Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 1 Apr 2020 15:53:16 +0200
From:   Jean-Philippe Brucker <jean-philippe@linaro.org>
To:     Jacob Pan <jacob.jun.pan@linux.intel.com>
Cc:     Joerg Roedel <joro@8bytes.org>,
        Alex Williamson <alex.williamson@redhat.com>,
        Lu Baolu <baolu.lu@linux.intel.com>,
        iommu@lists.linux-foundation.org,
        LKML <linux-kernel@vger.kernel.org>,
        David Woodhouse <dwmw2@infradead.org>,
        Jean-Philippe Brucker <jean-philippe@linaro.com>,
        Yi Liu <yi.l.liu@intel.com>,
        "Tian, Kevin" <kevin.tian@intel.com>,
        Raj Ashok <ashok.raj@intel.com>,
        Christoph Hellwig <hch@infradead.org>,
        Jonathan Cameron <jic23@kernel.org>,
        Eric Auger <eric.auger@redhat.com>
Subject: Re: [PATCH 05/10] iommu/ioasid: Create an IOASID set for host SVA use
Message-ID: <20200401135316.GF882512@myrica>
References: <1585158931-1825-1-git-send-email-jacob.jun.pan@linux.intel.com>
 <1585158931-1825-6-git-send-email-jacob.jun.pan@linux.intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1585158931-1825-6-git-send-email-jacob.jun.pan@linux.intel.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Mar 25, 2020 at 10:55:26AM -0700, Jacob Pan wrote:
> Bare metal SVA allocates IOASIDs for native process addresses. This
> should be separated from VM allocated IOASIDs thus under its own set.
> 
> This patch creates a system IOASID set with its quota set to PID_MAX.
> This is a reasonable default in that SVM capable devices can only bind
> to limited user processes.

Yes realistically there won't be more than PID_MAX_DEFAULT=0x8000 bound
address spaces. My machine uses a PID_MAX of 4 million though, so in
theory more than 0x8000 processes may want a bond. On Arm the limit of
shared contexts per VM is currently a little less than 0x10000 (which is
the number of CPU ASIDs).

But quotas are only necessary for VMs, when the host shares the PASID
space with them (which isn't a use-case for Arm systems as far as I know,
each VM gets its own PASID space). Could we have quota-free IOASID sets
for the host?

For the SMMU I'd like to allocate two sets, one SVA and one private for
auxiliary domains, and I don't think giving either a quota makes much
sense at the moment. There can be systems using only SVA and systems using
only private PASIDs. I think it should be first-come-first-served until
admins want a knob to define a policy themselves, based on cgroups for
example.

> Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
> ---
>  drivers/iommu/intel-iommu.c | 8 +++++++-
>  drivers/iommu/ioasid.c      | 9 +++++++++
>  include/linux/ioasid.h      | 9 +++++++++
>  3 files changed, 25 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
> index ec3fc121744a..af7a1ef7b31e 100644
> --- a/drivers/iommu/intel-iommu.c
> +++ b/drivers/iommu/intel-iommu.c
> @@ -3511,8 +3511,14 @@ static int __init init_dmars(void)
>  		goto free_iommu;
>  
>  	/* PASID is needed for scalable mode irrespective to SVM */
> -	if (intel_iommu_sm)
> +	if (intel_iommu_sm) {
>  		ioasid_install_capacity(intel_pasid_max_id);
> +		/* We should not run out of IOASIDs at boot */
> +		if (ioasid_alloc_system_set(PID_MAX_DEFAULT)) {
> +			pr_err("Failed to enable host PASID allocator\n");
> +			intel_iommu_sm = 0;
> +		}
> +	}
>  
>  	/*
>  	 * for each drhd
> diff --git a/drivers/iommu/ioasid.c b/drivers/iommu/ioasid.c
> index 6265d2dbbced..9135af171a7c 100644
> --- a/drivers/iommu/ioasid.c
> +++ b/drivers/iommu/ioasid.c
> @@ -39,6 +39,9 @@ struct ioasid_data {
>  static ioasid_t ioasid_capacity;
>  static ioasid_t ioasid_capacity_avail;
>  
> +int system_ioasid_sid;
> +static DECLARE_IOASID_SET(system_ioasid);
> +
>  /* System capacity can only be set once */
>  void ioasid_install_capacity(ioasid_t total)
>  {
> @@ -51,6 +54,12 @@ void ioasid_install_capacity(ioasid_t total)
>  }
>  EXPORT_SYMBOL_GPL(ioasid_install_capacity);
>  
> +int ioasid_alloc_system_set(int quota)
> +{
> +	return ioasid_alloc_set(&system_ioasid, quota, &system_ioasid_sid);
> +}
> +EXPORT_SYMBOL_GPL(ioasid_alloc_system_set);

I think this helper could stay in the VT-d driver for the moment. If the
SMMU driver ever implements auxiliary domains it will use a private IOASID
set, separate from the shared IOASID set managed by iommu-sva. Both could
qualify as "system set".

Thanks,
Jean

> +
>  /*
>   * struct ioasid_allocator_data - Internal data structure to hold information
>   * about an allocator. There are two types of allocators:
> diff --git a/include/linux/ioasid.h b/include/linux/ioasid.h
> index 8c82d2625671..097b1cc043a3 100644
> --- a/include/linux/ioasid.h
> +++ b/include/linux/ioasid.h
> @@ -29,6 +29,9 @@ struct ioasid_allocator_ops {
>  	void *pdata;
>  };
>  
> +/* Shared IOASID set for reserved for host system use */
> +extern int system_ioasid_sid;
> +
>  #define DECLARE_IOASID_SET(name) struct ioasid_set name = { 0 }
>  
>  #if IS_ENABLED(CONFIG_IOASID)
> @@ -41,6 +44,7 @@ int ioasid_register_allocator(struct ioasid_allocator_ops *allocator);
>  void ioasid_unregister_allocator(struct ioasid_allocator_ops *allocator);
>  int ioasid_attach_data(ioasid_t ioasid, void *data);
>  void ioasid_install_capacity(ioasid_t total);
> +int ioasid_alloc_system_set(int quota);
>  int ioasid_alloc_set(struct ioasid_set *token, ioasid_t quota, int *sid);
>  void ioasid_free_set(int sid, bool destroy_set);
>  int ioasid_find_sid(ioasid_t ioasid);
> @@ -88,5 +92,10 @@ static inline void ioasid_install_capacity(ioasid_t total)
>  {
>  }
>  
> +static inline int ioasid_alloc_system_set(int quota)
> +{
> +	return -ENOTSUPP;
> +}
> +
>  #endif /* CONFIG_IOASID */
>  #endif /* __LINUX_IOASID_H */
> -- 
> 2.7.4
>