valgrind-users — General discussion list for Valgrind users

Re: [Valgrind-users] [Valgrind-developers] sparc64/Linux and sparc64/Solaris ports

[Lionel C. <lio...@gm...>]

On 2 August 2017 at 18:26, Ivo Raisr <iv...@iv...> wrote:
> Dear Valgrind users,
>
> Please if you know of any users interested in sparc64/Linux or
> sparc64/Solaris Valgrind ports, let me know. We would like to
> determine how much time and effort to put into developing and
> maintaining them further in a foreseeable future.

Our development group at CERN would welcome
valgrind&memcheck&cachegrind on SPARC64 a *LOT*. Preferably Illumos.

Lionel

[Valgrind-users] sparc64/Linux and sparc64/Solaris ports

[Ivo R. <iv...@iv...>]

Dear Valgrind users,

Please if you know of any users interested in sparc64/Linux or
sparc64/Solaris Valgrind ports, let me know. We would like to
determine how much time and effort to put into developing and
maintaining them further in a foreseeable future.

More information here:
http://valgrind.org/info/platforms.html

Kind regards,
I.

[Valgrind-users] Undefined Behavior in 2017 - Embedded in Academia

[Ivo R. <iv...@iv...>]

An interesting article:
    https://blog.regehr.org/archives/1520

If the long list of undefined behaviours starts getting boring or
horrifying, then skip to the nice and concise "Summary" section for
soothing words of consolation, and the actionable countermeasures.

I.

Re: [Valgrind-users] Valgrind hangs when generating supression for Qt5 menus

[Nathan B. <dea...@gm...>]

Thank you for the feedback. Setting -gen-suppressions to all helped resolve
the issues.

On Fri, Jul 14, 2017 at 12:22 PM, Paul Floyd <pa...@fr...> wrote:

>
> On 14 Jul 2017, at 21:10, Nathan Bahr wrote:
>
> > Hi,
> >
> > I made a simple Qt5 application with a single menu item and valgrind is
> configured to print suppression code.
> >
> > If I open and close a menu, valgrind holds onto the QMenu object and
> prompts to print supression code. This causes the whole setup to hang where
> I cannot interact with the application or command line interface. I can
> force-quit the application, which exits out of the command line. The
> application does not hang if gen-suppressions=no.
> >
> > Is there any way to set valgrind to automatically generate suppression
> code instead of prompting? Or to make it from the following trace?
>
> Hi
>
> You can use the option
>
>     --gen-suppressions=no|yes|all    print suppressions for errors? [no]
>
> This will generate a suppression block after the usual stacktrace,
> something like this:
>
> {
>    <insert_a_suppression_name_here>
>    Memcheck:Leak
>    match-leak-kinds: reachable
>    fun:malloc
>    fun:main
> }
>
> If you do reuse the generated suppressions then you will need to generate
> some sort of unique name.
>
> A+
> Paul
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Valgrind-users mailing list
> Val...@li...
> https://lists.sourceforge.net/lists/listinfo/valgrind-users
>

Re: [Valgrind-users] Valgrind hangs when generating supression for Qt5 menus

[Paul F. <pa...@fr...>]

On 14 Jul 2017, at 21:10, Nathan Bahr wrote:

> Hi,
> 
> I made a simple Qt5 application with a single menu item and valgrind is configured to print suppression code.
> 
> If I open and close a menu, valgrind holds onto the QMenu object and prompts to print supression code. This causes the whole setup to hang where I cannot interact with the application or command line interface. I can force-quit the application, which exits out of the command line. The application does not hang if gen-suppressions=no. 
> 
> Is there any way to set valgrind to automatically generate suppression code instead of prompting? Or to make it from the following trace?

Hi

You can use the option

    --gen-suppressions=no|yes|all    print suppressions for errors? [no]

This will generate a suppression block after the usual stacktrace, something like this:

{
   <insert_a_suppression_name_here>
   Memcheck:Leak
   match-leak-kinds: reachable
   fun:malloc
   fun:main
}

If you do reuse the generated suppressions then you will need to generate some sort of unique name.

A+
Paul

Re: [Valgrind-users] Valgrind hangs when generating supression for Qt5 menus

[David F. <fa...@kd...>]

On vendredi 14 juillet 2017 21:10:50 CEST Nathan Bahr wrote:
> Hi,
> 
> I made a simple Qt5 application with a single menu item and valgrind is
> configured to print suppression code.
> 
> If I open and close a menu, valgrind holds onto the QMenu object and
> prompts to print supression code. This causes the whole setup to hang where
> I cannot interact with the application or command line interface. I can
> force-quit the application, which exits out of the command line. The
> application does not hang if gen-suppressions=no.

This is probably due to QMenu grabbing keyboard/mouse on X11?
You can disable this by passing -nograb to the application.

> Is there any way to set valgrind to automatically generate suppression code
> instead of prompting? 

Yes, --gen-suppressions=all

-- 
David Faure, fa...@kd..., http://www.davidfaure.fr
Working on KDE Frameworks 5

[Valgrind-users] Valgrind hangs when generating supression for Qt5 menus

[Nathan B. <dea...@gm...>]

Hi,

I made a simple Qt5 application with a single menu item and valgrind is
configured to print suppression code.

If I open and close a menu, valgrind holds onto the QMenu object and
prompts to print supression code. This causes the whole setup to hang where
I cannot interact with the application or command line interface. I can
force-quit the application, which exits out of the command line. The
application does not hang if gen-suppressions=no.

Is there any way to set valgrind to automatically generate suppression code
instead of prompting? Or to make it from the following trace?

==19392== Syscall param writev(vector[...]) points to uninitialised byte(s)
==19392==    at 0x624E40D: ??? (syscall-template.S:84)
==19392==    by 0x9204F28: ??? (in
/usr/lib/x86_64-linux-gnu/libxcb.so.1.1.0)
==19392==    by 0x920531C: ??? (in
/usr/lib/x86_64-linux-gnu/libxcb.so.1.1.0)
==19392==    by 0x9205A76: ??? (in
/usr/lib/x86_64-linux-gnu/libxcb.so.1.1.0)
==19392==    by 0x9205C43: xcb_flush (in
/usr/lib/x86_64-linux-gnu/libxcb.so.1.1.0)
==19392==    by 0x40FACB3: QXcbWindow::hide() (in
/usr/lib/x86_64-linux-gnu/libQt5XcbQpa.so.5.5.1)
==19392==    by 0x6614B17: QWindow::setVisible(bool) (in
/usr/lib/x86_64-linux-gnu/libQt5Gui.so.5.5.1)
==19392==    by 0x51E8816: QWidgetPrivate::hide_sys() (in
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.5.1)
==19392==    by 0x51EFE73: QWidgetPrivate::hide_helper() (in
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.5.1)
==19392==    by 0x51F47FF: QWidget::setVisible(bool) (in
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.5.1)
==19392==    by 0x533E78B: QMenuBar::mousePressEvent(QMouseEvent*) (in
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.5.1)
==19392==    by 0x51F540E: QWidget::event(QEvent*) (in
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.5.1)
==19392==  Address 0xd4c5721 is 4,545 bytes inside a block of size 21,152
alloc'd
==19392==    at 0x4C2FB55: calloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==19392==    by 0x92048DB: xcb_connect_to_fd (in
/usr/lib/x86_64-linux-gnu/libxcb.so.1.1.0)
==19392==    by 0x9208610: xcb_connect_to_display_with_auth_info (in
/usr/lib/x86_64-linux-gnu/libxcb.so.1.1.0)
==19392==    by 0x7003809: _XConnectXCB (in
/usr/lib/x86_64-linux-gnu/libX11.so.6.3.0)
==19392==    by 0x6FF4391: XOpenDisplay (in
/usr/lib/x86_64-linux-gnu/libX11.so.6.3.0)
==19392==    by 0x40E770E:
QXcbConnection::QXcbConnection(QXcbNativeInterface*, bool, unsigned int,
char const*) (in /usr/lib/x86_64-linux-gnu/libQt5XcbQpa.so.5.5.1)
==19392==    by 0x40EABAC: QXcbIntegration::QXcbIntegration(QStringList
const&, int&, char**) (in /usr/lib/x86_64-linux-gnu/libQt5XcbQpa.so.5.5.1)
==19392==    by 0x40293AC: ??? (in
/usr/lib/x86_64-linux-gnu/qt5/plugins/platforms/libqxcb.so)
==19392==    by 0x65FAD91: QPlatformIntegrationFactory::create(QString
const&, QStringList const&, int&, char**, QString const&) (in
/usr/lib/x86_64-linux-gnu/libQt5Gui.so.5.5.1)
==19392==    by 0x6606FC3:
QGuiApplicationPrivate::createPlatformIntegration() (in
/usr/lib/x86_64-linux-gnu/libQt5Gui.so.5.5.1)
==19392==    by 0x6607ECC: QGuiApplicationPrivate::createEventDispatcher()
(in /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5.5.1)
==19392==    by 0x596E7E5: QCoreApplication::init() (in
/usr/lib/x86_64-linux-gnu/libQt5Core.so.5.5.1)

Thanks

Re: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

[Beaman, T. <Tho...@xe...>]

Hi John,

Thank you for all the useful advice. 

My first test was to just run on a simple 32bit hello world file, and it generated the same type of warnings.

I had previously checked the 64bit and 32bit default.supp files existed and were identical. But from your ideas, I did notice that the errors I am getting did not fall under any of the rules in the default.supp file.  Another source on suppression files (https://wiki.wxwidgets.org/Valgrind_Suppression_File_Howto )   lead me to do this.

valgrind --gen-suppressions=all --log-file=min.log

and running this file through the grindmerge script, I was able to add a number of new rules. This allowed my helloworld file to run without warnings, and now I will go back to testing with my application.

Thank you for all your help on this,
Tom


>> When I run valgrind on a 64bit app it appears to work ok.
>> 
>> When I switch to running valgrind on a 32bit app I get lots of system 
>> related messages (ld-2.24.so, libc-2.24.so, pthread-2.24.so ... etc) 
>> like these
>   [[snip]]
>> As a test I have tried just installing 32bit valgrind and I get the same results.
>> 
>> What am I missing to get my 32bit app not to print 100's of these messages ? What other useful info can I provide ?

> 1) Check the default suppressions file; something like /usr/lib*/valgrind/default.supp ("strace -o strace.out -e trace=file valgrind /bin/date" to find the exact names).
> Make sure it has suppressions for ld-2.24 like on x86_64, such as:
> {
>    dl-hack3-cond-1
>    Memcheck:Cond
>    obj:*/lib*/ld-2.24*.so*
>    obj:*/lib*/ld-2.24*.so*
>    obj:*/lib*/ld-2.24*.so*
> }

Re: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

[John R. <jr...@bi...>]

> ... What other useful info can I provide ?

7) Find or construct the smallest program which triggers an unwanted complaint.
Identify it (such as /bin/date, /bin/who, /bin/ls, etc.), or post its source code, here.
Somebody else might be able to reproduce your experience, or figure out what the problem is.

-- 

Re: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

[John R. <jr...@bi...>]

> 2) Run with "--trace-origins=yes" to get better clues to what memcheck was thinking,
> or where the problem actually started.

s/trace-origins/track-origins/
Sorry.

Re: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

[John R. <jr...@bi...>]

> Wild guess: Build yourself a 32-bit Valgrind to use for this purpose.

The original poster claims to have tried that:
>> As a test I have tried just installing 32bit valgrind and I get the same results.

-- 

Re: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

[John R. <jr...@bi...>]

> When I run valgrind on a 64bit app it appears to work ok.
> 
> When I switch to running valgrind on a 32bit app I get lots of system related messages (ld-2.24.so, libc-2.24.so, pthread-2.24.so … etc) like these
   [[snip]]
> As a test I have tried just installing 32bit valgrind and I get the same results.
> 
> What am I missing to get my 32bit app not to print 100’s of these messages ? What other useful info can I provide ?

1) Check the default suppressions file; something like /usr/lib*/valgrind/default.supp
("strace -o strace.out -e trace=file valgrind /bin/date" to find the exact names).
Make sure it has suppressions for ld-2.24 like on x86_64, such as:
{
    dl-hack3-cond-1
    Memcheck:Cond
    obj:*/lib*/ld-2.24*.so*
    obj:*/lib*/ld-2.24*.so*
    obj:*/lib*/ld-2.24*.so*
}

2) Run with "--trace-origins=yes" to get better clues to what memcheck was thinking,
or where the problem actually started.

3) Install the debuginfo packages for glibc and other system libraries
in order to get more information in tracebacks.  You get line numbers
and source code (!!) to help you understand what's happening.

4) Run under vgdb:  valgrind --vgdb-error=0 --trace-origins=yes /bin/date
and have another terminal window available.  This may seem cumbersome at first,
but you get exquisite control and visibility to what is going on.  Hint:
after initialization, then the first "unscripted" command is:
     (gdb) continue
which can be abbreviated "c".

5) Either under vdbg or plain gdb, examine the instruction stream
that surrounds the pc of the fault:
     (gdb) x/15i $pc-10*4
Try to identify what's going on.

6) There may be only a dozen or so locations (tracebacks) that cause most
of the problems, so generate (and use) suppressions for them.
"valgrind --gen-suppressions=yes ..." and so on.

-- 

Re: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

[Fred S. <fs...@co...>]

Wild guess: Build yourself a 32-bit Valgrind to use for this purpose.

Fred Smith
Senior Programmer/Analyst
Computrition, Inc.
175 Middlesex Turnpike
Suite 3C
Bedford, MA

From: Beaman, Thomas [mailto:Tho...@xe...]
Sent: Monday, July 10, 2017 03:25 PM
To: val...@li...
Subject: [Valgrind-users] 32bit app on 64bit os generates lots of system related output

Hi,

I have 64bit and 32bit apps running on my 64bit ppc64 linux 4.1 OS

uname -a
Linux  4.1.30-rt34-yocto2.2.1_standard+ #1 SMP PREEMPT Wed Jul 5 04:17:14 EDT 2017 ppc64 GNU/Linux

When I run valgrind on a 64bit app it appears to work ok.


When I switch to running valgrind on a 32bit app I get lots of system related messages (ld-2.24.so, libc-2.24.so, pthread-2.24.so ... etc) like these
# export VALGRIND_LIB=/usr/lib/valgrind

==14722== Use of uninitialised value of size 4
==14722==    at 0x4004A98: ??? (in /lib/ld-2.24.so)

==14722== Use of uninitialised value of size 4
==14722==    at 0xFA2F78C: clock_gettime (in /lib/libc-2.24.so)

==14722== Use of uninitialised value of size 4
==14722==    at 0xFEC2A18: ??? (in /lib/libpthread-2.24.so)


I saw this on valgrind 3.12.0 and upgraded to 3.13.0 and see the same thing.
# valgrind --version
valgrind-3.13.0

I have both 64bit and 32bit versions installed
# ls /usr/lib64/valgrind/mem*
/usr/lib64/valgrind/memcheck-ppc64be-linux

# ls /usr/lib/valgrind/mem*
/usr/lib/valgrind/memcheck-ppc32-linux

As a test I have tried just installing 32bit valgrind and I get the same results.

What am I missing to get my 32bit app not to print 100's of these messages ? What other useful info can I provide ?

Thanks,
Tom

[Valgrind-users] 32bit app on 64bit os generates lots of system related output

[Beaman, T. <Tho...@xe...>]

Hi,

I have 64bit and 32bit apps running on my 64bit ppc64 linux 4.1 OS

uname -a
Linux  4.1.30-rt34-yocto2.2.1_standard+ #1 SMP PREEMPT Wed Jul 5 04:17:14 EDT 2017 ppc64 GNU/Linux

When I run valgrind on a 64bit app it appears to work ok.


When I switch to running valgrind on a 32bit app I get lots of system related messages (ld-2.24.so, libc-2.24.so, pthread-2.24.so ... etc) like these
# export VALGRIND_LIB=/usr/lib/valgrind

==14722== Use of uninitialised value of size 4
==14722==    at 0x4004A98: ??? (in /lib/ld-2.24.so)

==14722== Use of uninitialised value of size 4
==14722==    at 0xFA2F78C: clock_gettime (in /lib/libc-2.24.so)

==14722== Use of uninitialised value of size 4
==14722==    at 0xFEC2A18: ??? (in /lib/libpthread-2.24.so)


I saw this on valgrind 3.12.0 and upgraded to 3.13.0 and see the same thing.
# valgrind --version
valgrind-3.13.0

I have both 64bit and 32bit versions installed
# ls /usr/lib64/valgrind/mem*
/usr/lib64/valgrind/memcheck-ppc64be-linux

# ls /usr/lib/valgrind/mem*
/usr/lib/valgrind/memcheck-ppc32-linux

As a test I have tried just installing 32bit valgrind and I get the same results.

What am I missing to get my 32bit app not to print 100's of these messages ? What other useful info can I provide ?

Thanks,
Tom

[Valgrind-users] 答复: [hi] Do you provide any Commercial Support?

[Wuweijia <wuw...@hu...>]

Hi:
         Can you give me a reply. I am still waiting your answer.

BR
Owen


发件人: Wuweijia
发送时间: 2017年7月7日 15:52
收件人: 'val...@li...' <val...@li...>
抄送: Fanbohao <fan...@hu...>
主题: [hi] Do you provide any Commercial Support?

Hi
I have send the email to ju...@op...<mailto:ju...@op...> two weeks ago , but there is no reply till now.

There are some requires for exp-dhat.

Please give me the answer no matter you provide it.



BR
Owen

[Valgrind-users] [HELP] can you show me more how to develop the sub-tool for valgrind

[Wuweijia <wuw...@hu...>]

Hi:
         I am developing the sub-tool for valgrind in aarch64. I want to know more information about to develop. But there is no more details on the site.

         Is there any doc about how to develop the sub-tool for valgrind, and how to use api of coregrind,  what the api of coregrind is the usage for,  Can you show me more information .


BR
Owen

Re: [Valgrind-users] Source for previous versions of valgrind

[John R. <jr...@bi...>]

On 07/07/2017 @1120Z, Ivo Raisr wrote:

> Thank you, Mark.
> Now the question is whether http://valgrind.org/downloads/old.html
> should be updated to reflect this new reality?
> I.
> 

https://bugs.kde.org/show_bug.cgi?id=382099
    valgrind release archive is not maintained

Re: [Valgrind-users] Source for previous versions of valgrind

[Ivo R. <iv...@iv...>]

2017-07-07 9:58 GMT+02:00 Mark Wielaard <ma...@kl...>:
> On Thu, 2017-07-06 at 08:05 -0700, John Reiser wrote:
>> On 07/06/2017 @1457Z, Mark Wielaard wrote:
>> > On Thu, 2017-07-06 at 13:24 +0000, Phil Longstaff wrote:
>> >> Where can I find source tarballs for 3.11.0 and 3.12.0?
>> >
>> > http://valgrind.org/downloads/old.html
>>
>> Only for 3.4.1 back to 1.9.6.  Does not include 3.5 or anything newer.
>>
>> I have 3.5, 3.6, [no 3.7], 3.8, 3.8.1, 3.9, [no 3.10], 3.11, 3.12.  Ask me.
>
> I put everything I could find at ftp://soureceware.org/pub/valgrind aka
> https://sourceware.org/pub/valgrind/ together with the last 3.13.0
> release. This way everything is in one place. Of course most are only
> interesting for historical reasons. But please let me know if I missed a
> release.

Thank you, Mark.
Now the question is whether http://valgrind.org/downloads/old.html
should be updated to reflect this new reality?
I.

Re: [Valgrind-users] Source for previous versions of valgrind

[Mark W. <ma...@kl...>]

On Thu, 2017-07-06 at 08:05 -0700, John Reiser wrote:
> On 07/06/2017 @1457Z, Mark Wielaard wrote:
> > On Thu, 2017-07-06 at 13:24 +0000, Phil Longstaff wrote:
> >> Where can I find source tarballs for 3.11.0 and 3.12.0?
> > 
> > http://valgrind.org/downloads/old.html
> 
> Only for 3.4.1 back to 1.9.6.  Does not include 3.5 or anything newer.
> 
> I have 3.5, 3.6, [no 3.7], 3.8, 3.8.1, 3.9, [no 3.10], 3.11, 3.12.  Ask me.

I put everything I could find at ftp://soureceware.org/pub/valgrind aka
https://sourceware.org/pub/valgrind/ together with the last 3.13.0
release. This way everything is in one place. Of course most are only
interesting for historical reasons. But please let me know if I missed a
release.

Cheers,

Mark

[Valgrind-users] [hi] Do you provide any Commercial Support?

[Wuweijia <wuw...@hu...>]

Hi
I have send the email to ju...@op...<mailto:ju...@op...> two weeks ago , but there is no reply till now.

There are some requires for exp-dhat.

Please give me the answer no matter you provide it.



BR
Owen

Re: [Valgrind-users] Source for previous versions of valgrind

[Phil L. <plo...@sa...>]

Thanks

-----Original Message-----
From: Mark Wielaard [mailto:ma...@kl...] 
Sent: Thursday, July 6, 2017 11:00 AM
To: Phil Longstaff
Cc: Val...@li...
Subject: Re: [Valgrind-users] Source for previous versions of valgrind

On Thu, 2017-07-06 at 16:57 +0200, Mark Wielaard wrote:
> On Thu, 2017-07-06 at 13:24 +0000, Phil Longstaff wrote:
> > Where can I find source tarballs for 3.11.0 and 3.12.0?
> 
> http://valgrind.org/downloads/old.html

O, they are not linked from that page. Sorry.

But they are there:
http://valgrind.org/downloads/valgrind-3.11.0.tar.bz2
http://valgrind.org/downloads/valgrind-3.12.0.tar.bz2

Re: [Valgrind-users] Source for previous versions of valgrind

[John R. <jr...@bi...>]

On 07/06/2017 @1457Z, Mark Wielaard wrote:
> On Thu, 2017-07-06 at 13:24 +0000, Phil Longstaff wrote:
>> Where can I find source tarballs for 3.11.0 and 3.12.0?
> 
> http://valgrind.org/downloads/old.html

Only for 3.4.1 back to 1.9.6.  Does not include 3.5 or anything newer.


I have 3.5, 3.6, [no 3.7], 3.8, 3.8.1, 3.9, [no 3.10], 3.11, 3.12.  Ask me.

-- 

Re: [Valgrind-users] Source for previous versions of valgrind

[Mark W. <ma...@kl...>]

On Thu, 2017-07-06 at 16:57 +0200, Mark Wielaard wrote:
> On Thu, 2017-07-06 at 13:24 +0000, Phil Longstaff wrote:
> > Where can I find source tarballs for 3.11.0 and 3.12.0?
> 
> http://valgrind.org/downloads/old.html

O, they are not linked from that page. Sorry.

But they are there:
http://valgrind.org/downloads/valgrind-3.11.0.tar.bz2
http://valgrind.org/downloads/valgrind-3.12.0.tar.bz2

Re: [Valgrind-users] Source for previous versions of valgrind

[Mark W. <ma...@kl...>]

On Thu, 2017-07-06 at 13:24 +0000, Phil Longstaff wrote:
> Where can I find source tarballs for 3.11.0 and 3.12.0?

http://valgrind.org/downloads/old.html

[Valgrind-users] Source for previous versions of valgrind

[Phil L. <plo...@sa...>]

Where can I find source tarballs for 3.11.0 and 3.12.0?

Phil Longstaff
Sandvine Inc