Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Message-ID: <d58a6a64-a6b6-3238-00d7-573e4177f38c@linux.ibm.com>
Date:   Thu, 24 Feb 2022 16:47:19 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:91.0)
 Gecko/20100101 Thunderbird/91.6.1
Subject: Re: [PATCH v3 1/1] KVM: s390: pv: make use of ultravisor AIV support
Content-Language: en-US
To:     Christian Borntraeger <borntraeger@de.ibm.com>, kvm@vger.kernel.org
Cc:     cohuck@redhat.com, frankja@linux.ibm.com, thuth@redhat.com,
        pasic@linux.ibm.com, david@redhat.com, linux-s390@vger.kernel.org,
        linux-kernel@vger.kernel.org
References: <20220209152217.1793281-1-mimu@linux.ibm.com>
 <20220209152217.1793281-2-mimu@linux.ibm.com>
 <803275d5-58f4-21d9-8020-e56f05737450@de.ibm.com>
From:   Michael Mueller <mimu@linux.ibm.com>
In-Reply-To: <803275d5-58f4-21d9-8020-e56f05737450@de.ibm.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 22.02.22 09:13, Christian Borntraeger wrote:
> Am 09.02.22 um 16:22 schrieb Michael Mueller:
>> This patch enables the ultravisor adapter interruption vitualization
>> support indicated by UV feature BIT_UV_FEAT_AIV. This allows ISC
>> interruption injection directly into the GISA IPM for PV kvm guests.
>>
>> Hardware that does not support this feature will continue to use the
>> UV interruption interception method to deliver ISC interruptions to
>> PV kvm guests. For this purpose, the ECA_AIV bit for all guest cpus
>> will be cleared and the GISA will be disabled during PV CPU setup.
>>
>> In addition a check in __inject_io() has been removed. That reduces the
>> required instructions for interruption handling for PV and traditional
>> kvm guests.
>>
>> Signed-off-by: Michael Mueller <mimu@linux.ibm.com>
> 
> The CI said the following with gisa_disable in the calltrace.
> Will drop from next for now.

It turns out this is related to kvm_s390_set_tod_clock() which
is triggered by a kvm-unit-test (sac_PV) and not related directly
to this patch. Please re-apply.

> 
>     LOCKDEP_CIRCULAR (suite: kvm-unit-tests-kvm, case: -)
>                  WARNING: possible circular locking dependency detected
>                  
> 5.17.0-20220221.rc5.git1.b8f0356a093a.300.fc35.s390x+debug #1 Not tainted
>                  ------------------------------------------------------
>                  qemu-system-s39/161139 is trying to acquire lock:
>                  0000000280dc0b98 (&kvm->lock){+.+.}-{3:3}, at: 
> kvm_s390_set_tod_clock+0x36/0x220 [kvm]
>                  but task is already holding lock:
>                  0000000280f4e4b8 (&vcpu->mutex){+.+.}-{3:3}, at: 
> kvm_vcpu_ioctl+0x9a/0xa40 [kvm]
>                  which lock already depends on the new lock.
>                  the existing dependency chain (in reverse order) is:
>                  -> #1 (&vcpu->mutex){+.+.}-{3:3}:
>                         __lock_acquire+0x604/0xbd8
>                         lock_acquire.part.0+0xe2/0x250
>                         lock_acquire+0xb0/0x200
>                         __mutex_lock+0x9e/0x8a0
>                         mutex_lock_nested+0x32/0x40
>                         kvm_s390_gisa_disable+0xa4/0x130 [kvm]
>                         kvm_s390_handle_pv+0x718/0x778 [kvm]
>                         kvm_arch_vm_ioctl+0x4ac/0x5f8 [kvm]
>                         kvm_vm_ioctl+0x336/0x530 [kvm]
>                         __s390x_sys_ioctl+0xbe/0x100
>                         __do_syscall+0x1da/0x208
>                         system_call+0x82/0xb0
>                  -> #0 (&kvm->lock){+.+.}-{3:3}:
>                         check_prev_add+0xe0/0xed8
>                         validate_chain+0x736/0xb20
>                         __lock_acquire+0x604/0xbd8
>                         lock_acquire.part.0+0xe2/0x250
>                         lock_acquire+0xb0/0x200
>                         __mutex_lock+0x9e/0x8a0
>                         mutex_lock_nested+0x32/0x40
>                         kvm_s390_set_tod_clock+0x36/0x220 [kvm]
>                         kvm_s390_handle_b2+0x378/0x728 [kvm]
>                         kvm_handle_sie_intercept+0x13a/0x448 [kvm]
>                         vcpu_post_run+0x28e/0x560 [kvm]
>                         __vcpu_run+0x266/0x388 [kvm]
>                         kvm_arch_vcpu_ioctl_run+0x10a/0x270 [kvm]
>                         kvm_vcpu_ioctl+0x27c/0xa40 [kvm]
>                         __s390x_sys_ioctl+0xbe/0x100
>                         __do_syscall+0x1da/0x208
>                         system_call+0x82/0xb0
>                  other info that might help us debug this:
>                   Possible unsafe locking scenario:
>                         CPU0                    CPU1
>                         ----                    ----
>                    lock(&vcpu->mutex);
>                                                 lock(&kvm->lock);
>                                                 lock(&vcpu->mutex);
>                    lock(&kvm->lock);
>                   *** DEADLOCK ***
>                  2 locks held by qemu-system-s39/161139:
>                   #0: 0000000280f4e4b8 (&vcpu->mutex){+.+.}-{3:3}, at: 
> kvm_vcpu_ioctl+0x9a/0xa40 [kvm]
>                   #1: 0000000280dc47c8 (&kvm->srcu){....}-{0:0}, at: 
> __vcpu_run+0x1d4/0x388 [kvm]
>                  stack backtrace:
>                  CPU: 10 PID: 161139 Comm: qemu-system-s39 Not tainted 
> 5.17.0-20220221.rc5.git1.b8f0356a093a.300.fc35.s390x+debug #1
>                  Hardware name: IBM 8561 T01 701 (LPAR)
>                  Call Trace:
>                   [<00000001da4e89de>] dump_stack_lvl+0x8e/0xc8
>                   [<00000001d9876c56>] check_noncircular+0x136/0x158
>                   [<00000001d9877c70>] check_prev_add+0xe0/0xed8
>                   [<00000001d987919e>] validate_chain+0x736/0xb20
>                   [<00000001d987b23c>] __lock_acquire+0x604/0xbd8
>                   [<00000001d987c432>] lock_acquire.part.0+0xe2/0x250
>                   [<00000001d987c650>] lock_acquire+0xb0/0x200
>                   [<00000001da4f72ae>] __mutex_lock+0x9e/0x8a0
>                   [<00000001da4f7ae2>] mutex_lock_nested+0x32/0x40
>                   [<000003ff8070cd6e>] kvm_s390_set_tod_clock+0x36/0x220 
> [kvm]
>                   [<000003ff8071dd68>] kvm_s390_handle_b2+0x378/0x728 [kvm]
>                   [<000003ff8071146a>] 
> kvm_handle_sie_intercept+0x13a/0x448 [kvm]
>                   [<000003ff8070dd46>] vcpu_post_run+0x28e/0x560 [kvm]
>                   [<000003ff8070e27e>] __vcpu_run+0x266/0x388 [kvm]
>                   [<000003ff8070eba2>] 
> kvm_arch_vcpu_ioctl_run+0x10a/0x270 [kvm]
>                   [<000003ff806f4044>] kvm_vcpu_ioctl+0x27c/0xa40 [kvm]
>                   [<00000001d9b47ac6>] __s390x_sys_ioctl+0xbe/0x100
>                   [<00000001da4ec152>] __do_syscall+0x1da/0x208
>                   [<00000001da4fec42>] system_call+0x82/0xb0
>                  INFO: lockdep is turned off.