linux-hyperv.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Wei Liu <wei.liu@kernel.org>
To: Praveen Kumar <kumarpraveen@linux.microsoft.com>
Cc: linux-hyperv@vger.kernel.org, linux-kernel@vger.kernel.org,
	kys@microsoft.com, haiyangz@microsoft.com,
	sthemmin@microsoft.com, wei.liu@kernel.org, decui@microsoft.com,
	tglx@linutronix.de, mingo@redhat.com, bp@alien8.de,
	x86@kernel.org, hpa@zytor.com, viremana@linux.microsoft.com,
	sunilmut@microsoft.com, nunodasneves@linux.microsoft.com
Subject: Re: [PATCH] hyperv: root partition faults writing to VP ASSIST MSR PAGE
Date: Tue, 20 Jul 2021 11:20:11 +0000	[thread overview]
Message-ID: <20210720112011.7nxhiy6iyz4gz3j5@liuwe-devbox-debian-v2> (raw)
In-Reply-To: <20210719185126.3740-1-kumarpraveen@linux.microsoft.com>

The commit message needs a bit of work.

On Tue, Jul 20, 2021 at 12:21:26AM +0530, Praveen Kumar wrote:
> The root partition is not supposed to write to VP ASSIST PAGE as this MSR
> is specific to Guest VP, and thus below stack is observed.
> 

Yes, root kernel is supposed to write to this MSR, but that's not
because this MSR is specific to children (guest) partitions. It is just
that for root this is read-only.

You should mention VP assist pages for root are pre-determined by the
hypervisor. Root kernel is not allowed to change them to different
locations.

> [    2.778197] unchecked MSR access error: WRMSR to 0x40000073 (tried to write 0x0000000145ac5001) at rIP: 0xffffffff810c1084 (native_write_msr+0x4/0x30)
> [    2.784867] Call Trace:
> [    2.791507]  hv_cpu_init+0xf1/0x1c0
> [    2.798144]  ? hyperv_report_panic+0xd0/0xd0
> [    2.804806]  cpuhp_invoke_callback+0x11a/0x440
> [    2.811465]  ? hv_resume+0x90/0x90
> [    2.818137]  cpuhp_issue_call+0x126/0x130
> [    2.824782]  __cpuhp_setup_state_cpuslocked+0x102/0x2b0
> [    2.831427]  ? hyperv_report_panic+0xd0/0xd0
> [    2.838075]  ? hyperv_report_panic+0xd0/0xd0
> [    2.844723]  ? hv_resume+0x90/0x90
> [    2.851375]  __cpuhp_setup_state+0x3d/0x90
> [    2.858030]  hyperv_init+0x14e/0x410
> [    2.864689]  ? enable_IR_x2apic+0x190/0x1a0
> [    2.871349]  apic_intr_mode_init+0x8b/0x100
> [    2.878017]  x86_late_time_init+0x20/0x30
> [    2.884675]  start_kernel+0x459/0x4fb
> [    2.891329]  secondary_startup_64_no_verify+0xb0/0xbb
> 
> Root partition actually shares the VP ASSIST page with hypervisor, and

So do children partitions. This page is by design shared between
hypervisor and any partitions that use it.

> thus as a solution, this patch memremaps the memory from hypervisor
> during hv_cpu_init and unmaps during hv_cpu_die calls.
> 
> Further, this patch also resolve some error handling and checkpatch
> errors
> 
> Signed-off-by: Praveen Kumar <kumarpraveen@linux.microsoft.com>
> ---
>  arch/x86/hyperv/hv_init.c | 57 +++++++++++++++++++++++++++------------
>  1 file changed, 40 insertions(+), 17 deletions(-)
> 
> diff --git a/arch/x86/hyperv/hv_init.c b/arch/x86/hyperv/hv_init.c
> index 6f247e7e07eb..292b17e0b173 100644
> --- a/arch/x86/hyperv/hv_init.c
> +++ b/arch/x86/hyperv/hv_init.c
> @@ -44,7 +44,7 @@ EXPORT_SYMBOL_GPL(hv_vp_assist_page);
>  
>  static int hv_cpu_init(unsigned int cpu)
>  {
> -	struct hv_vp_assist_page **hvp = &hv_vp_assist_page[smp_processor_id()];
> +	struct hv_vp_assist_page **hvp = NULL;
>  	int ret;
>  
>  	ret = hv_common_cpu_init(cpu);
> @@ -54,25 +54,43 @@ static int hv_cpu_init(unsigned int cpu)
>  	if (!hv_vp_assist_page)
>  		return 0;
>  
> +	hvp = &hv_vp_assist_page[smp_processor_id()];
> +

Why is this needed? Is it because of checkpatch?

>  	/*
> -	 * The VP ASSIST PAGE is an "overlay" page (see Hyper-V TLFS's Section
> -	 * 5.2.1 "GPA Overlay Pages"). Here it must be zeroed out to make sure
> -	 * we always write the EOI MSR in hv_apic_eoi_write() *after* the
> -	 * EOI optimization is disabled in hv_cpu_die(), otherwise a CPU may
> -	 * not be stopped in the case of CPU offlining and the VM will hang.
> +	 * For Root partition we need to map the hypervisor VP ASSIST PAGE
> +	 * instead of allocating a new page.
>  	 */
> -	if (!*hvp) {
> -		*hvp = __vmalloc(PAGE_SIZE, GFP_KERNEL | __GFP_ZERO);
> -	}

This path suggests that it is possible to enter this function with *hvp
already set.

The new path for root is missing this check.

> +	if (hv_root_partition &&
> +	    ms_hyperv.features & HV_MSR_APIC_ACCESS_AVAILABLE) {

Is HV_MSR_APIC_ACCESS_AVAILABLE a root only flag? Shouldn't non-root
kernel check this too?

Wei.

  reply	other threads:[~2021-07-20 11:20 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-19 18:51 Praveen Kumar
2021-07-20 11:20 ` Wei Liu [this message]
2021-07-20 13:25   ` Praveen Kumar
2021-07-20 13:35     ` Wei Liu
2021-07-20 16:20       ` Michael Kelley
2021-07-20 16:29         ` Wei Liu
2021-07-21  4:10           ` Michael Kelley
2021-07-21  7:12             ` Praveen Kumar
2021-07-21 10:10               ` Wei Liu
2021-07-21 11:32                 ` Praveen Kumar
2021-07-21 15:07                   ` Michael Kelley

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210720112011.7nxhiy6iyz4gz3j5@liuwe-devbox-debian-v2 \
    --to=wei.liu@kernel.org \
    --cc=bp@alien8.de \
    --cc=decui@microsoft.com \
    --cc=haiyangz@microsoft.com \
    --cc=hpa@zytor.com \
    --cc=kumarpraveen@linux.microsoft.com \
    --cc=kys@microsoft.com \
    --cc=linux-hyperv@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=nunodasneves@linux.microsoft.com \
    --cc=sthemmin@microsoft.com \
    --cc=sunilmut@microsoft.com \
    --cc=tglx@linutronix.de \
    --cc=viremana@linux.microsoft.com \
    --cc=x86@kernel.org \
    --subject='Re: [PATCH] hyperv: root partition faults writing to VP ASSIST MSR PAGE' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).