[ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[jmerkey]

 Added local APIC and IOAPIC register support.  Removed all volatile 
 declarations from the code.  Instrumented barrier() compiler 
 directives.  streamlined mdb-list parser.   Added hard
 NMI interrupt capability to generate a hard NMI to a target 
 processor for debugging a hung processor via the APIC ASSERT_INT
 command.
 
 ISSUES:
 
 remote read of another processors local APIC registers has been
 instrumented, but I am having problems with it working on all 
 systems.   I have tried both physical and logical command modes 
 and I note that smptboot.c also relies on it for error reporting.
 
 At present, it is insufficiently tested and some other eyes and 
 testing would be helpful. 
 
ChangesLog:
2.6.27-rc5 1/29 mdb:  export ioapic read routines and detected ioapic count
 
 export ioapic read routines and detected ioapic count.
 

2.6.27-rc5 2/29 mdb:  export genapic and machine_restart functions
 
 export genapic and machine_restart functions
 

2.6.27-rc5 3/29 mdb:  add makefile for Merkey's Linux Kernel Debugger module
 
 add makefile for Merkey's Linux Kernel Debugger module
 

2.6.27-rc5 4/29 mdb:  add local Makefile for Merkey's Linux Kernel Debugger module
 
 add local Makefile for Merkey's Linux Kernel Debugger module
 

2.6.27-rc5 5/29 mdb:  add mdb-base.c architecure independent functions
 
 add mdb-base.c architecure independent functions
 

2.6.27-rc5 6/29 mdb:  add mdb-base.h architecure indepedent include
 
 add mdb-base.h architecure indepedent include
 

2.6.27-rc5 7/29 mdb:  add mdb.h architecure main include
 
 add mdb.h architecure main include
 

2.6.27-rc5 8/29 mdb:  add mdb-ia32-apic.c Intel x86 architecure functions
 
 add mdb-ia32-apic.c Intel x86 architecure functions.
 add support to dump local APIC and IOAPIC registers.  
 Instrument support for remote read of local apic registers.
 
 Add support for hardware NMI assertion for debugging a hard 
 hung processor.  
 
 

2.6.27-rc5 9/29 mdb:  add mdb-ia32.c Intel x86 architecure routines
 
 add mdb-ia32.c Intel x86 architecure routines
 
 remove ALL volatile declarations
 

2.6.27-rc5 10/29 mdb:  add mdb-ia32.h Intel x86 architecure includes
 
 add mdb-ia32.h Intel x86 architecure includes
 
 

2.6.27-rc5 11/29 mdb:  add mdb-ia32-proc.h Intel x86 architecure function includes
 
 add mdb-ia32-proc.h Intel x86 architecure function includes
 
 

2.6.27-rc5 12/29 mdb:  add mdb-ia32-support.c disassembler support
 
 add mdb-ia32-support.c disassembler support
 

2.6.27-rc5 13/29 mdb:  add mdb-keyboard.h keyboard function includes
 
 add mdb-keyboard.h keyboard function includes
 

2.6.27-rc5 14/29 mdb:  add mdb-list.c debugger parser functions
 
 add mdb-list.c debugger parser functions

2.6.27-rc5 15/29 mdb:  add mdb-list.h debugger parser function includes
 
 add mdb-list.h debugger parser function includes

2.6.27-rc5 16/29 mdb:  add mdb-logic.c math parser for Merkey's Linux Kernel Debugger
 
 add mdb-logic.c math parser for Merkey's Linux Kernel Debugger
 

2.6.27-rc5 17/29 mdb:  add mdb-main.c linux debugger entry functions
 
 add mdb-main.c linux debugger entry functions
 
 remove ALL volatile declarations
 

2.6.27-rc5 18/29 mdb:  add mdb-os.c operating system specific functions
 
 add mdb-os.c operating system specific functions
 

2.6.27-rc5 19/29 mdb:  add mdb-os.h operating system specific functions
 
 add mdb-os.h operating system specific functions
 

2.6.27-rc5 20/29 mdb:  add mdb-proc.h function prototype includes
 
 add mdb-proc.h function prototype includes
 

2.6.27-rc5 21/29 mdb:  add MAGIC_SYSRQ support for mdb to documentation
 
 add MAGIC_SYSRQ support for mdb to documentation
 

2.6.27-rc5 22/29 mdb:  export kmsg_redirect printk for console redirection 
 
 export kmsg_redirect printk for console redirection 
 

2.6.27-rc5 23/29 mdb:  export __kernel_text_address 
 
 export __kernel_text_address 
 

2.6.27-rc5 24/29 mdb:  add mdb_kallsyms function and export kallsyms lookup functions
 
 add mdb_kallsyms function and export kallsyms lookup functions
 

2.6.27-rc5 25/29 mdb:  add mdb_modules function 
 
 add mdb_modules function 
 

2.6.27-rc5 26/29 mdb:  add notify_die handler to system panic 
 
 add notify_die handler to system panic 
 

2.6.27-rc5 27/29 mdb:  export clocksource watchdog function
 
 export clocksource watchdog function
 

2.6.27-rc5 28/29 mdb:  add Kconfig.debug sections and documentation 
 
 add Kconfig.debug sections and documentation 
 
 updated help sections for MAGIC_SYSRQ keyboard support.
 Added X86 option restruction for X86 only features.
 

2.6.27-rc5 29/29 mdb:  add /debug directory to main Makefile 
 
 add /debug directory to main Makefile 
 
 TO-DO
 
 Add nested task gates to ia32 IDT for Linux to enable switching to a 
   good stack during exceptions.
 X86_64 disssembler support
 X86_64 register display support
 ia64 disassembler support.
 ia64 breakpoint support.
 ia64 register display support.
 

Release Site : http://www.wolfmountaingroup.org
FTP Address  : ftp://www.wolfmountaingroup.org/pub/mdb/mdb-2.6.27-rc5-ia32-08-28-08.patch

Re: [ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[Robin Holt]

Can this _PLEASE_ move to a different list?  I feel like these hourly
spam postings of 30 patches is really getting excessive.  If people are
interested in mdb, go to the mdb list or project page.

Thanks,
Robin


On Thu, Aug 28, 2008 at 11:47:42PM -0600, jmerkey@wolfmountaingroup.com wrote:
>  Added local APIC and IOAPIC register support.  Removed all volatile 
>  declarations from the code.  Instrumented barrier() compiler 
>  directives.  streamlined mdb-list parser.   Added hard
>  NMI interrupt capability to generate a hard NMI to a target 
>  processor for debugging a hung processor via the APIC ASSERT_INT
>  command.
>  
>  ISSUES:
>  
>  remote read of another processors local APIC registers has been
>  instrumented, but I am having problems with it working on all 
>  systems.   I have tried both physical and logical command modes 
>  and I note that smptboot.c also relies on it for error reporting.
>  
>  At present, it is insufficiently tested and some other eyes and 
>  testing would be helpful. 
>  
> ChangesLog:
> 2.6.27-rc5 1/29 mdb:  export ioapic read routines and detected ioapic count
>  
>  export ioapic read routines and detected ioapic count.
>  
> 
> 2.6.27-rc5 2/29 mdb:  export genapic and machine_restart functions
>  
>  export genapic and machine_restart functions
>  
> 
> 2.6.27-rc5 3/29 mdb:  add makefile for Merkey's Linux Kernel Debugger module
>  
>  add makefile for Merkey's Linux Kernel Debugger module
>  
> 
> 2.6.27-rc5 4/29 mdb:  add local Makefile for Merkey's Linux Kernel Debugger module
>  
>  add local Makefile for Merkey's Linux Kernel Debugger module
>  
> 
> 2.6.27-rc5 5/29 mdb:  add mdb-base.c architecure independent functions
>  
>  add mdb-base.c architecure independent functions
>  
> 
> 2.6.27-rc5 6/29 mdb:  add mdb-base.h architecure indepedent include
>  
>  add mdb-base.h architecure indepedent include
>  
> 
> 2.6.27-rc5 7/29 mdb:  add mdb.h architecure main include
>  
>  add mdb.h architecure main include
>  
> 
> 2.6.27-rc5 8/29 mdb:  add mdb-ia32-apic.c Intel x86 architecure functions
>  
>  add mdb-ia32-apic.c Intel x86 architecure functions.
>  add support to dump local APIC and IOAPIC registers.  
>  Instrument support for remote read of local apic registers.
>  
>  Add support for hardware NMI assertion for debugging a hard 
>  hung processor.  
>  
>  
> 
> 2.6.27-rc5 9/29 mdb:  add mdb-ia32.c Intel x86 architecure routines
>  
>  add mdb-ia32.c Intel x86 architecure routines
>  
>  remove ALL volatile declarations
>  
> 
> 2.6.27-rc5 10/29 mdb:  add mdb-ia32.h Intel x86 architecure includes
>  
>  add mdb-ia32.h Intel x86 architecure includes
>  
>  
> 
> 2.6.27-rc5 11/29 mdb:  add mdb-ia32-proc.h Intel x86 architecure function includes
>  
>  add mdb-ia32-proc.h Intel x86 architecure function includes
>  
>  
> 
> 2.6.27-rc5 12/29 mdb:  add mdb-ia32-support.c disassembler support
>  
>  add mdb-ia32-support.c disassembler support
>  
> 
> 2.6.27-rc5 13/29 mdb:  add mdb-keyboard.h keyboard function includes
>  
>  add mdb-keyboard.h keyboard function includes
>  
> 
> 2.6.27-rc5 14/29 mdb:  add mdb-list.c debugger parser functions
>  
>  add mdb-list.c debugger parser functions
> 
> 2.6.27-rc5 15/29 mdb:  add mdb-list.h debugger parser function includes
>  
>  add mdb-list.h debugger parser function includes
> 
> 2.6.27-rc5 16/29 mdb:  add mdb-logic.c math parser for Merkey's Linux Kernel Debugger
>  
>  add mdb-logic.c math parser for Merkey's Linux Kernel Debugger
>  
> 
> 2.6.27-rc5 17/29 mdb:  add mdb-main.c linux debugger entry functions
>  
>  add mdb-main.c linux debugger entry functions
>  
>  remove ALL volatile declarations
>  
> 
> 2.6.27-rc5 18/29 mdb:  add mdb-os.c operating system specific functions
>  
>  add mdb-os.c operating system specific functions
>  
> 
> 2.6.27-rc5 19/29 mdb:  add mdb-os.h operating system specific functions
>  
>  add mdb-os.h operating system specific functions
>  
> 
> 2.6.27-rc5 20/29 mdb:  add mdb-proc.h function prototype includes
>  
>  add mdb-proc.h function prototype includes
>  
> 
> 2.6.27-rc5 21/29 mdb:  add MAGIC_SYSRQ support for mdb to documentation
>  
>  add MAGIC_SYSRQ support for mdb to documentation
>  
> 
> 2.6.27-rc5 22/29 mdb:  export kmsg_redirect printk for console redirection 
>  
>  export kmsg_redirect printk for console redirection 
>  
> 
> 2.6.27-rc5 23/29 mdb:  export __kernel_text_address 
>  
>  export __kernel_text_address 
>  
> 
> 2.6.27-rc5 24/29 mdb:  add mdb_kallsyms function and export kallsyms lookup functions
>  
>  add mdb_kallsyms function and export kallsyms lookup functions
>  
> 
> 2.6.27-rc5 25/29 mdb:  add mdb_modules function 
>  
>  add mdb_modules function 
>  
> 
> 2.6.27-rc5 26/29 mdb:  add notify_die handler to system panic 
>  
>  add notify_die handler to system panic 
>  
> 
> 2.6.27-rc5 27/29 mdb:  export clocksource watchdog function
>  
>  export clocksource watchdog function
>  
> 
> 2.6.27-rc5 28/29 mdb:  add Kconfig.debug sections and documentation 
>  
>  add Kconfig.debug sections and documentation 
>  
>  updated help sections for MAGIC_SYSRQ keyboard support.
>  Added X86 option restruction for X86 only features.
>  
> 
> 2.6.27-rc5 29/29 mdb:  add /debug directory to main Makefile 
>  
>  add /debug directory to main Makefile 
>  
>  TO-DO
>  
>  Add nested task gates to ia32 IDT for Linux to enable switching to a 
>    good stack during exceptions.
>  X86_64 disssembler support
>  X86_64 register display support
>  ia64 disassembler support.
>  ia64 breakpoint support.
>  ia64 register display support.
>  
> 
> Release Site : http://www.wolfmountaingroup.org
> FTP Address  : ftp://www.wolfmountaingroup.org/pub/mdb/mdb-2.6.27-rc5-ia32-08-28-08.patch
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: [ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[jmerkey]

> Can this _PLEASE_ move to a different list?  I feel like these hourly
> spam postings of 30 patches is really getting excessive.  If people are
> interested in mdb, go to the mdb list or project page.
>
> Thanks,
> Robin

Every linux release is not hourly -- the last post was 7 days ago.   It
gets submitted for Andi and others to review the code, and it will be
submitted until Linus puts it into the kernel.   He needs to respond, not
one of his underlings.

I guess kdb downloads must be falling off or something.

Jeff

Re: [ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[jmerkey]

>
> Jeff, please post them all as a reply to a single, introductory message.
> That way, they get nicely grouped under that first introductory message,
> and
> anyone not interested can just delete the entire thread.
>
> If you're using git, git send-email can do it for you.
>


I will download and install git mail.  Actually, I wrote my own program to
create the patch series and submit it, but sounds like git is better.

I am getting a lot of emails back from people who are using mdb and
apparently have git email as well.

Thanks,

Jeff

Re: [ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[Henrique de Moraes Holschuh]

On Fri, 29 Aug 2008, jmerkey@wolfmountaingroup.com wrote:
> > Can this _PLEASE_ move to a different list?  I feel like these hourly
> > spam postings of 30 patches is really getting excessive.  If people are
> > interested in mdb, go to the mdb list or project page.
> >
> > Thanks,
> > Robin
> 
> Every linux release is not hourly -- the last post was 7 days ago.   It
> gets submitted for Andi and others to review the code, and it will be
> submitted until Linus puts it into the kernel.   He needs to respond, not
> one of his underlings.
> 
> I guess kdb downloads must be falling off or something.

Jeff, please post them all as a reply to a single, introductory message.
That way, they get nicely grouped under that first introductory message, and
anyone not interested can just delete the entire thread.

If you're using git, git send-email can do it for you.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Re: [ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[Peter Zijlstra]

On Fri, 2008-08-29 at 12:51 -0300, Henrique de Moraes Holschuh wrote:
> On Fri, 29 Aug 2008, jmerkey@wolfmountaingroup.com wrote:
> > > Can this _PLEASE_ move to a different list?  I feel like these hourly
> > > spam postings of 30 patches is really getting excessive.  If people are
> > > interested in mdb, go to the mdb list or project page.
> > >
> > > Thanks,
> > > Robin
> > 
> > Every linux release is not hourly -- the last post was 7 days ago.   It
> > gets submitted for Andi and others to review the code, and it will be
> > submitted until Linus puts it into the kernel.   He needs to respond, not
> > one of his underlings.
> > 
> > I guess kdb downloads must be falling off or something.
> 
> Jeff, please post them all as a reply to a single, introductory message.
> That way, they get nicely grouped under that first introductory message, and
> anyone not interested can just delete the entire thread.
> 
> If you're using git, git send-email can do it for you.

Can this git send-email thing sanely thread these days? It used to be
that each next email was threaded to the next, creating these insanely
deep threads.

quilt mail threads all subsequent emails from the first mail, not from
the previous - a mode I much prefer.

Re: [ANNOUNCE] mdb: Merkey's Linux Kernel Debugger 2.6.27-rc5 released

[Henrique de Moraes Holschuh]

On Sat, 30 Aug 2008, Peter Zijlstra wrote:
> > If you're using git, git send-email can do it for you.
> 
> Can this git send-email thing sanely thread these days? It used to be
> that each next email was threaded to the next, creating these insanely
> deep threads.

yes, use --no-chain-reply-to.  I hate deep threads as well.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh