Re: [PATCH v2] entry/kvm: Make vCPU tasks exit to userspace when a livepatch is pending

From: Eric W. Biederman
Date: Wed May 04 2022 - 10:53:55 EST


Petr Mladek <pmladek@xxxxxxxx> writes:

> On Wed 2022-05-04 08:50:22, Seth Forshee wrote:
>> On Wed, May 04, 2022 at 03:07:53PM +0200, Petr Mladek wrote:

>> > If "no" then I do not understand why TIF_NOTIFY_SIGNAL interrupts
>> > the syscall on the real hardware and not in kvm.
>>
>> It does interrupt, but xfer_to_guest_mode_handle_work() concludes it's
>> not necessary to return to userspace and resumes guest execution.
>
> In this case, we should revert the commit 8df1947c71ee53c7e21
> ("livepatch: Replace the fake signal sending with TIF_NOTIFY_SIGNAL
> infrastructure"). The flag TIF_NOTIFY_SIGNAL clearly does not guarantee
> restarting the syscall or exiting to the user space with -EINTR.
>
> It should solve this problem. And it looks like a cleaner solution
> to me.

Why not just?

diff --git a/kernel/entry/kvm.c b/kernel/entry/kvm.c
index 9d09f489b60e..cbb192aec13a 100644
--- a/kernel/entry/kvm.c
+++ b/kernel/entry/kvm.c
@@ -8,13 +8,7 @@ static int xfer_to_guest_mode_work(struct kvm_vcpu *vcpu, unsigned long ti_work)
do {
int ret;

- if (ti_work & (_TIF_SIGPENDING | _TIF_NOTIFY_SIGNAL)) {
- clear_notify_signal();
- if (task_work_pending(current))
- task_work_run();
- }
-
- if (ti_work & _TIF_SIGPENDING) {
+ if (ti_work & (_TIF_SIGPENDING | _TIF_NOTIFY_SIGNAL)) {
kvm_handle_signal_exit(vcpu);
return -EINTR;
}

As far as I can tell the only reason _TIF_NOTIFY_SIGNAL was handled any
differently than _TIF_SIGPENDING in xfer_to_guest_mode_work is because
of historical confusion.

Eric