| 1 : |
blume |
995 |
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
|
| 2 : |
|
|
<html>
|
| 3 : |
|
|
<head>
|
| 4 : |
|
|
<title>SML/NJ 110.37 NEWS</title>
|
| 5 : |
|
|
|
| 6 : |
|
|
<meta name="author" content="Matthias Blume">
|
| 7 : |
|
|
|
| 8 : |
|
|
<meta name="description" content="working release announcement for SML/NJ">
|
| 9 : |
|
|
</head>
|
| 10 : |
|
|
<body>
|
| 11 : |
|
|
|
| 12 : |
|
|
<div align="Center">
|
| 13 : |
|
|
<h1>S M L / N J 1 1 0 . 37
|
| 14 : |
|
|
N E W S</h1>
|
| 15 : |
|
|
</div>
|
| 16 : |
|
|
<br>
|
| 17 : |
|
|
|
| 18 : |
|
|
<div align="Center"><i>November 23, 2001</i><br>
|
| 19 : |
|
|
</div>
|
| 20 : |
|
|
<br>
|
| 21 : |
|
|
|
| 22 : |
|
|
<div align="Center"><b>WARNING This version is intended for compiler
|
| 23 : |
|
|
hackers. The version ought to be stable, however we have not run our full
|
| 24 : |
|
|
regression testing. <a href="http://cm.bell-labs.com/cm/cs/what/smlnj/index.html">
|
| 25 : |
|
|
http://cm.bell-labs.com/cm/cs/what/smlnj/index.html</a> </b><a href="http://cm.bell-labs.com/cm/cs/what/smlnj/index.html"><b><br>
|
| 26 : |
|
|
</b></a> </div>
|
| 27 : |
|
|
<br>
|
| 28 : |
|
|
|
| 29 : |
|
|
<hr width="100%" size="2" align="Left">
|
| 30 : |
|
|
<h3>Summary:</h3>
|
| 31 : |
|
|
|
| 32 : |
|
|
<ul>
|
| 33 : |
|
|
<li>Improvements to the MLRISC backend.</li>
|
| 34 : |
|
|
<li>Enhanced export-list syntax for .cm-files.</li>
|
| 35 : |
|
|
<li>Various assorted bug fixes and improvements to other components and
|
| 36 : |
|
|
subsystems.<br>
|
| 37 : |
|
|
</li>
|
| 38 : |
|
|
</ul>
|
| 39 : |
|
|
|
| 40 : |
|
|
<div align="Left">
|
| 41 : |
|
|
<hr width="100%" size="2" align="Left">
|
| 42 : |
|
|
<h3>Details:</h3>
|
| 43 : |
|
|
<h5>MLRISC:</h5>
|
| 44 : |
|
|
<ul>
|
| 45 : |
|
|
<li>Implemented a complete redesign of MLRISC pseudo-ops. Now there ought
|
| 46 : |
|
|
to never be any question of incompatabilities with pseudo-op syntax expected
|
| 47 : |
|
|
by host assemblers. For now, only modules supporting GAS syntax are implemented
|
| 48 : |
|
|
but more should follow, such as MASM, and vendor assembler syntax, e.g. IBM
|
| 49 : |
|
|
as, Sun as, etc. The new pseudo-ops design is used to add a string to the
|
| 50 : |
|
|
end of the code stream that represents the file name.</li>
|
| 51 : |
|
|
<li><br>
|
| 52 : |
|
|
</li>
|
| 53 : |
|
|
<li>Fix for a backpatching bug reported by Allen. Because the boundary
|
| 54 : |
|
|
between short and long span-dependent instructions is +/- 128, there are
|
| 55 : |
|
|
an astounding number of span-dependent instructions whose size is over estimated.
|
| 56 : |
|
|
Allen came up with the idea of letting the size of span dependent instructions
|
| 57 : |
|
|
be non-monotonic, for a maxIter number of times, after which the size must
|
| 58 : |
|
|
be monotonically increasing. This table shows the number of span-dependent
|
| 59 : |
|
|
instructions whose size was over-estimated as a function of maxIter, for
|
| 60 : |
|
|
the file Parse/parse/ml.grm.sml:
|
| 61 : |
|
|
<table cellpadding="2" cellspacing="2" border="1" width="40%" align="Center">
|
| 62 : |
|
|
<tbody>
|
| 63 : |
|
|
<tr>
|
| 64 : |
|
|
<td valign="Top" align="Left">maxIter<br>
|
| 65 : |
|
|
</td>
|
| 66 : |
|
|
<td valign="Top"># of instructions<br>
|
| 67 : |
|
|
</td>
|
| 68 : |
|
|
</tr>
|
| 69 : |
|
|
<tr>
|
| 70 : |
|
|
<td valign="Top">10<br>
|
| 71 : |
|
|
</td>
|
| 72 : |
|
|
<td valign="Top">687<br>
|
| 73 : |
|
|
</td>
|
| 74 : |
|
|
</tr>
|
| 75 : |
|
|
<tr>
|
| 76 : |
|
|
<td valign="Top">20<br>
|
| 77 : |
|
|
</td>
|
| 78 : |
|
|
<td valign="Top">438<br>
|
| 79 : |
|
|
</td>
|
| 80 : |
|
|
</tr>
|
| 81 : |
|
|
<tr>
|
| 82 : |
|
|
<td valign="Top">30<br>
|
| 83 : |
|
|
</td>
|
| 84 : |
|
|
<td valign="Top">198<br>
|
| 85 : |
|
|
</td>
|
| 86 : |
|
|
</tr>
|
| 87 : |
|
|
<tr>
|
| 88 : |
|
|
<td valign="Top">40<br>
|
| 89 : |
|
|
</td>
|
| 90 : |
|
|
<td valign="Top">0<br>
|
| 91 : |
|
|
</td>
|
| 92 : |
|
|
</tr>
|
| 93 : |
|
|
</tbody>
|
| 94 : |
|
|
</table>
|
| 95 : |
|
|
</li>
|
| 96 : |
|
|
<li>In compiling the compiler, there is no significant difference in compilation
|
| 97 : |
|
|
speed between maxIter=10 and maxIter=40. Actually, my measurements showed
|
| 98 : |
|
|
that maxIter=40 was a tad faster than maxIter=10! Also 96% of the files
|
| 99 : |
|
|
in the compiler reach a fix point within 13 iterations, so fixing maxIter
|
| 100 : |
|
|
at 40, while high, is okay.</li>
|
| 101 : |
|
|
<li>Fixed the bug described in blume-20010920-slowfp. The fix involves </li>
|
| 102 : |
|
|
<ol>
|
| 103 : |
|
|
<li>generating FCOPYs in FSTP in ia32-svid</li>
|
| 104 : |
|
|
<li>marking a CALL with the appropriate annotation</li>
|
| 105 : |
|
|
</ol>
|
| 106 : |
|
|
<li>(x86-fast-fp is still off by default.)</li>
|
| 107 : |
|
|
<li>X86RA now uses a valid (instead of dummy) PrintFlowgraph module.</li>
|
| 108 : |
|
|
<li>The representation of a program point never expected to see more than
|
| 109 : |
|
|
65536 instructions in a basic block! Fixed.</li>
|
| 110 : |
|
|
</ul>
|
| 111 : |
|
|
<h5>CM</h5>
|
| 112 : |
|
|
<ul>
|
| 113 : |
|
|
<li>avoid "/../" in filenames as much as possible (but only where it is
|
| 114 : |
|
|
safe)</li>
|
| 115 : |
|
|
<li>experimental implementation of "portable graphs" (See the proposal
|
| 116 : |
|
|
in <a href="http://cm.bell-labs.com/cm/cs/who/blume/pgraph/proposal.pdf">
|
| 117 : |
|
|
http://cm.bell-labs.com/cm/cs/who/blume/pgraph/proposal.pdf </a>for details
|
| 118 : |
|
|
on portable library graphs.)</li>
|
| 119 : |
|
|
<li>there are also new libraries $/pgraph.cm and $/pgraph-util.cm</li>
|
| 120 : |
|
|
<li>CM now implements a simple "set calculus" for specifying export lists.
|
| 121 : |
|
|
In particular, it is now possible to refer to the export lists of other libraries/groups/sources
|
| 122 : |
|
|
and form unions as well as differences. See the latest <a href="http://cm.bell-labs.com/cm/cs/what/smlnj/doc/CM/new.pdf">
|
| 123 : |
|
|
CM manual</a> for details.</li>
|
| 124 : |
|
|
<li>A separate notion of "proxy" libraries has been eliminated from CM's
|
| 125 : |
|
|
model. (Proxy libraries are now simply a special case of using the
|
| 126 : |
|
|
export list calculus.)</li>
|
| 127 : |
|
|
</ul>
|
| 128 : |
|
|
<h5>compiler</h5>
|
| 129 : |
|
|
<ul>
|
| 130 : |
|
|
<li>routed the name of the current source file to mlriscgen where it gets
|
| 131 : |
|
|
directly emitted into the code object (This uses the new pseudo-op infrastructure
|
| 132 : |
|
|
of MLRISC.)</li>
|
| 133 : |
|
|
</ul>
|
| 134 : |
|
|
<h5>runtime:</h5>
|
| 135 : |
|
|
<ul>
|
| 136 : |
|
|
<li>Underscore patch from Chris Richards (fixing problem with compiling
|
| 137 : |
|
|
runtime system under recent NetBSD).</li>
|
| 138 : |
|
|
<li>Removed handling of code object name strings from runtime code. (The
|
| 139 : |
|
|
only remaining code-name-string-related code in the runtime is for extracting/reading
|
| 140 : |
|
|
of same.)</li>
|
| 141 : |
|
|
</ul>
|
| 142 : |
|
|
<h5>CKIT:</h5>
|
| 143 : |
|
|
<ul>
|
| 144 : |
|
|
<li>Changed the "Function" constructor of type Ast.ctype to carry optional
|
| 145 : |
|
|
argument identifiers.</li>
|
| 146 : |
|
|
<li>Changed the return type of TypeUtil.getFunction accordingly.</li>
|
| 147 : |
|
|
<li>Type equality ignores the argument names.</li>
|
| 148 : |
|
|
<li>TypeUtil.composite tries to preserve argument names but gives up quickly
|
| 149 : |
|
|
if there is a mismatch.</li>
|
| 150 : |
|
|
</ul>
|
| 151 : |
|
|
<h5>installation script:</h5>
|
| 152 : |
|
|
<ul>
|
| 153 : |
|
|
<li>attempts to use "curl" if available (unless "wget" is available as
|
| 154 : |
|
|
well)</li>
|
| 155 : |
|
|
<li>put relative anchor names for tool commands into pathconfig</li>
|
| 156 : |
|
|
</ul>
|
| 157 : |
|
|
<h5>run script:</h5>
|
| 158 : |
|
|
<ul>
|
| 159 : |
|
|
<li>Changed config/_run-sml (resulting in a changed bin/.run-sml) so that
|
| 160 : |
|
|
arguments that contain delimiters are passed through correctly. This change
|
| 161 : |
|
|
also means that all "special" arguments of the form @SMLxxx... must come
|
| 162 : |
|
|
first.</li>
|
| 163 : |
|
|
</ul>
|
| 164 : |
|
|
<h5>NLFFI-LIB:</h5>
|
| 165 : |
|
|
<ul>
|
| 166 : |
|
|
<li>some cleanup (all cosmetic)</li>
|
| 167 : |
|
|
</ul>
|
| 168 : |
|
|
<h5>NLFFIGEN:</h5>
|
| 169 : |
|
|
<ul>
|
| 170 : |
|
|
<li>temporarily disabled the mechanism that suppresses ML output for C
|
| 171 : |
|
|
definitions whose identifiers start with an underscore character</li>
|
| 172 : |
|
|
<li>generate val bindings for enum constants</li>
|
| 173 : |
|
|
<li>user can request that only one style (light or heavy) is being used;
|
| 174 : |
|
|
default is to use both (command-line arguments: -heavy and -light)</li>
|
| 175 : |
|
|
<li>fixed bug in handling of function types involving incomplete pointers</li>
|
| 176 : |
|
|
<li>generate ML entry points that take record arguments (i.e., using named
|
| 177 : |
|
|
arguments) for C functions that have a prototype with named arguments (see
|
| 178 : |
|
|
changes to CKIT)</li>
|
| 179 : |
|
|
</ul>
|
| 180 : |
|
|
<h5>Basis implementation:</h5>
|
| 181 : |
|
|
<ul>
|
| 182 : |
|
|
<li>provided a non-hook implementation of exnName (at the toplevel) and
|
| 183 : |
|
|
made the "dummy" implementation of exnMessage (at the toplevel) more useful:
|
| 184 : |
|
|
if nothing gets "hooked in", then at least you are going to see the exception
|
| 185 : |
|
|
name and a message indicating why you don't see more.</li>
|
| 186 : |
|
|
<ul>
|
| 187 : |
|
|
<li>[For the time being, programs that need exnMessage and want to use
|
| 188 : |
|
|
ml-build should either use General.exnMessage (strongly recommended) or refer
|
| 189 : |
|
|
to structure General at some other point so that CM sees a static dependency.]</li>
|
| 190 : |
|
|
</ul>
|
| 191 : |
|
|
<ul>
|
| 192 : |
|
|
<li>[Similar remarks go for "print" and "use": If you want to use
|
| 193 : |
|
|
their functionality in stand-alone programs generated by ml-build, then use
|
| 194 : |
|
|
TextIO.output and Backend.Interact.useFile (from $smlnj/compiler.cm)</li>
|
| 195 : |
|
|
</ul>
|
| 196 : |
|
|
</ul>
|
| 197 : |
|
|
<h5>ml-yacc/ml-lex/ml-burg:</h5>
|
| 198 : |
|
|
<ul>
|
| 199 : |
|
|
<li>Don't use "exnMessage" -- use "General.exnMessage"! (see "Basis implementation")</li>
|
| 200 : |
|
|
</ul>
|
| 201 : |
|
|
<h4></h4>
|
| 202 : |
|
|
|
| 203 : |
|
|
<h4></h4>
|
| 204 : |
|
|
</div>
|
| 205 : |
|
|
|
| 206 : |
|
|
</body>
|
| 207 : |
|
|
</html>
|
Parent Directory
| 
Revision Log
