* [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD
@ 2013-04-17 14:33 Sergio Durigan Junior
2013-04-22 13:32 ` Sergio Durigan Junior
2013-04-22 14:29 ` Pedro Alves
0 siblings, 2 replies; 6+ messages in thread
From: Sergio Durigan Junior @ 2013-04-17 14:33 UTC (permalink / raw)
To: GDB Patches; +Cc: Andrew Haley, Tom Tromey
Hi,
Andrew Haley found a bug on GDB running on ARM when using
--enable-64-bit-bfd. Basically the issue happens when dealing with "bl"
instructions: GDB does branch destination calculation and (wrongly)
sign-extends the PC. Here is a piece of his original message explaining
the problem:
> next_pc = arm_get_next_pc (frame, get_frame_pc (frame));
>
> /* The Linux kernel offers some user-mode helpers in a high page. We can
> not read this page (as of 2.6.23), and even if we could then we couldn't
> set breakpoints in it, and even if we could then the atomic operations
> would fail when interrupted. They are all called as functions and return
> to the address in LR, so step to there instead. */
> if (next_pc > 0xffff0000)
> next_pc = get_frame_register_unsigned (frame, ARM_LR_REGNUM);
>
> arm_insert_single_step_breakpoint (gdbarch, aspace, next_pc);
>
> Unfortunately, branch destination addresses are SIGN EXTENDED to 64
> bits. So,
>
> (top-gdb) p/x next_pc
> $14 = 0xffffffffb6df2864
>
> Which triggers the next_pc = get_frame_register_unsigned(), and we
> cannot step into any branches because the destination PC is wrong.
Anyway, the fix is simple and Andrew himself provided it for us. It
took a while for me to figure out how to trigger the bug (in order to
write a testcase for it), but I finally made it.
The attached patch fixes the problem (by casting to `unsigned long'
instead of just `long'), and also includes a testcase to reproduce the
issue.
OK to apply? BTW, I guess this is trivial enough to be included in 7.6
as well, do you agree?
Thanks,
--
Sergio
gdb/
2013-04-11 Andrew Haley <aph@redhat.com>
* arm-tdep.c (BranchDest): Cast result as "unsigned long",
instead of "long".
gdb/testsuite
2013-04-11 Sergio Durigan Junior <sergiodj@redhat.com>
* gdb.arch/arm-bl-branch-dest.c: New file.
* gdb.arch/arm-bl-branch-dest.exp: Likewise.
diff --git a/gdb/arm-tdep.c b/gdb/arm-tdep.c
index 2a11890..16cb488 100644
--- a/gdb/arm-tdep.c
+++ b/gdb/arm-tdep.c
@@ -521,7 +521,7 @@ skip_prologue_function (struct gdbarch *gdbarch, CORE_ADDR pc, int is_thumb)
#define sbits(obj,st,fn) \
((long) (bits(obj,st,fn) | ((long) bit(obj,fn) * ~ submask (fn - st))))
#define BranchDest(addr,instr) \
- ((CORE_ADDR) (((long) (addr)) + 8 + (sbits (instr, 0, 23) << 2)))
+ ((CORE_ADDR) (((unsigned long) (addr)) + 8 + (sbits (instr, 0, 23) << 2)))
/* Extract the immediate from instruction movw/movt of encoding T. INSN1 is
the first 16-bit of instruction, and INSN2 is the second 16-bit of
diff --git a/gdb/testsuite/gdb.arch/arm-bl-branch-dest.c b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.c
new file mode 100644
index 0000000..98269d0
--- /dev/null
+++ b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.c
@@ -0,0 +1,29 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright 2013 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/>. */
+
+static void
+foo (int a)
+{
+ ++a;
+}
+
+int
+main (int argc, char *argv[])
+{
+ foo (10);
+ return 0;
+}
diff --git a/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp
new file mode 100644
index 0000000..1e433f3
--- /dev/null
+++ b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp
@@ -0,0 +1,37 @@
+# Copyright (C) 2013 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 { ![istarget "arm*-*-*"] } {
+ verbose "Skipping ${testfile}."
+ return
+}
+
+standard_testfile
+
+# We need to load the text segment in a high address. This is because
+# the bug we are dealing with happened when GDB sign-extended the PC
+# on ARM, causing the PC to acquire a wrong value. That's why we use
+# the "-Wl,-Ttext-segment" option compile the binary.
+
+if { [prepare_for_testing ${testfile}.exp ${testfile} ${srcfile} \
+ [list debug ldflags=-Wl,-Ttext-segment=0xb0000000]] } {
+ return -1
+}
+
+if { ![runto_main] } {
+ return -1
+}
+
+gdb_test "next" "\[0-9\]+\\s+return 0;"
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD
2013-04-17 14:33 [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD Sergio Durigan Junior
@ 2013-04-22 13:32 ` Sergio Durigan Junior
2013-04-22 14:29 ` Pedro Alves
1 sibling, 0 replies; 6+ messages in thread
From: Sergio Durigan Junior @ 2013-04-22 13:32 UTC (permalink / raw)
To: GDB Patches; +Cc: Andrew Haley, Tom Tromey
On Wednesday, April 17 2013, I wrote:
Ping.
> Andrew Haley found a bug on GDB running on ARM when using
> --enable-64-bit-bfd. Basically the issue happens when dealing with "bl"
> instructions: GDB does branch destination calculation and (wrongly)
> sign-extends the PC. Here is a piece of his original message explaining
> the problem:
>
>> next_pc = arm_get_next_pc (frame, get_frame_pc (frame));
>>
>> /* The Linux kernel offers some user-mode helpers in a high page. We can
>> not read this page (as of 2.6.23), and even if we could then we couldn't
>> set breakpoints in it, and even if we could then the atomic operations
>> would fail when interrupted. They are all called as functions and return
>> to the address in LR, so step to there instead. */
>> if (next_pc > 0xffff0000)
>> next_pc = get_frame_register_unsigned (frame, ARM_LR_REGNUM);
>>
>> arm_insert_single_step_breakpoint (gdbarch, aspace, next_pc);
>>
>> Unfortunately, branch destination addresses are SIGN EXTENDED to 64
>> bits. So,
>>
>> (top-gdb) p/x next_pc
>> $14 = 0xffffffffb6df2864
>>
>> Which triggers the next_pc = get_frame_register_unsigned(), and we
>> cannot step into any branches because the destination PC is wrong.
>
> Anyway, the fix is simple and Andrew himself provided it for us. It
> took a while for me to figure out how to trigger the bug (in order to
> write a testcase for it), but I finally made it.
>
> The attached patch fixes the problem (by casting to `unsigned long'
> instead of just `long'), and also includes a testcase to reproduce the
> issue.
>
> OK to apply? BTW, I guess this is trivial enough to be included in 7.6
> as well, do you agree?
>
> Thanks,
>
> --
> Sergio
>
> gdb/
> 2013-04-11 Andrew Haley <aph@redhat.com>
>
> * arm-tdep.c (BranchDest): Cast result as "unsigned long",
> instead of "long".
>
> gdb/testsuite
> 2013-04-11 Sergio Durigan Junior <sergiodj@redhat.com>
>
> * gdb.arch/arm-bl-branch-dest.c: New file.
> * gdb.arch/arm-bl-branch-dest.exp: Likewise.
>
> diff --git a/gdb/arm-tdep.c b/gdb/arm-tdep.c
> index 2a11890..16cb488 100644
> --- a/gdb/arm-tdep.c
> +++ b/gdb/arm-tdep.c
> @@ -521,7 +521,7 @@ skip_prologue_function (struct gdbarch *gdbarch, CORE_ADDR pc, int is_thumb)
> #define sbits(obj,st,fn) \
> ((long) (bits(obj,st,fn) | ((long) bit(obj,fn) * ~ submask (fn - st))))
> #define BranchDest(addr,instr) \
> - ((CORE_ADDR) (((long) (addr)) + 8 + (sbits (instr, 0, 23) << 2)))
> + ((CORE_ADDR) (((unsigned long) (addr)) + 8 + (sbits (instr, 0, 23) << 2)))
>
> /* Extract the immediate from instruction movw/movt of encoding T. INSN1 is
> the first 16-bit of instruction, and INSN2 is the second 16-bit of
> diff --git a/gdb/testsuite/gdb.arch/arm-bl-branch-dest.c b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.c
> new file mode 100644
> index 0000000..98269d0
> --- /dev/null
> +++ b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.c
> @@ -0,0 +1,29 @@
> +/* This testcase is part of GDB, the GNU debugger.
> +
> + Copyright 2013 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/>. */
> +
> +static void
> +foo (int a)
> +{
> + ++a;
> +}
> +
> +int
> +main (int argc, char *argv[])
> +{
> + foo (10);
> + return 0;
> +}
> diff --git a/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp
> new file mode 100644
> index 0000000..1e433f3
> --- /dev/null
> +++ b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp
> @@ -0,0 +1,37 @@
> +# Copyright (C) 2013 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 { ![istarget "arm*-*-*"] } {
> + verbose "Skipping ${testfile}."
> + return
> +}
> +
> +standard_testfile
> +
> +# We need to load the text segment in a high address. This is because
> +# the bug we are dealing with happened when GDB sign-extended the PC
> +# on ARM, causing the PC to acquire a wrong value. That's why we use
> +# the "-Wl,-Ttext-segment" option compile the binary.
> +
> +if { [prepare_for_testing ${testfile}.exp ${testfile} ${srcfile} \
> + [list debug ldflags=-Wl,-Ttext-segment=0xb0000000]] } {
> + return -1
> +}
> +
> +if { ![runto_main] } {
> + return -1
> +}
> +
> +gdb_test "next" "\[0-9\]+\\s+return 0;"
--
Sergio
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD
2013-04-17 14:33 [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD Sergio Durigan Junior
2013-04-22 13:32 ` Sergio Durigan Junior
@ 2013-04-22 14:29 ` Pedro Alves
2013-04-22 14:46 ` Sergio Durigan Junior
1 sibling, 1 reply; 6+ messages in thread
From: Pedro Alves @ 2013-04-22 14:29 UTC (permalink / raw)
To: Sergio Durigan Junior; +Cc: GDB Patches, Andrew Haley, Tom Tromey
On 04/17/2013 05:56 AM, Sergio Durigan Junior wrote:
> gdb/
> 2013-04-11 Andrew Haley <aph@redhat.com>
>
> * arm-tdep.c (BranchDest): Cast result as "unsigned long",
> instead of "long".
>
> gdb/testsuite
> 2013-04-11 Sergio Durigan Junior <sergiodj@redhat.com>
>
> * gdb.arch/arm-bl-branch-dest.c: New file.
> * gdb.arch/arm-bl-branch-dest.exp: Likewise.
OK.
--
Pedro Alves
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD
2013-04-22 14:29 ` Pedro Alves
@ 2013-04-22 14:46 ` Sergio Durigan Junior
2013-04-22 15:32 ` [commit] Replace additional_flags by ldflags (was Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD) Sergio Durigan Junior
2013-04-25 17:26 ` [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD Sergio Durigan Junior
0 siblings, 2 replies; 6+ messages in thread
From: Sergio Durigan Junior @ 2013-04-22 14:46 UTC (permalink / raw)
To: Pedro Alves; +Cc: GDB Patches, Andrew Haley, Tom Tromey
On Monday, April 22 2013, Pedro Alves wrote:
> On 04/17/2013 05:56 AM, Sergio Durigan Junior wrote:
>> gdb/
>> 2013-04-11 Andrew Haley <aph@redhat.com>
>>
>> * arm-tdep.c (BranchDest): Cast result as "unsigned long",
>> instead of "long".
>>
>> gdb/testsuite
>> 2013-04-11 Sergio Durigan Junior <sergiodj@redhat.com>
>>
>> * gdb.arch/arm-bl-branch-dest.c: New file.
>> * gdb.arch/arm-bl-branch-dest.exp: Likewise.
>
> OK.
Thanks, checked-in:
http://sourceware.org/ml/gdb-cvs/2013-04/msg00213.html
--
Sergio
^ permalink raw reply [flat|nested] 6+ messages in thread
* [commit] Replace additional_flags by ldflags (was Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD)
2013-04-22 14:46 ` Sergio Durigan Junior
@ 2013-04-22 15:32 ` Sergio Durigan Junior
2013-04-25 17:26 ` [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD Sergio Durigan Junior
1 sibling, 0 replies; 6+ messages in thread
From: Sergio Durigan Junior @ 2013-04-22 15:32 UTC (permalink / raw)
To: Pedro Alves; +Cc: GDB Patches, Andrew Haley, Tom Tromey
On Monday, April 22 2013, I wrote:
> Thanks, checked-in:
>
> http://sourceware.org/ml/gdb-cvs/2013-04/msg00213.html
Ops, what I committed was a previous version of the patch. They're
almost the same, with the difference that instead of using
"additional_flags" the new version uses "ldflags", a change suggested by
Pedro himself.
I am sorry about the confusion, it has been fixed by the patch below.
Checked-in.
Thanks,
--
Sergio
--- a/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp
+++ b/gdb/testsuite/gdb.arch/arm-bl-branch-dest.exp
@@ -26,7 +26,7 @@ standard_testfile
# the "-Wl,-Ttext-segment" option compile the binary.
if { [prepare_for_testing ${testfile}.exp ${testfile} ${srcfile} \
- [list debug additional_flags=-Wl,-Ttext-segment=0xb0000000]] } {
+ [list debug ldflags=-Wl,-Ttext-segment=0xb0000000]] } {
return -1
}
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD
2013-04-22 14:46 ` Sergio Durigan Junior
2013-04-22 15:32 ` [commit] Replace additional_flags by ldflags (was Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD) Sergio Durigan Junior
@ 2013-04-25 17:26 ` Sergio Durigan Junior
1 sibling, 0 replies; 6+ messages in thread
From: Sergio Durigan Junior @ 2013-04-25 17:26 UTC (permalink / raw)
To: Pedro Alves; +Cc: GDB Patches, Andrew Haley, Tom Tromey
On Monday, April 22 2013, I wrote:
> Thanks, checked-in:
>
> http://sourceware.org/ml/gdb-cvs/2013-04/msg00213.html
I forgot to check the patch in the 7.6 branch as well. Did this today.
http://sourceware.org/ml/gdb-cvs/2013-04/msg00239.html
--
Sergio
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2013-04-25 12:24 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-04-17 14:33 [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD Sergio Durigan Junior
2013-04-22 13:32 ` Sergio Durigan Junior
2013-04-22 14:29 ` Pedro Alves
2013-04-22 14:46 ` Sergio Durigan Junior
2013-04-22 15:32 ` [commit] Replace additional_flags by ldflags (was Re: [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD) Sergio Durigan Junior
2013-04-25 17:26 ` [PATCH] Fix ARM sign-extension on PC when using 64-bit BFD Sergio Durigan Junior
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox