Re: [Rosegarden-user] Crashes :(

[Will Godfrey]

On Wed, 15 May 2019 20:00:37 -0400
Ted Felix  wrote:

>On 5/15/19 2:36 PM, Will Godfrey wrote:
>> My understanding is that nullptr is the 'modern' replacement for NULL, so I
>> can't see why it would fail.  
>
>   It is.
>
>   In this case, the older version of QSharedPointer provides no way for 
>the compiler to convert a nullptr (which is actually a nullptr_t) to a 
>QSharedPointer which operator= requires.  It's a bit convoluted and only 
>happens in rare situations.  QSharedPointer prior to Qt 5.8 is 
>unfortunately one of those situations.
>
>   Gorey details are attached in a small test program.
>
>> As a cross-check I tried changing a few old-style NULL with nullptr in
>> Yoshimi, and it compiled without complaint, so I'm quite lost now :(  
>
>   In most cases, nullptr is a drop-in replacement for NULL.
>
>Ted.

So far so good {fingers crossed}

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 5/15/19 2:36 PM, Will Godfrey wrote:

My understanding is that nullptr is the 'modern' replacement for NULL, so I
can't see why it would fail.


  It is.

  In this case, the older version of QSharedPointer provides no way for 
the compiler to convert a nullptr (which is actually a nullptr_t) to a 
QSharedPointer which operator= requires.  It's a bit convoluted and only 
happens in rare situations.  QSharedPointer prior to Qt 5.8 is 
unfortunately one of those situations.


  Gorey details are attached in a small test program.


As a cross-check I tried changing a few old-style NULL with nullptr in
Yoshimi, and it compiled without complaint, so I'm quite lost now :(


  In most cases, nullptr is a drop-in replacement for NULL.

Ted.
// nullptr.cpp

// To Build:
// g++ -std=c++11 nullptr.cpp

// __cplusplus values:
//201103L = C++11
//199711L = C++98 (g++ 4.7 default)
#if __cplusplus >= 201103L
// This is C++11!
#else
#error Need to build with -std=c++11
#endif

class A
{
public:
explicit A(int *p) { m_p = p; }

// Ugh.  Avoid assignment operators if you can.
A =(const A )
{
m_p = rhs.m_p;
}

private:
int *m_p;
};

int main()
{
// nullptr is compatible with pointers in general.
int *p = nullptr;

// This is also OK.
A a(nullptr);

// This is not.  Because the ctor is explicit, we must 
// explicitly ask for the conversion for this to work.
//a = nullptr;

// Explicitly asking for the conversion works.
a = static_cast(nullptr);

// Adding a (non-explicit) ctor to A that takes a nullptr_t
// also fixes the problem.  Qt 5.8 adds this to QSharedPointer.

return 0;
}

___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 5/15/19 2:36 PM, Will Godfrey wrote:

Just tried to compile this, and it fails with the report:

error: no match for ‘operator=’ (operand types are
‘QSharedPointer’ and ‘std::nullptr_t’) m_baseStyle =
nullptr;


  Should be fixed in [r15493].

Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 5/15/19 2:36 PM, Will Godfrey wrote:

Just tried to compile this, and it fails with the report:

error: no match for ‘operator=’ (operand types are
‘QSharedPointer’ and ‘std::nullptr_t’) m_baseStyle =
nullptr;


  Looks like you may have an older version of the Qt5 development 
stuff.  5.7 and prior do not support nullptr.  David fixed some of that 
in [r15476].  Looks like I need to be more careful.  Wonder if I can 
install the Qt 5.7 development stuff somehow so I can stay on top of this...


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Will Godfrey]

On Sun, 12 May 2019 19:51:54 -0400
Ted Felix  wrote:

>On 5/3/19 5:06 AM, Will Godfrey wrote:
>> I haven't forgotten this - I'm still totally bogged down with work on 
>> Yoshimi.
>> There is a *lot* going on!
>
>   Just pushed r15492 which was a memory access issue related to 
>deleting segments and the segment parameter box.  This may have fixed 
>the problem.
>
>Ted.

Just tried to compile this, and it fails with the report:

error: no match for ‘operator=’ (operand types are
‘QSharedPointer’ and ‘std::nullptr_t’) m_baseStyle =
nullptr;

What is really odd is that build 15419 compiled without issues.

As far as I can see the only difference between now and then (on my computer) is
a number of system updates fetched from debian.

My understanding is that nullptr is the 'modern' replacement for NULL, so I
can't see why it would fail.

As a cross-check I tried changing a few old-style NULL with nullptr in
Yoshimi, and it compiled without complaint, so I'm quite lost now :(

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 5/3/19 5:06 AM, Will Godfrey wrote:

I haven't forgotten this - I'm still totally bogged down with work on Yoshimi.
There is a *lot* going on!


  Just pushed r15492 which was a memory access issue related to 
deleting segments and the segment parameter box.  This may have fixed 
the problem.


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Will Godfrey]

On Wed, 17 Apr 2019 09:42:42 -0400
Ted Felix  wrote:

>On 4/8/19 3:56 AM, Will Godfrey wrote:
>> I'm afraid I've still not had time to do anything with this - not fired up 
>> the
>> DAW at all lately - to much other stuff getting in the way :(  
>
>   I've updated the instructions for getting a stack trace on the wiki.
>
>https://www.rosegardenmusic.com/wiki/frequently_asked_questions#how_to_get_a_stack_trace_for_a_crash
>
>   I missed a few details before.  You'll need:
>
>1. A binary built with debugging turned on.
>2. ulimit set so that a core dump can be created.
>3. Run from command line.
>4. Use gdb's "bt" command to get the backtrace out of the core file.
>
>   I've been completely unable to reproduce this crash.  The core dump 
>and backtrace will be critical to tracking this down and fixing.
>
>Ted.

I'll try and find time this coming week - trouble is I've been really up to my
eyeballs lately :(

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 4/8/19 3:56 AM, Will Godfrey wrote:

I'm afraid I've still not had time to do anything with this - not fired up the
DAW at all lately - to much other stuff getting in the way :(


  I've updated the instructions for getting a stack trace on the wiki.

https://www.rosegardenmusic.com/wiki/frequently_asked_questions#how_to_get_a_stack_trace_for_a_crash

  I missed a few details before.  You'll need:

1. A binary built with debugging turned on.
2. ulimit set so that a core dump can be created.
3. Run from command line.
4. Use gdb's "bt" command to get the backtrace out of the core file.

  I've been completely unable to reproduce this crash.  The core dump 
and backtrace will be critical to tracking this down and fixing.


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[david]

Yes. Remembering also that even a good power supply running too long on 
a below-normal voltage source can be damaged by that over time.


Also keep in mind the mortal enemy of all computers and electronics: 
heat. Over time, that bakes the life out of electronic components like 
capacitors and power supplies.


Plus inserting and removing connections (such as PS cables, drive 
cables, etc) and changing components (such as cards and memory) can 
fatigue circuit boards around those points, eventually leading to failures.


I personally am always angry when electronics die. They're supposed to 
last forever! ;)


On 4/9/19 9:52 PM, Sami Jumppanen wrote:
For dying computer owners: before upgrading anything, use a known good 
PSU. I just lost a week worth of late nights trying to set up a new 
computer. I deducted that it was the PSU on the nearly functional 
computer that had caused CPU failure, by having something else than DC 
voltage. On the average the voltages were a bit low, but contained AC 
spiking.


That CPU didn't work with decent PSU (normal voltage levels) anymore, on 
any board I tried - the other boards didn't even boot into BIOS. I had 
three PSUs to test and always it was the old one that made the old board 
work so well that I could install Ubuntu Studio and upgrade it even 
once, but then I got jams and interrupted HDD writes.


I postponed indefinitely testing the other boards with another CPU 
because I had an older computer that had no problems - and still runs fine.


PSUs are dangerous! :)


ke 10. huhtik. 2019 klo 9.33 david (gn...@hawaii.rr.com 
) kirjoitti:


On 4/8/19 10:37 PM, D. Michael McIntyre wrote:
 > I haven't tested the timing thing either.  My computer keeps
freezing.
 > When I reboot it, I'm stuck at 1024x768, and the only way I have
managed
 > to make progress on that issue is to reinstall Kubuntu.  I've been
 > around that bush three times now, and I have grown tired of the
 > aggravation.  This machine is rather old, and I suspect the
hardware is
 > just flaking out.

My main desktop/server, also rather old, finally died. The onboard
video
went first. Then the machine just quit coming up at all.

The backup desktop/server still works. It's "newer" than the dead main
one, but has some peculiarities. For example, the BIOS doesn't
recognize
the presence of a keyboard unless it's plugged into the PS/2 port. Once
Linux comes up, it works just fine with my USB keyboard, so it's not a
problem with the keyboard.

Hoping to replace the guts with a Ryzen 7 motherboard. Could use the
speed for other things I do.

 > Either that, or Kubuntu 18.04 is less stable than its
 > many predecessors, which I find difficult to imagine.  Still, the
 > trouble started when I upgraded from 16.04, so that may be the
culprit.
 > So many things have changed it's more trouble than it's worth to
me to
 > try to keep track of any of it.

Well, in my experience, upgrading from Ubuntu 16 to Kubuntu 18 was a
big
hairy problem. For instance, after upgrade and restart, no working
network. I reinstalled fresh.

 > I had gotten in the habit of just buying a retail desktop PC off the
 > shelf, but you can't buy those anymore.  Getting Linux onto a retail
 > Windows PC is a huge pain in the ass anyway.

In my experience, only if you want to keep Windows around.

 > I guess I have to go
 > figure out what components to buy and roll my own like it's
1995.  I am
 > not looking forward to this at all.
Well, putting Linux on my wife's retail laptop PC that came with
Windows
on it worked just fine. Dell, Lenovo, HP apparently work. Only
ointment-infestation I can imagine might be support for fancy new
graphics cards?

At the moment, I like Asus motherboards. My wife's laptop and the mobo
in my backup desktop machine just work with Linux.

Since my present laptop experiences a few issues (doesn't think it
has a
battery, takes 3-5 minutes before the built-in display shows the POST
screen) due to having been dropped a few times, I'm thinking of jumping
to a Dell 15"/4K HDR display and Intel processor.

Looking for hardware that won't leave me crawling when I move to a
42mpixel camera and get to start making big panoramas.


--
David W. Jones
gn...@hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[D. Michael McIntyre]

On 4/10/19 3:52 AM, Sami Jumppanen wrote:

For dying computer owners: before upgrading anything, use a known good 
PSU.


That's a thought.  Just because I put a good PSU in a couple years ago 
doesn't mean the PSU isn't the issue now.  I don't remember what it is, 
but it's a ridiculously high grade component for this retail Walmart 
special.


--
D. Michael McIntyre


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[D. Michael McIntyre]

On 4/10/19 8:54 AM, rhkra...@gmail.com wrote:


I guess you mean a retail desktop without Windows off the shelf?


No, I mean a retail desktop anything off the shelf.  Nobody stocks them 
anymore.  It's all laptops and all-in-ones, peppered with the occasional 
ridiculous "gaming PC" that has crap specs at a high price with some red 
LEDs thrown in to make it look, I don't know, gamey.


--
D. Michael McIntyre


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[D. Michael McIntyre]

On 4/9/19 4:55 AM, Will Godfrey wrote:


Is there nothing like that where you are?


There is, but my father made a scene, and I'm not welcome there anymore. 
 They wouldn't take back a laptop they ordered for Dad, so he said, 
"Can I smash it on the counter then?"  They gave the go ahead, and he 
did.  Smashed the laptop, smashed the glass display case.  I'm 
overwhelmed with pride.


--
D. Michael McIntyre


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[D. Michael McIntyre]

On 4/9/19 7:49 AM, Sami Jumppanen wrote:


Out of curiosity, what is the problem?


I have no idea.  I'm in the middle of typing, and then the computer is 
just frozen rock solid.  When I get it back up, there's a good chance 
X11 or Weyland or whatever I'm using will be broken to a point where I 
can't figure out how to fix it.  In my current iteration of this, KDE 
boots with my session apps in a void, and then about five minutes later 
it finally throws the rest of my desktop around those windows.  There 
are no clues in the logs.  Damn if I know, basically.


--
D. Michael McIntyre


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[rhkramer]

On Tuesday, April 09, 2019 04:37:15 AM D. Michael McIntyre wrote:
> I had gotten in the habit of just buying a retail desktop PC off the
> shelf, but you can't buy those anymore.  

I guess you mean a retail desktop without Windows off the shelf?




___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 4/8/19 3:56 AM, Will Godfrey wrote:

My guess is that the delete
operation has left an invalid pointer somewhere.


  There are definitely bad pointers left around in the code. 
Intentionally.  (!?)  See the comments in 
SequenceManager::segmentDeleted().  If, as I explain there, the memory 
manager decides to hand you a pointer to the same block of memory again 
(a very reasonable possibility), then we've got a crash in the making. 
This might explain why some see it and some don't.  Different memory 
manager behavior.



However, if *immediately* after doing the delete, I save the file it seems to
sort everything out and is then OK.


  Save might be scrambling the memory manager enough that the same 
pointer address doesn't get reused.


  I'll see if I can reproduce this given all the new info.  It would be 
nice to have confirmation that SequenceManager is the culprit.  If I 
can't reproduce, I'll have a closer look at fixing all the intentional 
dangling pointers.  It's something that needs to be done anyway.


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Sami Jumppanen]

For dying computer owners: before upgrading anything, use a known good PSU.
I just lost a week worth of late nights trying to set up a new computer. I
deducted that it was the PSU on the nearly functional computer that had
caused CPU failure, by having something else than DC voltage. On the
average the voltages were a bit low, but contained AC spiking.

That CPU didn't work with decent PSU (normal voltage levels) anymore, on
any board I tried - the other boards didn't even boot into BIOS. I had
three PSUs to test and always it was the old one that made the old board
work so well that I could install Ubuntu Studio and upgrade it even once,
but then I got jams and interrupted HDD writes.

I postponed indefinitely testing the other boards with another CPU because
I had an older computer that had no problems - and still runs fine.

PSUs are dangerous! :)


ke 10. huhtik. 2019 klo 9.33 david (gn...@hawaii.rr.com) kirjoitti:

> On 4/8/19 10:37 PM, D. Michael McIntyre wrote:
> > I haven't tested the timing thing either.  My computer keeps freezing.
> > When I reboot it, I'm stuck at 1024x768, and the only way I have managed
> > to make progress on that issue is to reinstall Kubuntu.  I've been
> > around that bush three times now, and I have grown tired of the
> > aggravation.  This machine is rather old, and I suspect the hardware is
> > just flaking out.
>
> My main desktop/server, also rather old, finally died. The onboard video
> went first. Then the machine just quit coming up at all.
>
> The backup desktop/server still works. It's "newer" than the dead main
> one, but has some peculiarities. For example, the BIOS doesn't recognize
> the presence of a keyboard unless it's plugged into the PS/2 port. Once
> Linux comes up, it works just fine with my USB keyboard, so it's not a
> problem with the keyboard.
>
> Hoping to replace the guts with a Ryzen 7 motherboard. Could use the
> speed for other things I do.
>
> > Either that, or Kubuntu 18.04 is less stable than its
> > many predecessors, which I find difficult to imagine.  Still, the
> > trouble started when I upgraded from 16.04, so that may be the culprit.
> > So many things have changed it's more trouble than it's worth to me to
> > try to keep track of any of it.
>
> Well, in my experience, upgrading from Ubuntu 16 to Kubuntu 18 was a big
> hairy problem. For instance, after upgrade and restart, no working
> network. I reinstalled fresh.
>
> > I had gotten in the habit of just buying a retail desktop PC off the
> > shelf, but you can't buy those anymore.  Getting Linux onto a retail
> > Windows PC is a huge pain in the ass anyway.
>
> In my experience, only if you want to keep Windows around.
>
> > I guess I have to go
> > figure out what components to buy and roll my own like it's 1995.  I am
> > not looking forward to this at all.
> Well, putting Linux on my wife's retail laptop PC that came with Windows
> on it worked just fine. Dell, Lenovo, HP apparently work. Only
> ointment-infestation I can imagine might be support for fancy new
> graphics cards?
>
> At the moment, I like Asus motherboards. My wife's laptop and the mobo
> in my backup desktop machine just work with Linux.
>
> Since my present laptop experiences a few issues (doesn't think it has a
> battery, takes 3-5 minutes before the built-in display shows the POST
> screen) due to having been dropped a few times, I'm thinking of jumping
> to a Dell 15"/4K HDR display and Intel processor.
>
> Looking for hardware that won't leave me crawling when I move to a
> 42mpixel camera and get to start making big panoramas.
>
> --
> David W. Jones
> gn...@hawaii.rr.com
> authenticity, honesty, community
> http://dancingtreefrog.com
>
>
> ___
> Rosegarden-user mailing list
> Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
> https://lists.sourceforge.net/lists/listinfo/rosegarden-user
>


-- 
Sami Jumppanen
http://netti.nic.fi/some-e
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[david]

On 4/7/19 9:56 PM, Will Godfrey wrote:

On Thu, 28 Mar 2019 22:18:27 -0400
Ted Felix  wrote:


On 3/28/19 6:24 PM, Will Godfrey wrote:

Not forgotten this - just busy firefighting yoshimi :(


   No problem.  I've been working my way through everything valgrind is
showing me to see if I can find your crashes.  Nothing yet.  Lots of
memory leaks, but nothing crashy.


I'm afraid I've still not had time to do anything with this - not fired up the
DAW at all lately - to much other stuff getting in the way :(

However, I think I can pin down the sequence of events a bit better.

I'll have several verses on a recorded over 3-4 tracks, and decide I want to
add a counter melody to verse 2. I'll actually start the recording a verse
ahead and begin to play near the end of that verse to get in the grove and timed
right, going on into the wanted verse.

When I stop, I'll wind back, cut the track by time, and delete the unwanted
preamble. Next I'll want to go into the 'good' bit to do some fine tuning, and
that's is the point when the crash happens. My guess is that the delete
operation has left an invalid pointer somewhere.


Mine would happen when I cut off the end of a segment and deleted it, 
then tried to edit the rest of the segment.



However, if *immediately* after doing the delete, I save the file it seems to
sort everything out and is then OK.


Hmm, never tried that. My general practice with saving things is, "If 
something got messed up, don't risk saving something mangled over the 
'good' version I had before!" :)


--
David W. Jones
gn...@hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[David W. Jones]

On April 9, 2019 1:49:22 AM HST, Sami Jumppanen  
wrote:
> Out of curiosity, what is the problem?
> 
> 
> > >I had gotten in the habit of just buying a retail desktop PC off
> the
> > >shelf, but you can't buy those anymore.  Getting Linux onto a
> retail
> > >Windows PC is a huge pain in the ass anyway.  I guess I have to go
> > >figure out what components to buy and roll my own like it's 1995. 
> I am
> > >not looking forward to this at all.
> >
> 
> - Buy a computer
> - Put in another hard drive, I bought a cheap SSD
> - (Unplug the supplied Windows drive if you're not going to use it for
> now)
> - Install Linux
> 
> I did find the new Linux installer (with Ubuntu Studio) a bit strange.
> It
> seemed all logical with empty HDD, but when having another HDD in the
> computer with Ubuntu Studio, the installer says there were no other OS
> found and logical volume management options and all the talk of
> deletion
> look strange or scary :D
> 
> I am using now a 2006 era dual core and everything looks ok. Maybe the
> newest computers bring new problems?
> 
> Sami

My wife's Asus is 2014 or 2015. Booted it from a USB stick, got to see all the 
lovely partitions Window 10 scattered all over the drive. Xubuntu 18.04 
installed over it with no problems.

That wasn't Ubuntu Studio. Maybe they're behind the times?

--
David W. Jones
gn...@hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com

Sent from my Android device with F/LOSS K-9 Mail.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[david]

On 4/8/19 10:37 PM, D. Michael McIntyre wrote:
I haven't tested the timing thing either.  My computer keeps freezing. 
When I reboot it, I'm stuck at 1024x768, and the only way I have managed 
to make progress on that issue is to reinstall Kubuntu.  I've been 
around that bush three times now, and I have grown tired of the 
aggravation.  This machine is rather old, and I suspect the hardware is 
just flaking out.


My main desktop/server, also rather old, finally died. The onboard video 
went first. Then the machine just quit coming up at all.


The backup desktop/server still works. It's "newer" than the dead main 
one, but has some peculiarities. For example, the BIOS doesn't recognize 
the presence of a keyboard unless it's plugged into the PS/2 port. Once 
Linux comes up, it works just fine with my USB keyboard, so it's not a 
problem with the keyboard.


Hoping to replace the guts with a Ryzen 7 motherboard. Could use the 
speed for other things I do.


Either that, or Kubuntu 18.04 is less stable than its 
many predecessors, which I find difficult to imagine.  Still, the 
trouble started when I upgraded from 16.04, so that may be the culprit. 
So many things have changed it's more trouble than it's worth to me to 
try to keep track of any of it.


Well, in my experience, upgrading from Ubuntu 16 to Kubuntu 18 was a big 
hairy problem. For instance, after upgrade and restart, no working 
network. I reinstalled fresh.


I had gotten in the habit of just buying a retail desktop PC off the 
shelf, but you can't buy those anymore.  Getting Linux onto a retail 
Windows PC is a huge pain in the ass anyway.


In my experience, only if you want to keep Windows around.

I guess I have to go 
figure out what components to buy and roll my own like it's 1995.  I am 
not looking forward to this at all.
Well, putting Linux on my wife's retail laptop PC that came with Windows 
on it worked just fine. Dell, Lenovo, HP apparently work. Only 
ointment-infestation I can imagine might be support for fancy new 
graphics cards?


At the moment, I like Asus motherboards. My wife's laptop and the mobo 
in my backup desktop machine just work with Linux.


Since my present laptop experiences a few issues (doesn't think it has a 
battery, takes 3-5 minutes before the built-in display shows the POST 
screen) due to having been dropped a few times, I'm thinking of jumping 
to a Dell 15"/4K HDR display and Intel processor.


Looking for hardware that won't leave me crawling when I move to a 
42mpixel camera and get to start making big panoramas.


--
David W. Jones
gn...@hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Sami Jumppanen]

Out of curiosity, what is the problem?


> >I had gotten in the habit of just buying a retail desktop PC off the
> >shelf, but you can't buy those anymore.  Getting Linux onto a retail
> >Windows PC is a huge pain in the ass anyway.  I guess I have to go
> >figure out what components to buy and roll my own like it's 1995.  I am
> >not looking forward to this at all.
>

- Buy a computer
- Put in another hard drive, I bought a cheap SSD
- (Unplug the supplied Windows drive if you're not going to use it for now)
- Install Linux

I did find the new Linux installer (with Ubuntu Studio) a bit strange. It
seemed all logical with empty HDD, but when having another HDD in the
computer with Ubuntu Studio, the installer says there were no other OS
found and logical volume management options and all the talk of deletion
look strange or scary :D

I am using now a 2006 era dual core and everything looks ok. Maybe the
newest computers bring new problems?

Sami
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[D. Michael McIntyre]

I haven't tested the timing thing either.  My computer keeps freezing. 
When I reboot it, I'm stuck at 1024x768, and the only way I have managed 
to make progress on that issue is to reinstall Kubuntu.  I've been 
around that bush three times now, and I have grown tired of the 
aggravation.  This machine is rather old, and I suspect the hardware is 
just flaking out.  Either that, or Kubuntu 18.04 is less stable than its 
many predecessors, which I find difficult to imagine.  Still, the 
trouble started when I upgraded from 16.04, so that may be the culprit. 
So many things have changed it's more trouble than it's worth to me to 
try to keep track of any of it.


I had gotten in the habit of just buying a retail desktop PC off the 
shelf, but you can't buy those anymore.  Getting Linux onto a retail 
Windows PC is a huge pain in the ass anyway.  I guess I have to go 
figure out what components to buy and roll my own like it's 1995.  I am 
not looking forward to this at all.




--
D. Michael McIntyre


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Will Godfrey]

On Thu, 28 Mar 2019 22:18:27 -0400
Ted Felix  wrote:

>On 3/28/19 6:24 PM, Will Godfrey wrote:
>> Not forgotten this - just busy firefighting yoshimi :(  
>
>   No problem.  I've been working my way through everything valgrind is 
>showing me to see if I can find your crashes.  Nothing yet.  Lots of 
>memory leaks, but nothing crashy.

I'm afraid I've still not had time to do anything with this - not fired up the
DAW at all lately - to much other stuff getting in the way :(

However, I think I can pin down the sequence of events a bit better.

I'll have several verses on a recorded over 3-4 tracks, and decide I want to
add a counter melody to verse 2. I'll actually start the recording a verse
ahead and begin to play near the end of that verse to get in the grove and timed
right, going on into the wanted verse.

When I stop, I'll wind back, cut the track by time, and delete the unwanted
preamble. Next I'll want to go into the 'good' bit to do some fine tuning, and
that's is the point when the crash happens. My guess is that the delete
operation has left an invalid pointer somewhere.

However, if *immediately* after doing the delete, I save the file it seems to
sort everything out and is then OK.

HTH

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[david]

On 3/28/19 4:18 PM, Ted Felix wrote:

On 3/28/19 6:24 PM, Will Godfrey wrote:

Not forgotten this - just busy firefighting yoshimi :(


   No problem.  I've been working my way through everything valgrind is 
showing me to see if I can find your crashes.  Nothing yet.  Lots of 
memory leaks, but nothing crashy.


It was a few years ago, but I had the same problem: record a track, 
split the track apart somewhere (typically followed by deleting the 
unwanted part of the track), then play. RG would just vanish away. Never 
tried starting from command line to see if anything was being reported.


--
David W. Jones
gn...@hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 3/28/19 6:24 PM, Will Godfrey wrote:

Not forgotten this - just busy firefighting yoshimi :(


  No problem.  I've been working my way through everything valgrind is 
showing me to see if I can find your crashes.  Nothing yet.  Lots of 
memory leaks, but nothing crashy.


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Will Godfrey]

On Sun, 24 Mar 2019 21:30:31 -0400
Ted Felix  wrote:

>On 3/23/19 5:27 AM, Will Godfrey wrote:
>> Typically 3-4 minutes 8-12 tracks MIDI only.
>> 
>> Most commonly happens when splitting a track then deleting part of it. Actual
>> crash usually happens next time I go to 'Play'.  
>
>   Ok, thanks.  I'll see if I can reproduce with these sorts of edits.
>
>> Had a look at gdb in the past (ran away screaming) couldn't make a lot of 
>> sense
>> out of the information it presented.  
>
>   All I need is the backtrace which can be had by issuing the "bt" 
>command in gdb against a core dump.  Might need to do some configuring 
>to get your system to generate core dumps for rg.  E.g., run rg like 
>this from the command line:
>
>$ ulimit -c unlimited
>$ rosegarden
>
>   Then once you've got a core dump:
>
>$ gdb rosegarden core
>
>   And finally, use the "bt" (backtrace) command.
>
>(gdb) bt
>
>   Just need to copy/paste the output of the backtrace for me and I can 
>do the rest.  Might not pinpoint the exact problem, but will certainly 
>provide a data point to work with.
>
>> Some time back we had a segfault in
>> Yoshimi - gdb was no help at all and we found the problem by simply undoing
>> commits till we found the one where it changed :(  
>
>   A bisect is oftentimes a very useful tool in these situations.  But 
>only if it's an easy-to-reproduce problem.  Running gdb on a core dump 
>is best for intermittent or harder to reproduce issues.
>
>Ted.

OK, I'll have a go at that over the next day or two - bit busy at the mo.

P.S.
git bisect was what we used :)

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 3/23/19 5:27 AM, Will Godfrey wrote:

Typically 3-4 minutes 8-12 tracks MIDI only.

Most commonly happens when splitting a track then deleting part of it. Actual
crash usually happens next time I go to 'Play'.


  Ok, thanks.  I'll see if I can reproduce with these sorts of edits.


Had a look at gdb in the past (ran away screaming) couldn't make a lot of sense
out of the information it presented.


  All I need is the backtrace which can be had by issuing the "bt" 
command in gdb against a core dump.  Might need to do some configuring 
to get your system to generate core dumps for rg.  E.g., run rg like 
this from the command line:


$ ulimit -c unlimited
$ rosegarden

  Then once you've got a core dump:

$ gdb rosegarden core

  And finally, use the "bt" (backtrace) command.

(gdb) bt

  Just need to copy/paste the output of the backtrace for me and I can 
do the rest.  Might not pinpoint the exact problem, but will certainly 
provide a data point to work with.



Some time back we had a segfault in
Yoshimi - gdb was no help at all and we found the problem by simply undoing
commits till we found the one where it changed :(


  A bisect is oftentimes a very useful tool in these situations.  But 
only if it's an easy-to-reproduce problem.  Running gdb on a core dump 
is best for intermittent or harder to reproduce issues.


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[D. Michael McIntyre]

On 3/23/19 5:27 AM, Will Godfrey wrote:


Yoshimi - gdb was no help at all and we found the problem by simply undoing
commits till we found the one where it changed :(


Some bug has existed in some form since Richard Bown implemented track 
deletion, so we'd have to roll back to the commit before deleting tracks 
was possible.  That's a few revisions. lol


I haven't experienced a track deletion related crash in forever, which 
means you have a different usage habit.  You're probably going to have 
to be the one to reproduce and document it.  If you can't, well, I'm no 
good with any of that stuff either, so don't feel bad.


--
D. Michael McIntyre


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Will Godfrey]

On Wed, 20 Mar 2019 18:07:43 -0400
Ted Felix  wrote:
>
>   First would be to characterize the editing you are doing.  Just a 
>list of the sorts of things, like track deletes that seem to cause the 
>issue.  Also, how big are the compositions?  Number of segments, tracks. 
>  How many minutes?  Does it happen even if you never play the composition?

Sorry to be slow responding - been busy!

Typically 3-4 minutes 8-12 tracks MIDI only.

Most commonly happens when splitting a track then deleting part of it. Actual
crash usually happens next time I go to 'Play'.

>   Next, if you can analyze the core dumps for me when you crash, that 
>could be a big help.  Though sometimes with memory issues, the stack 
>traces are useless.  Have you worked with gdb and core dumps before?
>
>Ted.

Had a look at gdb in the past (ran away screaming) couldn't make a lot of sense
out of the information it presented. Some time back we had a segfault in
Yoshimi - gdb was no help at all and we found the problem by simply undoing
commits till we found the one where it changed :( 

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes :(

[Ted Felix]

On 3/20/19 4:53 PM, Will Godfrey wrote:

There seems to be a memory allocation problem that is triggered by deleting
tracks/sections. This has been there for quite a long time, but it has only
been a real problem for me lately because I'm doing a lot of quite heavy
editing, and have to resort to doing very frequent saves.


  I recently saw a crash when deleting segments related to the undo 
history.  Couldn't reproduce it, though.  And it only happened on exit.



I'm not certain, but I get the impression that the very fact of doing these
saves seems to reduce the likelihood of the crashes.


  Memory issues are always tricky and can easily be exacerbated or 
mitigated by almost anything.



Currently using build 15419, but as I said, it's been there a long time - years.


  I'm not surprised.  Last time I ran valgrind against rg there were 
too many problems to even begin to track them all down.  I gave up 
because no one was complaining.  That appears to have changed.


  There are a number of ways you can help me track these down.

  First would be to characterize the editing you are doing.  Just a 
list of the sorts of things, like track deletes that seem to cause the 
issue.  Also, how big are the compositions?  Number of segments, tracks. 
 How many minutes?  Does it happen even if you never play the composition?


  Next, if you can analyze the core dumps for me when you crash, that 
could be a big help.  Though sometimes with memory issues, the stack 
traces are useless.  Have you worked with gdb and core dumps before?


Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes when resizing audio segment

[Peter Desjardins]

On Fri, Jun 24, 2011 at 8:16 AM, Chris Cannam
can...@all-day-breakfast.com wrote:

 I had a quick try and couldn't reproduce this.  Can you try running
 Rosegarden under a debugger and asking for a stack trace?

 i.e.

 $ gdb /usr/bin/rosegarden                 (or wherever it lives)
 (gdb) run            (then use it until it crashes)
 (gdb) thread apply all bt

 and send the output of that.  It probably won't be very useful but
 might give a hint.

Hi. I started Rosegarden through gdb as you suggested. This time,
Rosegarden didn't immediately quit when I held down Shift and used the
resize tool. The Rosegarden UI remained visible but became
unresponsive. When other windows moved in front of it, Rosegarden no
longer refreshed the display and left black areas on the screen. I
forced it to quit.

Here's the output. I'm going to try removing the version that came
along with Ubuntu and install the latest Rosegarden release. I'm
surprised that Ubuntu is a major version behind here.

Thanks for your help.

Starting program: /usr/bin/rosegarden
[Thread debugging using libthread_db enabled]
Setting graphics system for Qt 4.5+ to: native
Thorn - 1
System Locale: en_US
Qt translations path: /usr/share/qt4/translations
Qt translations not loaded.
RG Translation: trying to load :locale/en_US
RG Translations loaded successfully.
Loaded application icon rg-rwb-rose3-16x16
Loaded application icon rg-rwb-rose3-32x32
Loaded application icon rg-rwb-rose3-48x48
Loaded application icon rg-rwb-rose3-64x64
Loaded application icon rg-rwb-rose3-128x128
NOTE: Found stylesheet at :/rosegarden.qss, applying it
QSettings::endGroup: No matching beginGroup()
[New Thread 0x7fffe714e700 (LWP 5434)]
AlsaDriver::AlsaDriver [begin]
Rosegarden 10.02 - AlsaDriver [ALSA library version 1.0.22, module
version 1.0.20, kernel version 2.6.31-11-rt]
[New Thread 0x7fffe694d700 (LWP 5435)]
[New Thread 0x7fffe614c700 (LWP 5436)]
PluginFactory::instance(dssi): creating new DSSIPluginFactory
LADSPAPluginFactory::discoverPlugins - discovering plugins; path is
[/home/peter/.dssi] [/usr/local/lib/dssi] [/usr/lib/dssi]
Enhanced3DNow! detected
SSE2 detected
[New Thread 0x7fffe5892700 (LWP 5439)]
[New Thread 0x7fffe5791700 (LWP 5440)]

JackDriver::initialiseAudio - JACK sample rate = 44100Hz, buffer size = 256
JackDriver::initialiseAudio - creating disk thread
JackDriver::initialiseAudio - found 3 JACK physical outputs
JackDriver::initialiseAudio - connecting from rosegarden:master out
L to system:playback_1
JackDriver::initialiseAudio - connecting from rosegarden:master out
R to system:playback_2
JackDriver::initialiseAudio - found 3 JACK physical inputs
JackDriver::initialiseAudio - connecting from system:capture_1 to
rosegarden:record in 1 L
JackDriver::initialiseAudio - connecting from system:capture_2 to
rosegarden:record in 1 R
JackDriver::initialiseAudio - initialised JACK audio subsystem

  ALSA Client information:

14,0 - (Midi Through, Midi Through Port-0)  (DUPLEX) [ctype 
2,
ptype 655362, cap 99]

AlsaDriver::setCurrentTimer((auto))
Current timer set to system timer with timer checks
AlsaDriver::initialiseMidi -  initialised MIDI subsystem

Composition::getTrackById(0) - WARNING - track id not found, this is
probably a BUG src/base/Composition.cpp:1560
Available track ids are:
AlsaDriver::addDevice(0,0)
CREATED OUTPUT PORT 3:out 1 - unnamed for device 0
Renamed 130:3 to General MIDI Device
AlsaDriver::addDevice(0,1)
AlsaDriver::setRecordDevice: device 1, action 1
WARNING: AlsaDriver::renameDevice: Cannot find device 1 in port map
audio file manager emitting fake setValue(100)
RosegardenDocument::openDocument: Successfully opened document
/home/peter/.local/share/rosegarden/autoload/autoload.rg
LADSPAPluginFactory::discoverPlugins - done
Object::connect: No such slot
Rosegarden::AudioInstrumentParameterPanel::updateAllBoxes() in
src/gui/editors/parameters/AudioInstrumentParameterPanel.cpp:81
Object::connect:  (receiver name: 'Audio Instrument Parameter Panel')
PluginFactory::instance(ladspa): creating new LADSPAPluginFactory
LADSPAPluginFactory::discoverPlugins - discovering plugins; path is
[/home/peter/.ladspa] [/usr/local/lib/ladspa] [/usr/lib/ladspa]

  ALSA Client information:

14,0 - (Midi Through, Midi Through Port-0)  (DUPLEX) [ctype 
2,
ptype 655362, cap 99]

LADSPAPluginFactory::discoverPlugins - done
Object::connect: No such signal
Rosegarden::CompositionView::contentsMoving (int, int) in
src/gui/editors/segment/compositionview/SegmentMover.cpp:58
Object::disconnect: No such signal
Rosegarden::CompositionView::contentsMoving (int, int) in
src/gui/editors/segment/compositionview/SegmentMover.cpp:65
Object::connect: No such signal
Rosegarden::CompositionView::contentsMoving (int, int) in
src/gui/editors/segment/compositionview/SegmentPencil.cpp:61
AlsaDriver::setPlausibleConnection: connection like  requested for device 0
AlsaDriver::setPlausibleConnection: nothing suitable available

Re: [Rosegarden-user] Crashes when resizing audio segment

[D. Michael McIntyre]

On Friday, June 24, 2011, Peter Desjardins wrote:

 I can arrange segments and play them without problems. However, when I
 hold down shift and drag to resize, Rosegarden instantly crashes. I
 can repeat this every time.

 Is there another configuration problem that might be causing the crashes?

My life is pretty much all work and sleep right now, so I can't guess when 
I'll get a chance to look into this.  I've flagged this message, and I will 
look at this eventually.

It sounds like you've probably discovered a bug, and I doubt you're doing 
anything wrong.
-- 
D. Michael McIntyre

--
All the data continuously generated in your IT infrastructure contains a 
definitive record of customers, application performance, security 
threats, fraudulent activity and more. Splunk takes this data and makes 
sense of it. Business sense. IT sense. Common sense.. 
http://p.sf.net/sfu/splunk-d2d-c1
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user

Re: [Rosegarden-user] Crashes when resizing audio segment

[Peter Desjardins]

Thank you for your work on Rosegarden, it's a wonderful program. Let
me know if I can help when you do get a chance.

Is there a more stable way to do what I am trying to do? I want to
arrange a sequence of samples, snippets of music and drum sounds.
Would most people use a separate sequencing program and control it
from Rosegarden through MIDI?

Peter

On Fri, Jun 24, 2011 at 7:15 AM, D. Michael McIntyre
michael.mcint...@rosegardenmusic.com wrote:
 On Friday, June 24, 2011, Peter Desjardins wrote:

 I can arrange segments and play them without problems. However, when I
 hold down shift and drag to resize, Rosegarden instantly crashes. I
 can repeat this every time.

 Is there another configuration problem that might be causing the crashes?

 My life is pretty much all work and sleep right now, so I can't guess when
 I'll get a chance to look into this.  I've flagged this message, and I will
 look at this eventually.

 It sounds like you've probably discovered a bug, and I doubt you're doing
 anything wrong.
 --
 D. Michael McIntyre

 --
 All the data continuously generated in your IT infrastructure contains a
 definitive record of customers, application performance, security
 threats, fraudulent activity and more. Splunk takes this data and makes
 sense of it. Business sense. IT sense. Common sense..
 http://p.sf.net/sfu/splunk-d2d-c1
 ___
 Rosegarden-user mailing list
 Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
 https://lists.sourceforge.net/lists/listinfo/rosegarden-user


--
All the data continuously generated in your IT infrastructure contains a 
definitive record of customers, application performance, security 
threats, fraudulent activity and more. Splunk takes this data and makes 
sense of it. Business sense. IT sense. Common sense.. 
http://p.sf.net/sfu/splunk-d2d-c1
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user