ICE while building wordsedit.cpp of agistudio 1.1 on alpha, happens with 2.95, 3.2.3, and 3.3. g++ -c -pipe -Wall -W -g -D_REENTRANT -DQT_THREAD_SUPPORT -I/usr/include/qt3 -I/usr/include/qt -o wordsedit.o wordsedit.cpp wordsedit.cpp: In member function `void WordsEdit::merge_file()': wordsedit.cpp:620: Internal compiler error in dwarf2out_frame_debug_expr, at dwarf2out.c:1328 Please submit a full bug report, 2.95 includes this as well, 3.2 and 3.3 don't: {standard input}: Assembler messages: {standard input}:9212: Warning: missing .end or .bend at end of file Release: 3.2.3, 3.3 20030410 (prerelease), 2.95.4 20011002 (Debian prerelease) Environment: Reading specs from /usr/lib/gcc-lib/alpha-linux/3.2.3/specs Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,objc,ada --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-gxx-include-dir=/usr/include/c++/3.2 --enable-shared --with-system-zlib --enable-nls --without-included-gettext --enable-__cxa_atexit --enable-clocale=gnu --enable-java-gc=boehm --enable-objc-gc alpha-linux Thread model: posix
From: Falk Hueffner <falk.hueffner@student.uni-tuebingen.de> To: rmurray@debian.org Cc: gcc-gnats@gcc.gnu.org, 192568@bugs.debian.org Subject: Re: c++/10695: 2.95/3.2/3.3: ICE on alpha while building agistudio Date: 09 May 2003 17:54:03 +0200 rmurray@debian.org writes: > g++ -c -pipe -Wall -W -g -D_REENTRANT -DQT_THREAD_SUPPORT -I/usr/include/qt3 -I/usr/include/qt -o wordsedit.o wordsedit.cpp > wordsedit.cpp: In member function `void WordsEdit::merge_file()': > wordsedit.cpp:620: Internal compiler error in dwarf2out_frame_debug_expr, at dwarf2out.c:1328 > Please submit a full bug report, This is acually a bug in the Debian package, because the file isn't compiled with -O2 as policy mandates, and the error only occurs at -O0 ;) Here's a test case: struct WordList { char c[320000]; }; void merge(struct WordList); void merge_file() { struct WordList w; merge(w); } The problem is still there in g++ 3.4 20030423. It also occurs with gcc -g. The author should probably also consider not passing 320k large structures by value... -- Falk
Don't know why bugzilla doesn't consider this confirmed... BTW, doesn't this bug more likely belong under "target"? Thanks, Dara
still happens with 3.3.1
Still present on mainline.
Still present on tree-ssa.
Subject: Bug 10695 CVSROOT: /cvs/gcc Module name: gcc Branch: gcc-3_4-branch Changes by: rth@gcc.gnu.org 2004-08-26 09:05:33 Modified files: gcc : ChangeLog gcc/config/alpha: alpha.c Log message: PR debug/10695 * config/alpha/alpha.c (emit_frame_store_1, emit_frame_store): New. (alpha_expand_prologue): Handle >32-bit frames. Generate proper unwind info for >16-bit frames. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=2.2326.2.597&r2=2.2326.2.598 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/alpha/alpha.c.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.342.4.10&r2=1.342.4.11
Subject: Bug 10695 CVSROOT: /cvs/gcc Module name: gcc Changes by: rth@gcc.gnu.org 2004-08-26 09:08:12 Modified files: gcc : ChangeLog gcc/config/alpha: alpha.c Log message: PR debug/10695 * config/alpha/alpha.c (emit_frame_store_1, emit_frame_store): New. (alpha_expand_prologue): Handle >32-bit frames. Generate proper unwind info for >16-bit frames. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.5101&r2=2.5102 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/alpha/alpha.c.diff?cvsroot=gcc&r1=1.391&r2=1.392
Fixed.
No, really. Fixed.