Linux – what is the difference between seek and skip in dd command

Why the second command outputs a different value?

For historical reasons, dd considers x to be a multiplication operator. So 0x3 is evaluated to be 0.

Is it possible to pass the skip|seek offset to dd as an hexadecimal value?

Not directly, as far as I know. As well as multiplication using the operator x, you can suffix any number with b to mean "multiply by 512" (0x200) and with K to mean "multiply by 1024" (0x400). With GNU dd you can also use suffixes M, G, T, P, E, Z and Y to mean multiply by 2 to the power of 20, 30, 40, 50, 60, 70, 80 or 90, respectively, and you can use upper or lower case except for the b suffix. (There are many other possible suffixes. For example, EB means "multiply by 10¹⁸" and PiB means "multiply by 2⁵⁰". See info coreutils "block size" for more information, if you have a GNU installation.)

You might find the above arcane, anachronistic, and geeky to the point of absurdity. Not to worry: you are not alone. Fortunately, you can just ignore it all and use your shell's arithmetic substitution instead (bash and other Posix compliant shells will work, as well as some non-Posix shells). The shell does understand hexadecimal numbers, and it allows a full range of arithmetic operators written in the normal way. You just need to surround the expression with $((...)):

# dd if=2013-Aug-uptime.csv bs=1 count=$((0x2B * 1024)) skip=$((0x37))

I really don't know how to explain this better than the manpage does.

bs= sets the blocksize, for example bs=1M would be 1MiB blocksize.

count= copies only this number of blocks (the default is for dd to keep going forever or until the input runs out). Ideally blocks are of bs= size but there may be incomplete reads, so if you use count= in order to copy a specific amount of data (count*bs), you should also supply iflag=fullblock.

seek= seeks this number of blocks in the output, instead of writing to the very beginning of the output device.

So, for example, this copies 1MiB worth of y\n to position 8MiB of the outputfile. So the total filesize will be 9MiB.

$ yes | dd bs=1M count=1 seek=8 iflag=fullblock of=outputfile
$ ls -alh outputfile
9.0M Jun  3 21:02 outputfile
$ hexdump -C outputfile
00000000  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
00800000  79 0a 79 0a 79 0a 79 0a  79 0a 79 0a 79 0a 79 0a  |y.y.y.y.y.y.y.y.|
*
00900000

Since you mention /dev/random and overwriting partitions... it will take forever since /dev/random (as well as /dev/urandom) is just too slow. You could just use shred -v -n 1 instead, that's fast and usually available anywhere.