doc/guide/src/Bugs20041109.adoc

   1 Bugs20041109
   2 ============
   3
   4 Here are the known bugs in <:Release20041109:MLton 20041109>, listed
   5 in reverse chronological order of date reported.
   6
   7 * <!Anchor(bug17)>
   8  `MLton.Finalizable.touch` doesn't necessarily keep values alive
   9  long enough.  Our SVN has a patch to the compiler.  You must rebuild
  10  the compiler in order for the patch to take effect.
  11 +
  12 Thanks to Florian Weimer for reporting this bug.
  13
  14 * <!Anchor(bug16)>
  15  A bug in an optimization pass may incorrectly transform a program
  16  to flatten ref cells into their containing data structure, yielding a
  17  type-error in the transformed program.  Our CVS has a
  18  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/mlton/ssa/ref-flatten.fun.diff?r1=1.35&r2=1.37[patch]
  19  to the compiler.  You must rebuild the compiler in order for the
  20  patch to take effect.
  21 +
  22 Thanks to <:VesaKarvonen:> for reporting this bug.
  23
  24 * <!Anchor(bug15)>
  25  A bug in the front end mistakenly allows unary constructors to be
  26  used without an argument in patterns.  For example, the following
  27  program is accepted, and triggers a large internal error.
  28 +
  29 [source,sml]
  30 ----
  31 fun f x = case x of SOME => true | _ => false
  32 ----
  33 +
  34 We have fixed the problem in our CVS.
  35 +
  36 Thanks to William Lovas for reporting this bug.
  37
  38 * <!Anchor(bug14)>
  39  A bug in `Posix.IO.{getlk,setlk,setlkw}` causes a link-time error:
  40  `undefined reference to Posix_IO_FLock_typ`
  41  Our CVS has a
  42  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/basis-library/posix/primitive.sml.diff?r1=1.34&r2=1.35[patch]
  43  to the Basis Library implementation.
  44 +
  45 Thanks to Adam Chlipala for reporting this bug.
  46
  47 * <!Anchor(bug13)>
  48  A bug can cause programs compiled with `-profile alloc` to
  49  segfault.  Our CVS has a
  50  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/mlton/backend/ssa-to-rssa.fun.diff?r1=1.106&r2=1.107[patch]
  51  to the compiler.  You must rebuild the compiler in order for the
  52  patch to take effect.
  53 +
  54 Thanks to John Reppy for reporting this bug.
  55
  56 * <!Anchor(bug12)>
  57  A bug in an optimization pass may incorrectly flatten ref cells
  58  into their containing data structure, breaking the sharing between
  59  the cells.  Our CVS has a
  60  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/mlton/ssa/ref-flatten.fun.diff?r1=1.32&r2=1.33[patch]
  61  to the compiler.  You must rebuild the compiler in order for the
  62  patch to take effect.
  63 +
  64 Thanks to Paul Govereau for reporting this bug.
  65
  66 * <!Anchor(bug11)>
  67  Some arrays or vectors, such as `(char * char) vector`, are
  68  incorrectly implemented, and will conflate the first and second
  69  components of each element.  Our CVS has a
  70  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/mlton/backend/packed-representation.fun.diff?r1=1.32&r2=1.33[patch]
  71  to the compiler.  You must rebuild the compiler in order for the
  72  patch to take effect.
  73 +
  74 Thanks to Scott Cruzen for reporting this bug.
  75
  76 * <!Anchor(bug10)>
  77  `Socket.Ctl.getLINGER` and `Socket.Ctl.setLINGER`
  78  mistakenly raise `Subscript`.
  79  Our CVS has a
  80  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/basis-library/net/socket.sml.diff?r1=1.14&r2=1.15[patch]
  81  to the Basis Library implementation.
  82 +
  83 Thanks to Ray Racine for reporting the bug.
  84
  85 * <!Anchor(bug09)>
  86  <:ConcurrentML: CML> `Mailbox.send` makes a call in the wrong atomic context.
  87  Our CVS has a http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/lib/cml/core-cml/mailbox.sml.diff?r1=1.3&r2=1.4[patch]
  88  to the CML implementation.
  89
  90 * <!Anchor(bug08)>
  91  `OS.Path.joinDirFile` and `OS.Path.toString` did not
  92  raise `InvalidArc` when they were supposed to.  They now do.
  93  Our CVS has a http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/basis-library/system/path.sml.diff?r1=1.8&r2=1.11[patch]
  94  to the Basis Library implementation.
  95 +
  96 Thanks to Andreas Rossberg for reporting the bug.
  97
  98 * <!Anchor(bug07)>
  99  The front end incorrectly disallows sequences of expressions
 100  (separated by semicolons) after a topdec has already been processed.
 101  For example, the following is incorrectly rejected.
 102 +
 103 [source,sml]
 104 ----
 105 val x = 0;
 106 ignore x;
 107 ignore x;
 108 ----
 109 +
 110 We have fixed the problem in our CVS.
 111 +
 112 Thanks to Andreas Rossberg for reporting the bug.
 113
 114 * <!Anchor(bug06)>
 115  The front end incorrectly disallows expansive `val`
 116  declarations that bind a type variable that doesn't occur in the
 117  type of the value being bound.   For example, the following is
 118  incorrectly rejected.
 119 +
 120 [source,sml]
 121 ----
 122 val 'a x = let exception E of 'a in () end
 123 ----
 124 +
 125 We have fixed the problem in our CVS.
 126 +
 127 Thanks to Andreas Rossberg for reporting this bug.
 128
 129 * <!Anchor(bug05)>
 130  The x86 codegen fails to account for the possibility that a 64-bit
 131  move could interfere with itself (as simulated by 32-bit moves).  We
 132  have fixed the problem in our CVS.
 133 +
 134 Thanks to Scott Cruzen for reporting this bug.
 135
 136 * <!Anchor(bug04)>
 137  `NetHostDB.scan` and `NetHostDB.fromString` incorrectly
 138  raise an exception on internet addresses whose last component is a
 139  zero, e.g `0.0.0.0`.  Our CVS has a
 140  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/basis-library/net/net-host-db.sml.diff?r1=1.12&r2=1.13[patch] to the Basis Library implementation.
 141 +
 142 Thanks to Scott Cruzen for reporting this bug.
 143
 144 * <!Anchor(bug03)>
 145  `StreamIO.inputLine` has an off-by-one error causing it to drop
 146  the first character after a newline in some situations.  Our CVS has a
 147  http://mlton.org/cgi-bin/viewcvs.cgi/mlton/mlton/basis-library/io/stream-io.fun.diff?r1=text&tr1=1.29&r2=text&tr2=1.30&diff_format=h[patch].
 148  to the Basis Library implementation.
 149 +
 150 Thanks to Scott Cruzen for reporting this bug.
 151
 152 * <!Anchor(bug02)>
 153  `BinIO.getInstream` and `TextIO.getInstream` are
 154  implemented incorrectly.  This also impacts the behavior of
 155  `BinIO.scanStream` and `TextIO.scanStream`.  If you (directly
 156  or indirectly) realize a `TextIO.StreamIO.instream` and do not
 157  (directly or indirectly) call `TextIO.setInstream` with a derived
 158  stream, you may lose input data.  We have fixed the problem in our
 159  CVS.
 160 +
 161 Thanks to <:WesleyTerpstra:> for reporting this bug.
 162
 163 * <!Anchor(bug01)>
 164  `Posix.ProcEnv.setpgid` doesn't work.  If you compile a program
 165  that uses it, you will get a link time error
 166 +
 167 ----
 168 undefined reference to `Posix_ProcEnv_setpgid'
 169 ----
 170 +
 171 The bug is due to `Posix_ProcEnv_setpgid` being omitted from the
 172  MLton runtime.  We fixed the problem in our CVS by adding the
 173  following definition to `runtime/Posix/ProcEnv/ProcEnv.c`
 174 +
 175 [source,c]
 176 ----
 177 Int Posix_ProcEnv_setpgid (Pid p, Gid g) {
 178         return setpgid (p, g);
 179 }
 180 ----
 181 +
 182 Thanks to Tom Murphy for reporting this bug.