I see the following fail on arm-none-eabi target,
(gdb) b 24^M
Breakpoint 1 at 0x4: file
../../../../git/gdb/testsuite/gdb.base/break-on-linker-gcd-function.cc,
line 24.^M
(gdb) FAIL: gdb.base/break-on-linker-gcd-function.exp: b 24
Currently, we are using flag has_section_at_zero to determine whether
address zero in debug info means the corresponding code has been
GC'ed, like this:
case DW_LNE_set_address:
address = read_address (abfd, line_ptr, cu, &bytes_read);
if (address == 0 && !dwarf2_per_objfile->has_section_at_zero)
{
/* This line table is for a function which has been
GCd by the linker. Ignore it. PR gdb/12528 */
However, this is incorrect on some bare metal targets, as .text
section is located at 0x0, so dwarf2_per_objfile->has_section_at_zero
is true. If a function is GC'ed by linker, the address is zero. GDB
thinks address zero is a function's address rather than this function
is GC'ed.
In this patch, we choose 'lowpc' got in read_file_scope to check
whether 'lowpc' is greater than zero. If it isn't, address zero really
means the function is GC'ed. In this patch, we pass 'lowpc' in
read_file_scope through handle_DW_AT_stmt_list and dwarf_decode_lines,
and to dwarf_decode_lines_1 finally.
This patch fixes the fail above. This patch also covers the path that
partial symbol isn't used, which is tested by starting gdb with
--readnow option.
It is regression tested on x86-linux with
target_board=dwarf4-gdb-index, and arm-none-eabi. OK to apply?
gdb:
2014-09-19 Yao Qi <yao@codesourcery.com>
* dwarf2read.c (dwarf_decode_lines): Update declaration.
(handle_DW_AT_stmt_list): Add argument 'lowpc'. Update
comments. Callers update.
(dwarf_decode_lines): Likewise.
(dwarf_decode_lines_1): Add argument 'lowpc'. Update
comments. Skip the line table if 'lowpc' is greater than
'address'. Don't check
dwarf2_per_objfile->has_section_at_zero.
gdb/testsuite:
2014-09-19 Yao Qi <yao@codesourcery.com>
* gdb.base/break-on-linker-gcd-function.exp: Move test into new
proc set_breakpoint_on_gcd_function. Invoke
set_breakpoint_on_gcd_function. Restart GDB with --readnow and
invoke set_breakpoint_on_gcd_function again.
This is just a testcase addition that I am proposing for upstream GDB.
We have this in our internal tree, and the related RH bug is:
<https://bugzilla.redhat.com/show_bug.cgi?id=809179>
(You might not be able to see all the comments without privileges.)
This bug is about a global variable that got incorrectly displayed by
GDB. This bug has already been fixed a long time ago by Joel's
commit:
commit 19630284f5
Author: Joel Brobecker <brobecker@gnat.com>
Date: Tue Jun 5 13:50:50 2012 +0000
But I think a testcase for it wouldn't hurt.
So, consider the following scenario:
$ cat solib1.c
int test;
void c_main (void)
{
test = 42;
}
$ cat solib2.c
int test;
void b_main (void)
{
test = 42;
}
$ cat main.c
int main (int argc, char *argv[])
{
c_main ();
b_main ();
return 0;
}
$ gcc -g -fPIC -shared -o libSO1.so -c solib1.c
$ gcc -g -fPIC -shared -o libSO2.so -c solib2.c
$ gcc -g -o main -L$PWD -lSO1 -lSO2 main.c
$ LD_LIBRARY_PATH=. gdb -q -batch -ex 'b c_main' -ex r -ex n -ex 'p test' ./main
...
$1 = 0
This happened with GDB before Joel's commit above. Now, things work
and GDB is able to correctly display the nested global variable:
$ LD_LIBRARY_PATH=. gdb -q -batch -ex 'b c_main' -ex r -ex n -ex 'p test' ./main
...
$1 = 42
The testcase attached tests this behavior.
gdb/testsuite/ChangeLog:
2014-09-16 Sergio Durigan Junior <sergiodj@redhat.com>
* gdb.base/global-var-nested-by-dso-solib1.c: New file.
* gdb.base/global-var-nested-by-dso-solib2.c: Likewise.
* gdb.base/global-var-nested-by-dso.c: Likewise.
* gdb.base/global-var-nested-by-dso.exp: Likewise.
Make test messages unique and a couple other tweaks.
gdb/testsuite/
2014-09-16 Sergio Durigan Junior <sergiodj@redhat.com>
Pedro Alves <palves@redhat.com>
* gdb.base/watch-bitfields.exp: Pass string other than test file
name to prepare_for_testing.
(watch): New procedure.
(expect_watchpoint): Use with_test_prefix.
(top level): Factor out tests to ...
(test_watch_location, test_regular_watch): ... these new
procedures, and use with_test_prefix and gdb_continue_to_end.
PR 12526 reports that -location watchpoints against bitfield arguments
trigger false positives when bits around the bitfield, but not the
bitfield itself, are modified.
This happens because -location watchpoints naturally operate at the
byte level, not at the bit level. When the address of a bitfield
lvalue is taken, information about the bitfield (i.e. its offset and
size) is lost in the process.
This information must first be retained throughout the lifetime of the
-location watchpoint. This patch achieves this by adding two new
fields to the watchpoint struct: val_bitpos and val_bitsize. These
fields are set when a watchpoint is first defined in watch_command_1.
They are both equal to zero if the watchpoint is not a -location
watchpoint or if the argument is not a bitfield.
Then these bitfield parameters are used inside update_watchpoint and
watchpoint_check to extract the actual value of the bitfield from the
watchpoint address, with the help of a local helper function
extract_bitfield_from_watchpoint_value.
Finally when creating a HW breakpoint pointing to a bitfield, we
optimize the address and length of the breakpoint. By skipping over
the bytes that don't cover the bitfield, this step reduces the
frequency at which a read watchpoint for the bitfield is triggered.
It also reduces the number of times a false-positive call to
check_watchpoint is triggered for a write watchpoint.
gdb/
PR breakpoints/12526
* breakpoint.h (struct watchpoint): New fields val_bitpos and
val_bitsize.
* breakpoint.c (watch_command_1): Use these fields to retain
bitfield information.
(extract_bitfield_from_watchpoint_value): New function.
(watchpoint_check): Use it.
(update_watchpoint): Use it. Optimize the address and length of a
HW watchpoint pointing to a bitfield.
* value.h (unpack_value_bitfield): New prototype.
* value.c (unpack_value_bitfield): Make extern.
gdb/testsuite/
PR breakpoints/12526
* gdb.base/watch-bitfields.exp: New file.
* gdb.base/watch-bitfields.c: New file.
Silly typo...
gdb/testsuite/
2014-09-16 Pedro Alves <palves@redhat.com>
* gdb.base/watchpoint-stops-at-right-insn.exp (test): Compare
software and hardware addresses, not software address against
itself.
This adds a test that makes sure GDB knows whether the target has
continuable, or non-continuable watchpoints.
That is, the test confirms that GDB presents a watchpoint value change
at the first instruction right after the instruction that changes
memory.
gdb/testsuite/ChangeLog:
2014-09-16 Pedro Alves <palves@redhat.com>
* gdb.base/watchpoint-stops-at-right-insn.c: New file.
* gdb.base/watchpoint-stops-at-right-insn.exp: New file.
In the recent review to my patch about copying files to remote host,
we find that we need a board file which is more closely mapped real
remote host testing to improve coverage. With the board file
local-remote-host-native.exp, DejaGNU copies files to
$build/gdb/testsuite/remote-host to emulate the effect of remote host.
Is it OK?
gdb/testsuite:
2014-09-16 Yao Qi <yao@codesourcery.com>
* boards/local-remote-host-native.exp: New file.
The test does a backtrace to see which thread (#2 or #3) is assigned
to which SIGUSR (1 or 2). If the main thread gets to all_threads_running
before the sigusr threads get to their entry point, then the function
name isn't in the backtrace and the test fails.
Alas this version of the code is within epsilon of what I started with,
and then over-simplified things.
If I want to change the signalled state of multiple threads
it's a bit cumbersome to do with the "signal" command.
What you really want is a way to set the signal state of the
desired threads and then just do "continue".
This patch adds a new command, queue-signal, to accomplish this.
Basically "signal N" == "queue-signal N" + "continue".
That's not precisely true in that "signal" can be used to inject
any signal, including signals set to "nopass"; whereas "queue-signal"
just queues the signal as if the thread stopped because of it.
"nopass" handling is done when the thread is resumed which
"queue-signal" doesn't do.
One could add extra complexity to allow queue-signal to be used to
deliver "nopass" signals like the "signal" command. I have no current
need for it so in the interests of incremental complexity, I have
left such support out and just have the code flag an error if one
tries to queue a nopass signal.
gdb/ChangeLog:
* NEWS: Mention new "queue-signal" command.
* infcmd.c (queue_signal_command): New function.
(_initialize_infcmd): Add new queue-signal command.
gdb/doc/ChangeLog:
* gdb.texinfo (Signaling): Document new queue-signal command.
gdb/testsuite/ChangeLog:
* gdb.threads/queue-signal.c: New file.
* gdb.threads/queue-signal.exp: New file.
I had occasion to use with_gdb_prompt in a test for the patch for PR 17314
and was passing the plain text prompt as the value, "(top-gdb)",
instead of a regexp, "\(top-gdb\)" (expressed as "\\(top-gdb\\)" in TCL).
I then discovered that in order to restore the prompt gdb passes the
original value of $gdb_prompt to "set prompt", which works because
"set prompt \(gdb\) " is equivalent to "set prompt (gdb) ".
Perhaps I'm being overly cautious but this feels a bit subtle,
but at any rate as an API choice I'd much rather pass the plain text
form to with_gdb_prompt.
I also discovered that the initial value of gdb_prompt is set in
two places to two different values.
At the global level gdb.exp sets it to "\[(\]gdb\[)\]"
and default_gdb_init sets it to "\\(gdb\\)".
The former form is undesirable as an argument to "set prompt",
but it's not clear to me that just deleting this code won't break
anything. Thus I just changed the value to be consistent and added
a comment.
gdb/testsuite/ChangeLog:
* lib/gdb.exp (gdb_prompt): Add comment and change initial value to
be consistent with what default_gdb_init uses.
(with_gdb_prompt): Change form of PROMPT argument from a regexp to
the plain text of the prompt. Add some logging printfs.
* gdb.perf/disassemble.exp: Update call to with_gdb_prompt.
See:
https://sourceware.org/ml/gdb-patches/2014-09/msg00404.html
We have a number of places that do gdb_run_cmd followed by gdb_expect,
when it would be better to use gdb_test_multiple or gdb_test.
This converts all that "grep gdb_run_cmd -A 2 | grep gdb_expect"
found.
Tested on x86_64 Fedora 20, native and gdbserver.
gdb/testsuite/
2014-09-12 Pedro Alves <palves@redhat.com>
* gdb.arch/gdb1558.exp: Replace uses of gdb_expect after
gdb_run_cmd with gdb_test_multiple or gdb_test throughout.
* gdb.arch/i386-size-overlap.exp: Likewise.
* gdb.arch/i386-size.exp: Likewise.
* gdb.arch/i386-unwind.exp: Likewise.
* gdb.base/a2-run.exp: Likewise.
* gdb.base/break.exp: Likewise.
* gdb.base/charset.exp: Likewise.
* gdb.base/chng-syms.exp: Likewise.
* gdb.base/commands.exp: Likewise.
* gdb.base/dbx.exp: Likewise.
* gdb.base/find.exp: Likewise.
* gdb.base/funcargs.exp: Likewise.
* gdb.base/jit-simple.exp: Likewise.
* gdb.base/reread.exp: Likewise.
* gdb.base/sepdebug.exp: Likewise.
* gdb.base/step-bt.exp: Likewise.
* gdb.cp/mb-inline.exp: Likewise.
* gdb.cp/mb-templates.exp: Likewise.
* gdb.objc/basicclass.exp: Likewise.
* gdb.threads/killed.exp: Likewise.
The problem is that rs6000_frame_cache attempts to read the stack backchain via
read_memory_unsigned_integer, which throws an exception if the stack pointer is
invalid. With this patch, it calls safe_read_memory_integer instead, which
doesn't throw an exception and allows for safe handling of that situation.
gdb/ChangeLog
2014-09-12 Edjunior Barbosa Machado <emachado@linux.vnet.ibm.com>
Ulrich Weigand <uweigand@de.ibm.com>
PR tdep/17379
* rs6000-tdep.c (rs6000_frame_cache): Use safe_read_memory_integer
instead of read_memory_unsigned_integer.
gdb/testcase/ChangeLog
2014-09-12 Edjunior Barbosa Machado <emachado@linux.vnet.ibm.com>
PR tdep/17379
* gdb.arch/powerpc-stackless.S: New file.
* gdb.arch/powerpc-stackless.exp: New file.
I have started seeing occasional runaway 'attach' processes these days.
I cannot be certain it is really caused by this patch, for example
grep 'FAIL.*cmdline attach run' does not show anything in my logs.
But as I remember this 'attach' runaway process always happened in GDB (but
I do not remember it in the past months) I think it would be most safe to just
solve it forever by [attached].
gdb/testsuite/ChangeLog
2014-09-12 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.base/attach.c: Include unistd.h.
(main): Call alarm. Add label postloop.
* gdb.base/attach.exp (do_attach_tests): Use gdb_get_line_number,
gdb_breakpoint, gdb_continue_to_breakpoint.
(test_command_line_attach_run): Kill ${testpid} in one exit path.
Doing:
gdb --pid=PID -ex run
Results in GDB getting a SIGTTIN, and thus ending stopped. That's
usually indicative of a missing target_terminal_ours call.
E.g., from the PR:
$ sleep 1h & p=$!; sleep 0.1; gdb -batch sleep $p -ex run
[1] 28263
[1] Killed sleep 1h
[2]+ Stopped gdb -batch sleep $p -ex run
The workaround is doing:
gdb -ex "attach $PID" -ex "run"
instead of
gdb [-p] $PID -ex "run"
With the former, gdb waits for the attach command to complete before
moving on to the "run" command, because the interpreter is in sync
mode at this point, within execute_command. But for the latter,
attach_command is called directly from captured_main, and thus misses
that waiting. IOW, "run" is running before the attach continuation
has run, before the program stops and attach completes. The broken
terminal settings are just one symptom of that. Any command that
queries or requires input results in the same.
The fix is to wait in catch_command_errors (which is specific to
main.c nowadays), just like we wait in execute_command.
gdb/ChangeLog:
2014-09-11 Pedro Alves <palves@redhat.com>
PR gdb/17347
* main.c: Include "infrun.h".
(catch_command_errors, catch_command_errors_const): Wait for the
foreground command to complete.
* top.c (maybe_wait_sync_command_done): New function, factored out
from ...
(maybe_wait_sync_command_done): ... here.
* top.h (maybe_wait_sync_command_done): New declaration.
gdb/testsuite/ChangeLog:
2014-09-11 Pedro Alves <palves@redhat.com>
PR gdb/17347
* lib/gdb.exp (gdb_spawn_with_cmdline_opts): New procedure.
* gdb.base/attach.exp (test_command_line_attach_run): New
procedure.
(top level): Call it.
Several places in the testsuite have a copy of a snippet of code that
spawns a test program, waits a bit, and then does some PID munging for
Cygwin. This is in order to have GDB attach to the spawned program.
This refactors all that to a common procedure.
(multi-attach.exp wants to spawn multiple processes, so this makes the
new procedure's interface work with lists.)
Tested on x86_64 Fedora 20.
gdb/testsuite/ChangeLog:
2014-09-11 Pedro Alves <palves@redhat.com>
* lib/gdb.exp (spawn_wait_for_attach): New procedure.
* gdb.base/attach.exp (do_attach_tests, do_call_attach_tests)
(do_command_attach_tests): Use spawn_wait_for_attach.
* gdb.base/solib-overlap.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
This fixes two FAIL results on this testcase which were caused by a
misplaced "continue" command. This testcase used to end inferior's
execution too soon, causing the following tests to fail. Now we break
right after inferior's loop and perform the rest of the tests there.
gdb/testsuite/ChangeLog:
* gdb.fortran/array-element.exp: Remove unexpected "continue"
command in testcase. Simplify testcase.
Trying to print the bounds or the length of a pointer to an array
whose bounds are dynamic results in the following error:
(gdb) p foo.three_ptr.all'first
Location address is not set.
(gdb) p foo.three_ptr.all'length
Location address is not set.
This is because, after having dereferenced our array pointer, we
use the type of the resulting array value, instead of the enclosing
type. The former is the original type where the bounds are unresolved,
whereas we need to get the actual array bounds.
Similarly, trying to apply those attributes to the array pointer
directly (without explicitly dereferencing it with the '.all'
operator) yields the same kind of error:
(gdb) p foo.three_ptr'first
Location address is not set.
(gdb) p foo.three_ptr'length
Location address is not set.
This is caused by the fact that the dereference was done implicitly
in this case, and perform at the type level only, which is not
sufficient in order to resolve the array type.
This patch fixes both issues, thus allowing us to get the expected output:
(gdb) p foo.three_ptr.all'first
$1 = 1
(gdb) p foo.three_ptr.all'length
$2 = 3
(gdb) p foo.three_ptr'first
$3 = 1
(gdb) p foo.three_ptr'length
$4 = 3
gdb/ChangeLog:
* ada-lang.c (ada_array_bound): If ARR is a TYPE_CODE_PTR,
dereference it first. Use value_enclosing_type instead of
value_type.
(ada_array_length): Likewise.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/dynarr-ptr.exp: Add 'first, 'last and 'length tests.
Consider a pointer to an array which dynamic bounds, described in
DWARF as follow:
<1><25>: Abbrev Number: 4 (DW_TAG_array_type)
<26> DW_AT_name : foo__array_type
[...]
<2><3b>: Abbrev Number: 5 (DW_TAG_subrange_type)
[...]
<40> DW_AT_lower_bound : 5 byte block: 97 38 1c 94 4
(DW_OP_push_object_address; DW_OP_lit8; DW_OP_minus;
DW_OP_deref_size: 4)
<46> DW_AT_upper_bound : 5 byte block: 97 34 1c 94 4
(DW_OP_push_object_address; DW_OP_lit4; DW_OP_minus;
DW_OP_deref_size: 4)
GDB is now able to correctly print the entire array, but not one
element of the array. Eg:
(gdb) p foo.three_ptr.all
$1 = (1, 2, 3)
(gdb) p foo.three_ptr.all(1)
Cannot access memory at address 0xfffffffff4123a0c
The problem occurs because we are missing a dynamic resolution of
the variable's array type when subscripting the array. What the current
code does is "fix"-ing the array type using the GNAT encodings, but
that operation ignores any of the array's dynamic properties.
This patch fixes the issue by using ada_value_ind to dereference
the array pointer, which takes care of the array type resolution.
It also continues to "fix" arrays described using GNAT encodings,
so backwards compatibility is preserved.
gdb/ChangeLog:
* ada-lang.c (ada_value_ptr_subscript): Remove parameter "type".
Adjust function implementation and documentation accordingly.
(ada_evaluate_subexp) <OP_FUNCALL>: Only assign "type" if
NOSIDE is EVAL_AVOID_SIDE_EFFECTS.
Update call to ada_value_ptr_subscript.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/dynarr-ptr.exp: Add subscripting tests.
Consider the following declaration:
type Array_Type is array (Natural range <>) of Integer;
type Array_Ptr is access all Array_Type;
for Array_Ptr'Size use 64;
Three_Ptr : Array_Ptr := new Array_Type'(1 => 1, 2 => 2, 3 => 3);
This creates a pointer to an array where the bounds are stored
in a memory region just before the array itself (aka a "thin pointer").
In DWARF, this is described as a the usual pointer type to an array
whose subrange has dynamic values for its bounds:
<1><25>: Abbrev Number: 4 (DW_TAG_array_type)
<26> DW_AT_name : foo__array_type
[...]
<2><3b>: Abbrev Number: 5 (DW_TAG_subrange_type)
[...]
<40> DW_AT_lower_bound : 5 byte block: 97 38 1c 94 4
(DW_OP_push_object_address; DW_OP_lit8; DW_OP_minus;
DW_OP_deref_size: 4)
<46> DW_AT_upper_bound : 5 byte block: 97 34 1c 94 4
(DW_OP_push_object_address; DW_OP_lit4; DW_OP_minus;
DW_OP_deref_size: 4)
GDB is currently printing the value of the array incorrectly:
(gdb) p foo.three_ptr.all
$1 = (26629472 => 1, 2,
value.c:819: internal-error: value_contents_bits_eq: [...]
The dereferencing (".all" operator) is done by calling ada_value_ind,
which itself calls value_ind. It first produces a new value where
the bounds of the array were correctly resolved to their actual value,
but then calls readjust_indirect_value_type which replaces the resolved
type by the original type.
The problem starts when ada_value_print does not take this situation
into account, and starts using the type of the resulting value, which
has unresolved array bounds, instead of using the value's enclosing
type.
After fixing this issue, the debugger now correctly prints:
(gdb) p foo.three_ptr.all
$1 = (1, 2, 3)
gdb/ChangeLog:
* ada-valprint.c (ada_value_print): Use VAL's enclosing type
instead of VAL's type.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/dynarr-ptr.c: New file.
* gdb.dwarf2/dynarr-ptr.exp: New file.
Similarly to the previous changes to gdb.reverse/sigall-reverse.exp and
gdb.reverse/until-precsave.exp this corrects the timeout tweak in
gdb.base/watchpoint-solib.exp.
This test case executes a large amount of code with a software watchpoint
enabled. This means single-stepping all the way through and takes a lot
of time, e.g. for an ARMv7 Panda board and a `-march=armv5te' multilib:
PASS: gdb.base/watchpoint-solib.exp: continue to foo again
elapsed: 714
for the same board and a `-mthumb -march=armv5te' multilib:
PASS: gdb.base/watchpoint-solib.exp: continue to foo again
elapsed: 1275
and for QEMU in the system emulation mode and a `-march=armv4t'
multilib:
PASS: gdb.base/watchpoint-solib.exp: continue to foo again
elapsed: 115
(values in seconds) -- all of which having the default timeout of 60s,
set based on the requirement of the remaining test cases (other than
gdb.reverse ones).
Here again the timeout extension to have a meaning should be calculated
by scaling rather than using an arbitrary constant, and a larger factor
of 30 will do, leaving some margin. Hopefully for everyone or otherwise
we'll probably have to come up with a smarter solution.
OTOH the other test cases in this script do not require the extension so
they can be moved outside its umbrella so as to avoid unnecessary delays
if something goes wrong and a genuine timeout triggers.
* gdb.base/watchpoint-solib.exp: Increase the timeout by a factor
of 30 rather than hardcoding 120 for a slow test case. Take the
`gdb,timeout' target setting into account for this calculation.
Don't extend the timeout for the test cases that don't need it.
There are three cases in two scripts in the gdb.reverse subset that
take a particularly long time. Two of them are already attempted to
take care of by extending the timeout from the default. The remaining
one has no precautions taken. The timeout extension is ineffective
though, it is done by adding a constant rather than by scaling and as
a result while it may work for target boards that get satisfied with
the detault test timeout of 10s, it does not serve its purpose for
slower ones.
Here are indicative samples of execution times (in seconds) observed
for these cases respectively, for an ARMv7 Panda board running Linux
and a `-march=armv5te' multilib:
PASS: gdb.reverse/sigall-reverse.exp: continue to signal exit
elapsed: 385
PASS: gdb.reverse/until-precsave.exp: run to end of main
elapsed: 4440
PASS: gdb.reverse/until-precsave.exp: save process recfile
elapsed: 965
for the same board and a `-mthumb -march=armv5te' multilib:
PASS: gdb.reverse/sigall-reverse.exp: continue to signal exit
elapsed: 465
PASS: gdb.reverse/until-precsave.exp: run to end of main
elapsed: 4191
PASS: gdb.reverse/until-precsave.exp: save process recfile
elapsed: 669
and for QEMU in the system emulation mode and a `-march=armv4t'
multilib:
PASS: gdb.reverse/sigall-reverse.exp: continue to signal exit
elapsed: 45
PASS: gdb.reverse/until-precsave.exp: run to end of main
elapsed: 433
PASS: gdb.reverse/until-precsave.exp: save process recfile
elapsed: 104
Based on the performance of other tests these two test configurations
have their default timeout set to 450s and 60s respectively.
The remaining two multilibs (`-mthumb -march=armv4t' and `-mthumb
-march=armv7-a') do not produce test results usable enough to have data
available for these cases.
Based on these results I have tweaked timeouts for these cases as
follows. This, together with a suitable board timeout setting, removes
timeouts for these cases. Note that for the default timeout of 10s the
new setting for the first case in gdb.reverse/until-precsave.exp is
compatible with the old one, just a bit higher to keep the convention
of longer timeouts to remain multiples of 30s. The second case there
does not need such a high setting so I have lowered it a bit to avoid
an unnecessary delay where this test case genuinely times out.
* gdb.reverse/sigall-reverse.exp: Increase the timeout by
a factor of 2 for a slow test case. Take the `gdb,timeout'
target setting into account for this calculation.
* gdb.reverse/until-precsave.exp: Increase the timeout by
a factor of 15 and 3 respectively rather than adding 120
for a pair of slow test cases. Take the `gdb,timeout'
target setting into account for this calculation.
The recent change to introduce `gdb_reverse_timeout' turned out
ineffective for board setups that set the `gdb,timeout' target variable.
A lower `gdb,timeout' setting takes precedence and defeats the effect of
`gdb_reverse_timeout'. This is because the global timeout is overridden
in gdb_test_multiple and then again in gdb_expect.
Three timeout variables are taken into account in these two places, in
this precedence:
1. The `gdb,timeout' target variable.
2. The caller's local `timeout' variable (upvar timeout)
3. The global `timeout' variable.
This precedence is obeyed by gdb_test_multiple strictly. OTOH
gdb_expect will select the higher of the two formers and will only take
the latter into account if none of the formers is present. However the
two timeout selections are conceptually the same and gdb_test_multiple
does its only for the purpose of passing it down to gdb_expect.
Therefore I decided there is no point to keep carrying on this
duplication and removed the sequence from gdb_test_multiple, however
retaining the `upvar timeout' variable definition. This way gdb_expect
will still access gdb_test_multiple's caller `timeout' variable (if any)
via its own `upvar timeout' reference.
Now as to the sequence in gdb_expect. In addition to the three
variables described above it also takes a timeout argument into account,
as the fourth value to choose from. It is currently used if it is
higher than the timeout selected from the variables as described above.
With the timeout selection code from gdb_test_multiple gone, gone is
also the most prominent use of this timeout argument, it's now used in
a couple of places only, mostly within this test framework library code
itself for preparatory commands or suchlike. With this being the case
this timeout selection code can be simplified as follows:
1. Among the three timeout variables, the highest is always chosen.
This is so that a test case doesn't inadvertently lower a high value
timeout needed by slow target boards. This is what all test cases
use.
2. Any timeout argument takes precedence. This is for special cases
such as within the framework library code, e.g. it doesn't make sense
to send `set height 0' with a timeout of 7200 seconds. This is a
local command that does not interact with the target and setting a
high timeout here only risks a test suite run taking ages if it goes
astray for some reason.
3. The fallback timeout of 60s remains.
* lib/gdb.exp (gdb_test_multiple): Remove code to select the
timeout, don't pass one down to gdb_expect.
(gdb_expect): Rework timeout selection.
As it happens we have a board that fails a gdb.base/gcore-relro.exp
test case reproducibly and moreover the case appears to trigger a
kernel bug making the it less than usable. Specifically the board
remains responsive to some extent, however processes do not appear
to be able to successfully complete termination anymore and perhaps
more importantly further gdbserver processes can be started, but they
never reach the stage of listening on the RSP socket.
This change handles timeouts in gdbserver start properly, by throwing
a TCL error exception when gdbserver does not report listening on the
RSP socket in time. This is then caught at the outer level and
reported, and 2 rather than 1 is returned so that the caller may tell
the failure to start gdbserver and other issues apart and act
accordingly (or do nothing).
I thought letting the exception unwind further on might be a good idea
for any test harnesses out there to break outright where a gdbserver
start error is silently ignored right now, however I figured out the
calls to gdbserver-support.exp are buried down too deep in the GDB test
suite for such a change to be made easily. I think returning a distinct
return value is good enough (the API says "non-zero", so 2 is as good as
1) and we can always make the error harder in a later step if required.
With config/gdbserver.exp being used this change remains transparent
to the target board, the return value is passed up by gdb_reload and
the error exception unwinds through gdbserver_gdb_load and is caught
and handled by mi_gdb_target_load. A call to perror is still made,
reporting the timeout, and in the case of mi_gdb_target_load the
procedure returns a value denoting unsuccessful completion. An
unsuccessful completion of gdb_reload is already handled elsewhere.
An alternative gdbserver board configuration can interpret the return
value in its gdb_reload implementation and catch the error in
gdbserver_gdb_load in an attempt to recover a target board that has
gone astray, for example by rebooting the board somehow. This has
proved effective with our failing board, that now completes the
remaining test cases with no further hiccups.
* lib/gdbserver-support.exp (gdbserver_start): Throw an error
exception on timeout.
(gdbserver_run): Catch any `gdbserver_spawn' error exceptions.
(gdbserver_start_extended): Catch any `gdbserver_start' error
exceptions.
(gdbserver_start_multi, mi_gdbserver_start_multi): Likewise.
* lib/mi-support.exp (mi_gdb_target_load): Catch any
`gdbserver_gdb_load' error exceptions.
Gdbserver support code uses the global timeout value to determine when
to stop waiting for a gdbserver process being started to respond before
continuing anyway. This timeout is usually as low as 10s and may not
be enough in this context, for example on the first run where the
filesystem cache is cold, even if it is elsewhere.
E.g. I observe this reliably with gdbserver started the first time in
QEMU running in the system emulation mode:
(gdb) file .../gdb.base/advance
Reading symbols from .../gdb.base/advance...done.
(gdb) delete breakpoints
(gdb) info breakpoints
No breakpoints or watchpoints.
(gdb) break main
Breakpoint 1 at 0x87f8: file .../gdb.base/advance.c,
line 41.
(gdb) set remotetimeout 15
(gdb) kill
The program is not being run.
(gdb)
[...]
.../bin/gdbserver --once :6014 advance
target remote localhost:6014
Remote debugging using localhost:6014
Remote communication error. Target disconnected.: Connection reset by peer.
(gdb) continue
The program is not being run.
(gdb) Process advance created; pid = 999
Listening on port 6014
FAIL: gdb.base/advance.exp: Can't run to main
-- notice how the test harness proceeded with the `target remote ...'
command even though gdbserver hasn't completed its startup yet. A
while later when it's finally ready it's too late already. I checked
the timing here and it takes gdbserver roughly 25 seconds to start in
this scenario. Subsequent gdbserver starts in the same test run take
less time and usually complete within 10 seconds although occasionally
`target remote ...' precedes the corresponding `Listening on port...'
message again.
Therefore I have fixed this problem by setting an explicit timeout to
120s on the expect call in question. If this turns out too arbitrary
sometime, then perhaps a separate `gdbserver_timeout' setting might be
due.
* lib/gdbserver-support.exp (gdbserver_start): Set timeout to
120 on waiting for the TCP socket to open.
Hi,
I see the following fail on arm-none-eabi target,
-var-evaluate-expression -f nat foo^M
^done,value="0x3 <_ftext+2>"^M
(gdb) ^M
FAIL: gdb.mi/mi-var-display.exp: eval variable -f nat foo
the "<_ftext+2>" isn't expected in the test, so "set print symbol off"
can prevent printing it. It is obvious and I'll commit it in three
days if no comments.
gdb/testsuite:
2014-09-09 Yao Qi <yao@codesourcery.com>
* gdb.mi/mi-var-display.exp: Set print symbol off.
have empty bodies.
User-defined commands that have empty bodies weren't being shown because
the print function returned too soon. Now, it prints the command's name
before checking if it has any body at all. This also fixes the same
problem on "show user <myemptycommand>", which wasn't being printed due
to a similar reason.
gdb/Changelog:
* cli/cli-cmds.c (show_user): Use cli_user_command_p to
decide whether we display the command on "show user".
* cli/cli-script.c (show_user_1): Only verify cmdlines after
printing command name.
* cli/cli-decode.h (cli_user_command_p): Declare new function.
* cli/cli-decode.c (cli_user_command_p): Create helper function
to verify whether cmd_list_element is a user-defined command.
gdb/testsuite/Changelog:
* gdb.base/commands.exp: Add tests to verify user-defined
commands with empty bodies.
* gdb.python/py-cmd.exp: Test that we don't show user-defined
python commands in `show user command`.
* gdb.python/scm-cmd.exp: Test that we don't show user-defined
scheme commands in `show user command`.
https://bugzilla.redhat.com/show_bug.cgi?id=1126177
ERROR: AddressSanitizer: SEGV on unknown address 0x000000000050 (pc 0x000000992bef sp 0x7ffff9039530 bp 0x7ffff9039540
T0)
#0 0x992bee in value_type .../gdb/value.c:925
#1 0x87c951 in py_print_single_arg python/py-framefilter.c:445
#2 0x87cfae in enumerate_args python/py-framefilter.c:596
#3 0x87e0b0 in py_print_args python/py-framefilter.c:968
It crashes because frame_arg::val is documented it may contain NULL
(frame_arg::error is then non-NULL) but the code does not handle it.
Another bug is that py_print_single_arg() calls goto out of its TRY_CATCH
which messes up GDB cleanup chain crashing GDB later.
It is probably 7.7 regression (I have not verified it) due to the introduction
of Python frame filters.
gdb/ChangeLog
PR python/17355
* python/py-framefilter.c (py_print_single_arg): Handle NULL FA->VAL.
Fix goto out of TRY_CATCH.
gdb/testsuite/ChangeLog
PR python/17355
* gdb.python/amd64-py-framefilter-invalidarg.S: New file.
* gdb.python/py-framefilter-invalidarg-gdb.py.in: New file.
* gdb.python/py-framefilter-invalidarg.exp: New file.
* gdb.python/py-framefilter-invalidarg.py: New file.
This patch is a fix to PR gdb/17235. The bug is about an unused
variable that got declared and set during one of the parsing phases of
an SDT probe's argument. I took the opportunity to rewrite some of the
code to improve the parsing. The bug was actually a thinko, because
what I wanted to do in the code was to discard the number on the string
being parsed.
During this portion, the code identifies that it is dealing with an
expression that begins with a sign ('+', '-' or '~'). This means that
the expression could be:
- a numeric literal (e.g., '+5')
- a register displacement (e.g., '-4(%rsp)')
- a subexpression (e.g., '-(2*3)')
So, after saving the sign and moving forward 1 char, now the code needs
to know if there is a digit followed by a register displacement prefix
operand (e.g., '(' on x86_64). If yes, then it is a register
operation. If not, then it will be handled recursively, and the code
will later apply the requested operation on the result (either a '+', a
'-' or a '~').
With the bug, the code was correctly discarding the digit (though using
strtol unnecessarily), but it wasn't properly dealing with
subexpressions when the register indirection prefix was '(', like on
x86_64. This patch also fixes this bug, and includes a testcase. It
passes on x86_64 Fedora 20.
This commit fixes the PR mentioned in $subject. It is about a set but
unused variable that refers to the output format of integer values
printed in Fortran.
This was probably a thinko (like most set-but-unused-vars), but it
could cause an internal error depending on the scenario. I am sending
a testcase which triggers this error as well.
gdb/ChangeLog:
2014-09-04 Sergio Durigan Junior <sergiodj@redhat.com>
PR fortran/17237
* f-valprint.c (f_val_print): Specify the correct print option to
use when printing integer values.
gdb/testsuite/ChangeLog:
2014-09-04 Sergio Durigan Junior <sergiodj@redhat.com>
PR fortran/17237
* gdb.fortran/print-formatted.exp: New file.
* gdb.fortran/print-formatted.f90: Likewise.
The ability to read registers is needed to use Frame Filter API to
display the frames created by JIT compilers.
gdb/ChangeLog:
2014-08-29 Sasha Smundak <asmundak@google.com>
* python/py-frame.c (frapy_read_register): New function.
gdb/doc/ChangeLog:
2014-08-26 Sasha Smundak <asmundak@google.com>
* python.texi (Frames in Python): Add read_register description.
gdb/testsuite/ChangeLog:
2014-08-26 Sasha Smundak <asmundak@google.com>
* gdb.python/py-frame.exp: Test Frame.read_register.
This PR came from a Red Hat bug that was filed recently. I checked and
it still exists on HEAD, so here's a proposed fix. Although this is
marked as a Python backend bug, this is really about the completion
mechanism used by GDB. Since this code reminds me of my first attempt
to make a good noodle, it took me quite some time to fix it in a
non-intrusive way.
The problem is triggered when one registers a completion method inside a
class in a Python script, rather than registering the command using a
completer class directly. For example, consider the following script:
class MyFirstCommand(gdb.Command):
def __init__(self):
gdb.Command.__init__(self,'myfirstcommand',gdb.COMMAND_USER,gdb.COMPLETE_FILENAME)
def invoke(self,argument,from_tty):
raise gdb.GdbError('not implemented')
class MySecondCommand(gdb.Command):
def __init__(self):
gdb.Command.__init__(self,'mysecondcommand',gdb.COMMAND_USER)
def invoke(self,argument,from_tty):
raise gdb.GdbError('not implemented')
def complete(self,text,word):
return gdb.COMPLETE_FILENAME
MyFirstCommand ()
MySecondCommand ()
When one loads this into GDB and tries to complete filenames for both
myfirstcommand and mysecondcommand, she gets:
(gdb) myfirstcommand /hom<TAB>
(gdb) myfirstcommand /home/
^
...
(gdb) mysecondcommand /hom<TAB>
(gdb) mysecondcommand /home
^
(The "^" marks the final position of the cursor after the TAB).
So we see that myfirstcommand honors the COMPLETE_FILENAME class (as
specified in the command creation), but mysecondcommand does not. After
some investigation, I found that the problem lies with the set of word
break characters that is used for each case. The set should be the same
for both commands, but it is not.
During the process of deciding which type of completion should be used,
the code in gdb/completer.c:complete_line_internal analyses the command
that requested the completion and tries to determine the type of
completion wanted by checking which completion function will be called
(e.g., filename_completer for filenames, location_completer for
locations, etc.).
This all works fine for myfirstcommand, because immediately after the
command registration the Python backend already sets its completion
function to filename_completer (which then causes the
complete_line_internal function to choose the right set of word break
chars). However, for mysecondcommand, this decision is postponed to
when the completer function is evaluated, and the Python backend uses an
internal completer (called cmdpy_completer). complete_line_internal
doesn't know about this internal completer, and can't choose the right
set of word break chars in time, which then leads to a bad decision when
completing the "/hom" word.
So, after a few attempts, I decided to create another callback in
"struct cmd_list_element" that will be responsible for handling the case
when there is an unknown completer function for complete_line_internal
to work with. So far, only the Python backend uses this callback, and
only when the user provides a completer method instead of registering
the command directly with a completer class. I think this is the best
option because it not very intrusive (all the other commands will still
work normally), but especially because the whole completion code is so
messy that it would be hard to fix this without having to redesign
things.
I have regtested this on Fedora 18 x86_64, without regressions. I also
included a testcase.
gdb/ChangeLog:
2014-09-03 Sergio Durigan Junior <sergiodj@redhat.com>
PR python/16699
* cli/cli-decode.c (set_cmd_completer_handle_brkchars): New
function.
(add_cmd): Set "completer_handle_brkchars" to NULL.
* cli/cli-decode.h (struct cmd_list_element)
<completer_handle_brkchars>: New field.
* command.h (completer_ftype_void): New typedef.
(set_cmd_completer_handle_brkchars): New prototype.
* completer.c (set_gdb_completion_word_break_characters): New
function.
(complete_line_internal): Call "completer_handle_brkchars"
callback from command.
* completer.h: Include "command.h".
(set_gdb_completion_word_break_characters): New prototype.
* python/py-cmd.c (cmdpy_completer_helper): New function.
(cmdpy_completer_handle_brkchars): New function.
(cmdpy_completer): Adjust to use cmdpy_completer_helper.
(cmdpy_init): Set completer_handle_brkchars to
cmdpy_completer_handle_brkchars.
gdb/testsuite/ChangeLog:
2014-09-03 Sergio Durigan Junior <sergiodj@redhat.com>
PR python/16699
* gdb.python/py-completion.exp: New file.
* gdb.python/py-completion.py: Likewise.
This commit renames nine files that contain code used by both 32- and
64-bit Intel ports such that their names are prefixed with "x86"
rather than "i386". All types, functions and variables within these
files are likewise renamed such that their names are prefixed with
"x86" rather than "i386". This makes GDB follow the convention used
by gdbserver such that 32-bit Intel code lives in files called
"i386-*", 64-bit Intel code lives in files called "amd64-*", and code
for both 32- and 64-bit Intel lives in files called "x86-*".
This commit only renames OS-independent files. The Linux ports of
both GDB and gdbserver now follow the i386/amd64/x86 convention fully.
Some ports still use the old convention where "i386" in file/function/
type/variable names can mean "32-bit only" or "32- and 64-bit" but I
don't want to touch ports I can't fully test except where absolutely
necessary.
gdb/ChangeLog:
* i386-nat.h: Renamed as...
* x86-nat.h: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
* i386-nat.c: Renamed as...
* x86-nat.c: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
* common/i386-xstate.h: Renamed as...
* common/x86-xstate.h: New file. All type, function and variable
name prefixes changed from "i386_" to "x86_". All references
updated.
* nat/i386-cpuid.h: Renamed as...
* nat/x86-cpuid.h: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
* nat/i386-gcc-cpuid.h: Renamed as...
* nat/x86-gcc-cpuid.h: New file. All type, function and variable
name prefixes changed from "i386_" to "x86_". All references
updated.
* nat/i386-dregs.h: Renamed as...
* nat/x86-dregs.h: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
* nat/i386-dregs.c: Renamed as...
* nat/x86-dregs.c: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
gdb/gdbserver/ChangeLog:
* i386-low.h: Renamed as...
* x86-low.h: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
* i386-low.c: Renamed as...
* x86-low.c: New file. All type, function and variable name
prefixes changed from "i386_" to "x86_". All references updated.
clang was using eax to construct %0 here:
asm ("mov %%eax, 0(%0)\n\t"
"mov %%ebx, 4(%0)\n\t"
"mov %%ecx, 8(%0)\n\t"
"mov %%edx, 12(%0)\n\t"
"mov %%esi, 16(%0)\n\t"
"mov %%edi, 20(%0)\n\t"
: /* no output operands */
: "r" (data)
: "eax", "ebx", "ecx", "edx", "esi", "edi");
which caused amd64-word.exp (and others similarly) to fail.
It's a perfectly legit thing for clang to do given the available data.
The patch fixes this by marking the registers as live from the
time of the preceding breakpoint.
gdb/testsuite/ChangeLog:
* gdb.arch/amd64-pseudo.c (main): Rewrite to better specify when
eax,etc. are live with values set by gdb and thus the compiler can't
use them.
* gdb.arch/i386-pseudo.c (main): Ditto.
I see the following fails on arm-none-eabi target,
print sn^M
$14 = 0x0 <_ftext>^M
(gdb) FAIL: gdb.python/py-value.exp: print sn
print sn^M
$14 = 0x0 <_ftext>^M
(gdb) FAIL: gdb.guile/scm-value.exp: print sn
as <_ftext> is unexpected. This patch is to set print symbol off to
avoid printing this.
gdb/testsuite:
2014-08-24 Yao Qi <yao@codesourcery.com>
* gdb.guile/scm-value.exp (test_lazy_strings): Set print
symbol off.
* gdb.python/py-value.exp (test_lazy_strings): Likewise.
See the description here:
https://sourceware.org/ml/gdb-patches/2014-08/msg00283.html
This patch keeps track of whether the current line has seen a
non-zero discriminator, and if so coalesces consecutive entries
for the same line (by ignoring all entries after the first).
gdb/ChangeLog:
PR 17276
* dwarf2read.c (dwarf_record_line_p): New function.
(dwarf_decode_lines_1): Ignore subsequent line number entries
for the same line if any entry had a non-zero discriminator.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/dw2-single-line-discriminators.S: New file.
* gdb.dwarf2/dw2-single-line-discriminators.c: New file.
* gdb.dwarf2/dw2-single-line-discriminators.exp: New file.
Some gdb.python/*.exp tests fail because the .py files aren't copied
to the (remote) host. This patch is to copy needed .py files to host.
Most of gdb.python/*.exp tests do this.
As it is still controversial to delete *.py files on host, we don't do
that in this patch.
gdb/testsuite:
2014-08-22 Yao Qi <yao@codesourcery.com>
* gdb.python/py-finish-breakpoint.exp: Copy .py file to host.
* gdb.python/py-finish-breakpoint2.exp: Likewise.
* gdb.python/python.exp: Likewise. Use .py file on the host
instead of the build.
Program received signal SIGABRT, Aborted.
[...]
(gdb) gcore foobar
Couldn't get registers: No such process.
(gdb) info threads
[...]
(gdb) gcore foobar
Saved corefile foobar
(gdb)
gcore tries to access the exited thread:
[Thread 0x7ffff7fce700 (LWP 6895) exited]
ptrace(PTRACE_GETREGS, 6895, 0, 0x7fff18167dd0) = -1 ESRCH (No such process)
Without the TRY_CATCH protection testsuite FAILs for:
gcore .../gdb/testsuite/gdb.threads/gcore-thread0.test
Cannot find new threads: debugger service failed
(gdb) FAIL: gdb.threads/gcore-thread.exp: save a zeroed-threads corefile
+
core .../gdb/testsuite/gdb.threads/gcore-thread0.test
".../gdb/testsuite/gdb.threads/gcore-thread0.test" is not a core dump: File format not recognized
(gdb) FAIL: gdb.threads/gcore-thread.exp: core0file: re-load generated corefile (bad file format)
Maybe the TRY_CATCH could be more inside update_thread_list().
Similar update_thread_list() call is IMO missing in procfs_make_note_section()
but I do not have where to verify that change.
gdb/ChangeLog
2014-08-21 Jan Kratochvil <jan.kratochvil@redhat.com>
* linux-tdep.c (linux_corefile_thread_callback): Ignore THREAD_EXITED.
(linux_make_corefile_notes): call update_thread_list, protected against
exceptions.
gdb/testsuite/ChangeLog
2014-08-21 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.threads/gcore-stale-thread.c: New file.
* gdb.threads/gcore-stale-thread.exp: New file.
Checking whether the gcore command is included in the GDB build as
proxy for checking whether core dumping is supported by the target is
useless, as gcore.o has been in COMMON_OBS since git 9b4eba8e:
2009-10-26 Michael Snyder <msnyder@vmware.com>
Hui Zhu <teawater@gmail.com>
* Makefile.in (SFILES): Add gcore.c.
(COMMON_OBS): Add gcore.o.
* config/alpha/alpha-linux.mh (NATDEPFILES): Delete gcore.o.
* config/alpha/fbsd.mh (NATDEPFILES): Ditto.
...
IOW, the command is always included in the build.
Instead, nowadays, tests bail out if actually trying to generate a
core fails with an indication the target doesn't support it. See
gdb_gcore_cmd and callers.
Tested on x86_64 Fedora 20.
gdb/testsuite/ChangeLog:
* gdb.base/gcore-buffer-overflow.exp: Remove "help gcore" test.
* gdb.base/gcore-relro-pie.exp: Likewise.
* gdb.base/gcore-relro.exp: Likewise.
* gdb.base/gcore.exp: Likewise.
* gdb.base/print-symbol-loading.exp: Likewise.
* gdb.threads/gcore-thread.exp: Likewise.
* lib/gdb.exp (gdb_gcore_cmd): Don't expect "Undefined command".
This integrates Jan Kratochvil's nice race reproducer from PR
testsuite/12649 into the testsuite infrustructure directly.
With this, one only has to do either 'make check-read1' or 'make check
READ1="1"' to preload the read1.so library into expect.
Currently only enabled for glibc/GNU systems, and if
build==host==target.
gdb/testsuite/ChangeLog:
* Makefile.in (EXTRA_RULES, CC): New variables, get from
configure.
(EXPECT): Handle READ1 being set.
(all): Depend on EXTRA_RULES.
(check-read1, expect-read1, read1.so, read1): New rules.
* README (Testsuite Parameters): Document the READ1 make variable.
(Race detection): New section.
* configure: Regenerate.
* configure.ac: If build==host==target, and running under a
GNU/glibc system, add read1 to the extra Makefile rules.
(EXTRA_RULES): AC_SUBST it.
* lib/read1.c: New file.
gdb/ChangeLog:
* Makefile.in (check-read1): New rule.
Consider an array described in the debugging information as being
a typedef of an array type for which there is a DW_AT_data_location
attribute. Trying to print the value of that array currently yields
incorrect element values. For instance:
(gdb) print foo.three_tdef
$1 = (6293760, 0, 6293772)
The problem occurs because we check for the data_location attribute
only on the typedef type, whereas we should be checking for the
typedef's target type. As a result, GDB erroneously thinks that
there is no data_location, and therefore starts reading the array's
content from the address of the descriptor instead of the data_location
address.
gdb/ChangeLog:
* value.c (value_from_contents_and_address): Strip resolved_type's
typedef layers before checking its TYPE_DATA_LOCATION.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/data-loc.exp: Add additional tests exercising
the handling of variables declared as a typedef to an array
which a DW_AT_data_location attribute.
This fixes PR symtab/14604, PR symtab/14605, and Jan's test at
https://sourceware.org/ml/gdb-patches/2014-07/msg00158.html, in a tree
with bddbbed reverted:
2014-07-22 Pedro Alves <palves@redhat.com>
* value.c (allocate_optimized_out_value): Don't mark value as
non-lazy.
The PRs are about variables described by the DWARF as being split over
multiple registers using DWARF piece information, but some of those
registers being marked as optimised out (not saved) by a later frame.
GDB currently incorrectly mishandles these partially-optimized-out
values.
Even though we can usually tell from the debug info whether a local or
global is optimized out, handling the case of a local living in a
register that was not saved in a frame requires fetching the variable.
GDB also needs to fetch a value to tell whether parts of it are
"<unavailable>". Given this, it's not worth it to try to avoid
fetching lazy optimized-out values based on debug info alone.
So this patch makes GDB track which chunks of a value's contents are
optimized out like it tracks <unavailable> contents. That is, it
makes value->optimized_out be a bit range vector instead of a boolean,
and removes the struct lval_funcs check_validity and check_any_valid
hooks.
Unlike Andrew's series which this is based on (at
https://sourceware.org/ml/gdb-patches/2013-08/msg00300.html, note some
pieces have gone in since), this doesn't merge optimized out and
unavailable contents validity/availability behind a single interface,
nor does it merge the bit range vectors themselves (at least yet).
While it may be desirable to have a single entry point that returns
existence of contents irrespective of what may make them
invalid/unavailable, several places want to treat optimized out /
unavailable / etc. differently, so each spot that potentially could
use it will need to be careful considered on case-by-case basis, and
best done as a separate change.
This fixes Jan's test, because value_available_contents_eq wasn't
considering optimized out value contents. It does now, and because of
that it's been renamed to value_contents_eq.
A new intro comment is added to value.h describing "<optimized out>",
"<not saved>" and "<unavailable>" values.
gdb/
PR symtab/14604
PR symtab/14605
* ada-lang.c (coerce_unspec_val_to_type): Use
value_contents_copy_raw.
* ada-valprint.c (val_print_packed_array_elements): Adjust.
* c-valprint.c (c_val_print): Use value_bits_any_optimized_out.
* cp-valprint.c (cp_print_value_fields): Let the common printing
code handle optimized out values.
(cp_print_value_fields_rtti): Use value_bits_any_optimized_out.
* d-valprint.c (dynamic_array_type): Use
value_bits_any_optimized_out.
* dwarf2loc.c (entry_data_value_funcs): Remove check_validity and
check_any_valid fields.
(check_pieced_value_bits): Delete and inline ...
(check_pieced_synthetic_pointer): ... here.
(check_pieced_value_validity): Delete.
(check_pieced_value_invalid): Delete.
(pieced_value_funcs): Remove check_validity and check_any_valid
fields.
(read_pieced_value): Use mark_value_bits_optimized_out.
(write_pieced_value): Switch to use
mark_value_bytes_optimized_out.
(dwarf2_evaluate_loc_desc_full): Copy the value contents instead
of assuming the whole value is optimized out.
* findvar.c (read_frame_register_value): Remove special handling
of optimized out registers.
(value_from_register): Use mark_value_bytes_optimized_out.
* frame-unwind.c (frame_unwind_got_optimized): Use
mark_value_bytes_optimized_out.
* jv-valprint.c (java_value_print): Adjust.
(java_print_value_fields): Let the common printing code handle
optimized out values.
* mips-tdep.c (mips_print_register): Remove special handling of
optimized out registers.
* opencl-lang.c (lval_func_check_validity): Delete.
(lval_func_check_any_valid): Delete.
(opencl_value_funcs): Remove check_validity and check_any_valid
fields.
* p-valprint.c (pascal_object_print_value_fields): Let the common
printing code handle optimized out values.
* stack.c (read_frame_arg): Remove special handling of optimized
out values. Fetch both VAL and ENTRYVAL before comparing
contents. Adjust to value_available_contents_eq rename.
* valprint.c (valprint_check_validity)
(val_print_scalar_formatted): Use value_bits_any_optimized_out.
(val_print_array_elements): Adjust.
* value.c (struct value) <optimized_out>: Now a VEC(range_s).
(value_bits_any_optimized_out): New function.
(value_entirely_covered_by_range_vector): New function, factored
out from value_entirely_unavailable.
(value_entirely_unavailable): Reimplement.
(value_entirely_optimized_out): New function.
(insert_into_bit_range_vector): New function, factored out from
mark_value_bits_unavailable.
(mark_value_bits_unavailable): Reimplement.
(struct ranges_and_idx): New struct.
(find_first_range_overlap_and_match): New function, factored out
from value_available_contents_bits_eq.
(value_available_contents_bits_eq): Rename to ...
(value_contents_bits_eq): ... this. Check both unavailable
contents and optimized out contents.
(value_available_contents_eq): Rename to ...
(value_contents_eq): ... this.
(allocate_value_lazy): Remove reference to the old optimized_out
boolean.
(allocate_optimized_out_value): Use
mark_value_bytes_optimized_out.
(require_not_optimized_out): Adjust to check whether the
optimized_out vec is empty.
(ranges_copy_adjusted): New function, factored out from
value_contents_copy_raw.
(value_contents_copy_raw): Also copy the optimized out ranges.
Assert the destination ranges aren't optimized out.
(value_contents_copy): Update comment, remove call to
require_not_optimized_out.
(value_contents_equal): Adjust to check whether the optimized_out
vec is empty.
(set_value_optimized_out, value_optimized_out_const): Delete.
(mark_value_bytes_optimized_out, mark_value_bits_optimized_out):
New functions.
(value_entirely_optimized_out, value_bits_valid): Delete.
(value_copy): Take a VEC copy of the 'optimized_out' field.
(value_primitive_field): Remove special handling of optimized out.
(value_fetch_lazy): Assert that lazy values have no unavailable
regions. Use value_bits_any_optimized_out. Remove some special
handling for optimized out values.
* value.h: Add intro comment about <optimized out> and
<unavailable>.
(struct lval_funcs): Remove check_validity and check_any_valid
fields.
(set_value_optimized_out, value_optimized_out_const): Remove.
(mark_value_bytes_optimized_out, mark_value_bits_optimized_out):
New declarations.
(value_bits_any_optimized_out): New declaration.
(value_bits_valid): Delete declaration.
(value_available_contents_eq): Rename to ...
(value_contents_eq): ... this, and extend comments.
gdb/testsuite/
PR symtab/14604
PR symtab/14605
* gdb.dwarf2/dw2-op-out-param.exp: Remove kfail branches and use
gdb_test.
This comment is no longer true for watchpoints since commit 31e77af2
(PR breakpoints/7143 - Watchpoint does not trigger when first set).
gdb/testsuite/
* gdb.base/watchpoint-hw-hit-once.c (main): Update comment.
In gdb.base/watchpoint-hw-hit-once.exp, test scans source and set
breakpoint on the line having "break-at-exit",
gdb_breakpoint [gdb_get_line_number "break-at-exit"]
However, in watchpoint-hw-hit-once.c, there are two lines having
this key word:
dummy = 1; /* Stub to catch break-at-exit after WATCHEE has been hit. */
dummy = 2; /* break-at-exit */
so the test sets breakpoint on the first one, while I think it is
expected to set breakpoint on the second one, as far as I can tell
from the comments in watchpoint-hw-hit-once.c:
/* Stub lines are present as no breakpoints/watchpoint gets hit if current PC
already stays on the line PC while entering "step"/"continue". */
This patch is to change the source matching pattern so that test
can correctly set breakpoint on the right line. This patch fixes
a fail we found on arm-none-eabi target.
(gdb) PASS: gdb.base/watchpoint-hw-hit-once.exp: continue
continue^M
Continuing.^M
^M
*** EXIT code 0^M
[Inferior 1 (Remote target) exited normally]^M
(gdb) FAIL: gdb.base/watchpoint-hw-hit-once.exp: continue to break-at-exit (the program exited)
Run it again on x86_64-linux, no result changes.
gdb/testsuite:
2014-08-19 Yao Qi <yao@codesourcery.com>
* gdb.base/watchpoint-hw-hit-once.exp: Set breakpoint on the
right line.
The testcase generates an assembly file where a second DW_AT_upper_bound
attribute gets generated in the array range. This was definitely
unintentional, and I only noticed this after pushing the testcase,
when dumping one more time the DWARF data using readelf.
This patch fixes it.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/data-loc.exp: Remove second DW_AT_upper bound
attribute in array range.
This testcase allows us to test the proper processing of both
DW_AT_data_location and DW_OP_push_object_address using a hand-crafted
testcase duplicating how we expect the Ada compiler to represent
unbounded arrays.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/data-loc.c, gdb.dwarf2/data-loc.exp: New files.
This is a fix for PR c++/17132.
If this new argument is set to EVAL_AVOID_SIDE_EFFECTS, then the object's
memory will not be read while picking the best overload match.
gdb/
* eval.c: Update all calls to find_overload_match.
* valarith.c: Likewise.
(value_user_defined_cpp_op, value_user_defined_op): New
argument NOSIDE. Update all callers.
* valops.c (find_overload_match): New argument NOSIDE.
* value.h (find_overload_match): Update signature.
gdb/testsuite
* gdb.cp/pr17132.cc: New file.
* gdb.cp/pr17132.exp: New file.
gdb/
* python/lib/gdb/command/xmethods.py (set_xm_status1): Use the
'items' methods instead of 'iteritems' method on dictionaries.
gdb/testsuite/
* gdb.python/py-xmethods.py (A_getarrayind)
(E_method_char_worker.__call__, E_method_int_worker.__call__):
Use 'print' with function call syntax.
(E_method_matcher.match): Fix tab vs space indentation mixup.
I added proc generate_tracefile in this patch
https://sourceware.org/ml/gdb-patches/2014-03/msg00591.html but
tfile.exp isn't skipped as changelog entry said:
* gdb.trace/tfile.exp: Skip the test if generate_tracefile
return 0.
it is a mistake I made at the last minute. Patch below fixed it.
gdb/testsuite:
2014-08-15 Yao Qi <yao@codesourcery.com>
* gdb.trace/tfile.exp: Return -1 if generate_tracefile returns
false.
GDB in default prints the symbol associated on an address, and tests
assume that there is no symbol on address zero. However, on bare
metal target, address may be mapped to zero and there may be a
symbol. Then, some tests fail as below:
print const_cast<void *> (0)^M
$8 = (void *) 0x0 <_ftext>^M
(gdb) FAIL: gdb.cp/casts.exp: const_cast of 0
p acp->c1^M
$9 = (A *) 0x0 <_ftext>^M
(gdb) FAIL: gdb.cp/class2.exp: p acp->c1
This patch is to set print symbol off in these tests, like what
I did previously https://sourceware.org/ml/gdb-patches/2014-07/msg00257.html
gdb/testsuite:
2014-08-15 Yao Qi <yao@codesourcery.com>
* gdb.cp/casts.exp: Set print symbol off.
* gdb.cp/class2.exp: Likewise.
* gdb.cp/overload.exp: Likewise.
* gdb.cp/templates.exp: Likewise.
* solib.c (solib_read_symbols): Delete "Loaded symbols for ..."
message, it is redundant with "Reading symbols from ..." message.
testsuite/
* gdb.base/print-symbol-loading.exp (test_load_core): Update.
(test_load_shlib): Update.
I find some gdb.python tests fail on arm-none-eabi target, because the
tests assume that memory on address 0x is inaccessible. Some tests
(in gdb.base) are aware of this, so do a "x 0" check first. However,
the code is copy-n-paste.
This patch is to move the "x 0" check to a procedure in lib/gdb.exp,
and get needed tests call it. The original code matches pattern
"0x0:\[ \t\]*Error accessing memory address 0x0\r\n$gdb_prompt $", but
I remove it from the new proc is_address_zero_readable, because GDB
doesn't emit such message any more.
gdb/testsuite:
2014-08-09 Yao Qi <yao@codesourcery.com>
* gdb.base/display.exp: Invoke is_address_zero_readable.
* gdb.guile/scm-value.exp (test_value_in_inferior): Likewise.
* gdb.python/py-value.exp (test_value_in_inferior): Likewise.
* gdb.base/hbreak-unmapped.exp: Return if
is_address_zero_readable returns true.
* gdb.base/signest.exp: Likewise.
* gdb.base/signull.exp: Likewise.
* gdb.base/sigbpt.exp: Likewise.
* gdb.guile/scm-disasm.exp: Do the test if
is_address_zero_readable returns false.
* gdb.guile/scm-pretty-print.exp (run_lang_tests): Likewise.
* gdb.python/py-arch.exp: Likewise.
* gdb.python/py-prettyprint.exp (run_lang_tests): Likewise.
* lib/gdb.exp (is_address_zero_readable): New proc.
When I fix a bug in gdb.mi/mi-var-display.exp, I find its test
messages aren't unique, which makes some confusions for me.
$ cat testsuite/gdb.sum | grep "PASS" | sort | uniq -c | sort -n
...
2 PASS: gdb.mi/mi-var-display.exp: set format variable bar
2 PASS: gdb.mi/mi-var-display.exp: set format variable foo
2 PASS: gdb.mi/mi-var-display.exp: set format variable weird.func_ptr
2 PASS: gdb.mi/mi-var-display.exp: set format variable weird.func_ptr_ptr
2 PASS: gdb.mi/mi-var-display.exp: show format variable foo
3 PASS: gdb.mi/mi-var-display.exp: eval variable foo
This patch is to make test messages in mi-var-display.exp unique.
gdb/testsuite:
2014-08-09 Yao Qi <yao@codesourcery.com>
PR testsuite/13443
* gdb.mi/mi-var-display.exp: Make test messages unique.
Right now, "set debug target" acts a bit strangely.
Most target APIs only notice that it has changed when the target stack
is changed in some way. This is because many methods implement the
setting using the special debug target. However, a few spots do
change their behavior immediately -- any place explicitly checking
"targetdebug".
Some of this peculiar behavior is documented. However, I think that
it just isn't very useful for it to work this way. So, this patch
changes "set debug target" to take effect immediately in all cases.
This is done by simply calling update_current_target when the setting
is changed.
This required one small change in the test suite. Here a test was
expecting the current behavior.
Built and regtested on x86-64 Fedora 20.
2014-08-04 Tom Tromey <tromey@redhat.com>
* target.c (set_targetdebug): New function.
(initialize_targets): Pass set_targetdebug when creating "set
debug target".
2014-08-04 Tom Tromey <tromey@redhat.com>
* gdb.texinfo (Debugging Output): Update for change to "set debug
target".
2014-08-04 Tom Tromey <tromey@redhat.com>
* gdb.base/sss-bp-on-user-bp-2.exp: Expect output from "set debug
target 0".
This fixes a test suite regession that Yao noticed.
This test checks for some specific "target debug" output
that has changed since the test was written.
2014-08-04 Tom Tromey <tromey@redhat.com>
* gdb.base/sss-bp-on-user-bp-2.exp: Match "to_resume", not
"target_resume".
In Ada, variable-sized field can be located at any position of
a structure. Consider for instance the following declarations:
Dyn_Size : Integer := 1;
type Table is array (Positive range <>) of Integer;
type Inner is record
T1 : Table (1 .. Dyn_Size) := (others => 1);
T2 : Table (1 .. Dyn_Size) := (others => 2);
end record;
type Inner_Array is array (1 .. 2) of Inner;
type Outer is
record
I0 : Integer := 0;
A1 : Inner_Array;
Marker : Integer := 16#01020304#;
end record;
Rt : Outer;
What this does is declare a variable "Rt" of type Outer, which
contains 3 fields where the second (A1) is of type Inner_Array.
type Inner_Array is an array with 2 elements of type Inner.
Because type Inner contains two arrays whose upper bound depend
on a variable, the size of the array, and therefore the size of
type Inner is dynamic, thus making field A1 a dynamically-size
field.
When trying to print the value of Rt, we hit the following limitation:
(gdb) print rt
Attempt to resolve a variably-sized type which appears in the interior of
a structure type
The limitation was somewhat making sense in C, but needs to be lifted
for Ada. This patch mostly lifts that limitation. As a result of this
patch, the type length computation had to be reworked a little bit.
gdb/ChangeLog:
* gdbtypes.c (resolve_dynamic_struct): Do not generate an error
if detecting a variable-sized field that is not the last field.
Fix struct type length computation.
gdb/testsuite/ChangeLog:
* gdb.base/vla-datatypes.c (vla_factory): Add new variable
inner_vla_struct_object_size.
* gdb.base/vla-datatypes.exp: Adjust last test, and mark it
as xfail.
As reported in PR 17206, an internal error is triggered when command
until is executed. In infcmd.c:until_next_command, step_range_end is
set to 'pc',
if (!func)
{
struct bound_minimal_symbol msymbol = lookup_minimal_symbol_by_pc (pc);
if (msymbol.minsym == NULL)
error (_("Execution is not within a known function."));
tp->control.step_range_start = BMSYMBOL_VALUE_ADDRESS (msymbol);
tp->control.step_range_end = pc;
}
and later in infrun.c:resume, the assert below is triggered in PR
17206.
if (tp->control.may_range_step)
{
/* If we're resuming a thread with the PC out of the step
range, then we're doing some nested/finer run control
operation, like stepping the thread out of the dynamic
linker or the displaced stepping scratch pad. We
shouldn't have allowed a range step then. */
gdb_assert (pc_in_thread_step_range (pc, tp));
}
In until_next_command, we set step range to [XXX, pc), so pc isn't
within the range. pc_in_thread_step_range returns false and the
assert is triggered. AFAICS, the range we want in until_next_command
is [XXX, pc] instead of [XXX, pc), because we want to program step
until greater than pc. This patch is to set step_range_end to
'pc + 1'. Running until-nodebug.exp with unpatched GDB will get the
following fail,
FAIL: gdb.base/until-nodebug.exp: until 2 (GDB internal error)
and the fail goes away when the fix is applied.
gdb:
2014-07-29 Yao Qi <yao@codesourcery.com>
PR gdb/17206
* infcmd.c (until_next_command): Set step_range_end to PC + 1.
gdb/testsuite:
2014-07-29 Yao Qi <yao@codesourcery.com>
PR gdb/17206
* gdb.base/until-nodebug.exp: New.
* guile/scm-param.c (pascm_parameter_defined_p): New function.
(gdbscm_register_parameter_x): Call it. Raise error for pre-existing
parameters.
testsuite/
* gdb.guile/scm-parameter.exp: Add tests for trying to create
previously existing parameter, and previously ambiguously spelled
parameter.
These tests used to fail on ARM but now pass, so remove the KFAIL.
gdb/testsuite/ChangeLog:
2014-07-28 Will Newton <will.newton@linaro.org>
* gdb.base/varargs.exp: Remove KFAILs for ARM.
* data-directory/Makefile.in (GUILE_FILES): Add support.scm.
* guile/lib/gdb/support.scm: New file.
* guile/guile.c (gdbscm_init_module_name): Change to "gdb".
* guile/lib/gdb.scm: Load gdb/init.scm as an include file.
All uses updated.
* guile/lib/gdb/init.scm (SCM_ARG1, SCM_ARG2): Moved to support.scm.
All uses updated.
(%assert-type): Ditto, and renamed to assert-type.
(%exception-print-style): Delete.
testsuite/
* gdb.guile/types-module.exp: Add tests for wrong type arguments.
Currently, GDB can pass a signal to the wrong thread in several
different but related scenarios.
E.g., if thread 1 stops for signal SIGFOO, the user switches to thread
2, and then issues "continue", SIGFOO is actually delivered to thread
2, not thread 1. This obviously messes up programs that use
pthread_kill to send signals to specific threads.
This has been a known issue for a long while. Back in 2008 when I
made stop_signal be per-thread (2020b7ab), I kept the behavior -- see
code in 'proceed' being removed -- wanting to come back to it later.
The time has finally come now.
The patch fixes this -- on resumption, intercepted signals are always
delivered to the thread that had intercepted them.
Another example: if thread 1 stops for a breakpoint, the user switches
to thread 2, and then issues "signal SIGFOO", SIGFOO is actually
delivered to thread 1, not thread 2, because 'proceed' first switches
to thread 1 to step over its breakpoint... If the user deletes the
breakpoint before issuing "signal FOO", then the signal is delivered
to thread 2 (the current thread).
"signal SIGFOO" can be used for two things: inject a signal in the
program while the program/thread had stopped for none, bypassing
"handle nopass"; or changing/suppressing a signal the program had
stopped for. These scenarios are really two faces of the same coin,
and GDB can't really guess what the user is trying to do. GDB might
have intercepted signals in more than one thread even (see the new
signal-command-multiple-signals-pending.exp test). At least in the
inject case, it's obviously clear to me that the user means to deliver
the signal to the currently selected thread, so best is to make the
command's behavior consistent and easy to explain.
Then, if the user is trying to suppress/change a signal the program
had stopped for instead of injecting a new signal, but, the user had
changed threads meanwhile, then she will be surprised that with:
(gdb) continue
Thread 1 stopped for signal SIGFOO.
(gdb) thread 2
(gdb) signal SIGBAR
... GDB actually delivers SIGFOO to thread 1, and SIGBAR to thread 2
(with scheduler-locking off, which is the default, because then
"signal" or any other resumption command resumes all threads).
So the patch makes GDB detect that, and ask for confirmation:
(gdb) thread 1
[Switching to thread 1 (Thread 10979)]
(gdb) signal SIGUSR2
Note:
Thread 3 previously stopped with signal SIGUSR2, User defined signal 2.
Thread 2 previously stopped with signal SIGUSR1, User defined signal 1.
Continuing thread 1 (the current thread) with specified signal will
still deliver the signals noted above to their respective threads.
Continue anyway? (y or n)
All these scenarios are covered by the new tests.
Tested on x86_64 Fedora 20, native and gdbserver.
gdb/
2014-07-25 Pedro Alves <palves@redhat.com>
* NEWS: Mention signal passing and "signal" command changes.
* gdbthread.h (struct thread_suspend_state) <stop_signal>: Extend
comment.
* breakpoint.c (until_break_command): Adjust clear_proceed_status
call.
* infcall.c (run_inferior_call): Adjust clear_proceed_status call.
* infcmd.c (proceed_thread_callback, continue_1, step_once)
(jump_command): Adjust clear_proceed_status call.
(signal_command): Warn if other thread that are resumed have
signals that will be delivered. Adjust clear_proceed_status call.
(until_next_command, finish_command)
(proceed_after_attach_callback, attach_command_post_wait)
(attach_command): Adjust clear_proceed_status call.
* infrun.c (proceed_after_vfork_done): Likewise.
(proceed_after_attach_callback): Adjust comment.
(clear_proceed_status_thread): Clear stop_signal if not in pass
state.
(clear_proceed_status_callback): Delete.
(clear_proceed_status): New 'step' parameter. Only clear the
proceed status of threads the command being prepared is about to
resume.
(proceed): If passed in an explicit signal, override stop_signal
with it. Don't pass the last stop signal to the thread we're
resuming.
(init_wait_for_inferior): Adjust clear_proceed_status call.
(switch_back_to_stepped_thread): Clear the signal if it should not
be passed.
* infrun.h (clear_proceed_status): New 'step' parameter.
(user_visible_resume_ptid): Add comment.
* linux-nat.c (linux_nat_resume_callback): Don't check whether the
signal is in pass state.
* remote.c (append_pending_thread_resumptions): Likewise.
* mi/mi-main.c (proceed_thread): Adjust clear_proceed_status call.
gdb/doc/
2014-07-25 Pedro Alves <palves@redhat.com>
Eli Zaretskii <eliz@gnu.org>
* gdb.texinfo (Signaling) <signal command>: Explain what happens
with multi-threaded programs.
gdb/testsuite/
2014-07-25 Pedro Alves <palves@redhat.com>
* gdb.threads/signal-command-handle-nopass.c: New file.
* gdb.threads/signal-command-handle-nopass.exp: New file.
* gdb.threads/signal-command-multiple-signals-pending.c: New file.
* gdb.threads/signal-command-multiple-signals-pending.exp: New file.
* gdb.threads/signal-delivered-right-thread.c: New file.
* gdb.threads/signal-delivered-right-thread.exp: New file.
Jan pointed out in
<https://sourceware.org/ml/gdb-patches/2014-07/msg00553.html> that
these testcases have racy results:
gdb.base/double-prompt-target-event-error.exp
gdb.base/paginate-after-ctrl-c-running.exp
gdb.base/paginate-bg-execution.exp
gdb.base/paginate-execution-startup.exp
gdb.base/paginate-inferior-exit.exp
This is easily reproducible with "read1" from:
[reproducer for races of expect incomplete reads]
http://sourceware.org/bugzilla/show_bug.cgi?id=12649
The '-notransfer -re "<return>" { exp_continue }' trick in the current
tests doesn't actually work.
The issue that led to the -notransfer trick was that
"---Type <return> to continue, or q <return> to quit---"
has two "<return>"s. If one wants gdb_test_multiple to not hit the
built-in "<return>" match that results in FAIL, one has to expect the
pagination prompt in chunks, first up to the first "<return>", then
again, up to the second. Something around these lines:
gdb_test_multiple "" $test {
-re "<return>" {
exp_continue
}
-re "to quit ---" {
pass $test
}
}
The intent was for -notransfer+exp_continue to make expect fetch more
input, and rerun the matches against the now potentially fuller
buffer, and then eventually the -re that includes the full pagination
prompt regex would match instead (because it's listed higher up, it
would match first). But, once that "<return>" -notransfer -re
matches, it keeps re-matching forever. It seems like with
exp_continue, expect immediately retries matching, instead of first
reading in more data into the buffer, if available.
Fix this like I should have done in the first place. There's actually
no good reason for gdb_test_multiple to only match "<return>". We can
make gdb_test_multiple expect the whole pagination prompt text
instead, which is store in the 'pagination_prompt' global (similar to
'gdb_prompt'). Then a gdb_test_multiple caller that doesn't want the
default match to trigger, because it wants to see one pagination
prompt, does simply:
gdb_test_multiple "" $test {
-re "$pagination_prompt$" {
pass $test
}
}
which is just like when we don't want the default $gdb_prompt match
within gdb_test_multiple to trigger, like:
gdb_test_multiple "" $test {
-re "$gdb_prompt $" {
pass $test
}
}
Tested on x86_64 Fedora 20. In addition, I've let the racy tests run
all in parallel in a loop for 30 minutes, and they never failed.
gdb/testsuite/
2014-07-25 Pedro Alves <palves@redhat.com>
* gdb.base/double-prompt-target-event-error.exp
(cancel_pagination_in_target_event): Remove '-notransfer <return>'
match.
(cancel_pagination_in_target_event): Rework double prompt
detection.
* gdb.base/paginate-after-ctrl-c-running.exp
(test_ctrlc_while_target_running_paginates): Remove '-notransfer
<return>' match.
* gdb.base/paginate-bg-execution.exp
(test_bg_execution_pagination_return)
(test_bg_execution_pagination_cancel): Remove '-notransfer
<return>' matches.
* gdb.base/paginate-execution-startup.exp
(test_fg_execution_pagination_return)
(test_fg_execution_pagination_cancel): Remove '-notransfer
<return>' matches.
* gdb.base/paginate-inferior-exit.exp
(test_paginate_inferior_exited): Remove '-notransfer <return>'
match.
* lib/gdb-utils.exp (string_to_regexp): Move here from lib/gdb.exp.
* lib/gdb.exp (pagination_prompt): Run text through
string_to_regexp.
(gdb_test_multiple): Match $pagination_prompt instead of
"<return>".
(string_to_regexp): Move to lib/gdb-utils.exp.
I noticed that the existing code casts a function's address to 'long',
but that doesn't work correctly on some ABIs, like Win64, where long
is 32-bit and while pointers are 64-bit:
func_addr = (long) &write_basic_trace_file;
Fixing that showed there's actually another place in the file that
writes a function address to file, and therefore should clear the
Thumb bit. This commit adds a macro+function pair to centralize the
Thumb bit handling, and uses it in both places.
The rest is just enough changes to make the file build without
warnings with "-Wall -Wextra" with x86_64-w64-mingw32-gcc and
i686-w64-mingw32-gcc cross compilers, and with -m32/-m64 on x86_64
GNU/Linux. Currently with x86_64-w64-mingw32-gcc we get:
$ x86_64-w64-mingw32-gcc tfile.c -Wall -DTFILE_DIR=\"\"
tfile.c: In function 'start_trace_file':
tfile.c:51:23: error: 'S_IRGRP' undeclared (first use in this function)
S_IRUSR|S_IWUSR|S_IRGRP|S_IROTH);
^
tfile.c:51:23: note: each undeclared identifier is reported only once for each function it appears in
tfile.c:51:31: error: 'S_IROTH' undeclared (first use in this function)
S_IRUSR|S_IWUSR|S_IRGRP|S_IROTH);
^
tfile.c: In function 'add_memory_block':
tfile.c:79:10: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
ll_x = (unsigned long) addr;
^
tfile.c: In function 'write_basic_trace_file':
tfile.c:113:15: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
func_addr = (long) &write_basic_trace_file;
^
tfile.c:137:3: warning: passing argument 1 of 'add_memory_block' from incompatible pointer type [enabled by default]
add_memory_block (&testglob, sizeof (testglob));
^
tfile.c:72:1: note: expected 'char *' but argument is of type 'int *'
add_memory_block (char *addr, int size)
^
tfile.c:139:3: warning: passing argument 1 of 'add_memory_block' from incompatible pointer type [enabled by default]
add_memory_block (&testglob2, 1);
^
tfile.c:72:1: note: expected 'char *' but argument is of type 'int *'
add_memory_block (char *addr, int size)
^
tfile.c: In function 'write_error_trace_file':
tfile.c:185:3: warning: implicit declaration of function 'alloca' [-Wimplicit-function-declaration]
char *hex = alloca (len * 2 + 1);
^
tfile.c:185:15: warning: incompatible implicit declaration of built-in function 'alloca' [enabled by default]
char *hex = alloca (len * 2 + 1);
^
tfile.c:211:6: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
(long) &write_basic_trace_file);
^
Tested on x86_64 Fedora 20, -m64 and -m32.
Tested by Yao on arm targets.
gdb/testsuite/
2014-07-16 Pedro Alves <palves@redhat.com>
* gdb.trace/tfile.c: Include unistd.h and stdint.h.
(start_trace_file): Guard S_IRGRP and S_IROTH uses behind #ifdef.
(tfile_write_64, tfile_write_16, tfile_write_8, tfile_write_addr)
(tfile_write_buf): New functions.
(add_memory_block): Rewrite using the above.
(adjust_function_address): New function.
(FUNCTION_ADDRESS): New macro.
(write_basic_trace_file): Remove short_x local, and use
tfile_write_16. Change type of func_addr local to unsigned long
long. Use FUNCTION_ADDRESS instead of handling the Thumb bit
here. Cast argument of add_memory_block to char pointer.
(write_error_trace_file): Avoid alloca. Use FUNCTION_ADDRESS.
(main): Remove parameters.
* gdb.trace/tfile.exp: Remove nowarnings.
As Joel pointed out in...
https://sourceware.org/ml/gdb-patches/2014-07/msg00391.html
...it would be nice to add a test for that.
Tested on Linux x86_64 (Ubuntu 14.10).
gdb/testsuite/ChangeLog
2014-07-15 Simon Marchi <simon.marchi@ericsson.com>
* gdb.base/debug-expr.exp: Test string evaluation with
"debug expression" on.
The other day I noticed that default_gdb_start reuses the GDB process
if it has been spawned already:
proc default_gdb_start { } {
...
if [info exists gdb_spawn_id] {
return 0
}
I was a bit surprised, and so I hacked in an error to check whether
anything is relying on it:
+ if [info exists gdb_spawn_id] {
+ error "GDB already spawned"
+ }
And lo, that tripped on a funny buglet (see below). The comment in
reread.exp says "Restart GDB entirely", but in reality, due to the
above, that's not what is happening, as a gdb_exit call is missing.
The test is proceeding with the previous GDB process...
I don't really want to go hunt for whether there's an odd setup out
there that assumes this in its board file or something, so for now,
I'm taking the simple route of just making the test do what it says it
does. I think this much makes it an obvious fix.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
(gdb) PASS: gdb.base/reread.exp: run to foo() second time
ERROR: tcl error sourcing ../src/gdb/testsuite/gdb.base/reread.exp.
ERROR: GDB already spawned
while executing
"error "GDB already spawned""
invoked from within
"if [info exists gdb_spawn_id] {
error "GDB already spawned"
}"
(procedure "default_gdb_start" line 22)
invoked from within
"default_gdb_start"
(procedure "gdb_start" line 2)
invoked from within
"gdb_start"
invoked from within
"if [is_remote target] {
unsupported "second pass: GDB should check for changes before running"
} else {
# Put the older executable back in pl..."
(file "../src/gdb/testsuite/gdb.base/reread.exp" line 114)
invoked from within
"source ../src/gdb/testsuite/gdb.base/reread.exp"
("uplevel" body line 1)
invoked from within
"uplevel #0 source ../src/gdb/testsuite/gdb.base/reread.exp"
invoked from within
"catch "uplevel #0 source $test_file_name""
testcase ../src/gdb/testsuite/gdb.base/reread.exp completed in 1 seconds
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
gdb/testsuite/
2014-07-15 Pedro Alves <palves@redhat.com>
* gdb.base/reread.exp: Use clean_restart.
The __flash qualifier is part of the named address spaces for AVR [1]. It
allows putting read-only data in the flash memory, normally reserved for
code.
When used together with a pointer, the DW_AT_address_class attribute is set
to 1 and allows GDB to detect that when it will be dereferenced, the data
will be loaded from the flash memory (with the LPM instruction).
We can now properly debug the following code:
~~~
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
~~~
Whereas previously, GDB would revert to the default address space which is
RAM and mapped in higher memory:
~~~
(gdb) print pointer_to_flash
$1 = 0x8001e8 ""
~~~
[1] https://gcc.gnu.org/onlinedocs/gcc/Named-Address-Spaces.html
2014-07-15 Pierre Langlois <pierre.langlois@embecosm.com>
gdb/
* avr-tdep.c (AVR_TYPE_ADDRESS_CLASS_FLASH): New macro.
(AVR_TYPE_INSTANCE_FLAG_ADDRESS_CLASS_FLASH): Likewise.
(avr_address_to_pointer): Check for AVR_TYPE_ADDRESS_CLASS_FLASH.
(avr_pointer_to_address): Likewise.
(avr_address_class_type_flags): New function.
(avr_address_class_type_flags_to_name): Likewise.
(avr_address_class_name_to_type_flags): Likewise.
(avr_gdbarch_init): Set address_class_type_flags,
address_class_type_flags_to_name and
address_class_name_to_type_flags.
gdb/testsuite/
* gdb.arch/avr-flash-qualifer.c: New.
* gdb.arch/avr-flash-qualifer.exp: New.
Put GDB's terminal settings into effect when paginating
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
* utils.c (prompt_for_continue): Call target_terminal_ours.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
* gdb.base/paginate-after-ctrl-c-running.c: New file.
* gdb.base/paginate-after-ctrl-c-running.exp: New file.
When the target is resumed in the foreground, we put the inferior's
terminal settings into effect, and remove stdin from the event loop.
When the target stops, we put GDB's terminal settings into effect
again, and re-register stdin in the event loop, ready for user input.
The former is done by target_terminal_inferior, and the latter by
target_terminal_ours.
There's an intermediate -- target_terminal_ours_for_output -- that is
called when printing output related to target events, and we don't
know yet whether we'll stop the program. That puts our terminal
settings into effect, enough to get proper results from our output,
but leaves input wired into the inferior.
If such output paginates, then we need the full target_terminal_ours
in order for the user to be able to provide input to answer the
pagination query.
The test in this commit hangs in async-capable targets without the fix
(as the user/test can't answer the pagination query). It doesn't hang
on sync targets because on those we don't unregister stdin from the
event loop while the target is running (because we block in
target_wait instead of in the event loop in that case).
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
* utils.c (prompt_for_continue): Call target_terminal_ours.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
* gdb.base/paginate-after-ctrl-c-running.c: New file.
* gdb.base/paginate-after-ctrl-c-running.exp: New file.
If an error is thrown while handling a target event (within
fetch_inferior_event), and, the interpreter is not async (but the
target is), then GDB prints the prompt twice.
One way to see that in action is throw a QUIT while in a pagination
prompt issued from within fetch_inferior_event (or one of its
callees). E.g. from the test:
---Type <return> to continue, or q <return> to quit---
^CQuit
(gdb) (gdb) p 1
^^^^^^^^^^^
$1 = 1
(gdb)
The issue is that inferior_event_handler swallows errors and notifies
the observers (the interpreters) about the command error, even if the
interpreter is forced sync while we're handling a nested event loop
(for execute_command). The observers print a prompt, and then when we
get back to the top event loop, we print another (in
start_event_loop).
I see no reason the error should be swallowed here. Just cancel the
execution related bits and let the error propagate to the top level
(start_event_loop), which re-enables stdin and notifies observers.
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
* inf-loop.c (inferior_event_handler): Use TRY_CATCH instead of
catch_errors. Don't re-enable stdin or notify observers where,
and rethrow error.
(fetch_inferior_event_wrapper): Delete.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
* gdb.base/double-prompt-target-event-error.c: New file.
* gdb.base/double-prompt-target-event-error.exp: New file.
If a pagination prompt triggers while the target is running, and the
target exits before the user responded to the pagination query, this
happens:
Starting program: foo
---Type <return> to continue, or q <return> to quit---No unwaited-for children left.
Couldn't get registers: No such process.
Couldn't get registers: No such process.
Couldn't get registers: No such process.
(gdb) Couldn't get registers: No such process.
(gdb)
To reiterate, the user hasn't replied to the pagination prompt above.
A pagination query nests an event loop (in gdb_readline_wrapper). In
async mode, in addition to stdin and signal handlers, we'll have the
target also installed in the event loop still. So if the target
reports an event, that wakes up the nested event loop, which calls
into fetch_inferior_event etc. to handle the event which generates
further output, all while we should be waiting for pagination
confirmation...
(TBC, any target event that generates output ends up spuriously waking
up the pagination, though exits seem to be the worse kind.)
I've played with a couple different approaches to fixing this, while
at the same time trying to avoid being invasive. Both revolve around
not listening to target events while in a pagination prompt (doing
anything else I think would be a much bigger change).
The approach taken just removes the target from the event loop while
within gdb_readline_wrapper. The other approach used gdb_select
directly, with only input_fd installed, but that had the issue that it
didn't handle the async signal handlers, and turned out to be a bit
more code than the first version.
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c: Include "inf-loop.h".
(struct gdb_readline_wrapper_cleanup) <target_is_async_orig>: New
field.
(gdb_readline_wrapper_cleanup): Make the target async again, if it
was async before.
(gdb_readline_wrapper): Store whether the target is async, and
make it sync.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* gdb.base/paginate-inferior-exit.c: New file.
* gdb.base/paginate-inferior-exit.exp: New file.
If pagination occurs as result of output sent as response to a target
event while the target is executing in the background, subsequent
input aborts readline/gdb:
$ gdb program
...
(gdb) continue&
Continuing.
(gdb)
---Type <return> to continue, or q <return> to quit---
*return*
---Type <return> to continue, or q <return> to quit---
Breakpoint 2, after_sleep () at paginate-bg-execution.c:21
---Type <return> to continue, or q <return> to quit---
21 return; /* after sleep */
p 1
readline: readline_callback_read_char() called with no handler!
*abort/SIGABRT*
$
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
We should do better with the prompt handling while the target is
running (I think we should coordinate with readline, and
hide/redisplay it around output), but that's a more invasive change
better done post 7.8, so this patch is conservative and just
reinstalls the handler as soon as we're out of the readline line
callback.
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* gdb.base/paginate-bg-execution.c: New file.
* gdb.base/paginate-bg-execution.exp: New file.
This fixes:
$ ./gdb program -ex "set height 2" -ex "start"
...
Reading symbols from /home/pedro/gdb/tests/threads...done.
---Type <return> to continue, or q <return> to quit---^CQuit << ctrl-c triggers a Quit
*type something*
readline: readline_callback_read_char() called with no handler!
Aborted
$
Usually, if an error propagates all the way to the top level, we'll
re-enable stdin, in case the command that was running was a
synchronous command. That's done in the event loop's actual loop
(event-loop.c:start_event_loop). However, if a foreground execution
command is run before the event loop starts and throws, nothing is
presently reenabling stdin, which leaves sync_execution set.
When we do start the event loop, because sync_execution is still
(mistakenly) set, display_gdb_prompt removes the readline input
callback, even though stdin is registered in the event loop. Any
input from here on results in readline aborting.
Such commands are run through catch_command_errors,
catch_command_errors_const, so add the tweak there.
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* main.c: Include event-top.h.
(handle_command_errors): New function.
(catch_command_errors, catch_command_errors_const): Use it.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* gdb.base/paginate-execution-startup.c: New file.
* gdb.base/paginate-execution-startup.exp: New file.
* lib/gdb.exp (pagination_prompt): New global.
(default_gdb_spawn): New procedure, factored out from
default_gdb_spawn.
(default_gdb_start): Adjust to call default_gdb_spawn.
(gdb_spawn): New procedure.
Often we'll do something like:
if {$ok} {
fail "whatever"
} else {
pass "whatever"
}
This adds a helper procedure for that, and converts one random place
to use it, as an example.
2014-07-14 Pedro Alves <palves@redhat.com>
* lib/gdb.exp (gdb_assert): New procedure.
* gdb.trace/backtrace.exp (gdb_backtrace_tdp_4): Use it.
The "call" and "print" commands presently always run synchronously, in
the foreground, but GDB currently forgets to put the inferior's
terminal settings into effect while running them, on async-capable
targets, resulting in:
(gdb) print func ()
hello world
Program received signal SIGTTOU, Stopped (tty output).
0x000000373bceb8d0 in __libc_tcdrain (fd=1) at ../sysdeps/unix/sysv/linux/tcdrain.c:29
29 return INLINE_SYSCALL (ioctl, 3, fd, TCSBRK, 1);
The program being debugged was signaled while in a function called from GDB.
GDB remains in the frame where the signal was received.
To change this behavior use "set unwindonsignal on".
Evaluation of the expression containing the function
(func) will be abandoned.
When the function is done executing, GDB will silently stop.
(gdb)
That's because target_terminal_inferior skips actually doing anything
if running in the background, and, nothing is setting sync_execution
while running infcalls:
void
target_terminal_inferior (void)
{
/* A background resume (``run&'') should leave GDB in control of the
terminal. Use target_can_async_p, not target_is_async_p, since at
this point the target is not async yet. However, if sync_execution
is not set, we know it will become async prior to resume. */
if (target_can_async_p () && !sync_execution)
return;
This would best be all cleaned up by making GDB not even call
target_terminal_inferior and try to pass the terminal to the inferior
if running in the background, but that's a more invasive fix that is
better done post-7.8.
This was originally caught by a patch later in this series that makes
catch_command_errors use exception_print instead of
print_any_exception. Note that print_flush calls serial_drain_output
while print_any_exception doesnt't have that bit. And,
gdb.gdb/python-selftest.exp does:
gdb_test "call catch_command_errors(execute_command, \"python print 5\", 0, RETURN_MASK_ALL)" \
"Python not initialized.* = 0"
which without this fix results in SIGTTOU...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
* infcall.c (run_inferior_call): Set 'sync_execution' while
running the inferior call.
gdb/testsuite/
2014-07-14 Pedro Alves <palves@redhat.com>
* gdb.base/execution-termios.c: New file.
* gdb.base/execution-termios.exp: New file.
This fixes PR 17106, a regression in printing.
The bug is that resolve_dynamic_type follows struct members and
references, but doesn't consider the possibility of infinite
recursion.
This patch fixes the problem by limiting reference following to the
topmost layer of calls -- that is, reference-typed struct members are
never considered as being VLAs.
Built and regtested on x86-64 Fedora 20.
New test case included.
2014-07-14 Tom Tromey <tromey@redhat.com>
PR exp/17106:
* gdbtypes.c (is_dynamic_type_internal): New function, from
is_dynamic_type.
(is_dynamic_type): Rewrite.
(resolve_dynamic_union): Use resolve_dynamic_type_internal.
(resolve_dynamic_struct): Likewise.
(resolve_dynamic_type_internal): New function, from
resolve_dynamic_type.
(resolve_dynamic_type): Rewrite.
2014-07-14 Tom Tromey <tromey@redhat.com>
* gdb.cp/vla-cxx.cc: New file.
* gdb.cp/vla-cxx.exp: New file.
This fixes the record "run" regression pointed out by Marc Khouzam:
https://sourceware.org/ml/gdb/2014-06/msg00096.html
The bug is that target_require_runnable must agree with the handling
of the "run" target, but currently it is out of sync. This patch
fixes the problem by changing target_require_runnable to also ignore
the record_stratum.
Built and regtested on x86-64 Fedora 20.
New test case included.
2014-07-14 Tom Tromey <tromey@redhat.com>
* target.c (target_require_runnable): Also check record_stratum.
Update comment.
2014-07-14 Tom Tromey <tromey@redhat.com>
* gdb.reverse/rerun-prec.c: New file.
* gdb.reverse/rerun-prec.exp: New file.
Right now we provide a board info entry, `gdb_init_command', that allows
one to send a single command to GDB before the program to be debugged is
started. This is useful e.g. for slow remote targets to change the
default "remotetimeout" setting. Occasionally I found a need to send
multiple commands instead, however this cannot be achieved with
`gdb_init_command'.
This change therefore extends the mechanism by adding a TCL list of GDB
commands to send, via a board info entry called `gdb_init_commands'.
There is no limit as to the number of commands put there. The old
`gdb_init_command' mechanism remains supported for compatibility with
existing people's environments.
* lib/gdb-utils.exp: New file.
* lib/gdb.exp (gdb_run_cmd): Call gdb_init_commands, replacing
inline `gdb_init_command' processing.
(gdb_start_cmd): Likewise.
* lib/mi-support.exp (mi_run_cmd): Likewise.
* README: Document `gdb_init_command' and `gdb_init_commands'.
We see the following fails on arm-none-eabi target,
print (void*)v_signed_char^M
$190 = (void *) 0x0 <_ftext>^M
(gdb) FAIL: gdb.base/exprs.exp: print (void*)v_signed_char (print
(void*)v_signed_char)
GDB behaves correctly but the test assumes there is no symbol on
address 0x0. This patch is set print symbol off, so that tests below
can match the address only.
gdb/testsuite:
2014-07-11 Yao Qi <yao@codesourcery.com>
* gdb.base/exprs.exp: "set print symbol off".
Here's an example, with the new test:
gdbserver :9999 gdb.threads/kill
gdb gdb.threads/kill
(gdb) b 52
Breakpoint 1 at 0x4007f4: file kill.c, line 52.
Continuing.
Breakpoint 1, main () at kill.c:52
52 return 0; /* set break here */
(gdb) k
Kill the program being debugged? (y or n) y
gdbserver :9999 gdb.threads/kill
Process gdb.base/watch_thread_num created; pid = 9719
Listening on port 1234
Remote debugging from host 127.0.0.1
Killing all inferiors
Segmentation fault (core dumped)
Backtrace:
(gdb) bt
#0 0x00000000004068a0 in find_inferior (list=0x66b060 <all_threads>, func=0x427637 <kill_one_lwp_callback>, arg=0x7fffffffd3fc) at src/gdb/gdbserver/inferiors.c:199
#1 0x00000000004277b6 in linux_kill (pid=15708) at src/gdb/gdbserver/linux-low.c:966
#2 0x000000000041354d in kill_inferior (pid=15708) at src/gdb/gdbserver/target.c:163
#3 0x00000000004107e9 in kill_inferior_callback (entry=0x6704f0) at src/gdb/gdbserver/server.c:2934
#4 0x0000000000406522 in for_each_inferior (list=0x66b050 <all_processes>, action=0x4107a6 <kill_inferior_callback>) at src/gdb/gdbserver/inferiors.c:57
#5 0x0000000000412377 in process_serial_event () at src/gdb/gdbserver/server.c:3767
#6 0x000000000041267c in handle_serial_event (err=0, client_data=0x0) at src/gdb/gdbserver/server.c:3880
#7 0x00000000004189ff in handle_file_event (event_file_desc=4) at src/gdb/gdbserver/event-loop.c:434
#8 0x00000000004181c6 in process_event () at src/gdb/gdbserver/event-loop.c:189
#9 0x0000000000418f45 in start_event_loop () at src/gdb/gdbserver/event-loop.c:552
#10 0x0000000000411272 in main (argc=3, argv=0x7fffffffd8d8) at src/gdb/gdbserver/server.c:3283
The problem is that linux_wait_for_event deletes lwps that have exited
(even those not passed in as lwps of interest), while the lwp/thread
list is being walked on with find_inferior. find_inferior can handle
the current iterated inferior being deleted, but not others.
When killing lwps, we don't really care about any of the pending
status handling of linux_wait_for_event. We can just waitpid the lwps
directly, which is also what GDB does (see
linux-nat.c:kill_wait_callback). This way the lwps are not deleted
while we're walking the list. They'll be deleted by linux_mourn
afterwards.
This crash triggers several times when running the testsuite against
GDBserver with the native-gdbserver board (target remote), but as GDB
can't distinguish between GDBserver crashing and "kill" being
sucessful, as in both cases the connection is closed (the 'k' packet
doesn't require a reply), and the inferior is gone, that results in no
FAIL.
The patch adds a generic test that catches the issue with
extended-remote mode (and works fine with native testing too). Here's
how it fails with the native-extended-gdbserver board without the fix:
(gdb) info threads
Id Target Id Frame
6 Thread 15367.15374 0x000000373bcbc98d in nanosleep () at ../sysdeps/unix/syscall-template.S:81
5 Thread 15367.15373 0x000000373bcbc98d in nanosleep () at ../sysdeps/unix/syscall-template.S:81
4 Thread 15367.15372 0x000000373bcbc98d in nanosleep () at ../sysdeps/unix/syscall-template.S:81
3 Thread 15367.15371 0x000000373bcbc98d in nanosleep () at ../sysdeps/unix/syscall-template.S:81
2 Thread 15367.15370 0x000000373bcbc98d in nanosleep () at ../sysdeps/unix/syscall-template.S:81
* 1 Thread 15367.15367 main () at .../gdb.threads/kill.c:52
(gdb) kill
Kill the program being debugged? (y or n) y
Remote connection closed
^^^^^^^^^^^^^^^^^^^^^^^^
(gdb) FAIL: gdb.threads/kill.exp: kill
Extended remote should remain connected after the kill.
gdb/gdbserver/
2014-07-11 Pedro Alves <palves@redhat.com>
* linux-low.c (kill_wait_lwp): New function, based on
kill_one_lwp_callback, but use my_waitpid directly.
(kill_one_lwp_callback, linux_kill): Use it.
gdb/testsuite/
2014-07-11 Pedro Alves <palves@redhat.com>
* gdb.threads/kill.c: New file.
* gdb.threads/kill.exp: New file.
We see the fail below happens on thumb related multi-libs
(-mthumb -march={armv4t,armv7-a}),
target tfile tfile-basic.tf ^M
warning: Uploaded tracepoint 1 has no source location, using raw address^M
warning: Breakpoint address adjusted from 0x00008959 to 0x00008958.^M
Tracepoint 3 at 0x8958: file /scratch/yqi/arm-none-linux-gnueabi/src/gdb-trunk/gdb/testsuite/gdb.trace/tfile.c, line 91.^M
Created tracepoint 3 for target's tracepoint 1 at 0x8959.^M
warning: Breakpoint address adjusted from 0x00008959 to 0x00008958.^M
warning: Breakpoint address adjusted from 0x00008959 to 0x00008958.^M
warning: Breakpoint address adjusted from 0x00008959 to 0x00008958.^M
(gdb) FAIL: gdb.trace/tfile.exp: complete-command 'target tfile'
The address of write_basic_trace_file is two-bytes aligned,
(gdb) p write_basic_trace_file
$1 = {void (void)} 0x8958 <write_basic_trace_file>
but the ld sets the LSB of every reference to the function address
(indicating the address is in thumb mode), so "&write_basic_trace_file"
in the program becomes 0x8959, which is saved in the trace file. That
is why the warnnings are emitted.
This patch is to clear the LSB of the function address written to trace
file in thumb and thumb2 mode. This patch fixes the fail above.
gdb/testsuite:
2014-07-10 Yao Qi <yao@codesourcery.com>
* gdb.trace/tfile.c (write_basic_trace_file)
[__thumb__||__thumb2__]: Clear the Thumb bit of the function
address written to trace file.
On async targets, a synchronous attach is done like this:
#1 - target_attach is called (PTRACE_ATTACH is issued)
#2 - a continuation is installed
#3 - we go back to the event loop
#4 - target reports stop (SIGSTOP), event loop wakes up, and
attach continuation is called
#5 - among other things, the continuation calls
target_terminal_inferior, which removes stdin from the event
loop
Note that in #3, GDB is still processing user input. If the user is
fast enough, e.g., with something like:
echo -e "attach PID\nset xxx=1" | gdb
... then the "set" command is processed before the attach completes.
We get worse behavior even, if input is a tty and therefore
readline/editing is enabled, with e.g.,:
(gdb) attach PID\nset xxx=1
we then crash readline/gdb, with:
Attaching to program: attach-wait-input, process 14537
readline: readline_callback_read_char() called with no handler!
Aborted
$
Fix this by calling target_terminal_inferior before #3 above.
The test covers both scenarios by running with editing/readline forced
to both on and off.
gdb/
2014-07-09 Pedro Alves <palves@redhat.com>
* infcmd.c (attach_command_post_wait): Don't call
target_terminal_inferior here.
(attach_command): Call it here instead.
gdb/testsuite/
2014-07-09 Pedro Alves <palves@redhat.com>
* gdb.base/attach-wait-input.exp: New file.
* gdb.base/attach-wait-input.c: New file.
https://sourceware.org/ml/gdb-patches/2014-05/msg00383.html
The MI command -var-info-path-expression currently does not handle
non-anonymous structs / unions nested within other structs / unions,
it will skip parts of the expression. Consider this example:
## START EXAMPLE ##
$ cat ex.c
#include <string.h>
int
main ()
{
struct s1
{
int a;
};
struct ss
{
struct s1 x;
};
struct ss an_ss;
memset (&an_ss, 0, sizeof (an_ss));
return 0;
}
$ gcc -g -o ex.x ex.c
$ gdb ex.x
(gdb) break 18
Breakpoint 1 at 0x80483ba: file ex.c, line 18.
(gdb) run
Starting program: /home/user/ex.x
Breakpoint 1, main () at ex.c:18
18 return 0;
(gdb) interpreter-exec mi "-var-create an_ss * an_ss"
(gdb) interpreter-exec mi "-var-list-children an_ss"
^done,numchild="1",children=[child={name="an_ss.x",exp="x",numchild="1",type="struct s1",thread-id="1"}],has_more="0"
(gdb) interpreter-exec mi "-var-list-children an_ss.x"
^done,numchild="1",children=[child={name="an_ss.x.a",exp="a",numchild="0",type="int",thread-id="1"}],has_more="0"
(gdb) interpreter-exec mi "-var-list-children an_ss.x.a"
^done,numchild="0",has_more="0"
(gdb) interpreter-exec mi "-var-info-path-expression an_ss.x.a"
^done,path_expr="(an_ss).a"
(gdb) print (an_ss).a
There is no member named a.
## END EXAMPLE ##
Notice that the path expression returned is wrong, and as a result
the print command fails.
This patch adds a new method to the varobj_ops structure called
is_path_expr_parent, to allow language specific control over finding
the parent varobj, the current logic becomes the C/C++ version and is
extended to handle the nested cases. No other language currently uses
this code, so all other languages just get a default method.
With this patch, the above example now finishes like this:
## START EXAMPLE ##
$ gdb ex.x
(gdb) break 18
Breakpoint 1 at 0x80483ba: file ex.c, line 18.
(gdb) run
Starting program: /home/user/ex.x
Breakpoint 1, main () at ex.c:18
18 return 0;
(gdb) interpreter-exec mi "-var-list-children an_ss"
^done,numchild="1",children=[child={name="an_ss.x",exp="x",numchild="1",type="struct s1",thread-id="1"}],has_more="0"
(gdb) interpreter-exec mi "-var-list-children an_ss.x"
^done,numchild="1",children=[child={name="an_ss.x.a",exp="a",numchild="0",type="int",thread-id="1"}],has_more="0"
(gdb) interpreter-exec mi "-var-list-children an_ss.x.a"
^done,numchild="0",has_more="0"
(gdb) interpreter-exec mi "-var-info-path-expression an_ss.x.a"
^done,path_expr="((an_ss).x).a"
(gdb) print ((an_ss).x).a
$1 = 0
## END EXAMPLE ##
Notice that the path expression is now correct, and the print is a
success.
gdb/ChangeLog:
* ada-varobj.c (ada_varobj_ops): Fill in is_path_expr_parent
field.
* c-varobj.c (c_is_path_expr_parent): New function, moved core
from varobj.c, with additional checks.
(c_varobj_ops): Fill in is_path_expr_parent field.
(cplus_varobj_ops): Fill in is_path_expr_parent field.
* jv-varobj.c (java_varobj_ops): Fill in is_path_expr_parent
field.
* varobj.c (is_path_expr_parent): Call is_path_expr_parent varobj
ops method.
(varobj_default_is_path_expr_parent): New function.
* varobj.h (lang_varobj_ops): Add is_path_expr_parent field.
(varobj_default_is_path_expr_parent): Declare new function.
gdb/testsuite/ChangeLog:
* gdb.mi/var-cmd.c (do_nested_struct_union_tests): New function
setting up test structures.
(main): Call new test function.
* gdb.mi/mi2-var-child.exp: Create additional breakpoint in new
test function, continue into test function and walk test
structures.
We see some fails in gdb.trace/entry-values.exp in thumb mode
(-mthumb -march={armv4t,armv7-a}).
In thumb mode, the lsb of references to 'foo' and 'bar' in the assembly
(produced by dwarf assember) is set, so the generated debug
information is incorrect.
This patch copies the approach used by
[PATCH 4/4] Fix dw2-ifort-parameter.exp on PPC64
https://sourceware.org/ml/gdb-patches/2014-03/msg00202.html
to introduce new labels 'foo_start' and 'bar_start' which are about
the correct function address (without lsb set). This patch fixes
these fails we've seen.
gdb/testsuite:
2014-07-08 Yao Qi <yao@codesourcery.com>
* gdb.trace/entry-values.c: Define labels 'foo_start' and
'bar_start' at the beginning of functions 'foo' and 'bar'
respectively.
* gdb.trace/entry-values.exp: Use 'foo_start' and 'bar_start'
instead of 'foo' and 'bar'.
The reverse-finish command results in an internal error if it cannot determine
the current function.
(gdb) c
Continuing.
Program received signal SIGSEGV, Segmentation fault.
0x0000000000000000 in ?? ()
(gdb) reverse-finish
Run back to call of #0 0x0000000000000000 in ?? ()
gdb/infcmd.c:1576: internal-error: Finish: couldn't find function.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) y
This is not an internal error case since the command may be used in scenarios
where there is no function at the current PC, e.g. after calling through a bad
function pointer.
Turn this into a normal error.
gdb/
* infcmd.c (finish_backward): Turn internal error into normal error.
testsuite/
* gdb.btrace/segv.c: New.
* gdb.btrace/segv.exp: New.
does in a way, because the arm/aarch64 branch instruction is the
same as powerpc's, but the target triplet pattern is not there.
In summary, the testcase fails to locate the branch offset and causes
a failure and the early termination of the test.
The following patch adds a separate conditional block for powerpc (to keep
things organized), allowing the testcase to continue.
2014-07-02 Luis Machado <lgustavo@codesourcery.com>
* gdb.trace/entry-values.exp: Handle powerpc-specific branch
instruction.
read_tag_const_type propagates the cv-qualifier to the array element type,
but read_tag_volatile_type didn't. Make sure that both cv-qualifiers that
apply to array types are handled the same.
gdb/ChangeLog
* dwarf2read.c (add_array_cv_type): New function.
(read_tag_const_type): Call add_array_cv_type for TYPE_CODE_ARRAY.
(read_tag_volatile_type): Likewise.
gdb/testsuite/ChangeLog
* gdb.base/constvars.c (violent, violet, vips, virgen, vulgar,
vulture, vilify, villar): New volatile array constants.
(vindictive, vegetation): New const volatile array constants.
* gdb.base/volatile.exp: Test volatile and const volatile array
types.
The test case "watchpoint-reuse-slot.exp" yields a lot of failures on
s390/s390x: all instances of awatch, rwatch, and hbreak are performed
even though they aren't supported on these targets. This is because
the test case ignores non-support error messages when probing for
support of these commands, like:
(gdb) rwatch buf.byte[0]
Target does not support this type of hardware watchpoint.
The patch adds handling for this case in the appropriate
gdb_test_multiple invocations.
gdb/testsuite/
* gdb.base/watchpoint-reuse-slot.exp: Handle the case that the
target lacks support for awatch, rwatch, or hbreak.
This patch is to add ptid into dummy_frame and extend frame_id to
dummy_frame_id (which has a ptid field). With this change, GDB uses
dummy_frame_id (thread ptid and frame_id) to find the dummy frames.
Currently, dummy frames are looked up by frame_id, which isn't
accurate in non-stop or multi-process mode. The test case
gdb.multi/dummy-frame-restore.exp shows the problem and this patch can
fix it.
Test dummy-frame-restore.exp makes two inferiors stop at
different functions, say, inferior 1 stops at f1 while inferior 2
stops at f2. Set a breakpoint to a function, do the inferior call
in two inferiors, and GDB has two dummy frames of the same frame_id.
When the inferior call is finished, GDB will look up a dummy frame
from its stack/list and restore the inferior's regcache. Two
inferiors are finished in different orders, the inferiors' states are
restored differently, which is wrong. Running dummy-frame-restore.exp
under un-patched GDB, we'll get two fails:
FAIL: gdb.multi/dummy-frame-restore.exp: inf 2 first: after infcall: bt in inferior 2
FAIL: gdb.multi/dummy-frame-restore.exp: inf 2 first: after infcall: bt in inferior 1
With this patch applied, GDB will choose the correct dummy_frame to
restore for a given inferior, because ptid is considered when looking up
dummy frames. Two fails above are fixed.
Regression tested on x86_64-linux, both native and gdbserver.
gdb:
2014-06-27 Yao Qi <yao@codesourcery.com>
* breakpoint.c (check_longjmp_breakpoint_for_call_dummy):
Change parameter type to 'struct thread_info *'. Caller
updated.
* breakpoint.h (check_longjmp_breakpoint_for_call_dummy):
Update declaration.
* dummy-frame.c (struct dummy_frame_id): New.
(dummy_frame_id_eq): New function.
(struct dummy_frame) <id>: Change its type to 'struct
dummy_frame_id'.
(dummy_frame_push): Add parameter ptid and save it in
dummy_frame_id.
(pop_dummy_frame_bpt): Use ptid of dummy_frame instead of
inferior_ptid.
(pop_dummy_frame): Assert that the ptid of dummy_frame equals
to inferior_ptid.
(lookup_dummy_frame): Change parameter type to 'struct
dummy_frame_id *'. Callers updated. Call dummy_frame_id_eq
instead of frame_id_eq.
(dummy_frame_pop): Add parameter ptid. Callers updated.
Update comments. Compose dummy_frame_id and pass it to
lookup_dummy_frame.
(dummy_frame_discard): Add parameter ptid.
(dummy_frame_sniffer): Compose dummy_frame_id and call
dummy_frame_id_eq instead of frame_id_eq.
(fprint_dummy_frames): Print ptid.
* dummy-frame.h: Remove comments.
(dummy_frame_push): Add ptid in declaration.
(dummy_frame_pop, dummy_frame_discard): Likewise.
gdb/testsuite:
2014-06-27 Yao Qi <yao@codesourcery.com>
* gdb.multi/dummy-frame-restore.exp: New.
* gdb.multi/dummy-frame-restore.c: New.
gdb/doc:
2014-06-27 Yao Qi <yao@codesourcery.com>
* gdb.texinfo (Maintenance Commands): Update the output of
'maint print dummy-frames' command.
When generating a core file using the "generate-core-file" command while
replaying with the btrace record target, we won't be able to access all
registers and all memory. This leads to the following assertion:
gdb/regcache.c:1034: internal-error: regcache_raw_supply: Assertion `regnum >= 0 && regnum < regcache->descr->nr_raw_registers' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) FAIL: gdb.btrace/gcore.exp: generate-core-file core (GDB internal error)
Resyncing due to internal error.
Pretend that we are not replaying while generating a core file. This will
forward fetch and store registers as well as xfer memory calls to the target
beneath.
gdb/
* record-btrace.c (record_btrace_generating_corefile)
(record_btrace_prepare_to_generate_core)
(record_btrace_done_generating_core): New.
(record_btrace_xfer_partial, record_btrace_fetch_registers)
(record_btrace_store_registers, record_btrace_prepare_to_store):
Forward request when generating a core file.
(record_btrace_open): Set record_btrace_generating_corefile to zero.
(init_record_btrace_ops): Set to_prepare_to_generate_core and
to_done_generating_core.
testsuite/
* gdb.btrace/gcore.exp: New.
This patch fixes this on x86 Linux:
(gdb) watch *buf@2
Hardware watchpoint 8: *buf@2
(gdb) si
0x00000000004005a7 34 for (i = 0; i < 100000; i++); /* stepi line */
(gdb) del
Delete all breakpoints? (y or n) y
(gdb) watch *(buf+1)@1
Hardware watchpoint 9: *(buf+1)@1
(gdb) si
0x00000000004005a7 in main () at ../../../src/gdb/testsuite/gdb.base/watchpoint-reuse-slot.c:34
34 for (i = 0; i < 100000; i++); /* stepi line */
Couldn't write debug register: Invalid argument.
(gdb)
In the example above the debug registers are being switched from this
state:
CONTROL (DR7): 0000000000050101 STATUS (DR6): 0000000000000000
DR0: addr=0x0000000000601040, ref.count=1 DR1: addr=0x0000000000000000, ref.count=0
DR2: addr=0x0000000000000000, ref.count=0 DR3: addr=0x0000000000000000, ref.count=0
to this:
CONTROL (DR7): 0000000000010101 STATUS (DR6): 0000000000000000
DR0: addr=0x0000000000601041, ref.count=1 DR1: addr=0x0000000000000000, ref.count=0
DR2: addr=0x0000000000000000, ref.count=0 DR3: addr=0x0000000000000000, ref.count=0
That is, before, DR7 was setup for watching a 2 byte region starting
at what's in DR0 (0x601040).
And after, DR7 is setup for watching a 1 byte region starting at
what's in DR0 (0x601041).
We always write DR0..DR3 before DR7, because if we enable a slot's
bits in DR7, you need to have already written the corresponding
DR0..DR3 registers -- the kernel rejects the DR7 write with EINVAL
otherwise.
The error shown above is the opposite scenario. When we try to write
0x601041 to DR0, DR7's bits still indicate intent of watching a 2-byte
region. That DR0/DR7 combination is invalid, because 0x601041 is
unaligned. To watch two bytes, we'd have to use two slots. So the
kernel errors out with EINVAL.
Fix this by always first clearing DR7, then writing DR0..DR3, and then
setting DR7's bits.
A little optimization -- if we're disabling the last watchpoint, then
we can clear DR7 just once. The changes to nat/i386-dregs.c make that
easier to detect, and as bonus, they make it a little easier to make
sense of DR7 in the debug logs, as we no longer need to remember we're
seeing stale bits.
Tested on x86_64 Fedora 20, native and GDBserver.
This adds an exhaustive test that switches between many different
combinations of watchpoint types and addresses and widths.
gdb/
2014-06-23 Pedro Alves <palves@redhat.com>
* amd64-linux-nat.c (amd64_linux_prepare_to_resume): Clear
DR_CONTROL before setting DR0..DR3.
* i386-linux-nat.c (i386_linux_prepare_to_resume): Likewise.
* nat/i386-dregs.c (i386_remove_aligned_watchpoint): Clear all
bits of DR_CONTROL related to the debug register slot being
disabled. If all slots are vacant, clear local slowdown as well,
and assert DR_CONTROL is 0.
gdb/gdbserver/
2014-06-23 Pedro Alves <palves@redhat.com>
* linux-x86-low.c (x86_linux_prepare_to_resume): Clear DR_CONTROL
before setting DR0..DR3.
gdb/testsuite/
2014-06-23 Pedro Alves <palves@redhat.com>
* gdb.base/watchpoint-reuse-slot.c: New file.
* gdb.base/watchpoint-reuse-slot.exp: New file.
Currently, the xmethod commands lookup xmethod matchers in the current
progspace even if the locus regular expression matches the progspace's
filename. Pretty printer commands do not match the current progspace's
filename.
gdb/
* python/lib/gdb/command/xmethods.py
(get_method_matchers_in_loci): Lookup xmethod matchers in the
current progspace only if the string "progspace" matches LOCUS_RE.
gdb/testsuite
* gdb.python/py-xmethods.exp: Use "progspace" instead of the
progspace's filename in 'info', 'enable' and 'disable' command
tests.
On x86_64 with -m32 or on i686 it will:
Running ./gdb.arch/amd64-stap-special-operands.exp ...
gdb compile failed, amd64-stap-triplet.c: Assembler messages:
amd64-stap-triplet.c:35: Error: bad register name `%rbp'
amd64-stap-triplet.c:38: Error: bad register name `%rsp'
amd64-stap-triplet.c:40: Error: bad register name `%rbp)'
amd64-stap-triplet.c:41: Error: bad register name `%rsi'
amd64-stap-triplet.c:42: Error: bad register name `%rbp)'
/tmp/ccjOdmpl.s:63: Error: bad register name `%rbp'
2014-06-23 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.arch/amd64-stap-special-operands.exp: Use is_lp64_target.
* gdb.arch/amd64-stap-optional-prefix.exp: Likewise.
* gdb.dwarf2/dw2-error.exp: Use istarget and is_lp64_target.
Message-ID: <20140622211401.GA3716@host2.jankratochvil.net>
gdb/
2014-06-05 Iain Buclaw <ibuclaw@gdcproject.org>
* Makefile.in (SFILES): Add d-exp.y.
(YYFILES): Add d-exp.c.
(YYOBJ): Add d-exp.o.
(local-maintainer-clean): Delete d-exp.c.
* d-exp.y: New file.
* d-lang.h (d_parse): New declaration.
(d_error): New declaration.
* d-lang.c (d_op_print_tab): Add entry for BINOP_CONCAT and BINOP_EXP.
Set BINOP_EQUAL and BINOP_NOTEQUAL to same precedence as other
PREC_ORDER operators.
(d_language_defn): Use d_parse, d_error instead of c_parse, c_error.
gdb/testsuite/
2014-06-05 Iain Buclaw <ibuclaw@gdcproject.org>
* gdb.dlang/expression.exp: New file.
Running gdb.threads/thread-execl.exp with scheduler-locking set to
"step" reveals a problem:
(gdb) next^M
[Thread 0x7ffff7fda700 (LWP 27168) exited]^M
[New LWP 27168]^M
[Thread 0x7ffff74ee700 (LWP 27174) exited]^M
process 27168 is executing new program: /home/jkratoch/redhat/gdb-clean/gdb/testsuite/gdb.threads/thread-execl^M
[Thread debugging using libthread_db enabled]^M
Using host libthread_db library "/lib64/libthread_db.so.1".^M
infrun.c:5225: internal-error: switch_back_to_stepped_thread: Assertion `!schedlock_applies (1)' failed.^M
A problem internal to GDB has been detected,^M
further debugging may prove unreliable.^M
Quit this debugging session? (y or n) FAIL: gdb.threads/thread-execl.exp: schedlock step: get to main in new image (GDB internal error)
The assertion is correct. The issue is that GDB is mistakenly trying
to switch back to an exited thread, that was previously stepping when
it exited. This is exactly the sort of thing the test wants to make
sure doesn't happen:
# Now set a breakpoint at `main', and step over the execl call. The
# breakpoint at main should be reached. GDB should not try to revert
# back to the old thread from the old image and resume stepping it
We don't see this bug with schedlock off only because a different
sequence of events makes GDB manage to delete the thread instead of
marking it exited.
This particular internal error can be fixed by making the loop over
all threads in switch_back_to_stepped_thread skip exited threads.
But, looking over other ALL_THREADS users, all either can or should be
skipping exited threads too. So for simplicity, this patch replaces
ALL_THREADS with a new macro that skips exited threads itself, and
updates everything to use it.
Tested on x86_64 Fedora 20.
gdb/
2014-06-19 Pedro Alves <palves@redhat.com>
* gdbthread.h (ALL_THREADS): Delete.
(ALL_NON_EXITED_THREADS): New macro.
* btrace.c (btrace_free_objfile): Use ALL_NON_EXITED_THREADS
instead of ALL_THREADS.
* infrun.c (find_thread_needs_step_over)
(switch_back_to_stepped_thread): Use ALL_NON_EXITED_THREADS
instead of ALL_THREADS.
* record-btrace.c (record_btrace_open)
(record_btrace_stop_recording, record_btrace_close)
(record_btrace_is_replaying, record_btrace_resume)
(record_btrace_find_thread_to_move, record_btrace_wait): Likewise.
* remote.c (append_pending_thread_resumptions): Likewise.
* thread.c (thread_apply_all_command): Likewise.
gdb/testsuite/
2014-06-19 Pedro Alves <palves@redhat.com>
* gdb.threads/thread-execl.exp (do_test): New procedure, factored
out from ...
(top level): ... here. Iterate running tests under different
scheduler-locking settings.
with type DW_FORM_string, which is wrong.
GDB was using that information to load data as strings, and then
proceeded to use the string pointers as addresses.
Even then, the test was passing just fine, because we were lucky
enough to have the low_pc string pointer smaller than the high_pc
string pointer.
Two issues are fixed. The first one is the DW_FORM_string type. The
second one is adjusting the addresses so that they are non-zero,
since GDB doesn't like seeing 0 in these fields due to a check
contained in dwarf2_get_pc_bounds:
if (low == 0 && !dwarf2_per_objfile->has_section_at_zero)
return 0;
With both fixes, the testcase passes deterministically.
2014-06-19 Luis Machado <lgustavo@codesourcery.com>
* gdb.cp/nsalias.exp: Set type of low_pc and high_pc entries
to DW_FORM_addr and use non-zero addresses.
gdb/testsuite/
PR gdb/17017
* gdb.python/py-xmethods.cc: Add global function call counters and
increment them in their respective functions. Remove "cout"
statements.
* gdb.python/py-xmethods.exp: Make tests check the global function
call counters instead of depending on inferior IO.
Using the test program gdb.base/foll-fork.c, with follow-fork-mode set to
"child" and detach-on-fork set to "off", stepping or running past the fork
call results in the child process running to completion, when it should
just finish the single step. In addition, the breakpoint is not removed
from the parent process, so if it is resumed it receives a SIGTRAP.
Cause:
No matter what the setting for detach-on-fork, when stepping past a fork,
the single-step breakpoint (step_resume_breakpoint) is not handled
correctly in the parent. The SR breakpoint is cloned for the child
process, but before the clone is associated with the child it is treated as
a duplicate of the original, associated wth the parent. This results in
the insertion state of the original SR breakpoint and the clone being
"swapped" by breakpoint.c:update_global_location_list, so that the clone is
marked as inserted.
In the case where the parent is not detached, the two breakpoints remain in
that state. The breakpoint is never inserted in the child, because
although the cloned SR breakpoint is associated with the child, it is
marked as inserted. When the child is resumed, it runs to completion. The
breakpoint is never removed from the parent, so that if it is resumed after
the child exits, it gets a SIGTRAP.
Here is the sequence of events:
1) handle_inferior_event: FORK event is recognized.
2) handle_inferior_event: detach_breakpoints removes all breakpoints
from the child.
3) follow_fork: the parent SR breakpoint is cloned. Part of this procedure
is to call update_global_location_list, which swaps the insertion state of
the original and cloned SR breakpoints as part of ensuring that duplicate
breakpoints are only inserted once. At this point the original SR
breakpoint is not marked as inserted, and the clone is. The breakpoint is
actually inserted in the parent but not the child.
4) follow_fork: the original breakpoint is deleted by calling
delete_step_resume_breakpoint. Since the original is not marked as
inserted, the actual breakpoint remains in the parent process.
update_global_location_list is called again as part of the deletion. The
clone is still associated with the parent, but since it is marked as
enabled and inserted, the breakpoint is left in the parent.
5) follow_fork: if detach-on-fork is 'on', the actual breakpoint will be
removed from the parent in target_detach, based on the cloned breakpoint
still associated with the parent. Then the clone is no longer marked as
inserted. In follow_inferior_reset_breakpoints the clone is associated
with the child, and can be inserted.
If detach-on-fork is 'off', the actual breakpoint in the parent is never
removed (although the breakpoint had been deleted from the list). Since
the clone continues to be marked 'inserted', the SR breakpoint is never
inserted in the child.
Fix:
Set the cloned breakpoint as disabled from the moment it is created. This
is done by modifying clone_momentary_breakpoint to take an additional
argument, LOC_ENABLED, which is used as the value of the
bp_location->enabled member. The clone must be disabled at that point
because clone_momentary_breakpoint calls update_global_location_list, which
will swap treat the clone as a duplicate of the original breakpoint if it
is enabled.
All the calls to clone_momentary_breakpoint had to be modified to pass '1'
or '0'. I looked at implementing an enum for the enabled member, but
concluded that readability would suffer because there are so many places it
is used as a boolean, e.g. "if (bl->enabled)".
In follow_inferior_reset_breakpoints the clone is set to enabled once it
has been associated with the child process. With this, the bp_location
'inserted' member is maintained correctly throughout the follow-fork
procedure and the behavior is as expected.
The same treatment is given to the exception_resume_breakpoint when
following a fork.
Testing:
Ran 'make check' on Linux x64.
Along with the fix above, the coverage of the follow-fork test
gdb.base/foll-fork.exp was expanded to:
1) cover all the combinations of values for
follow-fork-mode and detach-on-fork
2) make sure that both user breakpoints and
single-step breakpoints are propagated
correctly to the child
3) check that the inferior list has the
expected contents after following the fork.
4) check that unfollowed, undetached inferiors
can be resumed.
gdb/
2014-06-18 Don Breazeal <donb@codesourcery.com>
* breakpoint.c (set_longjmp_breakpoint): Call
momentary_breakpoint_from_master with additional argument.
(set_longjmp_breakpoint_for_call_dummy): Call
momentary_breakpoint_from_master with additional argument.
(set_std_terminate_breakpoint): Call
momentary_breakpoint_from_master with additional argument.
(momentary_breakpoint_from_master): Add argument to function
definition and use it to initialize structure member flag.
(clone_momentary_breakpoint): Call
momentary_breakpoint_from_master with additional argument.
* infrun.c (follow_inferior_reset_breakpoints): Clear structure
member flags set in momentary_breakpoint_from_master.
gdb/testsuite/
2014-06-18 Don Breazeal <donb@codesourcery.com>
* gdb.base/foll-fork.exp (default_fork_parent_follow):
Deleted procedure.
(explicit_fork_parent_follow): Deleted procedure.
(explicit_fork_child_follow): Deleted procedure.
(test_follow_fork): New procedure.
(do_fork_tests): Replace calls to deleted procedures with
calls to test_follow_fork and reset GDB for subsequent
procedure calls.
We find the following fails in gdb test on mingw host.
FAIL: gdb.base/wchar.exp: print repeat
FAIL: gdb.base/wchar.exp: print repeat_p
FAIL: gdb.base/wchar.exp: print repeat (print null on)
FAIL: gdb.base/wchar.exp: print repeat (print elements 3)
FAIL: gdb.base/wchar.exp: print repeat_p (print elements 3)
print repeat^M
$7 = L"A", '¢' <repeats 21 times>, "B", '\000' <repeats 104 times>^M
(gdb) FAIL: gdb.base/wchar.exp: print repeat
the \242 is expected in the test but cent sign is displayed.
In valprint.c:print_wchar, wchar_printable is called to determine
whether a wchar is printable. wchar_printable calls iswprint but
the iswprint's return value depends on LC_CTYPE setting of locale [1, 2].
The output may vary with different locale settings and OS. IMO, '¢'
(cent sign) is a correct output on Windows.
[1] http://pubs.opengroup.org/onlinepubs/009604499/functions/iswprint.html
[2] http://msdn.microsoft.com/en-us/library/ewx8s4kw.aspx
This patch is set $cent to cent sign if the GDB is running on a
Windows host.
gdb/testsuite:
2014-06-17 Yao Qi <yao@codesourcery.com>
* gdb.base/wchar.exp: Set $cent to \u00A2 if "host-charset" is
CP1252.
skip_type_update_when_not_use_rtti_test) the testcase assumes an
uninitialized object has a specific type. In particular, 'ptr' and
's'.
In reality the compiler is free to do what it wants with that
uninitialized variable, even initialize it beforehand with the future
assignment's value. This is exactly what happens on some targets.
ptr should have type 'Base *', but it really has type 'Derived *'
because it is already initialized (earlier) by the compiler. The same
thing happens to 's'.
The following patch addresses this by explicitly initializing those
variables so the compiler doesn't optimize their assignments and GDB
can print their correct values.
2014-06-17 Luis Machado <lgustavo@codesourcery.com>
* gdb.mi/mi-var-rtti.cc (type_update_when_use_rtti_test):
Initialize ptr and S explicitly.
(skip_type_update_when_not_use_rtti_test): Likewise.
If an MI client creates a varobj and attempts to update the root
/before/ the inferior is started, gdb will throw an internal error:
(gdb)
-var-create * - batch_flag
^done,name="var1",numchild="0",value="0",type="int",has_more="0"
(gdb)
-var-update var1
^done,changelist=[]
(gdb)
-var-update *
~"../../src/gdb/thread.c:628: internal-error: is_thread_state: Assertion `tp' failed.\nA problem internal to GDB has been detected,\nfurther debugging may prove unreliable.\nQuit this debugging session? "
~"(y or n) "
The function that handles the varobj update in the failing case,
mi_cmd_var_udpate_iter, checks if the thread/inferior is stopped before
attempting to update the varobj. It calls is_stopped (inferior_ptid)
which calls is_thread_state:
tp = find_thread_ptid (ptid);
gdb_assert (tp);
When there is no inferior, ptid is null_ptid, and find_thread_ptid (null_ptid)
returns NULL and the assertion is triggered.
This patch changes mi_cmd_var_update_iter to behave the same way
"-var-update var1" does: by calling the thread "stopped" if
there is no inferior (and thereby calling varobj_update_one).
ChangeLog
2014-06-16 Keith Seitz <keiths@redhat.com>
PR mi/15863
* mi/mi-cmd-var.c (mi_cmd_var_update_iter): Do not attempt
to update the varobj if inferior_ptid is null_ptid.
testsuite/ChangeLog
2014-06-16 Keith Seitz <keiths@redhat.com>
PR mi/15863
* gdb.mi/mi-var-cmd.exp: Add test for -var-update before
the inferior is started.
Turns out there's a difference between loading the program with "gdb
PROGRAM", vs loading it with "(gdb) file PROGRAM". The latter results
in the objfile ending up with OBJF_USERLOADED set, while not with the
former. (That difference seems bogus, but still that's not the point
of this patch. We can revisit that afterwards.)
The new code that suppresses breakpoint removal errors for
add-symbol-file objects ends up being too greedy:
/* In some cases, we might not be able to remove a breakpoint in
a shared library that has already been removed, but we have
not yet processed the shlib unload event. Similarly for an
unloaded add-symbol-file object - the user might not yet have
had the chance to remove-symbol-file it. shlib_disabled will
be set if the library/object has already been removed, but
the breakpoint hasn't been uninserted yet, e.g., after
"nosharedlibrary" or "remove-symbol-file" with breakpoints
always-inserted mode. */
if (val
&& (bl->loc_type == bp_loc_software_breakpoint
&& (bl->shlib_disabled
|| solib_name_from_address (bl->pspace, bl->address)
|| userloaded_objfile_contains_address_p (bl->pspace,
bl->address))))
val = 0;
as it turns out that OBJF_USERLOADED can be set for objfiles loaded by
some other means not add-symbol-file. In this case, symbol-file (or
"file", which is really just "exec-file"+"symbol-file").
Recall that add-symbol-file is documented as:
(gdb) help add-symbol-file
Load symbols from FILE, assuming FILE has been dynamically loaded.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
And it's the "dynamically loaded" aspect that the breakpoint.c code
cares about. So make add-symbol-file set OBJF_SHARED on its objfiles
too, and tweak the breakpoint.c code to look for OBJF_SHARED instead
of OBJF_USERLOADED.
This restores back the missing breakpoint removal warning when we let
sss-bp-on-user-bp-2.exp run on native GNU/Linux
(https://sourceware.org/ml/gdb-patches/2014-06/msg00335.html):
(gdb) PASS: gdb.base/sss-bp-on-user-bp-2.exp: define stepi_del_break
stepi_del_break
warning: Error removing breakpoint 3
(gdb) FAIL: gdb.base/sss-bp-on-user-bp-2.exp: stepi_del_break
I say "restores" because this was GDB's behavior in 7.7 and earlier.
And, likewise, "file" with no arguments only started turning
breakpoints set in the main executable to "<pending>" with the
remote-symbol-file patch (63644780). The old behavior is now
restored, and we break-unload-file.exp test now exercizes both "gdb;
file PROGRAM" and "gdb PROGRAM".
gdb/
2014-06-16 Pedro Alves <palves@redhat.com>
* breakpoint.c (insert_bp_location, remove_breakpoint_1): Adjust.
(disable_breakpoints_in_freed_objfile): Skip objfiles that don't
have OBJF_SHARED set.
* objfiles.c (userloaded_objfile_contains_address_p): Rename to...
(shared_objfile_contains_address_p): ... this. Check OBJF_SHARED
instead of OBJF_USERLOADED.
* objfiles.h (OBJF_SHARED): Update comment.
(userloaded_objfile_contains_address_p): Rename to ...
(shared_objfile_contains_address_p): ... this, and update
comments.
* symfile.c (add_symbol_file_command): Also set OBJF_SHARED in the
new objfile.
(remove_symbol_file_command): Skip objfiles that don't have
OBJF_SHARED set.
gdb/testsuite/
2014-06-16 Pedro Alves <palves@redhat.com>
* gdb.base/break-main-file-remove-fail.c: New file.
* gdb.base/break-main-file-remove-fail.exp: New file.
* gdb.base/break-unload-file.exp: Use build_executable instead of
prepare_for_testing.
(test_break): New parameter "initial_load". Handle it.
(top level): Add initial_load cmdline/file axis.
I noticed that a few tests in completion.exp put the directory name
into the name of the resulting test. While the directory name is
relative, this still makes for spurious differences depending on
whether the test was run in serial or parallel mode.
This patch fixes the problem. I'm checking it in.
2014-06-12 Tom Tromey <tromey@redhat.com>
* gdb.base/completion.exp: Don't use directory name in test.
gdb/
2014-06-09 Pedro Alves <palves@redhat.com>
* linux-nat.c (linux_child_follow_fork): Initialize status with
W_STOPCODE (0) instead of 0. Remove shodowing 'status' local from
inner block. Only pass the signal to PTRACE_DETACH if in pass
state.
I'm seeing a ton of new FAILs in fork-related tests. Like, these and
many more:
+FAIL: gdb.base/disp-step-syscall.exp: vfork: continue to vfork (2nd time) (timeout)
+FAIL: gdb.base/disp-step-syscall.exp: vfork: display/i $pc (timeout)
...
-PASS: gdb.base/foll-vfork.exp: exec: vfork parent follow, through step: step
+FAIL: gdb.base/foll-vfork.exp: exec: vfork parent follow, through step: step (timeout)
-PASS: gdb.base/foll-vfork.exp: exec: vfork parent follow, to bp: continue to bp
+FAIL: gdb.base/foll-vfork.exp: exec: vfork parent follow, to bp: continue to bp (timeout)
...
FAIL: gdb.threads/watchpoint-fork.exp: parent: multithreaded: breakpoint (A) after the first fork (timeout)
FAIL: gdb.threads/watchpoint-fork.exp: parent: multithreaded: watchpoint A after the first fork (timeout)
FAIL: gdb.base/fileio.exp: System(3) call (timeout)
FAIL: gdb.threads/watchpoint-fork.exp: parent: multithreaded: watchpoint B after the first fork (timeout)
-PASS: gdb.base/multi-forks.exp: run to exit 2
+FAIL: gdb.base/multi-forks.exp: run to exit 2 (timeout)
...
PASS: gdb.base/watch-vfork.exp: Watchpoint on global variable (hw)
-PASS: gdb.base/watch-vfork.exp: Watchpoint triggers after vfork (hw)
+FAIL: gdb.base/watch-vfork.exp: Watchpoint triggers after vfork (hw) (timeout)
PASS: gdb.base/watch-vfork.exp: Watchpoint on global variable (sw)
-PASS: gdb.base/watch-vfork.exp: Watchpoint triggers after vfork (sw)
+FAIL: gdb.base/watch-vfork.exp: Watchpoint triggers after vfork (sw) (timeout)
Three issues with
https://sourceware.org/ml/gdb-patches/2014-06/msg00348.html
(c077881a).
- The inner 'status' local is shadowing the outer 'status' local,
thus PTRACE_DETACH is never seeing the status it intends to pass on
the inferior.
- With that fixed, we then try to pass down the SIGTRAP that results
from the step to the inferior. Need to filter out signals that are
in nopass state.
- For software single-step archs, the current code is equivalent to:
int status = 0;
if (WIFSTOPPED (status))
ptrace (PTRACE_DETACH, child_pid, 0, WSTOPSIG (status));
... and status == 0 is WIFEXITED, not WIFSTOPPED, so we're never
detaching.
gdb/
2014-06-09 Pedro Alves <palves@redhat.com>
* linux-nat.c (linux_child_follow_fork): Initialize status with
W_STOPCODE (0) instead of 0. Remove shodowing 'status' local from
inner block. Only pass the signal to PTRACE_DETACH if in pass
state.
This commit reorders various pieces of code to separate ANSI-standard
signals from other signals that need checking. Comments are added to
document this, and to document the ordering of the signals.
gdb/
2014-06-09 Gary Benson <gbenson@redhat.com>
* common/signals.c (gdb_signal_from_host): Reorder to separate
the always-available ANSI-standard signals from the signals that
require checking.
(do_gdb_signal_to_host): Likewise.
* proc-events.c (signal_table): Likewise.
gdb/testsuite/
2014-06-09 Gary Benson <gbenson@redhat.com>
* gdb.base/sigall.c [Functions to send signals]: Reorder to
separate the always-available ANSI-standard signals from the
signals that require checking.
(main): Likewise.
* gdb.reverse/sigall-reverse.c [Functions to send signals]:
Likewise.
(main): Likewise.
I noticed that sss-bp-on-user-bp-2.exp is racy on native GNU/Linux. I
sometimes still see an int3 in the disassembly:
(gdb) PASS: gdb.base/sss-bp-on-user-bp-2.exp: set debug target 0
disassemble test
Dump of assembler code for function test:
0x0000000000400590 <+0>: push %rbp
0x0000000000400591 <+1>: mov %rsp,%rbp
0x0000000000400594 <+4>: nop
=> 0x0000000000400595 <+5>: int3
0x0000000000400596 <+6>: pop %rbp
0x0000000000400597 <+7>: retq
End of assembler dump.
(gdb) FAIL: gdb.base/sss-bp-on-user-bp-2.exp: before/after disassembly matches
Enabling infrun/target debug logs, we can see the problem.
Simplified, that's:
(gdb) PASS: gdb.base/sss-bp-on-user-bp-2.exp: define stepi_del_break
stepi_del_break
infrun: clear_proceed_status_thread (process 25311)
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current thread [process 25311] at 0x400594
LLR: PTRACE_SINGLESTEP process 25311, 0 (resume event thread)
target_resume (25311, step, 0)
native:target_xfer_partial (3, (null), 0x0, 0x32dce4c, 0x400595, 1) = 0, 0
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(gdb) linux_nat_wait: [process -1], [TARGET_WNOHANG]
0x400595 is the address of the breakpoint, and "= 0" is
TARGET_XFER_EOF. That's default_memory_remove_breakpoint trying to
remove the breakpoint, but failing.
The problem is that we had just resumed the target and the native
GNU/Linux target can't read memory off of a running thread. Most of
the time, we get "lucky", because we manage to read memory before the
kernel actually schedules the target to run.
So just give up and skip the test on any target that uses hardware
stepping, not just remote targets.
gdb/testsuite/
2014-06-06 Pedro Alves <palves@redhat.com>
* gdb.base/sss-bp-on-user-bp-2.exp: Look for target_resume(step)
in target debug output instead of looking at RSP packets,
disabling the test on any target that uses hardware stepping.
Update comments.
I see the following fail in some thumb multi-lib in arm-none-linux-gnueabi,
info function jit_function^M
All functions matching regular expression "jit_function":^M
^M
Non-debugging symbols:^M
0x00000790 __real_jit_function_XXXX^M
0x0000079c __jit_function_XXXX_from_arm^M
0x0000079c jit_function_0000^M
0x00000790 __real_jit_function_XXXX^M
0x0000079c __jit_function_XXXX_from_arm^M
0x0000079c jit_function_0001^M
(gdb) FAIL: gdb.base/jit.exp: one_jit_test-2: info function jit_function
the test expects to see only jit_function_0000 and jit_function_0001
one_jit_test 2 "${hex} jit_function_0000\[\r\n\]+${hex} jit_function_0001"
Symbols with the prefix "__real_" or suffix "_from_arm" is generated
by gcc/ld for arm/thumb interworking.
This patch is to restrict the pattern from "jit_function" to
"^jit_function", the output becomes expected.
gdb/testsuite:
2014-06-06 Yao Qi <yao@codesourcery.com>
* gdb.base/jit.exp (one_jit_test): Restrict the pattern
from "jit_function" to "^jit_function".
I see two fails in async.exp on arm-none-eabi target:
nexti&^M
(gdb) 0x000001ba 14 x = 5; x = 5;^M
completed.^M
FAIL: gdb.base/async.exp: nexti&
finish&^M
Run till exit from #0 0x000001ba in foo () at /scratch/yqi/arm-none-eabi-lite/src/gdb-trunk/gdb/testsuite/gdb.base/async.c:14^M
(gdb) 0x000001e6 in main () at /scratch/yqi/arm-none-eabi-lite/src/gdb-trunk/gdb/testsuite/gdb.base/async.c:32^M
32 y = foo ();^M
Value returned is $1 = 8^M
completed.^M
FAIL: gdb.base/async.exp: finish&
The corresponding test is "test_background "nexti&" "" ".*y = 3.*"",
and it assumes that GDB "nexti" into the next source line. It is wrong
on arm. After "nexti", it still stops at the same source line, and it
fails.
When gdb does "finish", if the PC is in the middle of a source line,
the PC address is printed too. See stack.c:print_frame,
if (opts.addressprint)
if (!sal.symtab
|| frame_show_address (frame, sal)
|| print_what == LOC_AND_ADDRESS)
{
annotate_frame_address ();
if (pc_p)
ui_out_field_core_addr (uiout, "addr", gdbarch, pc);
else
ui_out_field_string (uiout, "addr", "<unavailable>");
annotate_frame_address_end ();
ui_out_text (uiout, " in ");
}
frame_show_address checks whether PC is the middle of a source line.
Since after "nexti", the inferior stops at the middle of a source line,
when we do "finish" the PC address is displayed.
In sum, GDB works well, but test case needs update. This patch is to
add a statement at the same line to make sure "nexti" doesn't go to
the new line, match the next instruction address in the output and
match the hex address the output of "finish".
gdb/testsuite:
2014-06-06 Yao Qi <yao@codesourcery.com>
* gdb.base/async.c (foo): Add one statement.
* gdb.base/async.exp: Get the next instruction address and
match the output of "nexti" by instruction address. Match
the hex address in the output of "finish".
The six signals SIGINT, SIGILL, SIGABRT, SIGFPE, SIGSEGV and SIGTERM
are ANSI-standard and thus guaranteed to be available. This patch
removes all preprocessor conditionals relating to these symbols.
gdb/
2014-06-06 Gary Benson <gbenson@redhat.com>
* common/signals.c: Remove preprocessor conditionals for
always-defined signals SIGINT, SIGILL, SIGABRT, SIGFPE,
SIGSEGV and SIGTERM.
* proc-events.c: Likewise.
gdb/testsuite/
2014-06-06 Gary Benson <gbenson@redhat.com>
* gdb.base/call-signals.c: Remove preprocessor conditionals
for always-defined signals SIGINT, SIGILL, SIGABRT, SIGFPE,
SIGSEGV and SIGTERM.
* gdb.base/sigall.c: Likewise.
* gdb.base/unwindonsignal.c: Likewise.
* gdb.reverse/sigall-reverse.c: Likewise.
hbreak-unmapped.exp assumes that memory at address 0 is unmapped or
unreadable, but on bare metal or uclinux targets, memory at address
0 is readable. For example, on arm-none-eabi, the vector table base
address is 0x0.
hbreak *0^M
Hardware assisted breakpoint 3 at 0x0: file
/scratch/yqi/arm-none-eabi-lite/obj/cs3-2014.11-999999-arm-none-eabi-i686-pc-linux-gnu/generated/arm-vector.S,
line 25.^M
(gdb) FAIL: gdb.base/hbreak-unmapped.exp: hbreak *0
info break^M
Num Type Disp Enb Address What^M
3 hw breakpoint keep y 0x00000000
/scratch/yqi/arm-none-eabi-lite/obj/cs3-2014.11-999999-arm-none-eabi-i686-pc-linux-gnu/generated/arm-vector.S:25^M
(gdb) FAIL: gdb.base/hbreak-unmapped.exp: info break shows hw breakpoint
delete $bpnum
This patch is to check whether address 0 is readable via command 'x 0'.
If it is, skip the test.
gdb/testsuite:
2014-06-06 Yao Qi <yao@codesourcery.com>
* gdb.base/hbreak-unmapped.exp: Read memory at address 0. If
readable, skip the test.
The code in gdb.threads/staticthreads.exp about checking the value of
tlsvar in main thread is racy, because when child thread hits
breakpoint, the main thread may not go into pthread_join yet, and
may not be unwind to main.
This patch is to move the line setting breakpoint on after sem_wait,
so that the child thread will hit breakpoint after main thread calls
sem_post. IOW, when child thread hits breakpoint, the main thread is
in either sem_post or pthread_join. "up 10" can unwind main thread to
main.
gdb/testsuite:
2014-06-06 Yao Qi <yao@codesourcery.com>
* gdb.threads/staticthreads.c (thread_function): Move the line
setting breakpoint on forward.
* gdb.threads/staticthreads.exp: Update comments.
Original patch:
https://sourceware.org/ml/gdb-patches/2014-04/msg00552.html
New in v2:
* In remote.c:escape_buffer, pass '\\' to fputstrn_unfiltered/printchar to
make sure backslashes are escaped in remote debug output.
* Updated function documentation for printchar.
See updated ChangeLog below.
--------------------
The quoting in whatever goes in the event_channel of MI is little bit broken.
Link for the lazy:
https://sourceware.org/bugzilla/show_bug.cgi?id=15806
Here is an example of a =library-loaded event with an ill-named directory,
/tmp/how"are\you (the problem is present with every directory on Windows since
it uses backslashes as a path separator). The result will be the following:
=library-loaded,id="/tmp/how"are\\you/libexpat.so.1",...
The " between 'how' and 'are' should be escaped.
Another bad behavior is double escaping in =breakpoint-created, for example:
=breakpoint-created,bkpt={...,fullname="/tmp/how\\"are\\\\you/test.c",...}
The two backslashes before 'how' should be one and the four before 'you' should
be two.
The reason for this is that when sending something to an MI console, escaping
can take place at two different moments (the actual escaping work is always
done in the printchar function):
1. When generating the content, if ui_out_field_* functions are used. Here,
fields are automatically quoted with " and properly escaped. At least
mi_field_string does it, not sure about mi_field_fmt, I need to investigate
further.
2. When gdb_flush is called, to send the data in the buffer of the console to
the actual output (stdout). At this point, mi_console_raw_packet takes the
whole string in the buffer, quotes it, and escapes all occurences of the
quoting character and backslashes. The event_channel does not specify a quoting
character, so quotes are not escaped here, only backslashes.
The problem with =library-loaded is that it does use fprintf_unfiltered, which
doesn't do escaping (so, no #1). When gdb_flush is called, backslashes are
escaped (#2).
The problem with =breakpoint-created is that it first uses ui_out_field_*
functions to generate its output, so backslashes and quotes are escaped there
(#1). backslashes are escaped again in #2, leading to an overdose of
backslashes.
In retrospect, there is no way escaping can be done reliably in
mi_console_raw_packet for data that is already formatted, such as
event_channel. At this point, there is no way to differentiate quotes that
delimit field values from those that should be escaped. In the case of other MI
consoles, it is ok since mi_console_raw_packet receives one big string that
should be quoted and escaped as a whole.
So, first part of the fix: for the MI channels that specify no quoting
character, no escaping at all should be done in mi_console_raw_packet (that's
the change in printchar, thanks to Yuanhui Zhang for this). For those channels,
whoever generates the content is responsible for proper quoting and escaping.
This will fix the =breakpoint-created kind of problem.
Second part of the fix is to make =library-loaded generate content that is
properly escaped. For this, we use ui_out_field_* functions, instead of one big
fprintf_unfiltered. =library-unloaded suffered from the same problem so it is
modified as well. There might be other events that need fixing too, but that's
all I found with a quick scan. Those that use fprintf_unfiltered but whose sole
variable data is a %d are not critical, since it won't generate a " or a \.
Finally, a test has been fixed, as it was expecting an erroneous output.
Otherwise, all other tests that were previously passing still pass (x86-64
linux).
gdb/ChangeLog:
2014-06-02 Simon Marchi <simon.marchi@ericsson.com>
PR mi/15806
* utils.c (printchar): Don't escape at all if quoter is NUL.
Update function documentation to clarify effect of parameter
QUOTER.
* remote.c (escape_buffer): Pass '\\' as the quoter to
fputstrn_unfiltered.
* mi/mi-interp.c (mi_solib_loaded): Use ui_out_field_* functions to
generate the output.
(mi_solib_unloaded): Same.
gdb/testsuite/ChangeLog:
2014-06-02 Simon Marchi <simon.marchi@ericsson.com>
* gdb.mi/mi-breakpoint-changed.exp (test_insert_delete_modify): Fix
erroneous dprintf expected input.
Support for smobs as goops classes is changing in guile 2.2.
We may eventually switch to using structs instead of smobs,
so remove any claim we support goops or generics for now.
* gdb.guile/scm-generics.exp: Delete.
It is valid in GNU C to have a VLA in a struct or union type, but gdb
did not handle this.
This patch adds support for these cases in the obvious way.
Built and regtested on x86-64 Fedora 20.
New tests included.
2014-06-04 Tom Tromey <tromey@redhat.com>
* ada-lang.c (ada_template_to_fixed_record_type_1): Use
value_from_contents_and_address_unresolved.
(ada_template_to_fixed_record_type_1): Likewise.
(ada_which_variant_applies): Likewise.
* value.h (value_from_contents_and_address_unresolved): Declare.
* value.c (value_from_contents_and_address_unresolved): New
function.
* gdbtypes.c (is_dynamic_type, resolve_dynamic_type)
<TYPE_CODE_STRUCT, TYPE_CODE_UNION>: New cases.
(resolve_dynamic_struct, resolve_dynamic_union): New functions.
2014-06-04 Tom Tromey <tromey@redhat.com>
* gdb.base/vla-datatypes.exp: Add tests for VLA-in-structure and
VLA-in-union.
* gdb.base/vla-datatypes.c (vla_factory): Add vla_struct,
inner_vla_struct, vla_union types. Initialize objects of those
types and compute their sizes.
sss-bp-on-user-bp.c has an assumption that write to integer can be
compiled to a single instruction, which isn't true on some arch, such
as arm. This test requires setting two breakpoints on two consecutive
instructions, so this patch is to get the address of the next
instruction via disassemble and set the 2nd breakpoint there. This
approach is portable.
This patch fixes the fails in sss-bp-on-user-bp.exp on arm-none-abi
target. There is no change in x86 test results. I also revert the
patch to PR breakpoints/17000, and verified that the patched
sss-bp-on-user-bp.exp still trigger the fail on
x86-with-software-single-step.
gdb/testsuite:
2014-06-04 Yao Qi <yao@codesourcery.com>
* gdb.base/sss-bp-on-user-bp.c (main): Remove comments.
* gdb.base/sss-bp-on-user-bp.exp: Don't set breakpoint on
"set bar break here". Get the next instruction address and
set breakpoint there. Remove "bar break" from the regexp
patterns.
This patch is update version according to the discussion in
https://www.sourceware.org/ml/gdb-patches/2009-11/msg00090.html.
If test get the target doesn't support fileio system according to the
remote log. It will set this test as "unsupported".
Before I made this patch, I want add a check before all of tests in this
file. But I found that the target maybe support one call but not others.
For example: my target support Fwrite, Fopen and so on. But not Fgettimeofday.
And it doesn't support Fsystem NULL but it support Fsystem not NULL.
So I think if we want to check target support fileio, we need check them
one by one.
2014-06-04 Nathan Sidwell <nathan@codesourcery.com>
Hui Zhu <hui@codesourcery.com>
* gdb.base/fileio.exp: Add test for shell not available as well as
available.
* gdb.base/fileio.c (test_system): Check for shell twice.
When I test gdb head (for 7.8 release) on arm-none-eabi, I find the
following this failure, which are caused by the improper regexp
pattern in the test.
(gdb) help target native^M
Undefined target command: "native". Try "help target".^M
(gdb) FAIL: gdb.base/auto-connect-native-target.exp: help target native
The space in front of "$gdb_prompt $" looks redundant, and this patch
is to remove it from the regexp pattern.
gdb/testsuite:
2014-06-04 Yao Qi <yao@codesourcery.com>
* gdb.base/auto-connect-native-target.exp: Remove redundant
space from the regexp pattern.
I see this failure on arm-none-eabi gdb testing.
target native^M
Undefined target command: "native". Try "help target".^M
(gdb) FAIL: gdb.base/default.exp: target native
This patch is to update the regexp pattern to match "native" instead of
"child".
gdb/testsuite:
2014-06-04 Yao Qi <yao@codesourcery.com>
* gdb.base/default.exp: Replace "child" with "native" in
regexp pattern.
* python/py-xmethods.c: New file.
* python/py-objfile.c (objfile_object): New field 'xmethods'.
(objfpy_dealloc): XDECREF on the new xmethods field.
(objfpy_new, objfile_to_objfile_object): Initialize xmethods
field.
(objfpy_get_xmethods): New function.
(objfile_getset): New entry 'xmethods'.
* python/py-progspace.c (pspace_object): New field 'xmethods'.
(pspy_dealloc): XDECREF on the new xmethods field.
(pspy_new, pspace_to_pspace_object): Initialize xmethods
field.
(pspy_get_xmethods): New function.
(pspace_getset): New entry 'xmethods'.
* python/python-internal.h: Add declarations for new functions.
* python/python.c (_initialize_python): Invoke
gdbpy_initialize_xmethods.
* python/lib/gdb/__init__.py (xmethods): New
attribute.
* python/lib/gdb/xmethod.py: New file.
* python/lib/gdb/command/xmethods.py: New file.
testuite/
* gdb.python/py-xmethods.cc: New testcase to test xmethods.
* gdb.python/py-xmethods.exp: New tests to test xmethods.
* gdb.python/py-xmethods.py: Python script supporting the
new testcase and tests.
with the following code...
12 Nested; -- break #1
13 return I; -- break #2
14 end;
(line 12 is a call to function Nested)
... we have noticed the following errorneous behavior on ppc-aix,
where, after having inserted a breakpoint at line 12 and line 13,
and continuing from the breakpoint at line 12, the program never
stops at line 13, running away until the program terminates:
% gdb -q func
(gdb) b func.adb:12
Breakpoint 1 at 0x10000a24: file func.adb, line 12.
(gdb) b func.adb:13
Breakpoint 2 at 0x10000a28: file func.adb, line 13.
(gdb) run
Starting program: /[...]/func
Breakpoint 1, func () at func.adb:12
12 Nested; -- break #1
(gdb) c
Continuing.
[Inferior 1 (process 4128872) exited with code 02]
When resuming from the first breakpoint, GDB first tries to step out
of that first breakpoint. We rely on software single-stepping on this
platform, and it just so happens that the address of the first
software single-step breakpoint is the same as the user's breakpoint
#2 (0x10000a28). So, with infrun and target traces turned on (but
uninteresting traces snip'ed off), the "continue" operation looks like
this:
(gdb) c
### First, we insert the user breakpoints (the second one is an internal
### breakpoint on __pthread_init). The first user breakpoint is not
### inserted as we need to step out of it first.
target_insert_breakpoint (0x0000000010000a28, xxx) = 0
target_insert_breakpoint (0x00000000d03f3800, xxx) = 0
### Then we proceed with the step-out-of-breakpoint...
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=1, current thread [process 15335610] at 0x10000a24
### That's when we insert the SSS breakpoints...
target_insert_breakpoint (0x0000000010000a28, xxx) = 0
target_insert_breakpoint (0x00000000100009ac, xxx) = 0
### ... then let the inferior resume...
target_resume (15335610, continue, 0)
infrun: wait_for_inferior ()
target_wait (-1, status, options={}) = 15335610, status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: target_wait (-1, status) =
infrun: 15335610 [process 15335610],
infrun: status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x100009ac
### At this point, we stopped at the second SSS breakpoint...
target_stopped_by_watchpoint () = 0
### We remove the SSS breakpoints...
target_remove_breakpoint (0x0000000010000a28, xxx) = 0
target_remove_breakpoint (0x00000000100009ac, xxx) = 0
target_stopped_by_watchpoint () = 0
### We find that we're not done, so we resume....
infrun: no stepping, continue
### And thus insert the user breakpoints again, except we're not
### inserting the second breakpoint?!?
target_insert_breakpoint (0x0000000010000a24, xxx) = 0
infrun: resume (step=0, signal=GDB_SIGNAL_0), trap_expected=0, current thread [process 15335610] at 0x100009ac
target_resume (-1, continue, 0)
infrun: prepare_to_wait
target_wait (-1, status, options={}) = 15335610, status->kind = exited, status = 2
What happens is that the removal of the software single-step
breakpoints effectively removed the breakpoint instruction from
inferior memory. But because such breakpoints are inserted directly
as raw breakpoints rather than through the normal chain of
breakpoints, we fail to notice that one of the user breakpoints points
to the same address and that this user breakpoint is therefore
effectively un-inserted. When resuming after the single-step, GDB
thinks that the user breakpoint is still inserted and therefore does
not need to insert it again.
This patch teaches the insert and remove routines of both regular and
raw breakpoints to be aware of each other. Special care needs to be
applied in case the target supports evaluation of breakpoint
conditions or commands.
gdb/ChangeLog:
PR breakpoints/17000
* breakpoint.c (find_non_raw_software_breakpoint_inserted_here):
New function, extracted from software_breakpoint_inserted_here_p.
(software_breakpoint_inserted_here_p): Replace factored out code
by call to find_non_raw_software_breakpoint_inserted_here.
(bp_target_info_copy_insertion_state): New function.
(bkpt_insert_location): Handle the case of a single-step
breakpoint already inserted at the same address.
(bkpt_remove_location): Handle the case of a single-step
breakpoint still inserted at the same address.
(deprecated_insert_raw_breakpoint): Handle the case of non-raw
breakpoint already inserted at the same address.
(deprecated_remove_raw_breakpoint): Handle the case of a
non-raw breakpoint still inserted at the same address.
(find_single_step_breakpoint): New function, extracted from
single_step_breakpoint_inserted_here_p.
(find_single_step_breakpoint): New function,
factored out from single_step_breakpoint_inserted_here_p.
(single_step_breakpoint_inserted_here_p): Reimplement.
gdb/testsuite/ChangeLog:
PR breakpoints/17000
* gdb.base/sss-bp-on-user-bp.exp: Remove kfail.
* gdb.base/sss-bp-on-user-bp-2.exp: Remove kfail.
Tested on ppc-aix with AdaCore's testsuite. Tested on x86_64-linux,
(native and gdbserver) with the official testsuite. Also tested on
x86_64-linux through Pedro's branch enabling software single-stepping
on that platform (native and gdbserver).
The check for the source (or "from") directory snippet in listing
matching path substitution rules currently will not match anything
other than a direct match of the "from" field in a substitution rule,
resulting in the incorrect behavior below:
...
(gdb) set substitute-path /a/path /another/path
(gdb) show substitute-path
List of all source path substitution rules:
`/a/path' -> `/another/path'.
(gdb) show substitute-path /a/path/to/a/file.ext
Source path substitution rule matching `/a/path/to/a/file.ext':
(gdb) show substitute-path /a/path
Source path substitution rule matching `/a/path':
`/a/path' -> `/another/path'.
...
This change effects the following behavior by (sanely) checking
with the length of the "from" portion of a rule and ensuring that
the next character of the path considered for substitution is a path
delimiter (or NULL). With this change, the following behavior is
garnered:
...
(gdb) set substitute-path /a/path /another/path
(gdb) show substitute-path
List of all source path substitution rules:
`/a/path' -> `/another/path'.
(gdb) show substitute-path /a/path/to/a/file.ext
Source path substitution rule matching `/a/path/to/a/file.ext':
`/a/path' -> `/another/path'.
(gdb) show substitute-path /a/pathological/case/that/should/fail.err
Source path substitution rule matching `/a/pathological/case/that/should/fail.err':
(gdb)
Also included is a couple of tests added to subst.exp to verify
this behavior in the test suite.
gdb/ChangeLog:
* source.c (show_substitute_path_command): Fix display of matching
substitution rules.
gdb/testsuite/ChangeLog:
* gdb.ada/subst.exp: Add tests to verify partial path matching
output.
This was tested on x86_64 Linux.
gdb/testsuite/
2014-06-03 Pedro Alves <palves@redhat.com>
* gdb.base/sss-bp-on-user-bp-2.exp: Skip if testing with a remote
target that doesn't use software single-stepping.
GDB gets confused when removing a software single-step breakpoint that
is at the same address as another breakpoint. Add another kfailed
test.
gdb/testsuite/
2014-06-03 Pedro Alves <palves@redhat.com>
PR breakpoints/17000
* gdb.base/sss-bp-on-user-bp-2.c: New file.
* gdb.base/sss-bp-on-user-bp-2.exp: New file.
While the full fix for PR 15180 isn't in, it's best if we at least
make sure that GDB doesn't lose control when a breakpoint is set at
the same address as a dprintf.
gdb/
2014-06-02 Pedro Alves <palves@redhat.com>
* breakpoint.c (build_target_command_list): Don't build a command
list if we have any duplicate location that isn't a dprintf.
gdb/testsuite/
2014-06-02 Pedro Alves <palves@redhat.com>
* gdb.base/dprintf-bp-same-addr.c: New file.
* gdb.base/dprintf-bp-same-addr.exp: New file.
This patch removes some code in gdb.base/compilation.exp which
is aimed at restoring the original timeout global value after having
changed it for this testcase. Restoring the timeout global is not
necessary as this is taken care of by gdb_init, which is called
at the start of each testing.
gdb/testsuite/ChangeLog:
* gdb.base/completion.exp: Remove code aimed at restoring TIMEOUT.
gdb.base/watchpoint.exp has a test below which expects to see "Cannot
access memory at address 0x0" when a null pointer is dereferenced.
gdb_test "watch -location nullptr->p->x" \
"Cannot access memory at address 0x0"
This assumption is not true when the target is no-mmu, so we get
watch -location nullptr->p->x
Hardware watchpoint 28: -location nullptr->p->x
(gdb) FAIL: gdb.base/watchpoint.exp: watch -location nullptr->p->x
This patch is to check whether null pointer can be dereferenced first
and then do the test.
gdb/testsuite:
2014-06-01 Yao Qi <yao@codesourcery.com>
* gdb.base/watchpoint.exp (test_watch_location): Check null
pointer can be dereferenced. If not, do the test, otherwise
skip it.
https://sourceware.org/ml/gdb-patches/2014-05/msg00737.html
Currently a MEMORY_ERROR raised during unwinding a frame will cause the
unwind to stop with an error message, for example:
(gdb) bt
#0 breakpt () at amd64-invalid-stack-middle.c:27
#1 0x00000000004008f0 in func5 () at amd64-invalid-stack-middle.c:32
#2 0x0000000000400900 in func4 () at amd64-invalid-stack-middle.c:38
#3 0x0000000000400910 in func3 () at amd64-invalid-stack-middle.c:44
#4 0x0000000000400928 in func2 () at amd64-invalid-stack-middle.c:50
Cannot access memory at address 0x2aaaaaab0000
However, frame #4 is marked as being the end of the stack unwind, so a
subsequent request for the backtrace looses the error message, such as:
(gdb) bt
#0 breakpt () at amd64-invalid-stack-middle.c:27
#1 0x00000000004008f0 in func5 () at amd64-invalid-stack-middle.c:32
#2 0x0000000000400900 in func4 () at amd64-invalid-stack-middle.c:38
#3 0x0000000000400910 in func3 () at amd64-invalid-stack-middle.c:44
#4 0x0000000000400928 in func2 () at amd64-invalid-stack-middle.c:50
When fetching the backtrace, or requesting the stack depth using the MI
interface the situation is even worse, the first time a request is made
we encounter the memory error and so the MI returns an error instead of
the correct result, for example:
(gdb) -stack-info-depth
^error,msg="Cannot access memory at address 0x2aaaaaab0000"
Or,
(gdb) -stack-list-frames
^error,msg="Cannot access memory at address 0x2aaaaaab0000"
However, once one of these commands has been used gdb has, internally,
walked the stack and figured that out that frame #4 is the bottom of the
stack, so the second time an MI command is tried you'll get the "expected"
result:
(gdb) -stack-info-depth
^done,depth="5"
Or,
(gdb) -stack-list-frames
^done,stack=[frame={level="0", .. snip lots .. }]
After this patch the MEMORY_ERROR encountered during the frame unwind is
attached to frame #4 as the stop reason, and is displayed in the CLI each
time the backtrace is requested. In the MI, catching the error means that
the "expected" result is returned the first time the MI command is issued.
So, from the CLI the results of the backtrace will be:
(gdb) bt
#0 breakpt () at amd64-invalid-stack-middle.c:27
#1 0x00000000004008f0 in func5 () at amd64-invalid-stack-middle.c:32
#2 0x0000000000400900 in func4 () at amd64-invalid-stack-middle.c:38
#3 0x0000000000400910 in func3 () at amd64-invalid-stack-middle.c:44
#4 0x0000000000400928 in func2 () at amd64-invalid-stack-middle.c:50
Backtrace stopped: Cannot access memory at address 0x2aaaaaab0000
Each and every time that the backtrace is requested, while the MI output
will similarly be consistently:
(gdb) -stack-info-depth
^done,depth="5"
Or,
(gdb) -stack-list-frames
^done,stack=[frame={level="0", .. snip lots .. }]
gdb/ChangeLog:
* frame.c (struct frame_info): Add stop_string field.
(get_prev_frame_always_1): Renamed from get_prev_frame_always.
(get_prev_frame_always): Old content moved into
get_prev_frame_always_1. Call get_prev_frame_always_1 inside
TRY_CATCH, handle MEMORY_ERROR exceptions.
(frame_stop_reason_string): New function definition.
* frame.h (unwind_stop_reason_to_string): Extend comment to
mention frame_stop_reason_string.
(frame_stop_reason_string): New function declaration.
* stack.c (frame_info): Switch to frame_stop_reason_string.
(backtrace_command_1): Switch to frame_stop_reason_string.
* unwind_stop_reason.def: Add UNWIND_MEMORY_ERROR.
(LAST_ENTRY): Changed to UNWIND_MEMORY_ERROR.
* guile/lib/gdb.scm: Add FRAME_UNWIND_MEMORY_ERROR to export list.
gdb/doc/ChangeLog:
* guile.texi (Frames In Guile): Mention FRAME_UNWIND_MEMORY_ERROR.
* python.texi (Frames In Python): Mention
gdb.FRAME_UNWIND_MEMORY_ERROR.
gdb/testsuite/ChangeLog:
* gdb.arch/amd64-invalid-stack-middle.exp: Update expected results.
* gdb.arch/amd64-invalid-stack-top.exp: Likewise.
https://sourceware.org/ml/gdb-patches/2014-05/msg00712.html
If an error is thrown during computing a frame id then the frame is left
in existence but without a valid frame id, this will trigger internal
errors if/when the frame is later visited (for example in a backtrace).
This patch catches errors raised while computing the frame id, and
arranges for the new frame, the one without a frame id, to be removed
from the linked list of frames.
gdb/ChangeLog:
* frame.c (remove_prev_frame): New function.
(get_prev_frame_if_no_cycle): Create / discard cleanup using
remove_prev_frame.
gdb/testsuite/ChangeLog:
* gdb.arch/amd64-invalid-stack-middle.S: New file.
* gdb.arch/amd64-invalid-stack-middle.c: New file.
* gdb.arch/amd64-invalid-stack-middle.exp: New file.
* gdb.arch/amd64-invalid-stack-top.c: New file.
* gdb.arch/amd64-invalid-stack-top.exp: New file.
GDB gets confused when removing a software single-step breakpoint that
is at the same address as another breakpoint. Add a kfailed test.
gdb/testsuite/
2014-05-30 Pedro Alves <palves@redhat.com>
PR breakpoints/17000
* gdb.base/sss-bp-on-user-bp.c: New file.
* gdb.base/sss-bp-on-user-bp.exp: New file.
As suggested by Andrew Pinski.
gdb/testsuite/
* gdb.opt/inline-break.c: Fix clang compatibility by specifying
gnu_inline semantics via attribute.
* gdb.opt/inline-break.exp: Remove -std=c89 now that the test
source explicitly specifies the required semantics.
This finally makes background execution commands possible by default.
However, in order to do that, there's one last thing we need to do --
we need to separate the MI and target notions of "async". Unlike the
CLI, where the user explicitly requests foreground vs background
execution in the execution command itself (c vs c&), MI chose to treat
"set target-async" specially -- setting it changes the default
behavior of execution commands.
So, we can't simply "set target-async" default to on, as that would
affect MI frontends. Instead we have to make the setting MI-specific,
and teach MI about sync commands on top of an async target.
Because the "target" word in "set target-async" ends up as a potential
source of confusion, the patch adds a "set mi-async" option, and makes
"set target-async" a deprecated alias.
Rather than make the targets always async, this patch introduces a new
"maint set target-async" option so that the GDB developer can control
whether the target is async. This makes it simpler to debug issues
arising only in the synchronous mode; important because sync mode
seems unlikely to go away.
Unlike in previous revisions, "set target-async" does not affect this
new maint parameter. The rationale for this is that then one can
easily run the test suite in the "maint set target-async off" mode and
have tests that enable mi-async fail just like they fail on
non-async-capable targets. This emulation is exactly the point of the
maint option.
I had asked Tom in a previous iteration to split the actual change of
the target async default to a separate patch, but it turns out that
that is quite awkward in this version of the patch, because with MI
async and target async decoupled (unlike in previous versions), if we
don't flip the default at the same time, then just "set target-async
on" alone never actually manages to do anything. It's best to not
have that transitory state in the tree.
Given "set target-async on" now only has effect for MI, the patch goes
through the testsuite removing it from non-MI tests. MI tests are
adjusted to use the new and less confusing "mi-async" spelling.
2014-05-29 Pedro Alves <palves@redhat.com>
Tom Tromey <tromey@redhat.com>
* NEWS: Mention "maint set target-async", "set mi-async", and that
background execution commands are now always available.
* target.h (target_async_permitted): Update comment.
* target.c (target_async_permitted, target_async_permitted_1):
Default to 1.
(set_target_async_command): Rename to ...
(maint_set_target_async_command): ... this.
(show_target_async_command): Rename to ...
(maint_show_target_async_command): ... this.
(_initialize_target): Adjust.
* infcmd.c (prepare_execution_command): Make extern.
* inferior.h (prepare_execution_command): Declare.
* infrun.c (set_observer_mode): Leave target async alone.
* mi/mi-interp.c (mi_interpreter_init): Install
mi_on_sync_execution_done as sync_execution_done observer.
(mi_on_sync_execution_done): New function.
(mi_execute_command_input_handler): Don't print the prompt if we
just started a synchronous command with an async target.
(mi_on_resume): Check sync_execution before printing prompt.
* mi/mi-main.h (mi_async_p): Declare.
* mi/mi-main.c: Include gdbcmd.h.
(mi_async_p): New function.
(mi_async, mi_async_1): New globals.
(set_mi_async_command, show_mi_async_command, mi_async): New
functions.
(exec_continue): Call prepare_execution_command.
(run_one_inferior, mi_cmd_exec_run, mi_cmd_list_target_features)
(mi_execute_async_cli_command): Use mi_async_p.
(_initialize_mi_main): Install "set mi-async". Make
"target-async" a deprecated alias.
2014-05-29 Pedro Alves <palves@redhat.com>
Tom Tromey <tromey@redhat.com>
* gdb.texinfo (Non-Stop Mode): Remove "set target-async 1"
from example.
(Asynchronous and non-stop modes): Document '-gdb-set mi-async'.
Mention that target-async is now deprecated.
(Maintenance Commands): Document maint set/show target-async.
2014-05-29 Pedro Alves <palves@redhat.com>
Tom Tromey <tromey@redhat.com>
* gdb.base/async-shell.exp: Don't enable target-async.
* gdb.base/async.exp
* gdb.base/corefile.exp (corefile_test_attach): Remove 'async'
parameter. Adjust.
(top level): Don't test with "target-async".
* gdb.base/dprintf-non-stop.exp: Don't enable target-async.
* gdb.base/gdb-sigterm.exp: Don't test with "target-async".
* gdb.base/inferior-died.exp: Don't enable target-async.
* gdb.base/interrupt-noterm.exp: Likewise.
* gdb.mi/mi-async.exp: Use "mi-async" instead of "target-async".
* gdb.mi/mi-nonstop-exit.exp: Likewise.
* gdb.mi/mi-nonstop.exp: Likewise.
* gdb.mi/mi-ns-stale-regcache.exp: Likewise.
* gdb.mi/mi-nsintrall.exp: Likewise.
* gdb.mi/mi-nsmoribund.exp: Likewise.
* gdb.mi/mi-nsthrexec.exp: Likewise.
* gdb.mi/mi-watch-nonstop.exp: Likewise.
* gdb.multi/watchpoint-multi.exp: Adjust comment.
* gdb.python/py-evsignal.exp: Don't enable target-async.
* gdb.python/py-evthreads.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.reverse/break-precsave.exp: Don't test with "target-async".
* gdb.server/solib-list.exp: Don't enable target-async.
* gdb.threads/thread-specific-bp.exp: Likewise.
* lib/mi-support.exp: Adjust to use mi-async.
Ignoring expected and desired differences like whether the prompt is
output after *stoppped records, GDB MI output is still different in
sync and async modes.
In sync mode, when a CLI execution command is entered, the "reason"
field is missing in the *stopped async record. And in async mode, for
some events, like program exits, the corresponding CLI output is
missing in the CLI channel.
Vis, diff between sync vs async modes:
run
^running
*running,thread-id="1"
(gdb)
...
- ~"[Inferior 1 (process 15882) exited normally]\n"
=thread-exited,id="1",group-id="i1"
=thread-group-exited,id="i1",exit-code="0"
- *stopped
+ *stopped,reason="exited-normally"
si
...
(gdb)
~"0x000000000045e033\t29\t memset (&args, 0, sizeof args);\n"
- *stopped,frame=...,thread-id="1",stopped-threads="all",core="0"
+ *stopped,reason="end-stepping-range",frame=...,thread-id="1",stopped-threads="all",core="0"
(gdb)
In addition, in both cases, when a MI execution command is entered,
and a breakpoint triggers, the event is sent to the console too. But
some events like program exits have the CLI output missing in the CLI
channel:
-exec-run
^running
*running,thread-id="1"
(gdb)
...
=thread-exited,id="1",group-id="i1"
=thread-group-exited,id="i1",exit-code="0"
- *stopped
+ *stopped,reason="exited-normally"
We'll want to make background commands always possible by default.
IOW, make target-async be the default. But, in order to do that,
we'll need to emulate MI sync on top of an async target. That means
we'll have yet another combination to care for in the testsuite.
Rather than making the testsuite cope with all these differences, I
thought it better to just fix GDB to always have the complete output,
no matter whether it's in sync or async mode.
This is all related to interpreter-exec, and the corresponding uiout
switching. (Typing a CLI command directly in MI is shorthand for
running it through -interpreter-exec console.)
In sync mode, when a CLI command is active, normal_stop is called when
the current interpreter and uiout are CLI's. So print_XXX_reason
prints the stop reason to CLI uiout (only), and we don't show it in
MI.
In async mode the stop event is processed when we're back in the MI
interpreter, so the stop reason is printed directly to the MI uiout.
Fix this by making run control event printing roughly independent of
whatever is the current interpreter or uiout. That is, move these
prints to interpreter observers, that know whether to print or be
quiet, and if printing, which uiout to print to. In the case of the
console/tui interpreters, only print if the top interpreter. For MI,
always print.
Breakpoint hits / normal stops are already handled similarly -- MI has
a normal_stop observer that prints the event to both MI and the CLI,
though that could be cleaned up further in the direction of this
patch.
This also makes all of:
(gdb) foo
and
(gdb) interpreter-exec MI "-exec-foo"
and
(gdb)
-exec-foo
and
(gdb)
-interpreter-exec console "foo"
print as expected.
Tested on x86_64 Fedora 20, sync and async modes.
gdb/
2014-05-29 Pedro Alves <palves@redhat.com>
PR gdb/13860
* cli/cli-interp.c: Include infrun.h and observer.h.
(cli_uiout, cli_interp): New globals.
(cli_on_signal_received, cli_on_end_stepping_range)
(cli_on_signal_exited, cli_on_exited, cli_on_no_history): New
functions.
(cli_interpreter_init): Install them as 'end_stepping_range',
'signal_received' 'signal_exited', 'exited' and 'no_history'
observers.
(_initialize_cli_interp): Remove cli_interp local.
* infrun.c (handle_inferior_event): Call the several stop reason
observers instead of printing the stop reason directly.
(end_stepping_range): New function.
(print_end_stepping_range_reason, print_signal_exited_reason)
(print_exited_reason, print_signal_received_reason)
(print_no_history_reason): Make static, and add an uiout
parameter. Print to that instead of to CURRENT_UIOUT.
* infrun.h (print_end_stepping_range_reason)
(print_signal_exited_reason, print_exited_reason)
(print_signal_received_reason print_no_history_reason): New
declarations.
* mi/mi-common.h (struct mi_interp): Rename 'uiout' field to
'mi_uiout'.
<cli_uiout>: New field.
* mi/mi-interp.c (mi_interpreter_init): Adjust. Create the new
uiout for CLI output. Install 'signal_received',
'end_stepping_range', 'signal_exited', 'exited' and 'no_history'
observers.
(find_mi_interpreter, mi_interp_data, mi_on_signal_received)
(mi_on_end_stepping_range, mi_on_signal_exited, mi_on_exited)
(mi_on_no_history): New functions.
(ui_out_free_cleanup): Delete function.
(mi_on_normal_stop): Don't allocate a new uiout for CLI output,
instead use the one already stored in the MI interpreter data.
(mi_ui_out): Adjust.
* tui/tui-interp.c: Include infrun.h and observer.h.
(tui_interp): New global.
(tui_on_signal_received, tui_on_end_stepping_range)
(tui_on_signal_exited, tui_on_exited)
(tui_on_no_history): New functions.
(tui_init): Install them as 'end_stepping_range',
'signal_received' 'signal_exited', 'exited' and 'no_history'
observers.
(_initialize_tui_interp): Delete tui_interp local.
gdb/doc/
2014-05-29 Pedro Alves <palves@redhat.com>
PR gdb/13860
* observer.texi (signal_received, end_stepping_range)
(signal_exited, exited, no_history): New observer subjects.
gdb/testsuite/
2014-05-29 Pedro Alves <palves@redhat.com>
PR gdb/13860
* gdb.mi/mi-cli.exp: Always expect "end-stepping-range" stop
reason, even in sync mode.
If one sets a breakpoint with a condition that involves calling a
function in the inferior, and then the condition evaluates false, GDB
outputs one *running event for each time the program hits the
breakpoint. E.g.,
$ gdb return-false -i=mi
(gdb)
start
...
(gdb)
b 14 if return_false ()
&"b 14 if return_false ()\n"
~"Breakpoint 2 at 0x4004eb: file return-false.c, line 14.\n"
...
^done
(gdb)
c
&"c\n"
~"Continuing.\n"
^running
*running,thread-id=(...)
(gdb)
*running,thread-id=(...)
*running,thread-id=(...)
*running,thread-id=(...)
*running,thread-id=(...)
*running,thread-id=(...)
... repeat forever ...
An easy way a user can trip on this is with a dprintf with "set
dprintf-style call". In that case, a dprintf is just a breakpoint
that when hit GDB calls the printf function in the inferior, and then
resumes it, just like the case above.
If the breakpoint/dprintf is set in a loop, then these spurious events
can potentially slow down a frontend much, if it decides to refresh
its GUI whenever it sees this event (Eclipse is one such case).
When we run an infcall, we pretend we don't actually run the inferior.
This is already handled for the usual case of calling a function
directly from the CLI:
(gdb)
p return_false ()
&"p return_false ()\n"
~"$1 = 0"
~"\n"
^done
(gdb)
Note no *running, nor *stopped events. That's handled by:
static void
mi_on_resume (ptid_t ptid)
{
...
/* Suppress output while calling an inferior function. */
if (tp->control.in_infcall)
return;
and equivalent code on normal_stop.
However, in the cases of the PR, after finishing the infcall there's
one more resume, and mi_on_resume doesn't know that it should suppress
output then too, somehow.
The "running/stopped" state is a high level user/frontend state.
Internal stops are invisible to the frontend. If follows from that
that we should be setting the thread to running at a higher level
where we still know the set of threads the user _intends_ to resume.
Currently we mark a thread as running from within target_resume, a low
level target operation. As consequence, today, if we resume a
multi-threaded program while stopped at a breakpoint, we see this:
-exec-continue
^running
*running,thread-id="1"
(gdb)
*running,thread-id="all"
The first *running was GDB stepping over the breakpoint, and the
second is GDB finally resuming everything.
Between those two *running's, threads other than "1" still have their
state set to stopped. That's bogus -- in async mode, this opens a
tiny window between both resumes where the user might try to run
another execution command to threads other than thread 1, and very
much confuse GDB.
That is, the "step" below should fail the "step", complaining that the
thread is running:
(gdb) c -a &
(gdb) thread 2
(gdb) step
IOW, threads that GDB happens to not resume immediately (say, because
it needs to step over a breakpoint) shall still be marked as running.
Then, if we move marking threads as running to a higher layer,
decoupled from target_resume, plus skip marking threads as running
when running an infcall, the spurious *running events disappear,
because there will be no state transitions at all.
I think we might end up adding a new thread state -- THREAD_INFCALL or
some such, however since infcalls are always synchronous today, I
didn't find a need. There's no way to execute a CLI/MI command
directly from the prompt if some thread is running an infcall.
Tested on x86_64 Fedora 20.
gdb/
2014-05-29 Pedro Alves <palves@redhat.com>
PR PR15693
* infrun.c (resume): Determine how much to resume depending on
whether the caller wanted a step, not whether we can hardware step
the target. Mark all threads that we intend to run as running,
unless we're calling an inferior function.
(normal_stop): If the thread is running an infcall, don't finish
thread state.
* target.c (target_resume): Don't mark threads as running here.
gdb/testsuite/
2014-05-29 Pedro Alves <palves@redhat.com>
Hui Zhu <hui@codesourcery.com>
PR PR15693
* gdb.mi/mi-condbreak-call-thr-state-mt.c: New file.
* gdb.mi/mi-condbreak-call-thr-state-st.c: New file.
* gdb.mi/mi-condbreak-call-thr-state.c: New file.
* gdb.mi/mi-condbreak-call-thr-state.exp: New file.
This patch removes support for the "set/show remotebaud" command,
which were deprecated in GDB 7.7, and should be now be removed
ahead of cutting the GDB 7.8 branch.
gdb/ChangeLog:
* serial.c (_initialize_serial): Remove support for
the "set remotebaud" and "show remotebaud" commands.
* NEWS: Add entry documenting the removal of that command.
gdb/testsuite/ChangeLog:
* config/monitor.exp (gdb_target_monitor): Replace use of
"set remotebaud" by "set serial baud".
When I run no-thread-db.exp, the breakpoint is set on line 26.
However, the breakpoint is set to line 26 of dl-start.S rather than
no-thread-db.c, which is not intended.
(gdb) monitor set libthread-db-search-path /foo/bar^M
libthread-db-search-path set to `/foo/bar'^M
(gdb) PASS: gdb.server/no-thread-db.exp: libthread-db is now unresolvable
break 26^M
Breakpoint 1 at 0x48018078: file ../sysdeps/powerpc/powerpc32/dl-start.S, line 26.^M
(gdb) continue^M
Continuing.
This patch is to change the breakpoint setting with source file
specified, then it is correct now.
gdb/testsuite:
2014-05-26 Yao Qi <yao@codesourcery.com>
* gdb.server/no-thread-db.exp: Specify source file name
explicitly when setting a breakpoint.
When loading symbols for the vdso, also add its sections to target_sections.
This fixes an issue with record btrace where vdso instructions could not be
disassembled during replay.
* symfile-mem.c (symbol_file_add_from_memory): Add BFD sections.
testsuite/
* gdb.btrace/vdso.c: New.
* gdb.btrace/vdso.exp: New.
Allow gcore's capture_command_output function to be used by other tests.
testsuite/
* gdb.base/gcore.exp (capture_command_output): Move ...
* lib/gdb.exp (capture_command_output): ... here.
The btrace record target does not trace data. We therefore do not allow
accessing read-write memory during replay.
In some cases, this might be useful to advanced users, though, who we assume
to know what they are doing.
Add a set|show command pair to turn this memory access restriction off.
* record-btrace.c (record_btrace_allow_memory_access): Remove.
(replay_memory_access_read_only, replay_memory_access_read_write)
(replay_memory_access_types, replay_memory_access)
(set_record_btrace_cmdlist, show_record_btrace_cmdlist)
(cmd_set_record_btrace, cmd_show_record_btrace)
(cmd_show_replay_memory_access): New.
(record_btrace_xfer_partial, record_btrace_insert_breakpoint)
(record_btrace_remove_breakpoint): Replace
record_btrace_allow_memory_access with replay_memory_access.
(_initialize_record_btrace): Add commands.
* NEWS: Announce it.
testsuite/
* gdb.btrace/data.exp: Test it.
doc/
* gdb.texinfo (Process Record and Replay): Document it.
It should clear up confusion about the args parameter to mi_run_cmd_full.
Thanks to Joel for clear formulation. I also added a comment about the
impact of use_gdb_stub.
gdb/testsuite/ChangeLog:
2014-05-22 Simon Marchi <simon.marchi@ericsson.com>
* lib/mi-support.exp (mi_run_cmd_full): Add comments.
The other part of PR gdb/13860 is about console execution commands in
MI getting their output half lost. E.g., take the finish command,
executed on a frontend's GDB console:
sync:
finish
&"finish\n"
~"Run till exit from #0 usleep (useconds=10) at ../sysdeps/unix/sysv/linux/usleep.c:27\n"
^running
*running,thread-id="1"
(gdb)
~"0x00000000004004d7 in foo () at stepinf.c:6\n"
~"6\t usleep (10);\n"
~"Value returned is $1 = 0\n"
*stopped,reason="function-finished",frame={addr="0x00000000004004d7",func="foo",args=[],file="stepinf.c",fullname="/home/pedro/gdb/tests/stepinf.c",line="6"},thread-id="1",stopped-threads="all",core="1"
async:
finish
&"finish\n"
~"Run till exit from #0 usleep (useconds=10) at ../sysdeps/unix/sysv/linux/usleep.c:27\n"
^running
*running,thread-id="1"
(gdb)
*stopped,reason="function-finished",frame={addr="0x00000000004004d7",func="foo",args=[],file="stepinf.c",fullname="/home/pedro/gdb/tests/stepinf.c",line="6"},gdb-result-var="$1",return-value="0",thread-id="1",stopped-threads="all",core="0"
Note how all the "Value returned" etc. output is missing in async mode.
The same happens with e.g., catchpoints:
=breakpoint-modified,bkpt={number="1",type="catchpoint",disp="keep",enabled="y",what="22016",times="1"}
~"\nCatchpoint "
~"1 (forked process 22016), 0x0000003791cbd8a6 in __libc_fork () at ../nptl/sysdeps/unix/sysv/linux/fork.c:131\n"
~"131\t pid = ARCH_FORK ();\n"
*stopped,reason="fork",disp="keep",bkptno="1",newpid="22016",frame={addr="0x0000003791cbd8a6",func="__libc_fork",args=[],file="../nptl/sysdeps/unix/sysv/linux/fork.c",fullname="/usr/src/debug/glibc-2.14-394-g8f3b1ff/nptl/sysdeps/unix/sysv/linux/fork.c",line="131"},thread-id="1",stopped-threads="all",core="0"
where all those ~ lines are missing in async mode, or just the "step"
current line indication:
s
&"s\n"
^running
*running,thread-id="all"
(gdb)
~"13\t foo ();\n"
*stopped,frame={addr="0x00000000004004ef",func="main",args=[{name="argc",value="1"},{name="argv",value="0x7fffffffdd78"}],file="stepinf.c",fullname="/home/pedro/gdb/tests/stepinf.c",line="13"},thread-id="1",stopped-threads="all",core="3"
(gdb)
Or in the case of the PRs example, the "Stopped due to shared library
event" note:
start
&"start\n"
~"Temporary breakpoint 1 at 0x400608: file ../../../src/gdb/testsuite/gdb.mi/solib-main.c, line 21.\n"
=breakpoint-created,bkpt={number="1",type="breakpoint",disp="del",enabled="y",addr="0x0000000000400608",func="main",file="../../../src/gdb/testsuite/gdb.mi/solib-main.c",fullname="/home/pedro/gdb/mygit/src/gdb/testsuite/gdb.mi/solib-main.c",line="21",times="0",original-location="main"}
~"Starting program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.mi/solib-main \n"
=thread-group-started,id="i1",pid="21990"
=thread-created,id="1",group-id="i1"
^running
*running,thread-id="all"
(gdb)
=library-loaded,id="/lib64/ld-linux-x86-64.so.2",target-name="/lib64/ld-linux-x86-64.so.2",host-name="/lib64/ld-linux-x86-64.so.2",symbols-loaded="0",thread-group="i1"
~"Stopped due to shared library event (no libraries added or removed)\n"
*stopped,reason="solib-event",thread-id="1",stopped-threads="all",core="3"
(gdb)
IMO, if you're typing execution commands in a frontend's console, you
expect to see their output. Indeed it's what you get in sync mode. I
think async mode should do the same. Deciding what to mirror to the
console wrt to breakpoints and random stops gets messy real fast.
E.g., say "s" trips on a breakpoint. We'd clearly want to mirror the
event to the console in this case. But what about more complicated
cases like "s&; thread n; s&", and one of those steps spawning a new
thread, and that thread hitting a breakpoint? It's impossible in
general to track whether the thread had any relation to the commands
that had been executed. So I think we should just simplify and always
mirror breakpoints and random events to the console.
Notes:
- mi->out is the same as gdb_stdout when MI is the current
interpreter. I think that referring to that directly is cleaner.
An earlier revision of this patch made the changes that are now
done in mi_on_normal_stop directly in infrun.c:normal_stop, and so
not having an obvious place to put the new uiout by then, and not
wanting to abuse CLI's uiout, I made a temporary uiout when
necessary.
- Hopefuly the rest of the patch is more or less obvious given the
comments added.
Tested on x86_64 Fedora 20, no regressions.
2014-05-21 Pedro Alves <palves@redhat.com>
PR gdb/13860
* gdbthread.h (struct thread_control_state): New field
`command_interp'.
* infrun.c (follow_fork): Copy the new thread control field to the
child fork thread.
(clear_proceed_status_thread): Clear the new thread control field.
(proceed): Set the new thread control field.
* interps.h (command_interp): Declare.
* interps.c (command_interpreter): New global.
(command_interp): New function.
(interp_exec): Set `command_interpreter' while here.
* cli-out.c (cli_uiout_dtor): New function.
(cli_ui_out_impl): Install it.
* mi/mi-interp.c: Include cli-out.h.
(mi_cmd_interpreter_exec): Add comment.
(restore_current_uiout_cleanup): New function.
(ui_out_free_cleanup): New function.
(mi_on_normal_stop): If finishing an execution command started by
a CLI command, or any kind of breakpoint-like event triggered,
print the stop event to the output (CLI) stream.
* mi/mi-out.c (mi_ui_out_impl): Install NULL `dtor' handler.
2014-05-21 Pedro Alves <palves@redhat.com>
PR gdb/13860
* gdb.mi/mi-cli.exp (line_callee4_next_step): New global.
(top level): Test that output related to execution commands is
sent to the console with CLI commands, but not with MI commands.
Test that breakpoint events are always mirrored to the console.
Also expect the new source line to be output after a "next" in
async mode too. Make it a pass/fail test.
* gdb.mi/mi-solib.exp: Test that the CLI solib event note is
output.
* lib/mi-support.exp (mi_gdb_expect_cli_output): New procedure.
I noticed that "list" behaves differently in CLI vs MI. Particularly:
$ ./gdb -nx -q ./testsuite/gdb.mi/mi-cli
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.mi/mi-cli...done.
(gdb) start
Temporary breakpoint 1 at 0x40054d: file ../../../src/gdb/testsuite/gdb.mi/basics.c, line 62.
Starting program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.mi/mi-cli
Temporary breakpoint 1, main () at ../../../src/gdb/testsuite/gdb.mi/basics.c:62
62 callee1 (2, "A string argument.", 3.5);
(gdb) list
57 {
58 }
59
60 main ()
61 {
62 callee1 (2, "A string argument.", 3.5);
63 callee1 (2, "A string argument.", 3.5);
64
65 do_nothing (); /* Hello, World! */
66
(gdb)
Note the list started at line 57. IOW, the program stopped at line
62, and GDB centered the list on that.
compare with:
$ ./gdb -nx -q ./testsuite/gdb.mi/mi-cli -i=mi
=thread-group-added,id="i1"
~"Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.mi/mi-cli..."
~"done.\n"
(gdb)
start
&"start\n"
...
~"\nTemporary breakpoint "
~"1, main () at ../../../src/gdb/testsuite/gdb.mi/basics.c:62\n"
~"62\t callee1 (2, \"A string argument.\", 3.5);\n"
*stopped,reason="breakpoint-hit",disp="del",bkptno="1",frame={addr="0x000000000040054d",func="main",args=[],file="../../../src/gdb/testsuite/gdb.mi/basics.c",fullname="/home/pedro/gdb/mygit/src/gdb/testsuite/gdb.mi/basics.c",line="62"},thread-id="1",stopped-threads="all",core="0"
=breakpoint-deleted,id="1"
(gdb)
-interpreter-exec console list
~"62\t callee1 (2, \"A string argument.\", 3.5);\n"
~"63\t callee1 (2, \"A string argument.\", 3.5);\n"
~"64\t\n"
~"65\t do_nothing (); /* Hello, World! */\n"
~"66\t\n"
~"67\t callme (1);\n"
~"68\t callme (2);\n"
~"69\t\n"
~"70\t return 0;\n"
~"71\t}\n"
^done
(gdb)
Here the list starts at line 62, where the program was stopped.
This happens because print_stack_frame, called from both normal_stop
and mi_on_normal_stop, is the function responsible for setting the
current sal from the selected frame, overrides the PRINT_WHAT
argument, and only after that does it decide whether to center the
current sal line or not, based on the overridden value, and it will
always decide false.
(The print_stack_frame call in mi_on_normal_stop is a little different
from the call in normal_stop, in that it is an unconditional
SRC_AND_LOC call. A future patch will make those uniform.)
A previous version of this patch made MI uniform with CLI here, by
making print_stack_frame also center when MI is active. That changed
the output of a "list" command in mi-cli.exp, to expect line 57
instead of 62, as per the example above.
However, looking deeper, that list in question is the first "list"
after the program stops, and right after the stop, before the "list",
the test did "set listsize 1". Let's try the same thing with the CLI:
(gdb) start
62 callee1 (2, "A string argument.", 3.5);
(gdb) set listsize 1
(gdb) list
57 {
Huh, that's unexpected. Why the 57? It's because print_stack_frame,
called in reaction to the breakpoint stop, expecting the next "list"
to show 10 lines (the listsize at the time) around line 62, sets the
lines listed range to 57-67 (62 +/- 5). If the user changes the
listsize before "list", why would we still show that range? Looks
bogus to me.
So the fix for this whole issue should be delay trying to center the
listing to until actually listing, so that the correct listsize can be
taken into account. This makes MI and CLI uniform too, as it deletes
the center code from print_stack_frame.
A series of tests are added to list.exp to cover this. mi-cli.exp was
after all correct all along, but it now gains an additional test that
lists lines with listsize 10, to ensure the centering is consistent
with CLI's.
One related Python test changed related output -- it's a test that
prints the line number after stopping for a breakpoint, similar to the
new list.exp tests. Previously we'd print the stop line minus 5 (due
to the premature centering), now we print the stop line. I think
that's a good change.
Tested on x86_64 Fedora 20.
gdb/
2014-05-21 Pedro Alves <palves@redhat.com>
* cli/cli-cmds.c (list_command): Handle the first "list" after the
current source line having changed.
* frame.h (set_current_sal_from_frame): Remove 'center' parameter.
* infrun.c (normal_stop): Adjust call to
set_current_sal_from_frame.
* source.c (clear_lines_listed_range): New function.
(set_current_source_symtab_and_line, identify_source_line): Clear
the lines listed range.
(line_info): Handle the first "info line" after the current source
line having changed.
* stack.c (print_stack_frame): Remove center handling.
(set_current_sal_from_frame): Remove 'center' parameter. Don't
center sal.line.
gdb/testsuite/
2014-05-21 Pedro Alves <palves@redhat.com>
* gdb.base/list.exp (build_pattern, test_list): New procedures.
Use them to test variations of "list" after reaching a breakpoint.
* gdb.mi/mi-cli.exp (line_main_callme_2): New global.
Test "list" with listsize 10 after reaching a breakpoint.
* gdb.python/python.exp (decode_line current location line
number): Adjust expected line number.
This reverts commit 8c217a4b68.
Following this
https://sourceware.org/ml/gdb-patches/2014-05/msg00462.html
I suggest reverting my previous commit. I will follow with another
patch to add comments, to clarify some things as stated in the mail
thread.
I ran make check with on gdb.mi, and the test that the commit broke
passes again.
gdb/testsuite/ChangeLog:
2014-05-21 Simon Marchi <simon.marchi@ericsson.com>
* lib/mi-support.exp (mi_run_cmd_full): Revert to original
behavior for $args, pass it directly to "run".
This fixes:
PASS: gdb.base/info-macros.exp: info macro -a -- FOO
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: info macros 2
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: info macros 3
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: info macros 4
FAIL: gdb.base/info-macros.exp: info macros *$pc
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: next
FAIL: gdb.base/info-macros.exp: info macros
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: next
FAIL: gdb.base/info-macros.exp: info macros 6
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: next
FAIL: gdb.base/info-macros.exp: info macros 7
ERROR: internal buffer is full.
UNRESOLVED: gdb.base/info-macros.exp: info macros info-macros.c:42 (PRMS
gdb/NNNN)
with the arm-eabi target tested on the i686-mingw32 host where GCC
defines enough macros to exhaust expect's 30000 characters of buffer
space.
* lib/gdb.exp (default_gdb_init): Bump `match_max' up from
30000 to 65536.
Sometimes it's useful to be able to disable the automatic connection
to the native target. E.g., sometimes GDB disconnects from the
extended-remote target I was debugging, without me noticing it, and
then I do "run". That starts the program locally, and only after a
little head scratch session do I figure out the program is running
locally instead of remotely as intended. Same thing with "attach",
"info os", etc.
With the patch, we now can have this instead:
(gdb) set auto-connect-native-target off
(gdb) target extended-remote :9999
...
*gdb disconnects*
(gdb) run
Don't know how to run. Try "help target".
To still be able to connect to the native target with
auto-connect-native-target set to off, I've made "target native" work
instead of erroring out as today.
Before:
(gdb) target native
Use the "run" command to start a native process.
After:
(gdb) target native
Done. Use the "run" command to start a process.
(gdb) maint print target-stack
The current target stack is:
- native (Native process)
- exec (Local exec file)
- None (None)
(gdb) run
Starting program: ./a.out
...
I've also wanted this for the testsuite, when running against the
native-extended-gdbserver.exp board (runs against gdbserver in
extended-remote mode). With a non-native-target board, it's always a
bug to launch a program with the native target. Turns out we still
have one such case this patch catches:
(gdb) break main
Breakpoint 1 at 0x4009e5: file ../../../src/gdb/testsuite/gdb.base/coremaker.c, line 138.
(gdb) run
Don't know how to run. Try "help target".
(gdb) FAIL: gdb.base/corefile.exp: run: with core
On the patch itself, probably the least obvious bit is the need to go
through all targets, and move the unpush_target call to after the
generic_mourn_inferior call instead of before. This is what
inf-ptrace.c does too, ever since multi-process support was added.
The reason inf-ptrace.c does things in that order is that in the
current multi-process/single-target model, we shouldn't unpush the
target if there are still other live inferiors being debugged. The
check for that is "have_inferiors ()" (a misnomer nowadays...), which
does:
have_inferiors (void)
{
for (inf = inferior_list; inf; inf = inf->next)
if (inf->pid != 0)
return 1;
It's generic_mourn_inferior that ends up clearing inf->pid, so we need
to call it before the have_inferiors check. To make all native
targets behave the same WRT to explicit "target native", I've added an
inf_child_maybe_unpush_target function that targets call instead of
calling unpush_target directly, and as that includes the
have_inferiors check, I needed to adjust the targets.
Tested on x86_64 Fedora 20, native, and also with the
extended-gdbserver board.
Confirmed a cross build of djgpp gdb still builds.
Smoke tested a cross build of Windows gdb under Wine.
Untested otherwise.
gdb/
2014-05-21 Pedro Alves <palves@redhat.com>
* inf-child.c (inf_child_ops, inf_child_explicitly_opened): New
globals.
(inf_child_open_target): New function.
(inf_child_open): Use inf_child_open_target to push the target
instead of erroring out.
(inf_child_disconnect, inf_child_close)
(inf_child_maybe_unpush_target): New functions.
(inf_child_target): Install inf_child_disconnect and
inf_child_close. Store a pointer to the returned object.
* inf-child.h (inf_child_open_target, inf_child_maybe_unpush): New
declarations.
* target.c (auto_connect_native_target): New global.
(show_default_run_target): New function.
(find_default_run_target): Return NULL if automatically connecting
to the native target is disabled.
(_initialize_target): Install set/show auto-connect-native-target.
* NEWS: Mention "set auto-connect-native-target", and "target
native".
* linux-nat.c (super_close): New global.
(linux_nat_close): Call super_close.
(linux_nat_add_target): Store a pointer to the base class's
to_close method.
* inf-ptrace.c (inf_ptrace_mourn_inferior, inf_ptrace_detach): Use
inf_child_maybe_unpush.
* inf-ttrace.c (inf_ttrace_him): Don't push the target if it is
already pushed.
(inf_ttrace_mourn_inferior): Only unpush the target after mourning
the inferior. Use inf_child_maybe_unpush_target.
(inf_ttrace_attach): Don't push the target if it is already
pushed.
(inf_ttrace_detach): Use inf_child_maybe_unpush_target.
* darwin-nat.c (darwin_mourn_inferior): Only unpush the target
after mourning the inferior. Use inf_child_maybe_unpush_target.
(darwin_attach_pid): Don't push the target if it is already
pushed.
* gnu-nat.c (gnu_mourn_inferior): Only unpush the target after
mourning the inferior. Use inf_child_maybe_unpush_target.
(gnu_detach): Use inf_child_maybe_unpush_target.
* go32-nat.c (go32_create_inferior): Don't push the target if it
is already pushed.
(go32_mourn_inferior): Use inf_child_maybe_unpush_target.
* nto-procfs.c (procfs_is_nto_target): Adjust comment.
(procfs_open): Rename to ...
(procfs_open_1): ... this. Add target_ops parameter. Adjust
comments. Can target_preopen before changing node. Call
inf_child_open_target to push the target explicitly.
(procfs_attach): Don't push the target if it is already pushed.
(procfs_detach): Use inf_child_maybe_unpush_target.
(procfs_create_inferior): Don't push the target if it is already
pushed.
(nto_native_ops): New global.
(procfs_open): Reimplement.
(procfs_native_open): New function.
(init_procfs_targets): Install procfs_native_open as to_open of
"target native". Store a pointer to the "native" target in
nto_native_ops.
* procfs.c (procfs_attach): Don't push the target if it is already
pushed.
(procfs_detach): Use inf_child_maybe_unpush_target.
(procfs_mourn_inferior): Only unpush the target after mourning the
inferior. Use inf_child_maybe_unpush_target.
(procfs_init_inferior): Don't push the target if it is already
pushed.
* windows-nat.c (do_initial_windows_stuff): Don't push the target
if it is already pushed.
(windows_detach): Use inf_child_maybe_unpush_target.
(windows_mourn_inferior): Only unpush the target after mourning
the inferior. Use inf_child_maybe_unpush_target.
gdb/doc/
2014-05-21 Pedro Alves <palves@redhat.com>
* gdb.texinfo (Starting): Document "set/show
auto-connect-native-target".
(Target Commands): Document "target native".
gdb/testsuite/
2014-05-21 Pedro Alves <palves@redhat.com>
* boards/gdbserver-base.exp (GDBFLAGS): Set to "set
auto-connect-native-target off".
* gdb.base/auto-connect-native-target.c: New file.
* gdb.base/auto-connect-native-target.exp: New file.
I had been pondering renaming "target child" to something else.
"child" is a little lie in case of "attach", and not exactly very
clear to users, IMO. By best suggestion is "target native". If I
were to explain what "target child" is, I'd just start out with "it's
the native target" anyway. I was worrying a little that "native"
might be a lie too if some port comes up with a default target that
can run but is not really native, but I think that's a very minor
issue - we can consider that "native" really means the default built
in target that GDB supports, instead of saying that's the target that
debugs host native processes, if it turns out necessary.
This change doesn't affect users much, because "target child" results
in error today:
(gdb) target child
Use the "run" command to start a child process.
Other places "child" is visible:
(gdb) help target
...
List of target subcommands:
target child -- Child process (started by the "run" command)
target core -- Use a core file as a target
target exec -- Use an executable file as a target
...
(gdb) info target
Symbols from "/home/pedro/gdb/mygit/build/gdb/gdb".
Child process:
Using the running image of child Thread 0x7ffff7fc9740 (LWP 4818).
While running this, GDB does not access memory from...
...
These places will say "native" instead. I think that's a good thing.
gdb/
2014-05-21 Pedro Alves <palves@redhat.com>
* inf-child.c (inf_child_open): Remove mention of "child".
(inf_child_target): Rename target to "native" instead of "child".
gdb/testsuite/
2014-05-21 Pedro Alves <palves@redhat.com>
* gdb.base/default.exp: Test "target native" instead of "target
child".
I have posted:
TLS variables access for -static -lpthread executables
https://sourceware.org/ml/libc-help/2014-03/msg00024.html
and the GDB patch below has been confirmed as OK for current glibcs.
Further work should be done for newer glibcs:
Improve TLS variables glibc compatibility
https://sourceware.org/bugzilla/show_bug.cgi?id=16954
Still the patch below implements the feature in a fully functional way backward
compatible with current glibcs, it depends on the following glibc source line:
csu/libc-tls.c
main_map->l_tls_modid = 1;
gdb/
2014-05-21 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix TLS access for -static -pthread.
* linux-thread-db.c (struct thread_db_info): Add td_thr_tlsbase_p.
(try_thread_db_load_1): Initialize it.
(thread_db_get_thread_local_address): Call it if LM is zero.
* target.c (target_translate_tls_address): Remove LM_ADDR zero check.
* target.h (struct target_ops) (to_get_thread_local_address): Add
load_module_addr comment.
gdb/gdbserver/
2014-05-21 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix TLS access for -static -pthread.
* gdbserver/thread-db.c (struct thread_db): Add td_thr_tlsbase_p.
(thread_db_get_tls_address): Call it if LOAD_MODULE is zero.
(thread_db_load_search, try_thread_db_load_1): Initialize it.
gdb/testsuite/
2014-05-21 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix TLS access for -static -pthread.
* gdb.threads/staticthreads.c <HAVE_TLS> (tlsvar): New.
<HAVE_TLS> (thread_function, main): Initialize it.
* gdb.threads/staticthreads.exp: Try gdb_compile_pthreads for $have_tls.
Add clean_restart.
<$have_tls != "">: Check TLSVAR.
Message-ID: <20140410115204.GB16411@host2.jankratochvil.net>
The dcache (code/stack cache) is supposed to be transparent, but it's
actually not in one case. dcache tries to read chunks (cache lines)
at a time off of the target. This may end up trying to read
unaccessible or unavailable memory. Currently the caller gets an xfer
error in this case. But if the specific bits of memory the caller
actually wanted are available and accessible, then the caller should
get the memory it wanted, not an error.
gdb/
2014-05-21 Pedro Alves <palves@redhat.com>
* dcache.c (dcache_read_memory_partial): If reading the cache line
fails, fallback to reading just the memory the caller wanted.
gdb/testsuite/
2014-05-21 Pedro Alves <palves@redhat.com>
* gdb.base/dcache-line-read-error.c: New.
* gdb.base/dcache-line-read-error.exp: New.
This does two things:
1. Adds a test.
Recently compare-sections got a new "-r" switch, but given no test
existed for compare-sections, the patch was allowed in with no
testsuite addition. This now adds a test for both compare-sections
and compare-sections -r.
2. Makes the compare-sections command work against all targets.
Currently, compare-sections only works with remote targets, and only
those that support the qCRC packet. The patch makes it so that if the
target doesn't support accelerating memory verification, then GDB
falls back to comparing memory itself. This is of course slower, but
it's better than nothing, IMO. While testing against extended-remote
GDBserver I noticed that we send the qCRC request to the target if
we're connected, but not yet running a program. That can't work of
course -- the patch fixes that. This all also goes in the direction
of bridging the local/remote parity gap.
I didn't decouple 1. from 2., because that would mean that the test
would need to handle the case of the target not supporting the
command.
Tested on x86_64 Fedora 17, native, remote GDBserver, and
extended-remote GDBserver. I also hack-disabled qCRC support to make
sure the fallback paths in remote.c work.
gdb/doc/
2014-05-20 Pedro Alves <palves@redhat.com>
* gdb.texinfo (Memory) <compare-sections>: Generalize comments to
not be remote specific. Add cross reference to the qCRC packet.
(Separate Debug Files): Update cross reference to the qCRC packet.
(General Query Packets) <qCRC packet>: Add anchor.
gdb/
2014-05-20 Pedro Alves <palves@redhat.com>
* NEWS: Mention that compare-sections now works with all targets.
* remote.c (PACKET_qCRC): New enum value.
(remote_verify_memory): Don't send qCRC if the target has no
execution. Use packet_support/packet_ok. If the target doesn't
support the qCRC packet, fallback to a deep memory copy.
(compare_sections_command): Say "target image" instead of "remote
executable".
(_initialize_remote): Add PACKET_qCRC to the list of config
packets that have no associated command. Extend comment.
* target.c (simple_verify_memory, default_verify_memory): New
function.
* target.h (struct target_ops) <to_verify_memory>: Default to
default_verify_memory.
(simple_verify_memory): New declaration.
* target-delegates.c: Regenerate.
gdb/testsuite/
2014-05-20 Pedro Alves <palves@redhat.com>
* gdb.base/compare-sections.c: New file.
* gdb.base/compare-sections.exp: New file.
This patch fixes hardware breakpoint regressions exposed by my fix for
"PR breakpoints/7143 - Watchpoint does not trigger when first set", at
https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html
The testsuite caught them on Linux/x86_64, at least. gdb.sum:
gdb.sum:
FAIL: gdb.base/hbreak2.exp: next over recursive call
FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1)
FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test
gdb.log:
(gdb) next
Program received signal SIGTRAP, Trace/breakpoint trap.
factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113
113 if (value > 1) { /* set breakpoint 7 here */
(gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call
Actually, that patch just exposed a latent issue to "breakpoints
always-inserted off" mode, not really caused it. After that patch,
GDB no longer removes breakpoints at each internal event, thus making
some scenarios behave like breakpoint always-inserted on. The bug is
easy to trigger with always-inserted on.
The issue is that since the target-side breakpoint conditions support,
if the stub/server supports evaluating breakpoint conditions on the
target side, then GDB is sending duplicate Zx packets to the target
without removing them before, and GDBserver is not really expecting
that for Z packets other than Z0/z0. E.g., with "set breakpoint
always-inserted on" and "set debug remote 1":
(gdb) b main
Sending packet: $m410943,1#ff...Packet received: 48
Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028.
Sending packet: $Z0,410943,1#48...Packet received: OK
^^^^^^^^^^^^
(gdb) b main
Note: breakpoint 4 also set at pc 0x410943.
Sending packet: $m410943,1#ff...Packet received: 48
Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028.
Sending packet: $Z0,410943,1#48...Packet received: OK
^^^^^^^^^^^^
(gdb) b main
Note: breakpoints 4 and 5 also set at pc 0x410943.
Sending packet: $m410943,1#ff...Packet received: 48
Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028.
Sending packet: $Z0,410943,1#48...Packet received: OK
^^^^^^^^^^^^
(gdb) del
Delete all breakpoints? (y or n) y
Sending packet: $Z0,410943,1#48...Packet received: OK
Sending packet: $Z0,410943,1#48...Packet received: OK
Sending packet: $z0,410943,1#68...Packet received: OK
And for Z1, similarly:
(gdb) hbreak main
Sending packet: $m410943,1#ff...Packet received: 48
Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028.
Sending packet: $Z1,410943,1#49...Packet received: OK
^^^^^^^^^^^^
Packet Z1 (hardware-breakpoint) is supported
(gdb) hbreak main
Note: breakpoint 4 also set at pc 0x410943.
Sending packet: $m410943,1#ff...Packet received: 48
Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028.
Sending packet: $Z1,410943,1#49...Packet received: OK
^^^^^^^^^^^^
(gdb) hbreak main
Note: breakpoints 4 and 5 also set at pc 0x410943.
Sending packet: $m410943,1#ff...Packet received: 48
Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028.
Sending packet: $Z1,410943,1#49...Packet received: OK
^^^^^^^^^^^^
(gdb) del
Delete all breakpoints? (y or n) y
Sending packet: $Z1,410943,1#49...Packet received: OK
^^^^^^^^^^^^
Sending packet: $Z1,410943,1#49...Packet received: OK
^^^^^^^^^^^^
Sending packet: $z1,410943,1#69...Packet received: OK
^^^^^^^^^^^^
So GDB sent a bunch of Z1 packets, and then when finally removing the
breakpoint, only one z1 packet was sent. On the GDBserver side (with
monitor set debug-hw-points 1), in the Z1 case, we see:
$ ./gdbserver :9999 ./gdbserver
Process ./gdbserver created; pid = 8629
Listening on port 9999
Remote debugging from host 127.0.0.1
insert_watchpoint (addr=410943, len=1, type=instruction-execute):
CONTROL (DR7): 00000101 STATUS (DR6): 00000000
DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0
DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0
insert_watchpoint (addr=410943, len=1, type=instruction-execute):
CONTROL (DR7): 00000101 STATUS (DR6): 00000000
DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0
DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0
insert_watchpoint (addr=410943, len=1, type=instruction-execute):
CONTROL (DR7): 00000101 STATUS (DR6): 00000000
DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0
DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0
insert_watchpoint (addr=410943, len=1, type=instruction-execute):
CONTROL (DR7): 00000101 STATUS (DR6): 00000000
DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0
DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0
insert_watchpoint (addr=410943, len=1, type=instruction-execute):
CONTROL (DR7): 00000101 STATUS (DR6): 00000000
DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0
DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0
remove_watchpoint (addr=410943, len=1, type=instruction-execute):
CONTROL (DR7): 00000101 STATUS (DR6): 00000000
DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0
DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0
That's one insert_watchpoint call for each Z1 packet, and then one
remove_watchpoint call for the z1 packet. Notice how ref.count
increased for each insert_watchpoint call, and then in the end, after
GDB told GDBserver to forget about the hardware breakpoint, GDBserver
ends with the the first debug register still with ref.count=4! IOW,
the hardware breakpoint is left armed on the target, while on the GDB
end it's gone. If the program happens to execute 0x410943 afterwards,
then the CPU traps, GDBserver reports the trap to GDB, and GDB not
having a breakpoint set at that address anymore, reports to the user a
spurious SIGTRAP.
This is exactly what is happening in the hbreak2.exp test, though in
that case, it's a shared library event that triggers a
breakpoint_re_set, when breakpoints are still inserted (because
nowadays GDB doesn't remove breakpoints while handling internal
events), and that recreates breakpoint locations, which likewise
forces breakpoint reinsertion and Zx packet resends...
That is a lot of bogus Zx duplication that should possibly be
addressed on the GDB side. GDB resends Zx packets because the way to
change the target-side condition, is to resend the breakpoint to the
server with the new condition. (That's an option in the packet: e.g.,
"Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the
examples above are shorter because the breakpoints don't have
conditions attached). GDB doesn't remove the breakpoint first before
reinserting it because that'd be bad for non-stop, as it'd open a
window where the inferior could miss the breakpoint. The conditions
actually haven't changed between the resends, but GDB isn't smart
enough to realize that.
(TBC, if the target doesn't support target-side conditions, then GDB
doesn't trigger these resends (init_bp_location calls
mark_breakpoint_location_modified, and that does nothing if condition
evaluation is on the host side. The resends are caused by the
'loc->condition_changed = condition_modified.' line.)
But, even if GDB was made smarter, GDBserver should really still
handle the resends anyway. So target-side conditions also aren't
really to blame. The documentation of the Z/z packets says:
"To avoid potential problems with duplicate packets, the operations
should be implemented in an idempotent way."
As such, we may want to fix GDB, but we should definitely fix
GDBserver. The fix is a prerequisite for target-side conditions on
hardware breakpoints anyway (and while at it, on watchpoints too).
GDBserver indeed already treats duplicate Z0 packets in an idempotent
way. mem-break.c has the concept of high-level and low-level
breakpoints, somewhat similar to GDB's split of breakpoints vs
breakpoint locations, and keeps track of multiple breakpoints
referencing the same address/location, for the case of an internal
GDBserver breakpoint or a tracepoint being set at the same address as
a GDB breakpoint. But, it only allows GDB to ever contribute one
reference to a software breakpoint location. IOW, if gdbserver sees a
Z0 packet for the same address where it already had a GDB breakpoint
set, then GDBserver won't create another high-level GDB breakpoint.
However, mem-break.c only tracks GDB Z0 breakpoints. The same logic
should apply to all kinds of Zx packets. Currently, gdbserver passes
down each duplicate Zx (other than Z0) request directly to the
target->insert_point routine. The x86 watchpoint support itself
refcounts watchpoint / hw breakpoint requests, to handle overlapping
watchpoints, and save debug registers. But that code doesn't (and
really shouldn't) handle the duplicate requests, assuming that for
each insert there will be a corresponding remove.
So the fix is to generalize mem-break.c to track all kinds of Zx
breakpoints, and filter out duplicates. As mentioned, this ends up
adding support for target-side conditions on hardware breakpoints and
watchpoints too (though GDB itself doesn't support the latter yet).
Probably the least obvious change in the patch is that it kind of
turns the breakpoint insert/remove APIs inside out. Before, the
target methods were only called for GDB breakpoints. The internal
breakpoint set/delete methods inserted memory breakpoints directly
bypassing the insert/remove target methods. That's not good when the
target should use a debug API to set software breakpoints, instead of
relying on GDBserver patching memory with breakpoint instructions, as
is the case of NTO.
Now removal/insertion of all kinds of breakpoints/watchpoints, either
internal, or from GDB, always go through the target methods. The
insert_point/remove_point methods no longer get passed a Z packet
type, but an internal/raw breakpoint type. They're also passed a
pointer to the raw breakpoint itself (note that's still opaque outside
mem-break.c), so that insert_memory_breakpoint /
remove_memory_breakpoint have access to the breakpoint's shadow
buffer. I first tried passing down a new structure based on GDB's
"struct bp_target_info" (actually with that name exactly), but then
decided against it as unnecessary complication.
As software/memory breakpoints work by poking at memory, when setting
a GDB Z0 breakpoint (but not internal breakpoints, as those can assume
the conditions are already right), we need to tell the target to
prepare to access memory (which on Linux means stop threads). If that
operation fails, we need to return error to GDB. Seeing an error, if
this is the first breakpoint of that type that GDB tries to insert,
GDB would then assume the breakpoint type is supported, but it may
actually not be. So we need to check whether the type is supported at
all before preparing to access memory. And to solve that, the patch
adds a new target->supports_z_point_type method that is called before
actually trying to insert the breakpoint.
Other than that, hopefully the change is more or less obvious.
New test added that exercises the hbreak2.exp regression in a more
direct way, without relying on a breakpoint re-set happening before
main is reached.
Tested by building GDBserver for:
aarch64-linux-gnu
arm-linux-gnueabihf
i686-pc-linux-gnu
i686-w64-mingw32
m68k-linux-gnu
mips-linux-gnu
mips-uclinux
nios2-linux-gnu
powerpc-linux-gnu
sh-linux-gnu
tilegx-unknown-linux-gnu
x86_64-redhat-linux
x86_64-w64-mingw32
And also regression tested on x86_64 Fedora 20.
gdb/gdbserver/
2014-05-20 Pedro Alves <palves@redhat.com>
* linux-aarch64-low.c (aarch64_insert_point)
(aarch64_remove_point): No longer check whether the type is
supported here. Adjust to new interface.
(the_low_target): Install aarch64_supports_z_point_type as
supports_z_point_type method.
* linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function.
(arm_linux_hw_point_initialize): Take an enum raw_bkpt_type
instead of a Z packet char. Adjust.
(arm_supports_z_point_type): New function.
(arm_insert_point, arm_remove_point): Adjust to new interface.
(the_low_target): Install arm_supports_z_point_type.
* linux-crisv32-low.c (cris_supports_z_point_type): New function.
(cris_insert_point, cris_remove_point): Adjust to new interface.
Don't check whether the type is supported here.
(the_low_target): Install cris_supports_z_point_type.
* linux-low.c (linux_supports_z_point_type): New function.
(linux_insert_point, linux_remove_point): Adjust to new interface.
* linux-low.h (struct linux_target_ops) <insert_point,
remove_point>: Take an enum raw_bkpt_type instead of a char. Add
raw_breakpoint pointer parameter.
<supports_z_point_type>: New method.
* linux-mips-low.c (mips_supports_z_point_type): New function.
(mips_insert_point, mips_remove_point): Adjust to new interface.
Use mips_supports_z_point_type.
(the_low_target): Install mips_supports_z_point_type.
* linux-ppc-low.c (the_low_target): Install NULL as
supports_z_point_type method.
* linux-s390-low.c (the_low_target): Install NULL as
supports_z_point_type method.
* linux-sparc-low.c (the_low_target): Install NULL as
supports_z_point_type method.
* linux-x86-low.c (x86_supports_z_point_type): New function.
(x86_insert_point): Adjust to new insert_point interface. Use
insert_memory_breakpoint. Adjust to new
i386_low_insert_watchpoint interface.
(x86_remove_point): Adjust to remove_point interface. Use
remove_memory_breakpoint. Adjust to new
i386_low_remove_watchpoint interface.
(the_low_target): Install x86_supports_z_point_type.
* lynx-low.c (lynx_target_ops): Install NULL as
supports_z_point_type callback.
* nto-low.c (nto_supports_z_point_type): New.
(nto_insert_point, nto_remove_point): Adjust to new interface.
(nto_target_ops): Install nto_supports_z_point_type.
* mem-break.c: Adjust intro comment.
(struct raw_breakpoint) <raw_type, size>: New fields.
<inserted>: Update comment.
<shlib_disabled>: Delete field.
(enum bkpt_type) <gdb_breakpoint>: Delete value.
<gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2,
gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values.
(raw_bkpt_type_to_target_hw_bp_type): New function.
(find_enabled_raw_code_breakpoint_at): New function.
(find_raw_breakpoint_at): New type and size parameters. Use them.
(insert_memory_breakpoint): New function, based off
set_raw_breakpoint_at.
(remove_memory_breakpoint): New function.
(set_raw_breakpoint_at): Reimplement.
(set_breakpoint): New, based on set_breakpoint_at.
(set_breakpoint_at): Reimplement.
(delete_raw_breakpoint): Go through the_target->remove_point
instead of assuming memory breakpoints.
(find_gdb_breakpoint_at): Delete.
(Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions.
(find_gdb_breakpoint): New function.
(set_gdb_breakpoint_at): Delete.
(z_type_supported): New function.
(set_gdb_breakpoint_1): New function, loosely based off
set_gdb_breakpoint_at.
(check_gdb_bp_preconditions, set_gdb_breakpoint): New functions.
(delete_gdb_breakpoint_at): Delete.
(delete_gdb_breakpoint_1): New function, loosely based off
delete_gdb_breakpoint_at.
(delete_gdb_breakpoint): New function.
(clear_gdb_breakpoint_conditions): Rename to ...
(clear_breakpoint_conditions): ... this. Don't handle a NULL
breakpoint.
(add_condition_to_breakpoint): Make static.
(add_breakpoint_condition): Take a struct breakpoint pointer
instead of an address. Adjust.
(gdb_condition_true_at_breakpoint): Rename to ...
(gdb_condition_true_at_breakpoint_z_type): ... this, and add
z_type parameter.
(gdb_condition_true_at_breakpoint): Reimplement.
(add_breakpoint_commands): Take a struct breakpoint pointer
instead of an address. Adjust.
(gdb_no_commands_at_breakpoint): Rename to ...
(gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type
parameter. Return true if no breakpoint was found. Change debug
output.
(gdb_no_commands_at_breakpoint): Reimplement.
(run_breakpoint_commands): Rename to ...
(run_breakpoint_commands_z_type): ... this. Add z_type parameter,
and change return type to boolean.
(run_breakpoint_commands): New function.
(gdb_breakpoint_here): Also check for Z1 breakpoints.
(uninsert_raw_breakpoint): Don't try to reinsert a disabled
breakpoint. Go through the_target->remove_point instead of
assuming memory breakpoint.
(uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert
software and hardware breakpoints.
(reinsert_raw_breakpoint): Go through the_target->insert_point
instead of assuming memory breakpoint.
(reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert
software and hardware breakpoints.
(check_breakpoints, breakpoint_here, breakpoint_inserted_here):
Check both software and hardware breakpoints.
(validate_inserted_breakpoint): Assert the breakpoint is a
software breakpoint. Set the inserted flag to -1 instead of
setting shlib_disabled.
(delete_disabled_breakpoints): Adjust.
(validate_breakpoints): Only validate software breakpoints.
Adjust to inserted flag change.
(check_mem_read, check_mem_write): Skip breakpoint types other
than software breakpoints. Adjust to inserted flag change.
* mem-break.h (enum raw_bkpt_type): New enum.
(raw_breakpoint, struct process_info): Forward declare.
(Z_packet_to_target_hw_bp_type): Delete declaration.
(raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type)
(set_gdb_breakpoint, delete_gdb_breakpoint)
(clear_breakpoint_conditions): New declarations.
(set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete.
(breakpoint_inserted_here): Update comment.
(add_breakpoint_condition, add_breakpoint_commands): Replace
address parameter with a breakpoint pointer parameter.
(gdb_breakpoint_here): Update comment.
(delete_gdb_breakpoint_at): Delete.
(insert_memory_breakpoint, remove_memory_breakpoint): Declare.
* server.c (process_point_options): Take a struct breakpoint
pointer instead of an address. Adjust.
(process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and
delete_gdb_breakpoint.
* spu-low.c (spu_target_ops): Install NULL as
supports_z_point_type method.
* target.h: Include mem-break.h.
(struct target_ops) <prepare_to_access_memory>: Update comment.
<supports_z_point_type>: New field.
<insert_point, remove_point>: Take an enum raw_bkpt_type argument
instead of a char. Also take a raw breakpoint pointer.
* win32-arm-low.c (the_low_target): Install NULL as
supports_z_point_type.
* win32-i386-low.c (i386_supports_z_point_type): New function.
(i386_insert_point, i386_remove_point): Adjust to new interface.
(the_low_target): Install i386_supports_z_point_type.
* win32-low.c (win32_supports_z_point_type): New function.
(win32_insert_point, win32_remove_point): Adjust to new interface.
(win32_target_ops): Install win32_supports_z_point_type.
* win32-low.h (struct win32_target_ops):
<supports_z_point_type>: New method.
<insert_point, remove_point>: Take an enum raw_bkpt_type argument
instead of a char. Also take a raw breakpoint pointer.
gdb/testsuite/
2014-05-20 Pedro Alves <palves@redhat.com>
* gdb.base/break-idempotent.c: New file.
* gdb.base/break-idempotent.exp: New file.
When using a reverse execution command without execution history, GDB
might end up in a state where replaying has been started but remains
at the current instruction. This state is illegal.
Do not step if there is no execution history to avoid this.
2014-05-20 Markus Metzger <markus.t.metzger@intel.com>
* record-btrace.c (record_btrace_step_thread): Check for empty history.
testsuite/
* gdb.btrace/nohist.exp: New.
Hi,
This patch is to add a new board setting gdb_reverse_timeout, which is
used to set timeout for all gdb.reverse test cases, which are usually
very slow and cause some TIMEOUT failures, for example, on some arm
boards. We have some alternatives to this approach, but I am not
satisfied with them:
- Increase the timeout value. This is the global change, and it may
cause some delay where actual failures happen.
- Set timeout by gdb_reverse_timeout in every gdb.reverse/*.exp.
Then, we have to touch every file under gdb.reverse.
In this patch, we choose a central place to set timeout for all tests
in gdb.reverse, which is convenient.
gdb/testsuite:
2014-05-20 Yao Qi <yao@codesourcery.com>
* lib/gdb.exp (gdb_init): Set timeout if test file is under
gdb.reverse directory and gdb_reverse_timeout exists in board
setting.
* README: Document gdb_reverse_timeout.
The argument ARGS of gdb_init is passed from dejagnu is a string, the
test file name. In dejagnu/runtest.exp:
proc runtest { test_file_name } {
....
....
if [info exists tool] {
if { [info procs "${tool}_init"] != "" } {
${tool}_init $test_file_name;
}
}
....
}
but inn default_gdb_init (callee of gdb_init), we have
set gdb_test_file_name [file rootname [file tail [lindex $args 0]]]
In tcl, all actual arguments are combined to a list and assigned to
args. This code here isn't wrong, but unnecessary, because its caller
(proc runtest) only passes one string to it, and IMO, we don't need
such tricky tcl "args".
I doubt that "[lindex $args 0]" is to be backward compatible with old
dejagnu, but dejagnu-1.4 release started to pass $test_file_name to
${too}_init, as I showed above. dejagnu-1.4 was released in 2001, and
it should be old enough. I also tried to check whether gdb testusite
works with dejagnu-1.3 or not, but failed to build dejagnu-1.3 on my
machine. Supposing GDB testsuite requires at least dejagnu-1.4, this
change should be safe.
This patch is update default_gdb_init to treat ARGS as a string instead
of a list. Then, 'args' sounds like a list, and this patch also renames
it by 'test_file_name', to align with dejagnu.
gdb/testsuite:
2014-05-20 Yao Qi <yao@codesourcery.com>
* lib/gdb.exp (default_gdb_init): Rename argument 'args' by
'test_file_name'. Treat args as a string instead of a list.
(gdb_init): Rename argument 'args' by 'test_file_name'.
gdb/testsuite/
2014-05-16 Pedro Alves <palves@redhat.com>
* lib/mi-support.exp (mi_expect_stop): On timeout, say "timeout"
instead of "unknown output after running".
Some gdb.dwarf2/*.exp tests copy file1.txt to host via gdb_remote_download
but dw2-filename.exp and dw2-anonymous-func.exp don't do that. Looks like
an oversight in this patch
https://sourceware.org/ml/gdb-patches/2013-08/msg00365.html
There are some fails in remote host testing.
FAIL: gdb.dwarf2/dw2-anonymous-func.exp: list file1.txt
FAIL: gdb.dwarf2/dw2-filename.exp: interpreter-exec mi -file-list-exec-source-files
FAIL: gdb.dwarf2/dw2-filename.exp: info sources
This patch is to invoke gdb_remote_download to copy file1.txt to host
and remote it at the end. This patch fixes these fails above.
gdb/testsuite:
2014-05-16 Yao Qi <yao@codesourcery.com>
* gdb.dwarf2/dw2-filename.exp: Copy file1.txt to host. Remove
file1.txt from host at the end.
* gdb.dwarf2/dw2-anonymous-func.exp: Likewise.
Since we're not compiling with gcc, we don't know where the
DWO file will ultimately be built. It could be built in
testsuite/gdb.dwarf2/foo (non-parallel mode) or
testsuite/outputs/gdb.dwarf2/foo (parallel mode).
* gdb.dwarf2/fission-reread.S: Remove directory from .dwo file path.
* gdb.dwarf2/fission-reread.exp: Set debug-file-directory before
loading file. Add test for TU lookup.
Passing arguments did not work when use_mi_command was set.
gdb/testsuite/ChangeLog:
2014-05-13 Simon Marchi <simon.marchi@ericsson.com>
* lib/mi-support.exp (mi_run_cmd_full): Set arguments by calling
"-exec-arguments" or "set args" before running the inferior.
The message displayed by gdb is different when the inferior exits with
zero and non-zero values, this fix takes that into account.
gdb/testsuite/ChangeLog:
2014-05-13 Simon Marchi <simon.marchi@ericsson.com>
* lib/mi-support.exp (mi_expect_stop): Expect message for
inferiors that exit with non-zero exit code.
This patch fixes mi-file.exp fails on remote host. First, we can't
assume ${srcdir}/${subdir}/${srcfile} directory exists on remote host,
so this patch changes it to match ${srcfile} only on remote host.
Second, regexp pattern ".*/${srcfile}" isn't friendly to Windows path.
The file name is "basics.c" in my test env and can't match the pattern
due to "/" in it. Remove "/" from the pattern.
gdb/testsuite:
2014-05-14 Yao Qi <yao@codesourcery.com>
* gdb.mi/mi-file.exp (test_file_list_exec_source_file): Don't
match absolute path on remote host.
(test_file_list_exec_source_files): Remove "/" from the
pattern.
After I run test like this,
$ make check RUNTESTFLAGS='--host_board=local-remote-host dw2-basic.exp'
gdb.dwarf2/file1.txt in source tree was removed. In some gdb.dwarf2/*.exp,
file1.txt is copied to host and then removed at the end. However, in
local-remote-host-notty.exp, ${board}_download doesn't copy the file but
return the absolute path of the src file. 'remote_file host delete' at
the end will remove the file in source tree.
This patch is to overwrite ${board}_file, and specially make "delete"
option do nothing. This approach is used in gdbserver-base.exp and
remote-stdio-gdbserver.exp too.
gdb/testsuite:
2014-05-14 Yao Qi <yao@codesourcery.com>
* boards/local-remote-host-notty.exp (${board}_file): New
proc.
gdb/Changelog:
* aarch64-tdep.c (aarch64_software_single_step): New function.
(aarch64_gdbarch_init): Handle single stepping of atomic sequences
with aarch64_software_single_step.
gdb/testsuite/ChangeLog:
* gdb.arch/aarch64-atomic-inst.c: New file.
* gdb.arch/aarch64-atomic-inst.exp: New file.
Hi,
I recently see the fail in dwzbuildid.exp below on some targets,
(gdb) print the_int
No symbol "the_int" in current context.
(gdb) FAIL: gdb.dwarf2/dwzbuildid.exp: mismatch: print the_int
Looks the pattern expects to see "No symbol table is loaded", which
is emitted in c-exp.y,
variable: name_not_typename
....
if (msymbol.minsym != NULL)
write_exp_msymbol (pstate, msymbol);
else if (!have_full_symbols () && !have_partial_symbols ())
error (_("No symbol table is loaded. Use the \"file\" command."));
else
error (_("No symbol \"%s\" in current context."),
copy_name ($1.stoken));
it is expected to have no full symbols nor partial symbols, but something
brings full symbols or partial symbols in. I added "maint info symtabs"
and "maint info psymtabs" in dwzbuildid.exp, and it shows symbols are from
ld.so, which has debug information. Then, I reproduce the fail like this,
$ make check RUNTESTFLAGS="CFLAGS_FOR_TARGET='-Wl,-rpath=${glibc_build}:${glibc_build}/math -Wl,--dynamic-linker=${glibc_build}/elf/ld.so' dwzbuildid.exp"
${glibc_build} is the glibc build tree. Debug information is not striped,
so the test fail. However, if I strip debug information from libc.so, libm.so
and ld.so. The test passes.
This patch is to relax the pattern to match the both cases that glibc build
has and has not debug information.
gdb/testsuite:
2014-05-07 Yao Qi <yao@codesourcery.com>
* gdb.dwarf2/dwzbuildid.exp: Match output "No symbol "the_int" in
current context" too.
2014-05-05 Keith Seitz <keiths@redhat.com>
* linespec.c (linespec_parse_basic): Run cleanups if a convenience
variable or history value is successfully parsed.
2014-05-05 Keith Seitz <keiths@redhat.com>
* gdb.linespec/ls-dollar.exp: Add test for linespec
file:convenience_variable.
In gdb.trace/unavailable.exp, an action is defined to collect
struct_b.struct_a.array[2] and struct_b.struct_a.array[100],
struct StructB
{
int d, ef;
StructA struct_a;
int s:1;
static StructA static_struct_a;
const char *string;
};
and the other files are not collected.
When GDB examine traceframe collected by the action, "struct_b" is
unavailable completely, which is wrong.
(gdb) p struct_b
$1 = <unavailable>
When GDB reads 'struct_b', it will request to read memory at struct_b's address
of length LEN. Since struct_b.d is not collected, no 'M' block
includes the first part of the desired range, so tfile_xfer_partial returns
TARGET_XFER_UNAVAILABLE and GDB thinks the whole requested range is unavailable.
In order to fix this problem, in the iteration to 'M' blocks, we record the
lowest address of blocks within the request range. If it has, the requested
range isn't unavailable completely. This applies to ctf too. With this patch
applied, the result looks good and fails in unavailable.exp is fixed.
(gdb) p struct_b
$1 = {d = <unavailable>, ef = <unavailable>, struct_a = {a = <unavailable>, b = <unavailable>, array = {<unavailable>,
<unavailable>, -1431655766, <unavailable> <repeats 97 times>, -1431655766, <unavailable> <repeats 9899 times>}, ptr = <unavailable>, bitfield = <unavailable>}, s = <unavailable>, static static_struct_a = {a = <unavailable>, b = <unavailable>, array = {<unavailable> <repeats 10000 times>}, ptr = <unavailable>,
bitfield = <unavailable>}, string = <unavailable>}
gdb:
2014-05-05 Yao Qi <yao@codesourcery.com>
Pedro Alves <palves@redhat.com>
* tracefile-tfile.c (tfile_xfer_partial): Record the lowest
address of blocks that intersects the requested range. Trim
LEN up to LOW_ADDR_AVAILABLE if read from executable read-only
sections.
* ctf.c (ctf_xfer_partial): Likewise.
gdb/testsuite:
2014-05-05 Yao Qi <yao@codesourcery.com>
* gdb.trace/unavailable.exp (gdb_collect_args_test): Save
traceframes into tfile and ctf trace files. Read data from
trace file and test collected data.
(gdb_collect_locals_test): Likewise.
(gdb_unavailable_registers_test): Likewise.
(gdb_unavailable_floats): Likewise.
(gdb_collect_globals_test): Likewise.
(top-level): Append "ctf" to trace_file_targets if GDB
supports.
This patch moves traceframe checking code out of traceframe generation,
so that we can generation traceframe once, and do the checking in multiple
times (with target remote, tfile and ctf respectively). This is a
pure refactor, not functional changes in unavailable.exp.
gdb/testsuite:
2014-05-05 Yao Qi <yao@codesourcery.com>
* gdb.trace/unavailable.exp (gdb_collect_args_test): Move some
code to ...
(gdb_collect_args_test_1): ... it. New proc.
(gdb_collect_locals_test): Move some code to ...
(gdb_collect_locals_test_1): ... it. New proc.
(gdb_unavailable_registers_test): Move some code to ...
(gdb_unavailable_registers_test_1): ... it. New proc.
(gdb_unavailable_floats): Move some code to ...
(gdb_unavailable_floats_1): ... it. New proc.
This commit is actually an update to make the parser in
gdb/stap-probe.c be aware of all the possible prefixes that a probe
argument can have. According to the section "Argument Format" in:
<https://sourceware.org/systemtap/wiki/UserSpaceProbeImplementation>
The bitness of the arguments can be 8, 16, 32 or 64 bits, signed or
unsigned. Currently GDB recognizes only 32 and 64-bit arguments.
This commit extends this. It also provides a testcase, only for
x86_64 systems.
gdb/
2014-05-02 Sergio Durigan Junior <sergiodj@redhat.com>
* stap-probe.c (enum stap_arg_bitness): New enums to represent 8
and 16-bit signed and unsigned arguments. Update comment.
(stap_parse_probe_arguments): Extend code to handle such
arguments. Use warning instead of complaint to notify about
unrecognized bitness.
gdb/testsuite/
2014-05-02 Sergio Durigan Junior <sergiodj@redhat.com>
* gdb.arch/amd64-stap-optional-prefix.S (main): Add several
probes to test for bitness recognition.
* gdb.arch/amd64-stap-optional-prefix.exp
(test_probe_value_without_reg): New procedure.
Add code to test for different kinds of bitness.
This commit fixes PR breakpoints/16889, which is about a bug that
triggers when GDB tries to parse probes whose arguments do not contain
the initial (and optional) "N@" part. For reference sake, the de
facto format is described here:
<https://sourceware.org/systemtap/wiki/UserSpaceProbeImplementation>
Anyway, this PR actually uncovered two bugs (related) that were
happening while parsing the arguments. The first one was that the
parser *was* catching *some* arguments that were missing the "N@"
part, but it wasn't correctly setting the argument's type. This was
causing a NULL pointer being dereferenced, ouch...
The second bug uncovered was that the parser was not catching all of
the cases for a probe which did not provide the "N@" part. The fix
for that was to simplify the check that the code was making to
identify non-prefixed probes. The code is simpler and easier to read
now.
I am also providing a testcase for this bug, only for x86_64
architectures.
gdb/
2014-05-02 Sergio Durigan Junior <sergiodj@redhat.com>
PR breakpoints/16889
* stap-probe.c (stap_parse_probe_arguments): Simplify
check for non-prefixed probes (i.e., probes whose
arguments do not start with "N@"). Always set the
argument type to a sane value.
gdb/testsuite/
2014-05-02 Sergio Durigan Junior <sergiodj@redhat.com>
PR breakpoints/16889
* gdb.arch/amd64-stap-optional-prefix.S: New file.
* gdb.arch/amd64-stap-optional-prefix.exp: Likewise.
In a test I was writting, I needed a procedure that would connect to
the target, and do "load", or equivalent.
Years ago, boards would override gdb_load to implement that. Then
gdb_reload was added, and gdb_load was relaxed to allow boards avoid
the spawing and connecting to the target. This sped up gdbserver
testing. See
https://www.sourceware.org/ml/gdb-patches/2007-02/msg00318.html.
To actually spawn the target and load the executable on the target
side, gdb_reload was born:
# gdb_reload -- load a file into the target. Called before "running",
# either the first time or after already starting the program once,
# for remote targets. Most files that override gdb_load should now
# override this instead.
proc gdb_reload { } {
# For the benefit of existing configurations, default to gdb_load.
# Specifying no file defaults to the executable currently being
# debugged.
return [gdb_load ""]
}
Note the comment about specifying no file. Indeed looking at
config/sid.exp, or config/monitor.exp, we see examples of that.
However, the default gdb_load itself doesn't handle the case of no
file specified. When passed no file, it just calls gdb_file_cmd with
no file either, which ends up invocing the "file" command with no
argument, which means unloading the file and its symbols... That
means calling gdb_reload when testing against native targets is
broken. We don't see that today because the only call to gdb_reload
that exists today is guarded by target_info exists
gdb,do_reload_on_run.
The native-extended-gdbserver.exp board is likewise broken here. When
[gdb_load ""] is called, the board sets the remote exec-file to "" ...
Tested on x86_64 Fedora 17, native, remote gdbserver and
extended-remote gdbserver.
testsuite/
2014-05-01 Pedro Alves <palves@redhat.com>
* lib/gdb.exp (gdb_load): Extend comment. Skip calling
gdb_file_cmd if no file is specified.
* boards/native-extended-gdbserver.exp (gdb_load): Use the
last_loaded_file to set the remote exec-file.
This adds a variant of local-remote-host-notty.exp that forces
pseudo-tty allocation, so that readline/editing is enabled.
$ ssh localhost gdb -q
(gdb) show editing
Editing of command lines as they are typed is off.
(gdb)
vs:
$ ssh -t localhost gdb -q
(gdb) show editing
Editing of command lines as they are typed is on.
We now get, e.g.:
Running ../../../src/gdb/testsuite/gdb.base/filesym.exp ...
PASS: gdb.base/filesym.exp: complete on "filesy"
PASS: gdb.base/filesym.exp: completion list for "filesym"
PASS: gdb.base/filesym.exp: set breakpoint at filesym
gdb/testsuite/
2014-05-01 Pedro Alves <palves@redhat.com>
* boards/local-remote-host.exp: New file.
When testing with this board, stdin is not a tty, and so
readline/editing is disabled:
$ ssh localhost gdb -q
(gdb) show editing
Editing of command lines as they are typed is off.
(gdb)
Rename the file, to make room for a version of this board that forces a pseudo-tty.
gdb/testsuite/
2014-05-01 Pedro Alves <palves@redhat.com>
* boards/local-remote-host.exp: Rename to ...
* boards/local-remote-host-notty.exp: ... this.
This add a testcases that verifies correct handling of dynamicity
for lower bounds of arrays.
gdb/testsuite/ChangeLog:
* gdb.ada/dyn_arrayidx: New testcase.
In Dwarf::assemble in dwz.exp, 10 is hard-coded in it,
subprogram {
{name main}
{low_pc main addr}
{high_pc "main + 10" addr}
}
however, the length of main function varies on architectures. The
hard-coded 10 here causes dwz.exp fails on some targets, such as
nios2.
This patch is to add some code to compute the length of function main,
which is similar to what we are doing in entry-values.exp.
gdb/testsuite:
2014-04-26 Yao Qi <yao@codesourcery.com>
* gdb.dwarf2/dwz.exp: Compile main.c to object. Restart GDB
and compute the length of function main. Save it in
$main_length.
(Dwarf::assemble): Use $main_length instead of hard-coded 10.
(top-level): Use gdb_compile to compile objects into
executable and restart GDB. Remove invocation to
prepare_for_testing.
On Linux, we need to explicitly ptrace attach to all lwps of a
process. Because GDB might not be connected yet when an attach is
requested, and thus it may not be possible to activate thread_db, as
that requires access to symbols (IOW, gdbserver --attach), a while ago
we make linux_attach loop over the lwps as listed by /proc/PID/task to
find the lwps to attach to.
linux_attach_lwp_1 has:
...
if (initial)
/* If lwp is the tgid, we handle adding existing threads later.
Otherwise we just add lwp without bothering about any other
threads. */
ptid = ptid_build (lwpid, lwpid, 0);
else
{
/* Note that extracting the pid from the current inferior is
safe, since we're always called in the context of the same
process as this new thread. */
int pid = pid_of (current_inferior);
ptid = ptid_build (pid, lwpid, 0);
}
That "safe" comment referred to linux_attach_lwp being called by
thread-db.c. But this was clearly missed when a new call to
linux_attach_lwp_1 was added to linux_attach. As a result,
current_inferior will be set to some random process, and non-initial
lwps of the second inferior get assigned the pid of the wrong
inferior. E.g., in the case of attaching to two inferiors, for the
second inferior (and so on), non-initial lwps of the second inferior
get assigned the pid of the first inferior. This doesn't trigger on
the first inferior, when current_inferior is NULL, add_thread switches
the current inferior to the newly added thread.
Rather than making linux_attach switch current_inferior temporarily
(thus avoiding further reliance on global state), or making
linux_attach_lwp_1 get the tgid from /proc, which add extra syscalls,
and will be wrong in case of the user having originally attached
directly to a non-tgid lwp, and then that lwp spawning new clones (the
ptid.pid field of further new clones should be the same as the
original lwp's pid, which is not the tgid), we note that callers of
linux_attach_lwp/linux_attach_lwp_1 always have the right pid handy
already, so they can pass it down along with the lwpid.
The only other reason for the "initial" parameter is to error out
instead of warn in case of attach failure, when we're first attaching
to a process. There are only three callers of
linux_attach_lwp/linux_attach_lwp_1, and each wants to print a
different warn/error string, so we can just move the error/warn out of
linux_attach_lwp_1 to the callers, thus getting rid of the "initial"
parameter.
There really nothing gdbserver-specific about attaching to two
threaded processes, so this adds a new test under gdb.multi/. The
test passes cleanly against the native GNU/Linux target, but
fails/triggers the bug against GDBserver (before the patch), with the
native-extended-remote board (as plain remote doesn't support
multi-process).
Tested on x86_64 Fedora 17, with the native-extended-gdbserver board.
gdb/gdbserver/
2014-04-25 Pedro Alves <palves@redhat.com>
PR server/16255
* linux-low.c (linux_attach_fail_reason_string): New function.
(linux_attach_lwp): Delete.
(linux_attach_lwp_1): Rename to ...
(linux_attach_lwp): ... this. Take a ptid instead of a pid as
argument. Remove "initial" parameter. Return int instead of
void. Don't error or warn here.
(linux_attach): Adjust to call linux_attach_lwp. Call error on
failure to attach to the tgid. Call warning when failing to
attach to an lwp.
* linux-low.h (linux_attach_lwp): Take a ptid instead of a pid as
argument. Remove "initial" parameter. Return int instead of
void. Don't error or warn here.
(linux_attach_fail_reason_string): New declaration.
* thread-db.c (attach_thread): Adjust to linux_attach_lwp's
interface change. Use linux_attach_fail_reason_string.
gdb/
2014-04-25 Pedro Alves <palves@redhat.com>
PR server/16255
* common/linux-ptrace.c (linux_ptrace_attach_warnings): Rename to ...
(linux_ptrace_attach_fail_reason): ... this. Remove "warning: "
and newline from built string.
* common/linux-ptrace.h (linux_ptrace_attach_warnings): Rename to ...
(linux_ptrace_attach_fail_reason): ... this.
* linux-nat.c (linux_nat_attach): Adjust to use
linux_ptrace_attach_fail_reason.
gdb/testsuite/
2014-04-25 Simon Marchi <simon.marchi@ericsson.com>
Pedro Alves <palves@redhat.com>
PR server/16255
* gdb.multi/multi-attach.c: New file.
* gdb.multi/multi-attach.exp: New file.
For several RSP packets, there's a corresponding "set remote
foo-packet on/off/auto" command that one can use do bypass
auto-detection of support for the packet or feature. However, I
noticed that setting several of these commands to 'on' or 'off'
doesn't actually have any effect. These are, at least:
set remote breakpoint-commands-packet
set remote conditional-breakpoints-packet
set remote fast-tracepoints-packet
set remote static-tracepoints-packet
set remote install-in-trace-packet
These are commands that control a remote protocol feature that doesn't
have a corresponding regular packet, and because of that we cache the
knowledge of the remote side support as returned by the qSupported
packet in the remote_state object.
E.g., in the case of the 'set remote breakpoint-commands-packet'
command, whether the feature is supported is recorded in the
'breakpoint_commands' field of the remote_state object.
Whether to bypass packet support auto-detection or not is controlled
by the 'detect' field of the corresponding packet's packet_config
structure. That field is the variable associated directly with the
"set remote foo-packet" command. Actual remote stub support for the
packet (or feature) is recorded in the 'support' field of the same
structure.
However, when the user toggles the command, the 'support' field is
also correspondingly updated to PACKET_ENABLE/DISABLE/SUPPORT_UNKNOWN,
discarding the knowledge of whether the target actually supports the
feature. If one toggles back to 'auto', it's no big issue for real
packets, as they'll just end up re-probed the next time they might be
necessary. But features whose support is only reported through
qSupported don't get their corresponding (manually added/maintained)
fields in remote_state objected updated. As we lost the actual status
of the target support for the feature, GDB would need to probe the
qSupported features again, which GDB doesn't do.
But we can avoid that extra traffic, and clean things up, IMO.
Instead of going in that direction, this patch completely decouples
struct packet_config's 'detect' and 'support' fields. E.g., when the
user does "set remote foo-packet off", instead of setting the packet
config's 'support' field to PACKET_DISABLE, the 'support' field is not
touched at all anymore. That is, we end up respecting this simple
table:
| packet_config->detect | packet_config->support | should use packet/feature? |
|-----------------------+------------------------+----------------------------|
| auto | PACKET_ENABLE | PACKET_ENABLE |
| auto | PACKET_DISABLE | PACKET_DISABLE |
| auto | PACKET_UNKNOWN | PACKET_UNKNOWN |
| yes | don't care | PACKET_ENABLE |
| no | don't care | PACKET_DISABLE |
This is implemented by the new packet_support function. With that, we
need to update this pattern throughout:
if (remote_protocol_packets[PACKET_foo].support == PACKET_DISABLE)
to do this instead:
if (packet_support (PACKET_qAttached) == PACKET_DISABLE)
where as mentioned, the packet_support function takes struct
packet_config's 'detect' field into account, like in the table above.
As when the packet is force-disabled or force-enabled, the 'support'
field is just ignored, if the command is set back to auto, we'll
resume respecting whatever the target said it supports. IOW, the end
result is that the 'support' field always represents whether the
target actually supports the packet or not.
After all that, the manually maintained breakpoint_commands and
equivalent fields of struct remote_state can then be eliminated, with
references replaced by checking the result of calling the
packet_support function on the corresponding packet or feature. This
required adding new PACKET_foo enum values for several features that
didn't have it yet. (The patch does not add corresponding "set remote
foo-packet" style commands though, focusing only on bug fixing and
laying the groundwork).
Tested on x86_64 Fedora 17, native GDBserver. The new tests all fail
without this patch.
gdb/
2014-04-25 Pedro Alves <palves@redhat.com>
* remote.c (struct remote_state): Remove multi_process_aware,
non_stop_aware, cond_tracepoints, cond_breakpoints,
breakpoint_commands, fast_tracepoints, static_tracepoints,
install_in_trace, disconnected_tracing,
enable_disable_tracepoints, string_tracing, and
augmented_libraries_svr4_read fields.
(remote_multi_process_p): Move further below in the file.
(struct packet_config): Add comments.
(update_packet_config): Delete function.
(show_packet_config_cmd): Use packet_config_support.
(add_packet_config_cmd): Use NULL as set callback.
(packet_ok): "set remote foo-packet"-style commands no longer
change config->supported -- adjust.
(PACKET_ConditionalTracepoints, PACKET_ConditionalBreakpoints)
(PACKET_BreakpointCommands, PACKET_FastTracepoints)
(PACKET_StaticTracepoints, PACKET_InstallInTrace): Add comments.
(PACKET_QNonStop, PACKET_multiprocess_feature)
(PACKET_EnableDisableTracepoints_feature, PACKET_tracenz_feature)
(PACKET_DisconnectedTracing_feature)
(PACKET_augmented_libraries_svr4_read_feature): New enum values.
(set_remote_protocol_packet_cmd): Delete function.
(packet_config_support, packet_support): New functions.
(set_remote_protocol_Z_packet_cmd): Don't call
update_packet_config.
(remote_query_attached, remote_pass_signals)
(remote_program_signals, remote_threads_info)
(remote_threads_extra_info, remote_start_remote): Use
packet_support.
(remote_start_remote): Use packet_config_support and
packet_support.
(init_all_packet_configs): Set all packets to unknown support,
instead of calling update_packet_config.
(remote_check_symbols): Use packet_support.
(remote_supported_packet): Unconditionally set the packet config's
support status.
(remote_multi_process_feature, remote_non_stop_feature)
(remote_cond_tracepoint_feature, remote_cond_breakpoint_feature)
(remote_breakpoint_commands_feature)
(remote_fast_tracepoint_feature, remote_static_tracepoint_feature)
(remote_install_in_trace_feature)
(remote_disconnected_tracing_feature)
(remote_enable_disable_tracepoint_feature)
(remote_string_tracing_feature)
(remote_augmented_libraries_svr4_read_feature): Delete functions.
(remote_protocol_features): Adjust to use remote_supported_packet
for "augmented-libraries-svr4-read", "multiprocess", "QNonStop",
"ConditionalTracepoints", "ConditionalBreakpoints",
"BreakpointCommands", "FastTracepoints", "StaticTracepoints",
"InstallInTrace", "DisconnectedTracing", "DisconnectedTracing",
"EnableDisableTracepoints", and "tracenz".
(remote_query_supported): Use packet_support.
(remote_open_1): Adjust.
(extended_remote_attach_1): Use packet_support. Switch on the
result of packet_ok instead of checking whether the packet ended
up disabled.
(remote_vcont_resume): Use packet_support.
(remote_resume, remote_stop_ns, fetch_register_using_p)
(remote_prepare_to_store, store_register_using_P)
(check_binary_download, remote_write_bytes): Use packet_support.
(remote_vkill): Use packet_support. Switch on the result of
packet_ok instead of checking whether the packet ended up
disabled.
(extended_remote_supports_disable_randomization): Use
packet_support.
(extended_remote_run): Switch on the result of packet_ok instead
of checking whether the packet ended up disabled.
(remote_insert_breakpoint, remote_remove_breakpoint)
(remote_insert_watchpoint, remote_remove_watchpoint)
(remote_insert_hw_breakpoint, remote_remove_hw_breakpoint): Use
packet_support.
(remote_search_memory): Use packet_config_support.
(remote_get_thread_local_address, remote_get_tib_address)
(remote_hostio_send_command, remote_can_execute_reverse): Use
packet_support.
(remote_supports_cond_tracepoints)
(remote_supports_cond_breakpoints)
(remote_supports_fast_tracepoints)
(remote_supports_static_tracepoints)
(remote_supports_install_in_trace)
(remote_supports_enable_disable_tracepoint)
(remote_supports_string_tracing)
(remote_can_run_breakpoint_commands): Rewrite, checking whether
the packet config says the feature is enabled or disabled.
(remote_download_tracepoint, remote_trace_set_readonly_regions)
(remote_get_trace_status): Use packet_support.
(remote_set_disconnected_tracing): Adjust to check whether the
feature is enabled with packet_support.
(remote_set_trace_buffer_size, remote_use_agent)
(remote_can_use_agent, remote_supports_btrace): Use
packet_support.
(remote_enable_btrace, remote_disable_btrace, remote_read_btrace):
Use packet_config_support.
(remote_augmented_libraries_svr4_read): Rewrite, checking whether
the packet config says the feature is enabled or disabled.
(set_range_stepping): Use packet_support.
gdb/testsuite/
2014-04-25 Pedro Alves <palves@redhat.com>
* gdb.base/cond-eval-mode.exp (warning): Move trailing \r\n to
user.
(top level): Test that "set remote conditional-breakpoints-packet
off" works as intended.
* gdb.base/dprintf.exp: Test that "set remote
breakpoint-commands-packet off" works as intended.
* gdb.trace/change-loc.exp (tracepoint_install_in_trace_disabled):
New function.
(top level): Call it.
* gdb.trace/ftrace.exp (test_fast_tracepoints): Test that "set
remote fast-tracepoints-packet off" works as intended.
* gdb.trace/qtro.exp (gdb_is_target_remote): Moved ...
* lib/gdb.exp (gdb_is_target_remote): ... here.
gdb/testsuite/
* gdb.base/catch-syscall.c: Make unreferenced statics non-static to
ensure clang would not discard them.
* gdb.base/gdbvars.c: Ditto.
* gdb.base/memattr.c: Ditto.
* gdb.base/whatis.c: Ditto.
* gdb.python/py-prettyprint.c: Ditto.
* gdb.trace/actions.c: Ditto.
* gdb.cp/ptype-cv-cp.cc: Mark unused global const int as used to
ensure clang would not discard it.
gdb/testsuite/
* gdb.stabs/gdb11479.c (tag_dummy_enum): introduce a variable to cause
clang to emit the full definition of type required by the test
* gdb.stabs/gdb11479.exp (do_test): correct a typo in a test message
gdb/testsuite/
* gdb.cp/pr10728-x.cc: Return by value instead of pointer to coax
Clang into emitting the definition of the type.
* gdb.cp/pr10728-x.h: Ditto.
* gdb.cp/pr10728-y.cc: Ditto.
gdb/testsuite/
* gdb.base/label.exp: XFAIL label related tests under Clang.
* gdb.cp/cplabel.exp: Ditto.
* gdb.linespec/ls-errs.exp: Refactor tests to execute directly
and XFAIL under Clang those using labels.
I happen to see that 'double_label' isn't used in dwz.exp dwarf assembler.
Similarly, partial_label and double_label aren't used in dwzbuildid.exp.
This patch is to remove them.
gdb/testsuite:
2014-04-25 Yao Qi <yao@codesourcery.com>
* gdb.dwarf2/dwz.exp (Dwarf::assemble): Remove unused
double_label.
* gdb.dwarf2/dwzbuildid.exp (Dwarf::assemble): Remove
partial_label and double_label.
gdb/testsuite/
* gdb.cp/cpexprs.cc: Move braces to the same line as the start
of the function to work across GCC and Clang.
* gdb.cp/cpexprs.exp: Account for GCC/Clang difference in vtable
pointer types (const void ** const V void **).
This patch adds support for the Intel(R) Advanced Vector
Extensions 512 (Intel(R) AVX-512) registers. Native and remote
debugging are covered by this patch.
Intel(R) AVX-512 is an extension to AVX to support 512-bit wide
SIMD registers in 64-bit mode (XMM0-XMM31, YMM0-YMM31, ZMM0-ZMM31).
The number of available registers in 32-bit mode is still 8
(XMM0-7, YMM0-7, ZMM0-7). The lower 256-bits of the ZMM registers
are aliased to the respective 256-bit YMM registers. The lower
128-bits are aliased to the respective 128-bit XMM registers.
There are also 8 new, dedicated mask registers (K0-K7) in both 32-bit
mode and 64-bit mode.
For more information please see
Intel(R) Developer Zone: Intel(R) AVX
http://software.intel.com/en-us/intel-isa-extensions#pid-16007-1495
Intel(R) Architecture Instruction Set Extensions Programming Reference:
http://software.intel.com/en-us/file/319433-017pdf
2014-04-24 Michael Sturm <michael.sturm@mintel.com>
Walfred Tedeschi <walfred.tedeschi@intel.com>
* amd64-linux-nat.c (amd64_linux_gregset32_reg_offset): Add
AVX512 registers.
(amd64_linux_read_description): Add code to handle AVX512 xstate
mask and return respective tdesc.
* amd64-linux-tdep.c: Include features/i386/amd64-avx512-linux.c
and features/i386/x32-avx512-linux.c.
(amd64_linux_gregset_reg_offset): Add AVX512 registers.
(amd64_linux_core_read_description): Add code to handle AVX512
xstate mask and return respective tdesc.
(_initialize_amd64_linux_tdep): Initialize AVX512 tdesc.
* amd64-linux-tdep.h (AMD64_LINUX_ORIG_RAX_REGNUM): Adjust regnum
calculation.
(AMD64_LINUX_NUM_REGS): Adjust to new number of registers.
(tdesc_amd64_avx512_linux): New prototype.
(tdesc_x32_avx512_linux): Likewise.
* amd64-tdep.c: Include features/i386/amd64-avx512.c and
features/i386/x32-avx512.c.
(amd64_ymm_avx512_names): New register names for pseudo
registers YMM16-31.
(amd64_ymmh_avx512_names): New register names for raw registers
YMMH16-31.
(amd64_k_names): New register names for K registers.
(amd64_zmmh_names): New register names for ZMM raw registers.
(amd64_zmm_names): New registers names for ZMM pseudo registers.
(amd64_xmm_avx512_names): New register names for XMM16-31
registers.
(amd64_pseudo_register_name): Add code to return AVX512 pseudo
registers.
(amd64_init_abi): Add code to intitialize AVX512 tdep variables
if feature is present.
(_initialize_amd64_tdep): Call AVX512 tdesc initializers.
* amd64-tdep.h (enum amd64_regnum): Add AVX512 registers.
(AMD64_NUM_REGS): Adjust to new number of registers.
* i386-linux-nat.c (GETXSTATEREGS_SUPPLIES): Extend range of
registers supplied via XSTATE by AVX512 registers.
(i386_linux_read_description): Add case for AVX512.
* i386-linux-tdep.c: Include i386-avx512-linux.c.
(i386_linux_gregset_reg_offset): Add AVX512 registers.
(i386_linux_core_read_description): Add case for AVX512.
(i386_linux_init_abi): Install supported register note section
for AVX512.
(_initialize_i386_linux_tdep): Add call to tdesc init function for
AVX512.
* i386-linux-tdep.h (I386_LINUX_NUM_REGS): Set number of
registers to be number of zmm7h + 1.
(tdesc_i386_avx512_linux): Add tdesc for AVX512 registers.
* i386-tdep.c: Include features/i386/i386-avx512.c.
(i386_zmm_names): Add ZMM pseudo register names array.
(i386_zmmh_names): Add ZMM raw register names array.
(i386_k_names): Add K raw register names array.
(num_lower_zmm_regs): Add constant for the number of lower ZMM
registers. AVX512 has 16 more ZMM registers than there are YMM
registers.
(i386_zmmh_regnum_p): Add function to look up register number of
ZMM raw registers.
(i386_zmm_regnum_p): Likewise for ZMM pseudo registers.
(i386_k_regnum_p): Likewise for K raw registers.
(i386_ymmh_avx512_regnum_p): Likewise for additional YMM raw
registers added by AVX512.
(i386_ymm_avx512_regnum_p): Likewise for additional YMM pseudo
registers added by AVX512.
(i386_xmm_avx512_regnum_p): Likewise for additional XMM registers
added by AVX512.
(i386_register_name): Add code to hide YMMH16-31 and ZMMH0-31.
(i386_pseudo_register_name): Add ZMM pseudo registers.
(i386_zmm_type): Construct and return vector registers type for ZMM
registers.
(i386_pseudo_register_type): Return appropriate type for YMM16-31,
ZMM0-31 pseudo registers and K registers.
(i386_pseudo_register_read_into_value): Add code to read K, ZMM
and YMM16-31 registers from register cache.
(i386_pseudo_register_write): Add code to write K, ZMM and
YMM16-31 registers.
(i386_register_reggroup_p): Add code to include/exclude AVX512
registers in/from respective register groups.
(i386_validate_tdesc_p): Handle AVX512 feature, add AVX512
registers if feature is present in xcr0.
(i386_gdbarch_init): Add code to initialize AVX512 feature
variables in tdep structure, wire in pseudo registers and call
initialize_tdesc_i386_avx512.
* i386-tdep.h (struct gdbarch_tdep): Add AVX512 related
variables.
(i386_regnum): Add AVX512 registers.
(I386_SSE_NUM_REGS): New define for number of SSE registers.
(I386_AVX_NUM_REGS): Likewise for AVX registers.
(I386_AVX512_NUM_REGS): Likewise for AVX512 registers.
(I386_MAX_REGISTER_SIZE): Change to 64 bytes, ZMM registers are
512 bits wide.
(i386_xmm_avx512_regnum_p): New prototype for register look up.
(i386_ymm_avx512_regnum_p): Likewise.
(i386_k_regnum_p): Likewise.
(i386_zmm_regnum_p): Likewise.
(i386_zmmh_regnum_p): Likewise.
* i387-tdep.c : Update year in copyright notice.
(xsave_ymm_avx512_offset): New table for YMM16-31 offsets in
XSAVE buffer.
(XSAVE_YMM_AVX512_ADDR): New macro.
(xsave_xmm_avx512_offset): New table for XMM16-31 offsets in
XSAVE buffer.
(XSAVE_XMM_AVX512_ADDR): New macro.
(xsave_avx512_k_offset): New table for K register offsets in
XSAVE buffer.
(XSAVE_AVX512_K_ADDR): New macro.
(xsave_avx512_zmm_h_offset): New table for ZMM register offsets
in XSAVE buffer.
(XSAVE_AVX512_ZMM_H_ADDR): New macro.
(i387_supply_xsave): Add code to supply AVX512 registers to XSAVE
buffer.
(i387_collect_xsave): Add code to collect AVX512 registers from
XSAVE buffer.
* i387-tdep.h (I387_NUM_XMM_AVX512_REGS): New define for number
of XMM16-31 registers.
(I387_NUM_K_REGS): New define for number of K registers.
(I387_K0_REGNUM): New define for K0 register number.
(I387_NUM_ZMMH_REGS): New define for number of ZMMH registers.
(I387_ZMM0H_REGNUM): New define for ZMM0H register number.
(I387_NUM_YMM_AVX512_REGS): New define for number of YMM16-31
registers.
(I387_YMM16H_REGNUM): New define for YMM16H register number.
(I387_XMM16_REGNUM): New define for XMM16 register number.
(I387_YMM0_REGNUM): New define for YMM0 register number.
(I387_KEND_REGNUM): New define for last K register number.
(I387_ZMMENDH_REGNUM): New define for last ZMMH register number.
(I387_YMMH_AVX512_END_REGNUM): New define for YMM31 register
number.
(I387_XMM_AVX512_END_REGNUM): New define for XMM31 register
number.
* common/i386-xstate.h: Add AVX 3.1 feature bits, mask and XSTATE
size.
* features/Makefile: Add AVX512 related files.
* features/i386/32bit-avx512.xml: New file.
* features/i386/64bit-avx512.xml: Likewise.
* features/i386/amd64-avx512-linux.c: Likewise.
* features/i386/amd64-avx512-linux.xml: Likewise.
* features/i386/amd64-avx512.c: Likewise.
* features/i386/amd64-avx512.xml: Likewise.
* features/i386/i386-avx512-linux.c: Likewise.
* features/i386/i386-avx512-linux.xml: Likewise.
* features/i386/i386-avx512.c: Likewise.
* features/i386/i386-avx512.xml: Likewise.
* features/i386/x32-avx512-linux.c: Likewise.
* features/i386/x32-avx512-linux.xml: Likewise.
* features/i386/x32-avx512.c: Likewise.
* features/i386/x32-avx512.xml: Likewise.
* regformats/i386/amd64-avx512-linux.dat: New file.
* regformats/i386/amd64-avx512.dat: Likewise.
* regformats/i386/i386-avx512-linux.dat: Likewise.
* regformats/i386/i386-avx512.dat: Likewise.
* regformats/i386/x32-avx512-linux.dat: Likewise.
* regformats/i386/x32-avx512.dat: Likewise.
* NEWS: Add note about new support for AVX512.
testsuite/
* Makefile.in (EXECUTABLES): Added i386-avx512.
* gdb.arch/i386-avx512.c: New file.
* gdb.arch/i386-avx512.exp: Likewise.
gdbserver/
* Makefile.in: Added rules to handle new files
i386-avx512.c i386-avx512-linux.c amd64-avx512.c
amd64-avx512-linux.c x32-avx512.c x32-avx512-linux.c.
* configure.srv (srv_i386_regobj): Add i386-avx512.o.
(srv_i386_linux_regobj): Add i386-avx512-linux.o.
(srv_amd64_regobj): Add amd64-avx512.o and x32-avx512.o.
(srv_amd64_linux_regobj): Add amd64-avx512-linux.o and
x32-avx512-linux.o.
(srv_i386_32bit_xmlfiles): Add i386/32bit-avx512.xml.
(srv_i386_64bit_xmlfiles): Add i386/64bit-avx512.xml.
(srv_amd64_xmlfiles): Add i386/amd64-avx512.xml and
i386/x32-avx512.xml.
(srv_i386_linux_xmlfiles): Add i386/i386-avx512-linux.xml.
(srv_amd64_linux_xmlfiles): Add i386/amd64-avx512-linux.xml and
i386/x32-avx512-linux.xml.
* i387-fp.c (num_avx512_k_registers): New constant for number
of K registers.
(num_avx512_zmmh_low_registers): New constant for number of
lower ZMM registers (0-15).
(num_avx512_zmmh_high_registers): New constant for number of
higher ZMM registers (16-31).
(num_avx512_ymmh_registers): New contant for number of higher
YMM registers (ymm16-31 added by avx521 on x86_64).
(num_avx512_xmm_registers): New constant for number of higher
XMM registers (xmm16-31 added by AVX512 on x86_64).
(struct i387_xsave): Add space for AVX512 registers.
(i387_cache_to_xsave): Change raw buffer size to 64 characters.
Add code to handle AVX512 registers.
(i387_xsave_to_cache): Add code to handle AVX512 registers.
* linux-x86-low.c (init_registers_amd64_avx512_linux): New
prototypei from generated file.
(tdesc_amd64_avx512_linux): Likewise.
(init_registers_x32_avx512_linux): Likewise.
(tdesc_x32_avx512_linux): Likewise.
(init_registers_i386_avx512_linux): Likewise.
(tdesc_i386_avx512_linux): Likewise.
(x86_64_regmap): Add AVX512 registers.
(x86_linux_read_description): Add code to handle AVX512 XSTATE
mask.
(initialize_low_arch): Add code to initialize AVX512 registers.
doc/
* gdb.texinfo (i386 Features): Add description of AVX512
registers.
Change-Id: Ifc4c08c76b85dbec18d02efdbe6182e851584438
Signed-off-by: Michael Sturm <michael.sturm@intel.com>
breakpoint table handling. This is a patch in five parts (all committed
here in one commit).
----- 1/5: parse_args
parse_args is a very useful utility function which allows you to do
getopt-y kinds of things in Tcl.
Example:
proc myproc {foo args} {
parse_args {{bar} {baz "abc"} {qux}}
# ...
}
myproc ABC -bar -baz DEF peanut butter
will define the following variables in myproc:
foo (=ABC), bar (=1), baz (=DEF), and qux (=0)
args will be the list {peanut butter}
----- 2/5: mi_build_kv_pairs
build_kv_pairs simply does what it says: given the input list
and an option join string, it combines list elements into kv-pairs
for MI handling. It knows how to handle tuples and other special
MI types.
Example:
mi_build_kv_pairs {a b c d e f g \[.*\]}
returns a=\"b\",c=\"d\",e=\"f\",g=\[.*\]
----- 3/5: mi_make_breakpoint
This function builds breakpoint regexps, such as
"bkpt={number=\".*\", [snip]}".
Note that ONLY the options given to mi_make_breakpoint/mi_create_breakpoint
will actually be tested. So if -number is omitted, the regexp will allow
anything [number=\".*\"]
Examples:
mi_make_breakpoint -number 3
mi_create_breakpoint "myfile.c:21" -file myfile.c -line 21
----- 4/5: mi_make_breakpoint_table
This function builds MI breakpoint table regexps.
Example:
set bps {}
lappend bps [mi_make_breakpoint -number 1 -func "main" \
-file ".*/myfile.c" -line 42
lappend bps [mi_make_breakpoint -number 2 -func "marker" \
-file ".*myfile.c" -line 21
gdb_test "-break-info" "\\^done,[mi_make_breakpoint_table $bps]" \
"breakpoint list"
----- 5/5: Update all callers
Self-explanatory
testsuite/ChangeLog
2014-04-23 Keith Seitz <keiths@redhat.com>
* lib/mi-support.exp (mi_list_breakpoints): Delete.
(mi_make_breakpoint_table): New procedure.
(mi_create_breakpoint): Use mi_make_breakpoint
and return the result.
(mi_make_breakpoint): New procedure.
(mi_build_kv_pairs): New procedure.
* gdb.mi/mi-break.exp: Remove unused globals,
update mi_create_breakpoint usage, and use mi_make_breakpoint_table.
All callers updated.
* gdb.mi/mi-dprintf.exp: Use variable to track command
number.
Update all callers of mi_create_breakpoint and use
mi_make_breakpoint_table.
Remove any unused global variables.
* gdb.mi/mi-nonstop.exp: Likewise.
* gdb.mi/mi-nsintrall.exp: Likewise.
* gdb.mi/mi-nsmoribund.exp: Likewise.
* gdb.mi/mi-nsthrexec.exp: Likewise.
* gdb.mi/mi-reverse.exp: Likewise.
* gdb.mi/mi-simplerun.exp: Likewise.
* gdb.mi/mi-stepn.exp: Likewise.
* gdb.mi/mi-syn-frame.exp: Likewise.
* gdb.mi/mi-until.exp: Likewise.
* gdb.mi/mi-var-cp.exp: Likewise.
* gdb.mi/mi-var-display.exp: Likewise.
* gdb.mi/mi2-amd64-entry-value.exp: Likewise.
* gdb.mi/mi2-var-child.exp: Likewise.
* gdb.mi/mi-vla-c99.exp: Likewise.
* lib/mi-support.exp: Likewise.
From Ian Lance Taylor <iant@cygnus.com>:
* lib/gdb.exp (parse_args): New procedure.
Without the code portion of the patch, we get these failures:
FAIL: gdb.base/break-unload-file.exp: always-inserted on: break: continue
FAIL: gdb.base/break-unload-file.exp: always-inserted on: hbreak: continue
FAIL: gdb.base/sym-file.exp: stale bkpts: continue to breakpoint: end here
They all looks like random SIGTRAPs:
continue
Continuing.
Program received signal SIGTRAP, Trace/breakpoint trap.
0x0000000000400541 in foo () at ../../../src/gdb/testsuite/gdb.base/break-unload-file.c:21
21 }
(gdb) FAIL: gdb.base/break-unload-file.exp: always-inserted on: break: continue
(This is a regression caused by the remove-symbol-file command
series.)
break-unload-file.exp is about having breakpoints inserted, and then
doing "file". I caught this while writing a test that does "file
PROGRAM", while PROGRAM was already loaded, which internally does
"file" first, because I wanted to force a breakpoint_re_set, but the
test is more explicit in case GDB ever optimizes out that re-set.
The problem is that unloading the file with "file" ends up in
disable_breakpoints_in_freed_objfile, which marks all breakpoint
locations of the objfile as both shlib_disabled, _and_ clears the
inserted flag, without actually removing the breakpoints from the
inferior. Now, usually, in all-stop, breakpoints will already be
removed from the inferior before the user can issue the "file"
command, but, with non-stop, or breakpoints always-inserted on mode,
breakpoints stay inserted even while the user has the prompt. In the
latter case, then, if we let the program continue, and it executes the
address where we had previously set the breakpoint, it'll actually
execute the breakpoint instruction that we left behind...
Now, one issue is that the intent of
disable_breakpoints_in_freed_objfile is really to handle the unloading
of OBJF_USERLOADED objfiles. These are objfiles that were added with
add-symbol-file and that are removed with remove-symbol-file.
"add-symbol-file"'s docs in the manual clearly say these commands are
used to let GDB know about dynamically loaded code:
You would use this command when @var{filename} has been dynamically
loaded (by some other means) into the program that is running.
Similarly, the online help says:
(gdb) help add-symbol-file
Load symbols from FILE, assuming FILE has been dynamically loaded.
So it makes sense to, like when shared libraries are unloaded through
the generic solib machinery, mark the breakpoint locations as
shlib_disabled. But, the "file" command is not about dynamically
loaded code, it's about the main program. So the patch makes
disable_breakpoints_in_freed_objfile skip all objfiles but
OBJF_USERLOADED ones, thus skipping the main objfile.
Then, the reason that disable_breakpoints_in_freed_objfile was
clearing the inserted flag isn't clear, but likely to avoid breakpoint
removal errors, assuming remove-symbol-file was called after the
dynamic object was already unmapped from the inferior. In that case,
it'd okay to simply clear the inserted flag, but not so if the user
for example does remove-symbol-file to remove the library because he
made a mistake in the library's address, and wants to re-do
add-symbol-file with the correct address.
To address all that, I propose an alternative implementation, that
handles both cases. The patch includes changes to sym-file.exp to
cover them.
This implementation leaves the inserted flag alone, and handles
breakpoint insertion/removal failure gracefully when the locations are
in OBJF_USERLOADED objfiles, just like we handle insertion/removal
failure gracefully for locations in shared libraries.
To try to make sure we aren't patching back stale shadow memory
contents into the inferior, in case the program mapped a different
library at the same address where we had the breakpoint, without the
user having had a chance of remove-symbol-file'ing before, this adds a
new memory_validate_breakpoint function that checks if the breakpoint
instruction is still in memory. ppc_linux_memory_remove_breakpoint
does this unconditionally for all memory breakpoints, and questions
whether memory_remove_breakpoint should be changed to do this for all
breakpoints. Possibly yes, though I'm not certain, hence this
baby-steps patch.
Tested on x86_64 Fedora 17, native and gdbserver.
gdb/
2014-04-23 Pedro Alves <palves@redhat.com>
* breakpoint.c (insert_bp_location): Tolerate errors if the
breakpoint is set in a user-loaded objfile.
(remove_breakpoint_1): Likewise. Also tolerate errors if the
location is marked shlib_disabled. If the breakpoint is set in a
user-loaded objfile is a GDB-side memory breakpoint, validate it
before uninsertion. (disable_breakpoints_in_freed_objfile): Skip
non-OBJF_USERLOADED objfiles. Don't clear the location's inserted
flag.
* mem-break.c (memory_validate_breakpoint): New function.
* objfiles.c (userloaded_objfile_contains_address_p): New
function.
* objfiles.h (userloaded_objfile_contains_address_p): Declare.
* target.h (memory_validate_breakpoint): New declaration.
gdb/testsuite/
2014-04-23 Pedro Alves <palves@redhat.com>
* gdb.base/break-unload-file.c: New file.
* gdb.base/break-unload-file.exp: New file.
* gdb.base/sym-file-lib.c (baz): New function.
* gdb.base/sym-file-loader.c (struct segment) <mapped_size>: New
field.
(load): Store the segment's mapped size.
(unload): New function.
(unload_shlib): New function.
* gdb.base/sym-file-loader.h (unload_shlib): New declaration.
* gdb.base/sym-file-main.c (main): Unload, and reload the library,
set a breakpoint at baz, and call it.
* gdb.base/sym-file.exp: New tests for stale breakpoint
instructions.
libraries.
As explained in
https://sourceware.org/ml/gdb-patches/2008-08/msg00361.html, after a
shared library was unloaded, we can no longer insert or remove
breakpoints into/from its (no longer present) code segment. That'll
fail with memory errors. However, that concern does not apply to
hardware breakpoints. By definition, hardware breakpoints are
implemented using a mechanism that is not dependent on being able to
modify the target's memory. Usually, by setting up CPU debug
registers. IOW, we should be able to set hw breakpoints in an
unmapped address. We don't seem to have a test that exercises that,
so this patch adds one.
I noticed the error supression because of a related issue -- the
target_insert_hw_breakpoint/target_remove_hw_breakpoint interfaces
don't really distinguish "not supported" from "error" return, and so
remote.c returns -1 in both cases. This results in hardware
breakpoints set in shared libraries silently ending up pending forever
even though the target doesn't actually support hw breakpoints.
(gdb) set breakpoint always-inserted on
(gdb) set remote Z-packet off
(gdb) info breakpoints
No breakpoints or watchpoints.
(gdb) hbreak shrfunc
Hardware assisted breakpoint 3 at 0x7ffff7dfb657: file ../../../src/gdb/testsuite/gdb.base/hbreak-in-shr-unsupported-shr.c, line 21.
(gdb) info break
Num Type Disp Enb Address What
3 hw breakpoint keep y <PENDING> shrfunc
After the patch we get the expected:
(gdb) hbreak shrfunc
Hardware assisted breakpoint 3 at 0x7ffff7dfb657: file ../../../src/gdb/testsuite/gdb.base/hbreak-in-shr-unsupported-shr.c, line 21.
Warning:
Cannot insert hardware breakpoint 3.
Could not insert hardware breakpoints:
You may have requested too many hardware breakpoints/watchpoints.
(gdb) info break
Num Type Disp Enb Address What
3 hw breakpoint keep y 0x00007ffff7dfb657 in shrfunc at ../../../src/gdb/testsuite/gdb.base/hbreak-in-shr-unsupported-shr.c:21
(HW breakpoints set in the main executable, when the target doesn't
support HW breakpoints always resulted in the latter output.)
We probably should improve the insert/remove interface to return a
different error code for unsupported. But I chose to fix the error
supression first, as it's a deeper and wider issue.
Tested on x86_64 Fedora 17, native and gdbserver.
gdb/
2014-04-23 Pedro Alves <palves@redhat.com>
* breakpoint.c (insert_bp_location, remove_breakpoint_1): If
the breakpoint is set in a shared library, only suppress
errors for software breakpoints, not hardware breakpoints.
gdb/testsuite/
2014-04-23 Pedro Alves <palves@redhat.com>
* gdb.base/hbreak-in-shr-unsupported-shr.c: New file.
* gdb.base/hbreak-in-shr-unsupported.c: New file.
* gdb.base/hbreak-in-shr-unsupported.exp: New file.
* gdb.base/hbreak-unmapped.c: New file.
* gdb.base/hbreak-unmapped.exp: New file.
* gdb.trace/qtro.exp (gdb_is_target_remote): Move ...
* lib/gdb.exp (gdb_is_target_remote): ... here.
If a thread trips on a breakpoint that needs stepping over just after
finishing a step over, GDB currently fails an assertion. This is a
regression caused by the "Handle multiple step-overs." patch
(99619beac6) at
https://sourceware.org/ml/gdb-patches/2014-02/msg00765.html.
(gdb) x /4i $pc
=> 0x400540 <main+4>: movl $0x0,0x2003da(%rip) # 0x600924 <i>
0x40054a <main+14>: movl $0x1,0x2003d0(%rip) # 0x600924 <i>
0x400554 <main+24>: movl $0x2,0x2003c6(%rip) # 0x600924 <i>
0x40055e <main+34>: movl $0x3,0x2003bc(%rip) # 0x600924 <i>
(gdb) PASS: gdb.base/consecutive-step-over.exp: get breakpoint addresses
break *0x40054a
Breakpoint 2 at 0x40054a: file ../../../src/gdb/testsuite/gdb.base/consecutive-step-over.c, line 23.
(gdb) PASS: gdb.base/consecutive-step-over.exp: insn 1: set breakpoint
condition $bpnum condition
(gdb) PASS: gdb.base/consecutive-step-over.exp: insn 1: set condition
break *0x400554
Breakpoint 3 at 0x400554: file ../../../src/gdb/testsuite/gdb.base/consecutive-step-over.c, line 24.
(gdb) PASS: gdb.base/consecutive-step-over.exp: insn 2: set breakpoint
condition $bpnum condition
(gdb) PASS: gdb.base/consecutive-step-over.exp: insn 2: set condition
break *0x40055e
Breakpoint 4 at 0x40055e: file ../../../src/gdb/testsuite/gdb.base/consecutive-step-over.c, line 25.
(gdb) PASS: gdb.base/consecutive-step-over.exp: insn 3: set breakpoint
condition $bpnum condition
(gdb) PASS: gdb.base/consecutive-step-over.exp: insn 3: set condition
break 27
Breakpoint 5 at 0x400568: file ../../../src/gdb/testsuite/gdb.base/consecutive-step-over.c, line 27.
(gdb) continue
Continuing.
../../src/gdb/infrun.c:5200: internal-error: switch_back_to_stepped_thread: Assertion `!tp->control.trap_expected' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
FAIL: gdb.base/consecutive-step-over.exp: continue to breakpoint: break here (GDB internal error)
The assertion fails, because the code is not expecting that the event
thread itself might need another step over. IOW, not expecting that
TP in:
tp = find_thread_needs_step_over (stepping_thread != NULL,
stepping_thread);
could be the event thread.
A small fix for this would be to clear the event thread's
trap_expected earlier, before asserting. But looking deeper, although
currently_stepping_or_nexting_callback's intention is finding the
thread that is doing a step/next, it also returns the thread that is
doing a step-over dance, with trap_expected set. If there ever was a
reason for that (it was I who added
currently_stepping_or_nexting_callback , but I can't recall why I put
trap_expected there in the first place), the only remaining reason
nowadays is to aid in implementing switch_back_to_stepped_thread's
assertion that is now triggering, by piggybacking on the walk over all
threads, thus avoiding a separate walk. This is quite obscure, and I
think we can do even better, by merging the walks that look for the
stepping thread, and the walk that looks for some thread that might
need a step over.
Tested on x86_64 Fedora 17, native and gdbserver, and also native on
top of my "software single-step on x86_64" series.
gdb/
2014-04-22 Pedro Alves <palves@redhat.com>
* infrun.c (schedlock_applies): New function, factored out from
find_thread_needs_step_over.
(find_thread_needs_step_over): Use it.
(switch_back_to_stepped_thread): Always clear trap_expected if the
step over is finished. Return early if scheduler locking applies.
Look for the stepping thread and a potential step-over thread with
a single loop.
(currently_stepping_or_nexting_callback): Delete.
2014-04-22 Pedro Alves <palves@redhat.com>
* gdb.base/consecutive-step-over.c: New file.
* gdb.base/consecutive-step-over.exp: New file.
This switches the gdb_continue_to_breakpoint routine to use
gdb_test_multiple instead of send_gdb/gdb_expect, so that an internal
error is detected immediately, instead of failing on timeout.
gdb/testsuite/
2014-04-22 Pedro Alves <palves@redhat.com>
* lib/gdb.exp (gdb_continue_to_breakpoint): Use gdb_test_multiple
instead of send_gdb/gdb_expect.
In gdb.trace/tfile.exp, we execute binary to generate tracefile,
remote_exec target "$binfile"
however, this fails on bare metal target. This patch is to
handle binary execution failure by running binary in GDB.
The binary will do some io operation to generate tracefile, so
we need a check 'target_info exists gdb,nofileio'.
This patch is to check whether tracefile is generated. tfile.exp can
be skipped if generation is failed, while test_tfind_tfile in
mi-traceframe-changed.exp is skipped if generated failed. The rest of
the mi-traceframe-changed.exp can still be executed, because on some
bare metal targets, the remote stub supports tracepoint but doesn't
support fileio.
gdb/testsuite:
2014-04-22 Yao Qi <yao@codesourcery.com>
* lib/trace-support.exp (generate_tracefile): New procedure.
* gdb.trace/tfile.exp: Skip the test if generate_tracefile
return 0.
* gdb.trace/mi-traceframe-changed.exp: Invoke test_tfind_tfile
if generate_tracefile returns 1.
This PR is about an assertion failure in GDB that can be triggered by
setting "backtrace limit" to a value that causes GDB to stop unwinding
after an inline frame. In this case, an assertion in
inline_frame_this_id will trigger:
/* We need a valid frame ID, so we need to be based on a valid
frame. (...). */
gdb_assert (frame_id_p (*this_id));
Looking at the function:
static void
inline_frame_this_id (struct frame_info *this_frame,
void **this_cache,
struct frame_id *this_id)
{
struct symbol *func;
/* In order to have a stable frame ID for a given inline function,
we must get the stack / special addresses from the underlying
real frame's this_id method. So we must call get_prev_frame.
Because we are inlined into some function, there must be previous
frames, so this is safe - as long as we're careful not to
create any cycles. */
*this_id = get_frame_id (get_prev_frame (this_frame));
we see we're computing the frame id for the inline frame. If this is
an inline frame, which is a virtual frame constructed based on debug
info, on top of a real stack frame, we should _always_ be able to find
where the frame was inlined into, as that ultimately just means
peeling off the virtual frames on top of the real stack frame. If
there ultimately was no prev (real) stack frame, then we wouldn't have
been able to construct the inline frame either, by design. That's
what the assertion catches.
So we have an inline frame, we should _always_ be able to compute its
ID, even if that means bypassing the user backtrace limits to get at
the real stack frame's info. The problem is that inline_frame_id
calls get_prev_frame, and that takes user backtrace limits into
account. Code that wants to bypass the limits calls get_prev_frame_1
instead.
Note how get_prev_frame_1 already skips all checks for inline frames:
/* If we are unwinding from an inline frame, all of the below tests
were already performed when we unwound from the next non-inline
frame. We must skip them, since we can not get THIS_FRAME's ID
until we have unwound all the way down to the previous non-inline
frame. */
if (get_frame_type (this_frame) == INLINE_FRAME)
return get_prev_frame_if_no_cycle (this_frame);
And note how the related frame_unwind_caller_id function also uses
get_prev_frame_1:
struct frame_id
frame_unwind_caller_id (struct frame_info *next_frame)
{
struct frame_info *this_frame;
/* Use get_prev_frame_1, and not get_prev_frame. The latter will truncate
the frame chain, leading to this function unintentionally
returning a null_frame_id (e.g., when a caller requests the frame
ID of "main()"s caller. */
next_frame = skip_artificial_frames (next_frame);
this_frame = get_prev_frame_1 (next_frame);
if (this_frame)
return get_frame_id (skip_artificial_frames (this_frame));
else
return null_frame_id;
}
get_prev_frame_1 is currently static in frame.c. As a _1 suffix is
not a good name for an extern function, I've renamed it.
Tested on x86-64 Fedora 17.
gdb/
2014-04-18 Pedro alves <palves@redhat.com>
Tom Tromey <tromey@redhat.com>
PR backtrace/15558
* frame.c (get_prev_frame_1): Rename to ...
(get_prev_frame_always): ... this, and make extern. Adjust.
(skip_artificial_frames): Use get_prev_frame_always.
(frame_unwind_caller_id, frame_pop, get_prev_frame)
(get_frame_unwind_stop_reason): Adjust to rename.
* frame.h (get_prev_frame_always): Declare.
* inline-frame.c: Include frame.h.
(inline_frame_this_id): Use get_prev_frame_always.
gdb/testsuite/
2014-04-18 Tom Tromey <palves@redhat.com>
Pedro alves <tromey@redhat.com>
PR backtrace/15558
* gdb.opt/inline-bt.exp: Test backtracing from an inline function
with a backtrace limit.
* gdb.python/py-frame-inline.exp: Test running to an inline
function with a backtrace limit, and printing the newest frame.
* gdb.python/py-frame-inline.c (main): Call f.
Hi,
We find gdb.base/printcmds.exp fails a lot on windows host, like this,
p ctable1[163]
$204 = 163 '£'
(gdb) FAIL: gdb.base/printcmds.exp: p ctable1[163]
however, on linux host,
p ctable1[163]
$205 = 163 '\243'
(gdb) PASS: gdb.base/printcmds.exp: p ctable1[163]
The printing related code is in valprint.c:print_wchar,
if (gdb_iswprint (w) && (!need_escape || (!gdb_iswdigit (w)
&& w != LCST ('8')
&& w != LCST ('9'))))
{
gdb_wchar_t wchar = w;
if (w == gdb_btowc (quoter) || w == LCST ('\\'))
obstack_grow_wstr (output, LCST ("\\"));
obstack_grow (output, &wchar, sizeof (gdb_wchar_t));
}
else
{
// print W in hex or octal digits
}
When I debug gdb on different hosts, I find
on windows host, gdb_iswprint (iswprint) returns true if 'w' is 163.
However, on linux host, iswprint returns false if 'w' is 163. Looks
this difference is caused by the charset. On Linux host,
the target-charset is ANSI_X3.4-1968, while on windows host, the
target-charset is CP1252.
We can see how target-charset affects the output. On linux host,
(gdb) set target-charset ASCII
(gdb) p ctable1[163]
$1 = 163 '\243'
(gdb) set target-charset CP1252
(gdb) p ctable1[163]
$2 = 163 '£'
we can print the pound sign too, and it shows target-charset does
affect the output.
This patch is to set target-charset temporarily to ASCII for some
charset-sensitive tests. Tested on arm-none-eabi and
powerpc-linux-gnu on mingw32 host. More than one hundred fails are
fixed.
gdb/testsuite:
2014-04-17 Yao Qi <yao@codesourcery.com>
* lib/gdb.exp (with_target_charset): New proc.
* gdb.base/printcmds.exp (test_print_all_chars): Wrap tests with
with_target_charset.
(test_print_strings): Likewise.
(test_repeat_bytes): Likewise.
* gdb.base/setvar.exp: Set target-charset to ASCII temporarily
for some tests.
Install some sanity checks that sibling DIE offsets are not beyond the
defined limits of the DWARF input buffer in read_partial_die and skip_one_die.
2014-03-20 Keith Seitz <keiths@redhat.com>
PR gdb/15827
* dwarf2read.c (skip_one_die): Check that all relative-offset
sibling DIEs fall within range of the current reader's buffer.
(read_partial_die): Likewise.
2014-03-20 Keith Seitz <keiths@redhat.com>
PR gdb/15827
* gdb.dwarf2/corrupt.c: New file.
* gdb.dwarf2/corrupt.exp: New file.
[forgot to commit/push these with previous push]
If lookup_symbol_file tries to locate a member variable with NULL name:
/* A simple lookup failed. Check if the symbol was defined in
a base class. */
cleanup = make_cleanup (null_cleanup, NULL);
/* Find the name of the class and the name of the method,
variable, etc. */
prefix_len = cp_entire_prefix_len (name);
/* If no prefix was found, search "this". */
if (prefix_len == 0)
{
struct type *type;
struct symbol *this;
this = lookup_language_this (language_def (language_cplus), block);
if (this == NULL)
{
do_cleanups (cleanup);
return NULL;
}
type = check_typedef (TYPE_TARGET_TYPE (SYMBOL_TYPE (this)));
klass = xstrdup (TYPE_NAME (type));
nested = xstrdup (name);
}
TYPE_NAME (type) is NULL, so xstrdup (NULL) and boom!
This can happen, e.g., with clang++. See testsuite/gdb.cp/namelessclass.exp
or the bugzilla report.
This patch simply adds a fencepost against this case, allowing the caller
of lookup_symbol_file to search other blocks for the right symbol.
If lookup_symbol_file tries to locate a member variable with NULL name:
/* A simple lookup failed. Check if the symbol was defined in
a base class. */
cleanup = make_cleanup (null_cleanup, NULL);
/* Find the name of the class and the name of the method,
variable, etc. */
prefix_len = cp_entire_prefix_len (name);
/* If no prefix was found, search "this". */
if (prefix_len == 0)
{
struct type *type;
struct symbol *this;
this = lookup_language_this (language_def (language_cplus), block);
if (this == NULL)
{
do_cleanups (cleanup);
return NULL;
}
type = check_typedef (TYPE_TARGET_TYPE (SYMBOL_TYPE (this)));
klass = xstrdup (TYPE_NAME (type));
nested = xstrdup (name);
}
TYPE_NAME (type) is NULL, so xstrdup (NULL) and boom!
This can happen, e.g., with clang++. See testsuite/gdb.cp/namelessclass.exp
or the bugzilla report.
This patch simply adds a fencepost against this case, allowing the caller
of lookup_symbol_file to search other blocks for the right symbol.
* lib/gdbserver-support.exp (gdbserver_default_get_remote_address):
Add comment.
(gdbserver_default_get_comm_port): New function.
(gdbserver_start): Check if board file provided
"gdbserver,get_comm_port" and use it if so.
* boards/native-stdio-gdbserver.exp (sockethost): Set to "".
(gdb,socketport): Set to "stdio".
(gdbserver,get_comm_port): Set to ${board}_get_comm_port.
(stdio_gdbserver_template): Delete.
(${board}_get_remote_address): Update.
(${board}_build_remote_cmd): Delete.
(${board}_get_comm_port): New function.
(${board}_spawn): Update.
* boards/remote-stdio-gdbserver.exp (${board}_build_remote_cmd):
Delete.
(${board}_get_remote_address): Update.
(${board}_get_comm_port): New function.
https://sourceware.org/ml/gdb-patches/2014-04/msg00210.html
Improve the regexp used in the memattr.exp test so allow for different
memory regions (.data / .bss) being laid out in different orders.
gdb/testsuite/ChangeLog:
* gdb.base/memattr.exp: Improve regexps to handle memory regions
appearing in any order.
This test uses a simple custom elf loader, implemented in
gdb.base/sym-file-loader.h|c. This loader doesn't have a dlclose-like
function today, but I'll need one. But, I found that the guts of the
loader are exposed too much to the client, making the interface more
complicated than necessary. It's simpler if the loader just exports a
few dlopen/dlsym -style functions. That's what this patch does.
Tested on x86_86 Fedora 17, native and gdbserver.
gdb/testsuite/
2014-04-15 Pedro Alves <palves@redhat.com>
* gdb.base/sym-file-loader.h: Move inclusion of <inttypes.h>,
<ansidecl.h>, <elf/common.h> and <elf/external.h> to
sym-file-loader.c.
(Elf_External_Phdr, Elf_External_Ehdr, Elf_External_Shdr)
(Elf_External_Sym, Elf_Addr, GET, GETADDR, struct segment): Move
to sym-file-loader.c.
(struct library): Forward declare.
(load_shlib, lookup_function): Change prototypes.
(find_shstrtab, find_strtab, find_shdr, find_symtab)
(translate_offset): Remove declarations.
(get_text_addr): New declaration.
* gdb.base/sym-file-loader.c: Move inclusion of <inttypes.h>,
<ansidecl.h>, <elf/common.h> and <elf/external.h> here from
sym-file-loader.h.
(Elf_External_Phdr, Elf_External_Ehdr, Elf_External_Shdr)
(Elf_External_Sym, Elf_Addr, GET, GETADDR, struct segment): Move
here from sym-file-loader.h.
(struct library): New structure.
(load_shlib, lookup_function): Change prototypes and adjust to
work with a struct library.
(find_shstrtab, find_strtab, find_shdr, find_symtab)
(translate_offset): Make static.
(get_text_addr): New function.
* gdb.base/sym-file-main.c (main): Adjust to new loader interface.
SELF_LINK, not SELK_LINK...
gdb/testsuite/
2014-04-15 Pedro Alves <palves@redhat.com>
* gdb.base/sym-file-loader.c: Fix typo. SELF_LINK, not SELK_LINK.
The main issue here is that this test passes the host's absolute path
to the library to load to the "dlopen"-like routine, which doesn't
work when either the target or the host are remote, unless a shared
filesystem has been set up.
Tests that dynamically load a library solve this by dlopen'ing by
basename, and setting rpath to $ORIGIN. See gdb_compile.
This test doesn't use dlopen, but instead uses its own simple elf
loader. The fix is to pass this loader the library basename, and
teach it to look up the library by basename in the executable's
directory as well, i.e., assuming/emulating RPATH=$ORIGIN.
Tested on x86_64 Fedora 17, native and gdbserver.
I looked around in the web to figure out Linux's /proc/self/exe
equivalents in other ELF OSs. I think I covered all relevant, but if
not, I think it'll be simple enough to add more. (Note the test is
skipped on non-ELF targets.)
Tested on x86_64 Fedora 17, native and gdbserver.
gdb/testsuite/
2014-04-15 Pedro Alves <palves@redhat.com>
* gdb.base/sym-file-loader.c: Include <limits.h>.
(SELF_LINK): New define.
(get_origin): New function.
(load_shlib): Use it.
* gdb.base/sym-file.exp: Don't early return if the target is
remote. Use runto_main, and issue fail is that fails. Use
gdb_load_shlibs.
(shlib_name): Delete.
(lib_so, lib_syms, lib_dlopen): New globals. Use them throughout.
Remove regex characters from test message, and don't refer to
breakpoint numbers in test messages (subsequent patches will add more
breakpoints, changing these numbers). Result:
-PASS: gdb.base/sym-file.exp: add-symbol-file .*sym-file-lib\.so addr
+PASS: gdb.base/sym-file.exp: add-symbol-file sym-file-lib.so addr
-PASS: gdb.base/sym-file.exp: check if Breakpoint 2 is pending.
-PASS: gdb.base/sym-file.exp: check if Breakpoint 3 is pending.
+PASS: gdb.base/sym-file.exp: breakpoint at foo is pending
+PASS: gdb.base/sym-file.exp: breakpoint at bar is pending
gdb/testsuite/
2014-04-15 Pedro Alves <palves@redhat.com>
* gdb.base/sym-file.exp: Remove regex characters from test
message. Don't refer to breakpoint numbers in test messages.
PR c++/16253.
symbol_matches_domain was permitting searches for a VAR_DOMAIN
symbol to also match STRUCT_DOMAIN symbols for languages like C++
where STRUCT_DOMAIN symbols also define a typedef of the same name,
e.g., "struct foo {}" introduces a typedef of the name "foo".
Problems occur if there exists both a VAR_DOMAIN and STRUCT_DOMAIN
symbol of the same name. Then it is essentially a race between which
symbol is found first. The other symbol is obscurred.
[This is a relatively common idiom: enum e { ... } e;]
This patchset moves this "language defines a typedef" logic to
lookup_symbol[_in_language], looking first for a symbol in the given
domain and falling back to searching STRUCT_DOMAIN when/if appropriate.
2014-04-14 Keith Seitz <keiths@redhat.com>
PR c++/16253
* ada-lang.c (ada_symbol_matches_domain): Moved here and renamed
from symbol_matches_domain in symtab.c. All local callers
of symbol_matches_domain updated.
(standard_lookup): If DOMAIN is VAR_DOMAIN and no symbol is found,
search STRUCT_DOMAIN.
(ada_find_any_type_symbol): Do not search STRUCT_DOMAIN
independently. standard_lookup will do that automatically.
* cp-namespace.c (cp_lookup_symbol_nonlocal): Explain when/why
VAR_DOMAIN searches may return a STRUCT_DOMAIN match.
(cp_lookup_symbol_in_namespace): Likewise.
If no VAR_DOMAIN symbol is found, search STRUCT_DOMAIN.
(cp_lookup_symbol_exports): Explain when/why VAR_DOMAIN searches
may return a STRUCT_DOMAIN match.
(lookup_symbol_file): Search for the class name in STRUCT_DOMAIN.
* cp-support.c: Include language.h.
(inspect_type): Explicitly search STRUCT_DOMAIN before searching
VAR_DOMAIN.
* psymtab.c (match_partial_symbol): Compare the requested
domain with the symbol's domain directly.
(lookup_partial_symbol): Likewise.
* symtab.c (lookup_symbol_in_language): Explain when/why
VAR_DOMAIN searches may return a STRUCT_DOMAIN match.
If no VAR_DOMAIN symbol is found, search STRUCT_DOMAIN for
appropriate languages.
(symbol_matches_domain): Renamed `ada_symbol_matches_domain'
and moved to ada-lang.c
(lookup_block_symbol): Explain that this function only returns
symbol matching the requested DOMAIN.
Compare the requested domain with the symbol's domain directly.
(iterate_over_symbols): Compare the requested domain with the
symbol's domain directly.
* symtab.h (symbol_matches_domain): Remove.
2014-04-14 Keith Seitz <keiths@redhat.com>
PR c++/16253
* gdb.cp/var-tag.cc: New file.
* gdb.cp/var-tag.exp: New file.
* gdb.dwarf2/dw2-ada-ffffffff.exp: Set the language to C++.
* gdb.dwarf2/dw2-anon-mptr.exp: Likewise.
* gdb.dwarf2/dw2-double-set-die-type.exp: Likewise.
* gdb.dwarf2/dw2-inheritance.exp: Likewise.
This adds support for the C++11 "enum class" feature. This is
PR c++/15246.
I chose to use the existing TYPE_DECLARED_CLASS rather than introduce
a new type code. This seemed both simple and clear to me.
I made overloading support for the new enum types strict. This is how
it works in C++; and it didn't seem like an undue burden to keep this,
particularly because enum constants are printed symbolically by gdb.
Built and regtested on x86-64 Fedora 20.
2014-04-14 Tom Tromey <tromey@redhat.com>
PR c++/15246:
* c-exp.y (type_aggregate_p): New function.
(qualified_name, classify_inner_name): Use it.
* c-typeprint.c (c_type_print_base): Handle TYPE_DECLARED_CLASS
and TYPE_TARGET_TYPE of an enum type.
* dwarf2read.c (read_enumeration_type): Set TYPE_DECLARED_CLASS on
an enum type.
(determine_prefix) <case DW_TAG_enumeration_type>: New case;
handle TYPE_DECLARED_CLASS.
* gdbtypes.c (rank_one_type): Handle TYPE_DECLARED_CLASS on enum
types.
* gdbtypes.h (TYPE_DECLARED_CLASS): Update comment.
* valops.c (enum_constant_from_type): New function.
(value_aggregate_elt): Use it.
* cp-namespace.c (cp_lookup_nested_symbol): Handle
TYPE_CODE_ENUM.
2014-04-14 Tom Tromey <tromey@redhat.com>
* gdb.cp/classes.exp (test_enums): Handle underlying type.
* gdb.dwarf2/enum-type.exp: Add test for enum with underlying
type.
* gdb.cp/enum-class.exp: New file.
* gdb.cp/enum-class.cc: New file.
DWARF allows an enumeration type to have a DW_AT_type. GDB doesn't
recognize this, but there is a patch to change GCC to emit it, and a
DWARF proposal to further allow an enum type with a DW_AT_type to omit
the DW_AT_byte_size. This patch changes gdb to implement this.
Built and regtested on x86-64 Fedora 20.
2014-04-14 Tom Tromey <tromey@redhat.com>
* dwarf2read.c (read_enumeration_type): Handle DW_AT_type.
2014-04-14 Tom Tromey <tromey@redhat.com>
* gdb.dwarf2/enum-type.exp: New file.
The dwarf attribute DW_AT_count specifies the elements of a subrange.
This test covers subranges with present count but absent upper bound
attribute, both with static and dynamic attribute values.
testsuite/ChangeLog:
* gdb.dwarf2/count.exp: New file.
The c99 standard in "6.5.3.4 The sizeof operator" states:
If the type of the operand is a variable length array type, the operand
is evaluated;[...]
This patch mirrors the following c99 semantic in gdb:
1| int vla[n][m];
2| int i = 1;
3| sizeof(vla[i++][0]); // No sideffect
4| assert (i == 1);
5| sizeof(vla[i++]); // With sideffect
6| assert (i == 2);
Note: ptype/whatis still do not allow any sideeffects.
This patch was motivated by:
https://sourceware.org/ml/gdb-patches/2014-01/msg00732.html
gdb/ChangeLog:
* eval.c (evaluate_subexp_for_sizeof): Add enum noside argument.
(evaluate_subexp_standard): Pass noside argument.
(evaluate_subexp_for_sizeof) <BINOP_SUBSCRIPT>: Handle subscript case
if noside equals EVAL_NORMAL. If the subscript yields a vla type
re-evaluate subscript operation with EVAL_NORMAL to enable sideffects.
* gdbtypes.c (resolve_dynamic_bounds): Mark bound as evaluated.
* gdbtypes.h (enum range_flags): Add RANGE_EVALUATED case.
testsuite/ChangeLog:
* gdb.base/vla-sideeffect.c: New file.
* gdb.base/vla-sideeffect.exp: New file.
Clang defaults this warning to an error, breaking the build & causing
these tests not to run.
gdb/testsuite/
* gdb.mi/non-stop.c: Add return value for non-void function return
statement.
* gdb.threads/staticthreads.c: Ditto.
* guile/scm-value.c (gdbscm_value_dynamic_type): Use coerce_ref to
dereference TYPE_CODE_REF values.
testsuite/
* gdb.guile/scm-value.c: Improve test case.
* gdb.guile/scm-value.exp: Add new test.
