23 Nov
2005
23 Nov
'05
9:22 a.m.
I am wondering whether it would be worth to implement a SHA1
hash
calculation over given memory address range. In this way one
could
assure for example that the binary image version is correct
or that
binary image is intact.
Is the existing CRC32 checksum
not good enough for your purposes?
The CRC32 is not collision-proof in similar fashion that SHA1. In real life, if two binary images give the same SHA1 checksum, you can say that these binaries are identical. With CRC32 this is not so - or at least I would not count on that :-)
br,
Tom