Andrew Burgess [Tue, 23 Jun 2020 13:45:38 +0000 (14:45 +0100)]
gdb/python/mi: create MI commands using python
This commit allows a user to create custom MI commands using Python
similarly to what is possible for Python CLI commands.
A new subclass of mi_command is defined for Python MI commands,
mi_command_py. A new file, gdb/python/py-micmd.c contains the logic
for Python MI commands.
This commit is based on work linked too from this mailing list thread:
https://sourceware.org/pipermail/gdb/2021-November/049774.html
Which has also been previously posted to the mailing list here:
https://sourceware.org/pipermail/gdb-patches/2019-May/158010.html
And was recently reposted here:
https://sourceware.org/pipermail/gdb-patches/2022-January/185190.html
The version in this patch takes some core code from the previously
posted patches, but also has some significant differences, especially
after the feedback given here:
https://sourceware.org/pipermail/gdb-patches/2022-February/185767.html
A new MI command can be implemented in Python like this:
class echo_args(gdb.MICommand):
def invoke(self, args):
return { 'args': args }
echo_args("-echo-args")
The 'args' parameter (to the invoke method) is a list
containing (almost) all command line arguments passed to the MI
command (--thread and --frame are handled before the Python code is
called, and removed from the args list). This list can be empty if
the MI command was passed no arguments.
When used within gdb the above command produced output like this:
(gdb)
-echo-args a b c
^done,args=["a","b","c"]
(gdb)
The 'invoke' method of the new command must return a dictionary. The
keys of this dictionary are then used as the field names in the mi
command output (e.g. 'args' in the above).
The values of the result returned by invoke can be dictionaries,
lists, iterators, or an object that can be converted to a string.
These are processed recursively to create the mi output. And so, this
is valid:
class new_command(gdb.MICommand):
def invoke(self,args):
return { 'result_one': { 'abc': 123, 'def': 'Hello' },
'result_two': [ { 'a': 1, 'b': 2 },
{ 'c': 3, 'd': 4 } ] }
Which produces output like:
(gdb)
-new-command
^done,result_one={abc="123",def="Hello"},result_two=[{a="1",b="2"},{c="3",d="4"}]
(gdb)
I have required that the fields names used in mi result output must
match the regexp: "^[a-zA-Z][-_a-zA-Z0-9]*$" (without the quotes).
This restriction was never written down anywhere before, but seems
sensible to me, and we can always loosen this rule later if it proves
to be a problem. Much harder to try and add a restriction later, once
people are already using the API.
What follows are some details about how this implementation differs
from the original patch that was posted to the mailing list.
In this patch, I have changed how the lifetime of the Python
gdb.MICommand objects is managed. In the original patch, these object
were kept alive by an owned reference within the mi_command_py object.
As such, the Python object would not be deleted until the
mi_command_py object itself was deleted.
This caused a problem, the mi_command_py were held in the global mi
command table (in mi/mi-cmds.c), which, as a global, was not cleared
until program shutdown. By this point the Python interpreter has
already been shutdown. Attempting to delete the mi_command_py object
at this point was causing GDB to try and invoke Python code after
finalising the Python interpreter, and we would crash.
To work around this problem, the original patch added code in
python/python.c that would search the mi command table, and delete the
mi_command_py objects before the Python environment was finalised.
In contrast, in this patch, I have added a new global dictionary to
the gdb module, gdb._mi_commands. We already have several such global
data stores related to pretty printers, and frame unwinders.
The MICommand objects are placed into the new gdb.mi_commands
dictionary, and it is this reference that keeps the objects alive.
When GDB's Python interpreter is shut down gdb._mi_commands is deleted,
and any MICommand objects within it are deleted at this point.
This change avoids having to make the mi_cmd_table global, and walk
over it from within GDB's python related code.
This patch handles command redefinition entirely within GDB's python
code, though this does impose one small restriction which is not
present in the original code (detailed below), I don't think this is a
big issue. However, the original patch relied on being able to
finish executing the mi_command::do_invoke member function after the
mi_command object had been deleted. Though continuing to execute a
member function after an object is deleted is well defined, it is
also (IMHO) risky, its too easy for someone to later add a use of the
object without realising that the object might sometimes, have been
deleted. The new patch avoids this issue.
The one restriction that is added to avoid this, is that an MICommand
object can't be reinitialised with a different command name, so:
(gdb) python cmd = MyMICommand("-abc")
(gdb) python cmd.__init__("-def")
can't reinitialize object with a different command name
This feels like a pretty weird edge case, and I'm happy to live with
this restriction.
I have also changed how the memory is managed for the command name.
In the most recently posted patch series, the command name is moved
into a subclass of mi_command, the python mi_command_py, which
inherits from mi_command is then free to use a smart pointer to manage
the memory for the name.
In this patch, I leave the mi_command class unchanged, and instead
hold the memory for the name within the Python object, as the lifetime
of the Python object always exceeds the c++ object stored in the
mi_cmd_table. This adds a little more complexity in py-micmd.c, but
leaves the mi_command class nice and simple.
Next, this patch adds some extra functionality, there's a
MICommand.name read-only attribute containing the name of the command,
and a read-write MICommand.installed attribute that can be used to
install (make the command available for use) and uninstall (remove the
command from the mi_cmd_table so it can't be used) the command. This
attribute will be automatically updated if a second command replaces
an earlier command.
This patch adds additional error handling, and makes more use the
gdbpy_handle_exception function.
Co-Authored-By: Jan Vrany <jan.vrany@labware.com>
Andrew Burgess [Thu, 10 Mar 2022 11:18:18 +0000 (11:18 +0000)]
gdb/gdbarch: compare some fields against 0 verify_gdbarch
After the previous commit, which removes the predicate function
gdbarch_register_type_p, I noticed that the gdbarch->register_type
field was not checked at in the verify_gdbarch function.
More than not being checked, the field wasn't mentioned at all.
I find this strange, I would expect that every field would at least be
mentioned - we already generate comments for some fields saying that
this field is _not_ being checked, so the fact that this field isn't
being checked looks (to me), like this field is somehow slipping
through the cracks.
The comment at the top of gdbarch-components.py tries to explain how
the validation is done. I didn't understand this comment completely,
but, I think this final sentence:
"Otherwise, the check is done against 0 (really NULL for function
pointers, but same idea)."
Means that, if non of the other cases apply, then the field should be
checked against 0, with 0 indicating that the field is invalid (was
not set by the tdep code). However, this is clearly not being done.
Looking in gdbarch.py at the code to generate verify_gdbarch we do
find that there is a case that is not handled, the case where the
'invalid' field is set true True, but non of the other cases apply.
In this commit I propose two changes:
1. Handle the case where the 'invalid' field of a property is set to
True, this should perform a check for the field of gdbarch still
being set to 0, and
2. If the if/else series that generates verify_gdbarch doesn't handle
a property then we should raise an exception. This means that if a
property is added which isn't handled, we should no longer silently
ignore it.
After doing this, I re-generated the gdbarch files and saw that the
following gdbarch fields now had new validation checks:
register_type
believe_pcc_promotion
register_to_value
value_to_register
frame_red_zone_size
displaced_step_restore_all_in_ptid
solib_symbols_extension
Looking at how these are currently set in the various -tdep.c files, I
believe the only one of these that is required to be set for all
architectures is the register_type field.
And so, for all of the other fields, I've changed the property
definition on gdbarch-components.py, setting the 'invalid' field to
False.
Now, after re-generation, the register_type field is checked against
0, thus an architecture that doesn't set_gdbarch_register_type will
now fail during validation. For all the other fields we skip the
validation, in which case, it is find for an architecture to not set
this field.
My expectation is that there should be no user visible changes after
this commit. Certainly for all fields except register_type, all I've
really done is cause some extra comments to be generated, so I think
that's clearly fine.
For the register_type field, my claim is that any architecture that
didn't provide this would fail when creating its register cache, and I
couldn't spot an architecture that doesn't provide this hook. As
such, I think this change should be fine too.
Andrew Burgess [Thu, 10 Mar 2022 10:57:18 +0000 (10:57 +0000)]
gdb/gdbarch: remove the predicate function for gdbarch_register_type
I don't believe that the gdbarch_register_type_p predicate is called
anywhere in GDB, and the gdbarch_register_type function is called
without checking the gdbarch_register_type_p predicate function
everywhere it is used, for example in
init_regcache_descr (regcache.c).
My claim is that the gdbarch_register_type function is required for
every architecture, and GDB will not work if this function is not
supplied.
And so, in this commit, I remove the 'predicate=True' from
gdbarch-components.py for the 'register_type' field, and regenerate
the gdbarch files.
There should be no user visible changes after this commit.
Patrick Monnerat [Sat, 12 Mar 2022 12:41:47 +0000 (13:41 +0100)]
Replace deprecated_target_wait_hook by observers
Commit
b60cea7 (Make target_wait options use enum flags) broke
deprecated_target_wait_hook usage: there's a commit comment telling
this hook has not been converted.
Rather than trying to mend it, this patch replaces the hook by two
target_wait observers:
target_pre_wait (ptid_t ptid)
target_post_wait (ptid_t event_ptid)
Upon target_wait entry, target_pre_wait is notified with the ptid
passed to target_wait. Upon exit, target_post_wait is notified with
the event ptid returned by target_wait. Should an exception occur,
event_ptid is null_ptid.
This change benefits to Insight (out-of-tree): there's no real use of the
late hook in gdb itself.
Tom Tromey [Thu, 10 Mar 2022 16:23:45 +0000 (09:23 -0700)]
Correctly print subrange types in generic_value_print
I noticed that generic_value_print assumes that a subrange type is
always a subrange of an integer type. However, this isn't necessarily
the case. In Ada, for example, one has subranges of character and
enumeration types.
This code isn't often exercised, I think, because languages with real
subrange types tend to implement their own printers. However, it
still seemed worth fixing.
Luis Machado [Tue, 4 Jan 2022 17:06:36 +0000 (14:06 -0300)]
[aarch64/arm] Properly extract the return value returned in memory
When running gdb.cp/non-trivial-retval.exp, the following shows up for
both aarch64-linux and armhf-linux:
Breakpoint 3, f1 (i1=23, i2=100) at src/gdb/testsuite/gdb.cp/non-trivial-retval.cc:35
35 A a;
(gdb) finish
Run till exit from #0 f1 (i1=23, i2=100) at src/gdb/testsuite/gdb.cp/non-trivial-retval.cc:35
main () at /src/gdb/testsuite/gdb.cp/non-trivial-retval.cc:163
163 B b = f2 (i1, i2);
Value returned is $6 = {a = -11952}
(gdb)
The return value should be {a = 123} instead. This happens because the
backends don't extract the return value from the correct location. GDB should
fetch a pointer to the memory location from X8 for aarch64 and r0 for armhf.
With the patch, gdb.cp/non-trivial-retval.exp has full passes on
aarch64-linux and armhf-linux on Ubuntu 20.04/18.04.
The problem only shows up with the "finish" command. The "call" command
works correctly and displays the correct return value.
This is also related to PR gdb/28681
(https://sourceware.org/bugzilla/show_bug.cgi?id=28681) and fixes FAIL's in
gdb.ada/mi_var_array.exp.
A new testcase is provided, and it exercises GDB's ability to "finish" a
function that returns a large struct (> 16 bytes) and display the
contents of this struct correctly. This has always been incorrect for
these backends, but no testcase exercised this particular scenario.
GDB Administrator [Mon, 14 Mar 2022 00:00:15 +0000 (00:00 +0000)]
Automatic date update in version.in
Alan Modra [Sun, 13 Mar 2022 12:24:25 +0000 (22:54 +1030)]
PR28959, obdump doesn't disassemble mftb instruction
Without a -M cpu option given, powerpc objdump defaults currently to
-Mpower10 but -Many is also given. Commit
1ff6a3b8e562 regressed
-Many disassembly of instructions that are encoded differently
depending on cpu, such as mftb which has pre- and post-power4
encodings.
PR 28959
* ppc-dis.c (lookup_powerpc): Revert 2021-05-28 change. Instead
only look at deprecated PPC_OPCODE_RAW bit when -Many.
GDB Administrator [Sun, 13 Mar 2022 00:00:23 +0000 (00:00 +0000)]
Automatic date update in version.in
Tom Tromey [Sat, 12 Mar 2022 21:36:43 +0000 (14:36 -0700)]
Relax regexp in gdb.rust/unsized.exp
With nightly rustc, gdb.rust/unsized.exp fails:
(gdb) ptype *us
Structure has no component named operator*.
rustc changed to emit a bit more debug info for unsized types.
Because the original test is just to make sure that ptype of an
unsized array looks right, this patch relaxes the regexp and changes
the expression. I think this keeps the original test meaning, but
also works with nightly. I also tested stable and 1.48.
GDB Administrator [Sat, 12 Mar 2022 00:00:14 +0000 (00:00 +0000)]
Automatic date update in version.in
Tom Tromey [Mon, 13 Dec 2021 21:43:18 +0000 (14:43 -0700)]
Avoid crash with cross-linux core file
An internal test case creates a core file using gcore, then restarts
gdb with that core. When run with a cross-linux gdb (in this case,
x86-64 host with ppc64-linux target), the test fails:
| (gdb) core core
| [New LWP 18437]
| warning: `/lib64/libc.so.6': Shared library architecture unknown is not compatible with target architecture powerpc:common64.
| warning: Could not load shared library symbols for /lib64/ld64.so.1.
| Do you need "set solib-search-path" or "set sysroot"?
| ../../src/gdb/gdbarch.c:3388: internal-error: int gdbarch_elf_make_msymbol_special_p(gdbarch*): Assertion `gdbarch != NULL' failed.
| A problem internal to GDB has been detected,
| further debugging may prove unreliable.
| Quit this debugging session? (y or n) y
What's happening here is that the core file lists some shared
libraries. These aren't available via the solib search path, and so
gdb finds the local (x86-64) libraries. This is not ideal, but on the
other hand, it is what was asked for -- while the test does set
solib-search-path, it does not set the sysroot.
But, because gdb isn't configured to handle these libraries, it
crashes.
It seems to me that it's better to avoid the crash by having
solib_bfd_open fail in the case where a library is incompatible. That
is what this patch does. Now it looks like:
| [New LWP 15488]
| Error while mapping shared library sections:
| `/lib64/libc.so.6': Shared library architecture unknown is not compatible with target architecture powerpc:common64.
... and does not crash gdb.
I don't have a good setup for testing this using dejagnu, so I don't
know whether an existing gdb test covers this scenario.
Andrew Burgess [Thu, 10 Mar 2022 19:38:03 +0000 (19:38 +0000)]
gdb/testsuite: remove duplicates from gdb.base/stap-probe.exp
Remove the duplicate test names from gdb.base/stap-probe.exp, this is
done by actually passing a unique test name in a couple of
places (rather than using the command as the test name), and in
another couple of places, a test has a duplicate name due to a cut &
paste error, which I've fixed.
There's no change in what is actually being tested after this commit.
Vladimir Mezentsev [Fri, 11 Mar 2022 08:58:31 +0000 (08:58 +0000)]
gprofng: a new GNU profiler
top-level
* Makefile.def: Add gprofng module.
* configure.ac: Add --enable-gprofng option.
* src-release.sh: Add gprofng.
* Makefile.in: Regenerate.
* configure: Regenerate.
* gprofng: New directory.
binutils
* MAINTAINERS: Add gprofng maintainer.
* README-how-to-make-a-release: Add gprofng.
include.
* collectorAPI.h: New file.
* libcollector.h: New file.
* libfcollector.h: New file.
GDB Administrator [Fri, 11 Mar 2022 00:00:26 +0000 (00:00 +0000)]
Automatic date update in version.in
Aaron Merey [Thu, 10 Mar 2022 20:31:33 +0000 (15:31 -0500)]
gdb/auto-load: Remove repeating "auto-load" from debug message
Remove "auto-load:" from a format string passed to auto_load_debug_printf.
It is unnecessary since this function will prefix the string with "[auto-load]"
when printing it.
Tom Tromey [Thu, 17 Feb 2022 20:43:59 +0000 (13:43 -0700)]
Change how "print/x" displays floating-point value
Currently, "print/x" will display a floating-point value by first
casting it to an integer type. This yields weird results like:
(gdb) print/x 1.5
$1 = 0x1
This has confused users multiple times -- see PR gdb/16242, where
there are several dups. I've also seen some confusion from this
internally at AdaCore.
The manual says:
'x'
Regard the bits of the value as an integer, and print the integer
in hexadecimal.
... which seems more useful. So, perhaps what happened is that this
was incorrectly implemented (or maybe correctly implemented and then
regressed, as there don't seem to be any tests).
This patch fixes the bug.
There was a previous discussion where we agreed to preserve the old
behavior:
https://sourceware.org/legacy-ml/gdb-patches/2017-06/msg00314.html
However, I think it makes more sense to follow the manual.
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=16242
Tom Tromey [Fri, 4 Mar 2022 18:40:49 +0000 (11:40 -0700)]
Simplify the ui-out progress API
I noticed that 'progress' is a method on ui-out, but it seems to me
that it would be better if the only API were via the progress_meter
class. This patch makes this change, changing progress to be a method
on the meter itself.
Andrew Burgess [Thu, 10 Mar 2022 10:57:54 +0000 (10:57 +0000)]
gdb/gdbarch: fix typo in gdbarch-components.py
Fixes a minor typo, in a comment, in the gdbarch-components.py script.
There should be no user visible changes after this commit.
Lancelot SIX [Fri, 18 Feb 2022 15:05:51 +0000 (10:05 -0500)]
Process exit status is leader exit status testcase
This adds a multi-threaded testcase that has all threads in the
process exit with a different exit code, and ensures that GDB reports
the thread group leader's exit status as the whole-process exit
status. Before this set of patches, this would randomly report the
exit code of some other thread, and thus fail.
Tested on Linux-x86_64, native and gdbserver.
Co-Authored-By: Pedro Alves <pedro@palves.net>
Change-Id: I30cba2ff4576fb01b5169cc72667f3268d919557
Pedro Alves [Tue, 22 Feb 2022 10:15:07 +0000 (10:15 +0000)]
Re-add zombie leader on exit, gdbserver/linux
Same as the previous patch, but for GDBserver.
In summary, the current zombie leader detection code in linux-low.cc
has a race -- if a multi-threaded inferior exits just before
check_zombie_leaders finds that the leader is now zombie via checking
/proc/PID/status, check_zombie_leaders deletes the leader, assuming we
won't get an event for that exit (which we won't in some scenarios,
but not in this one), which is a false-positive scenario, where the
whole process is simply exiting. Later when we see the last LWP in
our list exit, we report that LWP's exit status as exit code, even
though for the (real) parent process, the exit code that counts is the
child's leader thread's exit code.
Like for GDB, the solution here is to:
- only report whole-process exit events for the leader.
- re-add the leader back to the LWP list when we finally see it
exit.
Change-Id: Id2d7bbb51a415534e1294fff1d555b7ecaa87f1d
Pedro Alves [Mon, 21 Feb 2022 20:07:20 +0000 (20:07 +0000)]
Re-add zombie leader on exit, gdb/linux
The current zombie leader detection code in linux-nat.c has a race --
if a multi-threaded inferior exits just before check_zombie_leaders
finds that the leader is now zombie via checking /proc/PID/status,
check_zombie_leaders deletes the leader, assuming we won't get an
event for that exit (which we won't in some scenarios, but not in this
one). That might seem mostly harmless, but it has some downsides:
- later when we continue pulling events out of the kernel, we will
collect the exit event of the non-leader threads, and once we see
the last lwp in our list exit, we return _that_ lwp's exit code as
whole-process exit code to infrun, instead of the leader's exit
code.
- this can cause a hang in stop_all_threads in infrun.c. Say there
are 2 threads in the process. stop_all_threads stops each of those
threads, and then waits for two stop or exit events, one for each
thread. If the whole process exits, and check_zombie_leaders hits
the false-positive case, linux-nat.c will only return one event to
GDB (the whole-process exit returned when we see the last thread,
the non-leader thread, exit), making stop_all_threads hang forever
waiting for a second event that will never come.
However, in this false-positive scenario, where the whole process is
exiting, as opposed to just the leader (with pthread_exit(), for
example), we _will_ get an exit event shortly for the leader, after we
collect the exit event of all the other non-leader threads. Or put
another way, we _always_ get an event for the leader after we see it
become zombie.
I tried a number of approaches to fix this:
#1 - My first thought to address the race was to make GDB always
report the whole-process exit status for the leader thread, not for
whatever is the last lwp in the list. We _always_ get a final exit
(or exec) event for the leader, and when the race triggers, we're not
collecting it.
#2 - My second thought was to try to plug the race in the first place.
I thought of making GDB call waitpid/WNOHANG for all non-leader
threads immediately when the zombie leader is detected, assuming there
would be an exit event pending for each of them waiting to be
collected. Turns out that that doesn't work -- you can see the leader
become zombie _before_ the kernel kills all other threads. Waitpid in
that small time window returns 0, indicating no-event. Thankfully we
hit that race window all the time, which avoided trading one race for
another. Looking at the non-leader thread's status in /proc doesn't
help either, the threads are still in running state for a bit, for the
same reason.
#3 - My next attempt, which seemed promising, was to synchronously
stop and wait for the stop for each of the non-leader threads. For
the scenario in question, this will collect all the exit statuses of
the non-leader threads. Then, if we are left with only the zombie
leader in the lwp list, it means we either have a normal while-process
exit or an exec, in which case we should not delete the leader. If
_only_ the leader exited, like in gdb.threads/leader-exit.exp, then
after pausing threads, we will still have at least one live non-leader
thread in the list, and so we delete the leader lwp. I got this
working and polished, and it was only after staring at the kernel code
to convince myself that this would really work (and it would, for the
scenario I considered), that I realized I had failed to account for
one scenario -- if any non-leader thread is _already_ stopped when
some thread triggers a group exit, like e.g., if you have some threads
stopped and then resume just one thread with scheduler-locking or
non-stop, and that thread exits the process. I also played with
PTRACE_EVENT_EXIT, see if it would help in any way to plug the race,
and I couldn't find a way that it would result in any practical
difference compared to looking at /proc/PID/status, with respect to
having a race.
So I concluded that there's no way to plug the race, we just have to
deal with it. Which means, going back to approach #1. That is the
approach taken by this patch.
Change-Id: I6309fd4727da8c67951f9cea557724b77e8ee979
Pedro Alves [Thu, 24 Feb 2022 11:35:43 +0000 (11:35 +0000)]
gdbserver: Reindent check_zombie_leaders
This fixes the indentation of
linux_process_target::check_zombie_leaders, which will help with
keeping its comments in sync with the gdb/linux-nat.c counterpart.
Change-Id: I37332343bd80423d934249e3de2d04feefad1891
Pedro Alves [Tue, 22 Feb 2022 10:15:07 +0000 (10:15 +0000)]
gdbserver: Reorganize linux_process_target::filter_event
Reorganize linux-low.cc:linux_process_target::filter_event such that
all the handling for events for LWPs not in the LWP list is together.
This helps make a following patch clearer. The comments and debug
messages have also been tweaked to have them synchronized with the GDB
counterpart.
Change-Id: If9019635f63a846607cfda44b454b4254a404019
Pedro Alves [Mon, 21 Feb 2022 20:07:20 +0000 (20:07 +0000)]
gdb: Reorganize linux_nat_filter_event
Reorganize linux-nat.c:linux_nat_filter_event such that all the
handling for events for LWPs not in the LWP list is together. This
helps make a following patch clearer. The comments and debug messages
have also been tweaked - the end goal is to have them synchronized
with the gdbserver counterpart.
Change-Id: I8586d8dcd76d8bd3795145e3056fc660e3b2cd22
Pedro Alves [Wed, 23 Feb 2022 11:17:26 +0000 (11:17 +0000)]
Fix gdb.threads/current-lwp-dead.exp race
If we make GDB report the process EXIT event for the leader thread, as
will be done in a latter patch of this series, then
gdb.threads/current-lwp-dead.exp starts failing:
(gdb) break fn_return
Breakpoint 2 at 0x5555555551b5: file /home/pedro/rocm/gdb/build/gdb/testsuite/../../../src/gdb/testsuite/gdb.threads/current-lwp-dead.c, line 45.
(gdb) continue
Continuing.
[New LWP
2138466]
[Inferior 1 (process
2138459) exited normally]
(gdb) FAIL: gdb.threads/current-lwp-dead.exp: continue to breakpoint: fn_return (the program exited)
The inferior exit reported is actually correct. The main thread has
indeed exited, and that's the thread that has the right exit code to
report to the user, as that's the exit code that is reported to the
program's parent. In this case, GDB managed to collect the exit code
for the leader thread before reaping the other thread, because in
reality, the testcase isn't creating standard threads, it is using raw
clone, and the new clones are put in their own thread group.
Fix it by making the main "thread" not exit until the scenario we're
exercising plays out. Also, run the program to completion for
completeness.
The original program really wanted the leader thread to exit before
the fn_return function was reached -- it was important that the
current thread as pointed by inferior_ptid was gone when infrun got
the breakpoint event. I've tweaked the testcase to ensure that that
condition is still held, though it is no longer the main thread that
exits. This required a bit of synchronization between the threads,
which required using CLONE_VM unconditionally. The #ifdef guards were
added as a fix for
https://sourceware.org/bugzilla/show_bug.cgi?id=11214, though I don't
think they were necessary because the program is not using TLS. If it
turns out they were necessary, we can link the testcase with "-z now"
instead, which was mentioned as an alternative workaround in that
Bugzilla.
Change-Id: I7be2f0da4c2fe8f80a60bdde5e6c623d8bd5a0aa
Pedro Alves [Wed, 23 Feb 2022 11:17:26 +0000 (11:17 +0000)]
Fix gdb.threads/clone-new-thread-event.exp race
If we make GDB report the process EXIT event for the leader thread,
instead of whatever is the last thread in the LWP list, as will be
done in a latter patch of this series, then
gdb.threads/current-lwp-dead.exp starts failing:
(gdb) FAIL: gdb.threads/clone-new-thread-event.exp: catch SIGUSR1 (the program exited)
This is a testcase race -- the main thread does not wait for the
spawned clone "thread" to finish before exiting, so the main program
may exit before the second thread is scheduled and reports its
SIGUSR1. With the change to make GDB report the EXIT for the leader,
the race is 100% reproducible by adding a sleep(), like so:
--- c/gdb/testsuite/gdb.threads/clone-new-thread-event.c
+++ w/gdb/testsuite/gdb.threads/clone-new-thread-event.c
@@ -51,6 +51,7 @@ local_gettid (void)
static int
fn (void *unused)
{
+ sleep (1);
tkill (local_gettid (), SIGUSR1);
return 0;
}
Resulting in:
Breakpoint 1, main (argc=1, argv=0x7fffffffd418) at gdb.threads/clone-new-thread-event.c:65
65 stack = malloc (STACK_SIZE);
(gdb) continue
Continuing.
[New LWP
3715562]
[Inferior 1 (process
3715555) exited normally]
(gdb) FAIL: gdb.threads/clone-new-thread-event.exp: catch SIGUSR1 (the program exited)
That inferior exit reported is actually correct. The main thread has
indeed exited, and that's the thread that has the right exit code to
report to the user, as that's the exit code that is reported to the
program's parent. In this case, GDB managed to collect the exit code
for the leader thread before reaping the other thread, because in
reality, the testcase isn't creating standard threads, it is using raw
clone, and the new clones are put in their own thread group.
Fix it by making the main thread wait for the child to exit. Also,
run the program to completion for completeness.
Change-Id: I315cd3dc2b9e860395dcab9658341ea868d7a6bf
Pedro Alves [Fri, 18 Feb 2022 16:42:06 +0000 (16:42 +0000)]
Fix gdbserver/linux target_waitstatus logging assert
Turning on debug output in gdbserver leads to an assertion failure if
gdbserver reports a non-signal event:
[threads] wait_1: LWP
3273770: extended event with waitstatus status->kind = EXECD, execd_pathname = gdb.threads/non-ldr-exc-1/non-ldr-exc-1
[threads] wait_1: Hit a non-gdbserver trap event.
../../src/gdbserver/../gdb/target/waitstatus.h:365: A problem internal to GDBserver has been detected.
sig: Assertion `m_kind == TARGET_WAITKIND_STOPPED || m_kind == TARGET_WAITKIND_SIGNALLED' failed.
Fix it in the obvious way, using target_waitstatus::to_string(),
resulting in, for example:
[threads] wait_1: ret = LWP
1542412.
1542412, status->kind = STOPPED, sig = GDB_SIGNAL_TRAP
Change-Id: Ia4832f9b4fa39f4af67fcaf21fd4d909a285a645
Nick Clifton [Thu, 10 Mar 2022 09:11:40 +0000 (09:11 +0000)]
Add option to objdump/readelf to disable access to debuginfod servers.
* dwarf.c (use_debuginfod): New variable. Set to 1.
(load_separate_debug_info): Only call
debuginfod_fetch_separate_debug_info is use_debuginfod is true.
(dwarf_select_sections_by_names): Add do-not-use-debuginfod and
use-debuginfod options.
(dwarf_select_sections_by_letters): Add D and E options.
* dwarf.h (use_debuginfod): New extern.
* objdump.c (usage): Mention the new options.
* readelf.c (usage): Likewise.
* doc/binutils.texi: Document the new options.
* doc/debug-options.texi: Describe the new options.
* NEWS: Mention the new feature.
* testsuite/binutils-all/debuginfod.exp: Add tests of the new
options.
Alan Modra [Thu, 10 Mar 2022 05:36:12 +0000 (16:06 +1030)]
Re: ld: Add a before_plugin_all_symbols_read hook
* testsuite/ld-plugin/pr28849.d: Adjust for powerpc64 function
descriptors.
H.J. Lu [Wed, 2 Feb 2022 22:40:03 +0000 (14:40 -0800)]
ld: Add a before_plugin_all_symbols_read hook
Add a before_plugin_all_symbols_read hook to load symbol references from
DT_NEEDED entries, included from --copy-dt-needed-entries, before reading
plugin symbols to properly resolve plugin symbol references.
bfd/
PR ld/28849
* elf-bfd.h (elf_link_hash_table): Add handling_dt_needed.
* elflink.c (_bfd_elf_merge_symbol): Don't set non_ir_ref_dynamic
before plugin 'all symbols read' hook is called.
ld/
PR ld/28849
* ldelf.c (ldelf_handle_dt_needed): New function.
(ldelf_before_plugin_all_symbols_read): Likewise.
(ldelf_after_open): Call ldelf_handle_dt_needed.
* ldelf.h (ldelf_before_plugin_all_symbols_read): New.
* ldemul.c (ldemul_before_plugin_all_symbols_read): Likewise.
* ldemul.h (ldemul_before_plugin_all_symbols_read): Likewise.
(ld_emulation_xfer_struct): Add before_plugin_all_symbols_read.
* ldlang.c (lang_process): Call
ldemul_before_plugin_all_symbols_read before calling
plugin_call_all_symbols_read.
* emultempl/elf.em
(gld${EMULATION_NAME}_before_plugin_all_symbols_read): New.
(LDEMUL_BEFORE_PLUGIN_ALL_SYMBOLS_READ): New.
* emultempl/emulation.em (ld_${EMULATION_NAME}_emulation):
Initialize the before_plugin_all_symbols_read field.
* testsuite/ld-plugin/lto.exp: Run PR ld/28849 tests.
* testsuite/ld-plugin/pr28849.d: New file.
* testsuite/ld-plugin/pr28849a.c: Likewise.
* testsuite/ld-plugin/pr28849b.c: Likewise.
GDB Administrator [Thu, 10 Mar 2022 00:00:14 +0000 (00:00 +0000)]
Automatic date update in version.in
Hans-Peter Nilsson [Wed, 9 Mar 2022 19:46:16 +0000 (20:46 +0100)]
toplevel: Makefile.def: Make configure-sim depend on all-readline
Without this, a "make all-sim" without the equivalent of
libreadline-dev installed on the build system, won't
properly pick up the in-tree readline build, and you'll see:
mkdir -p -- ./sim
Configuring in ./sim
configure: creating cache ./config.cache
checking build system type... x86_64-pc-linux-gnu
checking host system type... x86_64-pc-linux-gnu
checking target system type... cris-axis-elf
checking for x86_64-pc-linux-gnu-gcc... gcc
checking whether the C compiler works... yes
...
checking for library containing tgetent... -ltermcap
checking for readline in -lreadline... no
configure: error: the required "readline" library is missing
make[1]: *** [Makefile:11188: configure-sim] Error 1
make[1]: Leaving directory '/home/hp/sim/b'
The sim dependency on readline is apparently (nominally)
valid as there's a readline call in sim/erc32/sis.c.
2022-02-21 Hans-Peter Nilsson <hp@axis.com>
* Makefile.def (dependencies): Make configure-sim depend on
all-readline.
Maciej W. Rozycki [Wed, 9 Mar 2022 19:15:27 +0000 (19:15 +0000)]
GDB/testsuite: Fix a "displayed" typo in gdb.base/default.exp
Fix a typo, s/dislayed/displayed/ in default.exp at the top level.
Maciej W. Rozycki [Wed, 9 Mar 2022 19:15:27 +0000 (19:15 +0000)]
GDB/testsuite: Remove a stray backslash from gdb.base/settings.exp
Remove a stray trailing backslash from `test-integer' in settings.exp.
It is harmless as only white space follows in the next line before the
closing brace, so it merely swallows the newline character, but it may
look confusing to the reader.
Christina Schimpe [Wed, 9 Mar 2022 08:09:57 +0000 (09:09 +0100)]
* gdb/doc/gdb.texinfo (Requirements): Fix a typo.
Copyright-paperwork-exempt: yes
Alan Modra [Tue, 8 Mar 2022 12:18:51 +0000 (22:48 +1030)]
Constant fold view increment expressions
The idea here is to replace expressions like v + 1 + 1 + 1 with v + 3.
* dwarf2dbg.c (set_or_check_view): Remove useless assertion.
Resolve multiple view increments.
* testsuite/gas/elf/dwarf2-18.d: Don't xfail mep.
Alan Modra [Tue, 8 Mar 2022 12:19:52 +0000 (22:49 +1030)]
Reduce duplicated symbol_clone_if_forward_ref work
* symbol.c (struct symbol_flags): Add forward_resolved.
(symbol_entry_find): Update needle initialisation.
(symbol_clone_if_forward_ref): Do no work when forward_resolved
is already set. Set forward_resolved.
Aaron Merey [Tue, 8 Mar 2022 22:32:35 +0000 (17:32 -0500)]
gdb: Try searching for auto-load script using .gnu_debuglink
If an auto-load script cannot be found and objfile is a separate
debuginfo whose filename does not match the name found in the parent
file's .gnu_debuglink section, then repeat the search using the
parent's filename where the last component is replaced with the
.gnu_debuglink name.
For example if the parent's filename is "/usr/lib/libxyz.so" and the
name in its .gnu_debuglink section is "libxyz.so.debug", then
if no auto-load script is otherwise found the search will be
repeated with the filename "/usr/lib/libxyz.so.debug".
This helps gdb locate auto-load scripts when debuginfo files do not have
the expected filename, such as when they are aquired from debuginfod.
GDB Administrator [Wed, 9 Mar 2022 00:00:15 +0000 (00:00 +0000)]
Automatic date update in version.in
Aaron Merey [Sat, 20 Nov 2021 00:41:40 +0000 (19:41 -0500)]
PR gdb/27876 - debuginfod-downloaded source files don't pass proper fullname across mi / (gdb)info source
Source files downloaded from debuginfod currently use their original DWARF
filename as their "fullname". This causes a mismatch between the fullname
and the actual location of the source file in the debuginfod client cache.
MI consumers such as VSCode will fail to open debuginfod-downloaded
source files due to this. Also 'info source' will fail to include the
true paths of these files.
To fix this, use the debuginfod cache path as the fullname for debuginfod-
downloaded source files.
Jan Vrany [Wed, 2 Mar 2022 13:23:30 +0000 (13:23 +0000)]
gdb/mi: preserve user selected thread and frame when invoking MI commands
Fix for PR gdb/20684. When invoking MI commands with --thread and/or
--frame, the user selected thread and frame was not preserved:
(gdb)
info thread
&"info thread\n"
~" Id Target Id Frame \n"
~"* 1 Thread 0x7ffff7c30740 (LWP 19302) \"user-selected-c\" main () at /home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:60\n"
~" 2 Thread 0x7ffff7c2f700 (LWP 19306) \"user-selected-c\" child_sub_function () at /home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:30\n"
~" 3 Thread 0x7ffff742e700 (LWP 19307) \"user-selected-c\" child_sub_function () at /home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:30\n"
^done
(gdb)
info frame
&"info frame\n"
~"Stack level 0, frame at 0x7fffffffdf90:\n"
~" rip = 0x555555555207 in main (/home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:60); saved rip = 0x7ffff7c5709b\n"
~" source language c.\n"
~" Arglist at 0x7fffffffdf80, args: \n"
~" Locals at 0x7fffffffdf80, Previous frame's sp is 0x7fffffffdf90\n"
~" Saved registers:\n "
~" rbp at 0x7fffffffdf80, rip at 0x7fffffffdf88\n"
^done
(gdb)
-stack-info-depth --thread 3
^done,depth="4"
(gdb)
info thread
&"info thread\n"
~" Id Target Id Frame \n"
~" 1 Thread 0x7ffff7c30740 (LWP 19302) \"user-selected-c\" main () at /home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:60\n"
~" 2 Thread 0x7ffff7c2f700 (LWP 19306) \"user-selected-c\" child_sub_function () at /home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:30\n"
~"* 3 Thread 0x7ffff742e700 (LWP 19307) \"user-selected-c\" child_sub_function () at /home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:30\n"
^done
(gdb)
info frame
&"info frame\n"
~"Stack level 0, frame at 0x7ffff742dee0:\n"
~" rip = 0x555555555169 in child_sub_function (/home/uuu/gdb/gdb/testsuite/gdb.mi/user-selected-context-sync.c:30); saved rip = 0x555555555188\n"
~" called by frame at 0x7ffff742df00\n"
~" source language c.\n"
~" Arglist at 0x7ffff742ded0, args: \n"
~" Locals at 0x7ffff742ded0, Previous frame's sp is 0x7ffff742dee0\n"
~" Saved registers:\n "
~" rbp at 0x7ffff742ded0, rip at 0x7ffff742ded8\n"
^done
(gdb)
This caused problems for frontends that provide access to CLI because UI
may silently change the context for CLI commands (as demonstrated above).
This commit fixes the problem by restoring thread and frame in
mi_cmd_execute (). With this change, there are only two GDB/MI commands
that can change user selected context: -thread-select and -stack-select-frame.
This allows us to remove all and rather complicated logic of notifying
about user selected context change from mi_execute_command (), leaving it
to these two commands themselves to notify.
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=20684
Andrew Burgess [Tue, 8 Mar 2022 13:42:16 +0000 (13:42 +0000)]
gdb: announce upcoming removal of Python 2 support from gdb
As has been discussed here:
https://sourceware.org/pipermail/gdb-patches/2022-January/184910.html
Python 2 support will be removed from GDB after GDB 12 has branched.
This commit places an entry in the NEWS file to inform users of this
decision.
GDB Administrator [Tue, 8 Mar 2022 00:00:25 +0000 (00:00 +0000)]
Automatic date update in version.in
Andrew Burgess [Fri, 25 Feb 2022 11:03:03 +0000 (11:03 +0000)]
gdb/testsuite: add new test for comparing char types in Python
There's an interesting property of the 'char' type in C and C++, the
three types 'char', 'unsigned char', and 'signed char', are all
considered distinct.
In contrast, and 'int' is signed by default, and so 'int' and 'signed
int' are considered the same type.
This commit adds a test to ensure that this edge case is visible to a
user from Python.
It is worth noting that for any particular compiler implementation (or
the flags a compiler was invoked with), a 'char' will be either signed
or unsigned; it has to be one or the other, and a user can access this
information by using the Type.is_signed property. However, for
something like function overload resolution, the 'char' type is
considered distinct from the signed and unsigned variants.
There's no change to GDB with this commit, this is just adding a new
test to guard some existing functionality.
Andrew Burgess [Tue, 30 Nov 2021 14:35:44 +0000 (14:35 +0000)]
gdb/python: add Type.is_signed property
Add a new read-only property, Type.is_signed, which is True for signed
types, and False otherwise.
This property should only be read on types for which Type.is_scalar is
true, attempting to read this property for non-scalar types will raise
a ValueError.
I chose 'is_signed' rather than 'is_unsigned' in order to match the
existing Architecture.integer_type method, which takes a 'signed'
parameter. As far as I could find, that was the only existing
signed/unsigned selector in the Python API, so it seemed reasonable to
stay consistent.
Andrew Burgess [Fri, 25 Feb 2022 10:54:04 +0000 (10:54 +0000)]
gdb/python: add Type.is_scalar property
Add a new read-only property which is True for scalar types,
otherwise, it's False.
Andrew Burgess [Wed, 2 Mar 2022 11:11:47 +0000 (11:11 +0000)]
gdb/mi: add --no-connection to MI -add-inferior command
Following on from the previous commit, where the -add-inferior command
now uses the same connection as the current inferior, this commit adds
a --no-connection option to -add-inferior.
This new option matches the existing option of the same name for the
CLI version of add-inferior; the new inferior is created with no
connection.
I've added a new 'connection' field to the MI output of -add-inferior,
which includes the connection number and short name. I haven't
included the longer description field, this is the MI after all. My
expectation would be that if the frontend wanted to display all the
connection details then this would be looked up from 'info
connection' (or the MI equivalent if/when such a command is added).
The existing -add-inferior tests are updated, as are the docs.
Umair Sair [Thu, 24 Feb 2022 17:25:51 +0000 (22:25 +0500)]
gdb/mi: fix regression in mi -add-inferior command
Prior to the multi-target support commit:
commit
5b6d1e4fa4fc6827c7b3f0e99ff120dfa14d65d2
Date: Fri Jan 10 20:06:08 2020 +0000
Multi-target support
When a new inferior was added using the MI -add-inferior command, the
new inferior would be using the same target as all the other
inferiors. This makes sense, GDB only supported a single target stack
at a time.
After the above commit, each inferior has its own target stack.
To maintain backward compatibility, for the CLI add-inferior command,
when a new inferior is added the above commit has the new inferior
inherit a copy of the target stack from the current inferior.
Unfortunately, this same backward compatibility is missing for the MI.
This commit fixes this oversight.
Now, when the -add-inferior MI command is used, the new inferior will
inherit a copy of the target stack from the current inferior.
Tom Tromey [Thu, 3 Mar 2022 16:47:00 +0000 (09:47 -0700)]
Deprecate dbx mode
GDB has a dbx emulation mode that adds a few aliases and helper
commands. This mode is barely documented and is very superficial
besides. I suspect it is rarely used, and I would like to propose
deprecating it for GDB 12, and then removing it in GDB 13.
Pedro Alves [Mon, 7 Mar 2022 16:09:41 +0000 (16:09 +0000)]
Remove unnecessary inferior lookup in infrun:handle_one
infrun.c:handle_one calls find_inferior_ptid unnecessarily, since we
already have a thread pointer handy, and the thread has a pointer to
the inferior. This commit removes the unnecessary lookup.
Change-Id: I2ae18601dd75346c6c91068e9a4f9a6484fb3339
Tom Tromey [Wed, 16 Feb 2022 19:33:45 +0000 (12:33 -0700)]
Fix bug in ada_print_floating
ada_print_floating rewrites a floating-point string representation to
conform to Ada syntax. However, if you managed to get a floating
point error, you might see:
(gdb) print whatever
$2 = <invalid float valu.0e>
What's happening here is that ada_print_floating doesn't recognize
this error case, and proceeds to modify the error text.
This patch fixes this problem.
Tom Tromey [Wed, 16 Feb 2022 17:07:18 +0000 (10:07 -0700)]
Implement real literal extension for Ada
Sometimes it is convenient to be able to specify the exact bits of a
floating-point literal. For example, you may want to set a
floating-point register to a denormalized value, or to a particular
NaN.
In C, you can do this by combining the "{}" cast with an array
literal, like:
(gdb) p {double}{0x576488BDD2AE9FFE}
$1 = 9.
8765449999999996e+112
This patch adds a somewhat similar idea to Ada. It extends the lexer
to allow "l" and "f" suffixes in a based literal. The "f" indicates a
floating-point literal, and the "l"s control the size of the
floating-point type.
Note that this differs from Ada's based real literals. I believe
those can also be used to control the bits of a floating-point value,
but they are a bit more cumbersome to use (simplest is binary but
that's also very lengthy). Also, these aren't implemented in GDB.
I chose not to allow this extension to work with based integer
literals with exponents. That didn't seem very useful.
Tom Tromey [Wed, 16 Feb 2022 19:01:52 +0000 (12:01 -0700)]
Fix Ada integer literals with exponents
While working on another patch, I noticed that Ada integer literals
with exponents did not work. For example, with one form you get an
error:
(gdb) print 8e2
Invalid digit `e' in based literal
And with another form you get an incorrect value:
(gdb) print 16#8#e2
$2 = 8
This patch fixes the bugs and adds tests.
Tom Tromey [Mon, 28 Feb 2022 20:42:03 +0000 (13:42 -0700)]
Fix gdb.ada/arrayptr.exp results
PR ada/28115 points out that gdb.ada/arrayptr.exp works with GNAT 12,
but fails with minimal encodings in earlier versions.
This patch updates the test to try to report the results correctly. I
tried this with the Fedora 34 system gcc (GCC 11) and with a GCC 12
built from git trunk sometime relatively recently.
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=28115
Tom Tromey [Thu, 3 Feb 2022 17:42:07 +0000 (10:42 -0700)]
Handle non-ASCII identifiers in Ada
Ada allows non-ASCII identifiers, and GNAT supports several such
encodings. This patch adds the corresponding support to gdb.
GNAT encodes non-ASCII characters using special symbol names.
For character sets like Latin-1, where all characters are a single
byte, it uses a "U" followed by the hex for the character. So, for
example, thorn would be encoded as "Ufe" (0xFE being lower case
thorn).
For wider characters, despite what the manual says (it claims
Shift-JIS and EUC can be used), in practice recent versions only
support Unicode. Here, characters in the base plane are represented
using "Wxxxx" and characters outside the base plane using
"WWxxxxxxxx".
GNAT has some further quirks here. Ada is case-insensitive, and GNAT
emits symbols that have been case-folded. For characters in ASCII,
and for all characters in non-Unicode character sets, lower case is
used. For Unicode, however, characters that fit in a single byte are
converted to lower case, but all others are converted to upper case.
Furthermore, there is a bug in GNAT where two symbols that differ only
in the case of "Y WITH DIAERESIS" (and potentially others, I did not
check exhaustively) can be used in one program. I chose to omit
handling this case from gdb, on the theory that it is hard to figure
out the logic, and anyway if the bug is ever fixed, we'll regret
having a heuristic.
This patch introduces a new "ada source-charset" setting. It defaults
to Latin-1, as that is GNAT's default. This setting controls how "U"
characters are decoded -- W/WW are always handled as UTF-32.
The ada_tag_name_from_tsd change is needed because this function will
read memory from the inferior and interpret it -- and this caused an
encoding failure on PPC when running a test that tries to read
uninitialized memory.
This patch implements its own UTF-32-based case folder. This avoids
host platform quirks, and is relatively simple. A short Python
program to generate the case-folding table is included. It simply
relies on whatever version of Unicode is used by the host Python,
which seems basically acceptable.
Test cases for UTF-8, Latin-1, and Latin-3 are included. This
exercises most of the new code paths, aside from Y WITH DIAERESIS as
noted above.
Tom Tromey [Thu, 3 Feb 2022 19:04:36 +0000 (12:04 -0700)]
Define HOST_UTF32 in charset.h
rust-parse.c has a #define for the host-specific UTF-32 charset name.
A later patch needs the same thing, so this patch moves the definition
to charset.h for easier reuse.
Tom Tromey [Thu, 3 Feb 2022 20:18:25 +0000 (13:18 -0700)]
Let phex and phex_nz handle sizeof_l==1
Currently, neither phex nor phex_nz handle sizeof_l==1 -- they let
this case fall through to the default case. However, a subsequent
patch in this series needs this case to work correctly.
I looked at all calls to these functions that pass a 1 for the
sizeof_l parameter. The only such case seems to be correct with this
change.
Tom Tromey [Thu, 3 Feb 2022 19:24:12 +0000 (12:24 -0700)]
Don't pre-size result string in ada_decode
Currently, ada_decode pre-sizes the output string, filling it with 'X'
characters. However, it's a bit simpler and more flexible to let
std::string do the work here, and simply append characters to the
string as we go. This turns out to be useful for a subsequent patch.
Tom Tromey [Thu, 3 Feb 2022 20:12:21 +0000 (13:12 -0700)]
Simplify a regular expression in ada-lex.l
ada-lex.l uses "%option case-insensitive", so there is no need for
regular expressions to match upper case.
GDB Administrator [Mon, 7 Mar 2022 00:00:11 +0000 (00:00 +0000)]
Automatic date update in version.in
Maciej W. Rozycki [Sun, 6 Mar 2022 18:30:58 +0000 (18:30 +0000)]
MIPS/opcodes: Fix alias annotation for branch instructions
Correct issues with INSN2_ALIAS annotation for branch instructions:
- regular MIPS BEQZ/L and BNEZ/L assembly instructions are idioms for
BEQ/L and BNE/L respectively with the `rs' operand equal to $0,
- microMIPS 32-bit BEQZ and BNEZ assembly instructions are idioms for
BEQ and BNE respectively with the `rt' operand equal to $0,
- regular MIPS BAL assembly instruction is an idiom for architecture
levels of up to the MIPSr5 ISA and a machine instruction on its own
from the MIPSr6 ISA up.
Add missing annotation to BEQZ/L and BNEZ/L accordingly then and add a
new entry for BAL for the MIPSr6 ISA, correcting a disassembly bug:
$ mips-linux-gnu-objdump -m mips:isa64r6 -M no-aliases -d bal.o
bal.o: file format elf32-tradlittlemips
Disassembly of section .text:
00000000 <foo>:
0:
04110000 0x4110000
...
$
Add test cases accordingly.
Parts for regular MIPS BEQZ/L and BNEZ/L instructions from Sagar Patel.
2022-03-06 Maciej W. Rozycki <macro@orcam.me.uk>
binutils/
* testsuite/binutils-all/mips/mips1-branch-alias.d: New test.
* testsuite/binutils-all/mips/mips1-branch-noalias.d: New test.
* testsuite/binutils-all/mips/mips2-branch-alias.d: New test.
* testsuite/binutils-all/mips/mips2-branch-noalias.d: New test.
* testsuite/binutils-all/mips/mips32r6-branch-alias.d: New test.
* testsuite/binutils-all/mips/mips32r6-branch-noalias.d: New
test.
* testsuite/binutils-all/mips/micromips-branch-alias.d: New
test.
* testsuite/binutils-all/mips/micromips-branch-noalias.d: New
test.
* testsuite/binutils-all/mips/mips-branch-alias.s: New test
source.
* testsuite/binutils-all/mips/micromips-branch-alias.s: New test
source.
* testsuite/binutils-all/mips/mips.exp: Run the new tests.
2022-03-06 Sagar Patel <sagarmp@cs.unc.edu>
Maciej W. Rozycki <macro@orcam.me.uk>
opcodes/
* mips-opc.c (mips_builtin_opcodes): Fix INSN2_ALIAS annotation
for "bal", "beqz", "beqzl", "bnez" and "bnezl" instructions.
* micromips-opc.c (micromips_opcodes): Likewise for "beqz" and
"bnez" instructions.
Tom Tromey [Fri, 4 Mar 2022 03:25:32 +0000 (20:25 -0700)]
Use function view when iterating over block symbols
This changes iterate_over_block_local_vars and
iterate_over_block_arg_vars to take a gdb::function_view rather than a
function pointer and a user-data. In one spot, this allows us to
remove a helper structure and helper function. In another spot, this
looked more complicated, so I changed the helper function to be an
"operator()" -- also a simplification, just not as big.
Tom Tromey [Sat, 5 Mar 2022 00:14:44 +0000 (17:14 -0700)]
Simplify hppa-tdep.c use of objfile_key
I happened to notice a couple of unnecessary casts in hppa-tdep.c, and
then I saw that the use of objfile_key could be simplified -- removing
some code and using the default deleter rather than noop_deleter.
Tested by rebuilding. Let me know what you think.
Simon Marchi [Mon, 31 Jan 2022 20:57:58 +0000 (15:57 -0500)]
gdb: constify parameter of value_copy
In a following patch, I have a const value I want to copy using a
value_copy. However, value_copy takes a non-const source value, at the
moment. Change the paramter to be const,
If the source value is not lazy, we currently call
value_contents_all_raw, which calls allocate_value_contents, to get a
view on the contents. They both take a non-const value, that's a
problem. My first attempt at solving it was to add a const version of
value_contents_all_raw, make allocate_value_contents take a const value,
and either:
- make value::contents mutable
- make allocate_value_contents cast away the const
The idea being that allocating the value contents buffer does modify the
value at the bit level, but logically that doesn't change its state.
That was getting a bit complicated, so what I ended up doing is make
value_copy not call value_contents_all_raw. We know at this point that
the value is not lazy, so value::contents must have been allocate
already.
Change-Id: I3741ab362bce14315f712ec24064ccc17e3578d4
Simon Marchi [Mon, 31 Jan 2022 20:23:32 +0000 (15:23 -0500)]
gdb: remove internalvar_funcs::destroy
No kind of internal var uses it remove it. This makes the transition to
using a variant easier, since we don't need to think about where this
should be called (in a destructor or not), if it can throw, etc.
Change-Id: Iebbc867d1ce6716480450d9790410d6684cbe4dd
GDB Administrator [Sun, 6 Mar 2022 00:00:25 +0000 (00:00 +0000)]
Automatic date update in version.in
GDB Administrator [Sat, 5 Mar 2022 00:00:25 +0000 (00:00 +0000)]
Automatic date update in version.in
Tom Tromey [Fri, 4 Mar 2022 18:58:27 +0000 (11:58 -0700)]
Mark vDSO as not a file
The vDSO objfile is not a real file, so mark it as such. I noticed
this because, when playing with debuginfod, I saw:
Downloading 0.01 MB separate debug info for /tmp/system-supplied DSO at 0x7ffff7fc9000
That "/tmp" is wrong -- it's just gdb's cwd. This patch corrects the
problem, resulting in:
Downloading 0.01 MB separate debug info for system-supplied DSO at 0x7ffff7fc9000
Regression tested on x86-64 Fedora 34.
Simon Marchi [Fri, 4 Mar 2022 15:57:14 +0000 (10:57 -0500)]
binutils/readelf: fix indentation in process_dynamic_section
Clangd shows a warning about misleading indentation in this file, fix
it.
binutils/ChangeLog:
* readelf.c (process_dynamic_section): Fix indentation.
Change-Id: I43a7f4f4c75dd080af614222b980526f5debf297
Christina Schimpe [Mon, 25 Oct 2021 15:08:32 +0000 (17:08 +0200)]
gdb: Use a typedef's scoped type name to identify local typedefs
GDB prints the wrong type for typedefs in case there is another typedef
available for the same raw type (gdb/16040). The reason is that the
current hashmap based substitution mechanism always compares the target
type of a typedef and not its scoped name.
The original output of GDB for a program like
~~~~
namespace ns
{
typedef double scoped_double;
}
typedef double global_double;
class TypedefHolder
{
public:
double a;
ns::scoped_double b;
global_double c;
private:
typedef double class_double;
class_double d;
double method1(ns::scoped_double) { return 24.0; }
double method2(global_double) { return 24.0; }
};
int main()
{
TypedefHolder th;
return 0;
}
~~~~
is
~~~~
(gdb) b 27
Breakpoint 1 at 0x1131: file TypedefHolder.cc, line 27.
(gdb) r
Starting program: /tmp/typedefholder
Breakpoint 1, main () at TypedefHolder.cc:27
27 return 0;
(gdb) ptype th
type = class TypedefHolder {
public:
class_double a;
class_double b;
class_double c;
private:
class_double d;
class_double method1(class_double);
class_double method2(class_double);
typedef double class_double;
}
~~~~
Basically all attributes of a class which have the raw type "double" are
substituted by "class_double".
With the patch the output is the following
~~~~
type = class TypedefHolder {
public:
double a;
ns::scoped_double b;
global_double c;
private:
class_double d;
double method1(ns::scoped_double);
double method2(global_double);
typedef double class_double;
}
~~~~
Jan Beulich [Fri, 4 Mar 2022 12:37:59 +0000 (13:37 +0100)]
RISC-V: make .insn actually work for 64-bit insns
Presently in this case, due to an undefined behavior shift, at least
with x86 cross builds I'm observing:
Error: value conflicts with instruction length `8,0x0000003f'
Eliminate the UB and extend the respective testcase.
Jan Beulich [Fri, 4 Mar 2022 12:37:30 +0000 (13:37 +0100)]
x86: drop redundant x86-64-code16-2 test
The code16-2 test is already meaningless enough as a gas test, identical
to this one, and is run uniformly for all ELF targets anyway.
GDB Administrator [Fri, 4 Mar 2022 00:00:19 +0000 (00:00 +0000)]
Automatic date update in version.in
Roland McGrath [Thu, 3 Mar 2022 21:06:50 +0000 (13:06 -0800)]
Fix typo in last change.
Roland McGrath [Wed, 2 Mar 2022 00:03:58 +0000 (16:03 -0800)]
Avoid conflict with gnulib open/close macros.
On some systems, the gnulib configuration will decide to define open
and/or close as macros to replace the POSIX C functions. This
interferes with using those names in C++ class or namespace scopes.
gdbsupport/
* event-pipe.cc (event_pipe::open): Renamed to ...
(event_pipe::open_pipe): ... this.
(event_pipe::close): Renamed to ...
(event_pipe::close_pipe): ... this.
* event-pipe.h (class event_pipe): Updated.
gdb/
* inf-ptrace.h (async_file_open, async_file_close): Updated.
gdbserver/
* gdbserver/linux-low.cc (linux_process_target::async): Likewise.
Alan Modra [Fri, 25 Feb 2022 06:55:26 +0000 (17:25 +1030)]
Adjust ld ctf test for 32-bit targets
powerpc-linux, and I suspect other 32-bit targets, report "aligned at
0x4" for this test.
* testsuite/ld-ctf/nonrepresentable.d: Accept any alignment.
Luis Machado [Tue, 1 Mar 2022 13:33:42 +0000 (13:33 +0000)]
Update my e-mail address in the MAINTAINERS file
Update the information accordingly.
Tiezhu Yang [Thu, 3 Mar 2022 03:15:10 +0000 (11:15 +0800)]
gdb: testsuite: fix failed testcases in gdb.base/gdb-caching-proc.exp
When execute the following command:
make check-gdb TESTS="gdb.base/gdb-caching-proc.exp"
we can see there exist some failed testcases:
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 0: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 1: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 2: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 3: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 4: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 5: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 6: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 7: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 8: can spawn for attach (got interactive prompt)
FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 9: can spawn for attach (got interactive prompt)
here are the detailed messages in gdb/testsuite/gdb.log:
attach 873776
A program is being debugged already. Kill it? (y or n) n
Not killed.
(gdb) FAIL: gdb.base/gdb-caching-proc.exp: can_spawn_for_attach: 0: can spawn for attach (got interactive prompt)
so handle the case "A program is being debugged already. Kill it" in
can_spawn_for_attach to fix the failed testcases.
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Alan Modra [Thu, 3 Mar 2022 02:47:15 +0000 (13:17 +1030)]
comment typo fix
Alan Modra [Wed, 2 Mar 2022 13:34:57 +0000 (00:04 +1030)]
PowerPC64 DT_RELR relative reloc addresses
Section addresses can change between ppc64_elf_size_stubs and
ppc64_elf_build_stubs due to .eh_frame editing. The idea of stashing
r_offset final addresses calculated in ppc64_elf_size_stubs for use by
ppc64_elf_build_stubs was never a good idea. Instead, we need to keep
section/offset pairs.
* elf64-ppc.c (struct ppc_link_hash_table): Delete relr_addr.
Add relr section/offset array.
(append_relr_off): Rewrite. Update all callers.
(sort_relr): New function.
(ppc64_elf_size_stubs): Adjust to suit new relative reloc stash.
(ppc64_elf_build_stubs): Likewise.
GDB Administrator [Thu, 3 Mar 2022 00:00:26 +0000 (00:00 +0000)]
Automatic date update in version.in
John Baldwin [Wed, 2 Mar 2022 22:09:55 +0000 (14:09 -0800)]
configure: Stop checking for PT_GETXMMREGS.
This request is present on all modern *BSD/i386 systems (those
released since mid-2006), and the *BSD/i386 targets now assume it is
present unconditionally.
John Baldwin [Wed, 2 Mar 2022 22:09:55 +0000 (14:09 -0800)]
i386-bsd-nat: Assume PT_GETXMMREGS is present.
NetBSD has included PT_GETXMMREGS since 1.6 released in September
2002. OpenBSD has included PT_GETXMMREGS since 3.8 released in
November 2005.
John Baldwin [Wed, 2 Mar 2022 22:09:55 +0000 (14:09 -0800)]
i386-fbsd-nat: Assume PT_GETXMMREGS is present.
PT_GETXMMREGS was first added in FreeBSD 6.0 released in November 2005.
The last FreeBSD release without support was 5.5 released in May 2006.
John Baldwin [Wed, 2 Mar 2022 22:00:36 +0000 (14:00 -0800)]
fbsd-tdep: Implement the vsyscall_range gdbarch hook.
FreeBSD recently added a real vDSO in its shared page for the amd64
architecture. The vDSO is mapped at the address given by the
AT_KPRELOAD ELF auxiliary vector entry. To find the end of the
mapping range, parse the list of virtual map entries used by 'info
proc mappings' either from the NT_PROCSTAT_VMMAP core dump note, or
via the kinfo_getvmmap function for native targets (fetched from the
native target as the TARGET_OBJECT_FREEBSD_VMMAP object).
This silences warnings on recent FreeBSD/amd64 kernels due to not
finding symbols for the vdso:
warning: Could not load shared library symbols for [vdso].
Do you need "set solib-search-path" or "set sysroot"?
Tom Tromey [Mon, 14 Feb 2022 17:19:06 +0000 (10:19 -0700)]
Rewrite make-target-delegates in Python
I think gdb is probably better off having fewer languages involved
when generating code. 'sh' is unavoidable for build-time generation,
but for other things, let's use Python.
This rewrites make-target-delegates in Python. I've stuck pretty
closely to the original code in this rewrite, so it may look slightly
weird from a Python perspective.
The only output difference is that a copyright header is now
generated, using the code introduced in the previous patch.
make-target-delegates.py is simpler to invoke, as it knows the correct
input file to scan and it creates the output file itself.
Tom Tromey [Mon, 14 Feb 2022 15:26:32 +0000 (08:26 -0700)]
Move copyright code from gdbarch.py to new file
This moves the copyright code from gdbarch.py to a new Python source
file, gdbcopyright.py. The function in this file will find the
copyright dates by scanning the calling script. This will be reused
in a future patch.
This involved minor changes to the output of gdbarch.py. Also, I've
updated copyright.py to remove the reference to gdbarch.sh. We don't
need to mention gdbarch.py there, either.
GDB Administrator [Wed, 2 Mar 2022 00:00:18 +0000 (00:00 +0000)]
Automatic date update in version.in
Tom Tromey [Mon, 28 Feb 2022 00:31:54 +0000 (17:31 -0700)]
Some "distclean" fixes in gdb
PR build/12440 points out that "make distclean" is broken in gdb.
Most of the breakage comes from other projects in the tree, but we can
fix some of the issues, which is what this patch does.
Note that the yacc output files, like c-exp.c, are left alone. In a
source distribution, these are included in the tarball, and if the
user builds in-tree, we would not want to remove them.
While that seems a bit obscure, it seems to me that "distclean" is
only really useful for in-tree builds anyway -- out-of-tree I simply
delete the entire build directory and start over.
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=12440
Tom Tromey [Tue, 1 Mar 2022 23:50:39 +0000 (16:50 -0700)]
Fix typo in the "alias" example
PR cli/17332, filed around 8 years ago, points out a typo in the docs
-- in one example, the command and its output are obviously out of
sync. This patch fixes it. I'm checking this in as obvious.
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=17332
Nick Clifton [Tue, 1 Mar 2022 13:13:42 +0000 (13:13 +0000)]
Fix a typo in the previous delta to bfdio.c.
PR 25713
* bfdio.c (_bfd_real_fopen): Fix typo.
Alan Modra [Tue, 1 Mar 2022 11:24:34 +0000 (21:54 +1030)]
Revert "Check thin archive element file size against archive header"
This reverts commit
48e3e6aec8a4f37d00ea6c0da3ab45e76490e3db.
PR 28929
* archive.c (_bfd_get_elt_at_filepos): Don't check thin archive
element file size.
Nick Clifton [Tue, 1 Mar 2022 10:10:20 +0000 (10:10 +0000)]
Fix linker tests to compile with gcc-12.
PR 21964
* testsuite/ld-elf/pr21964-1a.c: Fix array comparisons.
* testsuite/ld-elf/pr21964-1b.c: Likewise.
* testsuite/ld-elf/pr21964-1c.c: Likewise.
* testsuite/ld-elf/pr21964-2a.c: Likewise.
* testsuite/ld-elf/pr21964-2b.c: Likewise.
* testsuite/ld-elf/pr21964-3a.c: Likewise.
Nick Clifton [Tue, 1 Mar 2022 09:51:59 +0000 (09:51 +0000)]
Prevent an assertion from being triggered when linking an ARM object file with incorrectly set build attributes.
PR 28848
PR 28859
* elf32-arm.c (elf32_arm_merge_eabi_attributes): If the first
input bfd has a Tag_ABI_HardFP_use set to 3 but does not also have
TAG_FP_arch set then reset the TAG_ABI_HardFP_use.
Tiezhu Yang [Tue, 1 Mar 2022 07:05:00 +0000 (15:05 +0800)]
gdb: testsuite: fix wrong expected result in attach-pie-noexec.exp
If /proc/sys/kernel/yama/ptrace_scope is 1, when execute the test case
gdb.base/attach-pie-noexec.exp without superuser, the gdb.log shows the
following info:
(gdb) attach 6500
Attaching to process 6500
ptrace: Operation not permitted.
(gdb) PASS: gdb.base/attach-pie-noexec.exp: attach
It is obviously wrong, the expected result should be UNSUPPORTED in such
a case.
It is better to make can_spawn_for_attach to return false for this case.
It would have to setup a small test program, compile it to exec, spawn it
and try to attach to it.
With this patch, we can see "Operation not permitted" in the log info,
and then we can do the following processes to test:
(1) set ptrace_scope as 0
$ echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
$ make check-gdb TESTS="gdb.base/attach-pie-noexec.exp"
(2) use sudo
$ sudo make check-gdb TESTS="gdb.base/attach-pie-noexec.exp"
Additionally, handle the other cases when test with RUNTESTFLAGS=
"--target_board=native-extended-gdbserver".
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Tiezhu Yang [Tue, 1 Mar 2022 07:03:00 +0000 (15:03 +0800)]
gdb: testsuite: print explicit test result in can_spawn_for_attach
In the current code, there is no test result when execute the following
commands:
$ make check-gdb TESTS="gdb.base/attach-pie-noexec.exp" RUNTESTFLAGS="--target_board=remote-gdbserver-on-localhost"
$ make check-gdb TESTS="gdb.base/attach-pie-noexec.exp" RUNTESTFLAGS="--target_board=native-gdbserver"
It is better to print explicit test result in can_spawn_for_attach.
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Tiezhu Yang [Tue, 1 Mar 2022 07:01:00 +0000 (15:01 +0800)]
gdb: add Tiezhu Yang as LoongArch maintainer
The patch series "gdb: Add basic support for LoongArch" has been
merged into master, list Tiezhu Yang as LoongArch maintainer.
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
GDB Administrator [Tue, 1 Mar 2022 00:00:26 +0000 (00:00 +0000)]
Automatic date update in version.in
Keith Seitz [Mon, 28 Feb 2022 19:55:51 +0000 (11:55 -0800)]
Fix "spawn id XYZ not open" errors in gdb.mi/mi-exec-run.exp
Running mi-exec-run.exp on native-extended-gdbserver/-m{32,64}
causes several Tcl errors to appear. For example,
(gdb)
ERROR: : spawn id exp20 not open
while executing
"expect {
-i exp11 -timeout 10
-i "$inferior_spawn_id"
-re ".*Cannot exec.*Permission denied" {
set saw_perm_error 1
verbose -log "saw..."
("uplevel" body line 1)
invoked from within
"uplevel $body" NONE : spawn id exp20 not open
UNRESOLVED: gdb.mi/mi-exec-run.exp: inferior-tty=separate: mi=separate: force-fail=1: run failure detected (eof)
This is happening because of the way this test is implemented:
while {1} {
gdb_expect {
-i "$inferior_spawn_id"
-re ".*Cannot exec.*Permission denied" {
set saw_perm_error 1
verbose -log "saw mi error"
}
-i "$gdb_spawn_id"
-re "\\^error,msg=\"During startup program exited with code 127" {
set saw_mi_error 1
verbose -log "saw mi error"
}
# and so on
}
}
The first time this loop is executed, `inferior_spawn_id' is valid. When the
first branch of the expect statement is reached, gdbserver has exited, closing
the spawn_id. Since we haven't seen the gdb-side error yet, the loop is executed
again. The first branch now refers to a non-existent spawn_id, leading to the error.
This can be fixed by using exp_continue to loop in expect instead of looping around
expect, which is the approach I have used[1]. Note I've had to update the expected
message for the "During startup..." error message when running with gdbserver.
One other small change I've made is to add a log entry which spills the values of
the two variables, saw_mi_error and saw_perm_error (and updated the log output
for the later). This should make the log output clearer about why the test failed.
With this patch installed, all the ERRORs disappear, leaving previously masked
FAILs (which I have not attempted to fix).
[1] Anyone know why this test doesn't simply use gdb_test_multiple? I can only
assume that it was intentionally written this way, and I've modified the code with
that assumption. I have tested a version using gdb_test_multiple, and that appears
to work fine, too, if that is preferred. [It still employs exp_continue to fix the
spawn_id errors.]