By inspection, this seems to be a bug in gdb-5.2/sim/ppc/hw_com.c. I have not tested it. The section of code is setting up output.file but the test for bad open is for input.file.
Matt diff -c hw_com.c-orig hw_com.c *** hw_com.c-orig Sun Oct 6 09:26:22 2002 --- hw_com.c Sun Oct 6 09:26:31 2002 *************** *** 249,255 **** if (device_find_property(me, "output-file") != NULL) { const char *output_file = device_find_string_property(me, "output-file"); com->output.file = fopen(output_file, "w"); ! if (com->input.file == NULL) device_error(me, "Problem opening output file %s\n", output_file); if (device_find_property(me, "output-buffering") != NULL) { const char *buffering = device_find_string_property(me, "output-buffering"); --- 249,255 ---- if (device_find_property(me, "output-file") != NULL) { const char *output_file = device_find_string_property(me, "output-file"); com->output.file = fopen(output_file, "w"); ! if (com->output.file == NULL) device_error(me, "Problem opening output file %s\n", output_file); if (device_find_property(me, "output-buffering") != NULL) { const char *buffering = device_find_string_property(me, "output-buffering"); _______________________________________________ Bug-gdb mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-gdb