binutils-gdb.git
3 years agogdbsupport: introduce struct observer
Simon Marchi [Sat, 24 Apr 2021 23:26:04 +0000 (19:26 -0400)]
gdbsupport: introduce struct observer

Instead of using a pair.  This allows keeping more data per observer in
a structured way, and using field names is clearer than first/second.

gdbsupport/ChangeLog:

* observable.h (class observable) <struct observer>: New.
<detach, notify>: Update.
<m_observers>: Change type to vector of observers.

Change-Id: Iadf7d1fa25049cfb089e6b1b429ddebc548825ab

3 years agosim: dv-cfi: fix printf format
Mike Frysinger [Sat, 24 Apr 2021 04:16:08 +0000 (00:16 -0400)]
sim: dv-cfi: fix printf format

Use the existing PRI constants to select the right format rather than
assume signed_cell is always %u.  Fixes building for riscv64.

3 years agosim: options: increase max option count
Mike Frysinger [Sat, 24 Apr 2021 04:10:44 +0000 (00:10 -0400)]
sim: options: increase max option count

As we turn on more modules by default for all ports, the number of
options has been increasing.  The sim-options module has a limit on
the number of options it can support, and if it's exceeded, it likes
to go into an infinite loop.  Increase the ceiling and add an assert
so we abort right away instead of hanging.

This will be needed to turn on hw support for v850 as it will then
exceed the current limit.

3 years agosim: simplify hardware m4 macro
Mike Frysinger [Sat, 24 Apr 2021 01:58:21 +0000 (21:58 -0400)]
sim: simplify hardware m4 macro

Every port using this sets the 1st arg to yes and the 2nd arg to "".
These are the defaults we probably want anyways in order to unify the
codebase, so move them to the macro and only allow ports to declare
extra hardware models.

3 years agoAutomatic date update in version.in
GDB Administrator [Sat, 24 Apr 2021 00:00:15 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agogdbsupport, gdb: change observer_debug to bool
Simon Marchi [Fri, 23 Apr 2021 20:28:26 +0000 (16:28 -0400)]
gdbsupport, gdb: change observer_debug to bool

gdb/ChangeLog:

* observable.c (observer_debug): Change to bool.

gdbsupport/ChangeLog:

* observable.h (observer_debug): Change to bool.

Change-Id: I58634235a20740a66eacb1c83bae3cf3304ae1fd

3 years agold: Properly create a symbolic link to tmpdir/ldscripts
H.J. Lu [Fri, 23 Apr 2021 19:36:27 +0000 (12:36 -0700)]
ld: Properly create a symbolic link to tmpdir/ldscripts

Don't create a symbolic link to tmpdir/ldscripts if it exists.

PR ld/27771
* testsuite/ld-bootstrap/bootstrap.exp: Create a symbolic link
to tmpdir/ldscripts only if it doesn't exist.

3 years agogdbsupport: include preprocessor.h in common-debug.h
Simon Marchi [Fri, 23 Apr 2021 18:03:30 +0000 (14:03 -0400)]
gdbsupport: include preprocessor.h in common-debug.h

While doing some changes, some code failed to compile because it used
the scoped_debug_start_end macro, but couldn't find the CONCAT macro.
Fix that by making common-debug.h include preprocessor.h, the header
file that provides CONCAT.

gdbsupport/ChangeLog:

* common-debug.h: Include preprocessor.h.

Change-Id: Ibf863a932a18cba9a57b4bd72df538ef52d39127

3 years agoFix type of .persistent.bss section
Eric Botcazou [Fri, 23 Apr 2021 17:07:59 +0000 (19:07 +0200)]
Fix type of .persistent.bss section

The change implementing the .persistent family of sections broke the
existing support for the .persistent.bss section in the compiler:

int a __attribute__ ((section (".persistent.bss")));

t.s: Assembler messages:
t.s:4: Warning: setting incorrect section type for .persistent.bss

The compiler encodes it as @nobits but the assembler expects @progbits.
The assembler is incorrect and should treat the section like the compiler.

bfd/
* elf.c (special_sections_p): Add .persistent.bss.
gas/
* testsuite/gas/elf/section25.d: Run it everywhere.
* testsuite/gas/elf/section26.d: Likewise.
* testsuite/gas/elf/section26.s: Add test for .persistent.bss.

3 years agoAllow .seh_pushframe to take an optional [code] parameter
J?r?me Gardou [Fri, 23 Apr 2021 15:33:47 +0000 (16:33 +0100)]
Allow .seh_pushframe to take an optional [code] parameter

* config/obj-coff-seh.c (obj_coff_seh_pushframe): Allow an
optional "code" argument.

3 years agoopcodes: xtensa: display loaded literal value
Max Filippov [Fri, 23 Apr 2021 09:03:36 +0000 (02:03 -0700)]
opcodes: xtensa: display loaded literal value

Display literal value loaded with l32r opcode as a part of disassembly.
This significantly simplifies reading of disassembly output.

2020-04-23  Max Filippov  <jcmvbkbc@gmail.com>
opcodes/
* xtensa-dis.c (print_xtensa_operand): For PC-relative operand
of l32r fetch and display referenced literal value.

3 years agoopcodes: xtensa: improve literal output
Max Filippov [Wed, 7 Apr 2021 01:44:50 +0000 (18:44 -0700)]
opcodes: xtensa: improve literal output

Output literals as 4-byte words, not as separate bytes.

2021-04-23  Max Filippov  <jcmvbkbc@gmail.com>
opcodes/
* xtensa-dis.c (print_insn_xtensa): Set info->bytes_per_chunk
to 4 for literal disassembly.

3 years agogdb: remove some caching from the dwarf reader
Andrew Burgess [Mon, 19 Apr 2021 12:14:41 +0000 (13:14 +0100)]
gdb: remove some caching from the dwarf reader

While working on some changes to 'info sources' I ran into a situation
where I was seeing the same source files reported twice in the output
of the 'info sources' command when using either .gdb_index or the
.debug_name index.

I traced the problem back to some caching in
dwarf2_base_index_functions::map_symbol_filenames; when called GDB
caches the set of filenames, but, filesnames are not removed as the
index entries are expanded into full symtabs.  As a result we can end
up seeing filenames reported both from a full symtab _and_ from
a (stale) previously cached index entry.

Now, obviously, when seeing a problem like this the "correct" fix is
to remove the stale entries from the cache, however, I ran a few
experiments to see why this wasn't really hitting us anywhere, and, as
far as I can tell, ::map_symbol_filenames is only called from three
places:

  1. The mi command -file-list-exec-source-files,
  2. The 'info sources' command, and
  3. Filename completion

However, the result of this "bug" is that we will see duplicate
filenames, and readline's completion mechanism already removes
duplicates, so for case #3 we will never see any problems.

Cases #1 and #2 are basically the same, and in each case, to see a
problem we need to ensure we craft the test in a particular way, start
up ensuring we have some unexpected symtabs, then run one of the
commands to populate the cache, then expand one of the symtabs, and
list the sources again.  At this point you'll see duplicate entries in
the results.  Hardly surprising we haven't randomly hit this situation
in testing.

So, considering that use cases #1 and #2 are certainly not "high
performance" code (i.e. I don't think these justify the need for
caching) this leaves use case #3.  Does this use justify the need for
caching?  Well the psymbol_functions::map_symbol_filenames function
doesn't seem to do any extra caching, and within
dwarf2_base_index_functions::map_symbol_filenames, the only expensive
bit appears to be the call to dw2_get_file_names, and this already
does its own caching via this_cu->v.quick->file_names.

The upshot of all this analysis was that I'm not convinced the need
for the additional caching is justified, and so, I propose that to fix
the bug in GDB, I just remove the extra caching (for now).

If we later find that the caching _was_ useful, then we can
reintroduce it, but add it back such that it doesn't reintroduce this
bug.

As I was changing dwarf2_base_index_functions::map_symbol_filenames I
replaced the use of htab_up with std::unordered_set.

Tested using target_boards cc-with-debug-names and dwarf4-gdb-index.

gdb/ChangeLog:

* dwarf2/read.c: Add 'unordered_set' include.
(dwarf2_base_index_functions::map_symbol_filenames): Replace
'visited' hash table with 'qfn_cache' unordered_set.  Remove use
of per_Bfd->filenames_cache cache, and use function local
filenames_cache instead.  Reindent.
* dwarf2/read.h (struct dwarf2_per_bfd) <filenames_cache>: Delete.

gdb/testsuite/ChangeLog:

* gdb.base/info_sources.exp: Add new tests.

3 years agox86: TLS desc call relocs target zero-size fields
Jan Beulich [Fri, 23 Apr 2021 07:18:48 +0000 (09:18 +0200)]
x86: TLS desc call relocs target zero-size fields

These are marker relocations, so together with their bit size being zero
their byte size should be zero as well. This is expressed by a value of
3, not 0.

3 years agox86-64: special case LEA when determining signedness of displacement
Jan Beulich [Fri, 23 Apr 2021 07:18:06 +0000 (09:18 +0200)]
x86-64: special case LEA when determining signedness of displacement

LEA behavior without a 64-bit destination is independent of address size
- in particular LEA with 32-bit addressing and 64-bit destination is the
same as LEA with 64-bit addressing and 32-bit destination. IOW checking
merely i.prefix[ADDR_PREFIX] is insufficient. This also means wrong
relocation types (R_X86_64_32S when R_X86_64_32 is needed) were used so
far in such cases.

Note that in one case in build_modrm_byte() the 64-bit check came too
early altogether, and hence gets dropped in favor of the one included in
the new helper. This is benign to non-64-bit code from all I can tell,
but the failure to clear disp16 could have been a latent problem.

3 years agox86-64: defer 32-bit signed displacement check
Jan Beulich [Fri, 23 Apr 2021 07:17:33 +0000 (09:17 +0200)]
x86-64: defer 32-bit signed displacement check

In preparation for extending the conditions here defer this check until
operands have been parsed, as certain further attributes will need to
be known for determinig applicability of this check to be correct to
LEA.

3 years agox86: re-order optimize_disp()
Jan Beulich [Fri, 23 Apr 2021 07:16:57 +0000 (09:16 +0200)]
x86: re-order optimize_disp()

While I can't point out any specific case where things break, it looks
wrong to have the consumer of a flag before its producer. Set .disp32
first, then do the possible conversion to signed 32-bit, and finally
check whether the value fits in a signed long.

3 years agox86: don't truncate values in diagnostics and alike
Jan Beulich [Fri, 23 Apr 2021 07:15:45 +0000 (09:15 +0200)]
x86: don't truncate values in diagnostics and alike

Truncating an expression's X_add_number to just "long" can result in
confusing output (e.g. an apparently in-range number claimed to be out
of range). Use the abstraction that bfd provides for this.

Take the opportunity and also insert a missing "of".

3 years agogdb: use function_view for iterate_over_bp_locations' callback
Simon Marchi [Fri, 23 Apr 2021 02:00:39 +0000 (22:00 -0400)]
gdb: use function_view for iterate_over_bp_locations' callback

Use a function_view instead of function pointer + data.  Actually,
nothing uses the data anyway, but that makes iterate_over_bp_locations
more like iterate_over_breakpoints.

gdb/ChangeLog:

* breakpoint.c (iterate_over_bp_locations): Change callback to
function view, remove data parameter.
* breakpoint.h (iterate_over_bp_locations): Likewise.
* record-full.c (record_full_sync_record_breakpoints): Remove
data parameter.

Change-Id: I66cdc94a505f67bc640bcc66865fb535ee939a57

3 years agoDo not check for sys/time.h or sys/times.h
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Do not check for sys/time.h or sys/times.h

This updates the sim so that it unconditionally uses sys/time.h.  This
is in agreement with existing code, and a recent change to BFD.

I also think that sys/times.h is never needed by the sim, so this
patch removes the check and the one spot that was conditionally
including it.

sim/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* m4/sim_ac_common.m4 (SIM_AC_COMMON): Don't check for sys/time.h
or sys/times.h.

sim/aarch64/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/arm/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/avr/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/bfin/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/bpf/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/common/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* sim-utils.c: Update includes.

sim/cr16/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* simops.c: Update includes.
* configure, config.in: Rebuild.

sim/cris/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* rvdummy.c: Update includes.
* dv-rv.c: Update includes.
* configure, config.in: Rebuild.

sim/d10v/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/erc32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/example-synacor/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/frv/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/ft32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/h8300/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/iq2000/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/lm32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/m32c/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/m32r/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/m68hc11/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/mcore/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/microblaze/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/mips/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/mn10300/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/moxie/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/msp430/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/or1k/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/ppc/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* mon.c: Update includes.
* emul_unix.c: Update includes.
(do_unix_gettimeofday): Update condition.

sim/pru/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/riscv/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/rl78/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/rx/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

sim/sh/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* interp.c: Update includes.
* configure, config.in: Rebuild.

sim/v850/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure, config.in: Rebuild.

3 years agoRemove LIBS from two sim Makefiles
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Remove LIBS from two sim Makefiles

A couple of sim Makefiles define LIBS, but don't use it.  This removes
these.

sim/m32c/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (LIBS): Remove

sim/rx/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (LIBS): Remove.

3 years agoRemove INCLUDE variable from some sim Makefiles
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Remove INCLUDE variable from some sim Makefiles

Some Makefiles in sim define INCLUDE but don't use it.  This removes
these instances.

sim/bfin/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (INCLUDE): Remove.

sim/m68hc11/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (INCLUDE): Remove.

sim/mn10300/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (INCLUDE): Remove.

sim/v850/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (INCLUDE): Remove.

3 years agoRemove and modernize dependencies in sim
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Remove and modernize dependencies in sim

Some spots in the sim build used manual dependencies, and some spots
did a compilation by hand but did not use the automatic dependency
tracking code.  This patch fixes these spots.

I didn't touch ppc, because it doesn't use the common Makefile code.
I also didn't touch objects that are for the build machine, because
automatic dependencies don't work for those.

sim/arm/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (armemu26.o, armemu32.o): Use COMPILE and
POSTCOMPILE.

sim/bpf/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (arch.o, cpu.o, sim-if.o, traps.o): Remove.
(mloop-le.o, mloop-be.o, decode-le.o, decode-be.o, sim-le.o)
(sim-be.o): Use COMPILE and POSTCOMPILE.
(SIM_EXTRA_DEPS): Add eng-le.h, eng-be.h.

sim/cr16/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (SIM_EXTRA_DEPS): New variable.
(simops.o): Remove.

sim/cris/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (sim-if.o, dv-cris.o, dv-rv.o, arch.o, traps.o)
(devices.o, crisv10f.o, mloopv10f.o, cpuv10.o, decodev10.o)
(modelv10.o, crisv32f.o, mloopv32f.o, cpuv32.o, decodev32.o)
(modelv32.o): Remove.
(SIM_EXTRA_DEPS): Add engv10.h.

sim/d10v/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (SIM_EXTRA_DEPS): New variable.
(simops.o): Remove.

sim/frv/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (arch.o, devices.o, frv.o, traps.o, pipeline.o)
(interrupts.o, memory.o, cache.o, options.o, reset.o)
(registers.o, profile.o, profile-fr400.o, profile-fr450.o)
(profile-fr500.o, profile-fr550.o, sim-if.o, mloop.o, cpu.o)
(decode.o, sem.o, model.o): Remove.
(SIM_EXTRA_DEPS): Add eng.h.

sim/iq2000/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (sim-if.o): Remove.
(arch.o): Use COMPILE and POSTCOMPILE.
(devices.o, iq2000.o, mloop.o, cpu.o, decode.o, sem.o, model.o):
Remove.
(SIM_EXTRA_DEPS): Add eng.h.

sim/lm32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (arch.o, traps.o, sim-if.o, lm32.o, mloop.o)
(cpu.o, decode.o, sem.o, model.o): Remove.
(SIM_EXTRA_DEPS): Add eng.h.

sim/m32r/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (sim-if.o, arch.o, traps.o, traps-linux.o)
(devices.o, m32r.o, mloop.o, cpu.o, decode.o, sem.o, model.o)
(m32rx.o, mloopx.o, cpux.o, decodex.o, semx.o, modelx.o)
(m32r2.o, mloop2.o, cpu2.o, decode2.o, sem2.o, model2.o): Remove.
(SIM_EXTRA_DEPS): Add eng.h, engx.h, eng2.h.

sim/m68hc11/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (interp.o): Remove.

sim/mips/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (interp.o, m16run.o, micromipsrun.o, multi-run.o):
Remove.
(SIM_EXTRA_DEPS): New variable.

sim/mn10300/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (interp.o): Remove.
(idecode.o op_utils.o semantics.o): Remove.

sim/or1k/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (mloop.o, arch.o, cpu.o, decode.o, sem.o)
(sem-switch.o, model.o): Remove.

sim/rl78/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (err.o, fpu.o, gdb-if.o, load.o, main.o, mem.o)
(reg.o, rl78.o): Remove.

sim/rx/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (err.o, fpu.o, gdb-if.o, load.o, main.o, mem.o)
(misc.o, reg.o, rx.o, syscalls.o, trace.o): Remove.

sim/sh/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (SIM_EXTRA_DEPS): New variable.
(interp.o): Remove.

sim/v850/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (interp.o, simops.o, semantics.o): Remove.

3 years agoRequire GNU make
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Require GNU make

GDB has required GNU make for quite some time, and this patch applies
this approach to the sim as well.  Requiring GNU make means that
automatic dependency tracking can be simple and reliable, and it also
makes other refactorings simpler.

sim/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* m4/sim_ac_common.m4 (SIM_AC_COMMON): Don't check for make.

sim/aarch64/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/arm/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/avr/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/bfin/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/common/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Make-common.in: Require GNU make.

sim/cr16/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/cris/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/d10v/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/erc32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/example-synacor/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/frv/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/ft32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/h8300/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/iq2000/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/lm32/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/m32c/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/m32r/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/m68hc11/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/mcore/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/microblaze/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/mips/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/mn10300/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/moxie/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/msp430/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/or1k/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/pru/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/riscv/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/rl78/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/rx/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/sh/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

sim/v850/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* configure: Rebuild.

3 years agoAdd stamp files for generated files in sim/ppc
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Add stamp files for generated files in sim/ppc

This changes the sim/ppc Makefile to use the stamp file idiom for a
couple of generated files, avoiding extra rebuilds.

sim/ppc/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Makefile.in (stamp-vals, stamp-map): New targets.
(targ-vals.h, targ-map.c): Update.
(clean): Remove files.

3 years agoIntroduce stamp file for hw-config.h
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Introduce stamp file for hw-config.h

This adds a stamp file for hw-config.h, to avoid unnecessary rebuilds.
It also arranges to remove hw-config.h in "mostlyclean", because the
file is created by "make".

sim/common/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Make-common.in (stamp-hw): New target.
(hw-config.h): Depend on stamp-hw.
(mostlyclean): Remove stamp-hw and hw-config.h.

3 years agoMake mostlyclean an alias for clean
Tom Tromey [Fri, 23 Apr 2021 01:51:54 +0000 (19:51 -0600)]
Make mostlyclean an alias for clean

I found out by accident that "mostlyclean" in a sim subdir removes all
the configure artifacts.  The usual rule is:

* If the maintainer built it, maintainer-clean should remove it;
* If configure built it, distclean should remove it;
* If make built it, "clean" should remove it;
* If there is a handy subset of "clean" that is "easy" to rebuild,
  "mostlyclean" should remove it; otherwise mostlyclean should be an
  alias for clean

This patch makes mostlyclean an alias for clean.

sim/common/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* Make-common.in (mostlyclean): Now an alias for clean, not
distclean.

3 years agoFix sim build failure
Tom Tromey [Thu, 22 Apr 2021 22:50:32 +0000 (16:50 -0600)]
Fix sim build failure

On x86-64 Fedora 32, the sim was failing to build.
sim_events_schedule was passing a 'dummy' argument to
sim_events_schedule_vtracef, which caused an error because the format
parameter was NULL.  However, removing this dummy argument caused an
error because too few arguments were being passed -- catch 22.

This patch fixes the build problem by using sim_events_schedule_tracef
instead.

sim/common/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* sim-events.c (sim_events_schedule): Use
sim_events_schedule_tracef.

3 years agoFix ptype/o bug with "<no data fields>"
Tom Tromey [Fri, 23 Apr 2021 01:39:56 +0000 (19:39 -0600)]
Fix ptype/o bug with "<no data fields>"

I noticed that when using ptype/o, the "<no data fields>" text that
may be emitted is indented incorrectly.  This patch fixes the bug and
adds a new test case.

I also removed a stray backslash from ptype-offsets.exp that I noticed
while writing the test.  This seemed too trivial to warrant a separate
patch.

gdb/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* c-typeprint.c (c_type_print_base_struct_union): Use
print_spaces_filtered_with_print_options.

gdb/testsuite/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* gdb.base/ptype-offsets.cc (struct empty_member): New.
(main): Use empty_member.
* gdb.base/ptype-offsets.exp: Add new test.

3 years agoAutomatic date update in version.in
GDB Administrator [Fri, 23 Apr 2021 00:00:15 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agogdb/testsuite: add Python support check in gdb.python/flexible-array-member.exp
Simon Marchi [Thu, 22 Apr 2021 19:45:34 +0000 (15:45 -0400)]
gdb/testsuite: add Python support check in gdb.python/flexible-array-member.exp

We don't want to execute this test if Python support is not compiled in
GDB, add the necessary check.

gdb/testsuite/ChangeLog:

* gdb.python/flexible-array-member.exp: Add check for Python
support.

Change-Id: I853b937d2a193a0bb216566bef1a35354264b1c5

3 years agogdb: fix getting range of flexible array member in Python
Simon Marchi [Thu, 22 Apr 2021 19:01:28 +0000 (15:01 -0400)]
gdb: fix getting range of flexible array member in Python

As reported in bug 27757, we get an internal error when doing:

    $ cat test.c
    struct foo {
        int len;
        int items[];
    };

    struct foo *p;

    int main() {
        return 0;
    }
    $ gcc test.c -g -O0 -o test
    $ ./gdb -q -nx --data-directory=data-directory ./test -ex 'python gdb.parse_and_eval("p").type.target()["items"].type.range()'
    Reading symbols from ./test...
    /home/simark/src/binutils-gdb/gdb/gdbtypes.h:435: internal-error: LONGEST dynamic_prop::const_val() const: Assertion `m_kind == PROP_CONST' failed.
    A problem internal to GDB has been detected,
    further debugging may prove unreliable.
    Quit this debugging session? (y or n)

This is because the Python code (typy_range) blindly reads the high
bound of the type of `items` as a constant value.  Since it is a
flexible array member, it has no high bound, the property is undefined.
Since commit 8c2e4e0689 ("gdb: add accessors to struct dynamic_prop"),
the getters check that you are not getting a property value of the wrong
kind, so this causes a failed assertion.

Fix it by checking if the property is indeed a constant value before
accessing it as such.  Otherwise, use 0.  This restores the previous GDB
behavior: because the structure was zero-initialized, this is what was
returned before.  But now this behavior is explicit and not accidental.

Add a test, gdb.python/flexible-array-member.exp, that is derived from
gdb.base/flexible-array-member.exp.  It tests the same things, but
through the Python API.  It also specifically tests getting the range
from the various kinds of flexible array member types (AFAIK it wasn't
possible to do the equivalent through the CLI).

gdb/ChangeLog:

PR gdb/27757
* python/py-type.c (typy_range): Check that bounds are constant
before accessing them as such.
* guile/scm-type.c (gdbscm_type_range): Likewise.

gdb/testsuite/ChangeLog:

PR gdb/27757
* gdb.python/flexible-array-member.c: New test.
* gdb.python/flexible-array-member.exp: New test.
* gdb.guile/scm-type.exp (test_range): Add test for flexible
array member.
* gdb.guile/scm-type.c (struct flex_member): New.
(main): Use it.

Change-Id: Ibef92ee5fd871ecb7c791db2a788f203dff2b841

3 years agosim: sprinkle some ATTRIBUTE_PRINTF
Simon Marchi [Thu, 22 Apr 2021 16:07:14 +0000 (12:07 -0400)]
sim: sprinkle some ATTRIBUTE_PRINTF

Add some ATTRIBUTE_PRINTF attributes to functions that take a format
string, to fix a few -Wformat-nonliteral warnings.  Use the
ATTRIBUTE_PRINTF macro like we use in GDB, instead of spelling out
__attribute__((format...)).  Use ATTRIBUTE_NULL_PRINTF at one place,
because callers expect to be able to pass NULL.

sim/common/ChangeLog:

* callback.c (os_printf_filtered, os_vprintf_filtered,
os_evprintf_filtered, os_error): Use ATTRIBUTE_PRINTF.
* sim-engine.h (sim_engine_abort, sim_engine_vabort): Likewise.
* sim-events.h (sim_events_schedule_tracef,
sim_events_schedule_vtracef): Use ATTRIBUTE_NULL_PRINTF.

Change-Id: Icd206f7b2c325e8b144f72eb129fb2a6b5af2fa3

3 years agogdb/continuations: turn continuation functions into inferior methods
Tankut Baris Aktemur [Thu, 22 Apr 2021 15:22:39 +0000 (17:22 +0200)]
gdb/continuations: turn continuation functions into inferior methods

Turn continuations-related functions into methods of the inferior
class.  This is a refactoring.

gdb/ChangeLog:
2021-04-22  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* Makefile.in (COMMON_SFILES): Remove continuations.c.
* inferior.c (inferior::add_continuation): New method, adapted
from 'add_inferior_continuation'.
(inferior::do_all_continuations): New method, adapted from
'do_all_inferior_continuations'.
(inferior::~inferior): Clear the list of continuations directly.
* inferior.h (class inferior) <continuations>: Rename into...
<m_continuations>: ...this and make private.
* continuations.c: Remove.
* continuations.h: Remove.
* event-top.c: Don't include "continuations.h".

Update the users below.
* inf-loop.c (inferior_event_handler)
* infcmd.c (attach_command)
(notice_new_inferior): Update.

3 years agogdb/continuations: use lambdas instead of function pointers
Tankut Baris Aktemur [Thu, 22 Apr 2021 15:22:39 +0000 (17:22 +0200)]
gdb/continuations: use lambdas instead of function pointers

Use lambdas and std::list to track inferior continuations.  This is a
refactoring.

gdb/ChangeLog:
2021-04-22  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* inferior.h (class inferior) <continuations>: Change the type
to be an std::list of std::function's.
Update the references and uses below.
* continuations.c (struct continuation): Delete.
(make_continuation): Delete.
(do_my_continuations_1): Delete.
(do_my_continuations): Delete.
(discard_my_continuations_1): Delete.
(discard_my_continuations): Delete.
(add_inferior_continuation): Update.
(do_all_inferior_continuations): Update.
(discard_all_inferior_continuations): Update.
* continuations.h (add_inferior_continuation): Update to take
an std::function as the parameter.
* infcmd.c (struct attach_command_continuation_args): Delete.
(attach_command_continuation): Delete.
(attach_command_continuation_free_args): Delete.
(attach_command): Update.
(notice_new_inferior): Update.

3 years agogdb/continuations: do minor cleanup
Tankut Baris Aktemur [Thu, 22 Apr 2021 15:22:39 +0000 (17:22 +0200)]
gdb/continuations: do minor cleanup

Inferior continuations are no longer used by the until and finish
command.  It is used only by the attach command and the remote target
upon detecting new inferiors.  Update the comment accordingly.

Also update another comment about non-existent thread continuations and
remove an unused #include.

gdb/ChangeLog:
2021-04-22  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* continuations.h: Update the general comment.
* inferior.h (class inferior) <continuations>: Update the comment.
* interps.c: Do not include "continuations.h".

3 years agogdb/continuations: remove the 'err' from 'do_all_inferior_continuations'
Tankut Baris Aktemur [Thu, 22 Apr 2021 15:22:38 +0000 (17:22 +0200)]
gdb/continuations: remove the 'err' from 'do_all_inferior_continuations'

The 'err' parameter of 'do_all_inferior_continuations' is effectively
unused.  There is only one place where the function is called, and
there the argument is a literal 0.  So, remove the parameter.

gdb/ChangeLog:
2021-04-22  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* continuations.h (do_all_inferior_continuations): Remove the 'err'
parameter.  Update the references below.
* continuations.c (do_my_continuations_1)
(do_my_continuations)
(do_all_inferior_continuations): Update.
* inf-loop.c (inferior_event_handler): Update.
* infcmd.c (attach_command_continuation): Update.

3 years agogdb/infcmd: update the comment for 'attach_post_wait'
Tankut Baris Aktemur [Thu, 22 Apr 2021 15:22:38 +0000 (17:22 +0200)]
gdb/infcmd: update the comment for 'attach_post_wait'

gdb/ChangeLog:
2021-04-22  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* infcmd.c (attach_post_wait): Update the function comment.

3 years agogdb/infcmd: remove the unused parameter 'args' in 'attach_post_wait'
Tankut Baris Aktemur [Thu, 22 Apr 2021 15:22:38 +0000 (17:22 +0200)]
gdb/infcmd: remove the unused parameter 'args' in 'attach_post_wait'

The 'arg' parameter of 'attach_post_wait' is unused.  Remove it.

gdb/ChangeLog:
2021-04-22  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* infcmd.c (attach_post_wait): Remove the unused parameter 'args'.
Update the references below.
(struct attach_command_continuation_args)
(attach_command_continuation)
(attach_command_continuation_free_args)
(attach_command)
(notice_new_inferior): Update to remove the reference to 'args'.

3 years ago[gdb] Fix assert in remote_async_get_pending_events_handler
Tom de Vries [Thu, 22 Apr 2021 15:01:00 +0000 (17:01 +0200)]
[gdb] Fix assert in remote_async_get_pending_events_handler

Occassionally I run into the following assert:
...
(gdb) PASS: gdb.multi/multi-target-continue.exp: inferior 5
Remote debugging from host ::1, port 49990^M
Process multi-target-continue created; pid = 31241^M
src/gdb/remote-notif.c:113: internal-error: \
  void remote_async_get_pending_events_handler(gdb_client_data): \
  Assertion `target_is_non_stop_p ()' failed.^M
...

The assert checks target_is_non_stop_p, which is related to the current
target.

Fix this by changing the assert such that it checks non-stopness related to
the event it's handling.

Tested on x86_64-linux.

gdb/ChangeLog:

2021-04-22  Simon Marchi  <simon.marchi@polymtl.ca>
    Tom de Vries  <tdevries@suse.de>

PR remote/27710
* remote.c (remote_target_is_non_stop_p): New function.
* remote.h (remote_target_is_non_stop_p): Declare.
* remote-notif.c (remote_async_get_pending_events_handler): Fix assert
to check non-stopness using notif_state->remote rather current target.

3 years agofix string table generation for XCOFF64 .debug section
Cl?ment Chigot [Thu, 22 Apr 2021 14:31:02 +0000 (15:31 +0100)]
fix string table generation for XCOFF64 .debug section

bfd * hash.c (struct bfd_strtab_hash): Remove xcoff field.
Add length_field_size field.
(_bfd_stringtab_init): Change prototype.
Adapt to new length_field_size.
(_bfd_xcoff_stringtab_init): Likewise.
(_bfd_stringtab_add): Likewise.
(_bfd_stringtab_emit): Likewise.
* libbfd-in.h (_bfd_xcoff_stringtab_init):
Change prototype.
* libbfd.h: Regenerate.
* xcofflink.c (_bfd_xcoff_bfd_link_hash_table_create):
Call _bfd_xcoff_stringtab_init with isxcoff64 value.

3 years agoImprove code coverage of Rust testing
Tom Tromey [Thu, 22 Apr 2021 13:16:36 +0000 (07:16 -0600)]
Improve code coverage of Rust testing

I enabled code coverage and ran the gdb test suite, and noticed that
the new Rust parser was missing testing on a few lines that were easy
to cover.  This patch mostly adds tests for certain syntax errors; but
this process also uncovered a couple of real bugs: I must have
cut-and-pasted the 'sizeof' parsing code from some other code, because
it is checking for KW_MUT (the old bison parser did not do this), and
the array length check is actually impossible because a negative
number like '-1' is parsed as two tokens.

gdb/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* rust-parse.c (rust_parser::parse_sizeof): Remove KW_MUT code.
(struct typed_val_int) <val>: Now ULONGEST.
(rust_parser::parse_array_type): Remove negative check.
(rust_lex_int_test): Change 'value' to ULONGEST.

gdb/testsuite/ChangeLog
2021-04-22  Tom Tromey  <tom@tromey.com>

* gdb.rust/modules.exp: Add checks for syntax errors.
* gdb.rust/expr.exp: Add checks for syntax errors.
* gdb.rust/simple.exp: Add checks for syntax errors.

3 years agoHarmonize and improve auxiliary entries support for XCOFF
Cl?ment Chigot [Thu, 22 Apr 2021 11:28:50 +0000 (12:28 +0100)]
Harmonize and improve auxiliary entries support for XCOFF

bfd/ChangeLog:
* coff-rs6000.c (_bfd_xcoff_swap_aux_in): Add errors for
unsupported storage class or auxialiry entries.
Improve and adapt to new aux structures.
Add C_DWARF support.
(_bfd_xcoff_swap_aux_out): Likewise.
* coff64-rs6000.c (_bfd_xcoff64_swap_aux_in): Likewise.
(_bfd_xcoff64_swap_aux_out): Likewise.

binutils/ChangeLog:
* od-xcoff.c (dump_xcoff32_symbols): Adapt to new
aux structures.

include/ChangeLog:
* coff/internal.h (union internal_auxent):
 Add x_sect structure.
* coff/rs6000.h (union external_auxent): Rework to
match official documentation.
* coff/rs6k64.h (union external_auxent): Likewise.
(_AUX_SECT): New define.

3 years agox86-64/PE: adjust PR ld/26659 testcase for Cygwin
Jan Beulich [Thu, 22 Apr 2021 08:55:07 +0000 (10:55 +0200)]
x86-64/PE: adjust PR ld/26659 testcase for Cygwin

While the testcase put in place by 74edb473c9ec ("PE/Windows x86_64: Fix
weak undef symbols after image base change") is fine for MingW, it fails
for Cygwin. This is because the default image base is different there
(for whatever reason).

3 years agogdb: allow default_addressable_memory_unit_size to handle more cases
Andrew Burgess [Wed, 21 Apr 2021 09:59:38 +0000 (10:59 +0100)]
gdb: allow default_addressable_memory_unit_size to handle more cases

Currently default_addressable_memory_unit_size always returns 1,
indicating 1 byte is 1 octet.  If a target has something other than
this (common) setup then the target should override the
default_addressable_memory_unit_size.

However, the bfd library already knows about each targets octets per
byte, so it seems redundant making targets override this method to
tell GDB something it already knows (through bfd).

In this commit I propose to make default_addressable_memory_unit_size
return a value based on bfd's bits per byte.  I checked, and for every
target that GDB currently supports the bits per byte in bfd is 8, so
the current behaviour will not change.

In fact, the only targets in bfd that have bits per byte set to
something other than 8 can be found in cpu-tic4x.c and cpu-tic54x.c, I
don't believe these are supported by GDB right now.

I don't propose to remove the ability to override
default_addressable_memory_unit_size, this allows targets additional
flexibility for how to handle weird combinations of byte sizes.

This change was motivated by an out of tree target I was working on,
but it seemed like it was a good change that others might benefit
from.

There should be no user visible changes after this commit.

gdb/ChangeLog:

* arch-utils.c (default_addressable_memory_unit_size): Return a
value based on bfd's bits per byte.

3 years agosim: regen against sim/m4/
Mike Frysinger [Thu, 22 Apr 2021 00:39:24 +0000 (20:39 -0400)]
sim: regen against sim/m4/

My scripts weren't updated to use new -I../m4 as the first arg.

3 years agoAutomatic date update in version.in
GDB Administrator [Thu, 22 Apr 2021 00:00:15 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agosim: use -Werror when probing for supported warning flags
Simon Marchi [Wed, 21 Apr 2021 18:51:10 +0000 (14:51 -0400)]
sim: use -Werror when probing for supported warning flags

When building with clang, we get:

    error: unknown warning option '-Wmissing-parameter-type' [-Werror,-Wunknown-warning-option]

This is because clang only warns by default when encountering an unknown
warning option, and the probe for supported warning flags is done
without -Werror.  All flags are therefore accepted by configure, but
then it breaks when actually compiling a source file with -Werror.

This is equivalent to this commit in gdb:

    3e019bdc20eb81b91ab5bc774386201b874c0bac
    gdb: Use -Werror when checking for (un)supported warning flags

We then see some other compilation errors when building with clang and
-Werror, they can be dealt with later.

3 years agoShrink size of dwarf2_per_cu_data
Tom Tromey [Wed, 21 Apr 2021 22:10:44 +0000 (16:10 -0600)]
Shrink size of dwarf2_per_cu_data

I noticed some holes in struct dwarf2_per_cu_data.  This patch
rearranges the type slightly, and shrinks the size of some fields.
This reduces it from 136 bytes to 112 bytes (on x86-64).

I also reduced the size of the DWARF "version" fields in a couple of
spots.  It seemed needless to use a short to hold a value that ranges
from 2 to 5, and this also helped the goal of shrinking
dwarf2_per_cu_data.

2021-04-21  Tom Tromey  <tom@tromey.com>

* dwarf2/read.h (struct dwarf2_per_cu_data) <dwarf_version>: Now
unsigned char.
(struct dwarf2_per_cu_data): Rearrange.
* dwarf2/comp-unit.h (struct comp_unit_head) <version>: Now
unsigned char.
(struct comp_unit_head): Rearrange.
* dwarf2/comp-unit.c (read_comp_unit_head): Update.

3 years agoFix test case gdb.base/valgrind-bt.exp.
Carl Love [Thu, 15 Apr 2021 22:21:51 +0000 (17:21 -0500)]
Fix test case gdb.base/valgrind-bt.exp.

gdb/testsuite/ChangeLog:

* gdb.base/valgrind-bt.exp: Add gdb_test "break main".
Update expected string for gdb_test "bt".

* lib/valgrind.exp: Add set remotetimeout 3.
Increase vgdb wait from 1 to 2.
Add max-invoke-ms option to vgdb command line.

3 years ago[gdb/build] Hardcode --with-included-regex
Tom de Vries [Wed, 21 Apr 2021 19:54:03 +0000 (21:54 +0200)]
[gdb/build] Hardcode --with-included-regex

Currently gdb has a configure option:
...
$ ./src/gdb/configure --help
  ...
  --without-included-regex
                          don't use included regex; this is the default on
                          systems with version 2 of the GNU C library (use
                          with caution on other system)
...

The configure option controls config.h macro USE_INCLUDED_REGEX, which is
used in gdb/gdb_regex.h to choose between:
- using regex from libiberty (which is included in the binutils-gdb.git repo,
  hence the 'included' in USE_INCLUDED_REGEX), or
- using regex.h.

In the former case, the symbol regcomp is remapped to a symbol xregcomp, which
is then provided by libiberty.

In the latter case, the symbol regcomp is resolved at runtime, usually binding
to libc.  However, there is no mechanism in place to enforce this.

PR27681 is an example of where that causes problems.  On openSUSE Tumbleweed,
the ncurses package got the --with-pcre2 configure switch enabled, and solved
the resulting dependencies using:
...
 $ cat /usr/lib64/libncursesw.so
 /* GNU ld script */
-INPUT(/lib64/libncursesw.so.6 AS_NEEDED(-ltinfo -ldl))
+INPUT(/lib64/libncursesw.so.6 AS_NEEDED(-ltinfo -ldl -lpcre2-posix -lpcre2-8))
...

This lead to regcomp being bound to libpcre2-posix instead of libc.

This causes problems in several ways:
- by compiling using regex.h, we've already chosen a specific regex_t
  implementation, and the one from pcre2-posix is not the same.
- in gdb_regex.c we use GNU regex function re_search, which pcre2-posix
  doesn't provide, so while regcomp binds to pcre2-posix, re_search binds to
  libc.

A note on the latter: it's actually a bug to compile a regex using regcomp and
then pass it to re_search.  The GNU regex interface requires one to use
re_compile_pattern or re_compile_fastmap.  But as long we're using one of the
GNU regex incarnations in gnulib, glibc or libiberty, we get away with this.

The PR could be fixed by adding -lc in a specific position in the link line,
to force regcomp to be bound to glibc.  But this solution was considered
in the discussion in the PR as being brittle, and possibly causing problems
elsewhere.

Another solution offered was to restrict regex usage to posix, and no longer
use the GNU regex API.  This however could mean having to reproduce some of
that functionality locally, which would mean maintaining the same
functionality in more than one place.

The solution chosen here, is to hardcode --with-included-regex, that is, using
libiberty.

The option of using glibc for regex was introduced because glibc became the
authorative source for GNU regex, so it offered the possibility to link
against a more up-to-date regex version.

In that aspect, this patch is a step back.  But we have the option of using a
more up-to-date regex version as a follow-up step: by using the regex from
gnulib.

Tested on x86_64-linux.

gdb/ChangeLog:

2021-04-21  Tom de Vries  <tdevries@suse.de>

PR build/27681
* configure.ac: Remove --without-included-regex/--with-included-regex.
* config.in: Regenerate.
* configure: Regenerate.
* gdb_regex.h: Assume USE_INCLUDED_REGEX is defined.

3 years agoExtend the description of PE header flags.
Eli Zaretskii [Wed, 21 Apr 2021 16:38:52 +0000 (17:38 +0100)]
Extend the description of PE header flags.

PR 27760
include * coff/pe.h (IMAGE_DLLCHARACTERISTICS_APPCONTAINER): Define.
(IMAGE_DLLCHARACTERISTICS_GUARD_CF): Define.

bfd * peXXigen.c (_bfd_XX_print_private_bfd_data_common): Add display
of IMAGE_FILE_REMOVABLE_RUN_FROM_SWAP,
IMAGE_FILE_NET_RUN_FROM_SWAP and IMAGE_FILE_UP_SYSTEM_ONLY flags.
Decode the contents of the DllCharacteristics field.

3 years agogdb/breakpoint: add a 'force_condition' parameter to 'create_breakpoint'
Tankut Baris Aktemur [Wed, 21 Apr 2021 14:42:40 +0000 (16:42 +0200)]
gdb/breakpoint: add a 'force_condition' parameter to 'create_breakpoint'

The 'create_breakpoint' function takes a 'parse_extra' argument that
determines whether the condition, thread, and force-condition
specifiers should be parsed from the extra string or be used from the
function arguments.  However, for the case when 'parse_extra' is
false, there is no way to pass the force-condition specifier.  This
patch adds it as a new argument.

Also, in the case when parse_extra is false, the current behavior is
as if the condition is being forced.  This is a bug.  The default
behavior should reject the breakpoint.  See below for a demo of this
incorrect behavior.  (The MI command '-break-insert' uses the
'create_breakpoint' function with parse_extra=0.)

  $ gdb -q --interpreter=mi3 /tmp/simple
  =thread-group-added,id="i1"
  =cmd-param-changed,param="history save",value="on"
  =cmd-param-changed,param="auto-load safe-path",value="/"
  ~"Reading symbols from /tmp/simple...\n"
  (gdb)
  -break-insert -c junk -f main
  &"warning: failed to validate condition at location 1, disabling:\n  "
  &"No symbol \"junk\" in current context.\n"
  ^done,bkpt={number="1",type="breakpoint",disp="keep",enabled="y",addr="<MULTIPLE>",cond="junk",times="0",original-location="main",locations=[{number="1.1",enabled="N",addr="0x000000000000114e",func="main",file="/tmp/simple.c",fullname="/tmp/simple.c",line="2",thread-groups=["i1"]}]}
  (gdb)
  break main if junk
  &"break main if junk\n"
  &"No symbol \"junk\" in current context.\n"
  ^error,msg="No symbol \"junk\" in current context."
  (gdb)
  break main -force-condition if junk
  &"break main -force-condition if junk\n"
  ~"Note: breakpoint 1 also set at pc 0x114e.\n"
  &"warning: failed to validate condition at location 1, disabling:\n  "
  &"No symbol \"junk\" in current context.\n"
  ~"Breakpoint 2 at 0x114e: file /tmp/simple.c, line 2.\n"
  =breakpoint-created,bkpt={number="2",type="breakpoint",disp="keep",enabled="y",addr="<MULTIPLE>",cond="junk",times="0",original-location="main",locations=[{number="2.1",enabled="N",addr="0x000000000000114e",func="main",file="/tmp/simple.c",fullname="/tmp/simple.c",line="2",thread-groups=["i1"]}]}
  ^done
  (gdb)

After applying this patch, we get the behavior below:

  (gdb)
  -break-insert -c junk -f main
  ^error,msg="No symbol \"junk\" in current context."

This restores the behavior that is present in the existing releases.

gdb/ChangeLog:
2021-04-21  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* breakpoint.h (create_breakpoint): Add a new parameter,
'force_condition'.
* breakpoint.c (create_breakpoint): Use the 'force_condition'
argument when 'parse_extra' is false to check if the condition
is invalid at all of the breakpoint locations.
Update the users below.
(break_command_1)
(dprintf_command)
(trace_command)
(ftrace_command)
(strace_command)
(create_tracepoint_from_upload): Update.
* guile/scm-breakpoint.c (gdbscm_register_breakpoint_x): Update.
* mi/mi-cmd-break.c (mi_cmd_break_insert_1): Update.
* python/py-breakpoint.c (bppy_init): Update.
* python/py-finishbreakpoint.c (bpfinishpy_init): Update.

gdb/testsuite/ChangeLog:
2021-04-21  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* gdb.mi/mi-break.exp: Extend with checks for invalid breakpoint
conditions.

3 years agotestsuite, gdb.mi: fix duplicate test names in mi-break.exp
Simon Marchi [Wed, 21 Apr 2021 14:42:39 +0000 (16:42 +0200)]
testsuite, gdb.mi: fix duplicate test names in mi-break.exp

gdb/testsuite/ChangeLog:
2021-04-21  Simon Marchi  <simon.marchi@polymtl.ca>
    Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* gdb.mi/mi-break.exp: Fix the duplicate test names.

3 years agogdb/breakpoint: display "N" on MI for disabled-by-condition locations
Tankut Baris Aktemur [Wed, 21 Apr 2021 14:42:39 +0000 (16:42 +0200)]
gdb/breakpoint: display "N" on MI for disabled-by-condition locations

For breakpoint locations that are disabled because of an invalid
condition, CLI displays "N*" in the 'enabled' field, where '*' refers
to the footnote below the table:

  (*): Breakpoint condition is invalid at this location.

This is not necessary for MI, where we shall simply print "N" without
the footnote.

Update the document to mention the "N" value for the MI.  Also remove
the line about the 'enable' field, because there is no such field for
locations.

gdb/ChangeLog:
2021-04-21  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* breakpoint.c (print_one_breakpoint_location): Display "N" for
disabled-by-condition locations on MI-like output.
(breakpoint_1): Do not display the disabled-by-condition footnote
if the output is MI-like.

gdb/doc/ChangeLog:
2021-04-21  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

* gdb.texinfo (GDB/MI Breakpoint Information): Update the
description for the 'enabled' field of breakpoint locations.

3 years agoAdd ability to select numeric base when displaying symbol values in readelf.
Nick Lott [Wed, 21 Apr 2021 14:33:13 +0000 (15:33 +0100)]
Add ability to select numeric base when displaying symbol values in readelf.

PR 27672
* readelf.c (sym_base): New variable.
(enum print_mode): Add more modes.
(print_vma): Add suport for new modes.
(options): Add sym-base.
(usage): Add sym-base.
(parse_args): Add support for --sym-base.
(print_dynamic_symbol_size): New function.
(print_dynamic_symbol): Use new function.
* doc/binutils.texi: Document the new feature.
* NEWS: Mention the new feature.

3 years agoupdate-netbsd.sh: fix script name, update year range in copyright.
Frederic Cambus [Wed, 21 Apr 2021 14:05:14 +0000 (08:05 -0600)]
update-netbsd.sh: fix script name, update year range in copyright.

Fix the script name and year range in update-netbsd.sh.

gdb/ChangeLog
2021-04-21  Frederic Cambus  <fred@statdns.com>

* syscalls/update-netbsd.sh: Fix script name display in usage, and
update year range in generated copyright notices.

3 years agoFix an assertion failure in the BFD library when parsing a corrupt SREC format file.
Nick Clifton [Wed, 21 Apr 2021 10:09:11 +0000 (11:09 +0100)]
Fix an assertion failure in the BFD library when parsing a corrupt SREC format file.

PR 27759
* srec.c (srec_read_section): Replace assertions with error
returns.

3 years agoAdjust readelf's output so that section symbols without a name as shown with their...
Nick Clifton [Wed, 21 Apr 2021 09:34:32 +0000 (10:34 +0100)]
Adjust readelf's output so that section symbols without a name as shown with their section name.

binutils* readelf.c (print_dynamic_symbol): Print the section name for
section symbols without a name of their own.
* testsuite/binutils-all/mips/global-local-symtab-sort-n64t.d:
Adjust expected output to allow for named section symbols.
* testsuite/binutils-all/mips/global-local-symtab-sort-o32t.d:
Likewise.
* testsuite/binutils-all/readelf.s-64: Likewise.
* testsuite/binutils-all/readelf.ss-64-unused: Likewise.
* testsuite/binutils-all/readelf.ss-tmips: Likewise.
* testsuite/binutils-all/readelf.ss-unused: Likewise.

ld * testsuite/ld-aarch64/variant_pcs-now.d: Adjust expected output
to allow for named section symbols.
* testsuite/ld-aarch64/variant_pcs-r.d: Likewise.
* testsuite/ld-aarch64/variant_pcs-shared.d: Likewise.
* testsuite/ld-alpha/tlsbin.rd: Likewise.
* testsuite/ld-alpha/tlsbinr.rd: Likewise.
* testsuite/ld-alpha/tlspic.rd: Likewise.
* testsuite/ld-arm/rodata-merge-map.sym: Likewise.
* testsuite/ld-arm/script-type.sym: Likewise.
* testsuite/ld-cris/libdso-2.d: Likewise.
* testsuite/ld-cris/pr16044.d: Likewise.
* testsuite/ld-elf/sec64k.exp: Likewise.
* testsuite/ld-ia64/tlsbin.rd: Likewise.
* testsuite/ld-ia64/tlspic.rd: Likewise.
* testsuite/ld-mips-elf/global-local-symtab-sort-n64t.d: Likewise.
* testsuite/ld-mips-elf/global-local-symtab-sort-o32t.d: Likewise.
* testsuite/ld-mmix/bspec1.d: Likewise.
* testsuite/ld-mmix/bspec2.d: Likewise.
* testsuite/ld-mmix/local1.d: Likewise.
* testsuite/ld-mmix/local3.d: Likewise.
* testsuite/ld-mmix/local5.d: Likewise.
* testsuite/ld-mmix/local7.d: Likewise.
* testsuite/ld-mmix/undef-3.d: Likewise.
* testsuite/ld-powerpc/tlsso.r: Likewise.
* testsuite/ld-powerpc/tlsso32.r: Likewise.
* testsuite/ld-powerpc/tlstocso.r: Likewise.
* testsuite/ld-s390/tlsbin.rd: Likewise.
* testsuite/ld-s390/tlsbin_64.rd: Likewise.
* testsuite/ld-s390/tlspic.rd: Likewise.
* testsuite/ld-s390/tlspic_64.rd: Likewise.
* testsuite/ld-sparc/gotop32.rd: Likewise.
* testsuite/ld-sparc/gotop64.rd: Likewise.
* testsuite/ld-sparc/tlssunbin32.rd: Likewise.
* testsuite/ld-sparc/tlssunbin64.rd: Likewise.
* testsuite/ld-sparc/tlssunnopic32.rd: Likewise.
* testsuite/ld-sparc/tlssunnopic64.rd: Likewise.
* testsuite/ld-sparc/tlssunpic32.rd: Likewise.
* testsuite/ld-sparc/tlssunpic64.rd: Likewise.
* testsuite/ld-tic6x/common.d: Likewise.
* testsuite/ld-tic6x/shlib-1.rd: Likewise.
* testsuite/ld-tic6x/shlib-1b.rd: Likewise.
* testsuite/ld-tic6x/shlib-1r.rd: Likewise.
* testsuite/ld-tic6x/shlib-1rb.rd: Likewise.
* testsuite/ld-tic6x/shlib-app-1.rd: Likewise.
* testsuite/ld-tic6x/shlib-app-1b.rd: Likewise.
* testsuite/ld-tic6x/shlib-app-1r.rd: Likewise.
* testsuite/ld-tic6x/shlib-app-1rb.rd: Likewise.
* testsuite/ld-tic6x/shlib-noindex.rd: Likewise.
* testsuite/ld-tic6x/static-app-1.rd: Likewise.
* testsuite/ld-tic6x/static-app-1b.rd: Likewise.
* testsuite/ld-tic6x/static-app-1r.rd: Likewise.
* testsuite/ld-tic6x/static-app-1rb.rd: Likewise.

gas * testsuite/gas/aarch64/dwarf.d: Adjust expected output to allow
for named section symbols.
* testsuite/gas/arm/thumbver.d: Likewise.
* testsuite/gas/bfin/loop_temps.d: Likewise.
* testsuite/gas/elf/section2.e-arc: Likewise.
* testsuite/gas/elf/section2.e-arm: Likewise.
* testsuite/gas/elf/section2.e-csky: Likewise.
* testsuite/gas/elf/section2.e-m32r: Likewise.
* testsuite/gas/elf/section2.e-mips: Likewise.
* testsuite/gas/elf/section2.e-msp430: Likewise.
* testsuite/gas/elf/section2.e-riscv: Likewise.
* testsuite/gas/elf/section2.e-rl78: Likewise.
* testsuite/gas/elf/section2.e-rx: Likewise.
* testsuite/gas/elf/section2.e-score: Likewise.
* testsuite/gas/elf/section2.e-tic6x: Likewise.
* testsuite/gas/elf/section2.e-unused: Likewise.
* testsuite/gas/elf/section2.e-v850: Likewise.
* testsuite/gas/elf/section2.e-xtensa: Likewise.
* testsuite/gas/ia64/alias-ilp32.d: Likewise.
* testsuite/gas/ia64/alias.d: Likewise.
* testsuite/gas/ia64/global.d: Likewise.
* testsuite/gas/microblaze/relax_size.elf: Likewise.
* testsuite/gas/microblaze/relax_size2.elf: Likewise.
* testsuite/gas/mips/global-local-symtab-sort-n64t.d: Likewise.
* testsuite/gas/mips/global-local-symtab-sort-o32t.d: Likewise.
* testsuite/gas/mmix/bspec-1.d: Likewise.
* testsuite/gas/mmix/byte-1.d: Likewise.
* testsuite/gas/mmix/comment-1.d: Likewise.
* testsuite/gas/mmix/loc-1.d: Likewise.
* testsuite/gas/mmix/loc-2.d: Likewise.
* testsuite/gas/mmix/loc-3.d: Likewise.
* testsuite/gas/mmix/loc-4.d: Likewise.
* testsuite/gas/mmix/loc-5.d: Likewise.
* testsuite/gas/tic6x/scomm-directive-4.d: Likewise.

3 years agoAutomatic date update in version.in
GDB Administrator [Wed, 21 Apr 2021 00:00:16 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agoRework the R_NEG support on both gas and ld for the PowerPC AIX targets, in order...
Cl?ment Chigot [Tue, 20 Apr 2021 13:40:43 +0000 (14:40 +0100)]
Rework the R_NEG support on both gas and ld for the PowerPC AIX targets, in order to manage C++ exceptions built with GCC.

bfd PR binutils/21700
* reloc.c (BFD_RELOC_PPC_NEG): New relocation.
* bfd-in2.h: Regenerate.
* libbfd.h: Regenerate.
* coff-rs6000.c (_bfd_xcoff_reloc_type_lookup): Add
BFD_RELOC_PPC_NEG handler.
(xcoff_reloc_type_neg): Correctly substract addend.
* coff64-rs6000.c (xcoff64_howto_table): Add R_NEG_32
howto.
(xcoff64_rtype2howto): Add handler for R_NEG_32.
(xcoff64_reloc_type_lookup): Add BFD_RELOC_PPC_NEG handler.
* xcofflink.c (xcoff_need_ldrel_p): Check output section
for R_POS-like relocations. New argument added.
(xcoff_mark): Adapt to new xcoff_need_ldrel_p argument.
(xcoff_link_input_bfd): Likewise.

gas * config/tc-ppc.c (ppc_get_csect_to_adjust): New function.
(ppc_fix_adjustable): Manage fx_subsy part.
(tc_gen_reloc): Create second relocation when both
fx_addsy and fx_subsy are provided.
* config/tc-ppc.h (RELOC_EXPANSION_POSSIBLE): New define.
(MAX_RELOC_EXPANSION): Likewise.
(TC_FORCE_RELOCATION_SUB_SAME): Likewise
(UNDEFINED_DIFFERENCE_OK): Likewise
* testsuite/gas/all/gas.exp: Skip difference between two
undefined symbols test.

ld * testsuite/ld-powerpc/aix52.exp: Add new test.
* testsuite/ld-powerpc/aix-neg-reloc-32.d: New test.
* testsuite/ld-powerpc/aix-neg-reloc-64.d: New test.
* testsuite/ld-powerpc/aix-neg-reloc.ex: New test.
* testsuite/ld-powerpc/aix-neg-reloc.s: New test.

3 years agosim: rl78/rx: drop unnecessary getopt.h probing
Mike Frysinger [Tue, 20 Apr 2021 13:21:28 +0000 (09:21 -0400)]
sim: rl78/rx: drop unnecessary getopt.h probing

Since getopt.h is provided by libiberty, there's no need to probe for
a system version of it.  Plus we already assume it exists in other
parts of the sim.

3 years agoMAINTAINERS: Remove Martin Schwidefsky as s390 maintainer
Andreas Krebbel [Tue, 20 Apr 2021 07:18:02 +0000 (09:18 +0200)]
MAINTAINERS: Remove Martin Schwidefsky as s390 maintainer

https://lwn.net/ml/linux-kernel/20190521162350.GA17107@osiris/

ChangeLog:

2021-04-20  Andreas Krebbel  <krebbel@linux.ibm.com>

* MAINTAINERS: Remove Martin Schwidefsky as s390 maintainer and
add him to Past Maintainers.
Update my email address.

3 years agogdb: Fix reduce/reduce conflicts for qualifier_seq_noopt in the C parser.
Felix Willgerodt [Tue, 20 Apr 2021 06:44:17 +0000 (08:44 +0200)]
gdb: Fix reduce/reduce conflicts for qualifier_seq_noopt in the C parser.

This fixes a problem with GDB's address space qualifier parsing.  GDB uses
'@' as a way to express an address space in expression evaluation.  This can
currently lead to a crash for "Add support for the __flash qualifier on AVR"
(487d975399dfcb2bb2f0998a7d12bd62acdd9fa1), the only user I am aware of.

Program:
~~~
const __flash char data_in_flash = 0xab;

int
main (void)
{
  const __flash char *pointer_to_flash = &data_in_flash;
}
~~~

Before:
~~~
(gdb) p data_in_flash
$1 = -85 '\253'
(gdb) p *(const char * @flash) pointer_to_flash
$2 = -85 '\253'
(gdb) p *(@flash const char *) pointer_to_flash
type-stack.c:201: internal-error: type* type_stack::follow_types(type*): unrecognized tp_ value in follow_types
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n)
~~~

After:
~~~
(gdb) p data_in_flash
$1 = -85 '\253'
(gdb) p *(const char *) pointer_to_flash
$2 = 0 '\000'
(gdb) p *(const char * @flash) pointer_to_flash
$3 = -85 '\253'
(gdb) p *(@flash const char *) pointer_to_flash
$4 = 0 '\000'
(gdb)
~~~

Note that how the binding of this qualifier is interpreted and resolved for an
address/pointer is target specific.  Hence only the prepended qualifier works
for AVR, even if it seems syntactically incorrect.  I won't change this for
AVR, as I am not familiar with that target.

Bison now also complains about less conflicts:

Before:
  YACC   c-exp.c
gdb/gdb/c-exp.y: warning: 153 shift/reduce conflicts [-Wconflicts-sr]
gdb/gdb/c-exp.y: warning: 70 reduce/reduce conflicts [-Wconflicts-rr]

After:
  YACC   c-exp.c
gdb/gdb/c-exp.y: warning: 60 shift/reduce conflicts [-Wconflicts-sr]
gdb/gdb/c-exp.y: warning: 69 reduce/reduce conflicts [-Wconflicts-rr]

gdb/ChangeLog:
2021-04-20  Felix Willgerodt  <felix.willgerodt@intel.com>

* c-exp.y (qualifier_seq_noopt): Replace qualifier_seq with
qualifier_seq_noopt.

3 years agogdb: Allow address space qualifier parsing in C++.
Felix Willgerodt [Tue, 20 Apr 2021 06:43:08 +0000 (08:43 +0200)]
gdb: Allow address space qualifier parsing in C++.

The goal of this patch is to allow target dependent address space qualifiers
in the C++ expression parser.  This can be useful for memory examination on
targets that actually use different address spaces in hardware without
having to deep-dive into implementation details of the whole solution.

GDB uses the @ symbol to parse address space qualifiers.  The only current
user that I am aware of is the __flash support for avr, which was added in
"Add support for the __flash qualifier on AVR"
(487d975399dfcb2bb2f0998a7d12bd62acdd9fa1)
and only works for C.

One use-case of the AVR patch is:

~~~
const __flash char data_in_flash = 0xab;

int
main (void)
{
  const __flash char *pointer_to_flash = &data_in_flash;
}
~~~

~~~
(gdb) print pointer_to_flash
$1 = 0x1e8 <data_in_flash> "\253"
(gdb) print/x *pointer_to_flash
$2 = 0xab
(gdb) x/x pointer_to_flash
0x1e8 <data_in_flash>: 0xXXXXXXab
(gdb)
(gdb) p/x *(char* @flash) pointer_to_flash
$3 = 0xab
~~~

I want to enable a similar usage of e.g. @local in C++.

Before this patch (using "set debug parser on"):

~~~
(gdb) p *(int* @local) 0x1234
(...)
Reading a token: Next token is token '@' ()
Shifting token '@' ()
Entering state 46
Reading a token: Next token is token UNKNOWN_CPP_NAME (ssym<name=local, sym=(null), field_of_this=0>)
A syntax error in expression, near `local) &x'.
~~~

After:
~~~
(gdb) p *(int* @local) 0x1234
(...)
Reading a token: Next token is token '@' ()
Shifting token '@' ()
Entering state 46
Reading a token: Next token is token UNKNOWN_CPP_NAME (ssym<name=local, sym=(null), field_of_this=0>)
Shifting token UNKNOWN_CPP_NAME (ssym<name=local, sym=(null), field_of_this=0>)
Entering state 121
Reducing stack by rule 278 (line 1773):
   $1 = token UNKNOWN_CPP_NAME (ssym<name=local, sym=(null), field_of_this=0>)
-> $$ = nterm name ()
Stack now 0 49 52 76 222 337 46
Entering state 167
Reducing stack by rule 131 (line 1225):
   $1 = token '@' ()
   $2 = nterm name ()
Unknown address space specifier: "local"
~~~

The "Unknown address space qualifier" is the right behaviour, as I ran this
on a target that doesn't have multiple address spaces and therefore obviously
no support for such qualifiers.

gdb/ChangeLog:
2021-04-20  Felix Willgerodt  <felix.willgerodt@intel.com>

* c-exp.y (single_qualifier): Handle UNKNOWN_CPP_NAME.

gdb/testsuite/ChangeLog:
2021-04-20  Felix Willgerodt  <felix.willgerodt@intel.com>

* gdb.base/address_space_qualifier.exp: New file.

3 years agoAutomatic date update in version.in
GDB Administrator [Tue, 20 Apr 2021 00:00:16 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agogdb: use compiled_regex instead of std::regex
Andrew Burgess [Thu, 15 Apr 2021 17:30:57 +0000 (18:30 +0100)]
gdb: use compiled_regex instead of std::regex

In GDB we should be using compiled_regex instead of std::regex.
Replace one use in producer.c.

There should be no user visible changes after this commit.

gdb/ChangeLog:

* producer.c: Replace 'regex' include with 'gdb_regex.h'.
(producer_is_icc): Replace use of std::regex with gdb's
compiled_regex.

3 years agoFix a problem running the archiver program in MRI mode on archives containing LTO...
Nick Clifton [Mon, 19 Apr 2021 16:38:39 +0000 (17:38 +0100)]
Fix a problem running the archiver program in MRI mode on archives containing LTO compiled objects.

PR 21702
* arsup.c (ar_addmod): Enable plugin support, if available.

3 years agoFix an assembler testuite failure when checking a toolchain configured with --target...
Nick Clifton [Mon, 19 Apr 2021 16:36:43 +0000 (17:36 +0100)]
Fix an assembler testuite failure when checking a toolchain configured with --target=rs6000-aix3.3.

* testsuite/gas/all/gas.exp: Add rs6000*-*-aix* to the list of
targets which should skip the undefined symbols test.

3 years agoaarch64: New instructions for maintenance of GPT entries cached in a TLB
Przemyslaw Wirkus [Mon, 19 Apr 2021 14:00:07 +0000 (15:00 +0100)]
aarch64: New instructions for maintenance of GPT entries cached in a TLB

This patch adds support to four new system registers (RPAOS, RPALOS, PAALLOS,
PAALL) in conjunction with TLBI instruction. This change is part of RME (Realm
Management Extension).

gas/ChangeLog:

2021-04-19  Przemyslaw Wirkus  <przemyslaw.wirkus@arm.com>

* NEWS: Update news.
* testsuite/gas/aarch64/rme.d: Update test.
* testsuite/gas/aarch64/rme.s: Update test.

opcodes/ChangeLog:

2021-04-19  Przemyslaw Wirkus  <przemyslaw.wirkus@arm.com>

* aarch64-opc.c: Add new registers (RPAOS, RPALOS, PAALLOS, PAALL) support for
TLBI instruction.

3 years agoaarch64: Add new data cache maintenance operations
Przemyslaw Wirkus [Mon, 19 Apr 2021 13:54:46 +0000 (14:54 +0100)]
aarch64: Add new data cache maintenance operations

This patch adds support to two new system registers (CIPAPA, CIGDPAPA) in
conjunction with DC instruction. This change is part of RME (Realm Management
Extension).

gas/ChangeLog:

2021-04-19  Przemyslaw Wirkus  <przemyslaw.wirkus@arm.com>

* testsuite/gas/aarch64/rme.d: Update test.
* testsuite/gas/aarch64/rme.s: Update test.

opcodes/ChangeLog:

2021-04-19  Przemyslaw Wirkus  <przemyslaw.wirkus@arm.com>

* aarch64-opc.c: Add new register (CIPAPA, CIGDPAPA) support for
DC instruction.

3 years agoarm64: add two initializers
Jan Beulich [Mon, 19 Apr 2021 13:41:35 +0000 (15:41 +0200)]
arm64: add two initializers

Old enough gcc can't cope and would warn about the variables potentially
remaining uninitialized.

3 years agoFix kfail patterns in inline-locals.exp
Tom Tromey [Mon, 19 Apr 2021 13:16:34 +0000 (07:16 -0600)]
Fix kfail patterns in inline-locals.exp

PR gdb/27742 points out that my recent change to
print_variable_and_value caused a regression in inline-locals.exp.  I
can't reproduce this, but I came up with this patch based on the
output shown in the bug.

gdb/testsuite/ChangeLog
2021-04-19  Tom Tromey  <tromey@adacore.com>

PR gdb/27742:
* gdb.opt/inline-locals.exp: Update kfail patterns.

3 years agogas: drop sprint_value()
Jan Beulich [Mon, 19 Apr 2021 11:28:35 +0000 (13:28 +0200)]
gas: drop sprint_value()

Its (documented) behavior is unhelpful in particular in 64-bit build
environments: While printing large 32-bit numbers in decimal already
isn't very meaningful to most people, this even more so goes for yet
larger 64-bit numbers. bfd_sprintf_vma() still tries to limit the number
of digits printed (without depending on a build system property), but
uniformly produces hex output.

3 years agoFix compile time warning about unused functions.
Nick Clifton [Mon, 19 Apr 2021 08:41:04 +0000 (09:41 +0100)]
Fix compile time warning about unused functions.

binutils * rename.c: (get_stat_atime_ns): Add prototype.
(get_stat_mtime_ns): Add prototype.

3 years agosim: switch to AC_CHECK_HEADERS_ONCE
Mike Frysinger [Mon, 19 Apr 2021 03:44:36 +0000 (23:44 -0400)]
sim: switch to AC_CHECK_HEADERS_ONCE

This avoids duplicate tests for headers between common m4, arches,
and any other sources that would trigger header tests.

3 years agosim: msp430: delete unused getopt.h probe
Mike Frysinger [Mon, 19 Apr 2021 03:33:31 +0000 (23:33 -0400)]
sim: msp430: delete unused getopt.h probe

This port doesn't include it anywhere, so drop the test.

3 years agosim: moxie: switch syscalls to common nltvals
Mike Frysinger [Mon, 19 Apr 2021 02:48:31 +0000 (22:48 -0400)]
sim: moxie: switch syscalls to common nltvals

Rather than hand duplicate the syscall constants, switch to the
common nltvals framework.  I made sure the constants have the
same values before & after too :).

3 years agosim: sh: switch syscalls to common nltvals
Mike Frysinger [Mon, 19 Apr 2021 02:43:06 +0000 (22:43 -0400)]
sim: sh: switch syscalls to common nltvals

Rather than hand duplicate the syscall table, switch to the common
nltvals framework.  We have to tweak the constant names, but we get
everything else for free.  I made sure the constants have the same
values before & after too :).

3 years agosim: rx: switch syscalls to common nltvals
Mike Frysinger [Mon, 19 Apr 2021 02:35:19 +0000 (22:35 -0400)]
sim: rx: switch syscalls to common nltvals

Rather than hand duplicate the syscall table, switch to the common
nltvals framework.  We have to tweak the constant names, but we get
everything else for free.  I made sure the constants have the same
values before & after too :).

3 years agosim: m32c: switch syscalls to common nltvals
Mike Frysinger [Mon, 19 Apr 2021 02:31:19 +0000 (22:31 -0400)]
sim: m32c: switch syscalls to common nltvals

Rather than hand duplicate the syscall table, switch to the common
nltvals framework.  We have to tweak the constant names, but we get
everything else for free.  I made sure the constants have the same
values before & after too :).

3 years agosim: iq2000: switch syscalls to common nltvals
Mike Frysinger [Mon, 19 Apr 2021 02:24:37 +0000 (22:24 -0400)]
sim: iq2000: switch syscalls to common nltvals

Rather than hand duplicate the syscall table, switch to the common
nltvals framework.  We have to tweak the constant names, but we get
everything else for free.  I made sure the constants have the same
values before & after too :).

3 years agosim: switch to AC_CHECK_FUNCS_ONCE & merge a little
Mike Frysinger [Mon, 19 Apr 2021 01:50:07 +0000 (21:50 -0400)]
sim: switch to AC_CHECK_FUNCS_ONCE & merge a little

This avoids duplicate tests for functions between common m4, arches,
and any other sources that would trigger func tests.

Also manually delete known duplicate function tests between the m4,
bfin, and v850 ports.

3 years agosim: mn10300: delete unused func & header tests
Mike Frysinger [Mon, 19 Apr 2021 01:30:24 +0000 (21:30 -0400)]
sim: mn10300: delete unused func & header tests

These appear to have been blindly copied from the v850 port many years
ago as the code has never been used.  Just delete it all and be done.

3 years agosim: syscall: add getpid support
Mike Frysinger [Mon, 19 Apr 2021 00:53:03 +0000 (20:53 -0400)]
sim: syscall: add getpid support

Hoist the Blackfin implementation up to the common one.

3 years agosim: d10v: fix build warnings
Mike Frysinger [Mon, 19 Apr 2021 01:13:34 +0000 (21:13 -0400)]
sim: d10v: fix build warnings

The printf fix uses our PRIxTA for our sim address type.

Then cast away the const (since the underlying code safely treats it
as such) even if it's ugly.

Finally touch up the argv iterator pointer to match the new func arg.

With this tidied up, we can delete the SIM_AC_OPTION_WARNINGS(no) call
to get the default common behavior where -Werror is enabled.

3 years agosim: cr16: fix build warnings
Mike Frysinger [Mon, 19 Apr 2021 01:01:00 +0000 (21:01 -0400)]
sim: cr16: fix build warnings

The printf fix is obvious enough, but the hash one is a real bug:
cr16/interp.c: In function 'sim_open':
cr16/interp.c:560:17: error: 'h' may be used uninitialized in this function [-Werror=maybe-uninitialized]
  560 |               h = h->next;
      |               ~~^~~~~~~~~

It happens to not cause a problem currently because the first entry in
the generated table that this loop operates matches a codepath where h
is initialized.  Then when later entries don't match, the previous value
is pointing at the end of a valid hash table already, and the rest of
the code does nothing.

With this tidied up, we can delete the SIM_AC_OPTION_WARNINGS(no) call
to get the default common behavior where -Werror is enabled.

3 years agoAutomatic date update in version.in
GDB Administrator [Mon, 19 Apr 2021 00:00:13 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agoAutomatic date update in version.in
GDB Administrator [Sun, 18 Apr 2021 00:00:15 +0000 (00:00 +0000)]
Automatic date update in version.in

3 years agoHandle unaligned mapping of .gdb_index
Tom Tromey [Sat, 17 Apr 2021 19:56:36 +0000 (13:56 -0600)]
Handle unaligned mapping of .gdb_index

The .gdb_index was designed such that all data would be aligned.
Unfortunately, we neglected to require this alignment in the objcopy
instructions in the manual.  As a result, in many cases, a .gdb_index
in the wild will not be properly aligned by mmap.  This yields
undefined behavior, which is PR gdb/23743.

This patch fixes the bug by always assuming that the mapping is
unaligned, and using extract_unsigned_integer when needed.  A new
helper class is introduced to make this less painful.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

PR gdb/23743:
* dwarf2/read.c (class offset_view): New.
(struct symbol_table_slot): Remove.
(struct mapped_index) <symbol_table, constant_pool>: Change type.
<symbol_name_index, symbol_vec_index>: New methods.
<symbol_name_slot_invalid, symbol_name_at, symbol_name_count>:
Rewrite.
(read_gdb_index_from_buffer): Update.
(struct dw2_symtab_iterator) <vec>: Change type.
(dw2_symtab_iter_init_common, dw2_symtab_iter_init)
(dw2_symtab_iter_next, dw2_expand_marked_cus): Update.
* dwarf2/index-write.c (class data_buf) <append_data>: Remove.
<append_array, append_offset>: New methods.
(write_hash_table, add_address_entry, write_gdbindex_1)
(write_debug_names): Update.
* dwarf2/index-common.h (byte_swap, MAYBE_SWAP): Remove.

3 years agoAvoid crash in write_psymtabs_to_index
Tom Tromey [Sat, 17 Apr 2021 19:40:04 +0000 (13:40 -0600)]
Avoid crash in write_psymtabs_to_index

If I try "save gdb-index" using the executable from
gdb.cp/cmpd-minsyms.exp, gdb will crash.  This happens due to a
missing NULL check.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* dwarf2/index-write.c (write_psymtabs_to_index): Check
partial_symtabs.

gdb/testsuite/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* gdb.dwarf2/gdb-index-nodebug.exp: New file.

3 years agoSimplify quick_symbol_functions::map_matching_symbols
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Simplify quick_symbol_functions::map_matching_symbols

quick_symbol_functions::map_matching_symbols is only used by the Ada
code.  Currently, it both expands certain psymtabs and then walks over
the full symtabs -- including any already-expanded ones -- calling a
callback.

It appears to work lazily as well, in that if the callback returns
false, iteration stops.  However, only the psymtab implementation does
this; the DWARF index implementations are not lazy.  It turns out,
though, that the only callback that is ever passed here never returns
false.

This patch simplifies this method by removing the callback.  The
method is also renamed.  In the new scheme, the caller is responsible
for walking the full symtabs, which removes some redundancy as well.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* psymtab.c (psymbol_functions::expand_matching_symbols): Rename
from map_matching_symbols.  Change parameters.
* psympriv.h (struct psymbol_functions) <expand_matching_symbols>:
Rename from map_matching_symbols.  Change parameters.
* dwarf2/read.c (struct dwarf2_gdb_index)
<expand_matching_symbols>: Rename from map_matching_symbols.
Change parameters.
(struct dwarf2_debug_names_index) <expand_matching_symbols>:
Rename from map_matching_symbols.  Change parameters.
(dwarf2_gdb_index::expand_matching_symbols): Rename from
dw2_map_matching_symbols.  Change parameters.
(dwarf2_gdb_index::expand_matching_symbols): Remove old
implementation.
(dwarf2_debug_names_index::expand_matching_symbols): Rename from
map_matching_symbols.  Change parameters.
* objfiles.h (struct objfile) <expand_matching_symbols>: Rename
from map_matching_symbols.  Change parameters.
* symfile-debug.c (objfile::expand_matching_symbols): Rename from
map_matching_symbols.  Change parameters.
* ada-lang.c (map_matching_symbols): New function.
(add_nonlocal_symbols): Update.

3 years agoRemove quick_symbol_functions::expand_symtabs_with_fullname
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Remove quick_symbol_functions::expand_symtabs_with_fullname

This removes quick_symbol_functions::expand_symtabs_with_fullname,
replacing it with a call to expand_symtabs_matching.  As with the
previous patches, the implementation is consolidated in the objfile
method.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* quick-symbol.h (struct quick_symbol_functions)
<expand_symtabs_with_fullname>: Remove.
* psymtab.c (psymbol_functions::expand_symtabs_with_fullname):
Remove.
* psympriv.h (struct psymbol_functions)
<expand_symtabs_with_fullname>: Remove.
* dwarf2/read.c (struct dwarf2_base_index_functions)
<expand_symtabs_with_fullname>: Remove.
(dwarf2_base_index_functions::expand_symtabs_with_fullname):
Remove.
* objfiles.h (struct objfile) <expand_symtabs_with_fullname>:
Update comment.
* symfile-debug.c (objfile::expand_symtabs_with_fullname):
Rewrite.

3 years agoRemove quick_symbol_functions::expand_symtabs_for_function
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Remove quick_symbol_functions::expand_symtabs_for_function

This removes quick_symbol_functions::expand_symtabs_for_function,
replacing it with a call to expand_symtabs_matching.  As with the
previous patches, the implementation is consolidated in the objfile
method.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* symfile-debug.c (objfile::expand_symtabs_for_function):
Rewrite.
* quick-symbol.h (struct quick_symbol_functions)
<expand_symtabs_for_function>: Remove.
* psymtab.c (psymbol_functions::expand_symtabs_for_function):
Remove.
* psympriv.h (struct psymbol_functions)
<expand_symtabs_for_function>: Remove.
* objfiles.h (struct objfile) <expand_symtabs_for_function>:
Update comment.
* dwarf2/read.c (struct dwarf2_gdb_index)
<expand_symtabs_for_function>: Remove.
(struct dwarf2_debug_names_index) <expand_symtabs_for_function>:
Remove.
(find_slot_in_mapped_hash): Remove.
(dw2_symtab_iter_init_common): Merge with dw2_symtab_iter_init.
(dw2_symtab_iter_init): Remove one overload.
(dwarf2_gdb_index::expand_symtabs_for_function)
(dwarf2_debug_names_index::expand_symtabs_for_function): Remove.

3 years agoRemove quick_symbol_functions::map_symtabs_matching_filename
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Remove quick_symbol_functions::map_symtabs_matching_filename

This replaces quick_symbol_functions::map_symtabs_matching_filename
with a call to expand_symtabs_matching.  As with the previous patch,
rather than update all callers, the implementation is consolidated in
objfile::map_symtabs_matching_filename.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* symfile-debug.c (objfile::map_symtabs_matching_filename):
Rewrite.
* quick-symbol.h (struct quick_symbol_functions)
<map_symtabs_matching_filename>: Remove.
* psymtab.c (partial_map_expand_apply)
(psymbol_functions::map_symtabs_matching_filename): Remove.
* psympriv.h (struct psymbol_functions)
<map_symtabs_matching_filename>: Remove.
* objfiles.h (struct objfile) <map_symtabs_matching_filename>:
Update comment.
* dwarf2/read.c (struct dwarf2_base_index_functions)
<map_symtabs_matching_filename>: Remove.
(dw2_map_expand_apply)
(dwarf2_base_index_functions::map_symtabs_matching_filename):
Remove.

3 years agoRemove quick_symbol_functions::lookup_symbol
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Remove quick_symbol_functions::lookup_symbol

This removes quick_symbol_functions, replacing it with calls to
expand_symtabs_matching.  Because the replacement is somewhat verbose,
objfile::lookup_symbol is not removed.  This consolidates some
duplicated code into this one spot.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* symfile-debug.c (objfile::lookup_symbol): Rewrite.
* quick-symbol.h (struct quick_symbol_functions) <lookup_symbol>:
Remove.
* psymtab.c (psymbol_functions::lookup_symbol): Remove.
* psympriv.h (struct psymbol_functions) <lookup_symbol>: Remove.
* objfiles.h (struct objfile) <lookup_symbol>: Add comment.
* dwarf2/read.c (struct dwarf2_gdb_index) <lookup_symbol>:
Remove.
(struct dwarf2_debug_names_index) <lookup_symbol>: Remove.
(dwarf2_gdb_index::lookup_symbol)
(dwarf2_debug_names_index::lookup_symbol): Remove.

3 years agoAdd 'domain' parameter to expand_symtabs_matching
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Add 'domain' parameter to expand_symtabs_matching

Currently, expand_symtabs_matching only accepts a search_domain
parameter.  However, lookup_symbol uses a domain_enum instead, and the
two, confusingly, do quite different things -- one cannot emulate the
other.  So, this patch adds a domain_enum parameter to
expand_symtabs_matching, with UNDEF_DOMAIN used as a wildcard.

This is another step toward replacing lookup_symbol with
expand_symtabs_matching.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* symtab.c (global_symbol_searcher::expand_symtabs): Update.
* symmisc.c (maintenance_expand_symtabs): Update.
* symfile.c (expand_symtabs_matching): Update.
* symfile-debug.c (objfile::expand_symtabs_matching): Add 'domain'
parameter.
* quick-symbol.h (struct quick_symbol_functions)
<expand_symtabs_matching>: Add 'domain' parameter.
* psymtab.c (recursively_search_psymtabs)
(psymbol_functions::expand_symtabs_matching): Add 'domain'
parameter.
* psympriv.h (struct psymbol_functions) <expand_symtabs_matching>:
Add 'domain' parameter.
* objfiles.h (struct objfile) <expand_symtabs_matching>: Add
'domain' parameter.
* linespec.c (iterate_over_all_matching_symtabs): Update.
* dwarf2/read.c (struct dwarf2_gdb_index)
<expand_symtabs_matching>: Add 'domain' parameter.
(struct dwarf2_debug_names_index) <expand_symtabs_matching>: Add
'domain' parameter.
(dw2_expand_symtabs_matching)
(dwarf2_gdb_index::expand_symtabs_matching)
(dw2_debug_names_iterator)
(dwarf2_debug_names_index::expand_symtabs_matching): Add 'domain'
parameter.

3 years agoAdd search_flags to expand_symtabs_matching
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Add search_flags to expand_symtabs_matching

This adds a block search flags parameter to expand_symtabs_matching.
All callers are updated to search both the static and global blocks,
as that was the implied behavior before this patch.

This is a step toward replacing lookup_symbol with
expand_symtabs_matching.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* symtab.c (global_symbol_searcher::expand_symtabs)
(default_collect_symbol_completion_matches_break_on): Update.
* symmisc.c (maintenance_expand_symtabs): Update.
* symfile.h (expand_symtabs_matching): Add search_flags
parameter.
* symfile.c (expand_symtabs_matching): Add search_flags
parameter.
* symfile-debug.c (objfile::expand_symtabs_matching): Add
search_flags parameter.
* quick-symbol.h (struct quick_symbol_functions)
<expand_symtabs_matching>: Add search_flags parameter.
* python/py-symbol.c (gdbpy_lookup_static_symbols): Update.
* psymtab.c (recursively_search_psymtabs)
(psymbol_functions::expand_symtabs_matching): Add search_flags
parameter.
* psympriv.h (struct psymbol_functions) <expand_symtabs_matching>:
Add search_flags parameter.
* objfiles.h (struct objfile) <expand_symtabs_matching>: Add
search_flags parameter.
* linespec.c (iterate_over_all_matching_symtabs): Update.
* dwarf2/read.c (struct dwarf2_gdb_index)
<expand_symtabs_matching>: Add search_flags parameter.
(struct dwarf2_debug_names_index) <expand_symtabs_matching>: Add
search_flags parameter.
(dw2_map_matching_symbols): Update.
(dw2_expand_marked_cus, dw2_expand_symtabs_matching)
(dwarf2_gdb_index::expand_symtabs_matching): Add search_flags
parameter.
(dw2_debug_names_iterator): Change block_index to search flags.
<m_block_index>: Likewise.
(dw2_debug_names_iterator::next)
(dwarf2_debug_names_index::lookup_symbol)
(dwarf2_debug_names_index::expand_symtabs_for_function)
(dwarf2_debug_names_index::map_matching_symbols)
(dwarf2_debug_names_index::map_matching_symbols): Update.
(dwarf2_debug_names_index::expand_symtabs_matching): Add
search_flags parameter.
* ada-lang.c (ada_add_global_exceptions)
(collect_symbol_completion_matches): Update.

3 years agoLet expand_symtabs_matching short-circuit
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Let expand_symtabs_matching short-circuit

This changes expand_symtabs_exp_notify_ftype to return bool, and
updates all the uses.  Now, if the notification function returns
false, the call is short-circuited and stops examining symtabs.  This
is a step toward replacing map_symtabs_matching_filename with
expand_symtabs_matching.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* symtab.c (default_collect_symbol_completion_matches_break_on):
Update.
* symfile.h (expand_symtabs_matching): Return bool.
* symfile.c (expand_symtabs_matching): Return bool.
* symfile-debug.c (objfile::expand_symtabs_matching): Return
bool.
* quick-symbol.h (expand_symtabs_exp_notify_ftype): Return bool.
(struct quick_symbol_functions) <expand_symtabs_matching>: Return
bool.
* psymtab.c (psymbol_functions::expand_symtabs_matching): Return
bool.
* psympriv.h (struct psymbol_functions)
<expand_symtabs_matching>: Return bool.
* objfiles.h (struct objfile) <expand_symtabs_matching>: Return
bool.
* dwarf2/read.c (struct dwarf2_gdb_index)
<expand_symtabs_matching>: Return bool.
(struct dwarf2_debug_names_index) <expand_symtabs_matching>:
Return bool.
(dw2_expand_symtabs_matching_symbol): Return bool.
(dw2_expand_symtabs_matching_one, dw2_expand_marked_cus)
(dw2_expand_symtabs_matching)
(dwarf2_gdb_index::expand_symtabs_matching)
(dwarf2_debug_names_index::expand_symtabs_matching)
(dwarf2_debug_names_index::expand_symtabs_matching): Return bool.

3 years agoAdd block_search_flags
Tom Tromey [Sat, 17 Apr 2021 15:35:04 +0000 (09:35 -0600)]
Add block_search_flags

This adds block_search_flags, a flag enum.  This will be used to by
certain search functions so that the caller can control which blocks
are searched more precisely.

gdb/ChangeLog
2021-04-17  Tom Tromey  <tom@tromey.com>

* quick-symbol.h (enum block_search_flag_values): New.
(block_search_flags): New enum flags type.