Skip to contentSkip to conversation

hw: Remove dynamic field width from trace events

PATCH Superseded Philippe Mathieu-Daudé 3 participants 10 messages Nov 18, 2019 – Nov 19, 2019

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

v3 cover hw: Remove dynamic field width from trace events

Eric noted in [1] the dtrace via stap backend can not support the dynamic '*' width format.
I'd really like to use dynamic width in trace event because the read/write accesses are easier to read but it is not a priority. Since next release is close, time to fix LP#1844817 [2].

Since v2:
- addressed Eric review comments from v2
- improved the documentation

Since v1:
- Do not update the qemu_log_mask() calls in hw/mips/gt64xxx_pci.c

[1] https://lists.gnu.org/archive/html/qemu-devel/2019-09/msg04720.html
[2] https://bugs.launchpad.net/qemu/+bug/1844817

Philippe Mathieu-Daudé (3):
hw/block/pflash: Remove dynamic field width from trace events hw/mips/gt64xxx: Remove dynamic field width from trace events
trace: Forbid dynamic field width in event format

 docs/devel/tracing.txt        |  3 ++-
 hw/block/pflash_cfi01.c       |  8 ++++----
 hw/block/pflash_cfi02.c       |  8 ++++----
 hw/mips/gt64xxx_pci.c         | 16 ++++++++--------
 hw/block/trace-events         |  8 ++++----
 hw/mips/trace-events          |  4 ++--
 scripts/tracetool/__init__.py |  3 +++
 7 files changed, 27 insertions(+), 23 deletions(-)
Signature · 1 line
2.21.0
#1 lore raw

PATCH v3 1/3 hw/block/pflash: Remove dynamic field width from trace events

Since not all trace backends support dynamic field width in format (dtrace via stap does not), replace by a static field width instead.

We previously passed to the trace API 'width << 1' as the number of hex characters to display (the dynamic field width). We don't need this anymore. Instead, display the size of bytes accessed.

---
v3: display size (in byte)

 hw/block/pflash_cfi01.c | 8 ++++----
 hw/block/pflash_cfi02.c | 8 ++++----
 hw/block/trace-events   | 8 ++++----
 3 files changed, 12 insertions(+), 12 deletions(-)
hw/block/pflash_cfi01.c +4 -4
index 566c0acb77..54e6ebd385 100644
@@ -276,7 +276,7 @@ static uint32_t pflash_data_read(PFlashCFI01 *pfl, hwaddr offset,
         DPRINTF("BUG in %s\n", __func__);
         abort();
     }
-    trace_pflash_data_read(offset, width << 1, ret);
+    trace_pflash_data_read(offset, width, ret);
     return ret;
 }
 
@@ -389,7 +389,7 @@ static uint32_t pflash_read(PFlashCFI01 *pfl, hwaddr offset,
 
         break;
     }
-    trace_pflash_io_read(offset, width, width << 1, ret, pfl->cmd, pfl->wcycle);
+    trace_pflash_io_read(offset, width, ret, pfl->cmd, pfl->wcycle);
 
     return ret;
 }
@@ -414,7 +414,7 @@ static inline void pflash_data_write(PFlashCFI01 *pfl, hwaddr offset,
 {
     uint8_t *p = pfl->storage;
 
-    trace_pflash_data_write(offset, width << 1, value, pfl->counter);
+    trace_pflash_data_write(offset, width, value, pfl->counter);
     switch (width) {
     case 1:
         p[offset] = value;
@@ -453,7 +453,7 @@ static void pflash_write(PFlashCFI01 *pfl, hwaddr offset,
 
     cmd = value;
 
-    trace_pflash_io_write(offset, width, width << 1, value, pfl->wcycle);
+    trace_pflash_io_write(offset, width, value, pfl->wcycle);
     if (!pfl->wcycle) {
         /* Set the device in I/O access mode */
         memory_region_rom_device_set_romd(&pfl->mem, false);
hw/block/pflash_cfi02.c +4 -4
index 4baca701b7..c7d92c3e79 100644
@@ -260,7 +260,7 @@ static uint64_t pflash_data_read(PFlashCFI02 *pfl, hwaddr offset,
 {
     uint8_t *p = (uint8_t *)pfl->storage + offset;
     uint64_t ret = pfl->be ? ldn_be_p(p, width) : ldn_le_p(p, width);
-    trace_pflash_data_read(offset, width << 1, ret);
+    trace_pflash_data_read(offset, width, ret);
     return ret;
 }
 
@@ -385,7 +385,7 @@ static uint64_t pflash_read(void *opaque, hwaddr offset, unsigned int width)
         }
         break;
     }
-    trace_pflash_io_read(offset, width, width << 1, ret, pfl->cmd, pfl->wcycle);
+    trace_pflash_io_read(offset, width, ret, pfl->cmd, pfl->wcycle);
 
     return ret;
 }
@@ -432,7 +432,7 @@ static void pflash_write(void *opaque, hwaddr offset, uint64_t value,
     uint8_t *p;
     uint8_t cmd;
 
-    trace_pflash_io_write(offset, width, width << 1, value, pfl->wcycle);
+    trace_pflash_io_write(offset, width, value, pfl->wcycle);
     cmd = value;
     if (pfl->cmd != 0xA0) {
         /* Reset does nothing during chip erase and sector erase. */
@@ -542,7 +542,7 @@ static void pflash_write(void *opaque, hwaddr offset, uint64_t value,
                 }
                 goto reset_flash;
             }
-            trace_pflash_data_write(offset, width << 1, value, 0);
+            trace_pflash_data_write(offset, width, value, 0);
             if (!pfl->ro) {
                 p = (uint8_t *)pfl->storage + offset;
                 if (pfl->be) {
hw/block/trace-events +4 -4
index 13d1b21dd4..c03e80c2c9 100644
@@ -8,10 +8,10 @@ fdc_ioport_write(uint8_t reg, uint8_t value) "write reg 0x%02x val 0x%02x"
 # pflash_cfi01.c
 pflash_reset(void) "reset"
 pflash_timer_expired(uint8_t cmd) "command 0x%02x done"
-pflash_io_read(uint64_t offset, int width, int fmt_width, uint32_t value, uint8_t cmd, uint8_t wcycle) "offset:0x%04"PRIx64" width:%d value:0x%0*x cmd:0x%02x wcycle:%u"
-pflash_io_write(uint64_t offset, int width, int fmt_width, uint32_t value, uint8_t wcycle) "offset:0x%04"PRIx64" width:%d value:0x%0*x wcycle:%u"
-pflash_data_read(uint64_t offset, int width, uint32_t value) "data offset:0x%04"PRIx64" value:0x%0*x"
-pflash_data_write(uint64_t offset, int width, uint32_t value, uint64_t counter) "data offset:0x%04"PRIx64" value:0x%0*x counter:0x%016"PRIx64
+pflash_io_read(uint64_t offset, unsigned size, uint32_t value, uint8_t cmd, uint8_t wcycle) "offset:0x%04"PRIx64" size:%u value:0x%04x cmd:0x%02x wcycle:%u"
+pflash_io_write(uint64_t offset, unsigned size, uint32_t value, uint8_t wcycle) "offset:0x%04"PRIx64" size:%u value:0x%04x wcycle:%u"
+pflash_data_read(uint64_t offset, unsigned size, uint32_t value) "data offset:0x%04"PRIx64" size:%u value:0x%04x"
+pflash_data_write(uint64_t offset, unsigned size, uint32_t value, uint64_t counter) "data offset:0x%04"PRIx64" size:%u value:0x%04x counter:0x%016"PRIx64
 pflash_manufacturer_id(uint16_t id) "Read Manufacturer ID: 0x%04x"
 pflash_device_id(uint16_t id) "Read Device ID: 0x%04x"
 pflash_device_info(uint64_t offset) "Read Device Information offset:0x%04"PRIx64
Signature · 1 line
2.21.0
Reviewed-by on v3 1/3
#2 lore raw

Re: [PATCH-for-4.2 v3 1/3] hw/block/pflash: Remove dynamic field width from trace events

Quoted 11 lines · Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), replace by a static field width instead.

We previously passed to the trace API 'width << 1' as the number of hex characters to display (the dynamic field width). We don't need this anymore. Instead, display the size of bytes accessed.

Is it worth a Fixes: XXX calling out the commit that introduced the '*'?

/me goes and searches

Fixes e8aa2d95ea
Fixes c1474acd5d
Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org
#3 lore raw

Re: [PATCH-for-4.2 v3 1/3] hw/block/pflash: Remove dynamic field width from trace events

Quoted 9 lines · Eric Blake
Quoted 8 lines · Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), replace by a static field width instead.

We previously passed to the trace API 'width << 1' as the number of hex characters to display (the dynamic field width). We don't need this anymore. Instead, display the size of bytes accessed.

Quoted 5 lines
Quoted 3 lines

Is it worth a Fixes: XXX calling out the commit that introduced the '*'?

Yes, good idea.

Quoted 5 lines

/me goes and searches

Fixes e8aa2d95ea
Fixes c1474acd5d

Similarly the next patch (hw/mips/gt64xxx_pci.c) fixes ab6bff424f4.

Quoted 1 line

Thanks!

#4 lore raw

PATCH v3 2/3 hw/mips/gt64xxx: Remove dynamic field width from trace events

In reply to Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), replace by a static field width instead.

We previously passed to the trace API 'width << 1' as the number of hex characters to display (the dynamic field width). We don't need this anymore. Instead, display the size of bytes accessed.

---
v2: Do not update qemu_log_mask()
v3: display size (in byte)

---
 hw/mips/gt64xxx_pci.c | 16 ++++++++--------
 hw/mips/trace-events  |  4 ++--
 2 files changed, 10 insertions(+), 10 deletions(-)
hw/mips/gt64xxx_pci.c +8 -8
index 5cab9c1ee1..f1af840d8e 100644
@@ -642,19 +642,19 @@ static void gt64120_writel(void *opaque, hwaddr addr,
         /* not really implemented */
         s->regs[saddr] = ~(~(s->regs[saddr]) | ~(val & 0xfffffffe));
         s->regs[saddr] |= !!(s->regs[saddr] & 0xfffffffe);
-        trace_gt64120_write("INTRCAUSE", size << 1, val);
+        trace_gt64120_write("INTRCAUSE", size, val);
         break;
     case GT_INTRMASK:
         s->regs[saddr] = val & 0x3c3ffffe;
-        trace_gt64120_write("INTRMASK", size << 1, val);
+        trace_gt64120_write("INTRMASK", size, val);
         break;
     case GT_PCI0_ICMASK:
         s->regs[saddr] = val & 0x03fffffe;
-        trace_gt64120_write("ICMASK", size << 1, val);
+        trace_gt64120_write("ICMASK", size, val);
         break;
     case GT_PCI0_SERR0MASK:
         s->regs[saddr] = val & 0x0000003f;
-        trace_gt64120_write("SERR0MASK", size << 1, val);
+        trace_gt64120_write("SERR0MASK", size, val);
         break;
 
     /* Reserved when only PCI_0 is configured. */
@@ -930,19 +930,19 @@ static uint64_t gt64120_readl(void *opaque,
     /* Interrupts */
     case GT_INTRCAUSE:
         val = s->regs[saddr];
-        trace_gt64120_read("INTRCAUSE", size << 1, val);
+        trace_gt64120_read("INTRCAUSE", size, val);
         break;
     case GT_INTRMASK:
         val = s->regs[saddr];
-        trace_gt64120_read("INTRMASK", size << 1, val);
+        trace_gt64120_read("INTRMASK", size, val);
         break;
     case GT_PCI0_ICMASK:
         val = s->regs[saddr];
-        trace_gt64120_read("ICMASK", size << 1, val);
+        trace_gt64120_read("ICMASK", size, val);
         break;
     case GT_PCI0_SERR0MASK:
         val = s->regs[saddr];
-        trace_gt64120_read("SERR0MASK", size << 1, val);
+        trace_gt64120_read("SERR0MASK", size, val);
         break;
 
     /* Reserved when only PCI_0 is configured. */
hw/mips/trace-events +2 -2
index 75d4c73f2e..321933283f 100644
@@ -1,4 +1,4 @@
 # gt64xxx.c
-gt64120_read(const char *regname, int width, uint64_t value) "gt64120 read %s value:0x%0*" PRIx64
-gt64120_write(const char *regname, int width, uint64_t value) "gt64120 write %s value:0x%0*" PRIx64
+gt64120_read(const char *regname, unsigned size, uint64_t value) "gt64120 read %s size:%u value:0x%08" PRIx64
+gt64120_write(const char *regname, unsigned size, uint64_t value) "gt64120 write %s size:%u value:0x%08" PRIx64
 gt64120_isd_remap(uint64_t from_length, uint64_t from_addr, uint64_t to_length, uint64_t to_addr) "ISD: 0x%08" PRIx64 "@0x%08" PRIx64 " -> 0x%08" PRIx64 "@0x%08" PRIx64
Signature · 1 line
2.21.0
Reviewed-by on v3 2/3
#5 lore raw

Re: [PATCH-for-4.2 v3 2/3] hw/mips/gt64xxx: Remove dynamic field width from trace events

Quoted 11 lines · Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), replace by a static field width instead.

We previously passed to the trace API 'width << 1' as the number of hex characters to display (the dynamic field width). We don't need this anymore. Instead, display the size of bytes accessed.

Fixes ab6bff424f
Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org
#6 lore raw

PATCH v3 3/3 trace: Forbid dynamic field width in event format

In reply to Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), forbid them.

Add a check to refuse field width in new formats:

  $ make
  [...]
    GEN     hw/block/trace.h
  Traceback (most recent call last):
    File "scripts/tracetool.py", line 152, in <module>
      main(sys.argv)
    File "scripts/tracetool.py", line 143, in main
      events.extend(tracetool.read_events(fh, arg))
    File "scripts/tracetool/__init__.py", line 371, in read_events
      event = Event.build(line)
    File "scripts/tracetool/__init__.py", line 285, in build
      raise ValueError("Event format must not contain field width '%*'")
  ValueError: Error at hw/block/trace-events:11: Event format must not contain field width '%*'

---
v3:
- use better regex provided by Eric,
- instead of re.match(), use re.search() which takes unanchored regex,
- added a comment in tracing.txt

---
 docs/devel/tracing.txt        | 3 ++-
 scripts/tracetool/__init__.py | 3 +++
 2 files changed, 5 insertions(+), 1 deletion(-)
docs/devel/tracing.txt +2 -1
index 8c0376fefa..6c01ce801e 100644
@@ -113,7 +113,8 @@ Format strings should reflect the types defined in the trace event.  Take
 special care to use PRId64 and PRIu64 for int64_t and uint64_t types,
 respectively.  This ensures portability between 32- and 64-bit platforms.
 Format strings must not end with a newline character.  It is the responsibility
-of backends to adapt line ending for proper logging.
+of backends to adapt line ending for proper logging.  Format strings must not
+use numeric field width dynamic precision (SystemTap does not support them).
 
 Each event declaration will start with the event name, then its arguments,
 finally a format string for pretty-printing. For example:
scripts/tracetool/__init__.py +3 -0
index 44c118bc2a..ec7fe9fa4a 100644
@@ -206,6 +206,7 @@ class Event(object):
                       "\s*"
                       "(?:(?:(?P<fmt_trans>\".+),)?\s*(?P<fmt>\".+))?"
                       "\s*")
+    _DFWRE = re.compile(r"%[\d\.\- +#']*\*") # dynamic width precision
 
     _VALID_PROPS = set(["disable", "tcg", "tcg-trans", "tcg-exec", "vcpu"])
 
@@ -280,6 +281,8 @@ class Event(object):
         if fmt.endswith(r'\n"'):
             raise ValueError("Event format must not end with a newline "
                              "character")
+        if Event._DFWRE.search(fmt):
+            raise ValueError("Event format must not contain field width '%*'")
 
         if len(fmt_trans) > 0:
             fmt = [fmt_trans, fmt]
Signature · 1 line
2.21.0
#7 lore raw

Re: [PATCH-for-4.2 v3 3/3] trace: Forbid dynamic field width in event format

Quoted 26 lines · Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), forbid them.

Add a check to refuse field width in new formats:

   $ make
   [...]
     GEN     hw/block/trace.h
   Traceback (most recent call last):
     File "scripts/tracetool.py", line 152, in <module>
       main(sys.argv)
     File "scripts/tracetool.py", line 143, in main
       events.extend(tracetool.read_events(fh, arg))
     File "scripts/tracetool/__init__.py", line 371, in read_events
       event = Event.build(line)
     File "scripts/tracetool/__init__.py", line 285, in build
       raise ValueError("Event format must not contain field width '%*'")
   ValueError: Error at hw/block/trace-events:11: Event format must not contain field width '%*'

---

+++ b/docs/devel/tracing.txt

docs/devel/tracing.txt Format strings should reflect the types defined in the trace event. Take · line 117

Quoted 6 lines
docs/devel/tracing.txt +2 -1
@@ -113,7 +113,8 @@ Format strings should reflect the types defined in the trace event.  Take
  special care to use PRId64 and PRIu64 for int64_t and uint64_t types,
  respectively.  This ensures portability between 32- and 64-bit platforms.
  Format strings must not end with a newline character.  It is the responsibility
-of backends to adapt line ending for proper logging.
+of backends to adapt line ending for proper logging.  Format strings must not
+use numeric field width dynamic precision (SystemTap does not support them).

Reads awkwardly - a dynamic precision is not numeric in the format string (but '*' instead). Better might be:

Format strings must not use dynamic field width or precision ('*'), as at least SystemTap does not support them.

Or even:

Format strings may use numeric field width or precision, but must not use dynamic forms ('*') as at least SystemTap does not support that.

scripts/tracetool/__init__.py class Event(object): · line 209

Quoted 12 lines
docs/devel/tracing.txt +0 -0
  Each event declaration will start with the event name, then its arguments,
  finally a format string for pretty-printing. For example:
scripts/tracetool/__init__.py +1 -0
index 44c118bc2a..ec7fe9fa4a 100644
@@ -206,6 +206,7 @@ class Event(object):
                        "\s*"
                        "(?:(?:(?P<fmt_trans>\".+),)?\s*(?P<fmt>\".+))?"
                        "\s*")
+    _DFWRE = re.compile(r"%[\d\.\- +#']*\*") # dynamic width precision

The comment is slightly off - this catches both dynamic field width (any '*' before '.') and dynamic precision (any '*' after '.'), maybe the fix is just s/width/width or/

scripts/tracetool/__init__.py class Event(object): · line 285

Quoted 9 lines
scripts/tracetool/__init__.py +2 -0
      _VALID_PROPS = set(["disable", "tcg", "tcg-trans", "tcg-exec", "vcpu"])
 
@@ -280,6 +281,8 @@ class Event(object):
          if fmt.endswith(r'\n"'):
              raise ValueError("Event format must not end with a newline "
                               "character")
+        if Event._DFWRE.search(fmt):
+            raise ValueError("Event format must not contain field width '%*'")

and I don't know if you want to tweak the error message, maybe:
Event format must not use dynamic '*'

If we're trying to get stuff in 4.2-rc2, patch 1 and 2 are actual bug fixes and deserve to go in; patch 3 is nice-to-have but doesn't affect the build if it is omitted (as there are no other offenders left), so slipping it into 5.0 for a v4 to clean it up slightly doesn't hurt. I don't know who would send the pull request, though, and slipping 1 and 2 into -rc3 just because of 3 is not ideal.

Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org
#8 lore raw

Re: [PATCH-for-4.2 v3 3/3] trace: Forbid dynamic field width in event format

Quoted 71 lines · Eric Blake
Quoted 32 lines · Philippe Mathieu-Daudé

Since not all trace backends support dynamic field width in format (dtrace via stap does not), forbid them.

Add a check to refuse field width in new formats:

   $ make
   [...]
     GEN     hw/block/trace.h
   Traceback (most recent call last):
     File "scripts/tracetool.py", line 152, in <module>
       main(sys.argv)
     File "scripts/tracetool.py", line 143, in main
       events.extend(tracetool.read_events(fh, arg))
     File "scripts/tracetool/__init__.py", line 371, in read_events
       event = Event.build(line)
     File "scripts/tracetool/__init__.py", line 285, in build
       raise ValueError("Event format must not contain field width '%*'")
   ValueError: Error at hw/block/trace-events:11: Event format must not contain field width '%*'

---

+++ b/docs/devel/tracing.txt

(quoted diff) +0 -0
@@ -113,7 +113,8 @@ Format strings should reflect the types defined in 

the trace event.  Take
  special care to use PRId64 and PRIu64 for int64_t and uint64_t types, respectively.  This ensures portability between 32- and 64-bit platforms. Format strings must not end with a newline character.  It is the responsibility

(quoted diff) +1 -1
-of backends to adapt line ending for proper logging.
+of backends to adapt line ending for proper logging.  Format strings 
Quoted 3 lines

must not
+use numeric field width dynamic precision (SystemTap does not support them).

Reads awkwardly - a dynamic precision is not numeric in the format string (but '*' instead).  Better might be:

Format strings must not use dynamic field width or precision ('*'), as at least SystemTap does not support them.

Or even:

Format strings may use numeric field width or precision, but must not use dynamic forms ('*') as at least SystemTap does not support that.

Quoted 11 lines

  Each event declaration will start with the event name, then its arguments, finally a format string for pretty-printing. For example:

scripts/tracetool/__init__.py +0 -0
index 44c118bc2a..ec7fe9fa4a 100644
@@ -206,6 +206,7 @@ class Event(object):

                        "\s*" "(?:(?:(?P<fmt_trans>\".+),)?\s*(?P<fmt>\".+))?" "\s*")
+    _DFWRE = re.compile(r"%[\d\.\- +#']*\*") # dynamic width precision

The comment is slightly off - this catches both dynamic field width (any '*' before '.') and dynamic precision (any '*' after '.'), maybe the fix is just s/width/width or/

Quoted 8 lines

      _VALID_PROPS = set(["disable", "tcg", "tcg-trans", "tcg-exec", "vcpu"])

scripts/tracetool/__init__.py +0 -0
@@ -280,6 +281,8 @@ class Event(object):

          if fmt.endswith(r'\n"'):
              raise ValueError("Event format must not end with a newline " "character")
+        if Event._DFWRE.search(fmt):
+            raise ValueError("Event format must not contain field width '%*'")

and I don't know if you want to tweak the error message, maybe:
 Event format must not use dynamic '*'

Quoted 1 line

If we're trying to get stuff in 4.2-rc2, patch 1 and 2 are actual bug fixes and deserve to go in; patch 3 is nice-to-have but doesn't affect the build if it is omitted (as there are no other offenders left), so slipping it into 5.0 for a v4 to clean it up slightly doesn't hurt.  I don't know who would send the pull request, though, and slipping 1 and 2 into -rc3 just because of 3 is not ideal.

Wise thought. I'll respin 1-2 with the Fixes: tag in case qemu-block@ or qemu-trivial@ have pending pull-request for tomorrow, they can queue it. Else they can go via the mips-next tree, since both patch affect the Malta board.

Thanks!

Phil.