tree ff4eb1b93ce96a2b30956ee7b8c1b9569c3879c5
parent 9fba70569d9c3c253dba10ebbe3359f2157e504c
author Jan Kara <jack@suse.cz> 1453838850 +0100
committer Jan Kara <jack@suse.cz> 1455019523 +0100

udf: Use separate buffer for copying split names

Code in udf_find_entry() and udf_readdir() used the same buffer for
storing filename that was split among blocks and for the resulting
filename in utf8. This worked because udf_get_filename() first
internally copied the name into a different buffer and only then
performed a conversion into the destination buffer. However we want to
get rid of intermediate buffers so use separate buffer for converted
name and name split between blocks so that we don't have the same source
and destination buffer when converting split names.

Signed-off-by: Jan Kara <jack@suse.cz>