Skip to content
This repository
Browse code

[GH #867] icu 4.4 uses a wrong uset_openEmpty(); function declaration

Since we cannot check icu in auto::warnings (it comes later), remove
-Werror=strict-prototypes from src/string/encoding/shared.c.

Note that I noticed an apprently wrong -I$(@D) macro in the suffix override rules, which looks wrong, but does no harm.
  • Loading branch information...
commit 1cc9caa17a19545df0cecf3800edf01f8e351fe5 1 parent 80b4ecd
Reini Urban rurban authored

Showing 2 changed files with 11 additions and 4 deletions. Show diff stats Hide diff stats

  1. +5 0 config/auto/warnings.pm
  2. +6 4 config/gen/makefiles/root.in
5 config/auto/warnings.pm
@@ -327,6 +327,11 @@ sub runstep {
327 327 push @{$self->{'warnings'}{$compiler}{'basic'}},
328 328 '-fvisibility=hidden';
329 329 };
  330 +
  331 + # icu4.4 has a uset_openEmpty() declaration without (void)
  332 + $self->{'warnings'}{$compiler}{'never'}{'-Werror=strict-prototypes'} =
  333 + [ 'src/string/encoding/shared.c' ];
  334 +
330 335 if ($conf->data->get('clang') and $compiler eq 'g++') { # clang++
331 336 $self->{'warnings'}{'g++'}{'override'} = {
332 337 '-Wno-parentheses-equality' => [ qw(
10 config/gen/makefiles/root.in
@@ -1959,11 +1959,13 @@ src/string/encoding/tables$(O) : \
1959 1959 src/string/encoding/tables.h \
1960 1960 src/string/encoding/tables.c
1961 1961
  1962 +## SUFFIX OVERRIDE
1962 1963 src/string/encoding/shared$(O) : \
1963   - $(PARROT_H_HEADERS) \
1964   - src/string/encoding/shared.h \
1965   - src/string/encoding/shared.c \
1966   - src/string/encoding/tables.h
  1964 + $(PARROT_H_HEADERS) \
  1965 + src/string/encoding/shared.h \
  1966 + src/string/encoding/shared.c \
  1967 + src/string/encoding/tables.h
  1968 + $(CC) $(CFLAGS) @optimize::src/string/encoding/shared.c@ @ccwarn::src/string/encoding/shared.c@ @cc_shared@ -I$(@D) @cc_o_out@$@ -c src/string/encoding/shared.c
1967 1969
1968 1970 src/string/encoding/null$(O) : \
1969 1971 $(PARROT_H_HEADERS) \

1 comment on commit 1cc9caa

Gerhard R.

see also #747 - imo the proper fix is to use -isystem for 3rd party headers

Please sign in to comment.
Something went wrong with that request. Please try again.