[PATCH v3 14/20] util: fix interleaving of error & trace output

Daniel P. Berrangé posted 20 patches 2 weeks, 3 days ago
Maintainers: Gerd Hoffmann <kraxel@redhat.com>, Christian Schoenebeck <qemu_oss@crudebyte.com>, Markus Armbruster <armbru@redhat.com>, Paolo Bonzini <pbonzini@redhat.com>, "Marc-André Lureau" <marcandre.lureau@redhat.com>, "Daniel P. Berrangé" <berrange@redhat.com>, "Philippe Mathieu-Daudé" <philmd@linaro.org>, "Dr. David Alan Gilbert" <dave@treblig.org>, Kevin Wolf <kwolf@redhat.com>, Hanna Reitz <hreitz@redhat.com>, Stefan Weil <sw@weilnetz.de>
There is a newer version of this series
[PATCH v3 14/20] util: fix interleaving of error & trace output
Posted by Daniel P. Berrangé 2 weeks, 3 days ago
The monitor_cur_hmp() function will acquire/release mutex locks, which
will trigger trace probes, which can in turn trigger qemu_log() calls.
vreport() calls monitor_cur() multiple times through its execution
both directly and indirectly via error_vprintf().

The result is that the prefix information printed by vreport() gets
interleaved with qemu_log() output, when run outside the context of
an HMP command dispatcher. This can be seen with:

 $ qemu-system-x86_64
     -msg timestamp=on,guest-name=on
     -display none
     -object tls-creds-x509,id=f,dir=fish
     -name fish
     -d trace:qemu_mutex*
   2025-09-10T16:30:42.514374Z qemu_mutex_unlock released mutex 0x560b0339b4c0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:30:42.514400Z qemu_mutex_lock waiting on mutex 0x560b033983e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:30:42.514402Z qemu_mutex_locked taken mutex 0x560b033983e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:30:42.514404Z qemu_mutex_unlock released mutex 0x560b033983e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:30:42.516716Z qemu_mutex_lock waiting on mutex 0x560b03398560 (../monitor/monitor.c:91)
   2025-09-10T16:30:42.516723Z qemu_mutex_locked taken mutex 0x560b03398560 (../monitor/monitor.c:91)
   2025-09-10T16:30:42.516726Z qemu_mutex_unlock released mutex 0x560b03398560 (../monitor/monitor.c:96)
   2025-09-10T16:30:42.516728Z qemu_mutex_lock waiting on mutex 0x560b03398560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842057Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842058Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
   2025-09-10T16:31:04.842055Z 2025-09-10T16:31:04.842060Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842061Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842062Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
   2025-09-10T16:31:04.842064Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842065Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842066Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
   fish 2025-09-10T16:31:04.842068Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842069Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842070Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
   2025-09-10T16:31:04.842072Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842097Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842099Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
   qemu-system-x86_64:2025-09-10T16:31:04.842100Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842102Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842103Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
    2025-09-10T16:31:04.842105Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842106Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842107Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
   Unable to access credentials fish/ca-cert.pem: No such file or directory2025-09-10T16:31:04.842109Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842110Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
   2025-09-10T16:31:04.842111Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)

To avoid this interleaving (as well as reduce the huge number of
mutex lock/unlock calls) we need to ensure that monitor_cur_hmp() is
only called once at the start of vreport(), and if no HMP is present,
no further monitor APIs can be called.

This implies error_[v]printf() cannot be called from vreport().

Instead a decision about whether output will go to monitor_[v]printf()
or fprintf() must be made upfront, and then those functions called
directly.

 $ qemu-system-x86_64
     -msg timestamp=on,guest-name=on
     -display none
     -object tls-creds-x509,id=f,dir=fish
     -name fish
     -d trace:qemu_mutex*
   2025-09-10T16:31:22.701691Z qemu_mutex_unlock released mutex 0x5626fd3b84c0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:31:22.701728Z qemu_mutex_lock waiting on mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:31:22.701730Z qemu_mutex_locked taken mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:31:22.701732Z qemu_mutex_unlock released mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
   2025-09-10T16:31:22.703989Z qemu_mutex_lock waiting on mutex 0x5626fd3b5560 (../monitor/monitor.c:91)
   2025-09-10T16:31:22.703996Z qemu_mutex_locked taken mutex 0x5626fd3b5560 (../monitor/monitor.c:91)
   2025-09-10T16:31:22.703999Z qemu_mutex_unlock released mutex 0x5626fd3b5560 (../monitor/monitor.c:96)
   2025-09-10T16:31:22.704000Z fish qemu-system-x86_64: Unable to access credentials fish/ca-cert.pem: No such file or directory

This change requires adding a stub for monitor_printf() since that
is now called directly by vreport(), where monitor_vprintf() was
only needed previously.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
 stubs/monitor-core.c           |  5 +++
 tests/unit/test-util-sockets.c |  1 +
 util/error-report.c            | 60 ++++++++++++++++++++++++----------
 3 files changed, 48 insertions(+), 18 deletions(-)

diff --git a/stubs/monitor-core.c b/stubs/monitor-core.c
index 1e0b11ec29..70b1f7a994 100644
--- a/stubs/monitor-core.c
+++ b/stubs/monitor-core.c
@@ -30,3 +30,8 @@ int monitor_vprintf(Monitor *mon, const char *fmt, va_list ap)
 {
     abort();
 }
+
+int monitor_printf(Monitor *mon, const char *fmt, ...)
+{
+    abort();
+}
diff --git a/tests/unit/test-util-sockets.c b/tests/unit/test-util-sockets.c
index d40813c682..53733163dd 100644
--- a/tests/unit/test-util-sockets.c
+++ b/tests/unit/test-util-sockets.c
@@ -76,6 +76,7 @@ Monitor *monitor_cur_hmp(void) { return cur_mon; }
 bool monitor_cur_is_qmp(void) { return false; }
 Monitor *monitor_set_cur(Coroutine *co, Monitor *mon) { abort(); }
 int monitor_vprintf(Monitor *mon, const char *fmt, va_list ap) { abort(); }
+int monitor_printf(Monitor *mon, const char *fmt, ...) { abort(); }
 
 #ifndef _WIN32
 static void test_socket_fd_pass_name_good(void)
diff --git a/util/error-report.c b/util/error-report.c
index c0fcf388de..6c887047e3 100644
--- a/util/error-report.c
+++ b/util/error-report.c
@@ -29,6 +29,9 @@ bool message_with_timestamp;
 bool error_with_guestname;
 const char *error_guest_name;
 
+typedef void (*error_print_func)(void *opaque, const char *fmt, ...);
+typedef void (*error_print_vfunc)(void *opaque, const char *fmt, va_list ap);
+
 /*
  * Print to current monitor if we have one, else to stderr.
  */
@@ -151,34 +154,34 @@ void loc_set_file(const char *fname, int lno)
 /*
  * Print current location to current monitor if we have one, else to stderr.
  */
-static void print_loc(void)
+static void print_loc(bool cur, error_print_func print_func, void *print_opaque)
 {
     const char *sep = "";
     int i;
     const char *const *argp;
 
-    if (!monitor_cur_hmp() && g_get_prgname()) {
-        error_printf("%s:", g_get_prgname());
+    if (!cur && g_get_prgname()) {
+        print_func(print_opaque, "%s:", g_get_prgname());
         sep = " ";
     }
     switch (cur_loc->kind) {
     case LOC_CMDLINE:
         argp = cur_loc->ptr;
         for (i = 0; i < cur_loc->num; i++) {
-            error_printf("%s%s", sep, argp[i]);
+            print_func(print_opaque, "%s%s", sep, argp[i]);
             sep = " ";
         }
-        error_printf(": ");
+        print_func(print_opaque, ": ");
         break;
     case LOC_FILE:
-        error_printf("%s:", (const char *)cur_loc->ptr);
+        print_func(print_opaque, "%s:", (const char *)cur_loc->ptr);
         if (cur_loc->num) {
-            error_printf("%d:", cur_loc->num);
+            print_func(print_opaque, "%d:", cur_loc->num);
         }
-        error_printf(" ");
+        print_func(print_opaque, " ");
         break;
     default:
-        error_printf("%s", sep);
+        print_func(print_opaque, "%s", sep);
     }
 }
 
@@ -199,34 +202,55 @@ real_time_iso8601(void)
 G_GNUC_PRINTF(2, 0)
 static void vreport(report_type type, const char *fmt, va_list ap)
 {
+    /*
+     * Calling monitor_cur_hmp() will acquire/release mutexes,
+     * which triggers trace probes, which can trigger
+     * qemu_log calls, which would interleave with output
+     * from this. Hence cache the monitor handle upfront
+     * so any tracing appears before we start outputting.
+     */
+    Monitor *cur = monitor_cur_hmp();
+    error_print_func print_func;
+    error_print_vfunc print_vfunc;
+    void *print_opaque;
     gchar *timestr;
 
-    if (message_with_timestamp && !monitor_cur_hmp()) {
+    if (cur) {
+        print_func = (error_print_func)monitor_printf;
+        print_vfunc = (error_print_vfunc)monitor_vprintf;
+        print_opaque = cur;
+    } else {
+        print_func = (error_print_func)fprintf;
+        print_vfunc = (error_print_vfunc)vfprintf;
+        print_opaque = stderr;
+    }
+
+    if (message_with_timestamp && !cur) {
         timestr = real_time_iso8601();
-        error_printf("%s ", timestr);
+        print_func(print_opaque, "%s ", timestr);
         g_free(timestr);
     }
 
     /* Only prepend guest name if -msg guest-name and -name guest=... are set */
-    if (error_with_guestname && error_guest_name && !monitor_cur_hmp()) {
-        error_printf("%s ", error_guest_name);
+    if (error_with_guestname && error_guest_name && !cur) {
+        print_func(print_opaque, "%s ", error_guest_name);
     }
 
-    print_loc();
+    print_loc(!!cur, print_func, print_opaque);
 
     switch (type) {
     case REPORT_TYPE_ERROR:
         break;
     case REPORT_TYPE_WARNING:
-        error_printf("warning: ");
+        print_func(print_opaque, "warning: ");
         break;
     case REPORT_TYPE_INFO:
-        error_printf("info: ");
+        print_func(print_opaque, "info: ");
         break;
     }
 
-    error_vprintf(fmt, ap);
-    error_printf("\n");
+    print_vfunc(print_opaque, fmt, ap);
+    print_func(print_opaque, "\n");
 }
 
 /*
-- 
2.50.1


Re: [PATCH v3 14/20] util: fix interleaving of error & trace output
Posted by Markus Armbruster 6 days, 4 hours ago
Daniel P. Berrangé <berrange@redhat.com> writes:

> The monitor_cur_hmp() function will acquire/release mutex locks, which
> will trigger trace probes, which can in turn trigger qemu_log() calls.
> vreport() calls monitor_cur() multiple times through its execution
> both directly and indirectly via error_vprintf().

This issue goes back to commit e69ee454b5f9 (monitor: Make current
monitor a per-coroutine property).  I've always disliked the hash table
(which is why we need to take the mutex), and now I dislike it even
more.  See also my review of PATCH 02.

I'd love to have a lockless monitor_cur() again.  But I don't have one
up my sleeve, so we need to work around the issue somehow.

> The result is that the prefix information printed by vreport() gets
> interleaved with qemu_log() output, when run outside the context of
> an HMP command dispatcher. This can be seen with:

[...]

> To avoid this interleaving (as well as reduce the huge number of
> mutex lock/unlock calls) we need to ensure that monitor_cur_hmp() is
> only called once at the start of vreport(), and if no HMP is present,
> no further monitor APIs can be called.
>
> This implies error_[v]printf() cannot be called from vreport().
>
> Instead a decision about whether output will go to monitor_[v]printf()
> or fprintf() must be made upfront, and then those functions called
> directly.
>
>  $ qemu-system-x86_64
>      -msg timestamp=on,guest-name=on
>      -display none
>      -object tls-creds-x509,id=f,dir=fish
>      -name fish
>      -d trace:qemu_mutex*
>    2025-09-10T16:31:22.701691Z qemu_mutex_unlock released mutex 0x5626fd3b84c0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>    2025-09-10T16:31:22.701728Z qemu_mutex_lock waiting on mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>    2025-09-10T16:31:22.701730Z qemu_mutex_locked taken mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>    2025-09-10T16:31:22.701732Z qemu_mutex_unlock released mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>    2025-09-10T16:31:22.703989Z qemu_mutex_lock waiting on mutex 0x5626fd3b5560 (../monitor/monitor.c:91)
>    2025-09-10T16:31:22.703996Z qemu_mutex_locked taken mutex 0x5626fd3b5560 (../monitor/monitor.c:91)
>    2025-09-10T16:31:22.703999Z qemu_mutex_unlock released mutex 0x5626fd3b5560 (../monitor/monitor.c:96)
>    2025-09-10T16:31:22.704000Z fish qemu-system-x86_64: Unable to access credentials fish/ca-cert.pem: No such file or directory
>
> This change requires adding a stub for monitor_printf() since that
> is now called directly by vreport(), where monitor_vprintf() was
> only needed previously.
>
> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
> ---
>  stubs/monitor-core.c           |  5 +++
>  tests/unit/test-util-sockets.c |  1 +
>  util/error-report.c            | 60 ++++++++++++++++++++++++----------
>  3 files changed, 48 insertions(+), 18 deletions(-)
>
> diff --git a/stubs/monitor-core.c b/stubs/monitor-core.c
> index 1e0b11ec29..70b1f7a994 100644
> --- a/stubs/monitor-core.c
> +++ b/stubs/monitor-core.c
> @@ -30,3 +30,8 @@ int monitor_vprintf(Monitor *mon, const char *fmt, va_list ap)
>  {
>      abort();
>  }
> +
> +int monitor_printf(Monitor *mon, const char *fmt, ...)
> +{
> +    abort();
> +}
> diff --git a/tests/unit/test-util-sockets.c b/tests/unit/test-util-sockets.c
> index d40813c682..53733163dd 100644
> --- a/tests/unit/test-util-sockets.c
> +++ b/tests/unit/test-util-sockets.c
> @@ -76,6 +76,7 @@ Monitor *monitor_cur_hmp(void) { return cur_mon; }
>  bool monitor_cur_is_qmp(void) { return false; }
>  Monitor *monitor_set_cur(Coroutine *co, Monitor *mon) { abort(); }
>  int monitor_vprintf(Monitor *mon, const char *fmt, va_list ap) { abort(); }
> +int monitor_printf(Monitor *mon, const char *fmt, ...) { abort(); }
>  
>  #ifndef _WIN32
>  static void test_socket_fd_pass_name_good(void)
> diff --git a/util/error-report.c b/util/error-report.c
> index c0fcf388de..6c887047e3 100644
> --- a/util/error-report.c
> +++ b/util/error-report.c
> @@ -29,6 +29,9 @@ bool message_with_timestamp;
>  bool error_with_guestname;
>  const char *error_guest_name;
>  
> +typedef void (*error_print_func)(void *opaque, const char *fmt, ...);
> +typedef void (*error_print_vfunc)(void *opaque, const char *fmt, va_list ap);
> +
>  /*
>   * Print to current monitor if we have one, else to stderr.
>   */
> @@ -151,34 +154,34 @@ void loc_set_file(const char *fname, int lno)
>  /*
>   * Print current location to current monitor if we have one, else to stderr.
>   */
> -static void print_loc(void)
> +static void print_loc(bool cur, error_print_func print_func, void *print_opaque)
>  {
>      const char *sep = "";
>      int i;
>      const char *const *argp;
>  
> -    if (!monitor_cur_hmp() && g_get_prgname()) {
> -        error_printf("%s:", g_get_prgname());
> +    if (!cur && g_get_prgname()) {
> +        print_func(print_opaque, "%s:", g_get_prgname());
>          sep = " ";
>      }
>      switch (cur_loc->kind) {
>      case LOC_CMDLINE:
>          argp = cur_loc->ptr;
>          for (i = 0; i < cur_loc->num; i++) {
> -            error_printf("%s%s", sep, argp[i]);
> +            print_func(print_opaque, "%s%s", sep, argp[i]);
>              sep = " ";
>          }
> -        error_printf(": ");
> +        print_func(print_opaque, ": ");
>          break;
>      case LOC_FILE:
> -        error_printf("%s:", (const char *)cur_loc->ptr);
> +        print_func(print_opaque, "%s:", (const char *)cur_loc->ptr);
>          if (cur_loc->num) {
> -            error_printf("%d:", cur_loc->num);
> +            print_func(print_opaque, "%d:", cur_loc->num);
>          }
> -        error_printf(" ");
> +        print_func(print_opaque, " ");
>          break;
>      default:
> -        error_printf("%s", sep);
> +        print_func(print_opaque, "%s", sep);
>      }
>  }
>  
> @@ -199,34 +202,55 @@ real_time_iso8601(void)
>  G_GNUC_PRINTF(2, 0)
>  static void vreport(report_type type, const char *fmt, va_list ap)
>  {
> +    /*
> +     * Calling monitor_cur_hmp() will acquire/release mutexes,
> +     * which triggers trace probes, which can trigger
> +     * qemu_log calls, which would interleave with output
> +     * from this. Hence cache the monitor handle upfront
> +     * so any tracing appears before we start outputting.
> +     */
> +    Monitor *cur = monitor_cur_hmp();
> +    error_print_func print_func;
> +    error_print_vfunc print_vfunc;
> +    void *print_opaque;
>      gchar *timestr;
>  
> -    if (message_with_timestamp && !monitor_cur_hmp()) {
> +    if (cur) {
> +        print_func = (error_print_func)monitor_printf;
> +        print_vfunc = (error_print_vfunc)monitor_vprintf;
> +        print_opaque = cur;
> +    } else {
> +        print_func = (error_print_func)fprintf;
> +        print_vfunc = (error_print_vfunc)vfprintf;
> +        print_opaque = stderr;
> +    }

print_func is is void (*)(void *, const char *, ...).

monitor_printf() is int (*)(Monitor *, const char *, ...).

fprintf() is int (*)(FILE *, const char *, ...)

ISO/IEC 9899:1999 §6.3.2.3 Pointers:

       [#8] A pointer to a function of one type may be converted to
       a pointer to a function of another type and back again;  the
       result  shall  compare  equal to the original pointer.  If a
       converted pointer is used to call a function whose  type  is
       not  compatible  with  the  pointed-to type, the behavior is
       undefined.

For this to work in practice, three conditions need to hold:

1. Both Monitor * and FILE * arguments are passed exactly like void *
arguments.  Holds on sane machines.

2. A function returning void is called the same way as a function
returning int when the value is ignored.  Holds on sane machines.  We
can easily eliminate this condition: just change print_func from void to
int.

3. The guys writing the compiler refrain from making use of their
self-granted license to ruin our day because our program has undefined
behavior.  Since improving some SPEC benchmark by half a percent is
clearly worth breaking any number of working programs, ...

If we want to stay out of this swamp, we need to define wrapper
functions taking void * and returning it.

> +
> +    if (message_with_timestamp && !cur) {
>          timestr = real_time_iso8601();
> -        error_printf("%s ", timestr);
> +        print_func(print_opaque, "%s ", timestr);
>          g_free(timestr);
>      }
>  
>      /* Only prepend guest name if -msg guest-name and -name guest=... are set */
> -    if (error_with_guestname && error_guest_name && !monitor_cur_hmp()) {
> -        error_printf("%s ", error_guest_name);
> +    if (error_with_guestname && error_guest_name && !cur) {
> +        print_func(print_opaque, "%s ", error_guest_name);
>      }
>  
> -    print_loc();
> +    print_loc(!!cur, print_func, print_opaque);
>  
>      switch (type) {
>      case REPORT_TYPE_ERROR:
>          break;
>      case REPORT_TYPE_WARNING:
> -        error_printf("warning: ");
> +        print_func(print_opaque, "warning: ");
>          break;
>      case REPORT_TYPE_INFO:
> -        error_printf("info: ");
> +        print_func(print_opaque, "info: ");
>          break;
>      }
>  
> -    error_vprintf(fmt, ap);
> -    error_printf("\n");
> +    print_vfunc(print_opaque, fmt, ap);
> +    print_func(print_opaque, "\n");
>  }
>  
>  /*

The code becomes more complicated.  Hmm.

As long as monitor_cur() hits trace points, we really want to call it
just once, before we start printing the error message.

Calls during vreport() (in master, because I can't be bothered to
recompile):

(1) At the beginning of vreport().  Fine.

(2) In the middle of vreport().  Easy to avoid: reuse the value gotten
at (1).

(3) In print_loc().  Easy to avoid: called only by vreport(), so pass
the value gotten at (1).

(4) In error_vprintf(), called multiple times from vreport() and
print_loc() directly and via error_printf().  Obvious, stupid solution:
have error_vprintf() and error_printf() take a Monitor *argument.  But
that's no good, because we have numerous callers, and we'd rather not
change these to pass monitor_cur().  So create two new functions, and
make the existing ones wrap around them.

Thoughts?
Re: [PATCH v3 14/20] util: fix interleaving of error & trace output
Posted by Richard Henderson 2 weeks, 2 days ago
On 9/10/25 18:03, Daniel P. Berrangé wrote:
> +    } else {
> +        print_func = (error_print_func)fprintf;
> +        print_vfunc = (error_print_vfunc)vfprintf;
> +        print_opaque = stderr;
> +    }
> +
> +    if (message_with_timestamp && !cur) {
>           timestr = real_time_iso8601();
> -        error_printf("%s ", timestr);
> +        print_func(print_opaque, "%s ", timestr);
>           g_free(timestr);
>       }
>   
>       /* Only prepend guest name if -msg guest-name and -name guest=... are set */
> -    if (error_with_guestname && error_guest_name && !monitor_cur_hmp()) {
> -        error_printf("%s ", error_guest_name);
> +    if (error_with_guestname && error_guest_name && !cur) {
> +        print_func(print_opaque, "%s ", error_guest_name);
>       }

It's worth pushing these second and third if into the first else,
so that !cur is already true.


r~

Re: [PATCH v3 14/20] util: fix interleaving of error & trace output
Posted by Daniel P. Berrangé 2 weeks, 2 days ago
On Thu, Sep 11, 2025 at 06:05:01PM +0000, Richard Henderson wrote:
> On 9/10/25 18:03, Daniel P. Berrangé wrote:
> > +    } else {
> > +        print_func = (error_print_func)fprintf;
> > +        print_vfunc = (error_print_vfunc)vfprintf;
> > +        print_opaque = stderr;
> > +    }
> > +
> > +    if (message_with_timestamp && !cur) {
> >           timestr = real_time_iso8601();
> > -        error_printf("%s ", timestr);
> > +        print_func(print_opaque, "%s ", timestr);
> >           g_free(timestr);
> >       }
> >       /* Only prepend guest name if -msg guest-name and -name guest=... are set */
> > -    if (error_with_guestname && error_guest_name && !monitor_cur_hmp()) {
> > -        error_printf("%s ", error_guest_name);
> > +    if (error_with_guestname && error_guest_name && !cur) {
> > +        print_func(print_opaque, "%s ", error_guest_name);
> >       }
> 
> It's worth pushing these second and third if into the first else,
> so that !cur is already true.

You've probably seen now that these 'if' blocks go away in the later
patch and the "qmessage_context_print(stderr);" that replaces them
is indeed inside the first 'else'.


With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|


Re: [PATCH v3 14/20] util: fix interleaving of error & trace output
Posted by Richard Henderson 2 weeks, 2 days ago
On 9/10/25 18:03, Daniel P. Berrangé wrote:
> The monitor_cur_hmp() function will acquire/release mutex locks, which
> will trigger trace probes, which can in turn trigger qemu_log() calls.
> vreport() calls monitor_cur() multiple times through its execution
> both directly and indirectly via error_vprintf().
> 
> The result is that the prefix information printed by vreport() gets
> interleaved with qemu_log() output, when run outside the context of
> an HMP command dispatcher. This can be seen with:
> 
>   $ qemu-system-x86_64
>       -msg timestamp=on,guest-name=on
>       -display none
>       -object tls-creds-x509,id=f,dir=fish
>       -name fish
>       -d trace:qemu_mutex*
>     2025-09-10T16:30:42.514374Z qemu_mutex_unlock released mutex 0x560b0339b4c0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:30:42.514400Z qemu_mutex_lock waiting on mutex 0x560b033983e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:30:42.514402Z qemu_mutex_locked taken mutex 0x560b033983e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:30:42.514404Z qemu_mutex_unlock released mutex 0x560b033983e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:30:42.516716Z qemu_mutex_lock waiting on mutex 0x560b03398560 (../monitor/monitor.c:91)
>     2025-09-10T16:30:42.516723Z qemu_mutex_locked taken mutex 0x560b03398560 (../monitor/monitor.c:91)
>     2025-09-10T16:30:42.516726Z qemu_mutex_unlock released mutex 0x560b03398560 (../monitor/monitor.c:96)
>     2025-09-10T16:30:42.516728Z qemu_mutex_lock waiting on mutex 0x560b03398560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842057Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842058Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>     2025-09-10T16:31:04.842055Z 2025-09-10T16:31:04.842060Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842061Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842062Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>     2025-09-10T16:31:04.842064Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842065Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842066Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>     fish 2025-09-10T16:31:04.842068Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842069Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842070Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>     2025-09-10T16:31:04.842072Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842097Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842099Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>     qemu-system-x86_64:2025-09-10T16:31:04.842100Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842102Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842103Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>      2025-09-10T16:31:04.842105Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842106Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842107Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
>     Unable to access credentials fish/ca-cert.pem: No such file or directory2025-09-10T16:31:04.842109Z qemu_mutex_lock waiting on mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842110Z qemu_mutex_locked taken mutex 0x564f5e401560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:04.842111Z qemu_mutex_unlock released mutex 0x564f5e401560 (../monitor/monitor.c:96)
> 
> To avoid this interleaving (as well as reduce the huge number of
> mutex lock/unlock calls) we need to ensure that monitor_cur_hmp() is
> only called once at the start of vreport(), and if no HMP is present,
> no further monitor APIs can be called.
> 
> This implies error_[v]printf() cannot be called from vreport().
> 
> Instead a decision about whether output will go to monitor_[v]printf()
> or fprintf() must be made upfront, and then those functions called
> directly.
> 
>   $ qemu-system-x86_64
>       -msg timestamp=on,guest-name=on
>       -display none
>       -object tls-creds-x509,id=f,dir=fish
>       -name fish
>       -d trace:qemu_mutex*
>     2025-09-10T16:31:22.701691Z qemu_mutex_unlock released mutex 0x5626fd3b84c0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:31:22.701728Z qemu_mutex_lock waiting on mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:31:22.701730Z qemu_mutex_locked taken mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:31:22.701732Z qemu_mutex_unlock released mutex 0x5626fd3b53e0 (/var/home/berrange/src/virt/qemu/include/qemu/lockable.h:56)
>     2025-09-10T16:31:22.703989Z qemu_mutex_lock waiting on mutex 0x5626fd3b5560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:22.703996Z qemu_mutex_locked taken mutex 0x5626fd3b5560 (../monitor/monitor.c:91)
>     2025-09-10T16:31:22.703999Z qemu_mutex_unlock released mutex 0x5626fd3b5560 (../monitor/monitor.c:96)
>     2025-09-10T16:31:22.704000Z fish qemu-system-x86_64: Unable to access credentials fish/ca-cert.pem: No such file or directory
> 
> This change requires adding a stub for monitor_printf() since that
> is now called directly by vreport(), where monitor_vprintf() was
> only needed previously.
> 
> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
> ---
>   stubs/monitor-core.c           |  5 +++
>   tests/unit/test-util-sockets.c |  1 +
>   util/error-report.c            | 60 ++++++++++++++++++++++++----------
>   3 files changed, 48 insertions(+), 18 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>


r~

> 
> diff --git a/stubs/monitor-core.c b/stubs/monitor-core.c
> index 1e0b11ec29..70b1f7a994 100644
> --- a/stubs/monitor-core.c
> +++ b/stubs/monitor-core.c
> @@ -30,3 +30,8 @@ int monitor_vprintf(Monitor *mon, const char *fmt, va_list ap)
>   {
>       abort();
>   }
> +
> +int monitor_printf(Monitor *mon, const char *fmt, ...)
> +{
> +    abort();
> +}
> diff --git a/tests/unit/test-util-sockets.c b/tests/unit/test-util-sockets.c
> index d40813c682..53733163dd 100644
> --- a/tests/unit/test-util-sockets.c
> +++ b/tests/unit/test-util-sockets.c
> @@ -76,6 +76,7 @@ Monitor *monitor_cur_hmp(void) { return cur_mon; }
>   bool monitor_cur_is_qmp(void) { return false; }
>   Monitor *monitor_set_cur(Coroutine *co, Monitor *mon) { abort(); }
>   int monitor_vprintf(Monitor *mon, const char *fmt, va_list ap) { abort(); }
> +int monitor_printf(Monitor *mon, const char *fmt, ...) { abort(); }
>   
>   #ifndef _WIN32
>   static void test_socket_fd_pass_name_good(void)
> diff --git a/util/error-report.c b/util/error-report.c
> index c0fcf388de..6c887047e3 100644
> --- a/util/error-report.c
> +++ b/util/error-report.c
> @@ -29,6 +29,9 @@ bool message_with_timestamp;
>   bool error_with_guestname;
>   const char *error_guest_name;
>   
> +typedef void (*error_print_func)(void *opaque, const char *fmt, ...);
> +typedef void (*error_print_vfunc)(void *opaque, const char *fmt, va_list ap);
> +
>   /*
>    * Print to current monitor if we have one, else to stderr.
>    */
> @@ -151,34 +154,34 @@ void loc_set_file(const char *fname, int lno)
>   /*
>    * Print current location to current monitor if we have one, else to stderr.
>    */
> -static void print_loc(void)
> +static void print_loc(bool cur, error_print_func print_func, void *print_opaque)
>   {
>       const char *sep = "";
>       int i;
>       const char *const *argp;
>   
> -    if (!monitor_cur_hmp() && g_get_prgname()) {
> -        error_printf("%s:", g_get_prgname());
> +    if (!cur && g_get_prgname()) {
> +        print_func(print_opaque, "%s:", g_get_prgname());
>           sep = " ";
>       }
>       switch (cur_loc->kind) {
>       case LOC_CMDLINE:
>           argp = cur_loc->ptr;
>           for (i = 0; i < cur_loc->num; i++) {
> -            error_printf("%s%s", sep, argp[i]);
> +            print_func(print_opaque, "%s%s", sep, argp[i]);
>               sep = " ";
>           }
> -        error_printf(": ");
> +        print_func(print_opaque, ": ");
>           break;
>       case LOC_FILE:
> -        error_printf("%s:", (const char *)cur_loc->ptr);
> +        print_func(print_opaque, "%s:", (const char *)cur_loc->ptr);
>           if (cur_loc->num) {
> -            error_printf("%d:", cur_loc->num);
> +            print_func(print_opaque, "%d:", cur_loc->num);
>           }
> -        error_printf(" ");
> +        print_func(print_opaque, " ");
>           break;
>       default:
> -        error_printf("%s", sep);
> +        print_func(print_opaque, "%s", sep);
>       }
>   }
>   
> @@ -199,34 +202,55 @@ real_time_iso8601(void)
>   G_GNUC_PRINTF(2, 0)
>   static void vreport(report_type type, const char *fmt, va_list ap)
>   {
> +    /*
> +     * Calling monitor_cur_hmp() will acquire/release mutexes,
> +     * which triggers trace probes, which can trigger
> +     * qemu_log calls, which would interleave with output
> +     * from this. Hence cache the monitor handle upfront
> +     * so any tracing appears before we start outputting.
> +     */
> +    Monitor *cur = monitor_cur_hmp();
> +    error_print_func print_func;
> +    error_print_vfunc print_vfunc;
> +    void *print_opaque;
>       gchar *timestr;
>   
> -    if (message_with_timestamp && !monitor_cur_hmp()) {
> +    if (cur) {
> +        print_func = (error_print_func)monitor_printf;
> +        print_vfunc = (error_print_vfunc)monitor_vprintf;
> +        print_opaque = cur;
> +    } else {
> +        print_func = (error_print_func)fprintf;
> +        print_vfunc = (error_print_vfunc)vfprintf;
> +        print_opaque = stderr;
> +    }
> +
> +    if (message_with_timestamp && !cur) {
>           timestr = real_time_iso8601();
> -        error_printf("%s ", timestr);
> +        print_func(print_opaque, "%s ", timestr);
>           g_free(timestr);
>       }
>   
>       /* Only prepend guest name if -msg guest-name and -name guest=... are set */
> -    if (error_with_guestname && error_guest_name && !monitor_cur_hmp()) {
> -        error_printf("%s ", error_guest_name);
> +    if (error_with_guestname && error_guest_name && !cur) {
> +        print_func(print_opaque, "%s ", error_guest_name);
>       }
>   
> -    print_loc();
> +    print_loc(!!cur, print_func, print_opaque);
>   
>       switch (type) {
>       case REPORT_TYPE_ERROR:
>           break;
>       case REPORT_TYPE_WARNING:
> -        error_printf("warning: ");
> +        print_func(print_opaque, "warning: ");
>           break;
>       case REPORT_TYPE_INFO:
> -        error_printf("info: ");
> +        print_func(print_opaque, "info: ");
>           break;
>       }
>   
> -    error_vprintf(fmt, ap);
> -    error_printf("\n");
> +    print_vfunc(print_opaque, fmt, ap);
> +    print_func(print_opaque, "\n");
>   }
>   
>   /*