[PATCH v5 5/6] dma: swiotlb: Add an overridable architecture pool opt-out
Robin Murphy
robin.murphy at arm.com
Wed Sep 23 06:04:00 PDT 2026
On 21/09/2026 7:36 am, Aneesh Kumar K.V (Arm) wrote:
> On x86, iommu=off historically prevented SWIOTLB initialization by not
> setting SWIOTLB_INIT_ADDRESSING_LIMIT. This overloads that flag: it
> controls both whether a pool is selected and where the pool is
> allocated. If another requirement, swiotlb=force selects a pool
> independently, omitting SWIOTLB_INIT_ADDRESSING_LIMIT may place it above
> the addressable range of devices which need bounce buffering.
Well then surely something is wrong with the design? If
SWIOTLB_INIT_ADDRESSING_LIMIT is set then SWIOTLB still needs to respect
it irrespective of any *additional* reasons for activating (either from
arch code or internally). And conversely, if devices have addressing
limits but SWIOTLB_INIT_ADDRESSING_LIMIT is *not* set, then the arch
code has clearly gone wrong or made a bad assumption, but that doesn't
seem like SWIOTLB's problem to worry about.
> Add SWIOTLB_INIT_FORCE_DISABLE to let an architecture opt out of normal
> pool initialization without discarding the pool placement constraints.
>
> Unlike swiotlb_force_disable, which is set by swiotlb=noforce and
> unconditionally prevents pool initialization, the new flag is an
> architecture default. Explicit requirements such as memory encryption,
> architecture remapping, or swiotlb=force take precedence over it.
Logically, that should already be implied by the architecture passing no
SWIOTLB_INIT flags.
Thanks,
Robin.
> iommu=soft retains SWIOTLB_INIT_ADDRESSING_LIMIT because it requests
> software bounce buffering for devices whose DMA masks cannot address all
> system memory. Such devices must be able to reach the bounce pool.
>
> Clear both flags for Xen because Xen explicitly requires a remapped
> SWIOTLB pool and may allocate it anywhere in directly mapped memory.
>
> With no override, iommu=off disables both hardware IOMMU translation and
> SWIOTLB bouncing. Direct DMA remains available; mappings outside a
> device's addressable range fail instead of being bounced.
>
> Reviewed-by: Catalin Marinas <catalin.marinas at arm.com>
> Signed-off-by: Aneesh Kumar K.V (Arm) <aneesh.kumar at kernel.org>
> ---
> arch/x86/kernel/pci-dma.c | 11 ++++++++---
> include/linux/swiotlb.h | 2 ++
> kernel/dma/swiotlb.c | 10 ++++++++--
> 3 files changed, 18 insertions(+), 5 deletions(-)
>
> diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
> index 0cff255827ba..ce257b20ea9c 100644
> --- a/arch/x86/kernel/pci-dma.c
> +++ b/arch/x86/kernel/pci-dma.c
> @@ -44,9 +44,13 @@ static unsigned int x86_swiotlb_flags;
> static void __init pci_swiotlb_detect(void)
> {
> /* don't initialize swiotlb if iommu=off (no_iommu=1) */
> - if (!no_iommu && max_possible_pfn > MAX_DMA32_PFN) {
> - x86_swiotlb_enable = true;
> + if (no_iommu)
> + x86_swiotlb_flags |= SWIOTLB_INIT_DEFAULT_OFF;
> +
> + if (max_possible_pfn > MAX_DMA32_PFN) {
> x86_swiotlb_flags |= SWIOTLB_INIT_ADDRESSING_LIMIT;
> + if (!no_iommu)
> + x86_swiotlb_enable = true;
> }
>
> /*
> @@ -84,7 +88,8 @@ static void __init pci_xen_swiotlb_init(void)
> return;
> x86_swiotlb_enable = true;
> /* Xen can use a SWIOTLB pool anywhere in directly mapped memory. */
> - x86_swiotlb_flags &= ~SWIOTLB_INIT_ADDRESSING_LIMIT;
> + x86_swiotlb_flags &= ~(SWIOTLB_INIT_ADDRESSING_LIMIT |
> + SWIOTLB_INIT_DEFAULT_OFF);
> x86_swiotlb_flags |= SWIOTLB_INIT_REMAP | SWIOTLB_ANY;
> swiotlb_init_remap(x86_swiotlb_flags, xen_swiotlb_fixup);
> dma_ops = &xen_swiotlb_dma_ops;
> diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
> index be3962a33fc6..466c33795db0 100644
> --- a/include/linux/swiotlb.h
> +++ b/include/linux/swiotlb.h
> @@ -20,6 +20,8 @@ struct scatterlist;
> #define SWIOTLB_INIT_ADDRESSING_LIMIT (1 << 2)
> /* Initialize a default-sized pool that requires architecture remapping. */
> #define SWIOTLB_INIT_REMAP (1 << 3)
> +/* Do not initialize a pool unless SWIOTLB is explicitly required. */
> +#define SWIOTLB_INIT_DEFAULT_OFF (1 << 4)
>
> /*
> * Maximum allowable number of contiguous slabs to map,
> diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
> index c19675846fc1..46b789eb0806 100644
> --- a/kernel/dma/swiotlb.c
> +++ b/kernel/dma/swiotlb.c
> @@ -563,10 +563,16 @@ swiotlb_select_pool_policy(unsigned int flags)
> if (flags & SWIOTLB_INIT_REMAP)
> return SWIOTLB_POOL_DEFAULT;
>
> - if (flags & SWIOTLB_INIT_ADDRESSING_LIMIT)
> + if (swiotlb_force_bounce)
> return SWIOTLB_POOL_DEFAULT;
>
> - if (swiotlb_force_bounce)
> + /*
> + * Explicit requirements above override an architecture's default opt-out.
> + */
> + if (flags & SWIOTLB_INIT_DEFAULT_OFF)
> + return SWIOTLB_POOL_NONE;
> +
> + if (flags & SWIOTLB_INIT_ADDRESSING_LIMIT)
> return SWIOTLB_POOL_DEFAULT;
>
> if (swiotlb_kmalloc_needs_bounce())
More information about the linux-arm-kernel
mailing list