| commit | 1c82528f8cef42b3f34534c25412c7fa39a66811 | [log] [tgz] |
|---|---|---|
| author | Markus Metzger <markus.t.metzger@intel.com> | Fri Jan 08 11:34:25 2016 +0100 |
| committer | Markus Metzger <markus.t.metzger@intel.com> | Fri Jan 08 11:35:11 2016 +0100 |
| tree | cb1d9abd88baf5c0ae27e27a5657cfa802859257 | |
| parent | ee6dd6340f6e7c9336f6ccce83fbc9fb8632831a [diff] |
libipt, ild: fix a0-a3 MOV decode The width of the displacement operand of the a0-a3 MOV instruction depends on the address-size, not the operand-size. Fix the respective case in compute_disp_dec() and add helper functions to determine the effective address size. We seem to have gotten this wrong also in our ild unit tests. Fix them and add some more. Change-Id: Ib2b5e57884ee8b2faeefb45959ea3ef2e433385a Signed-off-by: Markus Metzger <markus.t.metzger@intel.com>