From: Stefano Stabellini <sstabellini@kernel.org>
To: Julien Grall <julien.grall@arm.com>
Cc: andre.przywara@arm.com, sstabellini@kernel.org,
steve.capper@arm.com, wei.chen@arm.com, xen-devel@lists.xen.org
Subject: Re: [PATCH v3 14/16] xen/arm: arm64: Add cortex-A57 erratum 832075 workaround
Date: Wed, 22 Jun 2016 11:10:45 +0100 (BST) [thread overview]
Message-ID: <alpine.DEB.2.10.1606221109050.2575@sstabellini-ThinkPad-X260> (raw)
In-Reply-To: <1465315583-1278-15-git-send-email-julien.grall@arm.com>
On Tue, 7 Jun 2016, Julien Grall wrote:
> The ARM erratum 832075 applies to certain revisions of Cortex-A57, one
> of the workarounds is to change device loads into using load-acquire
> semantics.
>
> Use the alternative framework to enable the workaround only on affected
> cores.
>
> Whilst a guest could trigger the deadlock, it can be broken when the
> processor is receiving an interrupt. As the Xen scheduler will always setup
> a timer (firing to every 1ms to 300ms depending on the running time
> slice) on each processor, the deadlock would last only few milliseconds
> and only affects the guest time slice.
>
> Therefore a malicious guest could only hurt itself. Note that all the
> guests should implement/enable the workaround for the affected cores.
>
> Signed-off-by: Julien Grall <julien.grall@arm.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
> ---
> Changes in v2:
> - Update the commit message to explain why it is not necessary
> to take care of possible deadlock from the guest.
> ---
> docs/misc/arm/silicon-errata.txt | 1 +
> xen/arch/arm/Kconfig | 20 ++++++++++++++++++++
> xen/arch/arm/cpuerrata.c | 9 +++++++++
> xen/include/asm-arm/arm64/io.h | 21 +++++++++++++++++----
> xen/include/asm-arm/cpufeature.h | 3 ++-
> xen/include/asm-arm/processor.h | 2 ++
> 6 files changed, 51 insertions(+), 5 deletions(-)
>
> diff --git a/docs/misc/arm/silicon-errata.txt b/docs/misc/arm/silicon-errata.txt
> index fe9f52e..4b2c7c6 100644
> --- a/docs/misc/arm/silicon-errata.txt
> +++ b/docs/misc/arm/silicon-errata.txt
> @@ -46,3 +46,4 @@ stable hypervisors.
> | ARM | Cortex-A53 | #824069 | ARM64_ERRATUM_824069 |
> | ARM | Cortex-A53 | #819472 | ARM64_ERRATUM_819472 |
> | ARM | Cortex-A57 | #852523 | N/A |
> +| ARM | Cortex-A57 | #832075 | ARM64_ERRATUM_832075 |
> diff --git a/xen/arch/arm/Kconfig b/xen/arch/arm/Kconfig
> index a473d9c..e26c4c8 100644
> --- a/xen/arch/arm/Kconfig
> +++ b/xen/arch/arm/Kconfig
> @@ -122,6 +122,26 @@ config ARM64_ERRATUM_819472
> the kernel if an affected CPU is detected.
>
> If unsure, say Y.
> +
> +config ARM64_ERRATUM_832075
> + bool "Cortex-A57: 832075: possible deadlock on mixing exclusive memory accesses with device loads"
> + default y
> + depends on ARM_64
> + help
> + This option adds an alternative code sequence to work around ARM
> + erratum 832075 on Cortex-A57 parts up to r1p2.
> +
> + Affected Cortex-A57 parts might deadlock when exclusive load/store
> + instructions to Write-Back memory are mixed with Device loads.
> +
> + The workaround is to promote device loads to use Load-Acquire
> + semantics.
> + Please note that this does not necessarily enable the workaround,
> + as it depends on the alternative framework, which will only patch
> + the kernel if an affected CPU is detected.
> +
> + If unsure, say Y.
> +
> endmenu
>
> source "common/Kconfig"
> diff --git a/xen/arch/arm/cpuerrata.c b/xen/arch/arm/cpuerrata.c
> index 211b520..48fc87a 100644
> --- a/xen/arch/arm/cpuerrata.c
> +++ b/xen/arch/arm/cpuerrata.c
> @@ -34,6 +34,15 @@ static const struct arm_cpu_capabilities arm_errata[] = {
> MIDR_RANGE(MIDR_CORTEX_A53, 0x00, 0x01),
> },
> #endif
> +#ifdef CONFIG_ARM64_ERRATUM_832075
> + {
> + /* Cortex-A57 r0p0 - r1p2 */
> + .desc = "ARM erratum 832075",
> + .capability = ARM64_WORKAROUND_DEVICE_LOAD_ACQUIRE,
> + MIDR_RANGE(MIDR_CORTEX_A57, 0x00,
> + (1 << MIDR_VARIANT_SHIFT) | 2),
> + },
> +#endif
> {},
> };
>
> diff --git a/xen/include/asm-arm/arm64/io.h b/xen/include/asm-arm/arm64/io.h
> index f80156f..30bfc78 100644
> --- a/xen/include/asm-arm/arm64/io.h
> +++ b/xen/include/asm-arm/arm64/io.h
> @@ -22,6 +22,7 @@
>
> #include <asm/system.h>
> #include <asm/byteorder.h>
> +#include <asm/alternative.h>
>
> /*
> * Generic IO read/write. These perform native-endian accesses.
> @@ -49,28 +50,40 @@ static inline void __raw_writeq(u64 val, volatile void __iomem *addr)
> static inline u8 __raw_readb(const volatile void __iomem *addr)
> {
> u8 val;
> - asm volatile("ldrb %w0, [%1]" : "=r" (val) : "r" (addr));
> + asm volatile(ALTERNATIVE("ldrb %w0, [%1]",
> + "ldarb %w0, [%1]",
> + ARM64_WORKAROUND_DEVICE_LOAD_ACQUIRE)
> + : "=r" (val) : "r" (addr));
> return val;
> }
>
> static inline u16 __raw_readw(const volatile void __iomem *addr)
> {
> u16 val;
> - asm volatile("ldrh %w0, [%1]" : "=r" (val) : "r" (addr));
> + asm volatile(ALTERNATIVE("ldrh %w0, [%1]",
> + "ldarh %w0, [%1]",
> + ARM64_WORKAROUND_DEVICE_LOAD_ACQUIRE)
> + : "=r" (val) : "r" (addr));
> return val;
> }
>
> static inline u32 __raw_readl(const volatile void __iomem *addr)
> {
> u32 val;
> - asm volatile("ldr %w0, [%1]" : "=r" (val) : "r" (addr));
> + asm volatile(ALTERNATIVE("ldr %w0, [%1]",
> + "ldar %w0, [%1]",
> + ARM64_WORKAROUND_DEVICE_LOAD_ACQUIRE)
> + : "=r" (val) : "r" (addr));
> return val;
> }
>
> static inline u64 __raw_readq(const volatile void __iomem *addr)
> {
> u64 val;
> - asm volatile("ldr %0, [%1]" : "=r" (val) : "r" (addr));
> + asm volatile(ALTERNATIVE("ldr %0, [%1]",
> + "ldar %0, [%1]",
> + ARM64_WORKAROUND_DEVICE_LOAD_ACQUIRE)
> + : "=r" (val) : "r" (addr));
> return val;
> }
>
> diff --git a/xen/include/asm-arm/cpufeature.h b/xen/include/asm-arm/cpufeature.h
> index 474a778..78e2263 100644
> --- a/xen/include/asm-arm/cpufeature.h
> +++ b/xen/include/asm-arm/cpufeature.h
> @@ -36,8 +36,9 @@
> #define cpu_has_security (boot_cpu_feature32(security) > 0)
>
> #define ARM64_WORKAROUND_CLEAN_CACHE 0
> +#define ARM64_WORKAROUND_DEVICE_LOAD_ACQUIRE 1
>
> -#define ARM_NCAPS 1
> +#define ARM_NCAPS 2
>
> #ifndef __ASSEMBLY__
>
> diff --git a/xen/include/asm-arm/processor.h b/xen/include/asm-arm/processor.h
> index 5089bfd..1708253 100644
> --- a/xen/include/asm-arm/processor.h
> +++ b/xen/include/asm-arm/processor.h
> @@ -47,8 +47,10 @@
> #define ARM_CPU_IMP_ARM 0x41
>
> #define ARM_CPU_PART_CORTEX_A53 0xD03
> +#define ARM_CPU_PART_CORTEX_A57 0xD07
>
> #define MIDR_CORTEX_A53 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A53)
> +#define MIDR_CORTEX_A57 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A57)
>
> /* MPIDR Multiprocessor Affinity Register */
> #define _MPIDR_UP (30)
> --
> 1.9.1
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
next prev parent reply other threads:[~2016-06-22 10:10 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-07 16:06 [PATCH v3 00/16] xen/arm: Introduce alternative runtime patching for ARM64 Julien Grall
2016-06-07 16:06 ` [PATCH v3 01/16] xen/arm: Makefile: Sort the entries alphabetically Julien Grall
2016-06-07 16:06 ` [PATCH v3 02/16] xen/arm: Include the header asm-arm/system.h in asm-arm/page.h Julien Grall
2016-06-07 16:06 ` [PATCH v3 03/16] xen/arm: Add macros to handle the MIDR Julien Grall
2016-06-07 16:06 ` [PATCH v3 04/16] xen/arm: Add cpu_hwcap bitmap Julien Grall
2016-06-07 16:06 ` [PATCH v3 05/16] xen/arm64: Add an helper to invalidate all instruction caches Julien Grall
2016-06-07 16:06 ` [PATCH v3 06/16] xen/arm: arm64: Move the define BRK_BUG_FRAME into a separate header Julien Grall
2016-06-07 16:06 ` [PATCH v3 07/16] xen/arm: arm64: Reserve a brk immediate to fault on purpose Julien Grall
2016-06-07 16:06 ` [PATCH v3 08/16] xen/arm: arm64: Add helpers to decode and encode branch instructions Julien Grall
2016-06-07 16:06 ` [PATCH v3 09/16] xen/arm: Introduce alternative runtime patching Julien Grall
2016-06-07 17:24 ` Konrad Rzeszutek Wilk
2016-06-08 9:39 ` Julien Grall
2016-06-08 18:17 ` Konrad Rzeszutek Wilk
2016-06-08 18:22 ` Julien Grall
2016-06-08 18:35 ` Konrad Rzeszutek Wilk
2016-06-09 13:33 ` Julien Grall
2016-06-07 16:06 ` [PATCH v3 10/16] xen/arm: cpufeature: Provide an helper to check if a capability is supported Julien Grall
2016-06-22 9:59 ` Stefano Stabellini
2016-06-22 10:09 ` Julien Grall
2016-06-22 10:14 ` Stefano Stabellini
2016-06-07 16:06 ` [PATCH v3 11/16] xen/arm: Detect silicon revision and set cap bits accordingly Julien Grall
2016-06-22 10:00 ` Stefano Stabellini
2016-06-07 16:06 ` [PATCH v3 12/16] xen/arm: Document the errata implemented in Xen Julien Grall
2016-06-22 10:03 ` Stefano Stabellini
2016-06-07 16:06 ` [PATCH v3 13/16] xen/arm: arm64: Add Cortex-A53 cache errata workaround Julien Grall
2016-06-22 10:04 ` Stefano Stabellini
2016-06-07 16:06 ` [PATCH v3 14/16] xen/arm: arm64: Add cortex-A57 erratum 832075 workaround Julien Grall
2016-06-22 10:10 ` Stefano Stabellini [this message]
2016-06-07 16:06 ` [PATCH v3 15/16] xen/arm: traps: Don't inject a fault if the translation VA -> IPA fails Julien Grall
2016-06-07 16:06 ` [PATCH v3 16/16] xen/arm: arm64: Document Cortex-A57 erratum 834220 Julien Grall
2016-06-14 12:15 ` Julien Grall
2016-06-22 10:08 ` Stefano Stabellini
2016-06-22 10:18 ` Julien Grall
2016-06-22 10:37 ` Stefano Stabellini
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=alpine.DEB.2.10.1606221109050.2575@sstabellini-ThinkPad-X260 \
--to=sstabellini@kernel.org \
--cc=andre.przywara@arm.com \
--cc=julien.grall@arm.com \
--cc=steve.capper@arm.com \
--cc=wei.chen@arm.com \
--cc=xen-devel@lists.xen.org \
/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;
as well as URLs for NNTP newsgroup(s).