[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH] gdbstub: Add a missing case of signal number tr
From: |
Peter Maydell |
Subject: |
Re: [Qemu-devel] [PATCH] gdbstub: Add a missing case of signal number translation in gdbstub |
Date: |
Tue, 4 Nov 2014 19:09:43 +0000 |
On 4 November 2014 17:51, Martin Simmons <address@hidden> wrote:
> While using qemu with gdb "target remote" to debug an application that uses
> fork and exec, the qemu process receives SIGSTOP every time the forked process
> terminates (sending SIGCHLD).
>
> This is caused by a missing call to gdb_signal_to_target in gdbstub.c, which
> is fixed by this patch:
>
> Signed-off-by: Martin Simmons <address@hidden>
>
> diff --git a/gdbstub.c b/gdbstub.c
> index d1b5afd..6a73a35 100644
> --- a/gdbstub.c
> +++ b/gdbstub.c
> @@ -823,7 +823,9 @@ static int gdb_handle_packet(GDBState *s, const char
> *line_buf)
> action = *p++;
> signal = 0;
> if (action == 'C' || action == 'S') {
> - signal = strtoul(p, (char **)&p, 16);
> + signal = gdb_signal_to_target (strtoul(p, (char **)&p,
> 16));
> + if (signal == -1)
> + signal = 0;
> } else if (action != 'c' && action != 's') {
> res = 0;
> break;
The if() statement should have braces for our coding style,
and no space before the '(' in function calls; otherwise this
looks good to me.
I notice that gdb_signal_to_target() doesn't check for being
passed negative numbers, which means a malicious gdb could
make us crash here, but I assume nobody actually treats the
gdbstub as a security boundary... Anyway, that's a separate
issue for a different patch.
thanks
-- PMM