This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patch] aarch64: PR 19806: watchpoints: false negatives -> false positives
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: Pedro Alves <palves at redhat dot com>
- Cc: Yao Qi <qiyaoltc at gmail dot com>, gdb-patches at sourceware dot org
- Date: Sun, 19 Jun 2016 20:29:09 +0200
- Subject: Re: [patch] aarch64: PR 19806: watchpoints: false negatives -> false positives
- Authentication-results: sourceware.org; auth=none
- References: <20160606075945 dot GA19395 at host1 dot jankratochvil dot net> <86eg89w2sr dot fsf at gmail dot com> <48622de4-dc45-c48f-7172-495b669f2334 at redhat dot com> <86a8ixvx5k dot fsf at gmail dot com> <7fabd183-eb46-e916-77f2-f62d5c4e4ce7 at redhat dot com> <86oa7bvdi0 dot fsf at gmail dot com> <4f4d2f70-3931-6467-37c7-f97f99ad5c63 at redhat dot com> <380b5288-f46f-3e20-c9c3-8cc8738ee322 at redhat dot com>
On Wed, 08 Jun 2016 20:46:35 +0200, Pedro Alves wrote:
> I thought of a case where this is the wrong thing to do.
> (Alternative below.)
>
> The same example as before: e.g., a machine that only supports
> watching 32-bit-aligned words. Then with:
>
> union
> {
> char buf[4];
> uint32_t force_align;
> } global;
>
> (gdb) watch global.buf[1];
> Hardware watchpoint 1 ...
> (gdb) watch global.buf[3];
> Hardware watchpoint 2 ...
>
> ... if the program writes to global.buf[3], and the target
> reports a memory access to 'global.buf + 1' (because that's the
> first watchpoint in its own watchpoint list that overlaps
> the global.buf[0]..global.buf[3] range (what is really being
> watched)), gdb will believe that watchpoint 1 triggered, notice
> the value didn't change, and thus incorrectly ignore the watchpoint hit.
Here if the program really does 'global.buf[3] = 0xff;' then it still does
work as despite GDB places watchpoint at &global.buf[0] for 4 bytes aarch64
still reports the exact 1-byte location &global.buf[3]:
echo -e 'union { char buf[8]; unsigned long ul; } u; int main(void) {\n u.buf[3]=0xff;\n return 0; }'|gcc -Wall -g -x c -;./gdb -data-directory ./data-directory/ ./a.out -ex start -ex 'watch u.buf[1]' -ex 'watch u.buf[3]' -ex c -ex 'p u.buf[1]' -ex 'p u.buf[3]'
Hardware watchpoint 2: u.buf[1]
Hardware watchpoint 3: u.buf[3]
Continuing.
Hardware watchpoint 3: u.buf[3]
Old value = 0 '\000'
New value = 255 '\377'
main () at <stdin>:3
3 in <stdin>
$1 = 0 '\000'
$2 = 255 '\377'
But you are right the watchpoint gets missed if the program writes to
&global.buf[0] a 4-byte value not modifying the contents of global.buf[1] but
modifying the contents of global.buf[3]:
echo -e 'union { char buf[8]; unsigned long ul; } u; int main(void) {\n u.ul=0xff000000;\n return 0; }'|gcc -Wall -g -x c -;./gdb -data-directory ./data-directory/ ./a.out -ex start -ex 'watch u.buf[1]' -ex 'watch u.buf[3]' -ex 'b 3' -ex c -ex 'p u.buf[1]' -ex 'p u.buf[3]'
Hardware watchpoint 2: u.buf[1]
Hardware watchpoint 3: u.buf[3]
Breakpoint 4 at 0x400570: file <stdin>, line 3.
Continuing.
Breakpoint 4, main () at <stdin>:3
3 in <stdin>
$1 = 0 '\000'
$2 = 255 '\377'
I will follow with an add-on patch as I guess you think it makes sense to fix
GDB even for older kernels (assuming aarch64 kernel gets fixed in some time).
Thanks,
Jan