GNOME Bugzilla – Bug 172815
GCC 4 issues with parser code
Last modified: 2005-07-19 10:51:36 UTC
Distribution/Version: Fedora Core 4 Test 1 Compile gstreamer BRANCH-GSTREAMER-0_8 with GCC 4 and you get the following: gcc -DHAVE_CONFIG_H -I. -I. -I../.. -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -pthread -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/libx ml2 -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -fno-common -g -Wall -Werror -DGS T_DISABLE_DEPRECATED -I../.. -O2 -g -pipe -Wp,-D_FORTIFY_SOURCE=2 -fexceptions - m32 -march=i686 -mtune=pentium4 -fasynchronous-unwind-tables -MT libgstparse_la- grammar.tab.lo -MD -MP -MF .deps/libgstparse_la-grammar.tab.Tpo -c grammar.tab.c -o libgstparse_la-grammar.tab.o >/dev/null 2>&1 cc1: warnings being treated as errors lex._gst_parse_yy.c: In function '_gst_parse_yylex': parse.l:144: warning: ignoring return value of 'fwrite', declared with attribute warn_unused_result Company suggested a patch that fixes this problem. (will attach)
Created attachment 39750 [details] [review] fixes fwrite warn_unused_result
This is really a bug in flex, so someone should - file a bug against flex. - explain in the patch why the "#define ECHO" is needed, so it doesn't get lost. Other than that I'd wait until the warnings gcc4 reports are finalized before committing this patch.
I'll add some comments about need in the patch after the meeting i have coming up in a few minutes.
*** Bug 301097 has been marked as a duplicate of this bug. ***
I mailed Flex maintainer Vern Paxson today pointing him to this bugzilla entry.
Jeffrey, Benjamin, I'm totally confused about the echo, so can either of you add a patch explaining why it's there so this can be applied?
Jeffrey, is this still a problem for you?
I haven't been building CVS lately, as all of the fixes I had been looking for in CVS have been rolled into releases.
*** Bug 310106 has been marked as a duplicate of this bug. ***
Reopening, we still get things like bug #310106
Applied patch similar to attachment #1 [details], but actually printing out something. Seems to fix Christian's problem from bug #310106.