Skip to contentSkip to conversation

x86/swiotlb/amd: Half the size if allocation failed

PATCH Superseded Alexey Kardashevskiy 4 participants 5 messages Oct 27, 2022 – Oct 30, 2022

Subsystems (2)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86/swiotlb/amd: Half the size if allocation failed

At the moment the AMD encrypted platform reserves 6% of RAM for SWIOTLB or 1GB, whichever is less. However it is possible that there is no block big enough in the low memory which make SWIOTLB allocation fail and the kernel continues without DMA. In such case a VM hangs on DMA.

This divides the size in half and tries again reusing the existing remapping logic.

This updates default_nslabs on successful allocation which looks like an oversight as not doing so should have broken callers of swiotlb_size_or_default().

--

I hit the problem with
QEMU's "-m 16819M" where SWIOTLB was adjusted to 0x7e200 == 1,058,013,184 (slightly less than 1GB) while 0x7e180 still worked.

With guest errors enabled, there are many unassigned accesses from virtio.

---
 kernel/dma/swiotlb.c | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)
kernel/dma/swiotlb.c +13 -7
index 339a990554e7..d28c294320fd 100644
@@ -338,21 +338,27 @@ void __init swiotlb_init_remap(bool addressing_limit, unsigned int flags,
 	else
 		tlb = memblock_alloc_low(bytes, PAGE_SIZE);
 	if (!tlb) {
-		pr_warn("%s: failed to allocate tlb structure\n", __func__);
-		return;
-	}
-
-	if (remap && remap(tlb, nslabs) < 0) {
+		pr_warn("%s: Failed to allocate %zu bytes tlb structure\n",
+			__func__, bytes);
+	} else if (remap && remap(tlb, nslabs) < 0) {
 		memblock_free(tlb, PAGE_ALIGN(bytes));
+		tlb = NULL;
+		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
+	}
 
+	if (!tlb) {
 		nslabs = ALIGN(nslabs >> 1, IO_TLB_SEGSIZE);
 		if (nslabs >= IO_TLB_MIN_SLABS)
 			goto retry;
-
-		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
 		return;
 	}
 
+	if (default_nslabs != nslabs) {
+		pr_info("SWIOTLB bounce buffer size adjusted %lu -> %lu slabs",
+			default_nslabs, nslabs);
+		default_nslabs = nslabs;
+	}
+
 	alloc_size = PAGE_ALIGN(array_size(sizeof(*mem->slots), nslabs));
 	mem->slots = memblock_alloc(alloc_size, PAGE_SIZE);
 	if (!mem->slots) {
Signature · 1 line
2.37.3
#1 lore raw
Quoted 5 lines · Alexey Kardashevskiy

At the moment the AMD encrypted platform reserves 6% of RAM for SWIOTLB or 1GB, whichever is less. However it is possible that there is no block big enough in the low memory which make SWIOTLB allocation fail and the kernel continues without DMA. In such case a VM hangs on DMA.

Quoted 8 lines

This divides the size in half and tries again reusing the existing remapping logic.

This updates default_nslabs on successful allocation which looks like an oversight as not doing so should have broken callers of swiotlb_size_or_default().

Should this have a

Fixes e998879d4fb7 ("x86,swiotlb: Adjust SWIOTLB bounce buffer size for SEV guests")

?

Cascardo.

Quoted 56 lines

--

I hit the problem with
QEMU's "-m 16819M" where SWIOTLB was adjusted to 0x7e200 == 1,058,013,184 (slightly less than 1GB) while 0x7e180 still worked.

With guest errors enabled, there are many unassigned accesses from virtio.

---
 kernel/dma/swiotlb.c | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)
kernel/dma/swiotlb.c +13 -7
index 339a990554e7..d28c294320fd 100644
@@ -338,21 +338,27 @@ void __init swiotlb_init_remap(bool addressing_limit, unsigned int flags,
 	else
 		tlb = memblock_alloc_low(bytes, PAGE_SIZE);
 	if (!tlb) {
-		pr_warn("%s: failed to allocate tlb structure\n", __func__);
-		return;
-	}
-
-	if (remap && remap(tlb, nslabs) < 0) {
+		pr_warn("%s: Failed to allocate %zu bytes tlb structure\n",
+			__func__, bytes);
+	} else if (remap && remap(tlb, nslabs) < 0) {
 		memblock_free(tlb, PAGE_ALIGN(bytes));
+		tlb = NULL;
+		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
+	}
 
+	if (!tlb) {
 		nslabs = ALIGN(nslabs >> 1, IO_TLB_SEGSIZE);
 		if (nslabs >= IO_TLB_MIN_SLABS)
 			goto retry;
-
-		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
 		return;
 	}
 
+	if (default_nslabs != nslabs) {
+		pr_info("SWIOTLB bounce buffer size adjusted %lu -> %lu slabs",
+			default_nslabs, nslabs);
+		default_nslabs = nslabs;
+	}
+
 	alloc_size = PAGE_ALIGN(array_size(sizeof(*mem->slots), nslabs));
 	mem->slots = memblock_alloc(alloc_size, PAGE_SIZE);
 	if (!mem->slots) {
Signature · 1 line
2.37.3
#2 lore raw
Quoted 13 lines · Thadeu Lima de Souza Cascardo
Quoted 12 lines · Alexey Kardashevskiy

At the moment the AMD encrypted platform reserves 6% of RAM for SWIOTLB or 1GB, whichever is less. However it is possible that there is no block big enough in the low memory which make SWIOTLB allocation fail and the kernel continues without DMA. In such case a VM hangs on DMA.

This divides the size in half and tries again reusing the existing remapping logic.

This updates default_nslabs on successful allocation which looks like an oversight as not doing so should have broken callers of swiotlb_size_or_default().

Quoted 5 lines
Quoted 1 line

Should this have a

Fixes e998879d4fb7 ("x86,swiotlb: Adjust SWIOTLB bounce buffer size for SEV guests")

?

Well, the problem was there before that patch, the allocation failure was not handled while remap failure was. e998879d4fb7 just made it easier to see. But still worth mentioning I guess... Thanks,

Quoted 59 lines

Cascardo.

Quoted 56 lines

--

I hit the problem with
QEMU's "-m 16819M" where SWIOTLB was adjusted to 0x7e200 == 1,058,013,184 (slightly less than 1GB) while 0x7e180 still worked.

With guest errors enabled, there are many unassigned accesses from virtio.

---
  kernel/dma/swiotlb.c | 20 +++++++++++++-------
  1 file changed, 13 insertions(+), 7 deletions(-)
kernel/dma/swiotlb.c +13 -7
index 339a990554e7..d28c294320fd 100644
@@ -338,21 +338,27 @@ void __init swiotlb_init_remap(bool addressing_limit, unsigned int flags,
  	else
  		tlb = memblock_alloc_low(bytes, PAGE_SIZE);
  	if (!tlb) {
-		pr_warn("%s: failed to allocate tlb structure\n", __func__);
-		return;
-	}
-
-	if (remap && remap(tlb, nslabs) < 0) {
+		pr_warn("%s: Failed to allocate %zu bytes tlb structure\n",
+			__func__, bytes);
+	} else if (remap && remap(tlb, nslabs) < 0) {
  		memblock_free(tlb, PAGE_ALIGN(bytes));
+		tlb = NULL;
+		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
+	}
 
+	if (!tlb) {
  		nslabs = ALIGN(nslabs >> 1, IO_TLB_SEGSIZE);
  		if (nslabs >= IO_TLB_MIN_SLABS)
  			goto retry;
-
-		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
  		return;
  	}
 
+	if (default_nslabs != nslabs) {
+		pr_info("SWIOTLB bounce buffer size adjusted %lu -> %lu slabs",
+			default_nslabs, nslabs);
+		default_nslabs = nslabs;
+	}
+
  	alloc_size = PAGE_ALIGN(array_size(sizeof(*mem->slots), nslabs));
  	mem->slots = memblock_alloc(alloc_size, PAGE_SIZE);
  	if (!mem->slots) {
Signature · 1 line
2.37.3
Signature · 1 line
Alexey
Reviewed-by on v1 1/1
#3 lore raw
In reply to Alexey Kardashevskiy
Quoted 5 lines · Alexey Kardashevskiy

At the moment the AMD encrypted platform reserves 6% of RAM for SWIOTLB or 1GB, whichever is less. However it is possible that there is no block big enough in the low memory which make SWIOTLB allocation fail and the kernel continues without DMA. In such case a VM hangs on DMA.

Quoted 8 lines

This divides the size in half and tries again reusing the existing remapping logic.

This updates default_nslabs on successful allocation which looks like an oversight as not doing so should have broken callers of swiotlb_size_or_default().

Quoted 53 lines

--

I hit the problem with
QEMU's "-m 16819M" where SWIOTLB was adjusted to 0x7e200 == 1,058,013,184 (slightly less than 1GB) while 0x7e180 still worked.

With guest errors enabled, there are many unassigned accesses from virtio.

---
  kernel/dma/swiotlb.c | 20 +++++++++++++-------
  1 file changed, 13 insertions(+), 7 deletions(-)
kernel/dma/swiotlb.c +13 -7
index 339a990554e7..d28c294320fd 100644
@@ -338,21 +338,27 @@ void __init swiotlb_init_remap(bool addressing_limit, unsigned int flags,
  	else
  		tlb = memblock_alloc_low(bytes, PAGE_SIZE);
  	if (!tlb) {
-		pr_warn("%s: failed to allocate tlb structure\n", __func__);
-		return;
-	}
-
-	if (remap && remap(tlb, nslabs) < 0) {
+		pr_warn("%s: Failed to allocate %zu bytes tlb structure\n",
+			__func__, bytes);
+	} else if (remap && remap(tlb, nslabs) < 0) {
  		memblock_free(tlb, PAGE_ALIGN(bytes));
+		tlb = NULL;
+		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
+	}
 
+	if (!tlb) {
  		nslabs = ALIGN(nslabs >> 1, IO_TLB_SEGSIZE);
  		if (nslabs >= IO_TLB_MIN_SLABS)
  			goto retry;
-
-		pr_warn("%s: Failed to remap %zu bytes\n", __func__, bytes);
  		return;
  	}
 
+	if (default_nslabs != nslabs) {
+		pr_info("SWIOTLB bounce buffer size adjusted %lu -> %lu slabs",
+			default_nslabs, nslabs);
+		default_nslabs = nslabs;
+	}
+
  	alloc_size = PAGE_ALIGN(array_size(sizeof(*mem->slots), nslabs));
  	mem->slots = memblock_alloc(alloc_size, PAGE_SIZE);
  	if (!mem->slots) {
#4 lore raw
In reply to Alexey Kardashevskiy

The subject looks wrong - this just touches the swiotlb code, and nothing x86 or AMD specific.

The code flow now looks a little confusing. Can yu just add a little swiotlb_memblock_alloc helper that does the two variants of memblock allocation and remap, so that we can keep a simply
loop around them in the caller?