aboutsummaryrefslogtreecommitdiff
path: root/gdb
AgeCommit message (Collapse)AuthorFilesLines
1995-09-02 * gdb.base/setvar.exp: Add new testcases for truncation whenPeter Schauer1-0/+5
assigning invalid values to bitfields.
1995-09-02 * configure.in: Explicitly `exit 0' for broken shells.Peter Schauer4-0/+19
* configure: Rebuilt. * symtab.c (list_symbols): Add missing blank after `<function, no debug info>' output. * valops.c (value_assign): Handle truncation when assigning to bitfields. Use value_copy to construct the return value from toval. * values.c (value_copy): Copy VALUE_FRAME and VALUE_OPTIMIZED_OUT.
1995-09-02 * gdb.base/callfuncs.exp (do_function_calls): RemoveFred Fish2-3/+10
mips-sgi-irix* xfail for "call inferior func with struct - returns char *" and fix test so that an optional (unsigned char *) cast is accepted in the result.
1995-09-01 * gdb.base/funcargs.exp (float_and_integral_args): RemoveFred Fish3-4/+10
sparc-sun-solaris2* xfail for "print f1 after run to call2a". * gdb.c++/Makefile.in (SUFFIXES): Add .SUFFIXES and ".cc" suffix. * gdb.base/signals.exp: Remove duplicate "handle all print" test that accidentally got checked in.
1995-09-01Ideally the "remote-mips.c" prompt spotting code should be updated toJackie Smith Cashion9-3/+87
spot different prompts at run-time, so as to avoid having target specific versions of gdb. This is a pending TODO. Fri Sep 1 08:25:50 1995 James G. Smith <jsmith@beauty.cygnus.com> * configure (mips64*vr4300*-*-elf): Support added. * remote-mips.c (mips_readchar): Change to allow build-time prompt string. * config/mips/tm-mips.h: Added TARGET_MONITOR_PROMPT. * config/mips/{vr4300.mt, vr4300el.mt, tm-vr4300.h, tm-vr4300el.h}: Added.
1995-08-31 * config/sh/sh.mt (SIM): Add -lm.Jim Wilson2-1/+5
1995-08-31 * rmote-nindy.c (non_dle, nidy_resume, nindy_wait): Changes toKung Hsu2-48/+56
conform to GNU coding standards.
1995-08-29 * remote.c (remote_wait): Revert 19 July my change which should beKung Hsu1-0/+5
customer specific.
1995-08-28 * gdb.base/callfuncs.exp (do_function_calls): Add alpha-dec-osf2*Fred Fish8-3/+68
clear_xfail for "p t_float_values2(3.14159,float_val2)" for gcc compiled test. * gdb.base/opaque.exp (setup_xfail_on_opaque_pointer): Add mips-sgi-irix5* xfail for not gcc compiled. * gdb.base/Makefile.in (nodebug.o): Also create nodebug.ci. * gdb.base/nodebug.exp: Add mips-sgi-irix5 xfail when not gcc compiled for "p top", "whatis top", "p middle", and "whatis middle". * gdb.base/whatis.exp: Add mips-sgi-irix* xfail for "whatis signed char" for not gcc compiled. * gdb.base/setvar.exp: Add mips-sgi-irix4* xfail (works on irix5) for "set variable signed char=-1 (-1)" and "set variable signed char=0xFF (0xFF)" for not gcc compiled. * gdb.base/funcargs.exp (float_and_integral_args): Add mips-sgi-irix5* xfail for "run to call2a" for not gcc compiled. Add mips-sgi-irix* xfail when not gcc compiled for "continue to call2b". Add mips-sgi-irix4* xfail (works with irix5) when gcc compiled for "continue to call2g". (discard_and_shuffle): Add mips-sgi-irix5* xfail whn not gcc compiled for "backtrace from call6a" (shuffle_round_robin): Add mips-sgi-irix* xfail when not gcc compiled for "backtrace from call7k". Add mips-sgi-irix5* xfail when not gcc compiled for "backtrace from call7a". (localvars_after_alloca): Fix gdb_test cmds for "print * after runto ...". Remove rs6000-*-* xfails for "print i after runto localvars_after_alloca" and "print l after runto localvars_after_alloca" for all compilers. * gdb.base/exprs.exp: Add mips-sgi-irix4* xfails (works with irix5), when not compiled with gcc, for: "print signed char == (minus)", "print signed char != (minus)", "print signed char < (minus)", "print signed char > (minus)". * gdb.base/callfuncs.exp (do_function_calls): Add mips-sgi-irix* xfail, when compiled with native compiler, for "call inferior func with struct - returns char *". * gdb.base/return.exp (return_tests): Change xfail for "correct value returned double test" to include Solaris 2.4. * gdb.base/funcargs.exp (float_and_integral_args): Add sparc-sun-solaris2* xfail for "print f1 after run to call2a".
1995-08-26 * gdb.base/setvar.c, gdb.base/setvar.exp: Add new tests forPeter Schauer2-95/+62
enumeration bitfields if compiling with GNU C. * lib/gdb.exp: Consume `(y or n) ' in `Reinitialize source path to empty' prompt.
1995-08-26 * mdebugread.c (parse_symbol): Handle sh.value of zero for enums.Peter Schauer5-17/+142
Determine signedness of enum type from enumerators. (parse_type): Handle btIndirect types, handle fBitfield for some non-member types. (upgrade_type): Use TYPE_FLAG_TARGET_STUB for arrays with unknown length. (cross_ref): Handle stIndirect forward reference to btTypedef. * stabsread.c (read_enum_type): Determine signedness of enum type from enumerators. * top.c (execute_command): Remove trailing whitespace from command arguments, except for `set' and `complete' commands. (validate_comname): Allow underscores in user defined command names. * values.c (modify_field): Change `Value does not fit in %d bits' error to a warning. Exclude sign extension bits of negative field values from fit check.
1995-08-25Only build PowerPC simulator if target is powerpc*-*-eabisimMichael Meissner8-36/+102
1995-08-23Link in PowerPC simulator.Michael Meissner3-0/+14
1995-08-22 * gdb.base/scope.exp: xfail 'scope0.c'::filelocal_bss before runJeff Law2-0/+9
test for PRO targets.
1995-08-22 * tm-hppa.h (EXTRACT_RETURN_VALUE): Fix for FP values.Jeff Law1-0/+2
1995-08-22 * tm-hppa.h (STORE_RETURN_VALUE): Fix to work with -msoft-floatJeff Law2-3/+16
calling conventions too. Use the TYPE of the return value, not its length to determine if it should also be copied into the floating point registers.
1995-08-22 * tm-hppa.h (PROLOGUE_FIRSTLINE_OVERLAP): Delete. Causes moreJeff Law3-8/+32
problems than it fixes. * hppa-tdep.c (skip_prologue): If we exit the main loop without finding all the register saves, retry again without looking for the registers we could not find the first time. Better fix for 7768.
1995-08-22Opps. Fix thinko in last change.Jeff Law1-7/+7
1995-08-22 * gdb.base/funcargs.exp: Avoid ever setting more than 8Jeff Law2-7/+26
breakpoints in the inferior at any given time by making two groups of breakpoints for call2*, call6* and call7* tests.
1995-08-22 * hppa-tdep.c (frame_chain_valid): Handle systems where "$START$"Jeff Law2-1/+17
calls "main" directly. (skip_prologue): Always assume arguments were saved into the stack since GCC will do so without setting the magic Args_Saved bit in the unwind descriptor.
1995-08-21 * remote-udi.c (udi_wait): Mask off high bits of stop reason.Kung Hsu2-2/+9
* remote-udi.c (fetch_register): For unfetchable regs, pretend it's done. Fix a bug.
1995-08-21 * Makefile.in (install): Remove "brokensed" hack, unnecessary nowJeff Law1-0/+6
that we're using autoconf. (uninstall): Likewise.
1995-08-20 * gdb.base/funcargs.exp: Fix typos introduced by Aug 15 change.Peter Schauer3-17/+16
* gdb.base/callfuncs.c (main): Use struct_val1 to inhibit xlc from optimizing it away. * gdb.base/callfuncs.exp: Remove rs6000-*-* xfails for "call inferior func with struct".
1995-08-20Wire up the breakpoint window so that it works. Also, fix a typo thatMichael Tiemann2-7/+33
kept the breakpoints in the source window from updating when enabled or disabled.
1995-08-19 * gdbtypes.c (recursive_dump_type): Add dont_print_type_obstackPeter Schauer2-1/+36
to inhibit infinite recursion when printing aggregate types.
1995-08-19 * dcache.c (dcache_write_line): Write dirty lines right.Steve Chamberlain2-16/+20
1995-08-18 * sparc-tdep.c (isbranch): Fix typo which caused wrongPeter Schauer2-2/+7
target addresses for annulled branches.
1995-08-17 * config/pa/tm-hppa.h (PROLOGUE_FIRSTLINE_OVERLAP): Define.Jeff Law2-6/+44
Yes, this dusty old macro is really the right thing to do when you've got a port which performs scheduling on prologues... Fixes latest mentor gdb PR.
1995-08-16 * configure.in: Only configure gdb.chill for particular targets.Stan Shebs2-4/+8
1995-08-16 * gdb.base/Makefile.in: Add action to .c.o transformationFred Fish12-313/+407
rule that generates a .ci file for each .o file and remove explicit .c.o rules except for callfuncs.o. * gdb.c++/Makefile.in: Add .cc.o transformation rule that generates a .ci file for each .o file and remove explicit .cc.o rules. (clean): Remove *.tmp *.ci files. (EXECUTABLES): Remove templ-info.exp. * gdb.base/compiler.c: New file. * gdb.c++/compiler.cc: New file * gdb.base/scope.exp: Source gdb.base/scope.ci. Change rs6000-*-* xfail for "print 'scope0.c'::filelocal_ro" and "print 'scope1.c'::filelocal" and "print 'scope1.c'::filelocal_bss" and "print 'scope1.c'::filelocal_ro" and "print 'scope1.c'::foo::funclocal" and "print 'scope1.c'::foo::funclocal_ro" and "print 'scope1.c'::bar::funclocal" and "print 'scope0.c'::filelocal_ro" and "print 'scope1.c'::filelocal at foo" and "print 'scope1.c'::filelocal_bss at foo" and "print 'scope1.c'::filelocal_ro at foo" and "print 'scope1.c'::foo::funclocal at foo" and "print 'scope1.c'::foo::funclocal_bss at foo" and "print 'scope1.c'::foo::funclocal_ro at foo" and "print 'scope1.c'::bar::funclocal at foo" and "print 'scope0.c'::filelocal_ro" and "print 'scope1.c'::filelocal" and "print 'scope1.c'::filelocal_bss" and "print 'scope1.c'::filelocal_ro" and "print 'scope1.c'::foo::funclocal" and "print 'scope1.c'::foo::funclocal_bss" and "print 'scope1.c'::foo::funclocal_ro" and "print 'scope1.c'::bar::funclocal" and "print 'scope1.c'::bar::funclocal_bss" to only be xfail'd when not compiled with gcc. Add rs6000-*-* xfail for "print 'scope0.c'::filelocal_bss before run" when compiled with gcc. (test_at_main): Add rs6000-*-* xfail for "print filelocal_ro in test_at_main" when compiled with gcc. * gdb.base/ptype.exp: Source gdb.base/ptype.ci. Add rs6000-*-aix* xfail for "whatis unnamed typedef'd enum (compiler bug in IBM's xlc)" and "ptype t_char_array", not compiled with gcc. * gdb.base/list.exp (test_list_function): Add rs6000-*-* xfail for "list foo (in include file)" when gcc compiled. * gdb.base/funcargs.exp: Source gdb.base/funcargs.ci (integral_args): Add rs6000-*-* xfail for "run to call0a" if not compiled with gcc. (unsigned_integral_args): Add rs6000-*-* xfail for "run to call1a" if not compiled with gcc. (float_and_integral_args): Add rs6000-*-* xfail for "run to call2a" if not compiled with gcc and for "continue to call2b" for any compiler. Add rs6000-*-* xfail to "print f1 after run to call2a" for gcc compiled. (discard_and_shuffle): Add rs6000-*-* xfail for "run to call6a". (shuffle_round_robin): Add rs6000-*-* xfail for "backtrace from call7a" if not compiled with gcc. Add rs6000-*-* xfail for "backtrace from call7b" if compiled with gcc. (call_after_alloca): Add rs6000-*-* xfail for "print c in call_after_alloca" and "print s in call_after_alloca" and "backtrace from call_after_alloca_subr" if not compiled with gcc. (localvars_in_indirect_call): Add rs6000-*-* xfail for "print c in localvars_in_indirect_call" and "print c in localvars_in_indirect_call" and "backtrace in indirectly called function" and "stepping into indirectly called function" if not compiled with gcc. (localvars_after_alloca): Add rs6000-*-* xfail for "print c after runto localvars_after_alloca" and "print s after runto localvars_after_alloca" and "print i after runto localvars_after_alloca" and "print l after runto localvars_after_alloca" for all compilers. * gdb.base/whatis.exp: Source gdb.base/whatis.ci rather than whatis-info.exp. * gdb.base/opaque.exp: Source gdb.base/opaque0.ci rather than opaque-info.exp. Setup rs6000-*-* xfail for "ptype on opaque struct pointer (statically)" and "ptype on opaque struct pointer (dynamically)" when not compiled with gcc. * gdb.base/callfuncs.exp: Source gdb.base/callfuncs.ci rather than callf-info.exp. (do_function_calls): Add clear_xfail for rs6000-*-* for "p t_float_values2(3.14159,float_val2)". Seems to work fine there, both with xlc and gcc. Need to find out what it is doing right and fix other platforms. * gdb.base/callfuncs.exp: Add rs6000-*-* xfail for "call inferior func with struct - returns int", "call inferior func with struct - returns long", "call inferior func with struct - returns float", "call inferior func with struct - returns double", "call inferior func with struct - returns char *", but only if not gcc compiled (presumes xlc compiled). Change rs6000-*-* xfails for "call inferior func with struct - returns char" and " call inferior func with struct - returns short" to only xfail if not using gcc. (clean mostlyclean): Remove *.ci and *.tmp files. * gdb.c++/templates.exp: Source gdb.c++/templates.ci rather than templ-info.exp. * gdb.base/langs.exp: Source gdb.base/langs.ci. Add rs6000-*-* xfail for "up to foo in langs.exp" "up to cppsub_ in langs.exp" and "up to fsub in langs.exp" when not gcc compiled.
1995-08-15 * remote.c (remote_write_bytes): Chop up large transfers.Steve Chamberlain2-25/+39
1995-08-15Departure of gcc.patchStan Shebs1-1/+0
1995-08-15 * gcc.patch: Remove, relevant only to long-ago versions of GCC.Stan Shebs2-80/+8
1995-08-14 * blokcframe.c (inside_main_func): Add comments.Kung Hsu1-0/+2
1995-08-14 * blockframe.c (inside_main_func): Fix a typo in previous change.Kung Hsu2-1/+3
1995-08-14 * config/sparc/tm-sparclite.h: Define FRAME_CHAIN_VALID_ALTERNATE.Kung Hsu1-0/+2
1995-08-14 * config/sparc/tm-sparclite.h: Define FRAME_CHAIN_VALID_ALTERNATE.Kung Hsu2-104/+97
* blockframe.c (inside_main_func): If main func addr range not set, try to set it now.
1995-08-13 * config/powerpc/xm-aix.h (FIVE_ARG_PTRACE): Define.Jeff Law3-0/+7
* config/rs6000/xm-rs6000.h (FIVE_ARG_PTRACE): Likewise. So we can build under aix4 with an ansi compiler (should not effect aix3 since aix3 doesn't have a ptrace prototype).
1995-08-12 * configure.in: Recognize aix4 specially as some aspectsJeff Law15-5/+219
of aix4 need different handling than aix3. * configure: Updated. * config/powerpc/{aix4.mh,aix4.mt,tm-ppc-aix4.h}: New files specific to aix4 support on the power pc. * config/powerpc/tm-ppc-aix.h (DONT_RELOCATE_SYMFILE_OBJFILE): Do not defined. The aix4 specific target files will do that. * config/rs6000/{aix4,mh,aix4,mt,tm-rs6000-aix4.h}: New files specific to aix4 support on the rs6000. * config/rs6000/tm-rs6000.h (CONVERT_FROM_FUNC_PTR_ADDR): Don't do the conversion if the pointer is not a magic aix function pointer. * rs6000-tdep.c: Include objfiles.h and symtab.h. (is_magic_function_pointer): New function. * rs6000-tdep.c (skip_prologue): Refine check for frameless functions. Handle b .+4 emitted by aix4 compilers. Only allow one load of a minimal toc pointer. Handle aix4 compiler's code for alloca. * rs6000-tdep.c (find_toc_address): Report an error if no toc was found rather than possibly core dumping. * partial-stab.h: Handle extra field generated by the aix4 compiler for enumerations. * stabsread.c (read_enum_type): Likewise.
1995-08-12 * gdb.base/callfuncs.exp: Add xfails for the powerpc.Jeff Law8-290/+162
* gdb.base/corefile.exp: Likewise. * gdb.base/list.exp: Likewise. * gdb.base/scope.exp: Likewise. * gdb.base/siganls.exp: Likewise. * gdb.base/nodebug.exp: Add xfails for the powerpc. Handle aix4 compiler output. * gdb.base/whatis.exp: Likewise.
1995-08-12 * rs6000-tdep.c (extract_return_value): Fix returning of valuesPeter Schauer2-2/+14
whose length is less than the register size for big endian targets.
1995-08-11 * symtab.c (list_symbols): if break command set breakpoint on matchedKung Hsu2-1/+13
symbol.
1995-08-11 * gdb.disasm/sh3.exp (all_fp_misc_tests): No longer expect aJeff Law2-4/+7
failure (opcode table has been fixed to match reality).
1995-08-11 * gdb.base/a1-selftest.exp: Change "i*86-*-sysv4" xfail forFred Fish8-65/+162
"backtrace through signal handler" to "i*86-*-sysv4*". * gdb.base/signals.exp: Add xfail for "'next' behaved as continue" case. Add "known SVR4 bug" to fail message. Add "i*86-*-bsdi2.0" xfail for "handle all print". Add "i*86-*-bsdi2.0" xfail for "backtrace in signals_tests_1". * gdb.base/ptype.exp: Add "i*86-*-sysv4*" xfail for "whatis unnamed typedef'd enum..." and "ptype t_char_array". * gdb.base/langs.exp: Add "i*86-*-sysv4*" xfail for "up to foo in langs.exp", "show language at foo in langs.exp", "up to cppsub_ in langs.exp", "show language at cppsub_ in langs.exp", "up to fsub in langs.exp", and "show language at fsub in langs.exp". * gdb.base/corefile.exp: Add "i*86-*-sysv4*" to xfail for "bactrace in corefile.exp". * gdb.base/callfuncs.exp: Change xfail to "i*86-*-*" for "call inferior function with struct - returns float" and "call inferior function with struct - returns double".
1995-08-11 * defs.h (strchr, strrchr, strstr, strtok, strerror): Enclose inFred Fish1-0/+5
#ifndefs to protect against previous definitions as macros. * HOW_TO_RELEASE: Continue updating...
1995-08-09 * xcoffread.x (xcoff_symfile_offset): Revert an unwanted change getsKung Hsu2-1/+6
in accidentally from Aug 1 checkin.
1995-08-09 * remote-hms.c (hms_cmds): Get reg term right.Steve Chamberlain3-6/+25
* monitor.c (monitor_fetch_register): If we see a non-hex digit, just stop reading. * remote.c (remote_wait): Change way $O is handled.
1995-08-09Fix typo in changelog entry.Jeff Law1-1/+1
1995-08-09 * configure.in (powerpc-*-aix*): Recognize as a new gdb hostJeff Law8-13/+200
and target. (powerpc-*-eabi*): Don't set configdirs. (powerpcle-*-eabi*0: Likewise. * configure: Updated. * rs6000-nat.c (vmap_ldinfo): Don't relocate data addresses for the main objfile if DONT_RELOCATE_SYMFILE_OBJFILE is defined. * config/powerpc/{aix.mh,aix.mh}: Host and target makefile fragments for powerpc running aix4. * config/powerpc/{nm-aix.h, tm-ppc-aix.h, xm-aix.h}: Native, target and host include files for powerpc running aix4. More of Mike's powerpc work + minor changes.
1995-08-09 * top.c (target_output_hook): Really make it match defs.h (char *Stan Shebs2-1/+6
is not the same as unsigned char *).
generated by cgit v1.1 at 2025-03-10 11:01:03 +0000