Skip to contentSkip to conversation

gcc 8.2 reports stringop-truncation when building qemu

Qemu-devel Superseded Amir Gonnen 10 participants 33 messages Nov 18, 2018 – Apr 24, 2019 Block layer

v3 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

Public bug reported:

QEMU 3.0

block/sheepdog.c: In function 'find_vdi_name':
block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]

     strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If this is the intended behavior, please suppress the warning. For example:

#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wstringop-truncation"

    strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);

#pragma GCC diagnostic pop

** Affects: qemu
Importance: Undecided
Status: New

Signature · 28 lines
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  New

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  
  If this is the intended behavior, please suppress the warning. For example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
#1 lore raw

** Description changed:

QEMU 3.0

block/sheepdog.c: In function 'find_vdi_name':
block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
- strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
- ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
+      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-
- If this is the intended behavior, please suppress the warning. For example:
+ If this is the intended behavior, please suppress the warning. For
+ example:

#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wstringop-truncation"
- strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
+     strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN); #pragma GCC diagnostic pop
+
+ This also happens on other sources, for example hw/acpi/core.c, so
+ another option is to suppress it globally on CFLAGS (-Wno-stringop-
+ truncation)

Signature · 32 lines
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  New

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  If this is the intended behavior, please suppress the warning. For
  example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

  This also happens on other sources, for example hw/acpi/core.c, so
  another option is to suppress it globally on CFLAGS (-Wno-stringop-
  truncation)

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
Reviewed-by Marc-André Lureau
#2 lore raw

[Qemu-devel] [Bug 1803872] Re: [PATCH v2 3/3] migration: Replace strncpy() by strpadcpy(pad='\0')

In reply to Amir Gonnen
Quoted 20 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_store_running':
qemu/migration/global_state.c:45:5: error: 'strncpy' specified bound 100 equals destination size [-Werror=stringop-truncation]

       strncpy((char *)global_state.runstate, state, sizeof(global_state.runstate));
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

The runstate name doesn't require the strings to be NUL-terminated, therefore strncpy is the right function to use here.

Quoted 6 lines

We could add a #pragma GCC diagnostic ignored "-Wstringop-truncation" around, disable the warning globally using -Wno-stringop-truncation, but since QEMU provides the strpadcpy() which does the same purpose, simply use it to avoid the annoying warning.

Quoted 23 lines
---
 migration/global_state.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
migration/global_state.c +2 -3
index 8e8ab5c51e..c7e7618118 100644
@@ -42,8 +42,8 @@ int global_state_store(void)
 void global_state_store_running(void)
 {
     const char *state = RunState_str(RUN_STATE_RUNNING);
-    strncpy((char *)global_state.runstate,
-           state, sizeof(global_state.runstate));
+    strpadcpy((char *)global_state.runstate,
+              sizeof(global_state.runstate), state, '\0');
 }
 
 bool global_state_received(void)
--

2.17.2

Signature · 35 lines
Marc-André Lureau

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  New

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  If this is the intended behavior, please suppress the warning. For
  example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

  This also happens on other sources, for example hw/acpi/core.c, so
  another option is to suppress it globally on CFLAGS (-Wno-stringop-
  truncation)

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
author
#3 lore raw

[Qemu-devel] [Bug 1803872] Re: [PATCH v3 5/5] migration: Use strnlen() for fixed-size string

In reply to Amir Gonnen
Quoted 6 lines · Philippe Mathieu-Daudé

GCC 8 introduced the -Wstringop-overflow, which detect buffer overflow by string-modifying functions declared in <string.h>, such strncpy(), used in global_state_store_running().

Since the global_state.runstate does not necessarily contains a terminating NUL character, We had to use the QEMU_NONSTRING attribute.

s/We/we/

Quoted 39 lines

The GCC manual says about the nonstring attribute:

However, when the array is declared with the attribute the call to strlen is diagnosed because when the array doesn’t contain a NUL-terminated string the call is undefined. [...]
In addition, calling strnlen and strndup with such arrays is safe provided a suitable bound is specified, and not diagnosed.

GCC indeed found an incorrect use of strlen(), because this array is loaded by VMSTATE_BUFFER(runstate, GlobalState) then parsed using qapi_enum_parse which does not get the buffer length.

Use strnlen() which returns sizeof(s->runstate) if the array is not NUL-terminated.

This fixes:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_pre_save':

   qemu/migration/global_state.c:109:15: error: 'strlen' argument 1 declared attribute 'nonstring' [-Werror=stringop-overflow=]
        s->size = strlen((char *)s->runstate) + 1;
                  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
   qemu/migration/global_state.c:24:13: note: argument 'runstate' declared here
        uint8_t runstate[100] QEMU_NONSTRING;
                ^~~~~~~~
   cc1: all warnings being treated as errors
   make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1
---
  migration/global_state.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
migration/global_state.c +0 -0
index 6e19333422..c19030ef62 100644

migration/global_state.c static int global_state_pre_save(void *opaque) · line 109

Quoted 4 lines
migration/global_state.c +0 -1
@@ -106,7 +106,7 @@ static int global_state_pre_save(void *opaque)
      GlobalState *s = opaque;
 
      trace_migrate_global_state_pre_save((char *)s->runstate);
-    s->size = strlen((char *)s->runstate) + 1;

The old code sets s->size to the string length + space for the NUL byte (by assuming that a NUL byte was present), and accidentally sets it beyond the s->runstate array if there was no NUL byte (our existing runstate names are shorter than 100 bytes, so this could only happen on a malicious stream).

migration/global_state.c static int global_state_pre_save(void *opaque) · line 109

Quoted 1 line
migration/global_state.c +1 -0
+    s->size = strnlen((char *)s->runstate, sizeof(s->runstate)) + 1;

The new code can still end up setting s->size beyond the array. Is that intended, or would it be better to use strnlen(s->runstate, sizeof(s->runstate) - 1) + 1?

Also, as I argued on 4/5, why isn't this squashed in with the patch that marks the field NONSTRING?

Signature · 37 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  New

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  If this is the intended behavior, please suppress the warning. For
  example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

  This also happens on other sources, for example hw/acpi/core.c, so
  another option is to suppress it globally on CFLAGS (-Wno-stringop-
  truncation)

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
#4 lore raw

Re: [Qemu-devel] [PATCH v3 5/5] migration: Use strnlen() for fixed-size string

Quoted 6 lines · Eric Blake
Quoted 6 lines
migration/global_state.c +0 -0
index 6e19333422..c19030ef62 100644
@@ -106,7 +106,7 @@ static int global_state_pre_save(void *opaque)
       GlobalState *s = opaque;
Quoted 8 lines

migration/global_state.c static int global_state_pre_save(void *opaque) · line 108

Quoted 2 lines
migration/global_state.c +0 -1
         trace_migrate_global_state_pre_save((char *)s->runstate);
-    s->size = strlen((char *)s->runstate) + 1;

The old code sets s->size to the string length + space for the NUL byte (by assuming that a NUL byte was present), and accidentally sets it beyond the s->runstate array if there was no NUL byte (our existing runstate names are shorter than 100 bytes, so this could only happen on a malicious stream).

It cannot---this is a pre_save hook. A possible overflow bug exists, but it is in the call to qapi_enum_parse.

Paolo

author
#5 lore raw

[Qemu-devel] [Bug 1803872] Re: [PATCH v3 4/5] migration: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Amir Gonnen
Quoted 16 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_store_running':

   qemu/migration/global_state.c:45:5: error: 'strncpy' specified bound 100 equals destination size [-Werror=stringop-truncation]
        strncpy((char *)global_state.runstate, state, sizeof(global_state.runstate));
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1
Quoted 8 lines

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
  migration/global_state.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

Should this be squashed with 5/5?

Signature · 37 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  New

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  If this is the intended behavior, please suppress the warning. For
  example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

  This also happens on other sources, for example hw/acpi/core.c, so
  another option is to suppress it globally on CFLAGS (-Wno-stringop-
  truncation)

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
author
#6 lore raw

[Qemu-devel] [Bug 1803872] Re: [PATCH v3 4/5] migration: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Amir Gonnen
Quoted 21 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_store_running':

   qemu/migration/global_state.c:45:5: error: 'strncpy' specified bound 100 equals destination size [-Werror=stringop-truncation]
        strncpy((char *)global_state.runstate, state, sizeof(global_state.runstate));
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

  typedef struct {
      uint32_t size;

Review

Quoted 2 lines
(quoted diff) +1 -1
-    uint8_t runstate[100];
+    uint8_t runstate[100] QEMU_NONSTRING;

Since 100 bytes for runstate[] is larger than any string possible in our current enum string values, could we instead add an assert that strlen(state) < sizeof(global_state.runstate), and then use strpadcpy() to make our intent obvious while still shutting up the compiler warning, but without having to deal with the fallout of marking runstate as a non-string?

Signature · 37 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  New

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  If this is the intended behavior, please suppress the warning. For
  example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

  This also happens on other sources, for example hw/acpi/core.c, so
  another option is to suppress it globally on CFLAGS (-Wno-stringop-
  truncation)

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
#7 lore raw
In reply to Amir Gonnen

** Changed in: qemu
Status: Fix Committed => Fix Released

Signature · 32 lines
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1803872

Title:
  gcc 8.2 reports stringop-truncation when building qemu

Status in QEMU:
  Fix Released

Bug description:
  QEMU 3.0

  block/sheepdog.c: In function 'find_vdi_name':
  block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  If this is the intended behavior, please suppress the warning. For
  example:

  #pragma GCC diagnostic push
  #pragma GCC diagnostic ignored "-Wstringop-truncation"
      strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
  #pragma GCC diagnostic pop

  This also happens on other sources, for example hw/acpi/core.c, so
  another option is to suppress it globally on CFLAGS (-Wno-stringop-
  truncation)

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1803872/+subscriptions
#8 lore raw

v3 cover Fix strncpy() warnings for GCC8 new -Wstringop-truncation

GCC 8 new warning prevents builds to success since quite some time. First report on the mailing list is in July 2018:
https://lists.gnu.org/archive/html/qemu-devel/2018-07/msg03723.html

Various intents has been sent to fix this:
- Incorrectly using g_strlcpy()
https://lists.gnu.org/archive/html/qemu-devel/2018-08/msg03705.html
https://lists.gnu.org/archive/html/qemu-devel/2018-08/msg03706.html
- Using assert() and strpadcpy()
https://lists.gnu.org/archive/html/qemu-devel/2018-11/msg03938.html
- Use #pragma GCC diagnostic ignored "-Wstringop-truncation"
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- adding an inline wrapper with said pragma in there
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- -Wno-stringop-truncation is the makefile
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- Use the 'nonstring' attribute
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04493.html

This series add the QEMU_NONSTRING definition and use it.

Regards,

Phil.

Philippe Mathieu-Daudé (5):
qemu/compiler: Define QEMU_NONSTRING
block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays
migration: Use QEMU_NONSTRING for non NUL-terminated arrays
migration: Use strnlen() for fixed-size string

 block/sheepdog.c            |  2 +-
 hw/acpi/core.c              |  8 ++++----
 include/hw/acpi/acpi-defs.h |  8 ++++----
 include/qemu/compiler.h     | 15 +++++++++++++++
 migration/global_state.c    |  4 ++--
 5 files changed, 26 insertions(+), 11 deletions(-)
Signature · 1 line
2.17.2
#9 lore raw

PATCH v3 1/5 qemu/compiler: Define QEMU_NONSTRING

GCC 8 introduced the -Wstringop-truncation checker to detect truncation by the strncat and strncpy functions (closely related to -Wstringop-overflow, which detect buffer overflow by string-modifying functions declared in <string.h>).

Add the QEMU_NONSTRING macro which checks if the compiler supports this attribute.

Quoted 1 line

From the GCC manual [*]:

The nonstring variable attribute specifies that an object or member declaration with type array of char, signed char, or unsigned char, or pointer to such a type is intended to store character arrays that do not necessarily contain a terminating NUL. This is useful in detecting uses of such arrays or pointers with functions that expect NUL-terminated strings, and to avoid warnings when such an array or pointer is used as an argument to a bounded string manipulation function such as strncpy.

[*] https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-nonstring-variable-attribute

---
 include/qemu/compiler.h | 15 +++++++++++++++
 1 file changed, 15 insertions(+)
include/qemu/compiler.h +15 -0
index 261842beae..2d8f507c73 100644
@@ -151,6 +151,21 @@
 # define QEMU_ERROR(X)
 #endif
 
+/*
+ * The nonstring variable attribute specifies that an object or member
+ * declaration with type array of char or pointer to char is intended
+ * to store character arrays that do not necessarily contain a terminating
+ * NUL character. This is useful in detecting uses of such arrays or pointers
+ * with functions that expect NUL-terminated strings, and to avoid warnings
+ * when such an array or pointer is used as an argument to a bounded string
+ * manipulation function such as strncpy.
+ */
+#if __has_attribute(nonstring)
+# define QEMU_NONSTRING __attribute__((nonstring))
+#else
+# define QEMU_NONSTRING
+#endif
+
 /* Implement C11 _Generic via GCC builtins.  Example:
  *
  *    QEMU_GENERIC(x, (float, sinf), (long double, sinl), sin) (x)
Signature · 1 line
2.17.2
Reviewed-by on v3 1/5
#10 lore raw

Re: [Qemu-devel] [PATCH v3 1/5] qemu/compiler: Define QEMU_NONSTRING

Quoted 4 lines · Philippe Mathieu-Daudé

GCC 8 introduced the -Wstringop-truncation checker to detect truncation by the strncat and strncpy functions (closely related to -Wstringop-overflow, which detect buffer overflow by string-modifying functions declared in <string.h>).

This paragraph talks about a new warning checker, but makes no mention of an attribute.

Quoted 3 lines

Add the QEMU_NONSTRING macro which checks if the compiler supports this attribute.

Thus, "this attribute" has no antecedent; did you forget to add a sentence to the previous paragraph, or maybe put the mention of adding QEMU_NONSTRING after...

Quoted 10 lines
Quoted 2 lines

From the GCC manual [*]:

The nonstring variable attribute specifies that an object or member declaration with type array of char, signed char, or unsigned char, or pointer to such a type is intended to store character arrays that do not necessarily contain a terminating NUL. This is useful in detecting uses of such arrays or pointers with functions that expect NUL-terminated strings, and to avoid warnings when such an array or pointer is used as an argument to a bounded string manipulation function such as strncpy.

...the explanation of how the attribute was added in tandem with the new warning checker for silencing specific instances of the warning?

Quoted 9 lines

[*] https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-nonstring-variable-attribute

---
  include/qemu/compiler.h | 15 +++++++++++++++
  1 file changed, 15 insertions(+)
Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org
#11 lore raw

Re: [Qemu-devel] [PATCH v3 1/5] qemu/compiler: Define QEMU_NONSTRING

Quoted 33 lines · Eric Blake
Quoted 6 lines · Philippe Mathieu-Daudé

GCC 8 introduced the -Wstringop-truncation checker to detect truncation by
the strncat and strncpy functions (closely related to -Wstringop-overflow,
which detect buffer overflow by string-modifying functions declared in <string.h>).

This paragraph talks about a new warning checker, but makes no mention of an attribute.

Quoted 3 lines

Add the QEMU_NONSTRING macro which checks if the compiler supports this attribute.

Thus, "this attribute" has no antecedent; did you forget to add a sentence to the previous paragraph, or maybe put the mention of adding QEMU_NONSTRING after...

Quoted 13 lines
Quoted 2 lines

From the GCC manual [*]:

   The nonstring variable attribute specifies that an object or member declaration with type array of char, signed char, or unsigned char, or pointer to such a type is intended to store character arrays that do not necessarily contain a terminating NUL. This is useful in detecting
   uses of such arrays or pointers with functions that expect NUL-terminated
   strings, and to avoid warnings when such an array or pointer is used as
   an argument to a bounded string manipulation function such as strncpy.

Quoted 2 lines

...the explanation of how the attribute was added in tandem with the new warning checker for silencing specific instances of the warning?

Yes... I will rewrite this.

Quoted 6 lines
Quoted 6 lines

[*]
https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-nonstring-variable-attribute

Quoted 8 lines
Quoted 6 lines

---
  include/qemu/compiler.h | 15 +++++++++++++++

  1 file changed, 15 insertions(+)

Thanks!

#12 lore raw

PATCH v3 2/5 block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC block/sheepdog.o
qemu/block/sheepdog.c: In function 'find_vdi_name':
qemu/block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]

       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: block/sheepdog.o] Error 1

As described previous to the strncpy() calls, the use of strncpy() is correct here:

    /* This pair of strncpy calls ensures that the buffer is zero-filled,
     * which is desirable since we'll soon be sending those bytes, and
     * don't want the send_req to read uninitialized data.
     */
    strncpy(buf, filename, SD_MAX_VDI_LEN);
    strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
 block/sheepdog.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/sheepdog.c +1 -1
index 0125df9d49..d4ad6b119d 100644
@@ -1224,7 +1224,7 @@ static int find_vdi_name(BDRVSheepdogState *s, const char *filename,
     SheepdogVdiReq hdr;
     SheepdogVdiRsp *rsp = (SheepdogVdiRsp *)&hdr;
     unsigned int wlen, rlen = 0;
-    char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
+    QEMU_NONSTRING char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
 
     fd = connect_to_sdog(s, errp);
     if (fd < 0) {
Signature · 1 line
2.17.2
Reviewed-by on v3 2/5
#13 lore raw

Re: [Qemu-devel] [PATCH v3 2/5] block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 26 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC block/sheepdog.o
qemu/block/sheepdog.c: In function 'find_vdi_name':

   qemu/block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]
        strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   make: *** [qemu/rules.mak:69: block/sheepdog.o] Error 1

As described previous to the strncpy() calls, the use of strncpy() is correct here:

     /* This pair of strncpy calls ensures that the buffer is zero-filled,
      * which is desirable since we'll soon be sending those bytes, and
      * don't want the send_req to read uninitialized data.
      */
     strncpy(buf, filename, SD_MAX_VDI_LEN);
     strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
Quoted 8 lines

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
  block/sheepdog.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
Quoted 15 lines
block/sheepdog.c +1 -1
index 0125df9d49..d4ad6b119d 100644
@@ -1224,7 +1224,7 @@ static int find_vdi_name(BDRVSheepdogState *s, const char *filename,
      SheepdogVdiReq hdr;
      SheepdogVdiRsp *rsp = (SheepdogVdiRsp *)&hdr;
      unsigned int wlen, rlen = 0;
-    char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
+    QEMU_NONSTRING char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
 
      fd = connect_to_sdog(s, errp);
      if (fd < 0) {
Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org
#14 lore raw

Re: [Qemu-devel] [PATCH v3 2/5] block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Philippe Mathieu-Daudé
Quoted 40 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC block/sheepdog.o
qemu/block/sheepdog.c: In function 'find_vdi_name':
qemu/block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256 equals destination size [-Werror=stringop-truncation]

       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: block/sheepdog.o] Error 1

As described previous to the strncpy() calls, the use of strncpy() is correct here:

    /* This pair of strncpy calls ensures that the buffer is zero-filled,
     * which is desirable since we'll soon be sending those bytes, and
     * don't want the send_req to read uninitialized data.
     */
    strncpy(buf, filename, SD_MAX_VDI_LEN);
    strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
 block/sheepdog.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/sheepdog.c +0 -0
index 0125df9d49..d4ad6b119d 100644

block/sheepdog.c static int find_vdi_name(BDRVSheepdogState *s, const char *filename, · line 1227

Quoted 5 lines
block/sheepdog.c +1 -1
@@ -1224,7 +1224,7 @@ static int find_vdi_name(BDRVSheepdogState *s, const char *filename,
     SheepdogVdiReq hdr;
     SheepdogVdiRsp *rsp = (SheepdogVdiRsp *)&hdr;
     unsigned int wlen, rlen = 0;
-    char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
+    QEMU_NONSTRING char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];

In case you decide to respin anyway - this would be
a bit nicer as:
char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN] QEMU_NONSTRING

Quoted 4 lines
block/sheepdog.c +0 -0
     fd = connect_to_sdog(s, errp);
     if (fd < 0) {
Signature · 1 line
2.17.2
#15 lore raw

Re: [Qemu-devel] [PATCH v3 2/5] block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 49 lines · Michael S. Tsirkin
Quoted 7 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

Quoted 5 lines

This new warning leads to compilation failures:

CC block/sheepdog.o
qemu/block/sheepdog.c: In function 'find_vdi_name':
qemu/block/sheepdog.c:1239:5: error: 'strncpy' specified bound 256

equals destination size [-Werror=stringop-truncation]

block/sheepdog.c static int find_vdi_name(BDRVSheepdogState *s,

Quoted 28 lines
       strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: block/sheepdog.o] Error 1

As described previous to the strncpy() calls, the use of strncpy() is correct here:

    /* This pair of strncpy calls ensures that the buffer is zero-filled,
     * which is desirable since we'll soon be sending those bytes, and
     * don't want the send_req to read uninitialized data.
     */
    strncpy(buf, filename, SD_MAX_VDI_LEN);
    strncpy(buf + SD_MAX_VDI_LEN, tag, SD_MAX_VDI_TAG_LEN);

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
 block/sheepdog.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/sheepdog.c +0 -0
index 0125df9d49..d4ad6b119d 100644
@@ -1224,7 +1224,7 @@ static int find_vdi_name(BDRVSheepdogState *s,

const char *filename,

block/sheepdog.c static int find_vdi_name(BDRVSheepdogState *s, const char *filename, · line 1227

Quoted 5 lines
block/sheepdog.c +1 -1
     SheepdogVdiReq hdr;
     SheepdogVdiRsp *rsp = (SheepdogVdiRsp *)&hdr;
     unsigned int wlen, rlen = 0;
-    char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
+    QEMU_NONSTRING char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN];
Quoted 4 lines

In case you decide to respin anyway - this would be
a bit nicer as:
char buf[SD_MAX_VDI_LEN + SD_MAX_VDI_TAG_LEN] QEMU_NONSTRING

I'll have to, so OK.
(it looked to me more explicit prepended).

Quoted 8 lines
Quoted 7 lines
block/sheepdog.c +0 -1
     fd = connect_to_sdog(s, errp);
     if (fd < 0) {
--

2.17.2

#16 lore raw

PATCH v3 3/5 hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

---
 hw/acpi/core.c              | 8 ++++----
 include/hw/acpi/acpi-defs.h | 8 ++++----
 2 files changed, 8 insertions(+), 8 deletions(-)
hw/acpi/core.c +4 -4
index aafdc61648..f60f750c3d 100644
@@ -35,14 +35,14 @@
 struct acpi_table_header {
     uint16_t _length;         /* our length, not actual part of the hdr */
                               /* allows easier parsing for fw_cfg clients */
-    char sig[4];              /* ACPI signature (4 ASCII characters) */
+    char sig[4] QEMU_NONSTRING; /* ACPI signature (4 ASCII characters) */
     uint32_t length;          /* Length of table, in bytes, including header */
     uint8_t revision;         /* ACPI Specification minor version # */
     uint8_t checksum;         /* To make sum of entire table == 0 */
-    char oem_id[6];           /* OEM identification */
-    char oem_table_id[8];     /* OEM table identification */
+    char oem_id[6] QEMU_NONSTRING; /* OEM identification */
+    char oem_table_id[8] QEMU_NONSTRING; /* OEM table identification */
     uint32_t oem_revision;    /* OEM revision number */
-    char asl_compiler_id[4];  /* ASL compiler vendor ID */
+    char asl_compiler_id[4] QEMU_NONSTRING; /* ASL compiler vendor ID */
     uint32_t asl_compiler_revision; /* ASL compiler revision number */
 } QEMU_PACKED;
 
include/hw/acpi/acpi-defs.h +4 -4
index af8e023968..3bf0bec8ba 100644
@@ -43,7 +43,7 @@ enum {
 struct AcpiRsdpDescriptor {        /* Root System Descriptor Pointer */
     uint64_t signature;              /* ACPI signature, contains "RSD PTR " */
     uint8_t  checksum;               /* To make sum of struct == 0 */
-    uint8_t  oem_id [6];             /* OEM identification */
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */
     uint8_t  revision;               /* Must be 0 for 1.0, 2 for 2.0 */
     uint32_t rsdt_physical_address;  /* 32-bit physical address of RSDT */
     uint32_t length;                 /* XSDT Length in bytes including hdr */
@@ -62,10 +62,10 @@ typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;
     uint32_t length;                 /* Length of table, in bytes, including header */ \
     uint8_t  revision;               /* ACPI Specification minor version # */ \
     uint8_t  checksum;               /* To make sum of entire table == 0 */ \
-    uint8_t  oem_id [6];             /* OEM identification */ \
-    uint8_t  oem_table_id [8];       /* OEM table identification */ \
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */ \
+    uint8_t  oem_table_id [8] QEMU_NONSTRING; /* OEM table identification */ \
     uint32_t oem_revision;           /* OEM revision number */ \
-    uint8_t  asl_compiler_id [4];    /* ASL compiler vendor ID */ \
+    uint8_t  asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor ID */ \
     uint32_t asl_compiler_revision;  /* ASL compiler revision number */
 
 
Signature · 1 line
2.17.2
#17 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 54 lines · On Tue, 18 Dec 2018 18:51:20 +0100 Philippe Mathieu-Daudé <[email protected]>…

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

---
 hw/acpi/core.c              | 8 ++++----
 include/hw/acpi/acpi-defs.h | 8 ++++----
 2 files changed, 8 insertions(+), 8 deletions(-)
hw/acpi/core.c +4 -4
index aafdc61648..f60f750c3d 100644
@@ -35,14 +35,14 @@
 struct acpi_table_header {
     uint16_t _length;         /* our length, not actual part of the hdr */
                               /* allows easier parsing for fw_cfg clients */
-    char sig[4];              /* ACPI signature (4 ASCII characters) */
+    char sig[4] QEMU_NONSTRING; /* ACPI signature (4 ASCII characters) */
     uint32_t length;          /* Length of table, in bytes, including header */
     uint8_t revision;         /* ACPI Specification minor version # */
     uint8_t checksum;         /* To make sum of entire table == 0 */
-    char oem_id[6];           /* OEM identification */
-    char oem_table_id[8];     /* OEM table identification */
+    char oem_id[6] QEMU_NONSTRING; /* OEM identification */
+    char oem_table_id[8] QEMU_NONSTRING; /* OEM table identification */
     uint32_t oem_revision;    /* OEM revision number */
-    char asl_compiler_id[4];  /* ASL compiler vendor ID */
+    char asl_compiler_id[4] QEMU_NONSTRING; /* ASL compiler vendor ID */
     uint32_t asl_compiler_revision; /* ASL compiler revision number */
 } QEMU_PACKED;
 
include/hw/acpi/acpi-defs.h +0 -0
index af8e023968..3bf0bec8ba 100644

include/hw/acpi/acpi-defs.h enum { · line 49

Quoted 8 lines
include/hw/acpi/acpi-defs.h +1 -1
@@ -43,7 +43,7 @@ enum {
 struct AcpiRsdpDescriptor {        /* Root System Descriptor Pointer */
     uint64_t signature;              /* ACPI signature, contains "RSD PTR " */
     uint8_t  checksum;               /* To make sum of struct == 0 */
-    uint8_t  oem_id [6];             /* OEM identification */
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */
     uint8_t  revision;               /* Must be 0 for 1.0, 2 for 2.0 */
     uint32_t rsdt_physical_address;  /* 32-bit physical address of RSDT */
     uint32_t length;                 /* XSDT Length in bytes including hdr */

you'll need to rebase this on top the latest Michael's pull request. [PULL v2 25/30] hw: arm: Carry RSDP specific data through AcpiRsdpData [PULL v2 29/30] hw: acpi: Remove AcpiRsdpDescriptor and fix tests

Quoted 14 lines
include/hw/acpi/acpi-defs.h +3 -3
@@ -62,10 +62,10 @@ typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;
     uint32_t length;                 /* Length of table, in bytes, including header */ \
     uint8_t  revision;               /* ACPI Specification minor version # */ \
     uint8_t  checksum;               /* To make sum of entire table == 0 */ \
-    uint8_t  oem_id [6];             /* OEM identification */ \
-    uint8_t  oem_table_id [8];       /* OEM table identification */ \
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */ \
+    uint8_t  oem_table_id [8] QEMU_NONSTRING; /* OEM table identification */ \
     uint32_t oem_revision;           /* OEM revision number */ \
-    uint8_t  asl_compiler_id [4];    /* ASL compiler vendor ID */ \
+    uint8_t  asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor ID */ \
     uint32_t asl_compiler_revision;  /* ASL compiler revision number */
 
 
#18 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 71 lines · Igor Mammedov
Quoted 11 lines · On Tue, 18 Dec 2018 18:51:20 +0100 Philippe Mathieu-Daudé <[email protected]>…

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at

qemu/hw/acpi/core.c:296:5:

Quoted 1 line

qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals

destination size [-Werror=stringop-truncation]

Quoted 19 lines
           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

---
 hw/acpi/core.c              | 8 ++++----
 include/hw/acpi/acpi-defs.h | 8 ++++----
 2 files changed, 8 insertions(+), 8 deletions(-)
hw/acpi/core.c +0 -0
index aafdc61648..f60f750c3d 100644

hw/acpi/core.c line 36

Quoted 2 lines
hw/acpi/core.c +0 -0
@@ -35,14 +35,14 @@
 struct acpi_table_header {
     uint16_t _length;         /* our length, not actual part of the hdr

*/

hw/acpi/core.c line 37

Quoted 1 line
hw/acpi/core.c +0 -0
                               /* allows easier parsing for fw_cfg

clients */

hw/acpi/core.c line 38

Quoted 2 lines
hw/acpi/core.c +1 -1
-    char sig[4];              /* ACPI signature (4 ASCII characters) */
+    char sig[4] QEMU_NONSTRING; /* ACPI signature (4 ASCII characters)

*/

hw/acpi/core.c line 39

Quoted 1 line
hw/acpi/core.c +0 -0
     uint32_t length;          /* Length of table, in bytes, including

header */

Quoted 17 lines
hw/acpi/core.c +3 -3
     uint8_t revision;         /* ACPI Specification minor version # */
     uint8_t checksum;         /* To make sum of entire table == 0 */
-    char oem_id[6];           /* OEM identification */
-    char oem_table_id[8];     /* OEM table identification */
+    char oem_id[6] QEMU_NONSTRING; /* OEM identification */
+    char oem_table_id[8] QEMU_NONSTRING; /* OEM table identification */
     uint32_t oem_revision;    /* OEM revision number */
-    char asl_compiler_id[4];  /* ASL compiler vendor ID */
+    char asl_compiler_id[4] QEMU_NONSTRING; /* ASL compiler vendor ID */
     uint32_t asl_compiler_revision; /* ASL compiler revision number */
 } QEMU_PACKED;
 
include/hw/acpi/acpi-defs.h +0 -0
index af8e023968..3bf0bec8ba 100644

include/hw/acpi/acpi-defs.h enum { · line 44

Quoted 2 lines
include/hw/acpi/acpi-defs.h +0 -0
@@ -43,7 +43,7 @@ enum {
 struct AcpiRsdpDescriptor {        /* Root System Descriptor Pointer */
     uint64_t signature;              /* ACPI signature, contains "RSD

PTR " */

include/hw/acpi/acpi-defs.h enum {

Quoted 5 lines
include/hw/acpi/acpi-defs.h +1 -1
     uint8_t  checksum;               /* To make sum of struct == 0 */
-    uint8_t  oem_id [6];             /* OEM identification */
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */
     uint8_t  revision;               /* Must be 0 for 1.0, 2 for 2.0 */
     uint32_t rsdt_physical_address;  /* 32-bit physical address of RSDT
Quoted 8 lines

*/

include/hw/acpi/acpi-defs.h enum { · line 49

Quoted 1 line
include/hw/acpi/acpi-defs.h +0 -0
     uint32_t length;                 /* XSDT Length in bytes including

hdr */

you'll need to rebase this on top the latest Michael's pull request. [PULL v2 25/30] hw: arm: Carry RSDP specific data through AcpiRsdpData [PULL v2 29/30] hw: acpi: Remove AcpiRsdpDescriptor and fix tests

OK. Can I add your Ack-by then?

Quoted 21 lines
(quoted diff) +0 -0
@@ -62,10 +62,10 @@ typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;
 >      uint32_t length;                 /* Length of table, in bytes,
 including header */ \
Quoted 1 line

uint8_t revision; /* ACPI Specification minor

version # */ \

Quoted 1 line
     uint8_t  checksum;               /* To make sum of entire table ==

0 */ \

include/hw/acpi/acpi-defs.h typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;

Quoted 4 lines
include/hw/acpi/acpi-defs.h +2 -2
-    uint8_t  oem_id [6];             /* OEM identification */ \
-    uint8_t  oem_table_id [8];       /* OEM table identification */ \
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */ \
+    uint8_t  oem_table_id [8] QEMU_NONSTRING; /* OEM table

identification */ \

include/hw/acpi/acpi-defs.h

Quoted 3 lines
     uint32_t oem_revision;           /* OEM revision number */ \
include/hw/acpi/acpi-defs.h +1 -1
-    uint8_t  asl_compiler_id [4];    /* ASL compiler vendor ID */ \
+    uint8_t  asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor

ID */ \

Quoted 3 lines

uint32_t asl_compiler_revision; /* ASL compiler revision number */

Reviewed-by
#19 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 75 lines · On Wed, 19 Dec 2018 10:20:36 +0100 Philippe Mathieu-Daudé <[email protected]>…
Quoted 73 lines · Igor Mammedov
Quoted 11 lines · On Tue, 18 Dec 2018 18:51:20 +0100 Philippe Mathieu-Daudé <[email protected]>…

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at

qemu/hw/acpi/core.c:296:5:

Quoted 1 line

qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals

destination size [-Werror=stringop-truncation]

Quoted 19 lines
           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

---
 hw/acpi/core.c              | 8 ++++----
 include/hw/acpi/acpi-defs.h | 8 ++++----
 2 files changed, 8 insertions(+), 8 deletions(-)
hw/acpi/core.c +0 -0
index aafdc61648..f60f750c3d 100644

hw/acpi/core.c line 36

Quoted 2 lines
hw/acpi/core.c +0 -0
@@ -35,14 +35,14 @@
 struct acpi_table_header {
     uint16_t _length;         /* our length, not actual part of the hdr  

*/

hw/acpi/core.c line 37

Quoted 1 line
hw/acpi/core.c +0 -0
                               /* allows easier parsing for fw_cfg  

clients */

hw/acpi/core.c line 38

Quoted 2 lines
hw/acpi/core.c +1 -1
-    char sig[4];              /* ACPI signature (4 ASCII characters) */
+    char sig[4] QEMU_NONSTRING; /* ACPI signature (4 ASCII characters)  

*/

hw/acpi/core.c line 39

Quoted 1 line
hw/acpi/core.c +0 -0
     uint32_t length;          /* Length of table, in bytes, including  

header */

Quoted 17 lines
hw/acpi/core.c +3 -3
     uint8_t revision;         /* ACPI Specification minor version # */
     uint8_t checksum;         /* To make sum of entire table == 0 */
-    char oem_id[6];           /* OEM identification */
-    char oem_table_id[8];     /* OEM table identification */
+    char oem_id[6] QEMU_NONSTRING; /* OEM identification */
+    char oem_table_id[8] QEMU_NONSTRING; /* OEM table identification */
     uint32_t oem_revision;    /* OEM revision number */
-    char asl_compiler_id[4];  /* ASL compiler vendor ID */
+    char asl_compiler_id[4] QEMU_NONSTRING; /* ASL compiler vendor ID */
     uint32_t asl_compiler_revision; /* ASL compiler revision number */
 } QEMU_PACKED;
 
include/hw/acpi/acpi-defs.h +0 -0
index af8e023968..3bf0bec8ba 100644

include/hw/acpi/acpi-defs.h enum { · line 44

Quoted 2 lines
include/hw/acpi/acpi-defs.h +0 -0
@@ -43,7 +43,7 @@ enum {
 struct AcpiRsdpDescriptor {        /* Root System Descriptor Pointer */
     uint64_t signature;              /* ACPI signature, contains "RSD  

PTR " */

include/hw/acpi/acpi-defs.h enum {

Quoted 5 lines
include/hw/acpi/acpi-defs.h +1 -1
     uint8_t  checksum;               /* To make sum of struct == 0 */
-    uint8_t  oem_id [6];             /* OEM identification */
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */
     uint8_t  revision;               /* Must be 0 for 1.0, 2 for 2.0 */
     uint32_t rsdt_physical_address;  /* 32-bit physical address of RSDT  

*/

Quoted 1 line
include/hw/acpi/acpi-defs.h +0 -0
     uint32_t length;                 /* XSDT Length in bytes including  
Quoted 8 lines
Quoted 6 lines

hdr */

you'll need to rebase this on top the latest Michael's pull request. [PULL v2 25/30] hw: arm: Carry RSDP specific data through AcpiRsdpData [PULL v2 29/30] hw: acpi: Remove AcpiRsdpDescriptor and fix tests

OK. Can I add your Ack-by then?

pls note that new AcpiRsdpData has oem_id field that needs the same treatment

with rebase

Quoted 22 lines
Quoted 22 lines
(quoted diff) +0 -0
@@ -62,10 +62,10 @@ typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;  
 >      uint32_t length;                 /* Length of table, in bytes,  
 including header */ \  
Quoted 1 line

uint8_t revision; /* ACPI Specification minor

version # */ \

Quoted 1 line
     uint8_t  checksum;               /* To make sum of entire table ==  

0 */ \

include/hw/acpi/acpi-defs.h typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;

Quoted 4 lines
include/hw/acpi/acpi-defs.h +2 -2
-    uint8_t  oem_id [6];             /* OEM identification */ \
-    uint8_t  oem_table_id [8];       /* OEM table identification */ \
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */ \
+    uint8_t  oem_table_id [8] QEMU_NONSTRING; /* OEM table  

identification */ \

include/hw/acpi/acpi-defs.h

Quoted 3 lines
     uint32_t oem_revision;           /* OEM revision number */ \
include/hw/acpi/acpi-defs.h +1 -1
-    uint8_t  asl_compiler_id [4];    /* ASL compiler vendor ID */ \
+    uint8_t  asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor  

ID */ \

Quoted 3 lines

uint32_t asl_compiler_revision; /* ASL compiler revision number */

#20 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Philippe Mathieu-Daudé
Quoted 12 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

Quoted 6 lines
           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

Aren't we always starting with zero-initialized structures in ACPI code? If so, then we should be able to change the strncpy's to memcpy's.

Thanks,
drew

Quoted 62 lines
---
 hw/acpi/core.c              | 8 ++++----
 include/hw/acpi/acpi-defs.h | 8 ++++----
 2 files changed, 8 insertions(+), 8 deletions(-)
hw/acpi/core.c +4 -4
index aafdc61648..f60f750c3d 100644
@@ -35,14 +35,14 @@
 struct acpi_table_header {
     uint16_t _length;         /* our length, not actual part of the hdr */
                               /* allows easier parsing for fw_cfg clients */
-    char sig[4];              /* ACPI signature (4 ASCII characters) */
+    char sig[4] QEMU_NONSTRING; /* ACPI signature (4 ASCII characters) */
     uint32_t length;          /* Length of table, in bytes, including header */
     uint8_t revision;         /* ACPI Specification minor version # */
     uint8_t checksum;         /* To make sum of entire table == 0 */
-    char oem_id[6];           /* OEM identification */
-    char oem_table_id[8];     /* OEM table identification */
+    char oem_id[6] QEMU_NONSTRING; /* OEM identification */
+    char oem_table_id[8] QEMU_NONSTRING; /* OEM table identification */
     uint32_t oem_revision;    /* OEM revision number */
-    char asl_compiler_id[4];  /* ASL compiler vendor ID */
+    char asl_compiler_id[4] QEMU_NONSTRING; /* ASL compiler vendor ID */
     uint32_t asl_compiler_revision; /* ASL compiler revision number */
 } QEMU_PACKED;
 
include/hw/acpi/acpi-defs.h +4 -4
index af8e023968..3bf0bec8ba 100644
@@ -43,7 +43,7 @@ enum {
 struct AcpiRsdpDescriptor {        /* Root System Descriptor Pointer */
     uint64_t signature;              /* ACPI signature, contains "RSD PTR " */
     uint8_t  checksum;               /* To make sum of struct == 0 */
-    uint8_t  oem_id [6];             /* OEM identification */
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */
     uint8_t  revision;               /* Must be 0 for 1.0, 2 for 2.0 */
     uint32_t rsdt_physical_address;  /* 32-bit physical address of RSDT */
     uint32_t length;                 /* XSDT Length in bytes including hdr */
@@ -62,10 +62,10 @@ typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;
     uint32_t length;                 /* Length of table, in bytes, including header */ \
     uint8_t  revision;               /* ACPI Specification minor version # */ \
     uint8_t  checksum;               /* To make sum of entire table == 0 */ \
-    uint8_t  oem_id [6];             /* OEM identification */ \
-    uint8_t  oem_table_id [8];       /* OEM table identification */ \
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */ \
+    uint8_t  oem_table_id [8] QEMU_NONSTRING; /* OEM table identification */ \
     uint32_t oem_revision;           /* OEM revision number */ \
-    uint8_t  asl_compiler_id [4];    /* ASL compiler vendor ID */ \
+    uint8_t  asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor ID */ \
     uint32_t asl_compiler_revision;  /* ASL compiler revision number */
 
 
Signature · 1 line
2.17.2
#21 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Hi Drew,

Quoted 17 lines · Andrew Jones
Quoted 16 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Quoted 5 lines
Quoted 2 lines

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

Aren't we always starting with zero-initialized structures in ACPI code? If so, then we should be able to change the strncpy's to memcpy's.

The first call zero-initializes, but then we call realloc():

    /* We won't fail from here on. Initialize / extend the globals. */
    if (acpi_tables == NULL) {
        acpi_tables_len = sizeof(uint16_t);
        acpi_tables = g_malloc0(acpi_tables_len);
    }

    acpi_tables = g_realloc(acpi_tables, acpi_tables_len +
                                         ACPI_TABLE_PFX_SIZE +
                                         sizeof dfl_hdr + body_size);

    ext_hdr = (struct acpi_table_header *)(acpi_tables +
                                           acpi_tables_len);

So memcpy() isn't enough.

I can resend the previous patch which uses strpadcpy() if you prefer, Igor already reviewed it:

https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04406.html

Quoted 62 lines
Quoted 62 lines
---
 hw/acpi/core.c              | 8 ++++----
 include/hw/acpi/acpi-defs.h | 8 ++++----
 2 files changed, 8 insertions(+), 8 deletions(-)
hw/acpi/core.c +4 -4
index aafdc61648..f60f750c3d 100644
@@ -35,14 +35,14 @@
 struct acpi_table_header {
     uint16_t _length;         /* our length, not actual part of the hdr */
                               /* allows easier parsing for fw_cfg clients */
-    char sig[4];              /* ACPI signature (4 ASCII characters) */
+    char sig[4] QEMU_NONSTRING; /* ACPI signature (4 ASCII characters) */
     uint32_t length;          /* Length of table, in bytes, including header */
     uint8_t revision;         /* ACPI Specification minor version # */
     uint8_t checksum;         /* To make sum of entire table == 0 */
-    char oem_id[6];           /* OEM identification */
-    char oem_table_id[8];     /* OEM table identification */
+    char oem_id[6] QEMU_NONSTRING; /* OEM identification */
+    char oem_table_id[8] QEMU_NONSTRING; /* OEM table identification */
     uint32_t oem_revision;    /* OEM revision number */
-    char asl_compiler_id[4];  /* ASL compiler vendor ID */
+    char asl_compiler_id[4] QEMU_NONSTRING; /* ASL compiler vendor ID */
     uint32_t asl_compiler_revision; /* ASL compiler revision number */
 } QEMU_PACKED;
 
include/hw/acpi/acpi-defs.h +4 -4
index af8e023968..3bf0bec8ba 100644
@@ -43,7 +43,7 @@ enum {
 struct AcpiRsdpDescriptor {        /* Root System Descriptor Pointer */
     uint64_t signature;              /* ACPI signature, contains "RSD PTR " */
     uint8_t  checksum;               /* To make sum of struct == 0 */
-    uint8_t  oem_id [6];             /* OEM identification */
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */
     uint8_t  revision;               /* Must be 0 for 1.0, 2 for 2.0 */
     uint32_t rsdt_physical_address;  /* 32-bit physical address of RSDT */
     uint32_t length;                 /* XSDT Length in bytes including hdr */
@@ -62,10 +62,10 @@ typedef struct AcpiRsdpDescriptor AcpiRsdpDescriptor;
     uint32_t length;                 /* Length of table, in bytes, including header */ \
     uint8_t  revision;               /* ACPI Specification minor version # */ \
     uint8_t  checksum;               /* To make sum of entire table == 0 */ \
-    uint8_t  oem_id [6];             /* OEM identification */ \
-    uint8_t  oem_table_id [8];       /* OEM table identification */ \
+    uint8_t  oem_id [6] QEMU_NONSTRING; /* OEM identification */ \
+    uint8_t  oem_table_id [8] QEMU_NONSTRING; /* OEM table identification */ \
     uint32_t oem_revision;           /* OEM revision number */ \
-    uint8_t  asl_compiler_id [4];    /* ASL compiler vendor ID */ \
+    uint8_t  asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor ID */ \
     uint32_t asl_compiler_revision;  /* ASL compiler revision number */
 
 
Signature · 1 line
2.17.2
#22 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 41 lines · Philippe Mathieu-Daudé

Hi Drew,

Quoted 17 lines · Andrew Jones
Quoted 16 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Quoted 5 lines
Quoted 2 lines

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

Aren't we always starting with zero-initialized structures in ACPI code? If so, then we should be able to change the strncpy's to memcpy's.

The first call zero-initializes, but then we call realloc():

    /* We won't fail from here on. Initialize / extend the globals. */
    if (acpi_tables == NULL) {
        acpi_tables_len = sizeof(uint16_t);
        acpi_tables = g_malloc0(acpi_tables_len);
    }

    acpi_tables = g_realloc(acpi_tables, acpi_tables_len +
                                         ACPI_TABLE_PFX_SIZE +
                                         sizeof dfl_hdr + body_size);

    ext_hdr = (struct acpi_table_header *)(acpi_tables +
                                           acpi_tables_len);
Quoted 1 line

So memcpy() isn't enough.

Ah, thanks.

Quoted 6 lines

I can resend the previous patch which uses strpadcpy() if you prefer, Igor already reviewed it:

https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04406.html

I do like strpadcpy() better, but I'm not going to lose sleep about this either way it goes.

Thanks,
drew

#23 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 47 lines · On Wed, 19 Dec 2018 14:00:37 +0100 Andrew Jones <[email protected]> wrote:
Quoted 41 lines · Philippe Mathieu-Daudé

Hi Drew,

Quoted 17 lines · Andrew Jones
Quoted 16 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Quoted 5 lines
Quoted 2 lines

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

Aren't we always starting with zero-initialized structures in ACPI code? If so, then we should be able to change the strncpy's to memcpy's.

The first call zero-initializes, but then we call realloc():

    /* We won't fail from here on. Initialize / extend the globals. */
    if (acpi_tables == NULL) {
        acpi_tables_len = sizeof(uint16_t);
        acpi_tables = g_malloc0(acpi_tables_len);
    }

    acpi_tables = g_realloc(acpi_tables, acpi_tables_len +
                                         ACPI_TABLE_PFX_SIZE +
                                         sizeof dfl_hdr + body_size);

    ext_hdr = (struct acpi_table_header *)(acpi_tables +
                                           acpi_tables_len);
Quoted 1 line

So memcpy() isn't enough.

Ah, thanks.

Quoted 1 line
Quoted 8 lines
Quoted 5 lines

I can resend the previous patch which uses strpadcpy() if you prefer, Igor already reviewed it:

https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04406.html

I do like strpadcpy() better, but I'm not going to lose sleep about this either way it goes.

I'm ok with both ways, but v2 consensus was to use QEMU_NONSTRING if I got it right

Quoted 3 lines

Thanks,
drew

#24 lore raw

Re: [Qemu-devel] [PATCH v3 3/5] hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 54 lines · Igor Mammedov
Quoted 50 lines · On Wed, 19 Dec 2018 14:00:37 +0100 Andrew Jones <[email protected]> wrote:
Quoted 41 lines · Philippe Mathieu-Daudé

Hi Drew,

Quoted 17 lines · Andrew Jones
Quoted 16 lines · Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC hw/acpi/core.o
In function 'acpi_table_install', inlined from 'acpi_table_add' at qemu/hw/acpi/core.c:296:5:
qemu/hw/acpi/core.c:184:9: error: 'strncpy' specified bound 4 equals destination size [-Werror=stringop-truncation]

           strncpy(ext_hdr->sig, hdrs->sig, sizeof ext_hdr->sig);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: hw/acpi/core.o] Error 1

Quoted 5 lines
Quoted 2 lines

Use the QEMU_NONSTRING attribute, since ACPI tables don't require the strings to be NUL-terminated.

Aren't we always starting with zero-initialized structures in ACPI code? If so, then we should be able to change the strncpy's to memcpy's.

The first call zero-initializes, but then we call realloc():

    /* We won't fail from here on. Initialize / extend the globals. */
    if (acpi_tables == NULL) {
        acpi_tables_len = sizeof(uint16_t);
        acpi_tables = g_malloc0(acpi_tables_len);
    }

    acpi_tables = g_realloc(acpi_tables, acpi_tables_len +
                                         ACPI_TABLE_PFX_SIZE +
                                         sizeof dfl_hdr + body_size);

    ext_hdr = (struct acpi_table_header *)(acpi_tables +
                                           acpi_tables_len);
Quoted 1 line

So memcpy() isn't enough.

Ah, thanks.

Quoted 4 lines

I can resend the previous patch which uses strpadcpy() if you prefer, Igor already reviewed it:

Quoted 6 lines
Quoted 5 lines
Quoted 2 lines

https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04406.html

I do like strpadcpy() better, but I'm not going to lose sleep about this either way it goes.

I'm ok with both ways, but v2 consensus was to use QEMU_NONSTRING if I got it right

Yes, MST recommended it because this attribute is clever than strpadcpy().

#25 lore raw

PATCH v3 4/5 migration: Use QEMU_NONSTRING for non NUL-terminated arrays

In reply to Philippe Mathieu-Daudé

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_store_running':
qemu/migration/global_state.c:45:5: error: 'strncpy' specified bound 100 equals destination size [-Werror=stringop-truncation]

       strncpy((char *)global_state.runstate, state, sizeof(global_state.runstate));
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
 migration/global_state.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
migration/global_state.c +1 -1
index 8e8ab5c51e..6e19333422 100644
@@ -21,7 +21,7 @@
 
 typedef struct {
     uint32_t size;
-    uint8_t runstate[100];
+    uint8_t runstate[100] QEMU_NONSTRING;
     RunState state;
     bool received;
 } GlobalState;
Signature · 1 line
2.17.2
#26 lore raw

Re: [Qemu-devel] [PATCH v3 4/5] migration: Use QEMU_NONSTRING for non NUL-terminated arrays

Quoted 30 lines · * Philippe Mathieu-Daudé ([email protected]) wrote:

GCC 8 added a -Wstringop-truncation warning:

The -Wstringop-truncation warning added in GCC 8.0 via r254630 for bug 81117 is specifically intended to highlight likely unintended uses of the strncpy function that truncate the terminating NUL character from the source string.

This new warning leads to compilation failures:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_store_running':
qemu/migration/global_state.c:45:5: error: 'strncpy' specified bound 100 equals destination size [-Werror=stringop-truncation]

       strncpy((char *)global_state.runstate, state, sizeof(global_state.runstate));
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

Use the QEMU_NONSTRING attribute, since this array is intended to store character arrays that do not necessarily contain a terminating NUL.

---
 migration/global_state.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
migration/global_state.c +0 -0
index 8e8ab5c51e..6e19333422 100644

migration/global_state.c line 24

Quoted 5 lines
migration/global_state.c +1 -1
@@ -21,7 +21,7 @@
 
 typedef struct {
     uint32_t size;
-    uint8_t runstate[100];
+    uint8_t runstate[100] QEMU_NONSTRING;

Hmm; global_state_post_load needs to be fixed for this; it uses s->runsate and ends up passing it to both a trace and a qapi_enum_parse - so it's really treating it as a string. That code is unsafe anyway since it's assuming the received runstate would be terminated.

Dave

Quoted 6 lines
migration/global_state.c +0 -0
     RunState state;
     bool received;
 } GlobalState;
Signature · 1 line
2.17.2

--
Dr. David Alan Gilbert / [email protected] / Manchester, UK

#27 lore raw

PATCH v3 5/5 migration: Use strnlen() for fixed-size string

In reply to Philippe Mathieu-Daudé

GCC 8 introduced the -Wstringop-overflow, which detect buffer overflow by string-modifying functions declared in <string.h>, such strncpy(), used in global_state_store_running().

Since the global_state.runstate does not necessarily contains a terminating NUL character, We had to use the QEMU_NONSTRING attribute.

The GCC manual says about the nonstring attribute:

However, when the array is declared with the attribute the call to strlen is diagnosed because when the array doesn’t contain a NUL-terminated string the call is undefined. [...]
In addition, calling strnlen and strndup with such arrays is safe provided a suitable bound is specified, and not diagnosed.

GCC indeed found an incorrect use of strlen(), because this array is loaded by VMSTATE_BUFFER(runstate, GlobalState) then parsed using qapi_enum_parse which does not get the buffer length.

Use strnlen() which returns sizeof(s->runstate) if the array is not NUL-terminated.

This fixes:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_pre_save':
qemu/migration/global_state.c:109:15: error: 'strlen' argument 1 declared attribute 'nonstring' [-Werror=stringop-overflow=]

       s->size = strlen((char *)s->runstate) + 1;
                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~
  qemu/migration/global_state.c:24:13: note: argument 'runstate' declared here
       uint8_t runstate[100] QEMU_NONSTRING;
               ^~~~~~~~

cc1: all warnings being treated as errors
make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

---
 migration/global_state.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
migration/global_state.c +1 -1
index 6e19333422..c19030ef62 100644
@@ -106,7 +106,7 @@ static int global_state_pre_save(void *opaque)
     GlobalState *s = opaque;
 
     trace_migrate_global_state_pre_save((char *)s->runstate);
-    s->size = strlen((char *)s->runstate) + 1;
+    s->size = strnlen((char *)s->runstate, sizeof(s->runstate)) + 1;
 
     return 0;
 }
Signature · 1 line
2.17.2
#28 lore raw

Re: [Qemu-devel] [PATCH v3 5/5] migration: Use strnlen() for fixed-size string

Quoted 45 lines · Philippe Mathieu-Daudé

GCC 8 introduced the -Wstringop-overflow, which detect buffer overflow by string-modifying functions declared in <string.h>, such strncpy(), used in global_state_store_running().

Since the global_state.runstate does not necessarily contains a terminating NUL character, We had to use the QEMU_NONSTRING attribute.

The GCC manual says about the nonstring attribute:

However, when the array is declared with the attribute the call to strlen is diagnosed because when the array doesn’t contain a NUL-terminated string the call is undefined. [...]
In addition, calling strnlen and strndup with such arrays is safe provided a suitable bound is specified, and not diagnosed.

GCC indeed found an incorrect use of strlen(), because this array is loaded by VMSTATE_BUFFER(runstate, GlobalState) then parsed using qapi_enum_parse which does not get the buffer length.

Use strnlen() which returns sizeof(s->runstate) if the array is not NUL-terminated.

This fixes:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_pre_save':
qemu/migration/global_state.c:109:15: error: 'strlen' argument 1 declared attribute 'nonstring' [-Werror=stringop-overflow=]

       s->size = strlen((char *)s->runstate) + 1;
                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~
  qemu/migration/global_state.c:24:13: note: argument 'runstate' declared here
       uint8_t runstate[100] QEMU_NONSTRING;
               ^~~~~~~~

cc1: all warnings being treated as errors
make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

---
 migration/global_state.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
migration/global_state.c +0 -0
index 6e19333422..c19030ef62 100644

migration/global_state.c static int global_state_pre_save(void *opaque) · line 111

Quoted 7 lines
migration/global_state.c +1 -1
@@ -106,7 +106,7 @@ static int global_state_pre_save(void *opaque)
     GlobalState *s = opaque;
 
     trace_migrate_global_state_pre_save((char *)s->runstate);
-    s->size = strlen((char *)s->runstate) + 1;
+    s->size = strnlen((char *)s->runstate, sizeof(s->runstate)) + 1;
 
     return 0;

I don't think this works correctly if strnlen returns sizeof(s->runstate). Which never happens so we probably should jus add

	assert(e->size is <= sizeof(s->runstate));

But also I think this is not enough, there's a problem in post-load in the call to qapi_enum_parse. You probably want to force the last character to 0 there.

Quoted 3 lines
migration/global_state.c +0 -0
 }
Signature · 1 line
2.17.2
#29 lore raw

Re: [Qemu-devel] [PATCH v3 5/5] migration: Use strnlen() for fixed-size string

Quoted 54 lines · Michael S. Tsirkin
Quoted 19 lines · Philippe Mathieu-Daudé

GCC 8 introduced the -Wstringop-overflow, which detect buffer overflow by string-modifying functions declared in <string.h>, such strncpy(), used in global_state_store_running().

Since the global_state.runstate does not necessarily contains a terminating NUL character, We had to use the QEMU_NONSTRING attribute.

The GCC manual says about the nonstring attribute:

However, when the array is declared with the attribute the call to strlen is diagnosed because when the array doesn’t contain a NUL-terminated string the call is undefined. [...]
In addition, calling strnlen and strndup with such arrays is safe provided a suitable bound is specified, and not diagnosed.

GCC indeed found an incorrect use of strlen(), because this array is loaded by VMSTATE_BUFFER(runstate, GlobalState) then parsed using qapi_enum_parse which does not get the buffer length.

Quoted 8 lines

Use strnlen() which returns sizeof(s->runstate) if the array is not NUL-terminated.

This fixes:

CC migration/global_state.o
qemu/migration/global_state.c: In function 'global_state_pre_save':
qemu/migration/global_state.c:109:15: error: 'strlen' argument 1

declared attribute 'nonstring' [-Werror=stringop-overflow=]

Quoted 3 lines
       s->size = strlen((char *)s->runstate) + 1;
                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~

qemu/migration/global_state.c:24:13: note: argument 'runstate'

declared here

Quoted 21 lines
       uint8_t runstate[100] QEMU_NONSTRING;
               ^~~~~~~~

cc1: all warnings being treated as errors
make: *** [qemu/rules.mak:69: migration/global_state.o] Error 1

---
 migration/global_state.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
migration/global_state.c +1 -1
index 6e19333422..c19030ef62 100644
@@ -106,7 +106,7 @@ static int global_state_pre_save(void *opaque)
     GlobalState *s = opaque;
 
     trace_migrate_global_state_pre_save((char *)s->runstate);
-    s->size = strlen((char *)s->runstate) + 1;
+    s->size = strnlen((char *)s->runstate, sizeof(s->runstate)) + 1;
 
Quoted 8 lines

migration/global_state.c static int global_state_pre_save(void *opaque) · line 111

Quoted 1 line
migration/global_state.c +0 -0
     return 0;

I don't think this works correctly if strnlen returns sizeof(s->runstate). Which never happens so we probably should jus add

        assert(e->size is <= sizeof(s->runstate));

OK I'll resend Marc-André previous patch.

Quoted 4 lines

But also I think this is not enough, there's a problem in post-load in the call to qapi_enum_parse. You probably want to force the last character to 0 there.

OK, I'll have a look there.

Quoted 5 lines
Quoted 4 lines
migration/global_state.c +0 -1
 }
--

2.17.2

#30 lore raw

Re: [Qemu-devel] [PATCH v3 0/5] Fix strncpy() warnings for GCC8 new -Wstringop-truncation

In reply to Philippe Mathieu-Daudé
Quoted 10 lines · Philippe Mathieu-Daudé

GCC 8 new warning prevents builds to success since quite some time. First report on the mailing list is in July 2018:
https://lists.gnu.org/archive/html/qemu-devel/2018-07/msg03723.html

Various intents has been sent to fix this:
- Incorrectly using g_strlcpy()
https://lists.gnu.org/archive/html/qemu-devel/2018-08/msg03705.html
https://lists.gnu.org/archive/html/qemu-devel/2018-08/msg03706.html
- Using assert() and strpadcpy()
https://lists.gnu.org/archive/html/qemu-devel/2018-11/msg03938.html

And I forgot to add here:

This was the approch taken by the previous v2:
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04405.html

Quoted 14 lines

- Use #pragma GCC diagnostic ignored "-Wstringop-truncation"
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- adding an inline wrapper with said pragma in there
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- -Wno-stringop-truncation is the makefile
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- Use the 'nonstring' attribute
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04493.html

This series add the QEMU_NONSTRING definition and use it.

Regards,

Phil.

Reviewed-by on the v3 series
#31 lore raw

Re: [Qemu-devel] [PATCH v3 0/5] Fix strncpy() warnings for GCC8 new -Wstringop-truncation

In reply to Philippe Mathieu-Daudé
Quoted 19 lines · Philippe Mathieu-Daudé

GCC 8 new warning prevents builds to success since quite some time. First report on the mailing list is in July 2018:
https://lists.gnu.org/archive/html/qemu-devel/2018-07/msg03723.html

Various intents has been sent to fix this:
- Incorrectly using g_strlcpy()
https://lists.gnu.org/archive/html/qemu-devel/2018-08/msg03705.html
https://lists.gnu.org/archive/html/qemu-devel/2018-08/msg03706.html
- Using assert() and strpadcpy()
https://lists.gnu.org/archive/html/qemu-devel/2018-11/msg03938.html
- Use #pragma GCC diagnostic ignored "-Wstringop-truncation"
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- adding an inline wrapper with said pragma in there
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- -Wno-stringop-truncation is the makefile
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04261.html
- Use the 'nonstring' attribute
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg04493.html

Quoted 5 lines

This series add the QEMU_NONSTRING definition and use it.

Regards,

Phil.

Quoted 16 lines

Philippe Mathieu-Daudé (5):
qemu/compiler: Define QEMU_NONSTRING
block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays
migration: Use QEMU_NONSTRING for non NUL-terminated arrays
migration: Use strnlen() for fixed-size string

 block/sheepdog.c            |  2 +-
 hw/acpi/core.c              |  8 ++++----
 include/hw/acpi/acpi-defs.h |  8 ++++----
 include/qemu/compiler.h     | 15 +++++++++++++++
 migration/global_state.c    |  4 ++--
 5 files changed, 26 insertions(+), 11 deletions(-)
Signature · 1 line
2.17.2
#32 lore raw

Re: [Qemu-devel] [PATCH v3 0/5] Fix strncpy() warnings for GCC8 new -Wstringop-truncation

In reply to Philippe Mathieu-Daudé
Patchew URL: https://patchew.org/QEMU/[email protected]/

Patchew URL: https://patchew.org/QEMU/[email protected]/

Hi,

This series seems to have some coding style problems. See output below for more information:

Message-id: [email protected]
Type: series
Subject: [Qemu-devel] [PATCH v3 0/5] Fix strncpy() warnings for GCC8 new -Wstringop-truncation

=== TEST SCRIPT BEGIN ===
#!/bin/bash

BASE=base
n=1
total=$(git log --oneline $BASE.. | wc -l)
failed=0

git config --local diff.renamelimit 0
git config --local diff.renames True
git config --local diff.algorithm histogram

commits="$(git log --format=%H --reverse $BASE..)" for c in $commits; do

    echo "Checking PATCH $n/$total: $(git log -n 1 --format=%s $c)..."
    if ! git show $c --format=email | ./scripts/checkpatch.pl --mailback -; then
        failed=1
        echo
    fi
    n=$((n+1))

done

exit $failed
=== TEST SCRIPT END ===

Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
Switched to a new branch 'test'
df72d7a migration: Use strnlen() for fixed-size string 18211fe migration: Use QEMU_NONSTRING for non NUL-terminated arrays f5a08ba hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays 7d31f0c block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays 5ac217b qemu/compiler: Define QEMU_NONSTRING

=== OUTPUT BEGIN ===
Checking PATCH 1/5: qemu/compiler: Define QEMU_NONSTRING...
WARNING: architecture specific defines should be avoided #50: FILE: include/qemu/compiler.h:163:
+#if __has_attribute(nonstring)

total: 0 errors, 1 warnings, 21 lines checked

Your patch has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
Checking PATCH 2/5: block/sheepdog: Use QEMU_NONSTRING for non NUL-terminated arrays... Checking PATCH 3/5: hw/acpi: Use QEMU_NONSTRING for non NUL-terminated arrays...
ERROR: space prohibited before open square bracket '['
#64: FILE: include/hw/acpi/acpi-defs.h:46:
+ uint8_t oem_id [6] QEMU_NONSTRING; /* OEM identification */

WARNING: Block comments use a leading /* on a separate line #74: FILE: include/hw/acpi/acpi-defs.h:65:
+ uint8_t oem_id [6] QEMU_NONSTRING; /* OEM identification */ \

ERROR: space prohibited before open square bracket '[' #74: FILE: include/hw/acpi/acpi-defs.h:65:
+ uint8_t oem_id [6] QEMU_NONSTRING; /* OEM identification */ \

WARNING: Block comments use a leading /* on a separate line #75: FILE: include/hw/acpi/acpi-defs.h:66:
+ uint8_t oem_table_id [8] QEMU_NONSTRING; /* OEM table identification */ \

ERROR: space prohibited before open square bracket '[' #75: FILE: include/hw/acpi/acpi-defs.h:66:
+ uint8_t oem_table_id [8] QEMU_NONSTRING; /* OEM table identification */ \

WARNING: Block comments use a leading /* on a separate line #78: FILE: include/hw/acpi/acpi-defs.h:68:
+ uint8_t asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor ID */ \

ERROR: space prohibited before open square bracket '[' #78: FILE: include/hw/acpi/acpi-defs.h:68:
+ uint8_t asl_compiler_id [4] QEMU_NONSTRING; /* ASL compiler vendor ID */ \

total: 4 errors, 3 warnings, 39 lines checked

Your patch has style problems, please review. If any of these errors are false positives report them to the maintainer, see CHECKPATCH in MAINTAINERS.

Checking PATCH 4/5: migration: Use QEMU_NONSTRING for non NUL-terminated arrays... Checking PATCH 5/5: migration: Use strnlen() for fixed-size string... === OUTPUT END ===

Test command exited with code: 1

The full log is available at
http://patchew.org/logs/[email protected]/testing.checkpatch/?type=message.
---
Email generated automatically by Patchew [http://patchew.org/].
Please send your feedback to [email protected]