linux-kernel - Re: [PATCH v5 18/23] KVM: arm64: GICv4.1: Add direct injection capability to SGI registers

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <9b1ce453-f24f-7974-99ca-1608ad7c6a13@redhat.com>
Date:   Fri, 20 Mar 2020 11:56:48 +0100
From:   Auger Eric <eric.auger@...hat.com>
To:     Marc Zyngier <maz@...nel.org>
Cc:     linux-arm-kernel@...ts.infradead.org, kvmarm@...ts.cs.columbia.edu,
        kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
        Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
        Jason Cooper <jason@...edaemon.net>,
        Robert Richter <rrichter@...vell.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Zenghui Yu <yuzenghui@...wei.com>,
        James Morse <james.morse@....com>,
        Julien Thierry <julien.thierry.kdev@...il.com>,
        Suzuki K Poulose <suzuki.poulose@....com>
Subject: Re: [PATCH v5 18/23] KVM: arm64: GICv4.1: Add direct injection
 capability to SGI registers

Hi Marc,

On 3/20/20 11:05 AM, Marc Zyngier wrote:
> Hi Eric,
> 
> On 2020-03-20 08:11, Auger Eric wrote:
>> Hi Marc,
>> On 3/4/20 9:33 PM, Marc Zyngier wrote:
>>> Most of the GICv3 emulation code that deals with SGIs now has to be
>>> aware of the v4.1 capabilities in order to benefit from it.
>>>
>>> Add such support, keyed on the interrupt having the hw flag set and
>>> being a SGI.
>>>
>>> Signed-off-by: Marc Zyngier <maz@...nel.org>
>>> ---
>>>  virt/kvm/arm/vgic/vgic-mmio-v3.c | 15 +++++-
>>>  virt/kvm/arm/vgic/vgic-mmio.c    | 88 ++++++++++++++++++++++++++++++--
>>>  2 files changed, 96 insertions(+), 7 deletions(-)
>>>
> 
> [...]
> 
>>> @@ -113,7 +125,21 @@ void vgic_mmio_write_senable(struct kvm_vcpu *vcpu,
>>>          struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid +
>>> i);
>>>
>>>          raw_spin_lock_irqsave(&irq->irq_lock, flags);
>>> -        if (vgic_irq_is_mapped_level(irq)) {
>>> +        if (irq->hw && vgic_irq_is_sgi(irq->intid)) {
>>> +            if (!irq->enabled) {
>>> +                struct irq_data *data;
>>> +
>>> +                irq->enabled = true;
>>> +                data = &irq_to_desc(irq->host_irq)->irq_data;
>>> +                while (irqd_irq_disabled(data))
>>> +                    enable_irq(irq->host_irq);
>> could you explain me why the while() is requested?
> 
> Ah, interesting question: disable_irq() (and its variants) can nest. This
> means that if you have done two disable_irq(), you must do two enable_irq()
> to get back to the interrupt being enabled.
> 
> The locking should ensure that this nesting doesn't happen, but I'm
> paranoid
> (see the GICv4.0 doorbell handling). It also makes it easier to reason
> about
> the initial state.

OK! thank you for this explanation.

Thanks

Eric
> 
> [...]
> 
>> Reviewed-by: Eric Auger <eric.auger@...hat.com>
> 
> Thanks!
> 
>          M.