9.4 changes:

- updated version to 9.4

bintar README NOTICE makefiles/Mf-install.in scheme.1.in

c/Makefile.i3nt c/Makefile.a6nt c/Makefile.ti3nt c/Makefile.ta6nt

mats/bldnt.bat workarea c/scheme.rc s/7.ss s/cmacros.ss

release_notes/release_notes.stex csug/csug.stex

- added missing include

killme.c

- added new mat verifying that primitives raise exceptions for invalid

arguments based on primdata.ss signatures. fixed some of those

signatures, which weren't otherwise used except for arity checking.

fixed some issues turned up by the test with who reports errors

and how.

primdata.ss, 5_1.ss, 7.ss, compile.ss, cpnanopass.ss, fasl.ss,

interpret.ss, io.ss, record.ss, syntax.ss,

primvars.ms, 5_1.ms, 7.ms, 8.ms, record.ms, mats/Mf-base,

root-experr*, patch*

- string comparisons (string=?, string-ci<?, etc.) now get out fast

when handed eq arguments.

5_4.ss

- changed representation of most-negative iptr, I32, and I64 to make

Visual C compiler happy. updated windows make files, which had

fallen out of date. added missing hsrc= files sort.h and thread.h

so they show up in the Windows workarea c directory.

cmacros.ss,

fasl.c, number.c, c/Mf-base, c/Makefile.{t,}{i3,a6}nt

- The scheme.h definition of Sfixnum(x) now uses multiply rather than

left shift to avoid counting on the behavior of << on negative numbers,

which is undefined in C.

mkheader.ss

- Fixed a couple of casts, one harmless and the other causing

Sinteger64_value to return 0 for the most-negative I64 on 32-bit

builds.

number.c

- The configure-generated Makefile distclean target no longer removes

zlib and nanopass, since they are logically part of the git clone.

It does run make distclean in zlib.

makefiles/Makefile.in

- converted s_test_schlib shifts of -1 to equivalent shifts of 1 to

avoid undefined left-shift behavior on negative numbers.

prim5.c

- added if(---) {} wrapper around call to WRITE in display macro to

silence unused return-value warnings.

prim5.c

- liberalized get-mode check for ../mats. it's not our business whether

people make their directories group and/or other writeable.

6.ms

- make test now prints the actual relative path to summary in the

"check summary" message, whether invoked from the top-level directory

or from the workarea.

Makefile.in, Makefile-workarea.in

- configure now just uses cat to copy Makefile-workarea.in to $w/workarea,

since the file is presently the same regardless of the configuration.

configure

- fixed time-utc->date test in mat time&date-printing to work regardless of

what locale (and time zone) the host machine has set.

date.ms

- fixed date->time-utc to honor the zone-offset field when converting a date

object to a time-utc object.

stats.c,

date.ms

- fixed incorrect handling of library-extension when searching wpo files

compile.ss,

7.ms

- modified floatify_normalize to properly round denormalized results.

obviated scale_float in the process.

number.c,

ieee.ms

- fixed 0eNNNN for large NNNN to produce 0.0 rather than infinity

strnum.ss,

5_3.ms

- the reader now raises an exception with condition type implementation

restriction (among the other usual lexical condition types), and

string->number now raises #f, for #e<m>@<a>, where <m> and <a> are

nonzero integers, since Chez Scheme can't represent polar numbers other

than 0@<n> and <n>@0 exactly. <m>@<a> still produces an inexact result,

i.e., we're still extending the set of inexact numeric constants beyond

what R6RS dictates. doing this required a rework of $str->num, which

turned into a fairly extensive rewrite that fixed up a few other minor

issues (like r6rs:string->number improperly allowing 1/2e10) and

eliminated the need for consumers to call $str->num twice in cases

where it can actually produce a number. added some related new tests,

including several found missing by profiling. added a couple of

checks to number->string the absence of which was causing argument

errors to be reported by other routines.

strnum.ss, exceptions.ss, read.ss

5_3.ms, 6.ms, root-experr*, patch*

- added pdtml flag, which if set to t causes profile-dump-html to be

called at the end of a mat run.

mats/Mf-base

- compile-whole-program and compile-whole-library now copy the hash-bang

line from the wpo file (if it has one) to the object file.

compile.ss,

7.ms

- stex is now a submodule. csug/Makefile and release_notes/Makefile

set and use the required Scheme and STEXLIB variables accordingly.

they default the machine type to a6le, but this can be overridden

and is by the generated top-level Makefile. the generated top-level

Makefile now has a new docs target that runs make in both csug and

release_notes, and an updated distclean target that cleans the same.

the annoying csug Makefile .fig.pdf rule redefinition is now gone.

copyright.stex and csug.stex now list May 2016 as the revision month

and date; this will have to be updated for future releases.

configure, makefiles/Makefile.in,

csug/Makefile, copyright.stex, csug.stex,

release_notes/Makefile

- added custom install options. workarea creates an empty config.h,

and configure creates a config.h that sets the default scheme heap

path and scheme-script name based on the actual configuration.

configure, newrelease, workarea, checkin,

c/Mf-base, scheme.c, main.c,

Mf-install.in

- renamed the installed example directory from lib to examples.

Mf-install.in,

scheme.1.in

- added force option to gzip during man page install to prevent gzip from

asking for permission to overwrite existing man page files.

Mf-install.in

- removed ~/lib/csv%v/%m from the default scheme heap path on unix-like

systems. documented inclusion of %x\..\..\boot\%m in the Windows

default scheme heap path.

main.c,

use.stex

- added new configuration options: --installbin, --installlib,

--installschemename, --installpetitename, and --installscriptname.

configure

- updated the example library link to the nanopass framework.

CONTRIBUTING.md

- now cleaning up petite.1 and scheme.1 left behind by make install

Makefile-workarea.in, checkin

- now removing workarea after cleaning csug and release_notes so

Mf-stex (included from csug/Makefile and release_notes/Makefile)

doesn't complain trying to determine the machine type.

Makefile.in

- added installsh support for --ifdiff so the csug make file can use it

for the install target.

installsh,

csug/Makefile

- added instructions for building (cross-compiling) a boot file for

a supported machine type for which a boot file is not built by default.

BUILDING

- corrected CHEZSCHEMELIBS and CHEZSCHEMEEXTS index entries to be

CHEZSCHEMELIBDIRS and CHEZSCHEMELIBEXTS.

use.stex

- updated to curl stex version 1.2.1

configure

- updated the fix to S_mktime to work on windows. the struct tm

structure on windows does not have the tm_gmtoff field used in the

mac and linux version of the code.

stats.c

- updated the Windows makefiles for building and testing to remove links for

files that no longer exist, which was getting in the way of creating links

for files that do exist. Also updated the build batch file for Windows to

compile the nanopass framework separately before building the rest of the

scheme compiler and libraries.

s/Mf-{a6,i3,ta6,ti3}nt, s/bldnt.bat,

mats/Mf-{a6,i3,ta6,ti3}nt

- updated the read me to include a link to the Chez Scheme project page.

README.md

- fixed embarrassing typo in read me.

README.md

- profiler's html output refresh: mark the files as HTML5 rather

than HTML4; use target attributes rather than onclick events to

open links in specific windows; add a missing table row element;

replace the deprecated name attribute with an id attribute (and

replace the anchors with spans); and replace the deprecated valign

attribute with a style attribute.

pdhtml.ss

9.4.1 changes:

- updated version to 9.4.1

bintar BUILDING NOTICE makefiles/Mf-install.in scheme.1.in

c/Makefile.i3nt c/Makefile.a6nt c/Makefile.ti3nt c/Makefile.ta6nt

mats/bldnt.bat workarea c/scheme.rc s/7.ss s/cmacros.ss

release_notes/release_notes.stex csug/csug.stex

- updated newrelease to produce the correct log-entry format and

fixed the existing 9.4.1 log entry

newrelease, LOG

- replaced a couple of tabs

number.c

- updated the descriptions of statistics and related functions to

reflect the Version 9.3.1 change from sstats structures to sstats

records, with sstats times represented as time objects and sstats

bytes and counts represented as exact integers; also updated the

sstats-difference description to reflect that it no longer coerces

negative differences to zero. added a corresponding release note.

system.stex,

release_notes.stex

- added a cast to eliminate a warning

c/number.c

- fixed bug in Windows version of directory-separator-predicate when

path-* procedures are passed a path that is not a string.

s/6.ss

- fixed bug in cp0 on Windows with $foreign-wchar?.

s/cp0.ss

- Cygwin is now used on Windows, updated mats, eliminated unused killme

BUILDING c/*nt c/Mf-base c/scheme.exe.manifest configure

examples/Makefile mats/6.ms mats/Mf-* mats/foreign.ms mats/ftype.ms

mats/patch-* mats/windows.ms s/Mf-*nt s/Mf-base workarea

release_notes.stex

- fixed spelling of non-existent

s/syntax.ss

- now forcing zlib configuration before compiling files that depend on

the zlib header files, since zlib's configure script can modify the

header files. removed ZlibInclude variable, which no longer serves

a purpose.

c/Mf-*, c/Makefile.*nt

- removed unnecessary datestamp.c target

c/Mf.*nt

- fixed unnessesary blocking in expeditor on Windows.

c/expeditor.c

- eliminated a couple of thread-safety issues and limitations on the

sizes of pathnames produced by expansion of tilde (home-directory)

prefixes by replacing S_pathname, S_pathname_impl, and S_homedir

with S_malloc_pathname, which always mallocs space for the result.

one thread-safety issue involved the use of static strings for expanded

pathnames and affected various file-system operations. the other

affected the file open routines and involved use of the incoming

pathname while deactivated. the incoming pathname is sometimes if not

always a pointer into a Scheme bytevector, which can be overwritten if a

collection occurs while the thread is deactivated. the size limitation

corresponded to the use of the static strings, which were limited to

PATH_MAX bytes. (PATH_MAX typically isn't actually the maximum path

length in contemporary operating systems.) eliminated similar issues

for wide pathnames under Windows by adding S_malloc_wide_pathname.

consumers of the old routines have been modified to use the new

routines and to free the result strings. the various file operations

now consistently treat a pathname with an unresolvable home directory

as a pathname that happens to start with a tilde. eliminated unused

foreign-symbol binding of "(cs)pathname" to S_pathname.

io.c, externs.h, new_io.c, prim5.c, scheme.c, prim.c

- various places where a call to close or gzclose was retried when

the close operation was interrupted no longer do so, since this can

cause problems when another thread has reallocated the same file

descriptor.

new_io.c

- now using vcvarsall type x86_amd64 rather than amd64 when the

former appears to supported and the latter does not, as is the

case with VS Express 2015.

c/Mf-a6nt, c/Mf-ta6nt

- commented out one of the thread mats that consistently causes

indefinite delays under Windows and OpenBSD due to starvation.

thread.ms

- increased wait time for a couple of subprocess responses

6.ms

- added call to collector to close files opened during iconv mats

specifically for when mats are run under Windows with no iconv dll.

io.ms

- fixed typo: VC/bin/vcvars64.bat => VC/bin/amd64/vcvars64.bat

c/Mf-a6nt, c/Mf-ta6nt

- scheme_mutex_t now uses volatile keyword for owner and count fields

because these fields can be accessed from multiple threads

concurrently. Updated $yield and $thread-check in mats/thread.ms to

be more tolerant of timing variability.

c/types.h, mats/thread.ms, release_notes.stex

- fixed three instances of unchecked mallocs reported by laqrix in

github issue #77.

io.c, schlib.c, thread.c

- continue the profiler's html output refresh: refine the styling

(and palette) and update CSUG to match. update the CSUG screenshots

to reflect the refined look.

s/pdhtml.ss

csug/system.stex

csug/canned/profilehtml-orig.png

csug/canned/profilehtml.png

csug/canned/fatfibhtml-orig.png

csug/canned/fatfibhtml.png

- add unicode support to the expression editor. entry and display now work

except that combining characters are not treated correctly for

line-wrapping. this addresses github issue #32 and part of issue #81.

c/expeditor.c, s/expeditor.ss

- moved s_ee_write_char function within the WIN32 check to allow the unicode

change to compile on windows. unicode is not yet supported in the windows

version of the repl.

c/expeditor.c

- reworked the S_create_thread_object to print an error and exit when

allocating the thread context fails from Sactivate_thread. before

this change, the error was raised on the main thread, which resulted

in strange behavior at best. also added who argument to

S_create_thread_object to allow it to report either Sactivate_thread

or fork-thread led to the error.

externs.h, schsig.c, scheme.c, thread.c

- fixed a bug in cpvalid resulting in it leaving behind a cpvalid-defer

form for later passes to choke on. also fixed cp0 to print the correct

name for cpvalid when it does this.

cpvalid.ss, cp0.ss,

misc.ms

- updated the prototype for s_ee_write_char to match the definition

expeditor.c

- fixed a side-effect preservation bug with non-trivial test-context

not-like patterns.

cp0.ss,

cp0.ms, 4.ms

- instead of default-exception handler, new-cafe establishes a handler

that calls the current value of base-exception-handler so the handler

can be overridden, as we do in our own make files.

cafe.ss,

7.ms

- fixed a bug in case and exclusive-cond syntax-error calls causing an

exception in syntax-error instead of the intended error message.

syntax.ss

- added tests for the case and exclusive-cond syntax-error calls

4.ms, root-experr-compile-0-f-f-f

- added print-extended-identifiers parameter. when #t, symbols like

1+ and +++ print without escapes.

priminfo.ss, print.ss,

6.ms

- added descriptions of print-extended-identifiers to the user's guide

and release notes. updated the release notes to account for a couple

of other log entries.

release_notes.stex,

intro.stex, io.stex

- updated the sockets example to work with the current version of Chez.

Change the foreign procedure definitions to use the more portable int

rather than integer-32. Switch to a custom port

[make-custom-binary-input/output-port] rather than a generic port

[make-input/output-port], which resulted in deleting quite a bit of

code. Fix various compiler warnings in the C code, and along the way,

fix a signedness bug in c_write that could have resulted in not writing

the full buffer (but reporting that it did) in the case of errors from

write.

examples/csocket.c, examples/socket.ss

- use high-precision clock time on Windows 8 and up

c/stats.c

- fixed profiling code that keyed profiling locations off of only the

bfp to instead key off of both the bfp and efp.

pdhtml.ss

- added Windows installer using the WiX Toolset

BUILDING, install/* (new)

- fix typo in ordinal format for 12

format.ss,

format.ms

- renamed install directory to wininstall to avoid conflict with

top-level Makefile

BUILDING, install/* (removed), wininstall/* (new)

- updated zlib to version 1.2.11

configure

- added procedure-arity-mask to report the allowed argument counts of

a compiled function. On a procedure from interpret or from one of

the trace procedures or syntactic forms, procedure-arity-mask

may report counts that are not actually allowed by the source

procedure.

cmacros.ss, compile.ss, cpnanopass.ss, mkheader.ss, primdata.ss,

prims.ss, strip.ss,

fasl.c, gc.c, globals.h, prim.c, prim5.c, scheme.c, schsig.c,

misc.ms, root-experr*,

objects.stex

- for non-win32 systems, now setting al register to a count of the

floating-point register arguments as required for varargs functions

by the System V ABI.

x86_64.ss,

foreign.ms

- added a missing quote mark in new printf mat Windows case

foreign.ms

- added travis-ci automation script and a 'partialx' testing target to

allow us to perform more than a single run of testing without running

afoul of travis-ci's 50-minute build timeout. 'partialx' tests six

of the twelve configurations tested by 'allx'.

.travis.yml (new),

mats/Mf-base

- paired the 'partialx' down to just four test configurations, with one

interpreter run, to try to get the threaded builds into line with

travis-ci's timeout.

mats/Mf-base

- eliminated some direct assumptions that a vector's type/length field

is a fixnum and added meta-asserts to verify that it is in a couple of

others, to facilitate future changes to vector typing. vectors are

now treated essentially like fxvectors, strings, and bytevectors.

cmacros.ss, cpnanopass.ss, prims.ss, mkheader.ss,

alloc.c, gc.c, scheme.c

- fixed a few comments to refer to scheme.c rather than main.c

externs.h, globals.h, thread.c

- for 64-bit Windows systems, now copying foreign-procedure

double-precision floating-point register arguments to integer

registers as required for varargs functions. Windows does not

support single-precision floating-point arguments as varargs.

foreign.ms, np-languages.ss, x86_64.ss

- added an optional timeout argument to condition-wait

externs.h, stats.c, thread.c, thread.h, csug/threads.stex,

primvars.ms, thread.ms, release_notes.stex,

date.ss, primdata.ss, prims.ss

- added immutable strings, vectors, fxvector, bytevectors, and boxes

5_4.ss, 5_6.ss, bytevector.ss, cmacros.ss, cpnanopass.ss,

fasl.ss, library.ss, mkheader.ss, primdata.ss, prims.ss,

externs.h, types.h, alloc.c, fasl.c, gc.c, scheme.c,

5_5.ms, 5_6.ms, bytevector.ms, misc.ms, root-experr*

objects.stex

- various tweaks to the immutable object support; also taught cp0

to simplify ($fxu< (most-positive-fixnum) e) => (fx< e 0) so we

don't have any incentive in special casing length checks where

the maximum length happens to be (most-positive-fixnum).

5_4.ss, 5_6.ss, bytevector.ss, cmacros.ss, cp0.ss, cpnanopass.ss,

mkheader.ss, primdata.ss, prims.ss,

fasl.c, gc.c, types.h

root-experr*, patch*

- generated bytevector=? procedure now gets out quickly on eq

arguments. cp0 optimizes away a number of additional equality

operations at optimize-level 3 (including bytevector=?) when

applied to the same variable references, as it already did for

eq?, eqv?, and equal?, at all optimize levels.

cpnanopass.ss, cp0.ss, primdata.ss,

cp0.ms

- updated bullyx patches

patch*

- updated release notes and tweaked user's guide.

release-notes.stex, objects.stex

- fixed typo: fxvector-immutable-flag used in place of

bytevector-immutable-flag in computation of type-immutable-bytevector

cmacros.ss

- reallocated typed-object types, using previously unused tag #b010

for strings and giving bytevectors both #b001 and #b101 (the

latter for immutable bytevectors) so that the maximum bytevector

length on 32-bit machines is once again the most-positive fixnum.

treating bytevectors rather than strings or fxvectors (or even

vectors) special in this regard is appropriate since the maximum

number of bytes in a bytevector is maximum-length x 1 rather than

maximum-length x 4 for strings, fxvectors, and vectors on 32-bit

machines. with this change on 32-bit machines, a vector can

occupy up to 1/2 of virtual memory, strings and fxvectors 1/4,

and bytevectors 1/8.

cmacros.ss

- added record-type-equal-procedure, record-type-hash-procedure,

record-equal-procedure, and record-hash-procedure to enable

per-type customization of the behavior of equal? and equal-hash

for a record value

5_1.ss, newhash.ss, primdata.ss,

record.ms, root-experr*,

objects.stex

- adding dropped changes

record.ss,

root-experr*

- added entry for record-type-equal-procedure and friends

release_notes.stex

- changed copyright year to 2017

7.ss, scheme.1.in, comments of many files

- expanded the CSUG description of the handling of command-line

arguments not recognized as standard options and added a description

of the same to the COMMAND-LINE OPTIONS section of the man page.

did a bit of minor cleanup elsewhere in the man page.

use.stex, scheme.1.in

- destroy_thread now processes guardian entries

thread.c, 4.ms, release_notes.stex

- mutexes and conditions are now freed when no longer used via

$close-resurrected-mutexes&conditions, $keep-live primitive added

externs.h, prim5.c, thread.c, 4.ms, thread.ms, release_notes.stex,

7.ss, cpnanopass.ss, np-languages.ss, primdata.ss, prims.ss

- fix reduction for map and for-each with optimization level 3

to drop the expression, check that procedure has the correct

arity and is discardable or unsafe.

Also add a simplification for for-each with empty lists

with optimization level 2.

cp0.ss, 4.ms, primdata.ss

- fix invalid memory reference when enum-set-indexer procedure is not

passed a symbol

enum.ss, enum.ms, root-experr*, release_notes.stex

- fix overflow detection for fxsll, fxarithmetic-shift-left, and

fxarithmetic-shift

library.ss, fx.ms, release_notes.stex

- added ephemeron pairs and changed weak hashtables to use

ephemeron pairs for key--value mapping to avoid the key-in-value

problem

prims.ss, primdata.ss, newhash.ss, fasl.ss, mkheader.ss

cmacro.ss, prim5.c, fasl.c, gc.c, gcwrapper.c, types.h,

4.ms, hash.ms, objects.stex, smgmt.stex, csug.bib

- check_dirty_ephemeron now puts ephemerons whose keys haven't yet

been seen on the pending list rather than the trigger lists.

gc.c

- removed scan of space_ephemeron from check_heap because check_heap

as written can't handle the two link fields properly.

gcwrapper.c

- in the ephemerons mat that checks interaction between mutation and

collection, added generation arguments to the first two collect

calls so they always collect into the intended generation.

4.ms

- updated allx and bullyx patches

patch*

- fix strip-fasl-file for immutable strings and vectors,

fix an $oops call, and fix a vector-index increment in hashing

strip.ss, 7.ss, newhash.ss, misc.ms

- fix signature of fxbit-set?

primdata.ss

- more optimizations for map and for-each with explicit list

extend the reductions for map and for-each when the arguments are

explicit lists like (list 1 2 3 ...) or '(1 2 3 ...).

cp0.ss,

4.ms

- reverted to the preceding version of cp0 due to failure to preserve

the expected evaluation order in one of the mats; removed the

corresponding equivalent-expansion tests.

cp0.ss,

4.ms

- restored the map and for-each optimizations with a fix for the

evaluation-order bug.

cp0.ss,

4.ms

- added date-dst? to access the previously-hidden DST information in

date records, and added date-zone-name to provide a time zone name.

date.ss, primdata.ss,

stats.c,

date.ms, root-experr*, patch-compile*,

system.stex

- fixed a bug in flonum-extractor, which on 64-bit machines was using an

8-byte read instead of a 4-byte read to pick up the 4 highest-order

bytes of a little-endian flonum, potentially reading past the end of

mapped memory for flonums produced by taking the imaginary part of an

inexact complexnum (which, unlike other flonums, are not aligned on

16-byte boundaries). The 8-byte load would also have failed to produce

correct results on 64-bit big-endian machines (of which we presently

have none) because the offsets passed to flonum-extractor assume the

bits are in the lowest-order 4 bytes of the extracted field.

cp0.ss,

misc.ms,

release_notes.stex

- support Windows build on Bash/WSL

BUILDING, configure, workarea, c/vs.bat (new), mats/vs.bat (new),

c/Mf-*nt, mats/Mf-*, s/Mf-base

- fix c/version.h for FreeBDS (machine types i3fb, ti3fb, a6fb, ta6fb)

- fix reference to libc.so to be libc.so.7 for FreeBSD (machine types

i3fb, ti3fb, a6fb, ta6fb)

foreign.ms

- added CC option to configure for selecting the compiler

configure,

c/Mf-*

- Suppress warnings from implicit fallthrough in case labels.

Mf-{a6,arm32,i3,ppc,ta6,ti3,tpp32}le

- added bytevector-compress and bytevector-uncompress

bytevector.ss, primdata.ss, new-io.c, prim5.c, externs.h,

objects.stex, release_notes.stex,

bytevector.ms, root-experr*

- fixed typo in S_abnormal_exit

schsig.c

- don't remove the pariah form in the cp0 pass

cp0.ss,

misc.ms

- revert use of ephemerons in weak hashtables, add ephemeron

hashtables

newhash.ss, hashtable-types.ss, library.ss, primdata.ss,

fasl.ss, fasl.c, gc.c, globals.h,

hash.ms, objects.stex, release_notes.stex

- fixed pariah mat

misc.ms

- minor wordsmithing and fix for an overfull hbox

objects.stex, system.stex

- fix (define-values () ....) to expand to a definition

syntax.ss, 3.ms

- added optional line and column components to a source object, a

locate-source-object-source function that uses the new components,

a current-locate-source-object-source parameter to control looking up

line and column information, a current-make-source-object parameter to

control loation recording, an optional use-cache argument to

locate-source, and a 'source-object message for code and continuation

inspectors

read.ss, syntax.ss, 7.ss, compile.ss, cpnanopass.ss, exceptions.ss,

inspect.ss, primdata.ss, prims.ss, print.ss, cmacros.ss, types.ss,

mat.ss, 8.ms, root-experr*,

syntax.stex, debug.stex, system.stex, release_notes.stex

- fixed broken mats on Windows caused by Bash/WSL changes

7.ms, ftype.ms

- added "ez-grammar" example program

examples/ez-grammar.ss, examples/ez-grammar-test.ss,

examples/Makefile, examples.ms

- updated ez-grammar-test to write temp files to current directory and delete them when finished

examples/ez-grammar-test.ss

- added support for Microsoft Visual Studio 2017 on Windows

BUILDING, c/Mf-a6nt, c/Mf-ta6nt, c/vs.bat,

mats/Mf-a6nt, mats/Mf-ta6nt, mats/ftype.ms

- added support for building Windows installs with Bash/WSL

wininstall/Makefile, candle.bat, light.bat

- added support for building with Visual Studio 2017's BuildTools

c/vs.bat

- check for git before using to get submodules

configure

- fixed windows installer failure when vcredist is not preinstalled by

using the vcredist merge module, split the 32 and 64 bit MSIs and

added a wix bundle to combine the MSIs into a single exe installer,

added a batch script for locating Visual Studio's vcredist merge

modules, updated installer paths and names.

wininstall/*

- fixed np-normalize-context pass to process trivs list in mvset forms

in tail and predicate context and added regression tests. Thanks to

@marcomaggi for reporting the bug and @yjqww6 for providing a

simplified test and finding the initial solution.

cpnanopass.ss,

3.ms

- removed a useless check in foreign-alloc

record.ss

- fix cp0 reduction of fx[+-*]/carry and their signatures

cp0.ss

primdata.ss

fx.ms

- renamed s_gettime => S_gettime to remain consistent with the

convention that the only undocumented externs are prefixed with

S_.

externs.h, stats.c, thread.c

- added version number to scheme.1.in trailer; updated date.

scheme.1.in, newrelease

- removed version update of no-longer-existing bldnt.bat. "fixed"

sed patterns to replace \? with * for the benefit of the deficient

mac sed.

newrelease

9.5 changes:

- updated version to 9.5

bintar BUILDING NOTICE makefiles/Mf-install.in scheme.1.in

c/Makefile.i3nt c/Makefile.a6nt c/Makefile.ti3nt c/Makefile.ta6nt

workarea c/scheme.rc s/7.ss s/cmacros.ss

release_notes/release_notes.stex csug/csug.stex

- updated release notes and fixed user's guide overfull hbox.

release-notes.stex, syntax.stex

- updated install target to do something more sensible

release_notes/Makefile

9.5.1 changes:

- updated version to 9.5.1

bintar BUILDING NOTICE makefiles/Mf-install.in scheme.1.in

c/Makefile.i3nt c/Makefile.a6nt c/Makefile.ti3nt c/Makefile.ta6nt

workarea c/scheme.rc s/7.ss s/cmacros.ss

release_notes/release_notes.stex csug/csug.stex csug/use.stex

examples/ez-grammar-test.ss examples/socket.ss

wininstall/Makefile wininstall/*nt.wxs

- Added setting of CHEZSCHEMELIBDIRS to s and mats make files so that

an existing setting will not interfere with the build process, and

added a note to BUILDING that CHEZSCHEMELIBDIRS should be unset in

Version 9.5 and before.

s/Mf-base, mats/Mf-base, BUILDING

- the $case macro used by r6rs:case and case now unconditionally trims

redundant keys and expands into exclusive-cond rather than cond.

it catches references to => before expanding into exclusive-cond

to avoid supporting => as an undocumented and useless extension

of the case syntax. the r6rs:case and case macros now require

multiple clauses rather than leaving the enforcement to exclusive-cond,

and the exclusive-cond macro now requires multiple clauses rather

than leaving the enforcement to cond.

syntax.ss,

4.ms, root-experr*, patch*

- ifdef'd out include of xlocale.h for glibc, since the glibc

locale.h includes xlocale.h or, in glibc 2.26, its replacement.

expeditor.c

- Updated CSUG to replace \INSERTREVISIONMONTHSPACEYEAR with the current

month and year at the time of generation.

csug.stex, copyright.stex

- Updated configuration to set machine types in the CSUG and release notes

make files, and updated distclean target to remove these files.

configure, makefiles/Makefile-csug.in (renamed from csug/Makefile),

makefiles/Makefile-release_notes.in

(renamed from release_notes/Makefile),

makefiles/Makefile

- added pass-time tracking for pre-cpnanopass passes to compile.

compile.ss

- added inline handler for fxdiv-and-mod

cp0.ss, primdata.ss

- changed order in which return-point operations are done (adjust

sfp first, then store return values, then restore local saves) to

avoid storing return values to homes beyond the end of the stack

in cases where adjusting sfp might result in a call to dooverflood.

cpnanopass.ss, np-languages.ss

- removed unused {make-,}asm-return-registers bindings

cpnanopass.ss

- corrected the max-fv value field of the lambda produced by the

hand-coded bytevector=? handler.

cpnanopass.ss

- reduced live-pointer and inspector free-variable mask computation

overhead

cpnanopass.ss

- moved regvec cset copies to driver so they aren't copied each

time a uvar is assigned to a register. removed checks for

missing register csets, since registers always have csets.

cpnanopass.ss

- added closure-rep else clause in record-inspector-information!.

cpnanopass.ss

- augmented tree representation with a constant representation

for full trees to reduce the overhead of manipulating trees or

subtress with all bits set.

cpnanopass.ss

- tree-for-each now takes start and end offsets; this cuts the

cost of traversing and applying the action when the range of

applicable offsets is other than 0..tree-size.

cpnanopass.ss

- introduced the notion of poison variables to reduce the cost of

register/frame allocation for procedures with large sets of local

variables. When the number of local variables exceeds a given

limit (currently hardwired to 1000), each variable with a large

live range is considered poison. A reasonable set of variables

with large live ranges (the set of poison variables) is computed

by successive approximation to avoid excessive overhead. Poison

variables directly conflict with all spillables, and all non-poison

spillables indirectly conflict with all poison spillables through

a shared poison-cset. Thus poison variables cannot live in the

same location as any other variable, i.e., they poison the location.

Conflicts between frame locations and poison variables are handled

normally, which allows poison variables to be assigned to

move-related frame homes. Poison variables are spilled prior to

register allocation, so conflicts between registers and poison

variables are not represented. move relations between poison

variables and frame variables are recorded as usual, but other

move relations involving poison variables are not recorded.

cpnanopass.ss, np-languages.ss

- changed the way a uvar's degree is decremented by remove-victim!.

instead of checking for a conflict between each pair of victim

and keeper and decrementing when the conflict is found, remove-victim!

now decrements the degree of each var in each victim's conflict

set. while this might decrement other victims' degrees unnecessarily,

it can be much less expensive when large numbers of variables are

involved, since the number of conflicts between two non-poison

variables should be small due to the selection process for

(non-)poison variables and the fact that the unspillables introduced

by instruction selection should also have few conflicts. That

is, it reduces the worst-case complexity of decrementing degrees

from O(n^2) to O(n).

cpnanopass.ss

- took advice in compute-degree! comment to increment the uvars in

each registers csets rather than looping over the registers for

each uvar asking whether the register conflicts with the uvar.

cpnanopass.ss

- assign-new-frame! now zeros out save-weight for local saves, since

once they are explicitly saved and restored, they are no longer

call-live and thus have no save cost.

cpnanopass.ss

- desensitized the let-values source-caching timing test slightly

8.ms

- updated allx, bullyx patches

patch*

- attempt to stabilize timing tests let-values source-caching

test and ephemeron gc test while resensitizing the former

8.ms, 4.ms

- various formatting and comment corrections

workarea,

s/Mf-base, bytevector.ss, cpnanopass.ss, date.ss,

5_6.ms, examples.ms

- updated newrelease to handle mats/Mf-*nt

newrelease mats/Mf-a6nt mats/Mf-i3nt mats/Mf-ta6nt mats/Mf-ti3nt

- fixed gather-filedata's sort of profile entries. for any two

entries x and y in the list produced by the sort call, if x's

bfp = y's bfp, x should come before y if x's efp < y's efp.

The idea is that enclosing entries should always come later

in the list. this affects only languages where two expressions

can start at the same character position.

pdhtml.ss

- expanded capability of ez-grammar with support for simpl

parsing of binary operators w/precedence and associativity

and automatically generated markdown grammar descriptions.

ez-grammar-test.ss now also doubles as a test of pdhtml for

algebraic languages.

mats/examples.ms,

examples/ez-grammar.ss, examples/ez-grammar-test.ss,

examples/Makefile

- maybe-compile-{file,program,library} and automatic import

compilation now treat a malformed object file as if it were

not present and needs to be regenerated. A malformed object

file (particularly a truncated one) might occur if the compiling

processes is killed or aborts before it has a chance to delete

a partial object file.

syntax.ss,

7.ms