From: Peter Maydell <peter.maydell@linaro.org>
To: qemu-devel@nongnu.org
Cc: Richard Henderson <richard.henderson@linaro.org>,
Warner Losh <imp@bsdimp.com>, Kyle Evans <kevans@freebsd.org>,
libvir-list@redhat.com, Markus Armbruster <armbru@redhat.com>,
Laurent Vivier <laurent@vivier.eu>,
Eric Blake <eblake@redhat.com>
Subject: [PATCH v3 03/10] accel/tcg: Use one_insn_per_tb global instead of old singlestep global
Date: Mon, 17 Apr 2023 17:40:34 +0100 [thread overview]
Message-ID: <20230417164041.684562-4-peter.maydell@linaro.org> (raw)
In-Reply-To: <20230417164041.684562-1-peter.maydell@linaro.org>
The only place left that looks at the old 'singlestep' global
variable is the TCG curr_cflags() function. Replace the old global
with a new 'one_insn_per_tb' which is defined in tcg-all.c and
declared in accel/tcg/internal.h. This keeps it restricted to the
TCG code, unlike 'singlestep' which was available to every file in
the system and defined in multiple different places for softmmu vs
linux-user vs bsd-user.
While we're making this change, use qatomic_read() and qatomic_set()
on the accesses to the new global, because TCG will read it without
holding a lock.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
In discussion on v2, we talked about combining this with the
'nochain' flag so as to have a single 'tcg_cflags_global' that
held the flags for the current (one_insn_per_tb, nochain) state.
I have not attempted that here, because it's a little tricky:
* util/log.c is built into some binaries that don't have an
accelerator at all (the tools), so it can't simply call
current_accel() to get the TCG accelerator
* the initial value of the logging flags is set before the
TCG accelerator is even created
So I leave that to somebody else to have a go at if they like.
---
accel/tcg/internal.h | 2 ++
include/exec/cpu-common.h | 2 --
accel/tcg/cpu-exec.c | 2 +-
accel/tcg/tcg-all.c | 6 ++++--
bsd-user/main.c | 1 -
linux-user/main.c | 1 -
softmmu/globals.c | 1 -
7 files changed, 7 insertions(+), 8 deletions(-)
diff --git a/accel/tcg/internal.h b/accel/tcg/internal.h
index 96f198b28b4..7bb0fdbe149 100644
--- a/accel/tcg/internal.h
+++ b/accel/tcg/internal.h
@@ -67,4 +67,6 @@ static inline target_ulong log_pc(CPUState *cpu, const TranslationBlock *tb)
extern int64_t max_delay;
extern int64_t max_advance;
+extern bool one_insn_per_tb;
+
#endif /* ACCEL_TCG_INTERNAL_H */
diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
index 6feaa40ca7b..0be74f1e706 100644
--- a/include/exec/cpu-common.h
+++ b/include/exec/cpu-common.h
@@ -163,8 +163,6 @@ int cpu_memory_rw_debug(CPUState *cpu, vaddr addr,
void *ptr, size_t len, bool is_write);
/* vl.c */
-extern int singlestep;
-
void list_cpus(const char *optarg);
#endif /* CPU_COMMON_H */
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 8370c92c05e..bc0e1c3299a 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -159,7 +159,7 @@ uint32_t curr_cflags(CPUState *cpu)
*/
if (unlikely(cpu->singlestep_enabled)) {
cflags |= CF_NO_GOTO_TB | CF_NO_GOTO_PTR | CF_SINGLE_STEP | 1;
- } else if (singlestep) {
+ } else if (qatomic_read(&one_insn_per_tb)) {
cflags |= CF_NO_GOTO_TB | 1;
} else if (qemu_loglevel_mask(CPU_LOG_TB_NOCHAIN)) {
cflags |= CF_NO_GOTO_TB;
diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
index fcf361c8db6..a831f8d7c37 100644
--- a/accel/tcg/tcg-all.c
+++ b/accel/tcg/tcg-all.c
@@ -31,6 +31,7 @@
#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/accel.h"
+#include "qemu/atomic.h"
#include "qapi/qapi-builtin-visit.h"
#include "qemu/units.h"
#if !defined(CONFIG_USER_ONLY)
@@ -110,6 +111,7 @@ static void tcg_accel_instance_init(Object *obj)
}
bool mttcg_enabled;
+bool one_insn_per_tb;
static int tcg_init_machine(MachineState *ms)
{
@@ -219,8 +221,8 @@ static void tcg_set_one_insn_per_tb(Object *obj, bool value, Error **errp)
{
TCGState *s = TCG_STATE(obj);
s->one_insn_per_tb = value;
- /* For the moment, set the global also: this changes the behaviour */
- singlestep = value;
+ /* Set the global also: this changes the behaviour */
+ qatomic_set(&one_insn_per_tb, value);
}
static int tcg_gdbstub_supported_sstep_flags(void)
diff --git a/bsd-user/main.c b/bsd-user/main.c
index 09b84da190c..a9e5a127d38 100644
--- a/bsd-user/main.c
+++ b/bsd-user/main.c
@@ -49,7 +49,6 @@
#include "host-os.h"
#include "target_arch_cpu.h"
-int singlestep;
static bool opt_one_insn_per_tb;
uintptr_t guest_base;
bool have_guest_base;
diff --git a/linux-user/main.c b/linux-user/main.c
index 489694ad654..c7020b413bc 100644
--- a/linux-user/main.c
+++ b/linux-user/main.c
@@ -68,7 +68,6 @@
char *exec_path;
char real_exec_path[PATH_MAX];
-int singlestep;
static bool opt_one_insn_per_tb;
static const char *argv0;
static const char *gdbstub;
diff --git a/softmmu/globals.c b/softmmu/globals.c
index 39678aa8c58..e83b5428d12 100644
--- a/softmmu/globals.c
+++ b/softmmu/globals.c
@@ -43,7 +43,6 @@ int vga_interface_type = VGA_NONE;
bool vga_interface_created;
Chardev *parallel_hds[MAX_PARALLEL_PORTS];
int win2k_install_hack;
-int singlestep;
int fd_bootchk = 1;
int graphic_rotate;
QEMUOptionRom option_rom[MAX_OPTION_ROMS];
--
2.34.1
next prev parent reply other threads:[~2023-04-17 16:42 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-17 16:40 [PATCH v3 00/10] Deprecate/rename singlestep command line option, monitor interfaces Peter Maydell
2023-04-17 16:40 ` [PATCH v3 01/10] make one-insn-per-tb an accel option Peter Maydell
2023-04-17 16:40 ` [PATCH v3 02/10] softmmu: Don't use 'singlestep' global in QMP and HMP commands Peter Maydell
2023-04-20 8:54 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` Peter Maydell [this message]
2023-04-18 8:02 ` [PATCH v3 03/10] accel/tcg: Use one_insn_per_tb global instead of old singlestep global Richard Henderson
2023-04-18 8:05 ` Richard Henderson
2023-04-18 9:44 ` Peter Maydell
2023-04-20 8:55 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 04/10] linux-user: Add '-one-insn-per-tb' option equivalent to '-singlestep' Peter Maydell
2023-04-20 9:13 ` Philippe Mathieu-Daudé
2023-04-20 9:19 ` Peter Maydell
2023-04-20 10:05 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 05/10] bsd-user: " Peter Maydell
2023-04-20 9:13 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 06/10] Document that -singlestep command line option is deprecated Peter Maydell
2023-04-20 8:56 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 07/10] accel/tcg: Report one-insn-per-tb in 'info jit', not 'info status' Peter Maydell
2023-04-18 8:06 ` Richard Henderson
2023-04-20 9:02 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 08/10] hmp: Add 'one-insn-per-tb' command equivalent to 'singlestep' Peter Maydell
2023-04-20 9:03 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 09/10] qapi/run-state.json: Fix missing newline at end of file Peter Maydell
2023-04-20 9:04 ` Philippe Mathieu-Daudé
2023-04-17 16:40 ` [PATCH v3 10/10] hmp: Deprecate 'singlestep' member of StatusInfo Peter Maydell
2023-04-18 8:08 ` Richard Henderson
2023-04-20 9:05 ` Philippe Mathieu-Daudé
2023-04-24 10:06 ` Peter Maydell
[not found] ` <87jzy18oqv.fsf@pond.sub.org>
2023-04-25 12:13 ` Peter Maydell
2023-04-25 13:08 ` Markus Armbruster
2023-05-02 10:22 ` [PATCH v3 00/10] Deprecate/rename singlestep command line option, monitor interfaces Peter Maydell
2023-05-02 10:43 ` Markus Armbruster
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=20230417164041.684562-4-peter.maydell@linaro.org \
--to=peter.maydell@linaro.org \
--cc=armbru@redhat.com \
--cc=eblake@redhat.com \
--cc=imp@bsdimp.com \
--cc=kevans@freebsd.org \
--cc=laurent@vivier.eu \
--cc=libvir-list@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.org \
/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).