Re^3: Integer overflow

At least, on my 64-bit box, I get 2158052273

Yes, on my 64-bit build on linux (32-bit architecture) I'd have to use your solution.

On an architecture where sizeof(long)==8, I would think that neither solution would work. (This is probably, though not necessarily, an irrelevant observation in the context of what the op needs :-)

Cheers,

Comment on Re^3: Integer overflow

Replies are listed 'Best First'.
Re^4: Integer overflow by almut (Canon) on Apr 24, 2009 at 00:53 UTC
On an architecture where sizeof(long)==8, I would think that neither solution would work. I think (or actually, I'm pretty sure, because on my system, sizeof(long)==8) the pack/unpack solution would work, because `pack "l"` always (by definiton) is 32-bit, independent of what long might be on the system/C compiler, otherwise.	[reply] [d/l]
Re^5: Integer overflow by syphilis (Archbishop) on Apr 24, 2009 at 01:15 UTC
`pack "l"` always (by definiton) is 32-bit Yes, you're right (should'er known). According to the docs, the `i` and `I` specifiers take note of what the C compiler regards as an 'int', but the `l` and `L` types are always 32-bit. Sheesh ... you just can't take anything for granted :-) Cheers, Rob	[reply] [d/l] [select]