From: Jan Kratochvil <jan.kratochvil@redhat.com>
To: Philippe Waroquiers <philippe.waroquiers@skynet.be>,
Pedro Alves <palves@redhat.com>
Cc: Joel Brobecker <brobecker@adacore.com>,
gdb-patches@sourceware.org,
"Maciej W. Rozycki" <macro@codesourcery.com>
Subject: Re: [patchv2] Write bpt at the ON_STACK bpt address
Date: Fri, 27 Jul 2012 18:47:00 -0000 [thread overview]
Message-ID: <20120727184633.GA14182@host2.jankratochvil.net> (raw)
In-Reply-To: <5012B1B6.1030003@redhat.com> <1343339398.2258.119.camel@soleil>
On Thu, 26 Jul 2012 23:49:58 +0200, Philippe Waroquiers wrote:
> On Thu, 2012-07-26 at 23:23 +0200, Jan Kratochvil wrote:
> > I still have to write a testcase for it using valgrind.
>
> IIUC, this implies to write a new 'gdbserver board file' (or something
> like that) which will allow to access the Valgrind gdbsrv.
There is valgrind board file (without gdbsrv). But I have found running the
testsuite nightly with valgrind is not feasible, it is too costly resulting in
various testsuite issues.
I have added just regular testfile like there was already:
gdb.base/valgrind-db-attach.exp
> Note that this is covered by the Valgrind regression tests
> (that is how the change of behaviour with 7.4.91 was detected).
Somehow I was not notified soon enough in Fedora Rawhide, probably because
Fedora valgrind maintenance is now affected by the dwz incompatibility.
On Fri, 27 Jul 2012 17:20:22 +0200, Pedro Alves wrote:
> FWIW, the patch looks good to me, and I don't think that
> limiting to x86 is necessary:
OK, thanks for review. I will yet wait for the Maciej's MIPS test, not sure
what is the next 7.5 snapshot release.
> IMO, this comment should also mention that in addition to being nice for
> the user, this is actually _necessary_ for at least Valgrind v XXX.YYY.
Done.
Thanks,
Jan
gdb/
2012-07-27 Jan Kratochvil <jan.kratochvil@redhat.com>
* infcall.c (call_function_by_hand): Move BP_ADDR comment to
AT_ENTRY_POINT.
(call_function_by_hand) <ON_STACK>: Call write_memory with
gdbarch_breakpoint_from_pc, if possible.
(call_function_by_hand) <AT_ENTRY_POINT>: The BP_ADDR comment is moved
here.
gdb/doc/
2012-07-27 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdbint.texinfo (Defining Other Architecture Features): Clarify *pcptr
encoding for gdbarch_breakpoint_from_pc, bp_addr for
gdbarch_push_dummy_call and bp_addr for gdbarch_push_dummy_code.
gdb/testsuite/
2012-07-27 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.base/valgrind-db-attach.exp: Do not run in remote mode.
* gdb.base/valgrind-infcall.c: New file.
* gdb.base/valgrind-infcall.exp: New file.
diff --git a/gdb/doc/gdbint.texinfo b/gdb/doc/gdbint.texinfo
index 5e00f1f..b66f80b 100644
--- a/gdb/doc/gdbint.texinfo
+++ b/gdb/doc/gdbint.texinfo
@@ -4540,8 +4540,10 @@ contents and size of a breakpoint instruction. It returns a pointer to
a static string of bytes that encode a breakpoint instruction, stores the
length of the string to @code{*@var{lenptr}}, and adjusts the program
counter (if necessary) to point to the actual memory location where the
-breakpoint should be inserted. May return @code{NULL} to indicate that
-software breakpoints are not supported.
+breakpoint should be inserted. The program counter (@code{*@var{pcptr}}
+is inferior PC register encoded on the input and it is a plain address on the
+output. Function may return @code{NULL} to indicate that software breakpoints
+are not supported.
Although it is common to use a trap instruction for a breakpoint, it's
not required; for instance, the bit pattern could be an invalid
@@ -4821,7 +4823,7 @@ instead of value.
@anchor{gdbarch_push_dummy_call} Define this to push the dummy frame's call to
the inferior function onto the stack. In addition to pushing @var{nargs}, the
code should push @var{struct_addr} (when @var{struct_return} is non-zero), and
-the return address (@var{bp_addr}).
+the return address (@var{bp_addr}, in inferior PC register encoding).
@var{function} is a pointer to a @code{struct value}; on architectures that use
function descriptors, this contains the function descriptor value.
@@ -4835,12 +4837,14 @@ instruction sequence (including space for a breakpoint) to which the
called function should return.
Set @var{bp_addr} to the address at which the breakpoint instruction
-should be inserted, @var{real_pc} to the resume address when starting
-the call sequence, and return the updated inner-most stack address.
+should be inserted (in inferior PC register encoding), @var{real_pc} to the
+resume address when starting the call sequence, and return the updated
+inner-most stack address.
By default, the stack is grown sufficient to hold a frame-aligned
(@pxref{frame_align}) breakpoint, @var{bp_addr} is set to the address
-reserved for that breakpoint, and @var{real_pc} set to @var{funaddr}.
+reserved for that breakpoint (in inferior PC register encoding), and
+@var{real_pc} set to @var{funaddr}.
This method replaces @w{@code{gdbarch_call_dummy_location (@var{gdbarch})}}.
diff --git a/gdb/infcall.c b/gdb/infcall.c
index 51cd118..1b2c3d6 100644
--- a/gdb/infcall.c
+++ b/gdb/infcall.c
@@ -618,15 +618,38 @@ call_function_by_hand (struct value *function, int nargs, struct value **args)
not just the breakpoint but also an extra word containing the
size (?) of the structure being passed. */
- /* The actual breakpoint (at BP_ADDR) is inserted separatly so there
- is no need to write that out. */
-
switch (gdbarch_call_dummy_location (gdbarch))
{
case ON_STACK:
- sp = push_dummy_code (gdbarch, sp, funaddr,
- args, nargs, target_values_type,
- &real_pc, &bp_addr, get_current_regcache ());
+ {
+ const gdb_byte *bp_bytes;
+ CORE_ADDR bp_addr_as_address;
+ int bp_size;
+
+ /* Be careful BP_ADDR is in inferior PC encoding while
+ BP_ADDR_AS_ADDRESS is a plain memory address. */
+
+ sp = push_dummy_code (gdbarch, sp, funaddr, args, nargs,
+ target_values_type, &real_pc, &bp_addr,
+ get_current_regcache ());
+
+ /* Write a legitimate instruction at the point where the infcall
+ breakpoint is going to be inserted. While this instruction
+ is never going to be executed, a user investigating the
+ memory from GDB would see this instruction instead of random
+ uninitialized bytes. We chose the breakpoint instruction
+ as it may look as the most logical one to the user and also
+ valgrind 3.7.0 needs it for proper vgdb inferior calls.
+
+ If software breakpoints are unsupported for this target we
+ leave the user visible memory content uninitialized. */
+
+ bp_addr_as_address = bp_addr;
+ bp_bytes = gdbarch_breakpoint_from_pc (gdbarch, &bp_addr_as_address,
+ &bp_size);
+ if (bp_bytes != NULL)
+ write_memory (bp_addr_as_address, bp_bytes, bp_size);
+ }
break;
case AT_ENTRY_POINT:
{
@@ -634,8 +657,12 @@ call_function_by_hand (struct value *function, int nargs, struct value **args)
real_pc = funaddr;
dummy_addr = entry_point_address ();
+
/* A call dummy always consists of just a single breakpoint, so
- its address is the same as the address of the dummy. */
+ its address is the same as the address of the dummy.
+
+ The actual breakpoint is inserted separatly so there is no need to
+ write that out. */
bp_addr = dummy_addr;
break;
}
diff --git a/gdb/testsuite/gdb.base/valgrind-db-attach.exp b/gdb/testsuite/gdb.base/valgrind-db-attach.exp
index b14401f..313e4e0 100644
--- a/gdb/testsuite/gdb.base/valgrind-db-attach.exp
+++ b/gdb/testsuite/gdb.base/valgrind-db-attach.exp
@@ -13,6 +13,11 @@
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
+if [is_remote target] {
+ # The test always runs locally.
+ return 0
+}
+
set test valgrind-db-attach
set srcfile $test.c
set executable $test
diff --git a/gdb/testsuite/gdb.base/valgrind-infcall.c b/gdb/testsuite/gdb.base/valgrind-infcall.c
new file mode 100644
index 0000000..c119b7e
--- /dev/null
+++ b/gdb/testsuite/gdb.base/valgrind-infcall.c
@@ -0,0 +1,40 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright 2012 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <stdlib.h>
+
+static volatile int infcall_var;
+
+static int
+gdb_test_infcall (void)
+{
+ return ++infcall_var;
+}
+
+int
+main (void)
+{
+ void *p;
+
+ gdb_test_infcall ();
+ p = malloc (1);
+ if (p == NULL)
+ return 1;
+ free (p);
+ free (p); /* double-free */
+ return 0;
+}
diff --git a/gdb/testsuite/gdb.base/valgrind-infcall.exp b/gdb/testsuite/gdb.base/valgrind-infcall.exp
new file mode 100644
index 0000000..ede26f4
--- /dev/null
+++ b/gdb/testsuite/gdb.base/valgrind-infcall.exp
@@ -0,0 +1,115 @@
+# Copyright 2012 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+if [is_remote target] {
+ # The test always runs locally.
+ return 0
+}
+
+set test valgrind-infcall
+set srcfile $test.c
+set executable $test
+set binfile ${objdir}/${subdir}/${executable}
+if {[build_executable $test.exp $executable $srcfile {debug}] == -1} {
+ return -1
+}
+
+set test "spawn valgrind"
+set cmd "valgrind --vgdb-error=0 $binfile"
+set res [remote_spawn host $cmd];
+if { $res < 0 || $res == "" } {
+ verbose -log "Spawning $cmd failed."
+ unsupported $test
+ return -1
+}
+pass $test
+# Declare GDB now as running.
+set gdb_spawn_id -1
+
+# GDB started by vgdb stops already after the startup is executed, like with
+# non-extended gdbserver. It is also not correct to run/attach the inferior.
+set use_gdb_stub 1
+
+set test "valgrind started"
+# The trailing '.' differs for different memcheck versions.
+gdb_test_multiple "" $test {
+ -re "Memcheck, a memory error detector\\.?\r\n" {
+ pass $test
+ }
+ -re "valgrind: failed to start tool 'memcheck' for platform '.*': No such file or directory" {
+ unsupported $test
+ return -1
+ }
+ -re "valgrind: wrong ELF executable class" {
+ unsupported $test
+ return -1
+ }
+ -re "command not found" {
+ # The spawn succeeded, but then valgrind was not found - e.g. if
+ # we spawned SSH to a remote system.
+ unsupported $test
+ return -1
+ }
+ -re "valgrind: Bad option '--vgdb-error=0'" {
+ # valgrind is not >= 3.7.0.
+ unsupported $test
+ return -1
+ }
+}
+
+set test "vgdb prompt"
+# The trailing '.' differs for different memcheck versions.
+gdb_test_multiple "" $test {
+ -re " (target remote | \[^\r\n\]*/vgdb \[^\r\n\]*)\r\n" {
+ set vgdbcmd $expect_out(1,string)
+ pass $test
+ }
+}
+
+# Do not kill valgrind.
+unset gdb_spawn_id
+set board [host_info name]
+unset_board_info fileid
+
+clean_restart $executable
+
+gdb_test "$vgdbcmd" " in _start .*" "target remote for vgdb"
+
+gdb_test "monitor v.set gdb_output" "valgrind output will go to gdb.*"
+
+set continue_count 1
+while 1 {
+ set test "continue #$continue_count"
+ gdb_test_multiple "continue" "" {
+ -re "Invalid free\\(\\).*: main .*\r\n$gdb_prompt $" {
+ pass $test
+ break
+ }
+ -re "\r\n$gdb_prompt $" {
+ pass "$test (false warning)"
+ }
+ }
+ set continue_count [expr $continue_count + 1]
+}
+
+set test "p gdb_test_infcall ()"
+gdb_test_multiple $test $test {
+ -re "unhandled instruction bytes.*\r\n$gdb_prompt $" {
+ fail $test
+ }
+ -re "Continuing \\.\\.\\..*\r\n\\\$1 = 2\r\n$gdb_prompt $" {
+ pass $test
+ }
+}
next prev parent reply other threads:[~2012-07-27 18:47 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20120718163413.GA17548@adacore.com>
[not found] ` <1342739016.2220.32.camel@soleil>
[not found] ` <20120720071158.GA7053@host2.jankratochvil.net>
[not found] ` <1342817409.2149.41.camel@soleil>
[not found] ` <20120722173053.GA22036@host2.jankratochvil.net>
[not found] ` <1342983655.2301.55.camel@soleil>
2012-07-23 7:22 ` [patch] [i386] Put hlt at the ON_STACK breakpoint [Re: GDB 7.4.91 available for testing] Jan Kratochvil
2012-07-23 16:00 ` Joel Brobecker
2012-07-23 16:36 ` Jan Kratochvil
2012-07-23 20:07 ` Philippe Waroquiers
2012-07-23 20:16 ` Jan Kratochvil
2012-07-23 20:37 ` Philippe Waroquiers
2012-07-25 14:49 ` Joel Brobecker
2012-07-25 20:04 ` Philippe Waroquiers
2012-07-25 20:11 ` Jan Kratochvil
2012-07-25 20:39 ` Philippe Waroquiers
2012-07-25 14:59 ` Pedro Alves
2012-07-25 20:24 ` Philippe Waroquiers
2012-07-25 21:27 ` Joel Brobecker
2012-07-25 21:46 ` Philippe Waroquiers
2012-07-25 22:39 ` Joel Brobecker
2012-07-26 21:24 ` [patchv2] Write bpt at the ON_STACK bpt address Jan Kratochvil
2012-07-26 21:50 ` Philippe Waroquiers
2012-07-27 18:47 ` Jan Kratochvil [this message]
2012-07-28 7:28 ` Eli Zaretskii
2012-07-28 7:42 ` Jan Kratochvil
2012-07-31 7:37 ` [commit+7.5] " Jan Kratochvil
2012-08-01 9:08 ` [commit#2+7.5] testsuite: valgrind-infcall.exp UNSUPPORTED update [Re: [commit+7.5] [patchv2] Write bpt at the ON_STACK bpt address] Jan Kratochvil
2012-08-02 22:49 ` [commit+7.5] [patchv2] Write bpt at the ON_STACK bpt address Edjunior Barbosa Machado
2012-08-02 23:09 ` Sergio Durigan Junior
2012-08-03 0:15 ` Edjunior Barbosa Machado
2012-08-03 11:23 ` Jan Kratochvil
2012-08-03 12:09 ` Edjunior Barbosa Machado
2012-08-03 1:00 ` Pedro Alves
2012-08-03 1:48 ` Sergio Durigan Junior
2012-08-03 2:30 ` Edjunior Barbosa Machado
2012-08-03 21:45 ` Philippe Waroquiers
2012-08-03 14:23 ` Tom Tromey
2012-08-03 14:31 ` Jan Kratochvil
2012-08-03 15:02 ` Edjunior Barbosa Machado
2012-08-03 15:08 ` Jan Kratochvil
2012-08-03 16:43 ` Edjunior Barbosa Machado
2012-08-03 16:46 ` Jan Kratochvil
2012-08-03 18:00 ` Edjunior Barbosa Machado
2012-08-03 20:23 ` Jan Kratochvil
2012-08-03 21:46 ` Edjunior Barbosa Machado
2012-08-06 18:40 ` Tom Tromey
2012-07-31 7:40 ` [commit] valgrind-db-attach.exp: Do not run in remote mode [Re: [patchv2] Write bpt at the ON_STACK bpt address] Jan Kratochvil
2012-07-26 23:14 ` [patchv2] Write bpt at the ON_STACK bpt address Maciej W. Rozycki
2012-07-27 16:02 ` Jan Kratochvil
2012-07-28 0:10 ` Maciej W. Rozycki
2012-07-28 6:06 ` Jan Kratochvil
2012-07-30 18:09 ` Maciej W. Rozycki
2012-07-26 23:15 ` Philippe Waroquiers
2012-07-27 16:03 ` Jan Kratochvil
2012-07-27 15:21 ` Pedro Alves
2012-07-26 21:56 ` [patch] [i386] Put hlt at the ON_STACK breakpoint [Re: GDB 7.4.91 available for testing] Philippe Waroquiers
2012-07-26 22:41 ` Philippe Waroquiers
2012-07-26 5:13 ` Jan Kratochvil
2012-07-26 12:48 ` Pedro Alves
2012-07-26 22:21 ` Philippe Waroquiers
2012-07-27 14:59 ` Pedro Alves
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20120727184633.GA14182@host2.jankratochvil.net \
--to=jan.kratochvil@redhat.com \
--cc=brobecker@adacore.com \
--cc=gdb-patches@sourceware.org \
--cc=macro@codesourcery.com \
--cc=palves@redhat.com \
--cc=philippe.waroquiers@skynet.be \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox