Re: [Xen-devel] [PATCH 5/7] public / x86: introduce __HYPERCALL_iommu_op

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 12.02.18 at 11:47, <paul.durrant@xxxxxxxxxx> wrote:
> --- a/xen/arch/x86/Makefile
> +++ b/xen/arch/x86/Makefile
> @@ -33,6 +33,7 @@ obj-$(CONFIG_CRASH_DEBUG) += gdbstub.o
>  obj-y += hypercall.o
>  obj-y += i387.o
>  obj-y += i8259.o
> +obj-y += iommu_op.o

As mentioned in other contexts, I'd prefer if we stopped using
underscores in places where dashes (or other separators not
usable in C identifiers) are fine.

> --- /dev/null
> +++ b/xen/arch/x86/iommu_op.c
> @@ -0,0 +1,169 @@
> +/******************************************************************************
> + * x86/iommu_op.c
> + *
> + * Paravirtualised IOMMU functionality
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (C) 2018 Citrix Systems Inc
> + */
> +
> +#include <xen/event.h>
> +#include <xen/guest_access.h>
> +#include <xen/hypercall.h>
> +
> +static bool can_control_iommu(void)
> +{
> +    struct domain *currd = current->domain;
> +
> +    /*
> +     * IOMMU mappings cannot be manipulated if:
> +     * - the IOMMU is not enabled or,
> +     * - the IOMMU is passed through or,

"is passed through" isn't really a proper description of what
iommu_passthrough means, I'm afraid. The description of the
option says "Control whether to disable DMA remapping for
Dom0." Perhaps "is bypassed"? But then it would be better
to qualify the check with is_hardware_domain(), despite you
restricting things to Dom0 for now anyway.

> +     * - shared EPT configured or,
> +     * - Xen is maintaining an identity map.

Is this meant to describe ...

> +     */
> +    if ( !iommu_enabled || iommu_passthrough ||
> +         iommu_use_hap_pt(currd) || need_iommu(currd) )

... need_iommu() here? How is that implying an identity map?

> +        return false;
> +
> +    return true;

Please make this a singe return statement (with the expression as
operand).

> +long do_iommu_op(XEN_GUEST_HANDLE_PARAM(xen_iommu_op_t) uops,
> +                 unsigned int count)
> +{
> +    unsigned int i;
> +    int rc;
> +
> +    rc = xsm_iommu_op(XSM_PRIV, current->domain);
> +    if ( rc )
> +        return rc;
> +
> +    if ( !can_control_iommu() )
> +        return -EACCES;
> +
> +    for ( i = 0; i < count; i++ )
> +    {
> +        xen_iommu_op_t op;
> +
> +        if ( ((i & 0xff) == 0xff) && hypercall_preempt_check() )
> +        {
> +            rc = i;

For this to be correct for large enough values of "count", rc needs
to have long type.

> +            break;
> +        }
> +
> +        if ( copy_from_guest_offset(&op, uops, i, 1) )
> +        {
> +            rc = -EFAULT;
> +            break;
> +        }
> +
> +        iommu_op(&op);
> +
> +        if ( copy_to_guest_offset(uops, i, &op, 1) )

__copy_to_guest_offset()

Also do you really need to copy back other than the status?

> --- /dev/null
> +++ b/xen/include/public/iommu_op.h
> @@ -0,0 +1,55 @@
> +/*
> + * Permission is hereby granted, free of charge, to any person obtaining a 
> copy
> + * of this software and associated documentation files (the "Software"), to
> + * deal in the Software without restriction, including without limitation the
> + * rights to use, copy, modify, merge, publish, distribute, sublicense, 
> and/or
> + * sell copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 
> THE
> + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
> + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
> + * DEALINGS IN THE SOFTWARE.
> + *
> + * Copyright (C) 2018 Citrix Systems Inc
> + */
> +
> +#ifndef __XEN_PUBLIC_IOMMU_OP_H__
> +#define __XEN_PUBLIC_IOMMU_OP_H__

Please can you avoid introducing further name space violations
into the public headers?

> +#include "xen.h"
> +
> +struct xen_iommu_op {
> +    uint16_t op;
> +    uint16_t flags; /* op specific flags */
> +    int32_t status; /* op completion status: */
> +                    /* 0 for success otherwise, negative errno */
> +};

Peeking at patch 6, you need to add the union and a large enough
placeholder here right away, so that the struct size won't change
with future additions.

Jan

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