mutex_unlock

mutex_unlock

[Justin Mattock]

Hello with testing out git(very cool); my first test was with
2.6.25-rc8-00194-g4cac04d ran vary smoothly;
then I decided to pull the latest git (2.6.25-rc8-00208-g7180c4c)and
see what I might find.
upon reboot the system starts up giving me this:

Starting up
Decompressing Linux Done
Booting the kernel
__   <-------blinking

I waited a few seconds or minutes but nothing;
after reading earlier posts about something with a mutex_unlock maybe
this was what I was experiencing.
when loading a live cd and recompiling the same kernel, I noticed
under kernel hacking;

RT Mutex debugging
Built in scriptable tester for rt-mutexes
Spinlock and rw-lock debugging
Mutex debugging basic checks
Lock debugging detect incorrect freeing of live locks
Lock debugging prove locking correctness
lock usage statistics
Lock dependency engine debugging
spinlock debugging sleep-inside-spinlock checking
Locking API boot-time self-tests

With not knowing what I was doing I chose yes to all of these options,
then reboot -f,
The system booted up properly,

Is there a way where I can find out if this was what was going on?
could this be something different?



-- 
Justin P. Mattock

Re: mutex_unlock

[Rafael J. Wysocki]

On Wednesday, 9 of April 2008, Justin Mattock wrote:
> Hello with testing out git(very cool); my first test was with
> 2.6.25-rc8-00194-g4cac04d ran vary smoothly;
> then I decided to pull the latest git (2.6.25-rc8-00208-g7180c4c)and
> see what I might find.
> upon reboot the system starts up giving me this:
> 
> Starting up
> Decompressing Linux Done
> Booting the kernel
> __   <-------blinking
> 
> I waited a few seconds or minutes but nothing;
> after reading earlier posts about something with a mutex_unlock maybe
> this was what I was experiencing.
> when loading a live cd and recompiling the same kernel, I noticed
> under kernel hacking;
> 
> RT Mutex debugging
> Built in scriptable tester for rt-mutexes
> Spinlock and rw-lock debugging
> Mutex debugging basic checks
> Lock debugging detect incorrect freeing of live locks
> Lock debugging prove locking correctness
> lock usage statistics
> Lock dependency engine debugging
> spinlock debugging sleep-inside-spinlock checking
> Locking API boot-time self-tests
> 
> With not knowing what I was doing I chose yes to all of these options,
> then reboot -f,
> The system booted up properly,
> 
> Is there a way where I can find out if this was what was going on?
> could this be something different?

Hm, interesting.

It looks like our locking debugging code may hide some issues ...

Thanks,
Rafael

Re: mutex_unlock

[Justin Mattock]

On Wed, Apr 9, 2008 at 8:40 PM, Rafael J. Wysocki <rjw@sisk.pl> wrote:
>
> On Wednesday, 9 of April 2008, Justin Mattock wrote:
>  > Hello with testing out git(very cool); my first test was with
>  > 2.6.25-rc8-00194-g4cac04d ran vary smoothly;
>  > then I decided to pull the latest git (2.6.25-rc8-00208-g7180c4c)and
>  > see what I might find.
>  > upon reboot the system starts up giving me this:
>  >
>  > Starting up
>  > Decompressing Linux Done
>  > Booting the kernel
>  > __   <-------blinking
>  >
>  > I waited a few seconds or minutes but nothing;
>  > after reading earlier posts about something with a mutex_unlock maybe
>  > this was what I was experiencing.
>  > when loading a live cd and recompiling the same kernel, I noticed
>  > under kernel hacking;
>  >
>  > RT Mutex debugging
>  > Built in scriptable tester for rt-mutexes
>  > Spinlock and rw-lock debugging
>  > Mutex debugging basic checks
>  > Lock debugging detect incorrect freeing of live locks
>  > Lock debugging prove locking correctness
>  > lock usage statistics
>  > Lock dependency engine debugging
>  > spinlock debugging sleep-inside-spinlock checking
>  > Locking API boot-time self-tests
>  >
>  > With not knowing what I was doing I chose yes to all of these options,
>  > then reboot -f,
>  > The system booted up properly,
>  >
>  > Is there a way where I can find out if this was what was going on?
>  > could this be something different?
>
>  Hm, interesting.
>
>  It looks like our locking debugging code may hide some issues ...
>
>  Thanks,
>  Rafael
>

Maybe; weird  having something like that happen.(maybe it was
something like what appletouch is doing i.g. could not do mode request
error. The problem  is it very seldom, every X amount of boots)  I can
go back in the kernel and see if I can find out what option really
caused the system to boot properly(hopefully). Also with suspend; I
update to the latest openGl to see if suspend works, but it seems to
be worst with the latest release, i.g. a few days ago I update xorg to
7.3 with sid, suspend would give me a blank screen but no reboot(I
should of left my mouse on the terminal, then once the blank screen
shows up issue dmesg > suspend), now with the latest update with
openGL(today) I receive a blank screen and a reboot.
Maybe openGl needs to have DRI enabled? in order for this to work...

-- 
Justin P. Mattock

Re: mutex_unlock

[Ingo Molnar]

* Justin Mattock <justinmattock@gmail.com> wrote:

> >  > 2.6.25-rc8-00194-g4cac04d ran vary smoothly;
> >  > then I decided to pull the latest git (2.6.25-rc8-00208-g7180c4c)and
> >  > see what I might find.
> >  > upon reboot the system starts up giving me this:
> >  >
> >  > Starting up
> >  > Decompressing Linux Done
> >  > Booting the kernel
> >  > __   <-------blinking
> >  >
> >  > I waited a few seconds or minutes but nothing;

to perhaps get more kernel messages about this, try the earlyprintk=vga 
boot option - does it produce any messages?

	Ingo

Re: mutex_unlock

[Justin Mattock]

On Fri, Apr 11, 2008 at 9:07 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
>  * Justin Mattock <justinmattock@gmail.com> wrote:
>
>  > >  > 2.6.25-rc8-00194-g4cac04d ran vary smoothly;
>  > >  > then I decided to pull the latest git (2.6.25-rc8-00208-g7180c4c)and
>  > >  > see what I might find.
>  > >  > upon reboot the system starts up giving me this:
>  > >  >
>  > >  > Starting up
>  > >  > Decompressing Linux Done
>  > >  > Booting the kernel
>  > >  > __   <-------blinking
>  > >  >
>  > >  > I waited a few seconds or minutes but nothing;
>
>  to perhaps get more kernel messages about this, try the earlyprintk=vga
>  boot option - does it produce any messages?
>
>         Ingo
>

Thanks for the info, I can try that and see if I receive any kind of messages.
regards;

-- 
Justin P. Mattock