Same happens with Percona Server 5.7.10-3 GA as soon as I have:

innodb_flush_method=ALL_O_DIRECT

in my.cnf:

[root@centos ~]# service mysql stop
Stopping mysqld:                                           [  OK  ]
[root@centos ~]# vi /etc/my.cnf
[root@centos ~]# service mysql start
Percona MySQL Daemon failed to start.
Starting mysqld:                                           [FAILED]
[root@centos ~]# tail -100 /var/log/mysqld.log
...
2016-02-26T15:49:07.327649Z 0 [Note] /usr/sbin/mysqld: Shutdown complete

160226 17:49:07 mysqld_safe mysqld from pid file /var/run/mysqld/mysqld.pid ended
160226 17:50:41 mysqld_safe Transparent huge pages are already set to: never.
160226 17:50:41 mysqld_safe Starting mysqld daemon with databases from /var/lib/mysql
2016-02-26T15:50:42.607807Z 0 [Warning] TIMESTAMP with implicit DEFAULT value is deprecated. Please use --explicit_defaults_for_timestamp server option (see documentation for more details).
2016-02-26T15:50:42.608882Z 0 [Warning] 'NO_ZERO_DATE', 'NO_ZERO_IN_DATE' and 'ERROR_FOR_DIVISION_BY_ZERO' sql modes should be used with strict mode. They will be merged with strict mode in a future release.
2016-02-26T15:50:42.609278Z 0 [Warning] 'NO_AUTO_CREATE_USER' sql mode was not set.
2016-02-26T15:50:42.622279Z 0 [Note] /usr/sbin/mysqld (mysqld 5.7.10-3-log) starting as process 2958 ...
2016-02-26T15:50:42.667392Z 0 [Warning] No argument was provided to --log-bin, and --log-bin-index was not used; so replication may break when this MySQL server acts as a master and has his hostname changed!! Please use '--log-bin=centos-bin' to avoid this problem.
2016-02-26T15:50:42.680642Z 0 [Note] InnoDB: innodb_empty_free_list_algorithm has been changed to legacy because of small buffer pool size. In order to use backoff, increase buffer pool at least up to 20MB.

2016-02-26T15:50:42.681648Z 0 [Note] InnoDB: PUNCH HOLE support available
2016-02-26T15:50:42.681862Z 0 [Note] InnoDB: Mutexes and rw_locks use GCC atomic builtins
2016-02-26T15:50:42.681948Z 0 [Note] InnoDB: Uses event mutexes
2016-02-26T15:50:42.682280Z 0 [Note] InnoDB: GCC builtin __atomic_thread_fence() is used for memory barrier
2016-02-26T15:50:42.682376Z 0 [Note] InnoDB: Compressed tables use zlib 1.2.3
2016-02-26T15:50:42.682439Z 0 [Note] InnoDB: Using Linux native AIO
2016-02-26T15:50:42.684322Z 0 [Note] InnoDB: Number of pools: 1
2016-02-26T15:50:42.685172Z 0 [Note] InnoDB: Not using CPU crc32 instructions
2016-02-26T15:50:43.008759Z 0 [Note] InnoDB: Initializing buffer pool, total size = 5M, instances = 1, chunk size = 5M
2016-02-26T15:50:43.029818Z 0 [Note] InnoDB: Completed initialization of buffer pool
2016-02-26T15:50:43.085462Z 0 [Note] InnoDB: If the mysqld execution user is authorized, page cleaner thread priority can be changed. See the man page of setpriority().
2016-02-26T15:50:43.645092Z 0 [Note] InnoDB: Highest supported file format is Barracuda.
2016-02-26T15:50:43.651811Z 0 [Warning] InnoDB: Retry attempts for reading partial data failed.
2016-02-26T15:50:43.652176Z 0 [ERROR] InnoDB: Tried to read 2048 bytes at offset 0 was only able to read0
2016-02-26T15:50:43.652257Z 0 [ERROR] InnoDB: Operating system error number 22 in a file operation.
2016-02-26T15:50:43.652337Z 0 [ERROR] InnoDB: Error number 22 means 'Invalid argument'
2016-02-26T15:50:43.652383Z 0 [Note] InnoDB: Some operating system error numbers are described at http://dev.mysql.com/doc/refman/5.7/en/operating-system-error-codes.html
2016-02-26T15:50:43.652569Z 0 [ERROR] InnoDB: File (unknown): 'read' returned OS error 122. Cannot continue operation
2016-02-26T15:50:43.652608Z 0 [ERROR] InnoDB: Cannot continue operation.
[root@centos ~]# grep innodb_flush /etc/my.cnf
innodb_flush_method=ALL_O_DIRECT
[root@centos ~]#

So, I'd say we have a regression here, already noted on CentOS 6.7, Ubuntu 14.04 and 15.04.