Automake "strongly discourages" use of the serial driver, and switching
to the preferred parallel driver allows make -j4 check to run in about
half the time on a four core (not thread) host.
* Makefile.am (TESTS, TESTS_ENVIRONMENT): run ./check-guile from
test-suite/, not here.
* check-guile.in: let test harness handle progress output.
* configure.ac (AM_INIT_AUTOMAKE): allow parallel testing.
* test-suite/Makefile.am (SCM_TESTS): remove non-tests.
(EXTRA_DIST): move non-tests here.
(TESTS): include SCM_TESTS (now driven from here).
(TEST_EXTENSIONS): allow customization for .scm and .test.
(TESTS_ENVIRONMENT): stop defining (user only).
(AM_TESTS_ENVIRONMENT): replaces TESTS_ENVIRONMENT; drop guile.
(SCM_LOG_COMPILER): run .scm tests via meta/guile.
(AM_SCM_LOG_FLAGS): keep --no-auto-compile for .scm tests.
(TEST_LOG_DRIVER): run .test tests via custom automake ./driver.
* test-suite/driver: add automake test driver.
Support an optional --trs-file PATH argument that causes guile-test to
write the status information expected by the automake parallel test
harness to PATH.
In addition, when --trs-file is specified, suppress the final test
summary (via print-counts) since it would be repeated per-test-file when
running in parallel, the automake harness prints its own summary.
cf. https://www.gnu.org/software/automake/manual/html_node/API-for-Custom-Test-Drivers.html
* test-suite/guile-test (main): support --trs-file and --log-file.
* test-suite/test-suite/lib.scm: add count-summary-line.
* test-suite/test-suite/lib/automake.scm: add automake custom test driver.
Return 2 rather than 1 for errors so that 1 will be available for any
future boolean tests (as with say grep).
Direct error message to stderr rather than stdout.
* check-guile.in: send error message to stderr and exit 2 (not 1).
This allows users to adjust their pkg-config install location.
* configure.ac: Call PKG_INSTALLDIR.
* meta/Makefile.am (pkgconfigdir): Remove. It will be defined
automatically by the PKG_INSTALLDIR invocation.
* libguile/r6rs-ports.c (get_bytevector_all_var): New variable.
(init_bytevector_io_vars): New function.
(scm_get_bytevector_all): Rewrite as a proxy to ‘get-bytevector-all’
from (ice-9 binary-port).
* module/ice-9/binary-ports.scm (get-bytevector-all): New procedure.
* NEWS: Update.
Reported-by: Christopher Baines <mail@cbaines.net>
As suggested in
<https://lists.gnu.org/archive/html/guile-devel/2023-06/msg00008.html>.
* module/language/wisp.scm (wisp-uuid): Remove.
(repr-quote, repr-unquote, repr-quasiquote, repr-unquote-splicing)
(repr-syntax, repr-unsyntax, repr-quasisyntax, repr-unsyntax-splicing):
Turn into uninterned symbols.
(line-continues?, chunk-ends-with-period,
line-code-replace-inline-colons): Adjust comparisons accordingly.
(wisp-replace-paren-quotation-repr)[pred]: New procedure.
Use it to compare against the various ‘repr-’ values.
(wisp-make-improper)[dot?]: New procedure.
Use it to compare against ‘repr-dot’.
Fixes <https://bugs.gnu.org/67063>.
* doc/ref/api-io.texi (Venerable Port Interfaces): Bring unread-string
procedure documentation in line with other procedures in the section.
* libguile/ports.c (scm_unread_string): Make port argument optional.
* test-suite/tests/ports.test: Test unread-char and unread-string
without ports.
* NEWS: Update.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
The builtin primitive procedure `error` takes an optional message and
a list of arguments to include into the error message.
These args are formatted with `~S` and appended to the error message, so
that an example call of
`(error "Wrong argument: " 42)`
results in the output
"Wrong argument: 42"
If format strings occur in the message itself, however, they are
escaped. Thus a call like
`(error "Wrong argument: ~a" 42)`
is rendered as
"Wrong argument: ~a 42"
Some callers did not take this behavior into account, leading to
confusing error messages.
Changing the behavior of `error` to be
both backwards-compatible and accept also format strings inside messages
is not straightforward, because it would have to handle escaped `~`
characters as well. Therefore, fix `error` call sites using format
strings to use `format` before calling out to `error`.
The following files are affected:
* module/ice-9/format.scm (format)
* module/ice-9/r6rs-libraries.scm (resolve-r6rs-interface)
* module/oop/goops.scm (make)
* module/srfi/srfi-37.scm (Comment at the beginning of file)
* module/system/base/compile.scm (call-once)
* module/system/repl/command.scm (break, tracepoint)
* module/system/repl/common.scm (repl-default-options)
* module/system/vm/traps.scm (arg-check, trap-at-source-location)
There are a couple of further call sites that were left unchanged,
either because they are using their own `error` procedure:
* module/ice-9/read.scm
* module/ice-9/command-line.scm
or are not referenced from other modules:
* module/system/base/lalr.upstream.scm:
* module/sxml/upstream/assert.scm:
* module/sxml/sxml-match.ss:
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
gen-scmconfig.h is generated in libguile, not $(top_builddir).
* libguile/Makefile.am: Add '-I.' when compiling gen-scmconfig.o.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
Commit 08041d216f attempted to make the "invocation compatible with
BSD sed", but moving '-i' first does not solve the problem because
it still requires to pass an argument. Instead just redirect the
instantiated output into a temporary file and install that.
* libguile/Makefile.am: Remove '-i' from INSTANTIATE.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
Fixes <https://bugs.gnu.org/70144>.
Fixes a bug whereby ‘system*’ would change the handler of SIGINT and
SIGQUIT to SIG_IGN in a racy fashion, possibly competing with calls to
‘sigaction’ in Scheme in another thread.
This is a followup to 527c257d6e, which
witch to ‘posix_spawn’, ensuring signals are properly dealt with when
creating child processes.
* libguile/posix.c (restore_sigaction, scm_dynwind_sigaction): Remove.
(scm_system_star): Remove sigaction dynwind around call to
‘piped_process’.
* NEWS: Update.
Reported-by: Christopher Baines <mail@cbaines.net>
* module/language/tree-il/optimize.scm (make-optimizer): If the keyword
argument #:dump-optimized-tree-il? is present, print the Tree-IL that
will be handed to the next compiler. Also re-enable #:verify-tree-il?.
* libguile/scmsigs.c (scm_i_signals_post_fork): It's possible for
SCM_I_PTHREAD_ONCE_INIT to be a braced initializer. In that case when
this initializer is re-used as an assignment, it needs the type to
instead be treated as a C11 literal. Fixes compilation on Mac OS.
Thanks to Tony Garnock-Jones for the patch.
* lightening/aarch64-cpu.c (SWPAL, CASAL): New instructions.
(swap_atomic, cas_atomic): Use better instructions if we have LSE.
* lightening/aarch64.c (get_hwcap, jit_get_cpu): Arrange to detect LSE
availability on GNU/Linux and Darwin.
Based on a patch by Tony Garnock-Jones. Thanks!
Fixes a regression introduced in
4a0c2433d9: the strategy wouldn’t work
when writing to, say, a bytevector output port.
* module/system/vm/linker.scm (link-elf)[write-padding]: Reintroduce
loop for when PORT is not a file port. Remove first argument.
* libguile/ports.c (scm_seek): Let SEEK_DATA and SEEK_HOLE through.
(scm_init_ice_9_ports): Define ‘SEEK_DATA’ and ‘SEEK_HOLE’.
* module/ice-9/ports.scm: Export ‘SEEK_DATA’ and ‘SEEK_HOLE’ when
defined.
* test-suite/tests/ports.test ("size of sparse file")
("SEEK_DATA while on data", "SEEK_DATA while in hole")
("SEEK_HOLE while in hole"): New tests.
* NEWS: Update.
Since ‘*lcm-page-size*’ is 64 KiB, this saves disk space for small ‘.go’
files.
* module/system/vm/linker.scm (link-elf)[write-padding]: Rewrite in
terms of ‘seek’.
When generating the list of test files, ignore any whose names begin
with a dot. If nothing else, this avoids crashing on the symlinks that
Emacs creates for files with pending changes. In that case it creates a
symlink to nowhere until the content is saved or reverted, and those
symlinks produce test failures like this:
Backtrace:
3 (primitive-load "/home/rlb/src/guile/utf8-debug/test-su?")
In ice-9/eval.scm:
619:8 2 (_ #(#(#(#<directory (guile-user) 7f0239b32c80> #) #) #))
In ice-9/ports.scm:
450:11 1 (call-with-input-file "../../libguile/.#strings.c" #<p?> ?)
In unknown file:
0 (open-file "../../libguile/.#strings.c" "r" #:encoding # ?)
ERROR: In procedure open-file:
In procedure open-file: No such file or directory: "../../libguile/.#strings.c"
FAIL: test-bad-identifiers
* test-suite/standalone/test-bad-identifiers: ignore files with names
beginning with a dot.
Adjust guile-test to ignore any test files with names beginning with a
dot.
If nothing else, this avoids make check and ./check-guile failures when
Emacs has unsaved changes to a test file. In that case Emacs creates a
symlink to nowhere until the content is saved or reverted.
* test-suite/guile-test: ignore test files whose names begin with a dot.
* doc/ref/api-procedures.texi: remove make-binding, binding:name,
binding:boxed?, binding:index, binding:start, and binding:end since they
no longer exist.
cf. 1a2711a848
* module/language/cps/effects-analysis.scm (compute-clobber-map):
Previously a whole-object read would not be clobbered by a specific
field write. This crops up for the &read introduced at the site of
`cons` for the synthetic car and cdr definitions. This error was there
before but didn't cause bugs before 3.0.10 because cons got eagerly
lowered to separate allocation and initialization instructions.
