summaryrefslogtreecommitdiffstats
path: root/drivers/usb/serial/ezusb_convert.pl
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/usb/serial/ezusb_convert.pl')
-rw-r--r--drivers/usb/serial/ezusb_convert.pl50
1 files changed, 50 insertions, 0 deletions
diff --git a/drivers/usb/serial/ezusb_convert.pl b/drivers/usb/serial/ezusb_convert.pl
new file mode 100644
index 000000000000..13f11469116e
--- /dev/null
+++ b/drivers/usb/serial/ezusb_convert.pl
@@ -0,0 +1,50 @@
+#! /usr/bin/perl -w
+
+
+# convert an Intel HEX file into a set of C records usable by the firmware
+# loading code in usb-serial.c (or others)
+
+# accepts the .hex file(s) on stdin, a basename (to name the initialized
+# array) as an argument, and prints the .h file to stdout. Typical usage:
+# perl ezusb_convert.pl foo <foo.hex >fw_foo.h
+
+
+my $basename = $ARGV[0];
+die "no base name specified" unless $basename;
+
+while (<STDIN>) {
+ # ':' <len> <addr> <type> <len-data> <crc> '\r'
+ # len, type, crc are 2-char hex, addr is 4-char hex. type is 00 for
+ # normal records, 01 for EOF
+ my($lenstring, $addrstring, $typestring, $reststring, $doscrap) =
+ /^:(\w\w)(\w\w\w\w)(\w\w)(\w+)(\r?)$/;
+ die "malformed line: $_" unless $reststring;
+ last if $typestring eq '01';
+ my($len) = hex($lenstring);
+ my($addr) = hex($addrstring);
+ my(@bytes) = unpack("C*", pack("H".(2*$len), $reststring));
+ #pop(@bytes); # last byte is a CRC
+ push(@records, [$addr, \@bytes]);
+}
+
+@sorted_records = sort { $a->[0] <=> $b->[0] } @records;
+
+print <<"EOF";
+/*
+ * ${basename}_fw.h
+ *
+ * Generated from ${basename}.s by ezusb_convert.pl
+ * This file is presumed to be under the same copyright as the source file
+ * from which it was derived.
+ */
+
+EOF
+
+print "static const struct ezusb_hex_record ${basename}_firmware[] = {\n";
+foreach $r (@sorted_records) {
+ printf("{ 0x%04x,\t%d,\t{", $r->[0], scalar(@{$r->[1]}));
+ print join(", ", map {sprintf('0x%02x', $_);} @{$r->[1]});
+ print "} },\n";
+}
+print "{ 0xffff,\t0,\t{0x00} }\n";
+print "};\n";