From: "Alex Bennée" <alex.bennee@linaro.org>
To: Matheus Tavares Bernardino <quic_mathbern@quicinc.com>
Cc: qemu-devel@nongnu.org, bcain@quicinc.com, f4bug@amsat.org,
peter.maydell@linaro.org, tsimpson@quicinc.com,
"Philippe Mathieu-Daudé" <philmd@linaro.org>
Subject: Re: [PATCH v2 RESEND 1/7] gdbstub: only send stop-reply packets when allowed to
Date: Thu, 20 Apr 2023 14:41:08 +0100 [thread overview]
Message-ID: <87wn26d5zs.fsf@linaro.org> (raw)
In-Reply-To: <ec04a0d3cb1c1072703f776624e503ad6257dccd.1681993775.git.quic_mathbern@quicinc.com>
Matheus Tavares Bernardino <quic_mathbern@quicinc.com> writes:
> GDB's remote serial protocol allows stop-reply messages to be sent by
> the stub either as a notification packet or as a reply to a GDB command
> (provided that the cmd accepts such a response). QEMU currently does not
> implement notification packets, so it should only send stop-replies
> synchronously and when requested. Nevertheless, it still issues
> unsolicited stop messages through gdb_vm_state_change().
>
> Although this behavior doesn't seem to cause problems with GDB itself
> (the messages are just ignored), it can impact other debuggers that
> implement the GDB remote serial protocol, like hexagon-lldb. Let's
> change the gdbstub to send stop messages only as a response to a
> previous GDB command that accepts such a reply.
>
> Signed-off-by: Matheus Tavares Bernardino <quic_mathbern@quicinc.com>
> ---
> gdbstub/internals.h | 5 +++++
> gdbstub/gdbstub.c | 37 ++++++++++++++++++++++++++++---------
> gdbstub/softmmu.c | 13 +++++++++++--
> gdbstub/user.c | 17 +++++++++++------
> 4 files changed, 55 insertions(+), 17 deletions(-)
>
> diff --git a/gdbstub/internals.h b/gdbstub/internals.h
> index 94ddff4495..33d21d6488 100644
> --- a/gdbstub/internals.h
> +++ b/gdbstub/internals.h
> @@ -65,6 +65,11 @@ typedef struct GDBState {
> GByteArray *mem_buf;
> int sstep_flags;
> int supported_sstep_flags;
> + /*
> + * Whether we are allowed to send a stop reply packet at this moment.
> + * Must be set off after sending the stop reply itself.
> + */
> + bool allow_stop_reply;
> } GDBState;
>
> /* lives in main gdbstub.c */
> diff --git a/gdbstub/gdbstub.c b/gdbstub/gdbstub.c
> index 0760d78685..be18568d0a 100644
<snip>
> /*
> @@ -139,6 +140,10 @@ static void gdb_vm_state_change(void *opaque, bool running, RunState state)
> return;
> }
>
> + if (!gdbserver_state.allow_stop_reply) {
> + return;
> + }
> +
> gdb_append_thread_id(cpu, tid);
>
> switch (state) {
> @@ -205,6 +210,7 @@ static void gdb_vm_state_change(void *opaque, bool running, RunState state)
>
> send_packet:
> gdb_put_packet(buf->str);
> + gdbserver_state.allow_stop_reply = false;
>
> /* disable single step if it was enabled */
> cpu_single_step(cpu, 0);
> @@ -422,8 +428,11 @@ void gdb_exit(int code)
>
> trace_gdbstub_op_exiting((uint8_t)code);
>
> - snprintf(buf, sizeof(buf), "W%02x", (uint8_t)code);
> - gdb_put_packet(buf);
> + if (gdbserver_state.allow_stop_reply) {
> + snprintf(buf, sizeof(buf), "W%02x", (uint8_t)code);
> + gdb_put_packet(buf);
> + gdbserver_state.allow_stop_reply = false;
> + }
>
> qemu_chr_fe_deinit(&gdbserver_system_state.chr, true);
> }
> diff --git a/gdbstub/user.c b/gdbstub/user.c
> index 80488b6bb9..bb03622c83 100644
> --- a/gdbstub/user.c
> +++ b/gdbstub/user.c
> @@ -127,11 +127,14 @@ int gdb_handlesig(CPUState *cpu, int sig)
>
> if (sig != 0) {
> gdb_set_stop_cpu(cpu);
> - g_string_printf(gdbserver_state.str_buf,
> - "T%02xthread:", gdb_target_signal_to_gdb(sig));
> - gdb_append_thread_id(cpu, gdbserver_state.str_buf);
> - g_string_append_c(gdbserver_state.str_buf, ';');
> - gdb_put_strbuf();
> + if (gdbserver_state.allow_stop_reply) {
> + g_string_printf(gdbserver_state.str_buf,
> + "T%02xthread:", gdb_target_signal_to_gdb(sig));
> + gdb_append_thread_id(cpu, gdbserver_state.str_buf);
> + g_string_append_c(gdbserver_state.str_buf, ';');
> + gdb_put_strbuf();
> + gdbserver_state.allow_stop_reply = false;
> + }
> }
> /*
> * gdb_put_packet() might have detected that the peer terminated the
> @@ -174,12 +177,14 @@ void gdb_signalled(CPUArchState *env, int sig)
> {
> char buf[4];
>
> - if (!gdbserver_state.init || gdbserver_user_state.fd < 0) {
> + if (!gdbserver_state.init || gdbserver_user_state.fd < 0 ||
> + !gdbserver_state.allow_stop_reply) {
> return;
> }
>
> snprintf(buf, sizeof(buf), "X%02x", gdb_target_signal_to_gdb(sig));
> gdb_put_packet(buf);
> + gdbserver_state.allow_stop_reply = false;
Did I miss an equivalent for softmmu mode here?
Anyway:
Acked-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
Virtualisation Tech Lead @ Linaro
next prev parent reply other threads:[~2023-04-20 13:54 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-20 12:31 [PATCH v2 RESEND 0/7] Hexagon: add lldb support Matheus Tavares Bernardino
2023-04-20 12:31 ` [PATCH v2 RESEND 1/7] gdbstub: only send stop-reply packets when allowed to Matheus Tavares Bernardino
2023-04-20 13:41 ` Alex Bennée [this message]
2023-04-21 11:31 ` Matheus Tavares Bernardino
2023-04-20 12:31 ` [PATCH v2 RESEND 2/7] gdbstub: add test for untimely stop-reply packets Matheus Tavares Bernardino
2023-04-20 13:44 ` Alex Bennée
2023-04-20 12:31 ` [PATCH v2 RESEND 3/7] gdbstub: add support for the qRegisterInfo query Matheus Tavares Bernardino
2023-04-20 13:49 ` Alex Bennée
2023-04-21 11:34 ` Matheus Tavares Bernardino
2023-04-21 13:17 ` Alex Bennée
2023-04-20 14:04 ` Philippe Mathieu-Daudé
2023-04-20 14:05 ` Philippe Mathieu-Daudé
2023-04-20 12:31 ` [PATCH v2 RESEND 4/7] Hexagon: support qRegisterInfo at gdbstub Matheus Tavares Bernardino
2023-04-20 14:06 ` Philippe Mathieu-Daudé
2023-04-21 11:40 ` Matheus Tavares Bernardino
2023-04-20 12:31 ` [PATCH v2 RESEND 5/7] Hexagon (gdbstub): fix p3:0 read and write via stub Matheus Tavares Bernardino
2023-04-20 12:31 ` [PATCH v2 RESEND 6/7] Hexagon (gdbstub): add HVX support Matheus Tavares Bernardino
2023-04-20 12:31 ` [PATCH v2 RESEND 7/7] Hexagon (linux-user/hexagon): handle breakpoints Matheus Tavares Bernardino
2023-04-27 13:31 ` Richard Henderson
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=87wn26d5zs.fsf@linaro.org \
--to=alex.bennee@linaro.org \
--cc=bcain@quicinc.com \
--cc=f4bug@amsat.org \
--cc=peter.maydell@linaro.org \
--cc=philmd@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=quic_mathbern@quicinc.com \
--cc=tsimpson@quicinc.com \
/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;
as well as URLs for NNTP newsgroup(s).