of "witness"
102		values -- values that are not used by the operation but whose types
103		"testify" to their applicability.  On the implementation side, "convert"
104		is simply a projection (returning its second curried argument).  With
105		cross-module inlining, it should not result in any machine code being
106		generated.
107
108		----------------------------------------------------------------------
109		Name: Matthias Blume
110		Date: 2002/03/15 16:40:00 EST
111		Tag: blume-20020315-basis
112		Description:
113
114		Provided (preliminary?) implementations for
115
116		{String,Substring}.{concatWith,isSuffix,isSubstring}
117
118		and
119
120		Substring.full
121
122		Those are in the Basis spec but they were missing in SML/NJ.
123
124		----------------------------------------------------------------------
125		Name: Matthias Blume
126		Date: 2002/03/14 21:30:00 EST
127		Tag: blume-20020314-controls
128		Description:
129
130		Controls:
131		---------
132
133		1. Factored out the recently-added Controls : CONTROLS stuff and put
134		it into its own library $/controls-lib.cm.  The source tree for
135		this is under src/smlnj-lib/Controls.
136
137		2. Changed the names of types and functions in this interface, so they
138		make a bit more "sense":
139
140		module -> registry
141		'a registry -> 'a group
142
143		3. The interface now deals in ref cells only.  The getter/setter interface
144		is (mostly) gone.
145
146		4. Added a function that lets one register an already-existing ref cell.
147
148		5. Made the corresponding modifications to the rest of the code so that
149		everything compiles again.
150
151		6. Changed the implementation of Controls.MLRISC back to something closer
152		to the original.  In particular, this module (and therefore MLRISC)
153		does not depend on Controls.  There now is some link-time code in
154		int-sys.sml that registers the MLRISC controls with the Controls
155		module.
156
157		CM:
158		---
159
160		* One can now specify the lambda-split aggressiveness in init.cmi.
161
162		----------------------------------------------------------------------
163	Name: Allen Leung	Name: Allen Leung
164	Date: 2002/03/13 17:30:00 EST	Date: 2002/03/13 17:30:00 EST
165	Tag: leunga-20020313-x86-fp-unary	Tag: leunga-20020313-x86-fp-unary
#	Line 75	Line 222
222	provides a bit more "stability":  Once CM has seen a compilation	provides a bit more "stability":  Once CM has seen a compilation
223	unit, it keeps its identity constant (as long as you do not delete	unit, it keeps its identity constant (as long as you do not delete
224	those crucial CM/GUID/* files).  This means that when you change	those crucial CM/GUID/* files).  This means that when you change
225	and interface, compiler, then go back to the old interface, and	an interface, compile, then go back to the old interface, and
226	compile again, you arrive at the original pid.	compile again, you arrive at the original pid.
227
228	There now also is a mechanism that instructs CM to use the plain	There now also is a mechanism that instructs CM to use the plain

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.1142   changed lines   Added in v.1159

---

root@smlnj-gforge.cs.uchicago.edu	ViewVC Help
Powered by ViewVC 1.0.0