The driver handles the endian conversion already, so doing this from
user-space was resulting in the wrong value for big-endian machines.
Signed-off-by: Keith Busch <keith.busch@intel.com>
.data_len = data_len,
};
__u32 numd = (data_len >> 2) - 1;
- __le16 numdu, numdl;
+ __u16 numdu = numd >> 16, numdl = numd & 0xffff;
- numdu = htole16((numd & 0xFFFF0000) >> 16);
- numdl = htole16(numd & 0x0000FFFF);
cmd.cdw10 = log_id | (numdl << 16);
cmd.cdw11 = numdu;