Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Sat, 17 Jun 2023 08:35:45 +0100
Message-ID: <87zg4ypmlq.wl-maz@kernel.org>
From:   Marc Zyngier <maz@kernel.org>
To:     Kunkun Jiang <jiangkunkun@huawei.com>
Cc:     Thomas Gleixner <tglx@linutronix.de>,
        Zenghui Yu <yuzenghui@huawei.com>,
        "open list:IRQCHIP DRIVERS" <linux-kernel@vger.kernel.org>,
        <wanghaibin.wang@huawei.com>, <chenxiang66@hisilicon.com>,
        <tangnianyao@huawei.com>
Subject: Re: [PATCH] irqchipi/gic-v4: Ensure accessing the correct RD when and writing INVLPIR
In-Reply-To: <b073b1f7-0567-eb45-dec9-dcd2054e41c9@huawei.com>
References: <20230412041510.497-1-jiangkunkun@huawei.com>
        <86y1mxl9m4.wl-maz@kernel.org>
        <f618c540-879c-ca5b-31af-e55472d8208c@huawei.com>
        <86bkikmvjk.wl-maz@kernel.org>
        <b073b1f7-0567-eb45-dec9-dcd2054e41c9@huawei.com>
User-Agent: Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue)
 FLIM-LB/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL-LB/10.8 EasyPG/1.0.0 Emacs/28.2
 (x86_64-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO)
MIME-Version: 1.0 (generated by SEMI-EPG 1.14.7 - "Harue")
Content-Type: text/plain; charset=US-ASCII
Precedence: bulk

On Tue, 16 May 2023 13:01:59 +0100,
Kunkun Jiang <jiangkunkun@huawei.com> wrote:
> 
> Hi Marc,
> 
> On 2023/5/16 18:15, Marc Zyngier wrote:
> > On Thu, 13 Apr 2023 04:57:17 +0100,
> > Kunkun Jiang <jiangkunkun@huawei.com> wrote:
> >>> Wouldn't it be nice if irq_to_cpuid() could work out whether it is
> >>> dealing with a LPI or a VLPI like it does today, but also directly
> >>> with a VPE? We could then use the same code as derect_lpi_inv(). I
> >>> came up with this the hack below, which is totally untested as I don't
> >>> have access to GICv4.1 HW.
> >>> 
> >>> Could you give it a spin?
> >> Nice, I will test it as soon as possible.
> > Did you ever managed to test this?
> Sorry,I've only been coordinating the GICv4.1 environment in the last
> few days. I have tested it with GICv4 many times, and it doesn't recur.
> However, another call trace occurs with GICv4.1 when the device driver
> is loaded in Guest. I haven't found out why. Maybe you can help analyze it.

I also went back to my patch, and it is a bit bogus (it doesn't even
compile to start with). I've now posted new version[1] that should at
least fix the bug you initially reported.

Can you please test it and reply to it?

Thanks,

	M.

[1] https://lore.kernel.org/r/20230617073242.3199746-1-maz@kernel.org

-- 
Without deviation from the norm, progress is not possible.