Punctures, Loose headsets and a Bugcheck

It's been an interesting week as cycling goes.  Firstly, on Tuesday, the club ran one of its usual social meets at the Warrington Sports and Social club near Walton hall.  This week the theme was "Bike maintenance".  And given that my C2C adventure begins in two weeks I thought this would be a great opportunity to brush up on my current bike fixing skills.

This......is a bike.

The session was presented by one of my clubmates whom I only know by his NCC forum name.  So "MamilGrange" talked about fixing punctures, removing and replacing bottom brackets and tightening headsets. It lasted for about an hour and was informative and interesting.  Infact, given the material I wasn't nodding off at all.  Great lecture and after it had finished  I realised that my puncture removing skill, which I thought was actually really good was not very good and was infact one of his prime examples of how not to do it...ooops.
 

The XFactor auditions were becomming...tyresome

Other topics Grange covered were symptoms of failures, such as a screwed headset and chains that are about to snap and also what to do if your rear jocky pully assembly wound its way into your spokes.

 He also went into a lot of detail about the tools you should carry around with you. Thanks to him I've now ordered a few more goodies for my saddle bag.

Overall this was another enjoyable club social and I got the chance to meet some newbies who had just recently joined.  The club is getting bigger now and is attracting a lot more people....great stuff.

New delivery

On Wednesday my new panniers arrived.  I'm nearly there now with the kit needed for the my coast to coast tour and I'm really hoping that the weather is going to be kind. This week has been shit - weather wise that is - and I've been using the car more than I would have liked. Rain, the type that really gets you wet and a bracingly cold wind had put me off cycling in, however Wednesday and  today the weather changed and I finally got to work under my own steam..

According to the papers this weekend is supposed to be nice.  This will be great because on Sunday I'm going to ride my hybrid, fully loaded up, to Chester to get some mudguards.

This should give me an idea as to what to expect on the C2C and will give me some training on riding a loaded up bike.



Crashing

One of my other passions is gaming.  Even though I'm quickly approaching my 40th year, I still enjoy fighting aliens or roaming virtual dungeons either alone or online with my virtual buddies.  So this week I also ordered a copy of Deus Ex: Human Revolution for the PC.  I'd played the original about 11 years ago and loved every moment of it, so was really looking forward to this one.

I made sure all my drivers were up to date and installed it.  After a few minutes I was engaged in my first adventure..then...CRASH.  My machine hung......a few seconds later I was presented with bluescreen. 

WTF!!!!!

Windows 8 gives you a nicer way of saying "Your system is fucked"

I ignored it....rebooted and reloaded the game. An hour later...BANG!!!! another bluescreen.

Windows thankfully writes kernel minidumps to C:\Windows\Minidump when a bugcheck happens.  My day job as a software engineer involves a lot of kernel diagnosis so I decided that given I'd paid £20 for this game, I was going to figure out why the FUCK it was crashing. 

Firstly it had to be a driver because usermode apps don't cause bluescreens.  My guess was that the Nvidia drivers - which always crash - were the problem. So, loading up Windbg I opened up the minidump.

Buggered

The initial triage analysis yielded the following reason for the bluescreen..

WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
WHEA_ERROR_RECORD structure that describes the error conditon.
Arguments:
Arg1: 0000000000000000, Machine Check Exception
Arg2: fffffa80067b18f8, Address of the WHEA_ERROR_RECORD structure.
Arg3: 0000000000000000, High order 32-bits of the MCi_STATUS value.
Arg4: 0000000000000000, Low order 32-bits of the MCi_STATUS value.

Great..... An MCE (Machine Check Exception) meant that there was an actual hardware fault.  These are generated on certain class of cpus such as Intel and Amd 64bit cpus.  And as I'm running an Intel i7 950 on Windows 7 64bit, this is sort of normal....sort of. Basically as the CPU is executing code it runs checks on components that are part of the system as a whole.  If the math doesn't add up it halts further processing and tells the kernel to bugcheck.  This is what a bluescreen is and is a way to prevent further damage to your system.

So...

Using the address of the WHEA_ERROR_RECORD struct above in red I could actually look at what the error was

3: kd> !errrec fffffa80067b18f8
===============================================================================
Common Platform Error Record @ fffffa80067b18f8
-------------------------------------------------------------------------------
Record Id     : 01cc890ea9187ea3
Severity      : Fatal (1)
Length        : 928
Creator       : Microsoft
Notify Type   : Machine Check Exception
Timestamp     : 10/12/2011 18:42:18
Flags         : 0x00000002 PreviousError
===============================================================================
Section 0     : Processor Generic
-------------------------------------------------------------------------------
Descriptor    @ fffffa80067b1978
Section       @ fffffa80067b1a50
Offset        : 344
Length        : 192
Flags         : 0x00000001 Primary
Severity      : Fatal
Proc. Type    : x86/x64
Instr. Set    : x64
Error Type    : Micro-Architectural Error
Flags         : 0x00
CPU Version   : 0x00000000000106a5
Processor ID  : 0x0000000000000000
===============================================================================
Section 1     : x86/x64 Processor Specific
-------------------------------------------------------------------------------
Descriptor    @ fffffa80067b19c0
Section       @ fffffa80067b1b10
Offset        : 536
Length        : 128
Flags         : 0x00000000
Severity      : Fatal
Local APIC Id : 0x0000000000000000
CPU Id        : a5 06 01 00 00 08 10 00 - bd e3 98 00 ff fb eb bf
                00 00 00 00 00 00 00 00 - 00 00 00 00 00 00 00 00
                00 00 00 00 00 00 00 00 - 00 00 00 00 00 00 00 00
Proc. Info 0  @ fffffa80067b1b10
===============================================================================
Section 2     : x86/x64 MCA
-------------------------------------------------------------------------------
Descriptor    @ fffffa80067b1a08
Section       @ fffffa80067b1b90
Offset        : 664
Length        : 264
Flags         : 0x00000000
Severity      : Fatal
Error         : Internal timer (Proc 0 Bank 5)
  Status      : 0xbe00000000800400
  Address     : 0x000000000141b64f
  Misc.       : 0x0000000000000000

The last part details the Machine Check Architecture (MCA) record and this is where the CPU microcode stores the error. In this case it was an internal timer problem.  This is where my hardware crash diagnosis skills end and basically it could be down to anything.

Could this be down to my system being overclocked?  I'd not encountered any issues with Crysis 2 or other games.  Given that other people were also encountering issues with the game I concluded that it was Nvidia's fault.  After all, they are bastards and given that I had spent a fortune on one of their graphics cards you would expect the drivers to work..........

God I really hate PC's sometimes...and today was one of them!