🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

1 total messages Started by Jeff A. Earickson Tue, 29 Dec 1992 03:24
endless loop in ./genflags
#1494
Author: Jeff A. Earickson
Date: Tue, 29 Dec 1992 03:24
88 lines
4661 bytes
Dear GCC Developers,


    GCC Version: 2.3.3 (Dec 26 copy from prep.ai.mit.edu)
    Platform:  Sun 3/60, running SunOS 4.1.1
    Problem:   ./genflags goes into endless loop and creates
               a huge tmp-flags.h file (mostly blanks)


I did a "configure --target=sun3-sunos4.1" and then a 
"make LANGUAGES=c".  For this example, the makefile is using
Sun's /usr/bin/cc C compiler.  Here is the output from make:


cc  -DIN_GCC   -g     -I. -I. -I./config -DGCC_INCLUDE_DIR=\"/usr/local/lib/gcc-lib/sun3-sunos4.1/`sed -e 's/.*\"\([^ \"]*\)[ \"].*/\1/' < ./version.c`/include\" -DGPLUSPLUS_INCLUDE_DIR=\"/usr/local/lib/g++-include\" -DLOCAL_INCLUDE_DIR=\"/usr/local/inclu

de\" -DCROSS_INCLUDE_DIR=\"/usr/local/lib/gcc-lib/sun3-sunos4.1/`sed -e 's/.*\"\([^ \"]*\)[ \"].*/\1/' < ./version.c`/sys-include\" -DTOOLDIR=\"/usr/local/sun3-sunos4.1/\" -c `echo ./cccp.c | sed 's,^\./,,'`
cc  -DIN_GCC   -g     -I. -I. -I./config -c ./cexp.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config version.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config obstack.c
cc  -DIN_GCC   -g   -o cccp cccp.o cexp.o version.o obstack.o ` case "cc" in "cc") echo  ;; esac `  
rm -f cpp
ln cccp cpp
cc  -DIN_GCC   -g     -I. -I. -I./config -c ./c-parse.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-lang.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-lex.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-decl.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-typeck.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-convert.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-aux-info.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-common.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config c-pragma.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config ./genattr.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config rtl.c
cc  -DIN_GCC   -g   -o genattr genattr.o rtl.o ` case "obstack.o" in ?*) echo obstack.o ;; esac ` ` case "cc"@"" in "cc"@?*) echo  ;; esac ` ` case "" in ?*) echo  ;; esac ` 
./genattr md > tmp-attr.h
./move-if-change tmp-attr.h insn-attr.h
insn-attr.h is unchanged
touch stamp-attr
cc  -DIN_GCC   -g     -I. -I. -I./config  -DUSE_COLLECT2 -c `echo ./toplev.c | sed 's,^\./,,'`
cc -c  -DIN_GCC   -g     -I. -I. -I./config tree.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config print-tree.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config stor-layout.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config fold-const.c
cc -c  -DIN_GCC   -g     -I. -I. -I./config ./genflags.c
cc  -DIN_GCC   -g   -o genflags genflags.o rtl.o ` case "obstack.o" in ?*) echo obstack.o ;; esac ` ` case "cc"@"" in "cc"@?*) echo  ;; esac ` ` case "" in ?*) echo  ;; esac ` 
./genflags md > tmp-flags.h
/bin/sh: 5392 Killed


After genflags had churned along for a few minutes, I did an "ls -l"
on tmp-flags.h, and here is what I found:

(44)> ls -l tmp-flags.h
-rw-------  1 jeff     staff     1515520 Dec 29 09:14 tmp-flags.h


Doing a "more" on the file shows that it is ok down to:

(more stuff above)
#define HAVE_jump 1
#define HAVE_tablejump 1
#define HAVE_decrement_and_branch_until_zero (find_reg_note (insn, REG_NONNEG,0)) 
#define HAVE_call 1
#define HAVE_call_value 1
#define HAVE_nop 1
#define HAVE_probe (NEED_PROBE)
#define HAVE_return (USE_RETURN_INSN)
#define HAVE_indirect_jump 1

#ifndef NO_MD_PROTOTYPES
extern rtx gen_tstsi                                                            
(mucho blanks beyond here)


Doing an "od -bc" on the file revealed that everything else was blanks.
I did the same thing last night, then went to supper.  When I came back
two hours later, tmp-flags.h was a 179Mb file!


/*------------------------------------------------------------------------*/
/* Jeff A. Earickson                         e-mail: jeff@uaboeing.ua.edu */
/* Cray/nCUBE/Sun Consultant,                  Work Phone: (205) 348-3969 */
/* Alabama Supercomputer Network                 Work Fax: (205) 348-3993 */
/* Boeing Computer Services                                               */
/* Address: PO Box 870346, University of Alabama, Tuscaloosa, AL 35487    */
/* Alabama 12-0-0, AP #2, Sugar Bowl bound, soon to be #1.  Roll Tide!!!  */
/*------------------------------------------------------------------------*/
Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads