public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Sean Christopherson <sean.j.christopherson@intel.com>
To: nadav.amit@gmail.com
Cc: Paolo Bonzini <pbonzini@redhat.com>, kvm@vger.kernel.org
Subject: Re: [kvm-unit-tests PATCH] x86: Disable cache before relocating local APIC
Date: Tue, 30 Apr 2019 13:27:04 -0700	[thread overview]
Message-ID: <20190430202704.GB4523@linux.intel.com> (raw)
In-Reply-To: <20190430122701.41069-1-nadav.amit@gmail.com>

On Tue, Apr 30, 2019 at 05:27:01AM -0700, nadav.amit@gmail.com wrote:
> From: Nadav Amit <nadav.amit@gmail.com>
> 
> According to the SDM, during initialization, the BSP "Switches to
> protected mode and ensures that the APIC address space is mapped to the
> strong uncacheable (UC) memory type."
> 
> This requirement is not followed when the tests relocate the APIC. Set
> the cache-disable flag while the APIC base is reprogrammed. According
> to the SDM, the MTRRs should be modified as well, but it seems somewhat
> complicated to do that and probably unnecessary.

Alternatively, what about defining ALTERNATE_APIC_BASE to a sane value
that is guaranteed to be UC (assuming BIOS isn't being mean)?  Maybe a
use well-known address, e.g. 0xfed40000 from the TPM, so as to avoid any
other unwanted side effects.

> Cc: Sean Christopherson <sean.j.christopherson@intel.com>
> Signed-off-by: Nadav Amit <nadav.amit@gmail.com>
> ---
>  lib/x86/processor.h | 1 +
>  x86/apic.c          | 7 +++++++
>  2 files changed, 8 insertions(+)
> 
> diff --git a/lib/x86/processor.h b/lib/x86/processor.h
> index 916e67d..59137da 100644
> --- a/lib/x86/processor.h
> +++ b/lib/x86/processor.h
> @@ -29,6 +29,7 @@
>  #define X86_CR0_TS     0x00000008
>  #define X86_CR0_WP     0x00010000
>  #define X86_CR0_AM     0x00040000
> +#define X86_CR0_CD     0x40000000
>  #define X86_CR0_PG     0x80000000
>  #define X86_CR3_PCID_MASK 0x00000fff
>  #define X86_CR4_TSD    0x00000004
> diff --git a/x86/apic.c b/x86/apic.c
> index de5990c..de4a181 100644
> --- a/x86/apic.c
> +++ b/x86/apic.c
> @@ -165,8 +165,13 @@ static void test_apicbase(void)
>  {
>      u64 orig_apicbase = rdmsr(MSR_IA32_APICBASE);
>      u32 lvr = apic_read(APIC_LVR);
> +    u64 cr0 = read_cr0();
>      u64 value;
>  
> +    /* Disable caching to prevent the APIC from being cacheable */
> +    write_cr0(cr0 | X86_CR0_CD);
> +    asm volatile ("wbinvd" ::: "memory");
> +
>      wrmsr(MSR_IA32_APICBASE, orig_apicbase & ~(APIC_EN | APIC_EXTD));
>      wrmsr(MSR_IA32_APICBASE, ALTERNATE_APIC_BASE | APIC_BSP | APIC_EN);
>  
> @@ -186,6 +191,8 @@ static void test_apicbase(void)
>      wrmsr(MSR_IA32_APICBASE, orig_apicbase);
>      apic_write(APIC_SPIV, 0x1ff);
>  
> +    write_cr0(cr0);
> +
>      report_prefix_pop();
>  }
>  
> -- 
> 2.17.1
> 

  reply	other threads:[~2019-04-30 20:27 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-04-30 12:27 [kvm-unit-tests PATCH] x86: Disable cache before relocating local APIC nadav.amit
2019-04-30 20:27 ` Sean Christopherson [this message]
2019-04-30 21:13   ` Nadav Amit

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=20190430202704.GB4523@linux.intel.com \
    --to=sean.j.christopherson@intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=nadav.amit@gmail.com \
    --cc=pbonzini@redhat.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox