Replace parts of file with 0xFF?

MogensJ :

I want to modify a file so every byte from location 0x3000 to 0xDC000 is replaced with 0xFF, everything else should be unmodified.

How to accomplish this with standard Linux tools?

Socowi :

This is jhnc's answer with little improvements (explained at the end of this answer). If jhnc decides to turn their comment into an answer please upvote their answer instead/too.

#! /bin/bash
overwrite() {
    file="$1"; from="$2"; to="$3"; with="$4"
    yes '' | tr \\n "\\$(printf %o "$with")" |
    dd conv=notrunc bs=1 seek="$((from))" count="$((to-from))" of="$file"
}

In your case you would use the function from above like

overwrite yourFile 0x3000 0xDC000 0xFF

The start and end byte are both 0-based. The start is inclusive and the end is exclusive. Example:

$ printf 00000 > file
$ overwrite file 1 3 0x57
$ hexdump -C file
00000000  30 57 57 30 30   |0WW00|
00000005

Improvements made:

• Fixed wrong count=... and explained interpretation of start and end.

• Allow filling with null bytes.
  _{If you want to write null bytes 0x00 you cannot use yes $'\x00'. The null byte would represent the end of yes's argument string making the call equivalent to yes ''. Since yes '' | tr -d \\n produces no output dd will wait indefinitely.
  The command presented in this answer allows you to fill the region with any byte (choose one from 0x00 to 0xFF).}