Steve,
On Wed, 25 May 2005, Steve King wrote:
> Ok, I have sucessfully done 100,000,000.
>
> Please bear with me, as you saw, it takes several hours for each
> run, so I will have to run them overnight.
2^31*(log(2)/log(10)==646,456,993. That's where it *should* start
failing. Can you confirm t
Steve,
On Wed, 25 May 2005, Steve King wrote:
> Ok, I have sucessfully done 100,000,000.
>
> Please bear with me, as you saw, it takes several hours for each
> run, so I will have to run them overnight.
Ahhh, I'm beginning to suspect something: In CLN, the exponent is 32 Bits
wide, regardless of
Ok, I have sucessfully done 100,000,000.
Please bear with me, as you saw, it takes several hours for each
run, so I will have to run them overnight.
--
Steve
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
On Tue, 24 May 2005, Steve King wrote:
> I ran it again overnight,
>
> This is what we get:
>
> $ time pi 10 > pi
> Floating point overflow.
>
> real505m34.348s
> user314m7.021s
> sys 10m59.082s
>
>
> I don't have a full 20Gbytes of memory, so it is swapping.
> However, except f
I ran it again overnight,
This is what we get:
$ time pi 10 > pi
Floating point overflow.
real505m34.348s
user314m7.021s
sys 10m59.082s
I don't have a full 20Gbytes of memory, so it is swapping.
However, except for speed and killing my disk, it should not
affect the result
Thanks for your bugreport.
On Mon, 23 May 2005, Steve King wrote:
> Whilst messing with my new AMD64 I tried:
> pi 10
> and left it going over the weekend.
> Unfortunately when I came back, pi had failed with a floating point
> overflow.
Was "floatint point overflow" the actual failure?
Package: pi
Version: 1.1.9-1
Severity: minor
Whilst messing with my new AMD64 I tried:
pi 10
and left it going over the weekend.
Unfortunately when I came back, pi had failed with a floating point
overflow.
(With such a huge request the virtual address space was >20Gbytes
But is that not
7 matches
Mail list logo
