Standard ML of New Jersey
Release Notes

Version 110.92
August 10, 2019


This release has a major improvement in the compiler that makes the use of polymorphic arrays more efficient; namely, we have removed the dictionary passing for polymorphic array operations from FLINT. Besides that change, this release is primarily a checkpoint of work being done to support 64-bit targets, including new infrastructure for cross compilation.

We have had reports of a problem with using CM on Windows (bug #223). We have not been able to reproduce this problem, but we are leaving the bug open for now. If you encounter problems with this release on Windows, please submit a bug report using the bug form.



While the installer for 110.92 works on macOS 10.14 Mojave, building from source requires some extra steps. Another issue that you may encounter when building on macOS 10.14 Mojave is an error message for a shell script of the form

  /bin/sh: bad interpreter: Operation not permitted

This error arises because the attribute is set on the shell script. To fix the problem, remove the attribute using the command

  xattr -d shell-script

and resume the build.


We have removed the runtime-type passing from the FLINT IR. This mechanism was solely used to allow the real array type to be implemented using a packed representation (i.e., like RealArray.array), but it incurred an overhead on all polymorphic array code. By removing this mechanism, we have spead-up the compiler by almost 10% and we see significant improvements in most uses of polymorphic arrays. The downside is that code that uses the real array type heavily will be somewhat slower; in that case, we recommend switching to the monomorphic RealArray.array type.

We have a new version of the cross-compilation script (system/cmb-cross) thatsupports target-specific dependencies in the front-end (i.e., representation of numeric types and endianess). We use this new script in the allcross script too.


  • Some cleanup in the x86 backend. Removed the MULB, MULW, and MULL unsigned-multiplication instructions, since they are not binary operations. The MULL instruction is covered by the MULL1 constructor in the multDivOp datatype. The same change was applied to the amd64 backend.

  • Many changes to the amd64 machine description:

    • Removed the PUSHB, PUSHW, and PUSHL instructions, since the matching POP operations are not supported.

    • Removed the CALLQ operation, since it is the same as CALL.

    • Removed the CLTD and CQTO operations, since those names are just synonyms for CDQ and CDO.

    • Replaced the INTO operation (which is not valid in 64-bit mode) with INT of byte.

  • Restructured the amd64 machine-code generation implementation and filled in many of the missing encodings. It should be complete for SML/NJ code generation, but needs more work to support the full set of operations described in the amd64.mdl file.

Runtime System

Made a bunch of changes to fix compilation issues on 64-bit targets.


Here is a list of tracked bugs fixed (or closed) with this release, please see the bug tracker for more details.


Word64.fromLargeInt fails

Supported systems

We believe that SML/NJ will build and run on the following systems, but have only tested some of them:

Architecture Operating System Status

Power PC

Mac OS X 10.5 (Leopard)







x86 (32-bit)

Mac OS X 10.6 (Snow Leopard)

Mac OS X 10.7 (Lion)

Mac OS X 10.8 (Mountain Lion)

Mac OS X 10.9 (Mavericks)

Mac OS X 10.10 (Yosemite)

Mac OS X 10.11 (El Capitan)

macOS 10.12 (Sierra)


macOS 10.13 (High Sierra)


macOS 10.14 (Mojave)


Ubuntu 16.04.3 LTS


Other Linux variants

BSD variants

Windows 7

Windows 10

Cygwin (32-bit)

A note about 64-bit support

Apple is in the process of deprecating 32-bit support in macOS. With macOS 10.14 Mojave, compiling 32-bit applications, such as the SML/NJ runtime system, requires using an older SDK. The SML/NJ installer, however, works without issue on macOS 10.14 Mojave.

We have been working on 64-bit support since 110.82 we expect to have it in place before the release of macOS 10.15.