Remove a mention of "1989" at the end of the 4th paragraph in HISTORY section,
probably leftover because of copy-paste from the paragraph above.
Sort SEE ALSO.
Remove trailing white spaces at some places.
Remove Pp after Sh at few places as pointed out by mandoc.
ok christos@
* NEWS: doc: update NEWS for 2.6.3 release
2016-11-26 Explorer09 <explorer09@gmail.com>
* configure.ac, src/Makefile.am: build: New configure option
'--disable-bootstrap'. If configure is run with '--disable-bootstrap', then stage1flex
won't be built and stage1scan.c will be generated by sed'ing scan.c. This option is intended to workaround bootstrap bugs rather than to
fix the bootstrapping issues which are proving subtler and harder to
fix than anyone would like.
2016-11-26 Explorer09 <explorer09@gmail.com>
* src/Makefile.am: build: "make clean" deletes stage1scan.c,
stage1flex
2016-11-26 Explorer09 <explorer09@gmail.com>
* configure.ac, src/Makefile.am: build: allow building libfl even
with --disable-libfl. For various reasons, we may wish to build libfl explicitly even when
configure has been run with the --disable-libfl option. This is
possible, now, via 'make -C src libfl.la'.
2016-12-19 Alexis La Goutte <alexis.lagoutte@gmail.com>
* src/flex.skl: scanner: remove trailing whitespace in skeleton
2016-12-01 Demi Obenour <demiobenour@gmail.com>
* src/buf.c, src/main.c, src/parse.y: scanner: Disallow, overquote
'[' and ']' in prefix
2016-12-29 Samuel Thibault <samuel.thibault@ens-lyon.org>
* src/main.c: scanner: allocate correct buffer size for m4 path. Flex did not check the length of the m4 path which could lead to a
buffer overflow in some cases. Additionally, not all platforms
believe in PATH_MAX, so stop relying on it. Fixes#138
2016-11-24 Will Estes <westes575@gmail.com>
* NEWS: doc: note no more libcompat in NEWS
2016-11-24 Will Estes <westes575@gmail.com>
* lib/Makefile.am: build: explain empty lib/Makefile.am
2016-11-20 Explorer09 <explorer09@gmail.com>
* configure.ac: build: warn about cross compiling with *alloc
2016-11-19 Explorer09 <explorer09@gmail.com>
* configure.ac, lib/Makefile.am, lib/lib.c, src/Makefile.am: build:
Link $(LIBOBJS) from src/ dir, remove libcompat.la. The libcompat.la library was small and less friendly to
bootstrapping and cross compilation. Now, we will simply link
individual object files as needed, which is simpler.
2016-11-16 Explorer09 <explorer09@gmail.com>
* src/Makefile.am: build: Let stage1flex respect LFLAGS
2016-11-14 Will Estes <westes575@gmail.com>
* NEWS: doc: describe --disable-libfl in NEWS
2016-10-31 Explorer09 <explorer09@gmail.com>
* configure.ac, src/Makefile.am: build: Add --disable-libfl
configure option. Disabling libfl is useful when building flex for a cross-toolchain. Fixes: GH-99
2016-11-09 Demi Obenour <demiobenour@gmail.com>
* src/flex.skl, tests/array_r.l: fix backwards incompatible changes
in 2.6.2. This patch addscompatibility `#defines` for all macros affected by
`%prefix`. Fixes#113.
2016-11-08 Will Estes <westes575@gmail.com>
* tests/.gitignore: git: ignore quote_in_comment artifacts
2016-11-08 Demi Obenour <demiobenour@gmail.com>
* src/scan.l, tests/Makefile.am, tests/quote_in_comment.l,
tests/quote_in_comment.txt: Fixes a major bug in Flex's own lexing
of literals. My changes caused Flex to mishandle string and character literals in
line comments. This commit fixes them. Fixes#113.
2016-11-06 Thomas Klausner <wiz@NetBSD.org>
* configure.ac: Fix unportable test(1) operator. "==" is only supported by bash, "=" is the standard comparison
operator.
2016-11-02 Demi Obenour <demiobenour@gmail.com>
* tests/quotes.l: Add more escaping tests
2016-11-01 Demi Obenour <demiobenour@gmail.com>
* src/scan.l: Fix another escaping bug in non-indented verbatim section 2 code. I also did some reformatting.
2016-10-28 Alastair Hughes <hobbitalastair@gmail.com>
* configure.ac: build: fix false negatives for help2man and texi2dvi HELP2MAN and TEXI2DVI (or the corresponding ac_prog variables) will
never be zero length as they fall back to the missing script; check
for the fall back and warn on that instead of always warning.
2016-10-28 Explorer09 <explorer09@gmail.com>
* doc/Makefile.am: doc: Don't delete flex.1 during "make distclean". flex.1 is pre-generated in release tarball. If we delete it, the
next "configure and make" on the source directory will then require
help2man unnecessarily.
2016-10-27 Alastair Hughes <hobbitalastair@gmail.com>
* NEWS, doc/Makefile.am: Only regenerate the man page when required. Make the flex binary an order-only prerequisite, and add back the
prerequisites from before 7cfb440. This prevents rebuilding the man
page whenever the flex binary is rebuilt, which causes problems if
help2man is not installed and will never work when cross compiling. Fixes#108.
2016-10-26 Will Estes <westes575@gmail.com>
* NEWS, src/Makefile.am: build: no longer build PIC version of
libfl. The PIC version of libfl was not being built correctly. From the
lack of bug reports around this problem, we conclude that the PIC
version of libfl is not used and so we drop it from the build build
targets for flex.
2016-10-25 Explorer09 <explorer09@gmail.com>
* README.md: doc: README.md formatting fixes * Wrap everything in the raw document in 72 char per line limit. * Proper casing for terms "Git" and "GitHub" (don't try to look
lazy). * Add unordered list marks when needed. * Say `configure && make && make install` and quoted for fixed-width font. Signed-off-by: Kang-Che Sung <explorer09@gmail.com>
2016-10-25 Will Estes <westes575@gmail.com>
* NEWS, configure.ac: build: version 2.6.3 begins
2016-10-24 Will Estes <westes575@gmail.com>
* NEWS: doc: mark 2.6.2 release date
2016-10-24 Will Estes <westes575@gmail.com>
* .gitignore, NEWS, configure.ac: build: switch xz to lzip
2016-10-20 Demi Obenour <demiobenour@gmail.com>
* src/scan.l, tests/quotes.l: Fix M4 quotation in section 2 prologue
and refactor duplicated code
2016-10-21 Alex Kennedy <alexzanderkennedy@gmail.com>
* doc/flex.texi: Update flex.texi: Removed repedative wording "Flex used to" was written twice at the beginning of Chapter 9,
paragraph 2.
2016-09-27 Demi Obenour <demiobenour@gmail.com>
* src/dfa.c, src/gen.c, src/main.c, src/nfa.c, src/parse.y,
src/scan.l, tests/bison_yylval_scanner.l,
tests/include_by_push.direct.l: Improved M4 quotation This fixes M4 quotation of certain strings beginning with `yy` (in
section 3 of the input file only) and character literals. The new
quotation method is also less brittle and faster. Tests that relied on the old behavior were fixed. Also, `yyconst` is no longer defined; use `const` (which it
unconditionally was defined to) instead.
2016-10-19 Will Estes <westes575@gmail.com>
* Makefile.am, configure.ac: build: for automake, flex is foreign
2016-10-19 Will Estes <westes575@gmail.com>
* README, README.md: doc: mv README to README.md
2016-10-19 Will Estes <westes575@gmail.com>
* README: doc: touch up README
2016-10-19 Will Estes <westes575@gmail.com>
* README: doc: README no longer mentions sourceforge
2016-10-06 Will Estes <westes575@gmail.com>
* NEWS: doc: update NEWS with more post-2.6.1 additions
2016-09-27 Demi Obenour <demiobenour@gmail.com>
* src/filter.c, src/flexdef.h, src/main.c, src/misc.c,
src/options.c, src/options.h, src/scan.l, src/yylex.c,
tests/Makefile.am, tests/alloc_extra.l, tests/array_r.l,
tests/basic_r.l, tests/c_cxx_nr.lll, tests/c_cxx_r.lll,
tests/debug_r.l, tests/include_by_reentrant.direct.l,
tests/lineno_r.l, tests/mem_nr.l, tests/mem_r.l, tests/posix.l,
tests/posixly_correct.l, tests/pthread.l, tests/quotes.l,
tests/reject.l4, tests/rescan_nr.direct.l, tests/rescan_r.direct.l,
tests/string_nr.l, tests/string_r.l: Fix M4 quoting of section 3. This fixes M4 quoting of section 3 of the input file, including
escape sequences and character constants. Tests were added to verify the behavior in section 3 with respect to
quoting. Both escaping of quotes and quoting of potential
macro-start characters are tested. Existing tests were also fixed to
account for the new -- and now correct -- behavior. Many tests
relied on the old behavior of expanding M4 macros in section 3.
They needed to be updated for the new behavior.
2016-10-03 Mightyjo <mightyjo@gmail.com>
* src/flex.skl, tests/.gitignore, tests/Makefile.am,
tests/cxx_restart.ll, tests/cxx_restart.txt: Fix yyrestart(NULL)
SEGV.
2016-09-05 Demi Obenour <demiobenour@gmail.com>
* src/scan.l: scanner: M4 quoting fixes
2016-09-05 Demi Obenour <demiobenour@gmail.com>
* src/Makefile.am: Support `make indent` for out of source builds
2016-09-24 Demi Obenour <demiobenour@gmail.com>
* src/Makefile.am, src/mkskel.sh: Simplify some shell code
2016-09-23 Demi Obenour <demiobenour@gmail.com>
* doc/flex.texi, src/flex.skl, src/flexdef.h, src/main.c,
src/options.c, src/options.h, src/scan.l, tests/.gitignore,
tests/Makefile.am, tests/noansi_nr.l, tests/noansi_nr.txt,
tests/noansi_r.l, tests/noansi_r.txt: no longer generate K&R C
scanners
2016-09-04 Demi Obenour <demiobenour@gmail.com>
* src/scan.l: Fix escaping of `[[` and `]]` in strings Previously, `[[` and `]]` were not escaped in strings, which led to
bad interactions with m4. Also, don't break strings on newline, as GCC et al support
whitespace between a backslash and the subsequent newline.
2016-06-20 Translation Project <coordinator@translationproject.org>
* NEWS, po/sv.po: new sv translation
2016-05-23 Demetri Obenour <demiobenour@gmail.com>
* src/misc.c: Delete action_m4_define from misc.c This function was not used and always triggered a fatal error when
run.
2016-05-20 Will Estes <westes575@gmail.com>
* tests/Makefile.am, tests/testwrapper-direct.sh: test: run direct
tests from srcdir
2016-01-11 Egor Pugin <egor.pugin@gmail.com>
* tests/include_by_buffer.direct.l, tests/include_by_push.direct.l,
tests/include_by_reentrant.direct.l: Exited with error code on some
conditions in include tests
2016-05-20 Will Estes <westes575@gmail.com>
* tests/testwrapper.sh: test: drop .exe when making input file names
2016-05-20 Will Estes <westes575@gmail.com>
* tests/Makefile.am: test: removed dependencies for include tests
2016-04-28 Tobias Klauser <tklauser@distanz.ch>
* src/scanflags.c: Avoid realloc on every call of sf_push() Currently, every call to sf_push() realloc()'s _sf_stack, even if
the maximum size _sf_max wasn't changed. As the indentation beneath
the "if" clause already indicates, the realloc() should only be
executed if _sf_max was increased. Found by compiling flex with the -Wmisleading-indentation flags of
gcc, which leads to the following warning: scanflags.c: In function âsf_pushâ: scanflags.c:42:5: warning:
this âifâ clause does not guard... [-Wmisleading-indentation] if
(_sf_top_ix + 1 >= _sf_max) ^~ scanflags.c:44:9: note: ...this
statement, but the latter is misleadingly indented as if it is
guarded by the âifâ _sf_stk = realloc(_sf_stk,
sizeof(scanflags_t) * _sf_max); ^~~~~~~
2016-03-31 Will Estes <westes575@gmail.com>
* doc/Makefile.am: build: prettified command to build man page
2016-03-31 Will Estes <westes575@gmail.com>
* doc/Makefile.am: build: added man page to DISTCLEANFILES
2016-03-31 Will Estes <westes575@gmail.com>
* BUGS, TODO: remove unused TODO, BUGS files
2016-03-31 Will Estes <westes575@gmail.com>
* to.do/README, to.do/Wilhelms.todo, to.do/Wish-List,
to.do/flex.rmail, to.do/streams.mail, to.do/unicode/FlexLexer.h,
to.do/unicode/ccl.c, to.do/unicode/changes.txt,
to.do/unicode/ecs.c, to.do/unicode/flex.1, to.do/unicode/flex.skl,
to.do/unicode/flexdef.h, to.do/unicode/gen.c, to.do/unicode/main.c,
to.do/unicode/misc.c, to.do/unicode/scan.l, to.do/unicode/tblcmp.c:
removed to.do/ directory
2016-03-31 Will Estes <westes575@gmail.com>
* po/POTFILES.in: gettext: removed flex.skl from POTFILES.in file
list
2016-03-31 rlar <rlar>
* src/Makefile.am: fix skel.c dependency (missing tables_shared.c)
and polish build rule Also remove references to srcdir since skel.c is no longer mentioned
in po/POTFILES.in.
2016-03-31 rlar <rlar>
* doc/Makefile.am: fix flex.1 dependency git clean -fdx && ./autogen.sh && \ mkdir -p ../build && cd ../build && ../flex/configure && make -j10
&& \ make dist failed with: > help2man: can't get `--help' info from ../../flex/src/flex Note: There is no dependency except for the flex binary.
2016-03-31 rlar <rlar>
* src/Makefile.am: fix stage1scan.c and stage1scan.l dependency git clean -fdx && ./autogen.sh && \ mkdir -p ../build && cd ../build && ../flex/configure && make -j10 failed with: > ../src/stage1flex -o stage1scan.c stage1scan.l > stage1flex: can't
open stage1scan.l Note: stage1scan.c is not necessairy in the "make dist" generated tar.gz
file. stage1flex will be build from scan.c (which is
distributed), and this will then generate stage1scan.c from scan.l
2016-03-31 Will Estes <westes575@gmail.com>
* po/POTFILES.in: gettext: list src/flex.skl in po/POTFILES.in
2016-03-29 rlar <rlar>
* src/gen.c: warning in generated code, with -Ca warning: conversion to 'yy_state_type' from 'flex_uint32_t' may
change the sign of the result [-Wsign-conversion]
2016-03-29 rlar <rlar>
* tests/Makefile.am: suppress `WARNINGFLAGS' for the almost obsolete
`noansi' test cases
2016-03-29 rlar <rlar>
* configure.ac, src/Makefile.am, tests/Makefile.am: configure option
`--enable-warnings' and `WARNINGFLAGS' `WARNINGFLAGS' can be passed when invoking `configure' and when invoking `make' if configure switch `--enable-warnings' was given then default to something useful if we have `GCC' `WARNINGFLAGS' is not used when compiling `stage1flex' to avoid unnecessary clutter
2016-03-31 Tobias Klauser <tklauser@distanz.ch>
* src/main.c: Fix potential buffer overflow in strncat() When using clang/llvm 3.8 to compile flex, the following warning is
emitted: main.c:378:27: warning: the value of the size argument in 'strncat'
is too large, might lead to a buffer overflow [-Wstrncat-size]
strncat(m4_path, m4, sizeof(m4_path));
^~~~~~~~~~~~~~~
main.c:378:27: note: change the argument to be the free space in the
destination buffer minus the terminating null byte strncat(m4_path,
m4, sizeof(m4_path)); ^~~~~~~~~~~~~~~
sizeof(m4_path) - strlen(m4_path) - 1Fix it up by using the solution proposed by the warning message.
2016-03-29 Robert.Larice Robert Larice <Robert.Larice@t-online.de>
* src/Makefile.am: build: simplified dependency tracking so parallel
make runs succeed
2016-03-20 rlar <rlar>
* src/flex.skl, src/gen.c, src/main.c: avoid warning in generated
code, with -Cf warning: conversion to 'unsigned int' from 'int' may change the sign
of the result [-Wsign-conversion]
2016-03-18 Will Estes <westes575@gmail.com>
* NEWS, configure.ac: mention v2.6.2; summarize changes since 2.6.1
2016-03-18 Will Estes <westes575@gmail.com>
* doc/flex.texi: doc: corrected example in manual, gh#67
2016-03-16 rlar <rlar>
* tests/string_nr.l, tests/string_r.l, tests/yyextra.l: warning:
conversion to 'size_t' from 'int' may change the sign of the result
[-Wsign-conversion]
2016-03-16 rlar <rlar>
* tests/mem_nr.l, tests/mem_r.l, tests/pthread.l: warning:
conversion to 'long unsigned int' from 'int' may change the sign of
the result [-Wsign-conversion]
2016-03-16 rlar <rlar>
* src/flex.skl: warning: conversion to 'flex_uint32_t' from 'long
unsigned int' may alter its value [-Wconversion] struct yytbl_reader, member bread is of type flex_uint32_t
2016-03-16 rlar <rlar>
* tests/alloc_extra.l, tests/bison_nr_main.c,
tests/bison_nr_parser.y, tests/bison_yylloc_parser.y: warning:
redundant redeclaration of '...' [-Wredundant-decls]
2016-03-16 rlar <rlar>
* tests/bison_nr_parser.y, tests/bison_yylloc_parser.y,
tests/bison_yylval_parser.y, tests/pthread.l: warning: no previous
prototype for '...' [-Wmissing-prototypes]
2016-03-16 rlar <rlar>
* tests/multiple_scanners_r_main.c: warning: unused variable 'fp'
[-Wunused-variable]
2016-03-16 rlar <rlar>
* tests/pthread.l: main(), warning: redundant redeclaration of
'main' [-Wredundant-decls]
2016-03-16 rlar <rlar>
* tests/include_by_reentrant.direct.l, tests/lineno_nr.l,
tests/lineno_r.l, tests/lineno_trailing.l: main(), warning:
old-style function definition [-Wold-style-definition]
2016-03-16 rlar <rlar>
* tests/alloc_extra.l, tests/array_nr.l, tests/array_r.l,
tests/basic_nr.l, tests/basic_r.l, tests/ccl.l, tests/debug_nr.l,
tests/debug_r.l, tests/extended.l, tests/mem_nr.l, tests/mem_r.l,
tests/noansi_r.l, tests/posix.l, tests/posixly_correct.l,
tests/prefix_nr.l, tests/prefix_r.l, tests/quotes.l,
tests/string_nr.l, tests/string_r.l, tests/yyextra.l: main(),
warning: old-style function definition [-Wold-style-definition]
2016-03-01 rlar <rlar>
* src/ecs.c, src/tblcmp.c: Fix two more casts
2016-03-14 Bastian Köcher <git@kchr.de>
* src/gen.c: Fixes yyl compare with unsigned warning
2016-03-13 Robert Larice <Robert.Larice@t-online.de>
* configure.ac: Suppress portability warnings in Makefile generation
2016-02-27 rlar <rlar>
* src/flex.skl: generated code, in yyensure_buffer_stack(), change
type of local `num_to_alloc' which is exclusively used in yy_size_t context
2016-03-01 rlar <rlar>
* src/flex.skl: generated code, in yy_get_next_buffer(), change type
of local `number_to_move' suits better, because `yy_n_chars' and `yy_buf_size' are of type
`int'
2016-02-27 rlar <rlar>
* src/flex.skl: generated code, `_yybytes_len' is of type `int', fix
code accordingly
2016-02-28 rlar <rlar>
* src/gen.c: generated code, `max_size' seems to be of type `int',
fix casts accordingly
2016-02-28 rlar <rlar>
* src/flex.skl: generated code, here `new_size' is of type `int',
fix casts accordingly
2016-02-28 rlar <rlar>
* src/flex.skl: generated code, `yy_buf_size' is of type `int', fix
casts accordingly
2016-02-28 rlar <rlar>
* src/flex.skl: generated code, `offset' is of type `int'
2016-02-28 rlar <rlar>
* src/gen.c: generated code, `yy_more_len' is of type `int'
2016-02-28 rlar <rlar>
* src/scan.l: scan.l, rewrite two loops to avoid unneccesairy
casting
2016-03-01 rlar <rlar>
* src/regex.c, src/tables.c: improve readability
2016-02-28 rlar <rlar>
* src/tblcmp.c: another cast in tblcmp.c to avoid warning
2016-02-28 rlar <rlar>
* src/buf.c: casts in buf_append() to get rid of warnings
2016-02-28 rlar <rlar>
* src/dfa.c, src/gen.c: cast to suite type of flex_uint32_t td_lolen
2016-02-28 rlar <rlar>
* src/main.c, src/misc.c, src/regex.c, src/scan.l, src/scanopt.c,
src/tables_shared.c: cast to get rid of warnings
2016-03-01 rlar <rlar>
* src/buf.c, src/main.c: cast and fix usage of log10(), ceil to
prevent buffer overflow
2016-03-01 rlar <rlar>
* src/tables.c: tables.c, sprinkle casts to get rid of warnings
2016-03-01 rlar <rlar>
* src/tables.c: yytbl_data_compress(), change type of local newsz to
get rid of warnings
2016-03-01 rlar <rlar>
* src/tables.c, src/tables.h: change type of struct
yytbl_writer.total_written to get rid of warnings
2016-02-28 rlar <rlar>
* src/tables.c: change argument type of yytbl_writen() to get rid of
warnings
2016-03-01 rlar <rlar>
* src/tables.c: yytbl_write8/16/32(), change type of local variables
to get rid of warnings
2016-02-28 rlar <rlar>
* src/flexdef.h, src/misc.c, to.do/unicode/flexdef.h,
to.do/unicode/misc.c: change return type and rename int
htoi()/otoi() --> unsigned int htoui()/otoui()
2016-03-01 Will Estes <westes575@gmail.com>
* NEWS: Mention 2.6.1 release date
2016-02-28 rlar <rlar>
* src/gen.c: avoid warning, add (int) cast to the read() return
value For similiarity with the fread() case.
2016-02-28 rlar <rlar>
* src/scan.l: avoid warning, POSIX says yyless() has an `int'
argument
2016-02-29 rlar <rlar>
* src/filter.c: use type size_t in filter_tee_header() to avoid
warnings
2016-02-29 rlar <rlar>
* src/filter.c, src/misc.c, src/scanopt.c, src/tables.c: add
(size_t) casts to malloc invocations to prevent warnings
2016-02-29 rlar <rlar>
* src/buf.c, src/misc.c, src/scan.l, src/scanopt.c, src/tables.c:
add (int) casts to some strlen() invocations to prevent warnings
2016-02-29 rlar <rlar>
* src/flexdef.h, src/scan.l, src/sym.c: ndlookup(), char *, to get
rid of casts and warnings
2016-02-28 rlar <rlar>
* src/flexdef.h, src/scan.l, src/sym.c: ndinstal(), char *, to get
rid of casts and warnings
2016-02-28 rlar <rlar>
* src/flexdef.h, src/scan.l, src/sym.c: cclinstal() and ccllookup(),
char *, to get rid of casts and warnings
2016-02-28 rlar <rlar>
* src/gen.c: warning: redundant redeclaration of
âgen_next_stateâ [-Wredundant-decls]
2016-02-28 rlar <rlar>
* src/gen.c: warning: no previous prototype for âmkecstblâ
[-Wmissing-prototypes]
2016-02-28 rlar <rlar>
* src/main.c: warning: suggest parentheses around assignment used as
truth value [-Wparentheses]
2016-02-28 rlar <rlar>
* src/flexdef.h: warning: redundant redeclaration of âyywrapâ
[-Wredundant-decls]
2016-02-28 rlar <rlar>
* src/flexdef.h: warning: redundant redeclaration of âyylvalâ
[-Wredundant-decls]
2016-02-28 rlar <rlar>
* src/flexdef.h: warning: redundant redeclaration of âyyinâ
[-Wredundant-decls]
2016-02-28 rlar <rlar>
* src/flexdef.h, src/main.c: warning: redundant redeclaration of
âyyparseâ [-Wredundant-decls]
2016-02-28 rlar <rlar>
* src/parse.y, src/scan.l, src/yylex.c, to.do/unicode/scan.l:
unification, rename some more rename these too for improved similiarity: OPTION_OP OPT_HEADER OPT_EXTRA_TYPE OPT_TABLES
2016-02-28 rlar <rlar>
* src/parse.y, src/scan.l, src/yylex.c, to.do/unicode/scan.l: fix
name clash, OPT_OUTFILE from parse.y and from options.h these collide: OPT_OUTFILE OPT_PREFIX OPT_YYCLASS rename them TOK_... in the
parser
2016-02-28 Will Estes <westes575@gmail.com>
* lib/reallocarray.c: Prototyped reallocarray implementation
2016-02-28 Will Estes <westes575@gmail.com>
* src/scanopt.c: Removed custom strcasecmp() function
2016-02-28 rlar <rlar>
* src/flex.skl: fwrite wants a size_t, yyleng is int per posix
2016-02-28 rlar <rlar>
* src/scanopt.c: Remove some unneeded casts
2016-02-28 rlar <rlar>
* src/flexdef.h: warning: negative integer implicitly converted to
unsigned type [-Wsign-conversion]
2016-02-27 rlar <rlar>
* src/dfa.c, src/parse.y: dfa.c:157:24: warning: conversion to
'size_t' from 'int' may change the sign of the result
[-Wsign-conversion] most certainly safe cast
2016-02-27 rlar <rlar>
* src/ccl.c: ccl.c:86:19: warning: conversion to 'unsigned char'
from 'int' may alter its value [-Wconversion] ch seems to have been checked for proper range some lines above
2016-02-27 rlar <rlar>
* tests/bison_nr_parser.y: warning: deprecated directive, use
'%pure-parser' [-Wdeprecated]
2016-02-27 rlar <rlar>
* lib/lib.c: warning: no previous prototype for 'do_nothing'
[-Wmissing-prototypes]
2016-02-27 rlar <rlar>
* src/libmain.c: included stdlib.h header
2016-02-27 Will Estes <westes575@gmail.com>
* NEWS: Described more post v2.6.0 changes
2016-02-27 Will Estes <westes575@gmail.com>
* configure.ac: build: Removed bzip2 distribiution archive
2016-02-27 Will Estes <westes575@gmail.com>
* configure.ac: removed obsolete program check
2016-02-27 Will Estes <westes575@gmail.com>
* configure.ac: Made some program checks more robust
2016-02-27 rlar <rlar>
* src/flex.skl: Remove unneeded cast to int
2016-02-27 Will Estes <westes575@gmail.com>
* src/flex.skl: Fixed incorrect integer type
2016-02-27 Will Estes <westes575@gmail.com>
* src/flex.skl: Fix more integer types, resolves sf 184, 187
2016-02-26 Robert Larice <Robert.Larice@t-online.de>
* src/flex.skl, src/gen.c: Removed some type conversion warnings
2016-02-26 Will Estes <westes575@gmail.com>
* src/flex.skl: Changed another buffer size to int; resolves gh#61
2016-02-24 Will Estes <westes575@gmail.com>
* src/flex.skl: Changed type of yy_n_chars to int; gh#53, sf#160. The variable yy_n_chars had been of type yy_size_t which is
incorrect given its use in read(). While it might be adviseable to
look at defining a yy_ssize_t, there might be some issues doing this
and so, for now, at least, we'll punt back to int.
2016-02-24 Will Estes <westes575@gmail.com>
* src/buf.c: Fixed size of bufferallocation, resolved gh#54. The value of n_alloc was a count, not a size. Multiplying the value
by the element size was incorrect. That multiplication was already
being done and having it done twice was incorrect.
2016-02-23 Tobias Klauser <tklauser@distanz.ch>
* src/scan.l: Allow '%option noline' in flex input file, resolves
gh#56. Allow specifying '%option noline' in the input file, leading to the
same effect as calling flex with the command line option --noline. Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
2016-02-23 Tobias Klauser <tklauser@distanz.ch>
* src/buf.c, src/main.c: Emit no #line directives if gen_line_dirs
is false, resolves igh#55. There are two instances in the code which will print a #line
directive to the resulting lexer, regardless of the value of
gen_line_dirs. Fix them, so they also respect gen_line_dirs. Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
2016-02-16 Tobias Klauser <tklauser@distanz.ch>
* lib/lib.c, src/gen.c, src/misc.c, src/scan.l: Converted K&R style
function definitions to ANSI C style Consistently make use of the ANSI C function definition style
instead of the K&R style.
2016-01-29 Tobias Klauser <tklauser@distanz.ch>
* src/flex.skl: Used NULL constant instead of plain integer for NULL
pointer. The sparse static checker warns about using plain integer 0 as NULL
pointers in the generated lexer code. Fix this by using NULL
consistently for pointers.
2016-01-29 Tobias Klauser <tklauser@distanz.ch>
* src/flex.skl: Marked declaration and definition of yy_fatal_error
as noreturn. Only the declaration of yy_fatal_error is marked with
__attribute__((__noreturn__)) in case GCC >= 3 is used, but not the
definition. This leads to the sparse static checker to complain
about function declaration mismatch. Fix it by defining a macro yynoreturn and using it for both the
declaration and the definition of yy_fatal_error.
2016-01-27 Tobias Klauser <tklauser@distanz.ch>
* src/flex.skl: Fixed declaration mismatch in yy_fatal_error. The prototype declares yy_fatal_error parameter as "const char
msg[]" while the definition uses "const char* msg" (introduced by
commit e9d5fc713f61b) which causes the sparse static checkers to
produce an error. Fix this by adjusting the definition to use "const char* msg" as
well. Also change the C++ version accordingly so it matches the
declaration in FlexLexer.hpp.
2016-01-23 Will Estes <westes575@gmail.com>
* CODE_OF_CONDUCT.md: flex is for flex #NCoC
2016-01-18 Will Estes <westes575@gmail.com>
* doc/flex.texi: Referred to github for issue tracking, no longer sf
2016-01-10 Egor Pugin <egor.pugin@gmail.com>
* tests/reject.l4, tests/tableopts.l4: Opened files in binary mode
explicitly
2016-01-08 OBATA Akio <obache@NetBSD.org>
* src/Makefile.am: Linked flex binary against libintl, not libfl. Signed-off-by: Thomas Klausner <wiz@NetBSD.org>
2016-01-08 Michael van Elst <mlelstv@NetBSD.org>
* src/filter.c: Improved pipe-stdin hack behavior; resolves sf#198. Signed-off-by: Thomas <Klausner wiz@NetBSD.org>
2015-12-27 Will Estes <westes575@gmail.com>
* configure.ac, src/flexdef.h: Removed no longer needed header
checks
2015-12-27 Will Estes <westes575@gmail.com>
* configure.ac: Checked for reallocarray() with AC_REPLACE_FUNCS
2015-12-27 Will Estes <westes575@gmail.com>
* src/flexdef.h, src/main.c: include libgen.h from flexdef.h, not
main.c
2015-12-25 Michael Reed <m.reed@mykolab.com>
* src/main.c: Replace basename2() with basename(3). Given the following program: \#include <libgen.h> \#include <stdio.h> /* extracts basename from path, optionally stripping the extension
"\.*" * (same concept as /bin/sh `basename`, but different handling
of extension). */ static char *basename2 (char *path) { char *b; for (b = path; *path; path++) if (*path == '/') b = path + 1; return b; } static void basename_compare(char *path) { printf("basename: %s\n", basename(path)); printf("basename2: %s\n\n", basename2(path)); } int main (int argc, char *argv[]) { // From http://pubs.opengroup.org/onlinepubs/9699919799/ // ``Sample Input and Output Strings'' basename_compare("/usr/lib"); basename_compare("/usr/"); basename_compare("/"); basename_compare("///"); basename_compare("//usr//lib//"); return 0; } ... and the program's output: basename: lib basename2: lib basename: usr basename2: basename: / basename2: basename: / basename2: basename: lib basename2: ... we can see that basename2() behaves the same as basename(3) in
the average use case, but messes up pretty severely in others.
Besides that, basename(3) is mandated by POSIX so should be present
on modern Unix-like systems, so we shouldn't define it ourselves. Some notes: - it doesn't appear to be mentioned in POSIX, but OpenBSD's
basename(3) returns NULL if the returned path componenet is >
PATH_MAX, so add a check for that - basename(3) shouldn't return an empty string, so remove the program_name[0] != '\0' check
2015-12-25 Michael Reed <m.reed@mykolab.com>
* src/main.c: Simplify basename2(). It's only call site does not activate the `strip_ext` code path, so
the function can be simplified a lot. While here, remove a double
assignment.
2015-12-25 Michael Reed <m.reed@mykolab.com>
* src/flex.skl, src/misc.c: Cleaned up __STDC__ #ifdefs. Assuming a compiler conforming to the ISO C standard is used, i.e.,
__STDC__ is defined to 1, YY_USE_CONST is always defined and can be
eliminated.
2015-12-25 Michael Reed <m.reed@mykolab.com>
* src/flexdef.h, src/main.c, src/scanopt.c, src/scanopt.h,
src/sym.c: Remove remaining use of PROTO
2015-12-21 Serguey Parkhomovsky <xindigo@gmail.com>
* src/buf.c: buf.c: use snprintf
2015-12-19 Will Estes <westes575@gmail.com>
* configure.ac: build: reformatted AC_CHECK_FUNCS for readability
2015-12-17 Will Estes <westes575@gmail.com>
* src/scanopt.c: correct function prototype
2015-12-15 Michael Reed <m.reed@mykolab.com>
* src/main.c, src/scanopt.c, src/sym.c, src/tblcmp.c: Remove more
instances of PROTO
2015-12-15 Michael Reed <m.reed@mykolab.com>
* src/main.c: Removed prototype for main(). It's not called anywhere else so the prototype is not needed. See
the C99 standard [1], section 5.1.2.2.1 for more info. [1]: http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1256.pdf
2015-12-13 Michael McConville <mmcco@mykolab.com>
* src/dfa.c, src/gen.c, src/nfa.c, to.do/unicode/main.c,
to.do/unicode/tblcmp.c: Removed more instances of PROTO, ansifying.
2015-12-13 Will Estes <westes575@gmail.com>
* README: Noted github for issue tracking and pull requests
2015-12-13 Michael McConville <mmcco@mykolab.com>
* README: Reformatted README; removed sf bug tracking link.
2015-12-13 Michael McConville <mmcco@mykolab.com>
* src/gen.c, to.do/unicode/gen.c: Removed macros for indentation
level. Just increment or decrement the indentation counter. That's less to
remember and makes the code more readable. w# Please enter the commit message for your changes. Lines starting
2015-12-12 Mike Frysinger <vapier@gentoo.org>
* tests/Makefile.am, tests/testwrapper.sh: tests: fixed paths to
input files. The current test wrapper works only when the inputs are specified
using relative paths. If they're specified with absolute paths, the
driver fails to detect the inputs because it always prepends the
input dir name which itself is a relative path: $ cd tests $
./testwrapper.sh -d . -i $PWD/reject.txt -t ./reject_ver.table
<fails to open inputs> This normally doesn't show up because people run `./configure` or,
for out of tree builds, `../configure`. But if you happen to run
configure with an absolute path, then automake tends to generate
absolute paths as well leading to test failures. Fix all of this by dropping the implicit input directory prepending. - INPUT_NAME is often a list of files, not just a single one - the input directory is used to find the testname tables which are usually generated, so it's impossible to use files from both
source and build directories - most of the time, the full/correct path is already specified
2015-12-12 Mike Frysinger <vapier@gentoo.org>
* configure.ac: configure: fixed realloc test. The [] characters are used for quoting in m4, so the attempt to use
them in place of `test` fails yielding the warning at build time: .../flex/configure: line 20222: no: command not found
2015-12-12 Michael McConville <mmcco@mykolab.com>
* src/flexdef.h: Started removal of PROTO macro. The PROTO macro is no longer needed. Additionally, its usage is
inconsistent, so we began removing it.
2015-12-12 Michael Reed <m.reed@mykolab.com>
* Makefile.am, src/main.c: Removed SHORT_FILE_NAMES preprocessor
symbol. As a relic of MS-DOS, we don't need this. It's never defined; see 13b5b214f53d1c3354a7ab910bd160c126df1331. Removed additional MSDOS ifdef.
2015-12-11 Michael McConville <mmcco@mykolab.com>
* configure.ac, lib/reallocarray.c, src/flexdef.h: Added new
function reallocarray. This is taken from OpenSSH Portable, which in turn takes it from
OpenBSD. reallocarray wraps the stdlib's realloc function. It takes two size
arguments and checks for overflow, like calloc, but doesn't zero the
memory. Therefore, it allows us to do overflow-safe array
reallocations and overflow-safe unzeroed array allocations, which
the stdlib allocation functions don't. We have a bunch of specific array allocation macros, none of which
check for overflow. reallocarray should be able to replace them.
2015-12-11 Michael McConville <mmcco@mykolab.com>
* src/flexdef.h: Removed MS-DOS, VMS macros. Given the age of the MS-DOS and VMS platforms, it's likely that no
one is building flex on them any more. Additionally, the preferred
approach is to test for particular platform features rather than to
test for particular platforms.
2015-12-12 Will Estes <westes575@gmail.com>
* src/flexdef.h, src/main.c: Made search for m4 more explicit.
2015-12-12 Will Estes <westes575@gmail.com>
* src/flex.skl: Returned 0 from yywrap() instead of EOF
2015-12-11 Will Estes <westes575@gmail.com>
* src/libmain.c: Changed end of main() in libmain to exit(0)
2015-12-11 Will Estes <westes575@gmail.com>
* NEWS, configure.ac: Mentioned v2.6.1; documented some changes
since v2.6.0
2015-12-11 Will Estes <westes575@gmail.com>
* README: Updated build documentation; finished sf#155. Removed version numbers for build tools. Noted that version
requirements for build tools will be noted in configure.ac. Expanded
documentation of building texinfo based docs.
2015-12-11 Will Estes <westes575@gmail.com>
* doc/Makefile.am: Removed flex.pdf from distribution; partially
fixed sf#155
2015-12-11 Will Estes <westes575@gmail.com>
* src/flex.skl: Commented in C style in skeleton; fixed sf#195
2015-12-10 Will Estes <westes575@gmail.com>
* .gitignore: Ignored autoscan files
2015-12-09 Will Estes <westes575@gmail.com>
* configure.ac: Checked for ranlib and strdup() at configure time
2015-12-09 Michael McConville <mmcconville@mykolab.com>
* src/flexdef.h, src/misc.c, src/tblcmp.c, to.do/unicode/flexdef.h,
to.do/unicode/misc.c, to.do/unicode/tblcmp.c: Replace zero_out()
with stdlib's memset.
2015-12-09 Michael McConville <mmcconville@mykolab.com>
* src/flex.skl: Removed memory allocation casts.
2015-12-09 Michael McConville <mmcconville@mykolab.com>
* src/flexdef.h: Removed alloca() configuration. Since alloca() is dangerous, depricated, we remove it.
2015-12-09 Michael McConville <mmcconville@mykolab.com>
* src/flexdef.h: Removed implementation of isascii(). POSIX defines isascii(), so it's likely present on anything
weactually build flex on these days.
2015-12-09 Michael McConville <mmcconville@mykolab.com>
* src/main.c, src/scanopt.c, src/tables.c: Removed NULL-checks
before free()
2015-12-08 Michael McConville <mmcconville@mykolab.com>
* src/buf.c, src/dfa.c, src/flexdef.h, src/gen.c, src/scan.l,
to.do/unicode/flexdef.h, to.do/unicode/scan.l: Removed flex_free()i,
corrected buf_destroy logic. As with flex_alloc(), replace with direct calls to free(). The function buf_destroy is now null safe and the logic was
corrected to free() correctly.
2015-12-08 Michael McConville <mmcconville@mykolab.com>
* src/filter.c, src/flexdef.h, src/misc.c, src/scan.l,
src/scanflags.c, to.do/unicode/flexdef.h, to.do/unicode/misc.c,
to.do/unicode/scan.l: Removed flex_realloc(). As with flex_alloc(), replace calls to flex_realloc(), which was
just a wrapper around realloc().
2015-12-08 Michael McConville <mmcconville@mykolab.com>
* src/buf.c, src/filter.c, src/flexdef.h, src/main.c, src/misc.c,
src/regex.c, src/scan.l, src/scanflags.c, src/sym.c,
to.do/unicode/flexdef.h, to.do/unicode/misc.c, to.do/unicode/scan.l:
Removed flex_alloc; cleaned up style. The function flex_alloc() was just a wrapper around malloc(). Since
this only added unclarity, and the flex_alloc() function is likely a
legacy of olden times, remove it in favor of calls to malloc()
directly. Style elements cleaned up: * superfluous spacing around parentheses * non-constant initialization in variable declarations * needless casts * almost all uses of assignments as subexpressions
2015-12-07 Michael McConville <mmcconville@mykolab.com>
* src/dfa.c, src/flexdef.h, src/gen.c, src/main.c, src/scan.l,
src/scanopt.c, tests/bison_nr_scanner.l,
tests/bison_yylloc_scanner.l, tests/bison_yylval_scanner.l,
tests/mem_nr.l, tests/mem_r.l, tests/pthread.l, tests/string_nr.l,
tests/string_r.l, tests/yyextra.l, to.do/unicode/scan.l: Remove
allocation casts
2015-12-07 Will Estes <westes575@gmail.com>
* src/.gitignore, src/Makefile.am: Built flex with itself. Changes in scan.l need to be built into flex with the same version
of flex in some cases. Since this build requirement is minimal, we
simply bootstrap flex unconditionally. We intentionally exclude from version control the bootstrap
artifacts as the extra copy of the lexer, the intermediate scanner
and the bootstrap executable are not of interest.
2015-12-05 Michael McConville <mmcconville@mykolab.com>
* src/buf.c, src/dfa.c, src/main.c, src/nfa.c, src/parse.y,
src/scan.l, src/scanopt.c, src/sym.c, to.do/unicode/main.c,
to.do/unicode/scan.l: Use NULL rather than (type *) 0.
2015-12-05 Michael McConville <mmcconville@mykolab.com>
* src/flexdef.h, src/misc.c, src/sym.c, to.do/unicode/flexdef.h,
to.do/unicode/misc.c: Replace copy_unsigned_string() with xstrdup(). Like copy_string(), copy_unsigned_string() is just a clone of the
stlib's strdup(). We only use it twice. I'm pretty confident that
char signedness is irrelevant in this case.
2015-12-04 Akim Demaille <akim@lrde.epita.fr>
* src/FlexLexer.h: Cleaned up white space.
2015-12-04 Akim Demaille <akim@lrde.epita.fr>
* src/FlexLexer.h: Removed struct keyword before yy_buffer_state
2015-12-04 Akim Demaille <akim@lrde.epita.fr>
* examples/testxxLexer.l, src/FlexLexer.h, src/flex.skl, src/main.c:
Replaced FLEX_STD macro with std::. The std:: construct exists as of C++98, so we can simply assume it
is supported.
2015-12-04 Akim Demaille <akim@lrde.epita.fr>
* src/Makefile.am: Generated skel.c explicitly in srcdir. Rewrote the target for skel.c to explicitly mention the srcdir. This
should help when building flex from a directory outside the flex
tree. Spread the rule out over several lines to enhance readability.
2015-12-04 Akim Demaille <akim@lrde.epita.fr>
* autogen.sh: Called glibtoolize if libtoolize run fails. On Mac OS X, libtoolize is known as glibtoolize. In cases where
libtoolize is not present, then calling glibtoolize when
bootstrapping the build system gives more folks a shot at getting
flex built from the ground up.
2015-12-02 Mightyjo <mightyjo@gmail.com>
* configure.ac: Checked for (g)texi2dvi. Better bison, help2man
checks. Added test for presence of (g)texi2dvi program. Gave notice if
texi2dvi is unavailable and set TEXI2DVI=: to avoid giving users
headaches. Enhanced tests for bison and help2man with notices when the programs
aren't found. Set their program variables to use the missing script
in build-aux since it's compatible with them.
2015-12-02 Michael McConville <mmcconville@mykolab.com>
* src/flexdef.h, src/misc.c, src/parse.y, src/scan.l, src/sym.c,
src/tables.c, to.do/unicode/flexdef.h, to.do/unicode/misc.c,
to.do/unicode/scan.l: Made string copying more standard. copy_string() was a clone of the stdlib's strdup(). For safety,
simplicity, and speed, we should use that instead. We introduce
xstrdup() which wraps strdup() in a failure upon memory allocation
errors.
2015-11-30 Serguey Parkhomovsky <sergueyparkhomovsky@gmail.com>
* src/scan.l: Error on unbalanced parentheses in rules section.
2015-11-29 Will Estes <westes575@gmail.com>
* src/Makefile.am: Cleaned up BUILT_SOURCES list. Removed reference to skel.c as a built source since other make rules
cover this case.
2015-11-29 Will Estes <westes575@gmail.com>
* src/Makefile.am: Sorted file names in flex_SOURCES
2015-11-27 Mightyjo <mightyjo@gmail.com>
* src/ecs.c, src/flexdef.h, src/main.c, src/misc.c, src/scan.l,
src/scanopt.c, src/sym.c, src/tblcmp.c: Replaced CHAR macro with
unsigned char type. Thanks to Michael McConville for pointing out that the old Char
macro causes problems with static analysis. The macro has been
removed and replaced with 'unsigned char' throughout the flex
sources. The macro is not needed at best and was confusing at worst.
It was not used in any of the example files nor was it mentioned in
the manual at all.
2015-11-21 Will Estes <westes575@gmail.com>
* configure.ac: updated syntax of AC_INIT call as per autoupdate
2015-11-21 Will Estes <westes575@gmail.com>
* tests/Makefile.am: removed extra call to a _CPPFLAGS variable
2015-05-14 Stefan Reinauer <stefan.reinauer@coreboot.org>
* src/buf.c, src/ccl.c, src/dfa.c, src/ecs.c, src/gen.c,
src/main.c, src/misc.c, src/nfa.c, src/parse.y, src/scan.l,
src/scanopt.c, src/sym.c, src/tblcmp.c: Switch function definitions
from mixed K&R to consistent ANSI C. flex was using K&R function definitions for some functions and ANSI
C style in others, sometimes even in the same file. Change the code
to consistently use ANSI C. Signed-off-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2015-11-16 Mightyjo <mightyjo@gmail.com>
* tests/Makefile.am: Used in-tree header file for c++ tests.
Looks like the '%Q%' was supposed to be translated to an actual date through
some tool but neither the README nor the Makefile say anything. So just
setting the date manually.
ok wiz@
This change introduced ABI incompatible change with older versions shipped
on NetBSD. This back out code that is currently not working correctly due
to TLS-based std::call_once implementation in GNU libstdc++.
Error when starting gnuchash:
/usr/pkg/lib/libwebkitgtk-1.0.so.0: Undefined symbol "_ZSt15__once_callable" (symnum = 1705)
PR 51139
Reported by <wiz>
Currently std::call_once with libstdc++ works only with static linking.
Disable code path using __thread types and introduce FIXME_PR_51139.
Problem discussed in PR 51139
Functional std::call_once is required in LLVM and LLDB codebase.
Example code to test std::call_once:
#include <iostream>
#include <thread>
#include <mutex>
#include <cstdlib>
std::once_flag flag;
int main(int argc, char **argv)
{
std::call_once(flag, [](){ std::cout << "Simple example: called once\n"; });
return EXIT_SUCCESS;
}
Sponsored by <The NetBSD Foundation>
- sscr_insert(): sh_prompt should be compared with t, not p
- replace strnstr(3) with memcmp(3)
now, nvi is free from strnstr(3); it can be safely pulled-up into netbsd-7
if it is going to use it to set the time. This makes update work like checkout
with respect to time setting. The time of the updated file is set to the
repository modification time of the file as opposed to the the time that the
file was checked out.
backport fix from upstream:
URL: https://gcc.gnu.org/viewcvs?rev=238789&root=gcc&view=rev
Log:
rs6000: Fix logic for when to emit .machine (PR71216)
The current logic determining whether to use .machine in the generated
asm code puts it there if the compiler is not configured with a default
target cpu, _or_ no -mcpu= was given on the command line. It should
be "and" instead.
PR target/71216
* config/rs6000/rs6000.c (rs6000_file_start): Fix condition for
when to emit a ".machine" pseudo-op.
Modified:
branches/gcc-5-branch/gcc/ChangeLog
branches/gcc-5-branch/gcc/config/rs6000/rs6000.c
7265 works fine.
7260 and 3160 maybe work.
7265D, 3165 and 8260 not work yet.
if_iwm.c: r1.147.
if_iwmreg.h: r1.19 without r1.16.
if_iwmvar.h: r1.24.
firmware is taken from https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi
- iwlwifi-7260-ucode-16.242414.0.tgz
- iwlwifi-3160-ucode-16.242414.0.tgz
- iwlwifi-7265-ucode-16.242414.0.tgz
- iwlwifi-8000-ucode-16.242414.0.tgz
GNU ld for alpha is so clever that the redundant GP load in function
entrypoint is skipped. we must therefore skip initial GP loads; otherwise
breakpoints in function entrypoints can also be skipped.
Reported to upstream (Bug 20969):
https://sourceware.org/bugzilla/show_bug.cgi?id=20969
ok martin
