diff mbox series

[RFC,v1,2/6] swiotlb: convert variables to arrays

Message ID 20210203233709.19819-3-dongli.zhang@oracle.com (mailing list archive)
State Not Applicable
Headers show
Series swiotlb: 64-bit DMA buffer | expand

Commit Message

Dongli Zhang Feb. 3, 2021, 11:37 p.m. UTC 
This patch converts several swiotlb related variables to arrays, in
order to maintain stat/status for different swiotlb buffers. Here are
variables involved:

- io_tlb_start and io_tlb_end
- io_tlb_nslabs and io_tlb_used
- io_tlb_list
- io_tlb_index
- max_segment
- io_tlb_orig_addr
- no_iotlb_memory

There is no functional change and this is to prepare to enable 64-bit
swiotlb.

Cc: Joe Jin <joe.jin@oracle.com>
Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
---
 arch/powerpc/platforms/pseries/svm.c |   6 +-
 drivers/xen/swiotlb-xen.c            |   4 +-
 include/linux/swiotlb.h              |   5 +-
 kernel/dma/swiotlb.c                 | 257 ++++++++++++++-------------
 4 files changed, 140 insertions(+), 132 deletions(-)

Comments

Christoph Hellwig Feb. 4, 2021, 7:29 a.m. UTC | #1 
On Wed, Feb 03, 2021 at 03:37:05PM -0800, Dongli Zhang wrote:
> This patch converts several swiotlb related variables to arrays, in
> order to maintain stat/status for different swiotlb buffers. Here are
> variables involved:
> 
> - io_tlb_start and io_tlb_end
> - io_tlb_nslabs and io_tlb_used
> - io_tlb_list
> - io_tlb_index
> - max_segment
> - io_tlb_orig_addr
> - no_iotlb_memory
> 
> There is no functional change and this is to prepare to enable 64-bit
> swiotlb.

Claire Chang (on Cc) already posted a patch like this a month ago,
which looks much better because it actually uses a struct instead
of all the random variables.
Robin Murphy Feb. 4, 2021, 11:49 a.m. UTC | #2 
On 2021-02-04 07:29, Christoph Hellwig wrote:
> On Wed, Feb 03, 2021 at 03:37:05PM -0800, Dongli Zhang wrote:
>> This patch converts several swiotlb related variables to arrays, in
>> order to maintain stat/status for different swiotlb buffers. Here are
>> variables involved:
>>
>> - io_tlb_start and io_tlb_end
>> - io_tlb_nslabs and io_tlb_used
>> - io_tlb_list
>> - io_tlb_index
>> - max_segment
>> - io_tlb_orig_addr
>> - no_iotlb_memory
>>
>> There is no functional change and this is to prepare to enable 64-bit
>> swiotlb.
> 
> Claire Chang (on Cc) already posted a patch like this a month ago,
> which looks much better because it actually uses a struct instead
> of all the random variables.

Indeed, I skimmed the cover letter and immediately thought that this 
whole thing is just the restricted DMA pool concept[1] again, only from 
a slightly different angle.

Robin.

[1] 
https://lore.kernel.org/linux-iommu/20210106034124.30560-1-tientzu@chromium.org/
Konrad Rzeszutek Wilk Feb. 4, 2021, 7:31 p.m. UTC | #3 
On Thu, Feb 04, 2021 at 11:49:23AM +0000, Robin Murphy wrote:
> On 2021-02-04 07:29, Christoph Hellwig wrote:
> > On Wed, Feb 03, 2021 at 03:37:05PM -0800, Dongli Zhang wrote:
> > > This patch converts several swiotlb related variables to arrays, in
> > > order to maintain stat/status for different swiotlb buffers. Here are
> > > variables involved:
> > > 
> > > - io_tlb_start and io_tlb_end
> > > - io_tlb_nslabs and io_tlb_used
> > > - io_tlb_list
> > > - io_tlb_index
> > > - max_segment
> > > - io_tlb_orig_addr
> > > - no_iotlb_memory
> > > 
> > > There is no functional change and this is to prepare to enable 64-bit
> > > swiotlb.
> > 
> > Claire Chang (on Cc) already posted a patch like this a month ago,
> > which looks much better because it actually uses a struct instead
> > of all the random variables.
> 
> Indeed, I skimmed the cover letter and immediately thought that this whole
> thing is just the restricted DMA pool concept[1] again, only from a slightly
> different angle.


Kind of. Let me lay out how some of these pieces are right now:

+-----------------------+      +----------------------+
|                       |      |                      |
|                       |      |                      |
|   a)Xen-SWIOTLB       |      | b)SWIOTLB (for !Xen) |
|                       |      |                      |
+-----------XX----------+      +-------X--------------+
              XXXX             XXXXXXXXX
                 XXXX     XX XXX
                    X   XX
                    XXXX
         +----------XX-----------+
         |                       |
         |                       |
         |   c) SWIOTLB generic  |
         |                       |
         +-----------------------+

Dongli's patches modify the SWIOTLB generic c), and Xen-SWIOTLB a)
parts.

Also see the IOMMU_INIT logic which lays this a bit more deepth
(for example how to enable SWIOTLB on AMD boxes, or IBM with Calgary
IOMMU, etc - see iommu_table.h).

Furtheremore it lays the groundwork to allocate AMD SEV SWIOTLB buffers
later after boot (so that you can stich different pools together).
All the bits are kind of inside of the SWIOTLB code. And also it changes
the Xen-SWIOTLB to do something similar.

The mempool did it similarly by taking the internal parts (aka the
various io_tlb) of SWIOTLB and exposing them out and having
other code:

+-----------------------+      +----------------------+
|                       |      |                      |
|                       |      |                      |
| a)Xen-SWIOTLB         |      | b)SWIOTLB (for !Xen) |
|                       |      |                      |
+-----------XX----------+      +-------X--------------+
              XXXX             XXXXXXXXX
                 XXXX     XX XXX
                    X   XX
                    XXXX
         +----------XX-----------+         +------------------+
         |                       |         | Device tree      |
         |                       +<--------+ enabling SWIOTLB |
         |c) SWIOTLB generic     |         |                  |
         |                       |         | mempool          |
         +-----------------------+         +------------------+

What I was suggesting to Clarie to follow Xen model, that is
do something like this:

+-----------------------+      +----------------------+   +--------------------+
|                       |      |                      |   |                    |
|                       |      |                      |   |                    |
| a)Xen-SWIOTLB         |      | b)SWIOTLB (for !Xen) |   | e) DT-SWIOTLB      |
|                       |      |                      |   |                    |
+-----------XX----------+      +-------X--------------+   +----XX-X------------+
              XXXX             XXXXXXXXX        XXX X X XX X XX
                 XXXX     XX XXX        XXXXXXXX
                    X   XX XXXXXXXXXXXXX
                    XXXXXXXX
         +----------XXX----------+
         |                       |
         |                       |
         |c) SWIOTLB generic     |
         |                       |
         +-----------------------+


so using the SWIOTLB generic parts, and then bolt on top
of the device-tree logic, along with the mempool logic.



But Christopher has an interesting suggestion which is
to squash the all the existing code (a, b, c) all together
and pepper it with various jump-tables.


So:


-----------------------------+
| SWIOTLB:                   |
|                            |
|  a) SWIOTLB (for non-Xen)  |
|  b) Xen-SWIOTLB            |
|  c) DT-SWIOTLB             |
|                            |
|                            |
-----------------------------+


with all the various bits (M2P/P2M for Xen, mempool for ARM,
and normal allocation for BM) in one big file.
diff mbox series

Patch

diff --git a/arch/powerpc/platforms/pseries/svm.c b/arch/powerpc/platforms/pseries/svm.c
index 7b739cc7a8a9..9f8842d0da1f 100644
--- a/arch/powerpc/platforms/pseries/svm.c
+++ b/arch/powerpc/platforms/pseries/svm.c
@@ -55,9 +55,9 @@  void __init svm_swiotlb_init(void)
 	if (vstart && !swiotlb_init_with_tbl(vstart, io_tlb_nslabs, false))
 		return;
 
-	if (io_tlb_start)
-		memblock_free_early(io_tlb_start,
-				    PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
+	if (io_tlb_start[SWIOTLB_LO])
+		memblock_free_early(io_tlb_start[SWIOTLB_LO],
+				    PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT));
 	panic("SVM: Cannot allocate SWIOTLB buffer");
 }
 
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 2b385c1b4a99..3261880ad859 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -192,8 +192,8 @@  int __ref xen_swiotlb_init(int verbose, bool early)
 	/*
 	 * IO TLB memory already allocated. Just use it.
 	 */
-	if (io_tlb_start != 0) {
-		xen_io_tlb_start = phys_to_virt(io_tlb_start);
+	if (io_tlb_start[SWIOTLB_LO] != 0) {
+		xen_io_tlb_start = phys_to_virt(io_tlb_start[SWIOTLB_LO]);
 		goto end;
 	}
 
diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index ca125c1b1281..777046cd4d1b 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -76,11 +76,12 @@  dma_addr_t swiotlb_map(struct device *dev, phys_addr_t phys,
 
 #ifdef CONFIG_SWIOTLB
 extern enum swiotlb_force swiotlb_force;
-extern phys_addr_t io_tlb_start, io_tlb_end;
+extern phys_addr_t io_tlb_start[], io_tlb_end[];
 
 static inline bool is_swiotlb_buffer(phys_addr_t paddr)
 {
-	return paddr >= io_tlb_start && paddr < io_tlb_end;
+	return paddr >= io_tlb_start[SWIOTLB_LO] &&
+	       paddr < io_tlb_end[SWIOTLB_LO];
 }
 
 void __init swiotlb_exit(void);
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index 7c42df6e6100..1fbb65daa2dd 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -69,38 +69,38 @@  enum swiotlb_force swiotlb_force;
  * swiotlb_tbl_sync_single_*, to see if the memory was in fact allocated by this
  * API.
  */
-phys_addr_t io_tlb_start, io_tlb_end;
+phys_addr_t io_tlb_start[SWIOTLB_MAX], io_tlb_end[SWIOTLB_MAX];
 
 /*
  * The number of IO TLB blocks (in groups of 64) between io_tlb_start and
  * io_tlb_end.  This is command line adjustable via setup_io_tlb_npages.
  */
-static unsigned long io_tlb_nslabs;
+static unsigned long io_tlb_nslabs[SWIOTLB_MAX];
 
 /*
  * The number of used IO TLB block
  */
-static unsigned long io_tlb_used;
+static unsigned long io_tlb_used[SWIOTLB_MAX];
 
 /*
  * This is a free list describing the number of free entries available from
  * each index
  */
-static unsigned int *io_tlb_list;
-static unsigned int io_tlb_index;
+static unsigned int *io_tlb_list[SWIOTLB_MAX];
+static unsigned int io_tlb_index[SWIOTLB_MAX];
 
 /*
  * Max segment that we can provide which (if pages are contingous) will
  * not be bounced (unless SWIOTLB_FORCE is set).
  */
-static unsigned int max_segment;
+static unsigned int max_segment[SWIOTLB_MAX];
 
 /*
  * We need to save away the original address corresponding to a mapped entry
  * for the sync operations.
  */
 #define INVALID_PHYS_ADDR (~(phys_addr_t)0)
-static phys_addr_t *io_tlb_orig_addr;
+static phys_addr_t *io_tlb_orig_addr[SWIOTLB_MAX];
 
 /*
  * Protect the above data structures in the map and unmap calls
@@ -113,9 +113,9 @@  static int __init
 setup_io_tlb_npages(char *str)
 {
 	if (isdigit(*str)) {
-		io_tlb_nslabs = simple_strtoul(str, &str, 0);
+		io_tlb_nslabs[SWIOTLB_LO] = simple_strtoul(str, &str, 0);
 		/* avoid tail segment of size < IO_TLB_SEGSIZE */
-		io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+		io_tlb_nslabs[SWIOTLB_LO] = ALIGN(io_tlb_nslabs[SWIOTLB_LO], IO_TLB_SEGSIZE);
 	}
 	if (*str == ',')
 		++str;
@@ -123,40 +123,40 @@  setup_io_tlb_npages(char *str)
 		swiotlb_force = SWIOTLB_FORCE;
 	} else if (!strcmp(str, "noforce")) {
 		swiotlb_force = SWIOTLB_NO_FORCE;
-		io_tlb_nslabs = 1;
+		io_tlb_nslabs[SWIOTLB_LO] = 1;
 	}
 
 	return 0;
 }
 early_param("swiotlb", setup_io_tlb_npages);
 
-static bool no_iotlb_memory;
+static bool no_iotlb_memory[SWIOTLB_MAX];
 
 unsigned long swiotlb_nr_tbl(void)
 {
-	return unlikely(no_iotlb_memory) ? 0 : io_tlb_nslabs;
+	return unlikely(no_iotlb_memory[SWIOTLB_LO]) ? 0 : io_tlb_nslabs[SWIOTLB_LO];
 }
 EXPORT_SYMBOL_GPL(swiotlb_nr_tbl);
 
 unsigned int swiotlb_max_segment(void)
 {
-	return unlikely(no_iotlb_memory) ? 0 : max_segment;
+	return unlikely(no_iotlb_memory[SWIOTLB_LO]) ? 0 : max_segment[SWIOTLB_LO];
 }
 EXPORT_SYMBOL_GPL(swiotlb_max_segment);
 
 void swiotlb_set_max_segment(unsigned int val)
 {
 	if (swiotlb_force == SWIOTLB_FORCE)
-		max_segment = 1;
+		max_segment[SWIOTLB_LO] = 1;
 	else
-		max_segment = rounddown(val, PAGE_SIZE);
+		max_segment[SWIOTLB_LO] = rounddown(val, PAGE_SIZE);
 }
 
 unsigned long swiotlb_size_or_default(void)
 {
 	unsigned long size;
 
-	size = io_tlb_nslabs << IO_TLB_SHIFT;
+	size = io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT;
 
 	return size ? size : (IO_TLB_DEFAULT_SIZE);
 }
@@ -170,10 +170,10 @@  void __init swiotlb_adjust_size(unsigned long new_size)
 	 * architectures such as those supporting memory encryption to
 	 * adjust/expand SWIOTLB size for their use.
 	 */
-	if (!io_tlb_nslabs) {
+	if (!io_tlb_nslabs[SWIOTLB_LO]) {
 		size = ALIGN(new_size, 1 << IO_TLB_SHIFT);
-		io_tlb_nslabs = size >> IO_TLB_SHIFT;
-		io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+		io_tlb_nslabs[SWIOTLB_LO] = size >> IO_TLB_SHIFT;
+		io_tlb_nslabs[SWIOTLB_LO] = ALIGN(io_tlb_nslabs[SWIOTLB_LO], IO_TLB_SEGSIZE);
 
 		pr_info("SWIOTLB bounce buffer size adjusted to %luMB", size >> 20);
 	}
@@ -181,15 +181,16 @@  void __init swiotlb_adjust_size(unsigned long new_size)
 
 void swiotlb_print_info(void)
 {
-	unsigned long bytes = io_tlb_nslabs << IO_TLB_SHIFT;
+	unsigned long bytes = io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT;
 
-	if (no_iotlb_memory) {
+	if (no_iotlb_memory[SWIOTLB_LO]) {
 		pr_warn("No low mem\n");
 		return;
 	}
 
-	pr_info("mapped [mem %pa-%pa] (%luMB)\n", &io_tlb_start, &io_tlb_end,
-	       bytes >> 20);
+	pr_info("mapped [mem %pa-%pa] (%luMB)\n",
+		&io_tlb_start[SWIOTLB_LO], &io_tlb_end[SWIOTLB_LO],
+		bytes >> 20);
 }
 
 /*
@@ -203,11 +204,11 @@  void __init swiotlb_update_mem_attributes(void)
 	void *vaddr;
 	unsigned long bytes;
 
-	if (no_iotlb_memory || late_alloc)
+	if (no_iotlb_memory[SWIOTLB_LO] || late_alloc)
 		return;
 
-	vaddr = phys_to_virt(io_tlb_start);
-	bytes = PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT);
+	vaddr = phys_to_virt(io_tlb_start[SWIOTLB_LO]);
+	bytes = PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT);
 	set_memory_decrypted((unsigned long)vaddr, bytes >> PAGE_SHIFT);
 	memset(vaddr, 0, bytes);
 }
@@ -219,38 +220,38 @@  int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
 
 	bytes = nslabs << IO_TLB_SHIFT;
 
-	io_tlb_nslabs = nslabs;
-	io_tlb_start = __pa(tlb);
-	io_tlb_end = io_tlb_start + bytes;
+	io_tlb_nslabs[SWIOTLB_LO] = nslabs;
+	io_tlb_start[SWIOTLB_LO] = __pa(tlb);
+	io_tlb_end[SWIOTLB_LO] = io_tlb_start[SWIOTLB_LO] + bytes;
 
 	/*
 	 * Allocate and initialize the free list array.  This array is used
 	 * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
 	 * between io_tlb_start and io_tlb_end.
 	 */
-	alloc_size = PAGE_ALIGN(io_tlb_nslabs * sizeof(int));
-	io_tlb_list = memblock_alloc(alloc_size, PAGE_SIZE);
-	if (!io_tlb_list)
+	alloc_size = PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] * sizeof(int));
+	io_tlb_list[SWIOTLB_LO] = memblock_alloc(alloc_size, PAGE_SIZE);
+	if (!io_tlb_list[SWIOTLB_LO])
 		panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
 		      __func__, alloc_size, PAGE_SIZE);
 
-	alloc_size = PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t));
-	io_tlb_orig_addr = memblock_alloc(alloc_size, PAGE_SIZE);
-	if (!io_tlb_orig_addr)
+	alloc_size = PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] * sizeof(phys_addr_t));
+	io_tlb_orig_addr[SWIOTLB_LO] = memblock_alloc(alloc_size, PAGE_SIZE);
+	if (!io_tlb_orig_addr[SWIOTLB_LO])
 		panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
 		      __func__, alloc_size, PAGE_SIZE);
 
-	for (i = 0; i < io_tlb_nslabs; i++) {
-		io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
-		io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
+	for (i = 0; i < io_tlb_nslabs[SWIOTLB_LO]; i++) {
+		io_tlb_list[SWIOTLB_LO][i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
+		io_tlb_orig_addr[SWIOTLB_LO][i] = INVALID_PHYS_ADDR;
 	}
-	io_tlb_index = 0;
-	no_iotlb_memory = false;
+	io_tlb_index[SWIOTLB_LO] = 0;
+	no_iotlb_memory[SWIOTLB_LO] = false;
 
 	if (verbose)
 		swiotlb_print_info();
 
-	swiotlb_set_max_segment(io_tlb_nslabs << IO_TLB_SHIFT);
+	swiotlb_set_max_segment(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT);
 	return 0;
 }
 
@@ -265,25 +266,25 @@  swiotlb_init(int verbose)
 	unsigned char *vstart;
 	unsigned long bytes;
 
-	if (!io_tlb_nslabs) {
-		io_tlb_nslabs = (default_size >> IO_TLB_SHIFT);
-		io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+	if (!io_tlb_nslabs[SWIOTLB_LO]) {
+		io_tlb_nslabs[SWIOTLB_LO] = (default_size >> IO_TLB_SHIFT);
+		io_tlb_nslabs[SWIOTLB_LO] = ALIGN(io_tlb_nslabs[SWIOTLB_LO], IO_TLB_SEGSIZE);
 	}
 
-	bytes = io_tlb_nslabs << IO_TLB_SHIFT;
+	bytes = io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT;
 
 	/* Get IO TLB memory from the low pages */
 	vstart = memblock_alloc_low(PAGE_ALIGN(bytes), PAGE_SIZE);
-	if (vstart && !swiotlb_init_with_tbl(vstart, io_tlb_nslabs, verbose))
+	if (vstart && !swiotlb_init_with_tbl(vstart, io_tlb_nslabs[SWIOTLB_LO], verbose))
 		return;
 
-	if (io_tlb_start) {
-		memblock_free_early(io_tlb_start,
-				    PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
-		io_tlb_start = 0;
+	if (io_tlb_start[SWIOTLB_LO]) {
+		memblock_free_early(io_tlb_start[SWIOTLB_LO],
+				    PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT));
+		io_tlb_start[SWIOTLB_LO] = 0;
 	}
 	pr_warn("Cannot allocate buffer");
-	no_iotlb_memory = true;
+	no_iotlb_memory[SWIOTLB_LO] = true;
 }
 
 /*
@@ -294,22 +295,22 @@  swiotlb_init(int verbose)
 int
 swiotlb_late_init_with_default_size(size_t default_size)
 {
-	unsigned long bytes, req_nslabs = io_tlb_nslabs;
+	unsigned long bytes, req_nslabs = io_tlb_nslabs[SWIOTLB_LO];
 	unsigned char *vstart = NULL;
 	unsigned int order;
 	int rc = 0;
 
-	if (!io_tlb_nslabs) {
-		io_tlb_nslabs = (default_size >> IO_TLB_SHIFT);
-		io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+	if (!io_tlb_nslabs[SWIOTLB_LO]) {
+		io_tlb_nslabs[SWIOTLB_LO] = (default_size >> IO_TLB_SHIFT);
+		io_tlb_nslabs[SWIOTLB_LO] = ALIGN(io_tlb_nslabs[SWIOTLB_LO], IO_TLB_SEGSIZE);
 	}
 
 	/*
 	 * Get IO TLB memory from the low pages
 	 */
-	order = get_order(io_tlb_nslabs << IO_TLB_SHIFT);
-	io_tlb_nslabs = SLABS_PER_PAGE << order;
-	bytes = io_tlb_nslabs << IO_TLB_SHIFT;
+	order = get_order(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT);
+	io_tlb_nslabs[SWIOTLB_LO] = SLABS_PER_PAGE << order;
+	bytes = io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT;
 
 	while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
 		vstart = (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN,
@@ -320,15 +321,15 @@  swiotlb_late_init_with_default_size(size_t default_size)
 	}
 
 	if (!vstart) {
-		io_tlb_nslabs = req_nslabs;
+		io_tlb_nslabs[SWIOTLB_LO] = req_nslabs;
 		return -ENOMEM;
 	}
 	if (order != get_order(bytes)) {
 		pr_warn("only able to allocate %ld MB\n",
 			(PAGE_SIZE << order) >> 20);
-		io_tlb_nslabs = SLABS_PER_PAGE << order;
+		io_tlb_nslabs[SWIOTLB_LO] = SLABS_PER_PAGE << order;
 	}
-	rc = swiotlb_late_init_with_tbl(vstart, io_tlb_nslabs);
+	rc = swiotlb_late_init_with_tbl(vstart, io_tlb_nslabs[SWIOTLB_LO]);
 	if (rc)
 		free_pages((unsigned long)vstart, order);
 
@@ -337,10 +338,10 @@  swiotlb_late_init_with_default_size(size_t default_size)
 
 static void swiotlb_cleanup(void)
 {
-	io_tlb_end = 0;
-	io_tlb_start = 0;
-	io_tlb_nslabs = 0;
-	max_segment = 0;
+	io_tlb_end[SWIOTLB_LO] = 0;
+	io_tlb_start[SWIOTLB_LO] = 0;
+	io_tlb_nslabs[SWIOTLB_LO] = 0;
+	max_segment[SWIOTLB_LO] = 0;
 }
 
 int
@@ -350,9 +351,9 @@  swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
 
 	bytes = nslabs << IO_TLB_SHIFT;
 
-	io_tlb_nslabs = nslabs;
-	io_tlb_start = virt_to_phys(tlb);
-	io_tlb_end = io_tlb_start + bytes;
+	io_tlb_nslabs[SWIOTLB_LO] = nslabs;
+	io_tlb_start[SWIOTLB_LO] = virt_to_phys(tlb);
+	io_tlb_end[SWIOTLB_LO] = io_tlb_start[SWIOTLB_LO] + bytes;
 
 	set_memory_decrypted((unsigned long)tlb, bytes >> PAGE_SHIFT);
 	memset(tlb, 0, bytes);
@@ -362,37 +363,37 @@  swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
 	 * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
 	 * between io_tlb_start and io_tlb_end.
 	 */
-	io_tlb_list = (unsigned int *)__get_free_pages(GFP_KERNEL,
-	                              get_order(io_tlb_nslabs * sizeof(int)));
-	if (!io_tlb_list)
+	io_tlb_list[SWIOTLB_LO] = (unsigned int *)__get_free_pages(GFP_KERNEL,
+			get_order(io_tlb_nslabs[SWIOTLB_LO] * sizeof(int)));
+	if (!io_tlb_list[SWIOTLB_LO])
 		goto cleanup3;
 
-	io_tlb_orig_addr = (phys_addr_t *)
+	io_tlb_orig_addr[SWIOTLB_LO] = (phys_addr_t *)
 		__get_free_pages(GFP_KERNEL,
-				 get_order(io_tlb_nslabs *
+				 get_order(io_tlb_nslabs[SWIOTLB_LO] *
 					   sizeof(phys_addr_t)));
-	if (!io_tlb_orig_addr)
+	if (!io_tlb_orig_addr[SWIOTLB_LO])
 		goto cleanup4;
 
-	for (i = 0; i < io_tlb_nslabs; i++) {
-		io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
-		io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
+	for (i = 0; i < io_tlb_nslabs[SWIOTLB_LO]; i++) {
+		io_tlb_list[SWIOTLB_LO][i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
+		io_tlb_orig_addr[SWIOTLB_LO][i] = INVALID_PHYS_ADDR;
 	}
-	io_tlb_index = 0;
-	no_iotlb_memory = false;
+	io_tlb_index[SWIOTLB_LO] = 0;
+	no_iotlb_memory[SWIOTLB_LO] = false;
 
 	swiotlb_print_info();
 
 	late_alloc = 1;
 
-	swiotlb_set_max_segment(io_tlb_nslabs << IO_TLB_SHIFT);
+	swiotlb_set_max_segment(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT);
 
 	return 0;
 
 cleanup4:
-	free_pages((unsigned long)io_tlb_list, get_order(io_tlb_nslabs *
+	free_pages((unsigned long)io_tlb_list[SWIOTLB_LO], get_order(io_tlb_nslabs[SWIOTLB_LO] *
 	                                                 sizeof(int)));
-	io_tlb_list = NULL;
+	io_tlb_list[SWIOTLB_LO] = NULL;
 cleanup3:
 	swiotlb_cleanup();
 	return -ENOMEM;
@@ -400,23 +401,23 @@  swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
 
 void __init swiotlb_exit(void)
 {
-	if (!io_tlb_orig_addr)
+	if (!io_tlb_orig_addr[SWIOTLB_LO])
 		return;
 
 	if (late_alloc) {
-		free_pages((unsigned long)io_tlb_orig_addr,
-			   get_order(io_tlb_nslabs * sizeof(phys_addr_t)));
-		free_pages((unsigned long)io_tlb_list, get_order(io_tlb_nslabs *
-								 sizeof(int)));
-		free_pages((unsigned long)phys_to_virt(io_tlb_start),
-			   get_order(io_tlb_nslabs << IO_TLB_SHIFT));
+		free_pages((unsigned long)io_tlb_orig_addr[SWIOTLB_LO],
+			   get_order(io_tlb_nslabs[SWIOTLB_LO] * sizeof(phys_addr_t)));
+		free_pages((unsigned long)io_tlb_list[SWIOTLB_LO],
+			   get_order(io_tlb_nslabs[SWIOTLB_LO] * sizeof(int)));
+		free_pages((unsigned long)phys_to_virt(io_tlb_start[SWIOTLB_LO]),
+			   get_order(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT));
 	} else {
-		memblock_free_late(__pa(io_tlb_orig_addr),
-				   PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t)));
-		memblock_free_late(__pa(io_tlb_list),
-				   PAGE_ALIGN(io_tlb_nslabs * sizeof(int)));
-		memblock_free_late(io_tlb_start,
-				   PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
+		memblock_free_late(__pa(io_tlb_orig_addr[SWIOTLB_LO]),
+				   PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] * sizeof(phys_addr_t)));
+		memblock_free_late(__pa(io_tlb_list[SWIOTLB_LO]),
+				   PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] * sizeof(int)));
+		memblock_free_late(io_tlb_start[SWIOTLB_LO],
+				   PAGE_ALIGN(io_tlb_nslabs[SWIOTLB_LO] << IO_TLB_SHIFT));
 	}
 	swiotlb_cleanup();
 }
@@ -465,7 +466,7 @@  phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr,
 		size_t mapping_size, size_t alloc_size,
 		enum dma_data_direction dir, unsigned long attrs)
 {
-	dma_addr_t tbl_dma_addr = phys_to_dma_unencrypted(hwdev, io_tlb_start);
+	dma_addr_t tbl_dma_addr = phys_to_dma_unencrypted(hwdev, io_tlb_start[SWIOTLB_LO]);
 	unsigned long flags;
 	phys_addr_t tlb_addr;
 	unsigned int nslots, stride, index, wrap;
@@ -475,7 +476,7 @@  phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr,
 	unsigned long max_slots;
 	unsigned long tmp_io_tlb_used;
 
-	if (no_iotlb_memory)
+	if (no_iotlb_memory[SWIOTLB_LO])
 		panic("Can not allocate SWIOTLB buffer earlier and can't now provide you with the DMA bounce buffer");
 
 	if (mem_encrypt_active())
@@ -518,11 +519,11 @@  phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr,
 	 */
 	spin_lock_irqsave(&io_tlb_lock, flags);
 
-	if (unlikely(nslots > io_tlb_nslabs - io_tlb_used))
+	if (unlikely(nslots > io_tlb_nslabs[SWIOTLB_LO] - io_tlb_used[SWIOTLB_LO]))
 		goto not_found;
 
-	index = ALIGN(io_tlb_index, stride);
-	if (index >= io_tlb_nslabs)
+	index = ALIGN(io_tlb_index[SWIOTLB_LO], stride);
+	if (index >= io_tlb_nslabs[SWIOTLB_LO])
 		index = 0;
 	wrap = index;
 
@@ -530,7 +531,7 @@  phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr,
 		while (iommu_is_span_boundary(index, nslots, offset_slots,
 					      max_slots)) {
 			index += stride;
-			if (index >= io_tlb_nslabs)
+			if (index >= io_tlb_nslabs[SWIOTLB_LO])
 				index = 0;
 			if (index == wrap)
 				goto not_found;
@@ -541,39 +542,42 @@  phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr,
 		 * contiguous buffers, we allocate the buffers from that slot
 		 * and mark the entries as '0' indicating unavailable.
 		 */
-		if (io_tlb_list[index] >= nslots) {
+		if (io_tlb_list[SWIOTLB_LO][index] >= nslots) {
 			int count = 0;
 
 			for (i = index; i < (int) (index + nslots); i++)
-				io_tlb_list[i] = 0;
-			for (i = index - 1; (OFFSET(i, IO_TLB_SEGSIZE) != IO_TLB_SEGSIZE - 1) && io_tlb_list[i]; i--)
-				io_tlb_list[i] = ++count;
-			tlb_addr = io_tlb_start + (index << IO_TLB_SHIFT);
+				io_tlb_list[SWIOTLB_LO][i] = 0;
+			for (i = index - 1;
+			     (OFFSET(i, IO_TLB_SEGSIZE) != IO_TLB_SEGSIZE - 1) &&
+			     io_tlb_list[SWIOTLB_LO][i];
+			     i--)
+				io_tlb_list[SWIOTLB_LO][i] = ++count;
+			tlb_addr = io_tlb_start[SWIOTLB_LO] + (index << IO_TLB_SHIFT);
 
 			/*
 			 * Update the indices to avoid searching in the next
 			 * round.
 			 */
-			io_tlb_index = ((index + nslots) < io_tlb_nslabs
+			io_tlb_index[SWIOTLB_LO] = ((index + nslots) < io_tlb_nslabs[SWIOTLB_LO]
 					? (index + nslots) : 0);
 
 			goto found;
 		}
 		index += stride;
-		if (index >= io_tlb_nslabs)
+		if (index >= io_tlb_nslabs[SWIOTLB_LO])
 			index = 0;
 	} while (index != wrap);
 
 not_found:
-	tmp_io_tlb_used = io_tlb_used;
+	tmp_io_tlb_used = io_tlb_used[SWIOTLB_LO];
 
 	spin_unlock_irqrestore(&io_tlb_lock, flags);
 	if (!(attrs & DMA_ATTR_NO_WARN) && printk_ratelimit())
 		dev_warn(hwdev, "swiotlb buffer is full (sz: %zd bytes), total %lu (slots), used %lu (slots)\n",
-			 alloc_size, io_tlb_nslabs, tmp_io_tlb_used);
+			 alloc_size, io_tlb_nslabs[SWIOTLB_LO], tmp_io_tlb_used);
 	return (phys_addr_t)DMA_MAPPING_ERROR;
 found:
-	io_tlb_used += nslots;
+	io_tlb_used[SWIOTLB_LO] += nslots;
 	spin_unlock_irqrestore(&io_tlb_lock, flags);
 
 	/*
@@ -582,7 +586,7 @@  phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr,
 	 * needed.
 	 */
 	for (i = 0; i < nslots; i++)
-		io_tlb_orig_addr[index+i] = orig_addr + (i << IO_TLB_SHIFT);
+		io_tlb_orig_addr[SWIOTLB_LO][index+i] = orig_addr + (i << IO_TLB_SHIFT);
 	if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
 	    (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL))
 		swiotlb_bounce(orig_addr, tlb_addr, mapping_size, DMA_TO_DEVICE);
@@ -599,8 +603,8 @@  void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
 {
 	unsigned long flags;
 	int i, count, nslots = ALIGN(alloc_size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
-	int index = (tlb_addr - io_tlb_start) >> IO_TLB_SHIFT;
-	phys_addr_t orig_addr = io_tlb_orig_addr[index];
+	int index = (tlb_addr - io_tlb_start[SWIOTLB_LO]) >> IO_TLB_SHIFT;
+	phys_addr_t orig_addr = io_tlb_orig_addr[SWIOTLB_LO][index];
 
 	/*
 	 * First, sync the memory before unmapping the entry
@@ -619,23 +623,26 @@  void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
 	spin_lock_irqsave(&io_tlb_lock, flags);
 	{
 		count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
-			 io_tlb_list[index + nslots] : 0);
+			 io_tlb_list[SWIOTLB_LO][index + nslots] : 0);
 		/*
 		 * Step 1: return the slots to the free list, merging the
 		 * slots with superceeding slots
 		 */
 		for (i = index + nslots - 1; i >= index; i--) {
-			io_tlb_list[i] = ++count;
-			io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
+			io_tlb_list[SWIOTLB_LO][i] = ++count;
+			io_tlb_orig_addr[SWIOTLB_LO][i] = INVALID_PHYS_ADDR;
 		}
 		/*
 		 * Step 2: merge the returned slots with the preceding slots,
 		 * if available (non zero)
 		 */
-		for (i = index - 1; (OFFSET(i, IO_TLB_SEGSIZE) != IO_TLB_SEGSIZE -1) && io_tlb_list[i]; i--)
-			io_tlb_list[i] = ++count;
+		for (i = index - 1;
+		     (OFFSET(i, IO_TLB_SEGSIZE) != IO_TLB_SEGSIZE - 1) &&
+		     io_tlb_list[SWIOTLB_LO][i];
+		     i--)
+			io_tlb_list[SWIOTLB_LO][i] = ++count;
 
-		io_tlb_used -= nslots;
+		io_tlb_used[SWIOTLB_LO] -= nslots;
 	}
 	spin_unlock_irqrestore(&io_tlb_lock, flags);
 }
@@ -644,8 +651,8 @@  void swiotlb_tbl_sync_single(struct device *hwdev, phys_addr_t tlb_addr,
 			     size_t size, enum dma_data_direction dir,
 			     enum dma_sync_target target)
 {
-	int index = (tlb_addr - io_tlb_start) >> IO_TLB_SHIFT;
-	phys_addr_t orig_addr = io_tlb_orig_addr[index];
+	int index = (tlb_addr - io_tlb_start[SWIOTLB_LO]) >> IO_TLB_SHIFT;
+	phys_addr_t orig_addr = io_tlb_orig_addr[SWIOTLB_LO][index];
 
 	if (orig_addr == INVALID_PHYS_ADDR)
 		return;
@@ -716,7 +723,7 @@  bool is_swiotlb_active(void)
 	 * When SWIOTLB is initialized, even if io_tlb_start points to physical
 	 * address zero, io_tlb_end surely doesn't.
 	 */
-	return io_tlb_end != 0;
+	return io_tlb_end[SWIOTLB_LO] != 0;
 }
 
 #ifdef CONFIG_DEBUG_FS
@@ -726,8 +733,8 @@  static int __init swiotlb_create_debugfs(void)
 	struct dentry *root;
 
 	root = debugfs_create_dir("swiotlb", NULL);
-	debugfs_create_ulong("io_tlb_nslabs", 0400, root, &io_tlb_nslabs);
-	debugfs_create_ulong("io_tlb_used", 0400, root, &io_tlb_used);
+	debugfs_create_ulong("io_tlb_nslabs", 0400, root, &io_tlb_nslabs[SWIOTLB_LO]);
+	debugfs_create_ulong("io_tlb_used", 0400, root, &io_tlb_used[SWIOTLB_LO]);
 	return 0;
 }