fix potential overflow in exponent reading
authorRich Felker <dalias@aerifal.cx>
Wed, 11 Apr 2012 03:05:16 +0000 (23:05 -0400)
committerRich Felker <dalias@aerifal.cx>
Wed, 11 Apr 2012 03:05:16 +0000 (23:05 -0400)
commit633a26c1e69b6a977d16086834f2b937e0378002
treee4f51cf1854e3416f9423c5c15bb779f4eeb5004
parentc5ff29699cb3cccf8dcd8e44ddf4dcb7599e585c
fix potential overflow in exponent reading

note that there's no need for a precise cutoff, because exponents this
large will always result in overflow or underflow (it's impossible to
read enough digits to compensate for the exponent magnitude; even at a
few nanoseconds per digit it would take hundreds of years).
src/internal/floatscan.c