Received-SPF: pass (google.com: domain of linux-kernel+bounces-145847-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Message-ID: <81d05bd9-ef8f-4a3a-a5df-384a82be49f5@linux.intel.com>
Date: Mon, 15 Apr 2024 12:48:53 -0700
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH v3 13/14] x86/sev: Hide SVSM attestation entries if not
 running under an SVSM
To: Tom Lendacky <thomas.lendacky@amd.com>, linux-kernel@vger.kernel.org,
 x86@kernel.org, linux-coco@lists.linux.dev, svsm-devel@coconut-svsm.dev
Cc: Thomas Gleixner <tglx@linutronix.de>, Ingo Molnar <mingo@redhat.com>,
 Borislav Petkov <bp@alien8.de>, Dave Hansen <dave.hansen@linux.intel.com>,
 "H. Peter Anvin" <hpa@zytor.com>, Andy Lutomirski <luto@kernel.org>,
 Peter Zijlstra <peterz@infradead.org>,
 Dan Williams <dan.j.williams@intel.com>, Michael Roth
 <michael.roth@amd.com>, Ashish Kalra <ashish.kalra@amd.com>,
 Joel Becker <jlbec@evilplan.org>, Christoph Hellwig <hch@lst.de>
References: <cover.1711405593.git.thomas.lendacky@amd.com>
 <67893f352bc54de61160bfe251cba0bdf0431f37.1711405593.git.thomas.lendacky@amd.com>
 <1f6f4477-0f2a-434a-8c89-3b5d51d61581@linux.intel.com>
 <19d69960-e548-a2e6-87d9-c463f2851613@amd.com>
 <e286883d-40ef-d749-26a8-7ec6fbd81eae@amd.com>
Content-Language: en-US
From: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
In-Reply-To: <e286883d-40ef-d749-26a8-7ec6fbd81eae@amd.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Hi,

On 4/15/24 12:16 PM, Tom Lendacky wrote:
> On 4/12/24 10:52, Tom Lendacky wrote:
>> On 4/9/24 13:12, Kuppuswamy Sathyanarayanan wrote:
>>> On 3/25/24 3:26 PM, Tom Lendacky wrote:
>>>> Config-fs provides support to hide individual attribute entries. Using
>>>> this support, base the display of the SVSM related entries on the presence
>>>> of an SVSM.
>>>>
>>>> Cc: Joel Becker <jlbec@evilplan.org>
>>>> Cc: Christoph Hellwig <hch@lst.de>
>>>> Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
>>>> ---
>>>>   arch/x86/coco/core.c        |  4 ++++
>>>>   drivers/virt/coco/tsm.c     | 14 ++++++++++----
>>>>   include/linux/cc_platform.h |  8 ++++++++
>>>>   3 files changed, 22 insertions(+), 4 deletions(-)
>>>>
>
>>>
>>> Any comment about the following query? I think introducing a CC flag for this use
>>> case is over kill.
>>>
>>> https://lore.kernel.org/lkml/6b90b223-46e0-4e6d-a17c-5caf72e3c949@linux.intel.com/
>>
>> If you don't think TDX will be able to make use of the SVSM attribute I can look at adding a callback. But I was waiting to see if anyone else had comments, for or against, before re-doing it all.
>>
>
> What about something like this (applied on top of patch 13):
>
> diff --git a/arch/x86/coco/core.c b/arch/x86/coco/core.c
> index efa0f648f754..d07be9d05cd0 100644
> --- a/arch/x86/coco/core.c
> +++ b/arch/x86/coco/core.c
> @@ -12,7 +12,6 @@
>  
>  #include <asm/coco.h>
>  #include <asm/processor.h>
> -#include <asm/sev.h>
>  
>  enum cc_vendor cc_vendor __ro_after_init = CC_VENDOR_NONE;
>  u64 cc_mask __ro_after_init;
> @@ -79,9 +78,6 @@ static bool noinstr amd_cc_platform_has(enum cc_attr attr)
>      case CC_ATTR_GUEST_STATE_ENCRYPT:
>          return sev_status & MSR_AMD64_SEV_ES_ENABLED;
>  
> -    case CC_ATTR_GUEST_SVSM_PRESENT:
> -        return snp_get_vmpl();
> -
>      /*
>       * With SEV, the rep string I/O instructions need to be unrolled
>       * but SEV-ES supports them through the #VC handler.
> diff --git a/drivers/virt/coco/sev-guest/sev-guest.c b/drivers/virt/coco/sev-guest/sev-guest.c
> index 0d2c9926a97c..68c881a50026 100644
> --- a/drivers/virt/coco/sev-guest/sev-guest.c
> +++ b/drivers/virt/coco/sev-guest/sev-guest.c
> @@ -1036,6 +1036,17 @@ static int sev_report_new(struct tsm_report *report, void *data)
>      return 0;
>  }
>  
> +static bool sev_tsm_visibility(enum tsm_type type)
> +{
> +    /* Check for SVSM-related attributes */
> +    switch (type) {
> +    case TSM_TYPE_SERVICE_PROVIDER:
> +        return snp_get_vmpl();
> +    default:
> +        return false;
> +    }
> +}
> +
>  static struct tsm_ops sev_tsm_ops = {
>      .name = KBUILD_MODNAME,
>      .report_new = sev_report_new,
> @@ -1126,7 +1137,8 @@ static int __init sev_guest_probe(struct platform_device *pdev)
>      /* Set the privlevel_floor attribute based on the current VMPL */
>      sev_tsm_ops.privlevel_floor = snp_get_vmpl();
>  
> -    ret = tsm_register(&sev_tsm_ops, snp_dev, &tsm_report_extra_type);
> +    ret = tsm_register(&sev_tsm_ops, snp_dev, &tsm_report_extra_type,
> +               sev_tsm_visibility);
>      if (ret)
>          goto e_free_cert_data;
>  
> diff --git a/drivers/virt/coco/tdx-guest/tdx-guest.c b/drivers/virt/coco/tdx-guest/tdx-guest.c
> index 1253bf76b570..0fd8e60d7bee 100644
> --- a/drivers/virt/coco/tdx-guest/tdx-guest.c
> +++ b/drivers/virt/coco/tdx-guest/tdx-guest.c
> @@ -301,7 +301,7 @@ static int __init tdx_guest_init(void)
>          goto free_misc;
>      }
>  
> -    ret = tsm_register(&tdx_tsm_ops, NULL, NULL);
> +    ret = tsm_register(&tdx_tsm_ops, NULL, NULL, NULL);
>      if (ret)
>          goto free_quote;
>  
> diff --git a/drivers/virt/coco/tsm.c b/drivers/virt/coco/tsm.c
> index d30471874e87..e73840aed13d 100644
> --- a/drivers/virt/coco/tsm.c
> +++ b/drivers/virt/coco/tsm.c
> @@ -16,6 +16,7 @@ static struct tsm_provider {
>      const struct tsm_ops *ops;
>      const struct config_item_type *type;
>      void *data;
> +    tsm_visibility_t visibility;
>  } provider;
>  static DECLARE_RWSEM(tsm_rwsem);
>  
> @@ -64,10 +65,13 @@ static struct tsm_report_state *to_state(struct tsm_report *report)
>      return container_of(report, struct tsm_report_state, report);
>  }
>  
> -static bool provider_visibility(const struct config_item *item,
> -                const struct configfs_attribute *attr)
> +static bool service_provider_visibility(const struct config_item *item,
> +                    const struct configfs_attribute *attr)
>  {
> -    return cc_platform_has(CC_ATTR_GUEST_SVSM_PRESENT);
> +    if (!provider.visibility)
> +        return true;
> +
> +    return provider.visibility(TSM_TYPE_SERVICE_PROVIDER);
>  }
>  
>  static int try_advance_write_generation(struct tsm_report *report)
> @@ -150,7 +154,7 @@ static ssize_t tsm_report_service_provider_store(struct config_item *cfg,
>  
>      return len;
>  }
> -CONFIGFS_ATTR_VISIBLE_WO(tsm_report_, service_provider, provider_visibility);
> +CONFIGFS_ATTR_VISIBLE_WO(tsm_report_, service_provider, service_provider_visibility);
>  
>  static ssize_t tsm_report_service_guid_store(struct config_item *cfg,
>                           const char *buf, size_t len)
> @@ -171,7 +175,7 @@ static ssize_t tsm_report_service_guid_store(struct config_item *cfg,
>  
>      return len;
>  }
> -CONFIGFS_ATTR_VISIBLE_WO(tsm_report_, service_guid, provider_visibility);
> +CONFIGFS_ATTR_VISIBLE_WO(tsm_report_, service_guid, service_provider_visibility);
>  
>  static ssize_t tsm_report_service_manifest_version_store(struct config_item *cfg,
>                               const char *buf, size_t len)
> @@ -192,7 +196,7 @@ static ssize_t tsm_report_service_manifest_version_store(struct config_item *cfg
>  
>      return len;
>  }
> -CONFIGFS_ATTR_VISIBLE_WO(tsm_report_, service_manifest_version, provider_visibility);
> +CONFIGFS_ATTR_VISIBLE_WO(tsm_report_, service_manifest_version, service_provider_visibility);
>  
>  static ssize_t tsm_report_inblob_write(struct config_item *cfg,
>                         const void *buf, size_t count)
> @@ -339,7 +343,8 @@ static ssize_t tsm_report_manifestblob_read(struct config_item *cfg, void *buf,
>  
>      return tsm_report_read(report, buf, count, TSM_MANIFEST);
>  }
> -CONFIGFS_BIN_ATTR_VISIBLE_RO(tsm_report_, manifestblob, NULL, TSM_OUTBLOB_MAX, provider_visibility);
> +CONFIGFS_BIN_ATTR_VISIBLE_RO(tsm_report_, manifestblob, NULL, TSM_OUTBLOB_MAX,
> +                 service_provider_visibility);
>  
>  #define TSM_DEFAULT_ATTRS() \
>      &tsm_report_attr_generation, \
> @@ -449,7 +454,8 @@ static struct configfs_subsystem tsm_configfs = {
>  };
>  
>  int tsm_register(const struct tsm_ops *ops, void *priv,
> -         const struct config_item_type *type)
> +         const struct config_item_type *type,
> +         tsm_visibility_t visibility)
>  {
>      const struct tsm_ops *conflict;
>  
> @@ -468,6 +474,7 @@ int tsm_register(const struct tsm_ops *ops, void *priv,
>      provider.ops = ops;
>      provider.data = priv;
>      provider.type = type;
> +    provider.visibility = visibility;
>      return 0;
>  }
>  EXPORT_SYMBOL_GPL(tsm_register);
> @@ -480,6 +487,7 @@ int tsm_unregister(const struct tsm_ops *ops)
>      provider.ops = NULL;
>      provider.data = NULL;
>      provider.type = NULL;
> +    provider.visibility = NULL;
>      return 0;
>  }
>  EXPORT_SYMBOL_GPL(tsm_unregister);
> diff --git a/include/linux/cc_platform.h b/include/linux/cc_platform.h
> index f1b4266c1484..cb0d6cd1c12f 100644
> --- a/include/linux/cc_platform.h
> +++ b/include/linux/cc_platform.h
> @@ -90,14 +90,6 @@ enum cc_attr {
>       * Examples include TDX Guest.
>       */
>      CC_ATTR_HOTPLUG_DISABLED,
> -
> -    /**
> -     * @CC_ATTR_GUEST_SVSM_PRESENT: Guest is running under an SVSM
> -     *
> -     * The platform/OS is running as a guest/virtual machine and is
> -     * running under a Secure VM Service Module (SVSM).
> -     */
> -    CC_ATTR_GUEST_SVSM_PRESENT,
>  };
>  
>  #ifdef CONFIG_ARCH_HAS_CC_PLATFORM
> diff --git a/include/linux/tsm.h b/include/linux/tsm.h
> index 27cc97fe8dcd..5aaf626d178d 100644
> --- a/include/linux/tsm.h
> +++ b/include/linux/tsm.h
> @@ -74,7 +74,20 @@ extern const struct config_item_type tsm_report_default_type;
>  /* publish @privlevel, @privlevel_floor, and @auxblob attributes */
>  extern const struct config_item_type tsm_report_extra_type;
>  
> +/*
> + * Used to indicate the attribute group type to the visibility callback to
> + * avoid the callback having to examine the attribute name.

Checking the attribute name will give more flexibility, right? Since it is one time
check, it should not be costly, right?

> +enum tsm_type {
> +    TSM_TYPE_SERVICE_PROVIDER,
> +
> +    TSM_TYPE_MAX
> +};
> +
> +typedef bool (*tsm_visibility_t)(enum tsm_type type);
> +
>  int tsm_register(const struct tsm_ops *ops, void *priv,
> -         const struct config_item_type *type);
> +         const struct config_item_type *type,
> +         tsm_visibility_t visibility);
>  int tsm_unregister(const struct tsm_ops *ops);
>  #endif /* __TSM_H */
>

Why not add a callback in tsm_ops?

>> Thanks,
>> Tom
>>
-- 
Sathyanarayanan Kuppuswamy
Linux Kernel Developer