Re: [PATCH v6 6/8] dma: swiotlb: Centralize memory-encryption pool sizing

From
Kohei Enju <hidden>
Date
2026-10-30 10:50:35
Also in
linux-arm-kernel, linux-iommu, linux-mips, linux-riscv, linux-s390, lkml, loongarch
Reply to
Aneesh Kumar K.V, 2026-10-08
On 10/08 20:02, Aneesh Kumar K.V wrote:
Catalin Marinas [off-list ref] writes:
quoted
On Thu, Oct 08, 2026 at 11:03:27AM +0530, Aneesh Kumar K.V wrote:
quoted
Aneesh Kumar K.V [off-list ref] writes:
quoted
Will Deacon [off-list ref] writes:
quoted
On Wed, Oct 07, 2026 at 11:04:05AM +0100, Catalin Marinas wrote:
quoted
On Tue, Oct 06, 2026 at 10:49:17PM +0100, Will Deacon wrote:
quoted
On Thu, Sep 24, 2026 at 11:37:54AM +0530, Aneesh Kumar K.V (Arm) wrote:
quoted
@@ -496,7 +516,8 @@ swiotlb_select_pool_policy(unsigned int flags)
 	if (swiotlb_force_disable)
 		return SWIOTLB_POOL_NONE;
 
-	if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT))
+	if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT) &&
+	    !restricted_dma_pool_present)
 		return SWIOTLB_POOL_CC_GUEST;
I think this check on the restricted DMA pool is too general -- the pool
could be tied to a specific DMA-capable peripheral and so treating its
presence as a global property isn't right.
I agree it's a hack but that was the simplest way to avoid the pVMs
getting a bounce buffer after this patch. More than happy to leave it
out and reduce the buffer on cmdline or we come up with some better
heuristics.
Hrm, that does mean that reverting just this part will regress pVMs
because they'll suddenly be allocating a tonne more memory for an
entirely unused swiotlb buffer. So I think I'd prefer to drop the entire
series until this has been worked out properly.
quoted
Another option could be the arch code passing another flag that it
doesn't want an encrypted pool (e.g. when running in a pKVM guest) but I
don't particularly this either. The arch code doesn't know whether
there's an alternative pool.
At that point, the default size may as well be driven by the
drivers/virt/coco driver.
quoted
That said, such heuristics should have been a separate patch to make it
easier to review/drop.
I think the only right way to get a semi-accurate heuristic is to take
into account the set of dma-capable devices that will use the swiotlb
pool, but that's fiddly and should probably be tackled as a separate
series. Maybe a simpler hack in that direction would be to take the
SWIOTLB_POOL_CC_GUEST if _any_ device is going to use swiotlb? You'll
run into the usual problem of not being able to tell if a device is
DMA-capable or not, but you could probably look for a global restricted
DMA pool and, if that doesn't exist, check for per-device restricted pools
on dma-coherent devices (since restricted DMA isn't supported by ACPI) as
a reasonable approximation.
So, something like this?

	if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT) &&
	    swiotlb_cc_guest_needs_default_pool())
		return SWIOTLB_POOL_CC_GUEST;
Detecting a DMA-capable device is not straightforward, and if we get it
wrong, we will enable SWIOTLB_POOL_CC_GUEST unnecessarily. Would the
code below be a reasonable approximation of what you suggested?

Another option would be to make swiotlb_cc_guest_needs_default_pool() a
weak function that architectures can override. arm64 pKVM could then use
a different scheme (for this patch series default to false). Would that
be preferable?
Even the rmem check for each device is still a hack that may bite us in
the future (private devices for example would not need swiotlb). I'm
thinking more and more of leaving the sizing an arch-specific decision,
don't bother generalising it at all.

On pKVM vs CCA guests, there's really nothing specific here to pKVM
guests. The only difference is that confidential guests that so far have
run without a swiotlb buffer will regress if their memory is tight. For
confidential guests without dedicated rmem (either CCA or pKVM), I think
our options are either command line swiotlb sizing or dynamic swiotlb.

Could you respin your series while leaving out the generic sizing? IOW,
no x86 code generalisation. We can discuss the best strategy on sizing
later (I haven't checked how much of this series still makes sense
without the generic sizing).
It would mostly consist of the first three cleanup patches, followed by
three patches that replace the addressing_limit argument with flags.

#define SWIOTLB_VERBOSE	(1 << 0) /* verbose initialization */
/* Initialize a pool for devices with limited DMA addressing. */
#define SWIOTLB_INIT_ADDRESSING_LIMIT	(1 << 1)
/* Initialize a pool that requires architecture remapping. */
#define SWIOTLB_INIT_REMAP		(1 << 2)
/* Initialize a pool for DMA to memory-encrypted host or guest memory. */
#define SWIOTLB_INIT_MEM_ENCRYPT		(1 << 3)
/* Initialize a pool for unaligned kmalloc bouncing. */
#define SWIOTLB_INIT_KMALLOC		(1 << 4)
/* Do not initialize a pool unless SWIOTLB is explicitly required. */
#define SWIOTLB_INIT_DEFAULT_OFF		(1 << 5)

This results in the large change below. I'm not sure we want to do this
for no real benefit other than making swiotlb_should_init() slightly
easier to follow.

static bool __init swiotlb_should_init(unsigned int flags)
{
	if (swiotlb_force_disable)
		return false;

	if (swiotlb_force_bounce)
		return true;

	if (flags & (SWIOTLB_INIT_REMAP | SWIOTLB_INIT_MEM_ENCRYPT)))
		return true;

	/* Explicit requirements override an architecture's default opt-out. */
	if (flags & SWIOTLB_INIT_DEFAULT_OFF)
		return false;

	return flags & (SWIOTLB_INIT_ADDRESSING_LIMIT | SWIOTLB_INIT_KMALLOC);
}
 [...]
Hi Aneesh,

I can reproduce the CCA regression discussed in this thread:
  https://lore.kernel.org/all/yq5azeyxyfol.fsf@kernel.org/#t (local)

The guest runs out of SWIOTLB slots, causing disk I/O to fail and
preventing the guest from booting.

I would like to test the fix on real CCA hardware. Could you please Cc
me when you respin the series? I follow LKML, but I may still miss the
posting by chance.

Thanks,
Kohei
Thread (29 messages) 29 messages, 5 authors, last reply 5d ago
Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help