| commit | ab5ea55805f89a26e35a8becb072049fec0f9d01 | |
| author | Jan Kiszka <jan.kiszka@web.de> | |
| Wed, 2 Jun 2010 06:49:14 +0000 (2 08:49 +0200) | ||
| committer | Aurelien Jarno <aurelien@aurel32.net> | |
| Wed, 30 Jun 2010 18:59:53 +0000 (30 20:59 +0200) | ||
| tree | 3760e9e37f101f7abca8d22dcaff9abc14db4018 | treesnapshot (tar.gz zip) |
| parent | 6881dd5f199e283b88c69fe1a3f58331031a3752 | commitdiff |
x86: svm: Always clear event_inj on vmexit
We currently only clear SVM_EVTINJ_VALID after successful interrupt
delivery. This apparently does not match real hardware which clears the
whole event_inj field on every vmexit, including unsuccessful interrupt
delivery.
Reported-by: Erik van der Kouwe <vdkouwe@cs.vu.nl>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
We currently only clear SVM_EVTINJ_VALID after successful interrupt
delivery. This apparently does not match real hardware which clears the
whole event_inj field on every vmexit, including unsuccessful interrupt
delivery.
Reported-by: Erik van der Kouwe <vdkouwe@cs.vu.nl>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
| target-i386/op_helper.c | diffblobblamehistory |