Page 1 of 2

bluescreen

Posted: Tue Apr 30, 2013 1:59 pm
by Mikemk
Type 2^34359738368 into a calculator on a 32-bit computer.

Re: instant bluescreen

Posted: Tue Apr 30, 2013 7:48 pm
by ASMMan
Are you talking about Microsoft Windows's calc? I'm not on MS Windows currently; in my current system it just given an error message.

Re: instant bluescreen

Posted: Wed May 01, 2013 2:55 am
by AJ
On a fully patched Windows XP box:
Windows XP Calculator with Result of 2^34359738368
Windows XP Calculator with Result of 2^34359738368
Cheers,
Adam

Re: instant bluescreen

Posted: Wed May 01, 2013 8:19 am
by Brendan
Hi,
m12 wrote:Type 2^34359738368 into a calculator on a 32-bit computer.
I tried it on Kcalc (running on 64-bit Gentoo); and it caused the calculator to crash (segmentation fault in "__gmpn_copyi()"). :)

I suspect an out of memory problem - e.g. someone forgetting to test if "malloc()" returned NULL (or possibly someone using C++, where "new" exists to ensure software crashes unexpectedly).


Cheers,

Brendan

Re: instant bluescreen

Posted: Wed May 01, 2013 8:46 am
by Mikemk
Ubuntu 12.04. Not sure the name of the program I used

Re: instant bluescreen

Posted: Wed May 01, 2013 9:17 am
by Love4Boobies
So you started the thread by explaining how to make the OS crash without mentioning the OS or the program that did it...

Re: instant bluescreen

Posted: Wed May 01, 2013 9:29 am
by Mikemk
Love4Boobies wrote:So you started the thread by explaining how to make the OS crash without mentioning the OS or the program that did it...
Good point. I was wondering what the number would be if my entire ram was filled with 0xff. I didn't realize that it would take my entire ram to calculate it.

Re: instant bluescreen

Posted: Wed May 01, 2013 12:13 pm
by xenos
It's 2^(RAM bits) - 1 ;)

Re: instant bluescreen

Posted: Wed May 01, 2013 2:36 pm
by darkinsanity
Brendan wrote:I tried it on Kcalc (running on 64-bit Gentoo); and it caused the calculator to crash (segmentation fault in "__gmpn_copyi()"). :)
I used Kcalc on netrunner 12.12, it hung for a couple of seconds, took 1.680680 gigabytes of RAM and then said "6.48067178374e+2082970051".

Re: instant bluescreen

Posted: Wed May 01, 2013 2:56 pm
by xenos
darkinsanity wrote:6.48067178374e+2082970051
Doesn't look right. 34359738368 ln 2 / ln 10 is a bit less than 10343311892, so one should expect something between 1e+10343311891 and 1e+10343311892. To be more precise:

2^34359738368 = 10^(34359738368 Log[2]/Log[10] - 10343311891) * 1e+10343311891,

which is approximately 8.6013834471e+10343311891. Indeed, Mathematica agrees:

Code: Select all

In[1]:= Timing[2.0^34359738368.0]
Out[1]= {0., 8.6014*10^10343311891}
...and calculates this even too fast to measure the time it takes.

Re: instant bluescreen

Posted: Wed May 01, 2013 3:09 pm
by darkinsanity
XenOS wrote:Doesn't look right.
Then Kcalc either wanted to troll me, or I mistyped something.

Re: instant bluescreen

Posted: Thu May 02, 2013 10:56 am
by CWood
Tried it in KCalc just now, got 1. What?

Re: instant bluescreen

Posted: Thu May 02, 2013 12:12 pm
by Mikemk
I used gcalctool.

Re: instant bluescreen

Posted: Fri May 03, 2013 12:52 am
by dozniak
m12 wrote:I used gcalctool.
Based on you description (should take all the RAM), the bluescreen cannot possibly be instant because it takes some time to fill all the ram, and then a system would probbaly swap for a while.

And then the calculator process will be terminated by oom killer; still no bsod.

Re: bluescreen

Posted: Fri May 03, 2013 7:17 am
by Mikemk
dozniak wrote:
m12 wrote:I used gcalctool.
Based on you description (should take all the RAM), the bluescreen cannot possibly be instant because it takes some time to fill all the ram, and then a system would probbaly swap for a while..
OK, fine. I'll take the word instant out.