Re: [PATCH for-4.11 00/10] x86: Improvements and fixes to Spectre handling

From: Juergen Gross <jgross@suse.com>
To: Andrew Cooper <andrew.cooper3@citrix.com>,
	Xen-devel <xen-devel@lists.xen.org>
Cc: "Wei Liu" <wei.liu2@citrix.com>,
	"Anthony Liguori" <aliguori@amazon.com>,
	"Zhenzhong Duan" <zhenzhong.duan@oracle.com>,
	"Martin Pohlack" <mpohlack@amazon.de>,
	"Jan Beulich" <JBeulich@suse.com>,
	"Boris Ostrovsky" <boris.ostrovsky@oracle.com>,
	"David Woodhouse" <dwmw@amazon.co.uk>,
	"Roger Pau Monné" <roger.pau@citrix.com>
Subject: Re: [PATCH for-4.11 00/10] x86: Improvements and fixes to Spectre handling
Date: Tue, 15 May 2018 20:25:48 +0200	[thread overview]
Message-ID: <4531d062-1db4-6e5d-3e2f-df33d2ec4515@suse.com> (raw)
In-Reply-To: <1526035094-14343-1-git-send-email-andrew.cooper3@citrix.com>

On 11/05/18 12:38, Andrew Cooper wrote:
> In hindsight, the end result of the Spectre mitigations aren't as great as I'd
> hoped, and have several inefficiencies.  Also, the `bti=` command line option
> isn't as flexible as intended.
> 
> This series does four things:
> 
>   1) Some internal cleanup, for clarity and to help the other features
>   2) Introduce `spec-ctrl=no-pv` mode.  XenServer's performance measurements
>      see a 10% net/disk performance improvement in some production scenarios.
>   3) Introduce the ability to use IBPB-only mode for guests.  This was
>      discussed by Amazon during the Spectre work, but I don't have any
>      performance numbers to hand.
>   4) Avoid imposing IBRS mode while dom0 is booting.  This was reported by
>      Oracle on the list, and speeds up boot time on some servers by 50s.
> 
> I know this series is rather late for 4.11, but seeing as I've managed to
> complete it before 4.12 opens, it should be considered at this point, as all
> of the Spectre code is new in 4.11.
> 
> Andrew Cooper (10):
>   x86/spec_ctrl: Read MSR_ARCH_CAPABILITIES only once
>   x86/spec_ctrl: Express Xen's choice of MSR_SPEC_CTRL value as a variable
>   x86/spec_ctrl: Merge bti_ist_info and use_shadow_spec_ctrl into spec_ctrl_flags
>   x86/spec_ctrl: Fold the XEN_IBRS_{SET,CLEAR} ALTERNATIVES together
>   x86/spec_ctrl: Rename bits of infrastructure to avoid NATIVE and VMEXIT
>   x86/spec_ctrl: Split X86_FEATURE_SC_MSR into PV and HVM variants
>   x86/spec_ctrl: Explicitly set Xen's default MSR_SPEC_CTRL value
>   x86/cpuid: Improvements to guest policies for speculative sidechannel features
>   x86/spec_ctrl: Introduce a new `spec-ctrl=` command line argument to replace `bti=`
>   x86/spec_ctrl: Elide MSR_SPEC_CTRL handling in idle context when possible
> 
>  docs/misc/xen-command-line.markdown |  49 +++++++
>  xen/arch/x86/acpi/power.c           |   4 +-
>  xen/arch/x86/cpuid.c                |  60 +++++----
>  xen/arch/x86/hvm/svm/entry.S        |   4 +-
>  xen/arch/x86/hvm/vmx/entry.S        |   4 +-
>  xen/arch/x86/setup.c                |   7 +
>  xen/arch/x86/smpboot.c              |   8 ++
>  xen/arch/x86/spec_ctrl.c            | 258 ++++++++++++++++++++++++++++--------
>  xen/arch/x86/x86_64/asm-offsets.c   |   4 +-
>  xen/arch/x86/x86_64/compat/entry.S  |   2 +-
>  xen/arch/x86/x86_64/entry.S         |   2 +-
>  xen/include/asm-x86/cpufeatures.h   |   9 +-
>  xen/include/asm-x86/current.h       |   4 +-
>  xen/include/asm-x86/spec_ctrl.h     |  20 +--
>  xen/include/asm-x86/spec_ctrl_asm.h | 131 +++++++++---------
>  15 files changed, 396 insertions(+), 170 deletions(-)
> 

For the series:

Release-acked-by: Juergen Gross <jgross@suse.com>

Juergen

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel