Blue Screen errors with WireGuardNT using two peers

Blue Screen errors with WireGuardNT using two peers

[Hiroshi Shirosaki]

Hi,

I tried to use WireGuard 0.4.0 with ExperimentalKernelDriver on.
When I use only one peer, it seems to work fine.
When I used another peer to ssh at the same time, Windows 10 became
blue screen and rebooted.
Windows 10 Pro 21H1.
Is using multiple peers at the same time not supported?

Thanks,
Hiroshi Shirosaki

Re: Blue Screen errors with WireGuardNT using two peers

[Jason A. Donenfeld]

Hi Hiroshi,

On 8/3/21 7:06 AM, Hiroshi Shirosaki wrote:
> I tried to use WireGuard 0.4.0 with ExperimentalKernelDriver on.
> When I use only one peer, it seems to work fine.
> When I used another peer to ssh at the same time, Windows 10 became
> blue screen and rebooted.
> Windows 10 Pro 21H1.
> Is using multiple peers at the same time not supported?

Multiple peers _should_ be supported, but you've evidently hit some kind 
of bug. Do you think you could send me the files in the 
C:\windows\minidump\ directory? It's probably best to mail these to me 
personally rather than sending them to the mailing list.

Thanks,
Jason

Re: Blue Screen errors with WireGuardNT using two peers

[Jason A. Donenfeld]

On 8/3/21 5:01 PM, Jason A. Donenfeld wrote:
> On 8/3/21 7:06 AM, Hiroshi Shirosaki wrote:
>> I tried to use WireGuard 0.4.0 with ExperimentalKernelDriver on.
>> When I use only one peer, it seems to work fine.
>> When I used another peer to ssh at the same time, Windows 10 became
>> blue screen and rebooted.
>> Windows 10 Pro 21H1.
>> Is using multiple peers at the same time not supported?
> 
> Multiple peers _should_ be supported, but you've evidently hit some kind 
> of bug. Do you think you could send me the files in the 
> C:\windows\minidump\ directory? It's probably best to mail these to me 
> personally rather than sending them to the mailing list.

v0.4.1 is now released. Please see if that fixes the issue.

Jason

Re: Blue Screen errors with WireGuardNT using two peers

[Hiroshi Shirosaki]

On Wed, Aug 4, 2021 at 2:25 AM Jason A. Donenfeld <Jason@zx2c4.com> wrote:
>
> v0.4.1 is now released. Please see if that fixes the issue.

Thank you for the fix.
I tried and it works stably for now.

FYI
I confirmed windbg minidump result of the previous(v0.4.0) failure.
It shows the following.

2: kd> !analyze -v

DRIVER_IRQL_NOT_LESS_OR_EQUAL (d1)
An attempt was made to access a pageable (or completely invalid) address at an
interrupt request level (IRQL) that is too high.  This is usually
caused by drivers using improper addresses.
If kernel debugger is available get stack backtrace.
Arguments:
Arg1: 0000000000000024, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000000, value 0 = read operation, 1 = write operation
Arg4: fffff801245e2d54, address which referenced memory

Debugging Details:
------------------
(snip)

Thanks,
Hiroshi Shirosaki

Re: Blue Screen errors with WireGuardNT using two peers

[Jason A. Donenfeld]

Hi Hiroshi,

> I confirmed windbg minidump result of the previous(v0.4.0) failure.
> It shows the following.
>
> 2: kd> !analyze -v
>
> DRIVER_IRQL_NOT_LESS_OR_EQUAL (d1)
> An attempt was made to access a pageable (or completely invalid) address at an
> interrupt request level (IRQL) that is too high.  This is usually
> caused by drivers using improper addresses.
> If kernel debugger is available get stack backtrace.
> Arguments:
> Arg1: 0000000000000024, memory referenced
> Arg2: 0000000000000002, IRQL
> Arg3: 0000000000000000, value 0 = read operation, 1 = write operation
> Arg4: fffff801245e2d54, address which referenced memory
>
> Debugging Details:
> ------------------
> (snip)

Thank you for letting me know that 0.4.1 fixes the issue. With regards
to your windbg - could you send the snipped part, just to be sure?

Jason