Il 27/04/2014 13:40, Toralf Förster ha scritto:
On 04/27/2014 12:45 PM, Paolo Bonzini wrote:
Il 27/04/2014 12:30, Toralf Förster ha scritto:
Signed-off-by: Toralf Förster <toralf.foerster@xxxxxx>
---
arch/x86/kvm/x86.c | 23 ++++++++++-------------
1 file changed, 10 insertions(+), 13 deletions(-)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 8b8fc0b..93cf454 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -5680,20 +5680,17 @@ int kvm_hv_hypercall(struct kvm_vcpu *vcpu)
kvm_x86_ops->get_cs_db_l_bits(vcpu, &cs_db, &cs_l);
longmode = is_long_mode(vcpu) && cs_l == 1;
- if (!longmode) {
- param = ((u64)kvm_register_read(vcpu, VCPU_REGS_RDX) << 32) |
- (kvm_register_read(vcpu, VCPU_REGS_RAX) & 0xffffffff);
- ingpa = ((u64)kvm_register_read(vcpu, VCPU_REGS_RBX) << 32) |
- (kvm_register_read(vcpu, VCPU_REGS_RCX) & 0xffffffff);
- outgpa = ((u64)kvm_register_read(vcpu, VCPU_REGS_RDI) << 32) |
- (kvm_register_read(vcpu, VCPU_REGS_RSI) & 0xffffffff);
- }
#ifdef CONFIG_X86_64
- else {
- param = kvm_register_read(vcpu, VCPU_REGS_RCX);
- ingpa = kvm_register_read(vcpu, VCPU_REGS_RDX);
- outgpa = kvm_register_read(vcpu, VCPU_REGS_R8);
- }
+ param = kvm_register_read(vcpu, VCPU_REGS_RCX);
+ ingpa = kvm_register_read(vcpu, VCPU_REGS_RDX);
+ outgpa = kvm_register_read(vcpu, VCPU_REGS_R8);
+#else
+ param = ((u64)kvm_register_read(vcpu, VCPU_REGS_RDX) << 32) |
+ (kvm_register_read(vcpu, VCPU_REGS_RAX) & 0xffffffff);
+ ingpa = ((u64)kvm_register_read(vcpu, VCPU_REGS_RBX) << 32) |
+ (kvm_register_read(vcpu, VCPU_REGS_RCX) & 0xffffffff);
+ outgpa = ((u64)kvm_register_read(vcpu, VCPU_REGS_RDI) << 32) |
+ (kvm_register_read(vcpu, VCPU_REGS_RSI) & 0xffffffff);
#endif
code = param & 0xffff;
No, wait, it's only superfluous in the sense that you don't need
longmode for !CONFIG_X86_64. It's definitely needed for CONFIG_X86_64,
because the guest can run in 32-bit mode.
Paolo
Ah, so the following would work, but looks too ugly, right ? :
#ifdef CONFIG_X86_64
if (!longmode) {
#endif
param = ((u64)kvm_register_read(vcpu, VCPU_REGS_RDX) << 32) |
(kvm_register_read(vcpu, VCPU_REGS_RAX) & 0xffffffff);
ingpa = ((u64)kvm_register_read(vcpu, VCPU_REGS_RBX) << 32) |
(kvm_register_read(vcpu, VCPU_REGS_RCX) & 0xffffffff);
outgpa = ((u64)kvm_register_read(vcpu, VCPU_REGS_RDI) << 32) |
(kvm_register_read(vcpu, VCPU_REGS_RSI) & 0xffffffff);
#ifdef CONFIG_X86_64
}
else {
param = kvm_register_read(vcpu, VCPU_REGS_RCX);
ingpa = kvm_register_read(vcpu, VCPU_REGS_RDX);
outgpa = kvm_register_read(vcpu, VCPU_REGS_R8);
}
#endif
Yep. :) Note that GCC correctly reports no warning, because it looks
through is_long_mode. In the end, #ifdef in the code can just be
removed. Please compile-test it on 32-bit though (well, I will too
before committing but...).
Paolo
--
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
