Hello Nadav Amit,
The patch d1442d85cc30: "KVM: x86: Handle errors when RIP is set
during far jumps" from Sep 18, 2014, leads to the following static
checker warning:
arch/x86/kvm/emulate.c:2015 em_jmp_far()
warn: add some parenthesis here?
arch/x86/kvm/emulate.c
2013 rc = assign_eip_far(ctxt, ctxt->src.val, new_desc.l);
2014 if (rc != X86EMUL_CONTINUE) {
2015 WARN_ON(!ctxt->mode != X86EMUL_MODE_PROT64);
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
No idea what was intended here. The negate has higher precedence than
the !=.
2016 /* assigning eip failed; restore the old cs */
2017 ops->set_segment(ctxt, old_sel, &old_desc, 0, VCPU_SREG_CS);
2018 return rc;
2019 }
There are a couple other static checker warnings as well:
arch/x86/kvm/emulate.c:579 assign_eip_far()
warn: bitwise AND condition is false here
567 static inline int assign_eip_far(struct x86_emulate_ctxt *ctxt, ulong dst,
568 int cs_l)
569 {
570 switch (ctxt->op_bytes) {
571 case 2:
572 ctxt->_eip = (u16)dst;
573 break;
574 case 4:
575 ctxt->_eip = (u32)dst;
576 break;
577 case 8:
578 if ((cs_l && is_noncanonical_address(dst)) ||
579 (!cs_l && (dst & ~(u32)-1)))
^^^^^^^^
This is a very complicated way of saying zero.
580 return emulate_gp(ctxt, 0);
581 ctxt->_eip = dst;
582 break;
583 default:
584 WARN(1, "unsupported eip assignment size\n");
585 }
586 return X86EMUL_CONTINUE;
587 }
arch/x86/kvm/emulate.c:2112 em_ret_far()
warn: add some parenthesis here?
2110 rc = assign_eip_far(ctxt, eip, new_desc.l);
2111 if (rc != X86EMUL_CONTINUE) {
2112 WARN_ON(!ctxt->mode != X86EMUL_MODE_PROT64);
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2113 ops->set_segment(ctxt, old_cs, &old_desc, 0, VCPU_SREG_CS);
2114 }
2115 return rc;
2116 }
regards,
dan carpenter
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
