[PULL 0/5] Tracing patches

[PULL 0/5] Tracing patches

[Stefan Hajnoczi]

The following changes since commit 928173659d6e5dc368284f73f90ea1d129e1f57d:

  Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200130' into staging (2020-01-30 16:19:04 +0000)

are available in the Git repository at:

  https://github.com/stefanha/qemu.git tags/tracing-pull-request

for you to fetch changes up to e144a605a614d22165000c69e8e1dc6986d45cd8:

  qemu_set_log_filename: filename argument may be NULL (2020-01-30 21:33:50 +0000)

----------------------------------------------------------------
Pull request

----------------------------------------------------------------

Peter Maydell (3):
  docs/devel/tracing.txt: Recommend only
    trace_event_get_state_backends()
  memory.c: Use trace_event_get_state_backends()
  hw/display/qxl.c: Use trace_event_get_state_backends()

Philippe Mathieu-Daudé (1):
  Makefile: Keep trace-events-subdirs ordered

Salvador Fandino (1):
  qemu_set_log_filename: filename argument may be NULL

 Makefile.objs          |  4 ++--
 docs/devel/tracing.txt | 12 +++++-------
 hw/display/qxl.c       |  2 +-
 memory.c               |  8 ++++----
 trace/control.c        |  4 +---
 util/log.c             | 28 ++++++++++++++++------------
 vl.c                   |  5 +----
 7 files changed, 30 insertions(+), 33 deletions(-)

-- 
2.24.1

[PULL 1/5] Makefile: Keep trace-events-subdirs ordered

[Stefan Hajnoczi]

From: Philippe Mathieu-Daudé <philmd@redhat.com>

Adding the same directory multiple times to trace-events-subdirs
might trigger build failures, in particular when using the LTTng
Userspace Tracer library as backend.

For example when using two times the hw/core/ directory:

    $ ./configure --enable-trace-backends=ust && make
   [...]
     CC      trace-ust-all.o
   In file included from trace-ust-all.h:13,
                    from trace-ust-all.c:13:
   trace-ust-all.h:35151:1: error: redefinition of ‘__tracepoint_cb_qemu___loader_write_rom’
   35151 | TRACEPOINT_EVENT(
         | ^~~~~~~~~~~~~~~~
   trace-ust-all.h:31791:1: note: previous definition of ‘__tracepoint_cb_qemu___loader_write_rom’ was here
   31791 | TRACEPOINT_EVENT(
         | ^~~~~~~~~~~~~~~~

To ease review and reduce likelihood of merge failures (see [*]),
keep trace-events-subdirs ordered when possible, following eb7ccb3c0.

[*] https://www.mail-archive.com/qemu-devel@nongnu.org/msg671007.html
    Duplicate trace-events-subdirs entries generates duplicated
    symbols when using the LTTng Userspace Tracer library.

Reviewed-by: Damien Hedde <damien.hedde@greensocs.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200116114339.30670-1-philmd@redhat.com
Message-Id: <20200116114339.30670-1-philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 Makefile.objs | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/Makefile.objs b/Makefile.objs
index f2da7241fc..26b9cff954 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -132,8 +132,8 @@ trace-events-subdirs += nbd
 trace-events-subdirs += scsi
 endif
 ifeq ($(CONFIG_SOFTMMU),y)
-trace-events-subdirs += chardev
 trace-events-subdirs += audio
+trace-events-subdirs += chardev
 trace-events-subdirs += hw/9pfs
 trace-events-subdirs += hw/acpi
 trace-events-subdirs += hw/alpha
@@ -181,6 +181,7 @@ trace-events-subdirs += migration
 trace-events-subdirs += net
 trace-events-subdirs += ui
 endif
+trace-events-subdirs += hw/core
 trace-events-subdirs += hw/display
 trace-events-subdirs += qapi
 trace-events-subdirs += qom
@@ -193,7 +194,6 @@ trace-events-subdirs += target/riscv
 trace-events-subdirs += target/s390x
 trace-events-subdirs += target/sparc
 trace-events-subdirs += util
-trace-events-subdirs += hw/core
 
 trace-events-files = $(SRC_PATH)/trace-events $(trace-events-subdirs:%=$(SRC_PATH)/%/trace-events)
 
-- 
2.24.1

[PULL 2/5] docs/devel/tracing.txt: Recommend only trace_event_get_state_backends()

[Stefan Hajnoczi]

From: Peter Maydell <peter.maydell@linaro.org>

Instead of recommending checking the TRACE_FOO_ENABLED macro to
skip expensive computations needed only for tracing, recommend
only using trace_event_get_state_backends(). This works for both
compile-time and run-time disabling of events, and has no extra
performance impact if the event is compile-time disabled.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200120151142.18954-2-peter.maydell@linaro.org
Message-Id: <20200120151142.18954-2-peter.maydell@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 docs/devel/tracing.txt | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/docs/devel/tracing.txt b/docs/devel/tracing.txt
index 8c0376fefa..cb5f685de9 100644
--- a/docs/devel/tracing.txt
+++ b/docs/devel/tracing.txt
@@ -342,8 +342,10 @@ edit the "trace-events-all" file).
 
 In addition, there might be cases where relatively complex computations must be
 performed to generate values that are only used as arguments for a trace
-function. In these cases you can use the macro 'TRACE_${EVENT_NAME}_ENABLED' to
-guard such computations and avoid its compilation when the event is disabled:
+function. In these cases you can use 'trace_event_get_state_backends()' to
+guard such computations, so they are skipped if the event has been either
+compile-time disabled or run-time disabled. If the event is compile-time
+disabled, this check will have no performance impact.
 
     #include "trace.h"  /* needed for trace event prototype */
     
@@ -356,7 +358,7 @@ guard such computations and avoid its compilation when the event is disabled:
             align = getpagesize();
         }
         ptr = qemu_memalign(align, size);
-        if (TRACE_QEMU_VMALLOC_ENABLED) { /* preprocessor macro */
+        if (trace_event_get_state_backends(TRACE_QEMU_VMALLOC)) {
             void *complex;
             /* some complex computations to produce the 'complex' value */
             trace_qemu_vmalloc(size, ptr, complex);
@@ -364,10 +366,6 @@ guard such computations and avoid its compilation when the event is disabled:
         return ptr;
     }
 
-You can check both if the event has been disabled and is dynamically enabled at
-the same time using the 'trace_event_get_state_backends' routine (see header
-"trace/control.h" for more information).
-
 === "tcg" ===
 
 Guest code generated by TCG can be traced by defining an event with the "tcg"
-- 
2.24.1

[PULL 3/5] memory.c: Use trace_event_get_state_backends()

[Stefan Hajnoczi]

From: Peter Maydell <peter.maydell@linaro.org>

The preferred way to test whether a trace event is enabled is to
use trace_event_get_state_backends(), because this will give the
correct answer (allowing expensive computations to be skipped)
whether the trace event is compile-time or run-time disabled.
Convert the four old-style direct uses of TRACE_FOO_ENABLED in
memory.c.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20200120151142.18954-3-peter.maydell@linaro.org
Message-Id: <20200120151142.18954-3-peter.maydell@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 memory.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/memory.c b/memory.c
index 854798791e..aeaa8dcc9e 100644
--- a/memory.c
+++ b/memory.c
@@ -434,7 +434,7 @@ static MemTxResult  memory_region_read_accessor(MemoryRegion *mr,
     tmp = mr->ops->read(mr->opaque, addr, size);
     if (mr->subpage) {
         trace_memory_region_subpage_read(get_cpu_index(), mr, addr, tmp, size);
-    } else if (TRACE_MEMORY_REGION_OPS_READ_ENABLED) {
+    } else if (trace_event_get_state_backends(TRACE_MEMORY_REGION_OPS_READ)) {
         hwaddr abs_addr = memory_region_to_absolute_addr(mr, addr);
         trace_memory_region_ops_read(get_cpu_index(), mr, abs_addr, tmp, size);
     }
@@ -456,7 +456,7 @@ static MemTxResult memory_region_read_with_attrs_accessor(MemoryRegion *mr,
     r = mr->ops->read_with_attrs(mr->opaque, addr, &tmp, size, attrs);
     if (mr->subpage) {
         trace_memory_region_subpage_read(get_cpu_index(), mr, addr, tmp, size);
-    } else if (TRACE_MEMORY_REGION_OPS_READ_ENABLED) {
+    } else if (trace_event_get_state_backends(TRACE_MEMORY_REGION_OPS_READ)) {
         hwaddr abs_addr = memory_region_to_absolute_addr(mr, addr);
         trace_memory_region_ops_read(get_cpu_index(), mr, abs_addr, tmp, size);
     }
@@ -476,7 +476,7 @@ static MemTxResult memory_region_write_accessor(MemoryRegion *mr,
 
     if (mr->subpage) {
         trace_memory_region_subpage_write(get_cpu_index(), mr, addr, tmp, size);
-    } else if (TRACE_MEMORY_REGION_OPS_WRITE_ENABLED) {
+    } else if (trace_event_get_state_backends(TRACE_MEMORY_REGION_OPS_WRITE)) {
         hwaddr abs_addr = memory_region_to_absolute_addr(mr, addr);
         trace_memory_region_ops_write(get_cpu_index(), mr, abs_addr, tmp, size);
     }
@@ -496,7 +496,7 @@ static MemTxResult memory_region_write_with_attrs_accessor(MemoryRegion *mr,
 
     if (mr->subpage) {
         trace_memory_region_subpage_write(get_cpu_index(), mr, addr, tmp, size);
-    } else if (TRACE_MEMORY_REGION_OPS_WRITE_ENABLED) {
+    } else if (trace_event_get_state_backends(TRACE_MEMORY_REGION_OPS_WRITE)) {
         hwaddr abs_addr = memory_region_to_absolute_addr(mr, addr);
         trace_memory_region_ops_write(get_cpu_index(), mr, abs_addr, tmp, size);
     }
-- 
2.24.1

[PULL 4/5] hw/display/qxl.c: Use trace_event_get_state_backends()

[Stefan Hajnoczi]

From: Peter Maydell <peter.maydell@linaro.org>

The preferred way to test whether a trace event is enabled is to
use trace_event_get_state_backends(), because this will give the
correct answer (allowing expensive computations to be skipped)
whether the trace event is compile-time or run-time disabled.
Convert the old-style direct use of TRACE_FOO_ENABLED.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20200120151142.18954-4-peter.maydell@linaro.org
Message-Id: <20200120151142.18954-4-peter.maydell@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 hw/display/qxl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/hw/display/qxl.c b/hw/display/qxl.c
index 944c02ce56..c33b1915a5 100644
--- a/hw/display/qxl.c
+++ b/hw/display/qxl.c
@@ -1764,7 +1764,7 @@ async_common:
         qxl_set_mode(d, val, 0);
         break;
     case QXL_IO_LOG:
-        if (TRACE_QXL_IO_LOG_ENABLED || d->guestdebug) {
+        if (trace_event_get_state_backends(TRACE_QXL_IO_LOG) || d->guestdebug) {
             /* We cannot trust the guest to NUL terminate d->ram->log_buf */
             char *log_buf = g_strndup((const char *)d->ram->log_buf,
                                       sizeof(d->ram->log_buf));
-- 
2.24.1

[PULL 5/5] qemu_set_log_filename: filename argument may be NULL

[Stefan Hajnoczi]

From: Salvador Fandino <salvador@qindel.com>

NULL is a valid log filename used to indicate we want to use stderr
but qemu_set_log_filename (which is called by bsd-user/main.c) was not
handling it correctly.

That also made redundant a couple of NULL checks in calling code which
have been removed.

Signed-off-by: Salvador Fandino <salvador@qindel.com>
Message-Id: <20200123193626.19956-1-salvador@qindel.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 trace/control.c |  4 +---
 util/log.c      | 28 ++++++++++++++++------------
 vl.c            |  5 +----
 3 files changed, 18 insertions(+), 19 deletions(-)

diff --git a/trace/control.c b/trace/control.c
index 0fb8124160..6c775e68eb 100644
--- a/trace/control.c
+++ b/trace/control.c
@@ -229,9 +229,7 @@ void trace_init_file(const char *file)
     /* If both the simple and the log backends are enabled, "--trace file"
      * only applies to the simple backend; use "-D" for the log backend.
      */
-    if (file) {
-        qemu_set_log_filename(file, &error_fatal);
-    }
+    qemu_set_log_filename(file, &error_fatal);
 #else
     if (file) {
         fprintf(stderr, "error: --trace file=...: "
diff --git a/util/log.c b/util/log.c
index 867264da8d..47f2827397 100644
--- a/util/log.c
+++ b/util/log.c
@@ -148,25 +148,29 @@ void qemu_log_needs_buffers(void)
  * Allow the user to include %d in their logfile which will be
  * substituted with the current PID. This is useful for debugging many
  * nested linux-user tasks but will result in lots of logs.
+ *
+ * filename may be NULL. In that case, log output is sent to stderr
  */
 void qemu_set_log_filename(const char *filename, Error **errp)
 {
-    char *pidstr;
     g_free(logfilename);
     logfilename = NULL;
 
-    pidstr = strstr(filename, "%");
-    if (pidstr) {
-        /* We only accept one %d, no other format strings */
-        if (pidstr[1] != 'd' || strchr(pidstr + 2, '%')) {
-            error_setg(errp, "Bad logfile format: %s", filename);
-            return;
-        } else {
-            logfilename = g_strdup_printf(filename, getpid());
-        }
-    } else {
-        logfilename = g_strdup(filename);
+    if (filename) {
+            char *pidstr = strstr(filename, "%");
+            if (pidstr) {
+                /* We only accept one %d, no other format strings */
+                if (pidstr[1] != 'd' || strchr(pidstr + 2, '%')) {
+                    error_setg(errp, "Bad logfile format: %s", filename);
+                    return;
+                } else {
+                    logfilename = g_strdup_printf(filename, getpid());
+                }
+            } else {
+                logfilename = g_strdup(filename);
+            }
     }
+
     qemu_log_close();
     qemu_set_log(qemu_loglevel);
 }
diff --git a/vl.c b/vl.c
index 24951b51a9..7dcb0879c4 100644
--- a/vl.c
+++ b/vl.c
@@ -3903,10 +3903,7 @@ int main(int argc, char **argv, char **envp)
 
     /* Open the logfile at this point and set the log mask if necessary.
      */
-    if (log_file) {
-        qemu_set_log_filename(log_file, &error_fatal);
-    }
-
+    qemu_set_log_filename(log_file, &error_fatal);
     if (log_mask) {
         int mask;
         mask = qemu_str_to_log_mask(log_mask);
-- 
2.24.1

Re: [PULL 0/5] Tracing patches

[Peter Maydell]

On Thu, 30 Jan 2020 at 21:38, Stefan Hajnoczi <stefanha@redhat.com> wrote:
>
> The following changes since commit 928173659d6e5dc368284f73f90ea1d129e1f57d:
>
>   Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200130' into staging (2020-01-30 16:19:04 +0000)
>
> are available in the Git repository at:
>
>   https://github.com/stefanha/qemu.git tags/tracing-pull-request
>
> for you to fetch changes up to e144a605a614d22165000c69e8e1dc6986d45cd8:
>
>   qemu_set_log_filename: filename argument may be NULL (2020-01-30 21:33:50 +0000)
>
> ----------------------------------------------------------------
> Pull request
>
> ----------------------------------------------------------------
>
> Peter Maydell (3):
>   docs/devel/tracing.txt: Recommend only
>     trace_event_get_state_backends()
>   memory.c: Use trace_event_get_state_backends()
>   hw/display/qxl.c: Use trace_event_get_state_backends()
>
> Philippe Mathieu-Daudé (1):
>   Makefile: Keep trace-events-subdirs ordered
>
> Salvador Fandino (1):
>   qemu_set_log_filename: filename argument may be NULL

Here's a weird one -- with this pullreq applied I
see a new warning running check-tcg on the linux-user
static config build:

  TEST    linux-test on aarch64
  TEST    testthread on aarch64
  TEST    float_madds on aarch64
  DIFF    float_madds.out with
/home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/float_ma
dds.ref
  TEST    fcvt on aarch64
warning: TCG temporary leaks before 0000000000400a0c
  DIFF    fcvt.out with
/home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/fcvt.ref
  TEST    pauth-1 on aarch64
  TEST    pauth-2 on aarch64

but I'm not sure why any of the patches here would have
provoked that, unless they're now causing a log message
that would previously have been suppressed or directed
somewhere else to be emitted.

thanks
-- PMM

Re: [PULL 0/5] Tracing patches

[Philippe Mathieu-Daudé]

On 1/31/20 2:57 PM, Peter Maydell wrote:
> On Thu, 30 Jan 2020 at 21:38, Stefan Hajnoczi <stefanha@redhat.com> wrote:
>>
>> The following changes since commit 928173659d6e5dc368284f73f90ea1d129e1f57d:
>>
>>    Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200130' into staging (2020-01-30 16:19:04 +0000)
>>
>> are available in the Git repository at:
>>
>>    https://github.com/stefanha/qemu.git tags/tracing-pull-request
>>
>> for you to fetch changes up to e144a605a614d22165000c69e8e1dc6986d45cd8:
>>
>>    qemu_set_log_filename: filename argument may be NULL (2020-01-30 21:33:50 +0000)
>>
>> ----------------------------------------------------------------
>> Pull request
>>
>> ----------------------------------------------------------------
>>
>> Peter Maydell (3):
>>    docs/devel/tracing.txt: Recommend only
>>      trace_event_get_state_backends()
>>    memory.c: Use trace_event_get_state_backends()
>>    hw/display/qxl.c: Use trace_event_get_state_backends()
>>
>> Philippe Mathieu-Daudé (1):
>>    Makefile: Keep trace-events-subdirs ordered
>>
>> Salvador Fandino (1):
>>    qemu_set_log_filename: filename argument may be NULL
> 
> Here's a weird one -- with this pullreq applied I
> see a new warning running check-tcg on the linux-user
> static config build:
> 
>    TEST    linux-test on aarch64
>    TEST    testthread on aarch64
>    TEST    float_madds on aarch64
>    DIFF    float_madds.out with
> /home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/float_ma
> dds.ref
>    TEST    fcvt on aarch64
> warning: TCG temporary leaks before 0000000000400a0c
>    DIFF    fcvt.out with
> /home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/fcvt.ref
>    TEST    pauth-1 on aarch64
>    TEST    pauth-2 on aarch64
> 
> but I'm not sure why any of the patches here would have
> provoked that, unless they're now causing a log message
> that would previously have been suppressed or directed
> somewhere else to be emitted.

Looking at the last patch "qemu_set_log_filename: filename argument may 
be NULL", maybe these were previously logged into a ./'(null)' logfile?

Re: [PULL 0/5] Tracing patches

[Philippe Mathieu-Daudé]

On 1/31/20 3:38 PM, Philippe Mathieu-Daudé wrote:
> On 1/31/20 2:57 PM, Peter Maydell wrote:
>> On Thu, 30 Jan 2020 at 21:38, Stefan Hajnoczi <stefanha@redhat.com> 
>> wrote:
>>>
>>> The following changes since commit 
>>> 928173659d6e5dc368284f73f90ea1d129e1f57d:
>>>
>>>    Merge remote-tracking branch 
>>> 'remotes/pmaydell/tags/pull-target-arm-20200130' into staging 
>>> (2020-01-30 16:19:04 +0000)
>>>
>>> are available in the Git repository at:
>>>
>>>    https://github.com/stefanha/qemu.git tags/tracing-pull-request
>>>
>>> for you to fetch changes up to e144a605a614d22165000c69e8e1dc6986d45cd8:
>>>
>>>    qemu_set_log_filename: filename argument may be NULL (2020-01-30 
>>> 21:33:50 +0000)
>>>
>>> ----------------------------------------------------------------
>>> Pull request
>>>
>>> ----------------------------------------------------------------
>>>
>>> Peter Maydell (3):
>>>    docs/devel/tracing.txt: Recommend only
>>>      trace_event_get_state_backends()
>>>    memory.c: Use trace_event_get_state_backends()
>>>    hw/display/qxl.c: Use trace_event_get_state_backends()
>>>
>>> Philippe Mathieu-Daudé (1):
>>>    Makefile: Keep trace-events-subdirs ordered
>>>
>>> Salvador Fandino (1):
>>>    qemu_set_log_filename: filename argument may be NULL
>>
>> Here's a weird one -- with this pullreq applied I
>> see a new warning running check-tcg on the linux-user
>> static config build:
>>
>>    TEST    linux-test on aarch64
>>    TEST    testthread on aarch64
>>    TEST    float_madds on aarch64
>>    DIFF    float_madds.out with
>> /home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/float_ma
>> dds.ref
>>    TEST    fcvt on aarch64
>> warning: TCG temporary leaks before 0000000000400a0c
>>    DIFF    fcvt.out with
>> /home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/fcvt.ref
>>    TEST    pauth-1 on aarch64
>>    TEST    pauth-2 on aarch64
>>
>> but I'm not sure why any of the patches here would have
>> provoked that, unless they're now causing a log message
>> that would previously have been suppressed or directed
>> somewhere else to be emitted.
> 
> Looking at the last patch "qemu_set_log_filename: filename argument may 
> be NULL", maybe these were previously logged into a ./'(null)' logfile?

Forget that comment, I need some caffeine.

Re: [PULL 0/5] Tracing patches

[Alex Bennée]

Peter Maydell <peter.maydell@linaro.org> writes:

> On Thu, 30 Jan 2020 at 21:38, Stefan Hajnoczi <stefanha@redhat.com> wrote:
>>
>> The following changes since commit 928173659d6e5dc368284f73f90ea1d129e1f57d:
>>
>>   Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200130' into staging (2020-01-30 16:19:04 +0000)
>>
>> are available in the Git repository at:
>>
>>   https://github.com/stefanha/qemu.git tags/tracing-pull-request
>>
>> for you to fetch changes up to e144a605a614d22165000c69e8e1dc6986d45cd8:
>>
>>   qemu_set_log_filename: filename argument may be NULL (2020-01-30 21:33:50 +0000)
>>
>> ----------------------------------------------------------------
>> Pull request
>>
>> ----------------------------------------------------------------
>>
>> Peter Maydell (3):
>>   docs/devel/tracing.txt: Recommend only
>>     trace_event_get_state_backends()
>>   memory.c: Use trace_event_get_state_backends()
>>   hw/display/qxl.c: Use trace_event_get_state_backends()
>>
>> Philippe Mathieu-Daudé (1):
>>   Makefile: Keep trace-events-subdirs ordered
>>
>> Salvador Fandino (1):
>>   qemu_set_log_filename: filename argument may be NULL
>
> Here's a weird one -- with this pullreq applied I
> see a new warning running check-tcg on the linux-user
> static config build:
>
>   TEST    linux-test on aarch64
>   TEST    testthread on aarch64
>   TEST    float_madds on aarch64
>   DIFF    float_madds.out with
> /home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/float_ma
> dds.ref
>   TEST    fcvt on aarch64
> warning: TCG temporary leaks before 0000000000400a0c
>   DIFF    fcvt.out with
> /home/petmay01/linaro/qemu-for-merges/tests/tcg/aarch64/fcvt.ref
>   TEST    pauth-1 on aarch64
>   TEST    pauth-2 on aarch64
>
> but I'm not sure why any of the patches here would have
> provoked that, unless they're now causing a log message
> that would previously have been suppressed or directed
> somewhere else to be emitted.

That must be it - before hand we wouldn't have called
qemu_set_log_filename which in turn doesn't call qemu_set_log.

Anyway I have replicated the warning.

>
> thanks
> -- PMM


-- 
Alex Bennée

Re: [PULL 0/5] Tracing patches

[Peter Maydell]

On Thu, 30 Jan 2020 at 21:38, Stefan Hajnoczi <stefanha@redhat.com> wrote:
>
> The following changes since commit 928173659d6e5dc368284f73f90ea1d129e1f57d:
>
>   Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200130' into staging (2020-01-30 16:19:04 +0000)
>
> are available in the Git repository at:
>
>   https://github.com/stefanha/qemu.git tags/tracing-pull-request
>
> for you to fetch changes up to e144a605a614d22165000c69e8e1dc6986d45cd8:
>
>   qemu_set_log_filename: filename argument may be NULL (2020-01-30 21:33:50 +0000)
>
> ----------------------------------------------------------------
> Pull request
>
> ----------------------------------------------------------------
>


Applied, thanks.

Please update the changelog at https://wiki.qemu.org/ChangeLog/5.0
for any user-visible changes.

-- PMM

Re: [PULL 5/5] qemu_set_log_filename: filename argument may be NULL

[Richard Henderson]

On 1/30/20 1:38 PM, Stefan Hajnoczi wrote:
> From: Salvador Fandino <salvador@qindel.com>
> 
> NULL is a valid log filename used to indicate we want to use stderr
> but qemu_set_log_filename (which is called by bsd-user/main.c) was not
> handling it correctly.
> 
> That also made redundant a couple of NULL checks in calling code which
> have been removed.
> 
> Signed-off-by: Salvador Fandino <salvador@qindel.com>
> Message-Id: <20200123193626.19956-1-salvador@qindel.com>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
>  trace/control.c |  4 +---
>  util/log.c      | 28 ++++++++++++++++------------
>  vl.c            |  5 +----
>  3 files changed, 18 insertions(+), 19 deletions(-)

This patch has broken -D <filename> for *-linux-user.
After e144a605a, all logging goes to stderr.

> +    if (filename) {
> +            char *pidstr = strstr(filename, "%");
> +            if (pidstr) {

Also, the indentation is off.


r~

Re: [PULL 5/5] qemu_set_log_filename: filename argument may be NULL

[Alex Bennée]

Richard Henderson <richard.henderson@linaro.org> writes:

> On 1/30/20 1:38 PM, Stefan Hajnoczi wrote:
>> From: Salvador Fandino <salvador@qindel.com>
>> 
>> NULL is a valid log filename used to indicate we want to use stderr
>> but qemu_set_log_filename (which is called by bsd-user/main.c) was not
>> handling it correctly.
>> 
>> That also made redundant a couple of NULL checks in calling code which
>> have been removed.
>> 
>> Signed-off-by: Salvador Fandino <salvador@qindel.com>
>> Message-Id: <20200123193626.19956-1-salvador@qindel.com>
>> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
>> ---
>>  trace/control.c |  4 +---
>>  util/log.c      | 28 ++++++++++++++++------------
>>  vl.c            |  5 +----
>>  3 files changed, 18 insertions(+), 19 deletions(-)
>
> This patch has broken -D <filename> for *-linux-user.
> After e144a605a, all logging goes to stderr.

I posted:

  Subject: [PATCH] tracing: only allow -trace to override -D if set
  Date: Tue, 11 Feb 2020 11:10:54 +0000
  Message-Id: <20200211111054.27538-1-alex.bennee@linaro.org>

as a fix which partially reverted this.

>
>> +    if (filename) {
>> +            char *pidstr = strstr(filename, "%");
>> +            if (pidstr) {
>
> Also, the indentation is off.
>
>
> r~


-- 
Alex Bennée

Re: [PULL 5/5] qemu_set_log_filename: filename argument may be NULL

[Richard Henderson]

On 2/12/20 2:45 AM, Alex Bennée wrote:
> 
> Richard Henderson <richard.henderson@linaro.org> writes:
> 
>> On 1/30/20 1:38 PM, Stefan Hajnoczi wrote:
>>> From: Salvador Fandino <salvador@qindel.com>
>>>
>>> NULL is a valid log filename used to indicate we want to use stderr
>>> but qemu_set_log_filename (which is called by bsd-user/main.c) was not
>>> handling it correctly.
>>>
>>> That also made redundant a couple of NULL checks in calling code which
>>> have been removed.
>>>
>>> Signed-off-by: Salvador Fandino <salvador@qindel.com>
>>> Message-Id: <20200123193626.19956-1-salvador@qindel.com>
>>> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
>>> ---
>>>  trace/control.c |  4 +---
>>>  util/log.c      | 28 ++++++++++++++++------------
>>>  vl.c            |  5 +----
>>>  3 files changed, 18 insertions(+), 19 deletions(-)
>>
>> This patch has broken -D <filename> for *-linux-user.
>> After e144a605a, all logging goes to stderr.
> 
> I posted:
> 
>   Subject: [PATCH] tracing: only allow -trace to override -D if set
>   Date: Tue, 11 Feb 2020 11:10:54 +0000
>   Message-Id: <20200211111054.27538-1-alex.bennee@linaro.org>
> 
> as a fix which partially reverted this.

Thanks, that fixes it for me.


r~