Diff of /sml/trunk/HISTORY

Parent Directory | Revision Log | Patch

	revision 1486, Tue May 11 17:04:01 2004 UTC	revision 1572, Fri Jul 30 22:55:23 2004 UTC
#	Line 14	Line 14
14
15	----------------------------------------------------------------------	----------------------------------------------------------------------
16	Name: Matthias Blume (blume (at) tti - c (dot) org)	Name: Matthias Blume (blume (at) tti - c (dot) org)
17		Date: 2004/07/30 17:55:00 CDT
18		Tag: blume-20040730-various
19		Description:
20
21		Gearing up towards 110.47...
22
23		- various minor bugfixes to ml-nlffigen
24		- a beginning of a manual for nlffi
25
26		- eliminated 'export name=value' in config/install.sh as this does
27		not work with certain versions of /bin/sh
28		(Thanks to David King at Motorola for catching this.)
29
30		- several bugfixes provided or suggested by David Hansel at Reactive Systems:
31		- added a test for tm==NULL to gmtime.c and localtime.c
32		- applied patch for incorrect GetFileTime under win32
33		- toSeconds -> toMilliseconds in Win32/win32-process.sml
34
35		----------------------------------------------------------------------
36		Name: Matthias Blume (blume (at) tti - c (dot) org)
37		Date: 2004/07/21 18:20:00 CDT
38		Tag: blume-20040721-nlffigen
39		Description:
40
41		- Fixed minor issue in ml-nlffigen: Now generate structure T_foo
42		for a typedef to an incomplete type, but leave out the "typ" member.
43		(This is just for consistency.)
44		- Started to produce what is supposed to become better (i.e., comprehensive)
45		documentation of what ml-nlffigen does and produces.
46
47		----------------------------------------------------------------------
48		Name: Matthias Blume (blume (at) tti - c (dot) org)
49		Date: 2004/07/14 16:25:00 CDT
50		Tag: blume-20040714-union
51		Description:
52
53		Added C_UNION to c-calls/c-types.sml and updated the machinery
54		(ml-nlffigen, cproto.sml) that conveys C function interface
55		information to the code generator.
56
57		However, the actual architecture-specific implementation of function
58		arguments and results that are C unions is still not implemented.
59
60		----------------------------------------------------------------------
61		Name: Allen Leung (leunga (at) reservoir (dot) com)
62		Date: 2004/07/14 14:38:00 EST
63		Tag: leunga-110_46_1-ppc-lwzu
64		Description:
65
66		Added these instructions to the PowerPC architecture:
67		LBZU(X), LHZU(X), LWZU(X),
68		STWU(X), STFDU, STFSU
69
70		etc...
71
72		Note: I haven't added their instruction encoding into the description.
73
74		----------------------------------------------------------------------
75		Name: Allen Leung (leunga (at) reservoir (dot) com)
76		Date: 2004/07/13 15:04:00 EST
77		Tag: leunga-110_46_1-ppc-lwarx
78		Description:
79
80		Added the two instructions LWARX and STWCX to the PowerPC
81		instruction set.
82
83		A (untested) rewrite of loop-structure.sml.  The old version
84		is completely broken.
85
86		----------------------------------------------------------------------
87		Name: Matthias Blume (blume (at) tti - c (dot) org)
88		Date: 2004/07/13 13:50:00 CDT
89		Tag: blume-20040713-nlffi
90		Description:
91
92		- use paramAlloc to report c-calls with too many arguments
93		(for PPC version where parameter area is pre-allocated)
94		- added ccall_maxargspace to machspec (to implement the above)
95		- made "make" commend in CM's "make" tool configurable
96		- added option (default: on) for passing the name of the SML/NJ's "bin"
97		directory to "make"; the call looks like this:
98
99		make <options> SMLNJ_BINDIR=<dir> <target>
100
101		This can be used by the Makefile to, e.g., pick the "right" version
102		of ml-nlffigen.
103		- minor code tweaks
104
105		----------------------------------------------------------------------
106		Name: Matthias Blume (blume (at) tti - c (dot) org)
107		Date: 2004/07/12 22:50:00 CDT
108		Tag: blume-110_46_1-macosx-nlffi
109		Description:
110
111		NLFFI under Mac OS X now working (sort of).  This is largely untested,
112		though.
113
114		Note:  1. You have to make a new, clean build of the runtime system.
115		2. There are new BOOTFILES, you have to use them!
116		(Doing the bootstrap process yourself would be *very* painful!
117		If you absolutely have to do it, build the system under
118		a different architecture and then cross-compile.)
119
120		Version bumped to 110.46.1 to account for runtime data format changes.
121
122		----------------------------------------------------------------------
123		Name: Matthias Blume (blume (at) tti - c (dot) org)
124		Date: 2004/06/18 14:30:00 CDT
125		Tag: blume-20040618-unix
126		Description:
127
128		Changed the implementation of structure Unix so that the same stream
129		is returned every time one of the {text,bin}{In,Out}streamOf functions
130		is invoked on the same proc.  This is not what the spec currently
131		says -- although IMO it arguably should.  (See discussion below.)
132
133		----------------------------------------------------------------------
134		Name: Matthias Blume (blume (at) tti - c (dot) org)
135		Date: 2004/06/17 18:15:00 CDT
136		Tag: Release_110_46
137		Description:
138
139		New working version (110.46).  NEW BOOTFILES!
140
141		----------------------------------------------------------------------
142		Name: Matthias Blume (blume (at) tti - c (dot) org)
143		Date: 2004/06/17 17:20:00 CDT
144		Tag: blume-20040617-timer-unix
145		Description:
146
147		Changed the interface of structures Timer and Unix to match the most
148		recent Basis spec.
149
150		In the case of Unix there still seems to be an open/weird issue:
151
152		The {text,bin}{In,Out}streamOf functions are supposed to create
153		fresh streams whenever they are called -- as opposed to have them
154		return the same stream every time.  This design is supposed to
155		prevent space leaks caused by proc values hanging on to streams.
156
157		The reap function, on the other hand, is supposed to close the
158		streams.  This cannot be done without having a handle on the
159		stream in proc after all...
160
161		I took the liberty to implement the following stopgap solution:
162
163		The proc value hangs on to the most recently created stream(s).
164		Reap closes those.  If either or both of the two streams hadn't
165		been created at all yet, then reap will close the corresponding
166		file descriptors directly.
167
168		PS: I don't understand the original space leak argument anymore.  If
169		a proc hangs on to the imperative stream, then I/O operations on those
170		will advance the state of the cached stream and avoid the space leak.
171
172		----------------------------------------------------------------------
173		Name: Matthias Blume (blume (at) tti - c (dot) org)
174		Date: 2004/05/28 16:45:00 CDT
175		Tag: blume-20040528-basis
176		Description:
177
178		Added signature PACK_REAL and exported functor PrimIO.
179
180		----------------------------------------------------------------------
181		Name: Matthias Blume (blume (at) tti - c (dot) org)
182		Date: 2004/05/25 16:00:00 CDT
183		Tag: blume-20040525-group-owner
184		Description:
185
186		CM now ignores (but still accepts) the "owner" information in group
187		descriptions.  The owner of a group is its next enclosing
188		library. Each group must have a unique owner.  (There is a virtual
189		"toplevel" library that owns groups which are not nested within a real
190		library.)  Previously, each group had to explicitly declare its owner,
191		and CM would check that such a declaration is correct.  The new scheme
192		is to have CM check that for each group there is precisely one owning
193		library.
194
195		The advantage of the new scheme is that the programmer no longer needs
196		to maintain the somewhat annoying owner information.  The downside is
197		that CM cannot enforce the ownership rule across multiple runs of
198		CM.make.  Fortunately, enclosing the same group in two different
199		libraries A and B which are not part of the same program does not
200		cause real problems.
201
202		----------------------------------------------------------------------
203		Name: Matthias Blume (blume (at) tti - c (dot) org)
204		Date: 2004/05/20 16:00:00 CDT
205		Tag: blume-20040520-win32
206		Description:
207
208		Made the win32 version work again.  (Strangely, a misplaced comma had
209		slipped into win32-process.c which prevented the runtime from being
210		compiled correctly.)
211
212		Also, included a minor addition to ml-build.bat analogous to what was
213		done in blume-20040519-ml-build.
214
215		----------------------------------------------------------------------
216		Name: Matthias Blume (blume (at) tti - c (dot) org)
217		Date: 2004/05/19 22:10:00 CDT
218		Tag: blume-20040519-ml-build
219		Description:
220
221		Arranged for ml-build to clean up after itself a little bit better.
222		The script generates a temporary SML source file and compiles it using
223		CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
224		gets rid of those at the end, so they don't accumulate under .cm.
225
226		This required a minor change to install.sh because the name of the
227		metadata directory (default: .cm) is actually configurable at
228		installation time.
229
230		----------------------------------------------------------------------
231		Name: Matthias Blume (blume (at) tti - c (dot) org)
232		Date: 2004/05/18 15:50:00 CDT
233		Tag: blume-20040518-mkreader
234		Description:
235
236		Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
237		implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
238
239		----------------------------------------------------------------------
240		Name: Matthias Blume (blume (at) tti - c (dot) org)
241		Date: 2004/05/11 14:35:00 CDT
242		Tag: blume-20040511-win32sock
243		Description:
244
245		Added previously missing support for many socket-related functions
246		under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
247		for the voluminous patch!
248
249		(I have not tested this patch under win32 yet.)
250
251		Here is David's e-mail:
252
253		Hi,
254
255		Attached to this email you find a diff against sml/nj 110.45
256		that will enable socket support under Windows.
257
258		To apply the patch (using unix or cygwin)
259		1) gunzip runtime.diff.gz
260		2) "cd" into "src/runtime" in the source tree of a fresh
261		110.45 installation.
262		3) patch -p 1 < [your/path/to]runtime.diff
263
264		The code compiles fine but has NOT yet been extensively tested.
265		I only ran a few tests for basic socket client functionality
266		(which worked fine).  Especially the functions that use ioctl
267		are not tested at all and might not work (see below).
268
269		I implemented this since we want to move to a newer version of sml/nj
270		but need socket support in order to use it.  This is the first time I
271		even had a look at the sml/nj source,  so please review my changes
272		before making this part of the distribution!  Here are a few issues
273		that I think might be better for someone to solve who is more
274		familiar with the sml/nj source (and socket programming):
275
276		- getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
277		exception since I could not figure out what the windows equivalent
278		of these functions is
279
280		- In sockets-osdep.h there are a some #include statements that are
281		only used in a few files that include sockets-osdep.h
282
283		- In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
284		does not process its return value since I don't know how to
285		report an error upwards.
286
287		- It would probably be good to have a call to WSACleanup() when
288		the library is unloaded (if there is such a possibility).
289		Otherwise I think Windows will take care of this automatically
290		when the process finishes.
291
292		- I used ioctlsocket() as a replacement for ioctl() but I have
293		no idea if that is actually the proper replacement on Windows.
294
295		- All these issues are marked in the code by "FIXME" comments.
296
297		We use sml/nj extensively in our products and are quite happy
298		with it.  I hope this contribution will help you.
299
300		Keep up the good work!
301
302		David
303
304		----------------------------------------------------------------------
305		Name: Matthias Blume (blume (at) tti - c (dot) org)
306		Date: 2004/05/11 14:20:00 CDT
307		Tag: blume-20040511-installml
308		Description:
309
310		Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
311		the patch.)
312
313		----------------------------------------------------------------------
314		Name: Matthias Blume (blume (at) tti - c (dot) org)
315		Date: 2004/05/11 14:05:00 CDT
316		Tag: blume-20040511-nlffi-netbsd
317		Description:
318
319		Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
320		the patch.)
321
322		----------------------------------------------------------------------
323		Name: Matthias Blume (blume (at) tti - c (dot) org)
324	Date: 2004/05/11 12:05:00 CDT	Date: 2004/05/11 12:05:00 CDT
325	Tag: blume-20040511-exports	Tag: blume-20040511-exports
326	Description:	Description:

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.1486   changed lines   Added in v.1572

---

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