diff --git a/MAINTAINERS b/MAINTAINERS
index f5526da33d4c..9278c30ef1d5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -11841,6 +11841,7 @@ T:	git git://git.kernel.org/pub/scm/linux/kernel/git/iommu/linux.git
 F:	drivers/iommu/dma-iommu.c
 F:	drivers/iommu/dma-iommu.h
 F:	drivers/iommu/iova.c
+F:	include/linux/iommu-dma.h
 F:	include/linux/iova.h
 
 IOMMU SUBSYSTEM
diff --git a/arch/Kconfig b/arch/Kconfig
index 4e2eaba9e305..405c85ab86f2 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -17,6 +17,15 @@ config CPU_MITIGATIONS
 	def_bool y
 endif
 
+#
+# Selected by architectures that need custom DMA operations for e.g. legacy
+# IOMMUs not handled by dma-iommu.  Drivers must never select this symbol.
+#
+config ARCH_HAS_DMA_OPS
+	depends on HAS_DMA
+	select DMA_OPS_HELPERS
+	bool
+
 menu "General architecture-dependent options"
 
 config ARCH_HAS_SUBPAGE_FAULTS
diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig
index 50ff06d5b799..109a4cddcd13 100644
--- a/arch/alpha/Kconfig
+++ b/arch/alpha/Kconfig
@@ -4,12 +4,12 @@ config ALPHA
 	default y
 	select ARCH_32BIT_USTAT_F_TINODE
 	select ARCH_HAS_CURRENT_STACK_POINTER
+	select ARCH_HAS_DMA_OPS if PCI
 	select ARCH_MIGHT_HAVE_PC_PARPORT
 	select ARCH_MIGHT_HAVE_PC_SERIO
 	select ARCH_NO_PREEMPT
 	select ARCH_NO_SG_CHAIN
 	select ARCH_USE_CMPXCHG_LOCKREF
-	select DMA_OPS if PCI
 	select FORCE_PCI
 	select PCI_DOMAINS if PCI
 	select PCI_SYSCALL if PCI
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 0ec034933cae..749179a1d162 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -10,6 +10,7 @@ config ARM
 	select ARCH_HAS_CURRENT_STACK_POINTER
 	select ARCH_HAS_DEBUG_VIRTUAL if MMU
 	select ARCH_HAS_DMA_ALLOC if MMU
+	select ARCH_HAS_DMA_OPS
 	select ARCH_HAS_DMA_WRITE_COMBINE if !ARM_DMA_MEM_BUFFERABLE
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_FORTIFY_SOURCE
@@ -54,7 +55,6 @@ config ARM
 	select DCACHE_WORD_ACCESS if HAVE_EFFICIENT_UNALIGNED_ACCESS
 	select DMA_DECLARE_COHERENT
 	select DMA_GLOBAL_POOL if !MMU
-	select DMA_OPS
 	select DMA_NONCOHERENT_MMAP if MMU
 	select EDAC_SUPPORT
 	select EDAC_ATOMIC_SCRUB
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 8f2df6ed745d..690e90b62d92 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -24,6 +24,7 @@ config ARM64
 	select ARCH_HAS_CURRENT_STACK_POINTER
 	select ARCH_HAS_DEBUG_VIRTUAL
 	select ARCH_HAS_DEBUG_VM_PGTABLE
+	select ARCH_HAS_DMA_OPS if XEN
 	select ARCH_HAS_DMA_PREP_COHERENT
 	select ARCH_HAS_ACPI_TABLE_UPGRADE if ACPI
 	select ARCH_HAS_FAST_MULTIPLIER
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index a0400b9aa814..27a32ff15412 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -114,36 +114,33 @@ static void __init arch_reserve_crashkernel(void)
 				    low_size, high);
 }
 
-/*
- * Return the maximum physical address for a zone accessible by the given bits
- * limit. If DRAM starts above 32-bit, expand the zone to the maximum
- * available memory, otherwise cap it at 32-bit.
- */
-static phys_addr_t __init max_zone_phys(unsigned int zone_bits)
+static phys_addr_t __init max_zone_phys(phys_addr_t zone_limit)
 {
-	phys_addr_t zone_mask = DMA_BIT_MASK(zone_bits);
-	phys_addr_t phys_start = memblock_start_of_DRAM();
+	/**
+	 * Information we get from firmware (e.g. DT dma-ranges) describe DMA
+	 * bus constraints. Devices using DMA might have their own limitations.
+	 * Some of them rely on DMA zone in low 32-bit memory. Keep low RAM
+	 * DMA zone on platforms that have RAM there.
+	 */
+	if (memblock_start_of_DRAM() < U32_MAX)
+		zone_limit = min(zone_limit, U32_MAX);
 
-	if (phys_start > U32_MAX)
-		zone_mask = PHYS_ADDR_MAX;
-	else if (phys_start > zone_mask)
-		zone_mask = U32_MAX;
-
-	return min(zone_mask, memblock_end_of_DRAM() - 1) + 1;
+	return min(zone_limit, memblock_end_of_DRAM() - 1) + 1;
 }
 
 static void __init zone_sizes_init(void)
 {
 	unsigned long max_zone_pfns[MAX_NR_ZONES]  = {0};
-	unsigned int __maybe_unused acpi_zone_dma_bits;
-	unsigned int __maybe_unused dt_zone_dma_bits;
-	phys_addr_t __maybe_unused dma32_phys_limit = max_zone_phys(32);
+	phys_addr_t __maybe_unused acpi_zone_dma_limit;
+	phys_addr_t __maybe_unused dt_zone_dma_limit;
+	phys_addr_t __maybe_unused dma32_phys_limit =
+		max_zone_phys(DMA_BIT_MASK(32));
 
 #ifdef CONFIG_ZONE_DMA
-	acpi_zone_dma_bits = fls64(acpi_iort_dma_get_max_cpu_address());
-	dt_zone_dma_bits = fls64(of_dma_get_max_cpu_address(NULL));
-	zone_dma_bits = min3(32U, dt_zone_dma_bits, acpi_zone_dma_bits);
-	arm64_dma_phys_limit = max_zone_phys(zone_dma_bits);
+	acpi_zone_dma_limit = acpi_iort_dma_get_max_cpu_address();
+	dt_zone_dma_limit = of_dma_get_max_cpu_address(NULL);
+	zone_dma_limit = min(dt_zone_dma_limit, acpi_zone_dma_limit);
+	arm64_dma_phys_limit = max_zone_phys(zone_dma_limit);
 	max_zone_pfns[ZONE_DMA] = PFN_DOWN(arm64_dma_phys_limit);
 #endif
 #ifdef CONFIG_ZONE_DMA32
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 60077e576935..023ad33a7e94 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -8,6 +8,7 @@ config MIPS
 	select ARCH_HAS_CPU_FINALIZE_INIT
 	select ARCH_HAS_CURRENT_STACK_POINTER if !CC_IS_CLANG || CLANG_VERSION >= 140000
 	select ARCH_HAS_DEBUG_VIRTUAL if !64BIT
+	select ARCH_HAS_DMA_OPS if MACH_JAZZ
 	select ARCH_HAS_FORTIFY_SOURCE
 	select ARCH_HAS_KCOV
 	select ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE if !EVA
@@ -393,7 +394,6 @@ config MACH_JAZZ
 	select ARC_PROMLIB
 	select ARCH_MIGHT_HAVE_PC_PARPORT
 	select ARCH_MIGHT_HAVE_PC_SERIO
-	select DMA_OPS
 	select FW_ARC
 	select FW_ARC32
 	select ARCH_MAY_HAVE_PC_FDC
diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig
index db3edd54d6f2..aa6a3cad275d 100644
--- a/arch/parisc/Kconfig
+++ b/arch/parisc/Kconfig
@@ -10,6 +10,7 @@ config PARISC
 	select ARCH_WANT_FRAME_POINTERS
 	select ARCH_HAS_CPU_CACHE_ALIASING
 	select ARCH_HAS_DMA_ALLOC if PA11
+	select ARCH_HAS_DMA_OPS
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_STRICT_KERNEL_RWX
 	select ARCH_HAS_STRICT_MODULE_RWX
@@ -23,7 +24,6 @@ config PARISC
 	select ARCH_HAS_CACHE_LINE_SIZE
 	select ARCH_HAS_DEBUG_VM_PGTABLE
 	select HAVE_RELIABLE_STACKTRACE
-	select DMA_OPS
 	select RTC_CLASS
 	select RTC_DRV_GENERIC
 	select INIT_ALL_POSSIBLE
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 0b8b2e3a6381..8094a01974cc 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -133,6 +133,7 @@ config PPC
 	select ARCH_HAS_DEBUG_WX		if STRICT_KERNEL_RWX
 	select ARCH_HAS_DEVMEM_IS_ALLOWED
 	select ARCH_HAS_DMA_MAP_DIRECT 		if PPC_PSERIES
+	select ARCH_HAS_DMA_OPS			if PPC64
 	select ARCH_HAS_FORTIFY_SOURCE
 	select ARCH_HAS_GCOV_PROFILE_ALL
 	select ARCH_HAS_KCOV
@@ -185,7 +186,6 @@ config PPC
 	select CPUMASK_OFFSTACK			if NR_CPUS >= 8192
 	select DCACHE_WORD_ACCESS		if PPC64 && CPU_LITTLE_ENDIAN
 	select DMA_OPS_BYPASS			if PPC64
-	select DMA_OPS				if PPC64
 	select DYNAMIC_FTRACE			if FUNCTION_TRACER
 	select EDAC_ATOMIC_SCRUB
 	select EDAC_SUPPORT
diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c
index da606ef18eae..1221c561b43a 100644
--- a/arch/powerpc/mm/mem.c
+++ b/arch/powerpc/mm/mem.c
@@ -216,7 +216,7 @@ static int __init mark_nonram_nosave(void)
  * everything else. GFP_DMA32 page allocations automatically fall back to
  * ZONE_DMA.
  *
- * By using 31-bit unconditionally, we can exploit zone_dma_bits to inform the
+ * By using 31-bit unconditionally, we can exploit zone_dma_limit to inform the
  * generic DMA mapping code.  32-bit only devices (if not handled by an IOMMU
  * anyway) will take a first dip into ZONE_NORMAL and get otherwise served by
  * ZONE_DMA.
@@ -230,6 +230,7 @@ void __init paging_init(void)
 {
 	unsigned long long total_ram = memblock_phys_mem_size();
 	phys_addr_t top_of_ram = memblock_end_of_DRAM();
+	int zone_dma_bits;
 
 #ifdef CONFIG_HIGHMEM
 	unsigned long v = __fix_to_virt(FIX_KMAP_END);
@@ -256,6 +257,8 @@ void __init paging_init(void)
 	else
 		zone_dma_bits = 31;
 
+	zone_dma_limit = DMA_BIT_MASK(zone_dma_bits);
+
 #ifdef CONFIG_ZONE_DMA
 	max_zone_pfns[ZONE_DMA]	= min(max_low_pfn,
 				      1UL << (zone_dma_bits - PAGE_SHIFT));
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index b0d0b3a8d196..b632112611cc 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -70,6 +70,7 @@ config S390
 	select ARCH_HAS_DEBUG_VM_PGTABLE
 	select ARCH_HAS_DEBUG_WX
 	select ARCH_HAS_DEVMEM_IS_ALLOWED
+	select ARCH_HAS_DMA_OPS if PCI
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_FORCE_DMA_UNENCRYPTED
 	select ARCH_HAS_FORTIFY_SOURCE
@@ -137,7 +138,6 @@ config S390
 	select BUILDTIME_TABLE_SORT
 	select CLONE_BACKWARDS2
 	select DCACHE_WORD_ACCESS if !KMSAN
-	select DMA_OPS if PCI
 	select DYNAMIC_FTRACE if FUNCTION_TRACER
 	select FUNCTION_ALIGNMENT_8B if CC_IS_GCC
 	select FUNCTION_ALIGNMENT_16B if !CC_IS_GCC
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index e3d258f9e726..688abc65c79e 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -97,7 +97,7 @@ void __init paging_init(void)
 
 	vmem_map_init();
 	sparse_init();
-	zone_dma_bits = 31;
+	zone_dma_limit = DMA_BIT_MASK(31);
 	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
 	max_zone_pfns[ZONE_DMA] = virt_to_pfn(MAX_DMA_ADDRESS);
 	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index 11bf9d312318..dcfdb7f1dae9 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -14,9 +14,9 @@ config SPARC
 	bool
 	default y
 	select ARCH_HAS_CPU_CACHE_ALIASING
+	select ARCH_HAS_DMA_OPS
 	select ARCH_MIGHT_HAVE_PC_PARPORT if SPARC64 && PCI
 	select ARCH_MIGHT_HAVE_PC_SERIO
-	select DMA_OPS
 	select OF
 	select OF_PROMTREE
 	select HAVE_ASM_MODVERSIONS
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 8f07de9e2406..961472357248 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -79,6 +79,7 @@ config X86
 	select ARCH_HAS_DEBUG_VIRTUAL
 	select ARCH_HAS_DEBUG_VM_PGTABLE	if !X86_PAE
 	select ARCH_HAS_DEVMEM_IS_ALLOWED
+	select ARCH_HAS_DMA_OPS			if GART_IOMMU || XEN
 	select ARCH_HAS_EARLY_DEBUG		if KGDB
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_FAST_MULTIPLIER
@@ -944,7 +945,6 @@ config DMI
 
 config GART_IOMMU
 	bool "Old AMD GART IOMMU support"
-	select DMA_OPS
 	select IOMMU_HELPER
 	select SWIOTLB
 	depends on X86_64 && PCI && AMD_NB
diff --git a/drivers/accel/qaic/qaic_drv.c b/drivers/accel/qaic/qaic_drv.c
index 580b29ed1902..bf10156c334e 100644
--- a/drivers/accel/qaic/qaic_drv.c
+++ b/drivers/accel/qaic/qaic_drv.c
@@ -447,9 +447,7 @@ static int init_pci(struct qaic_device *qdev, struct pci_dev *pdev)
 	ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
 	if (ret)
 		return ret;
-	ret = dma_set_max_seg_size(&pdev->dev, UINT_MAX);
-	if (ret)
-		return ret;
+	dma_set_max_seg_size(&pdev->dev, UINT_MAX);
 
 	qdev->bar_0 = devm_ioremap_resource(&pdev->dev, &pdev->resource[0]);
 	if (IS_ERR(qdev->bar_0))
diff --git a/drivers/dma/idma64.c b/drivers/dma/idma64.c
index e3505e56784b..1398814d8fbb 100644
--- a/drivers/dma/idma64.c
+++ b/drivers/dma/idma64.c
@@ -598,9 +598,7 @@ static int idma64_probe(struct idma64_chip *chip)
 
 	idma64->dma.dev = chip->sysdev;
 
-	ret = dma_set_max_seg_size(idma64->dma.dev, IDMA64C_CTLH_BLOCK_TS_MASK);
-	if (ret)
-		return ret;
+	dma_set_max_seg_size(idma64->dma.dev, IDMA64C_CTLH_BLOCK_TS_MASK);
 
 	ret = dma_async_device_register(&idma64->dma);
 	if (ret)
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index 60c4de8dac1d..82a9fe88ad54 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -3163,10 +3163,7 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 	 * This is the limit for transfers with a buswidth of 1, larger
 	 * buswidths will have larger limits.
 	 */
-	ret = dma_set_max_seg_size(&adev->dev, 1900800);
-	if (ret)
-		dev_err(&adev->dev, "unable to set the seg size\n");
-
+	dma_set_max_seg_size(&adev->dev, 1900800);
 
 	init_pl330_debugfs(pl330);
 	dev_info(&adev->dev,
diff --git a/drivers/dma/qcom/bam_dma.c b/drivers/dma/qcom/bam_dma.c
index 5e7d332731e0..368ffaa40037 100644
--- a/drivers/dma/qcom/bam_dma.c
+++ b/drivers/dma/qcom/bam_dma.c
@@ -1325,11 +1325,7 @@ static int bam_dma_probe(struct platform_device *pdev)
 
 	/* set max dma segment size */
 	bdev->common.dev = bdev->dev;
-	ret = dma_set_max_seg_size(bdev->common.dev, BAM_FIFO_SIZE);
-	if (ret) {
-		dev_err(bdev->dev, "cannot set maximum segment size\n");
-		goto err_bam_channel_exit;
-	}
+	dma_set_max_seg_size(bdev->common.dev, BAM_FIFO_SIZE);
 
 	platform_set_drvdata(pdev, bdev);
 
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index 40482cb73d79..1094a2f82164 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -1868,9 +1868,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
 
 	dmac->dev = &pdev->dev;
 	platform_set_drvdata(pdev, dmac);
-	ret = dma_set_max_seg_size(dmac->dev, RCAR_DMATCR_MASK);
-	if (ret)
-		return ret;
+	dma_set_max_seg_size(dmac->dev, RCAR_DMATCR_MASK);
 
 	ret = dma_set_mask_and_coherent(dmac->dev, DMA_BIT_MASK(40));
 	if (ret)
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index 2c489299148e..d52e1685aed5 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -3632,11 +3632,7 @@ static int __init d40_probe(struct platform_device *pdev)
 	if (ret)
 		goto destroy_cache;
 
-	ret = dma_set_max_seg_size(base->dev, STEDMA40_MAX_SEG_SIZE);
-	if (ret) {
-		d40_err(dev, "Failed to set dma max seg size\n");
-		goto destroy_cache;
-	}
+	dma_set_max_seg_size(base->dev, STEDMA40_MAX_SEG_SIZE);
 
 	d40_hw_init(base);
 
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
index 77b50c56c124..3e807195a0d0 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
@@ -559,11 +559,7 @@ static int mtk_drm_kms_init(struct drm_device *drm)
 	 * Configure the DMA segment size to make sure we get contiguous IOVA
 	 * when importing PRIME buffers.
 	 */
-	ret = dma_set_max_seg_size(dma_dev, UINT_MAX);
-	if (ret) {
-		dev_err(dma_dev, "Failed to set DMA segment size\n");
-		goto err_component_unbind;
-	}
+	dma_set_max_seg_size(dma_dev, UINT_MAX);
 
 	ret = drm_vblank_init(drm, MAX_CRTC);
 	if (ret < 0)
diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig
index 22addaedf64d..b3aa1f5d5321 100644
--- a/drivers/iommu/Kconfig
+++ b/drivers/iommu/Kconfig
@@ -151,7 +151,7 @@ config OF_IOMMU
 # IOMMU-agnostic DMA-mapping layer
 config IOMMU_DMA
 	def_bool ARM64 || X86 || S390
-	select DMA_OPS
+	select DMA_OPS_HELPERS
 	select IOMMU_API
 	select IOMMU_IOVA
 	select IRQ_MSI_IOMMU
diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index 7b1dfa0665df..3672d619bcb6 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -17,6 +17,7 @@
 #include <linux/gfp.h>
 #include <linux/huge_mm.h>
 #include <linux/iommu.h>
+#include <linux/iommu-dma.h>
 #include <linux/iova.h>
 #include <linux/irq.h>
 #include <linux/list_sort.h>
@@ -1037,9 +1038,8 @@ out_unmap:
 	return NULL;
 }
 
-static struct sg_table *iommu_dma_alloc_noncontiguous(struct device *dev,
-		size_t size, enum dma_data_direction dir, gfp_t gfp,
-		unsigned long attrs)
+struct sg_table *iommu_dma_alloc_noncontiguous(struct device *dev, size_t size,
+	       enum dma_data_direction dir, gfp_t gfp, unsigned long attrs)
 {
 	struct dma_sgt_handle *sh;
 
@@ -1055,7 +1055,7 @@ static struct sg_table *iommu_dma_alloc_noncontiguous(struct device *dev,
 	return &sh->sgt;
 }
 
-static void iommu_dma_free_noncontiguous(struct device *dev, size_t size,
+void iommu_dma_free_noncontiguous(struct device *dev, size_t size,
 		struct sg_table *sgt, enum dma_data_direction dir)
 {
 	struct dma_sgt_handle *sh = sgt_handle(sgt);
@@ -1066,8 +1066,8 @@ static void iommu_dma_free_noncontiguous(struct device *dev, size_t size,
 	kfree(sh);
 }
 
-static void iommu_dma_sync_single_for_cpu(struct device *dev,
-		dma_addr_t dma_handle, size_t size, enum dma_data_direction dir)
+void iommu_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle,
+		size_t size, enum dma_data_direction dir)
 {
 	phys_addr_t phys;
 
@@ -1081,8 +1081,8 @@ static void iommu_dma_sync_single_for_cpu(struct device *dev,
 	swiotlb_sync_single_for_cpu(dev, phys, size, dir);
 }
 
-static void iommu_dma_sync_single_for_device(struct device *dev,
-		dma_addr_t dma_handle, size_t size, enum dma_data_direction dir)
+void iommu_dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle,
+		size_t size, enum dma_data_direction dir)
 {
 	phys_addr_t phys;
 
@@ -1096,9 +1096,8 @@ static void iommu_dma_sync_single_for_device(struct device *dev,
 		arch_sync_dma_for_device(phys, size, dir);
 }
 
-static void iommu_dma_sync_sg_for_cpu(struct device *dev,
-		struct scatterlist *sgl, int nelems,
-		enum dma_data_direction dir)
+void iommu_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sgl,
+		int nelems, enum dma_data_direction dir)
 {
 	struct scatterlist *sg;
 	int i;
@@ -1112,9 +1111,8 @@ static void iommu_dma_sync_sg_for_cpu(struct device *dev,
 			arch_sync_dma_for_cpu(sg_phys(sg), sg->length, dir);
 }
 
-static void iommu_dma_sync_sg_for_device(struct device *dev,
-		struct scatterlist *sgl, int nelems,
-		enum dma_data_direction dir)
+void iommu_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sgl,
+		int nelems, enum dma_data_direction dir)
 {
 	struct scatterlist *sg;
 	int i;
@@ -1129,9 +1127,9 @@ static void iommu_dma_sync_sg_for_device(struct device *dev,
 			arch_sync_dma_for_device(sg_phys(sg), sg->length, dir);
 }
 
-static dma_addr_t iommu_dma_map_page(struct device *dev, struct page *page,
-		unsigned long offset, size_t size, enum dma_data_direction dir,
-		unsigned long attrs)
+dma_addr_t iommu_dma_map_page(struct device *dev, struct page *page,
+	      unsigned long offset, size_t size, enum dma_data_direction dir,
+	      unsigned long attrs)
 {
 	phys_addr_t phys = page_to_phys(page) + offset;
 	bool coherent = dev_is_dma_coherent(dev);
@@ -1189,7 +1187,7 @@ static dma_addr_t iommu_dma_map_page(struct device *dev, struct page *page,
 	return iova;
 }
 
-static void iommu_dma_unmap_page(struct device *dev, dma_addr_t dma_handle,
+void iommu_dma_unmap_page(struct device *dev, dma_addr_t dma_handle,
 		size_t size, enum dma_data_direction dir, unsigned long attrs)
 {
 	struct iommu_domain *domain = iommu_get_dma_domain(dev);
@@ -1342,8 +1340,8 @@ out_unmap:
  * impedance-matching, to be able to hand off a suitably-aligned list,
  * but still preserve the original offsets and sizes for the caller.
  */
-static int iommu_dma_map_sg(struct device *dev, struct scatterlist *sg,
-		int nents, enum dma_data_direction dir, unsigned long attrs)
+int iommu_dma_map_sg(struct device *dev, struct scatterlist *sg, int nents,
+		enum dma_data_direction dir, unsigned long attrs)
 {
 	struct iommu_domain *domain = iommu_get_dma_domain(dev);
 	struct iommu_dma_cookie *cookie = domain->iova_cookie;
@@ -1462,8 +1460,8 @@ out:
 	return ret;
 }
 
-static void iommu_dma_unmap_sg(struct device *dev, struct scatterlist *sg,
-		int nents, enum dma_data_direction dir, unsigned long attrs)
+void iommu_dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nents,
+		enum dma_data_direction dir, unsigned long attrs)
 {
 	dma_addr_t end = 0, start;
 	struct scatterlist *tmp;
@@ -1512,7 +1510,7 @@ static void iommu_dma_unmap_sg(struct device *dev, struct scatterlist *sg,
 		__iommu_dma_unmap(dev, start, end - start);
 }
 
-static dma_addr_t iommu_dma_map_resource(struct device *dev, phys_addr_t phys,
+dma_addr_t iommu_dma_map_resource(struct device *dev, phys_addr_t phys,
 		size_t size, enum dma_data_direction dir, unsigned long attrs)
 {
 	return __iommu_dma_map(dev, phys, size,
@@ -1520,7 +1518,7 @@ static dma_addr_t iommu_dma_map_resource(struct device *dev, phys_addr_t phys,
 			dma_get_mask(dev));
 }
 
-static void iommu_dma_unmap_resource(struct device *dev, dma_addr_t handle,
+void iommu_dma_unmap_resource(struct device *dev, dma_addr_t handle,
 		size_t size, enum dma_data_direction dir, unsigned long attrs)
 {
 	__iommu_dma_unmap(dev, handle, size);
@@ -1557,7 +1555,7 @@ static void __iommu_dma_free(struct device *dev, size_t size, void *cpu_addr)
 		dma_free_contiguous(dev, page, alloc_size);
 }
 
-static void iommu_dma_free(struct device *dev, size_t size, void *cpu_addr,
+void iommu_dma_free(struct device *dev, size_t size, void *cpu_addr,
 		dma_addr_t handle, unsigned long attrs)
 {
 	__iommu_dma_unmap(dev, handle, size);
@@ -1601,8 +1599,8 @@ out_free_pages:
 	return NULL;
 }
 
-static void *iommu_dma_alloc(struct device *dev, size_t size,
-		dma_addr_t *handle, gfp_t gfp, unsigned long attrs)
+void *iommu_dma_alloc(struct device *dev, size_t size, dma_addr_t *handle,
+		gfp_t gfp, unsigned long attrs)
 {
 	bool coherent = dev_is_dma_coherent(dev);
 	int ioprot = dma_info_to_prot(DMA_BIDIRECTIONAL, coherent, attrs);
@@ -1635,7 +1633,7 @@ static void *iommu_dma_alloc(struct device *dev, size_t size,
 	return cpu_addr;
 }
 
-static int iommu_dma_mmap(struct device *dev, struct vm_area_struct *vma,
+int iommu_dma_mmap(struct device *dev, struct vm_area_struct *vma,
 		void *cpu_addr, dma_addr_t dma_addr, size_t size,
 		unsigned long attrs)
 {
@@ -1666,7 +1664,7 @@ static int iommu_dma_mmap(struct device *dev, struct vm_area_struct *vma,
 			       vma->vm_page_prot);
 }
 
-static int iommu_dma_get_sgtable(struct device *dev, struct sg_table *sgt,
+int iommu_dma_get_sgtable(struct device *dev, struct sg_table *sgt,
 		void *cpu_addr, dma_addr_t dma_addr, size_t size,
 		unsigned long attrs)
 {
@@ -1693,19 +1691,19 @@ static int iommu_dma_get_sgtable(struct device *dev, struct sg_table *sgt,
 	return ret;
 }
 
-static unsigned long iommu_dma_get_merge_boundary(struct device *dev)
+unsigned long iommu_dma_get_merge_boundary(struct device *dev)
 {
 	struct iommu_domain *domain = iommu_get_dma_domain(dev);
 
 	return (1UL << __ffs(domain->pgsize_bitmap)) - 1;
 }
 
-static size_t iommu_dma_opt_mapping_size(void)
+size_t iommu_dma_opt_mapping_size(void)
 {
 	return iova_rcache_range();
 }
 
-static size_t iommu_dma_max_mapping_size(struct device *dev)
+size_t iommu_dma_max_mapping_size(struct device *dev)
 {
 	if (dev_is_untrusted(dev))
 		return swiotlb_max_mapping_size(dev);
@@ -1713,32 +1711,6 @@ static size_t iommu_dma_max_mapping_size(struct device *dev)
 	return SIZE_MAX;
 }
 
-static const struct dma_map_ops iommu_dma_ops = {
-	.flags			= DMA_F_PCI_P2PDMA_SUPPORTED |
-				  DMA_F_CAN_SKIP_SYNC,
-	.alloc			= iommu_dma_alloc,
-	.free			= iommu_dma_free,
-	.alloc_pages_op		= dma_common_alloc_pages,
-	.free_pages		= dma_common_free_pages,
-	.alloc_noncontiguous	= iommu_dma_alloc_noncontiguous,
-	.free_noncontiguous	= iommu_dma_free_noncontiguous,
-	.mmap			= iommu_dma_mmap,
-	.get_sgtable		= iommu_dma_get_sgtable,
-	.map_page		= iommu_dma_map_page,
-	.unmap_page		= iommu_dma_unmap_page,
-	.map_sg			= iommu_dma_map_sg,
-	.unmap_sg		= iommu_dma_unmap_sg,
-	.sync_single_for_cpu	= iommu_dma_sync_single_for_cpu,
-	.sync_single_for_device	= iommu_dma_sync_single_for_device,
-	.sync_sg_for_cpu	= iommu_dma_sync_sg_for_cpu,
-	.sync_sg_for_device	= iommu_dma_sync_sg_for_device,
-	.map_resource		= iommu_dma_map_resource,
-	.unmap_resource		= iommu_dma_unmap_resource,
-	.get_merge_boundary	= iommu_dma_get_merge_boundary,
-	.opt_mapping_size	= iommu_dma_opt_mapping_size,
-	.max_mapping_size       = iommu_dma_max_mapping_size,
-};
-
 void iommu_setup_dma_ops(struct device *dev)
 {
 	struct iommu_domain *domain = iommu_get_domain_for_dev(dev);
@@ -1746,19 +1718,15 @@ void iommu_setup_dma_ops(struct device *dev)
 	if (dev_is_pci(dev))
 		dev->iommu->pci_32bit_workaround = !iommu_dma_forcedac;
 
-	if (iommu_is_dma_domain(domain)) {
-		if (iommu_dma_init_domain(domain, dev))
-			goto out_err;
-		dev->dma_ops = &iommu_dma_ops;
-	} else if (dev->dma_ops == &iommu_dma_ops) {
-		/* Clean up if we've switched *from* a DMA domain */
-		dev->dma_ops = NULL;
-	}
+	dev->dma_iommu = iommu_is_dma_domain(domain);
+	if (dev->dma_iommu && iommu_dma_init_domain(domain, dev))
+		goto out_err;
 
 	return;
 out_err:
-	 pr_warn("Failed to set up IOMMU for device %s; retaining platform DMA ops\n",
-		 dev_name(dev));
+	pr_warn("Failed to set up IOMMU for device %s; retaining platform DMA ops\n",
+		dev_name(dev));
+	dev->dma_iommu = false;
 }
 
 static struct iommu_dma_msi_page *iommu_dma_get_msi_page(struct device *dev,
diff --git a/drivers/iommu/intel/Kconfig b/drivers/iommu/intel/Kconfig
index f52fb39c968e..88fd32a9323c 100644
--- a/drivers/iommu/intel/Kconfig
+++ b/drivers/iommu/intel/Kconfig
@@ -12,7 +12,6 @@ config DMAR_DEBUG
 config INTEL_IOMMU
 	bool "Support for Intel IOMMU using DMA Remapping Devices"
 	depends on PCI_MSI && ACPI && X86
-	select DMA_OPS
 	select IOMMU_API
 	select IOMMU_IOVA
 	select IOMMUFD_DRIVER if IOMMUFD
diff --git a/drivers/macintosh/macio_asic.c b/drivers/macintosh/macio_asic.c
index 13626205530d..bede200e32e8 100644
--- a/drivers/macintosh/macio_asic.c
+++ b/drivers/macintosh/macio_asic.c
@@ -387,7 +387,7 @@ static struct macio_dev * macio_add_one_device(struct macio_chip *chip,
 	dma_set_max_seg_size(&dev->ofdev.dev, 65536);
 	dma_set_seg_boundary(&dev->ofdev.dev, 0xffffffff);
 
-#if defined(CONFIG_PCI) && defined(CONFIG_DMA_OPS)
+#if defined(CONFIG_PCI) && defined(CONFIG_ARCH_HAS_DMA_OPS)
 	/* Set the DMA ops to the ones from the PCI device, this could be
 	 * fishy if we didn't know that on PowerMac it's always direct ops
 	 * or iommu ops that will work fine
@@ -396,7 +396,7 @@ static struct macio_dev * macio_add_one_device(struct macio_chip *chip,
 	 */
 	dev->ofdev.dev.archdata = chip->lbus.pdev->dev.archdata;
 	dev->ofdev.dev.dma_ops = chip->lbus.pdev->dev.dma_ops;
-#endif /* CONFIG_PCI && CONFIG_DMA_OPS */
+#endif /* CONFIG_PCI && CONFIG_ARCH_HAS_DMA_OPS */
 
 #ifdef DEBUG
 	printk("preparing mdev @%p, ofdev @%p, dev @%p, kobj @%p\n",
diff --git a/drivers/media/common/videobuf2/videobuf2-dma-contig.c b/drivers/media/common/videobuf2/videobuf2-dma-contig.c
index 3d4fd4ef5310..bb0b7fa67b53 100644
--- a/drivers/media/common/videobuf2/videobuf2-dma-contig.c
+++ b/drivers/media/common/videobuf2/videobuf2-dma-contig.c
@@ -854,8 +854,7 @@ int vb2_dma_contig_set_max_seg_size(struct device *dev, unsigned int size)
 		return -ENODEV;
 	}
 	if (dma_get_max_seg_size(dev) < size)
-		return dma_set_max_seg_size(dev, size);
-
+		dma_set_max_seg_size(dev, size);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(vb2_dma_contig_set_max_seg_size);
diff --git a/drivers/media/pci/intel/ipu6/Kconfig b/drivers/media/pci/intel/ipu6/Kconfig
index 40e20f0aa5ae..49e4fb696573 100644
--- a/drivers/media/pci/intel/ipu6/Kconfig
+++ b/drivers/media/pci/intel/ipu6/Kconfig
@@ -4,8 +4,13 @@ config VIDEO_INTEL_IPU6
 	depends on VIDEO_DEV
 	depends on X86 && X86_64 && HAS_DMA
 	depends on IPU_BRIDGE || !IPU_BRIDGE
+	#
+	# This driver incorrectly tries to override the dma_ops.  It should
+	# never have done that, but for now keep it working on architectures
+	# that use dma ops
+	#
+	depends on ARCH_HAS_DMA_OPS
 	select AUXILIARY_BUS
-	select DMA_OPS
 	select IOMMU_IOVA
 	select VIDEO_V4L2_SUBDEV_API
 	select MEDIA_CONTROLLER
diff --git a/drivers/media/pci/intel/ipu6/ipu6.c b/drivers/media/pci/intel/ipu6/ipu6.c
index bbd646378ab3..83e70c692d95 100644
--- a/drivers/media/pci/intel/ipu6/ipu6.c
+++ b/drivers/media/pci/intel/ipu6/ipu6.c
@@ -576,9 +576,7 @@ static int ipu6_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	if (ret)
 		return dev_err_probe(dev, ret, "Failed to set DMA mask\n");
 
-	ret = dma_set_max_seg_size(dev, UINT_MAX);
-	if (ret)
-		return dev_err_probe(dev, ret, "Failed to set max_seg_size\n");
+	dma_set_max_seg_size(dev, UINT_MAX);
 
 	ret = ipu6_pci_config_setup(pdev, isp->hw_ver);
 	if (ret)
diff --git a/drivers/mmc/host/mmci_stm32_sdmmc.c b/drivers/mmc/host/mmci_stm32_sdmmc.c
index f5da7f9baa52..9dc51859c2e5 100644
--- a/drivers/mmc/host/mmci_stm32_sdmmc.c
+++ b/drivers/mmc/host/mmci_stm32_sdmmc.c
@@ -213,7 +213,8 @@ static int sdmmc_idma_setup(struct mmci_host *host)
 		host->mmc->max_seg_size = host->mmc->max_req_size;
 	}
 
-	return dma_set_max_seg_size(dev, host->mmc->max_seg_size);
+	dma_set_max_seg_size(dev, host->mmc->max_seg_size);
+	return 0;
 }
 
 static int sdmmc_idma_start(struct mmci_host *host, unsigned int *datactrl)
diff --git a/drivers/net/ethernet/microsoft/mana/gdma_main.c b/drivers/net/ethernet/microsoft/mana/gdma_main.c
index ddb8f68d80a2..ca4ed58f1206 100644
--- a/drivers/net/ethernet/microsoft/mana/gdma_main.c
+++ b/drivers/net/ethernet/microsoft/mana/gdma_main.c
@@ -1496,11 +1496,7 @@ static int mana_gd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (err)
 		goto release_region;
 
-	err = dma_set_max_seg_size(&pdev->dev, UINT_MAX);
-	if (err) {
-		dev_err(&pdev->dev, "Failed to set dma device segment size\n");
-		goto release_region;
-	}
+	dma_set_max_seg_size(&pdev->dev, UINT_MAX);
 
 	err = -ENOMEM;
 	gc = vzalloc(sizeof(*gc));
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index e1dfa96c2a55..50620918becd 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -13861,12 +13861,7 @@ fcponly:
 	if (sli4_params->sge_supp_len > LPFC_MAX_SGE_SIZE)
 		sli4_params->sge_supp_len = LPFC_MAX_SGE_SIZE;
 
-	rc = dma_set_max_seg_size(&phba->pcidev->dev, sli4_params->sge_supp_len);
-	if (unlikely(rc)) {
-		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
-				"6400 Can't set dma maximum segment size\n");
-		return rc;
-	}
+	dma_set_max_seg_size(&phba->pcidev->dev, sli4_params->sge_supp_len);
 
 	/*
 	 * Check whether the adapter supports an embedded copy of the
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 3958a6d14bf4..7f0394c44920 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1988,8 +1988,15 @@ void scsi_init_limits(struct Scsi_Host *shost, struct queue_limits *lim)
 	if (shost->no_highmem)
 		lim->features |= BLK_FEAT_BOUNCE_HIGH;
 
-	dma_set_seg_boundary(dev, shost->dma_boundary);
-	dma_set_max_seg_size(dev, shost->max_segment_size);
+	/*
+	 * Propagate the DMA formation properties to the dma-mapping layer as
+	 * a courtesy service to the LLDDs.  This needs to check that the buses
+	 * actually support the DMA API first, though.
+	 */
+	if (dev->dma_parms) {
+		dma_set_seg_boundary(dev, shost->dma_boundary);
+		dma_set_max_seg_size(dev, shost->max_segment_size);
+	}
 }
 EXPORT_SYMBOL_GPL(scsi_init_limits);
 
diff --git a/drivers/vdpa/Kconfig b/drivers/vdpa/Kconfig
index 5265d09fc1c4..559fb9d3271f 100644
--- a/drivers/vdpa/Kconfig
+++ b/drivers/vdpa/Kconfig
@@ -11,8 +11,7 @@ if VDPA
 
 config VDPA_SIM
 	tristate "vDPA device simulator core"
-	depends on RUNTIME_TESTING_MENU && HAS_DMA
-	select DMA_OPS
+	depends on RUNTIME_TESTING_MENU
 	select VHOST_RING
 	select IOMMU_IOVA
 	help
@@ -36,7 +35,12 @@ config VDPA_SIM_BLOCK
 config VDPA_USER
 	tristate "VDUSE (vDPA Device in Userspace) support"
 	depends on EVENTFD && MMU && HAS_DMA
-	select DMA_OPS
+	#
+	# This driver incorrectly tries to override the dma_ops.  It should
+	# never have done that, but for now keep it working on architectures
+	# that use dma ops
+	#
+	depends on ARCH_HAS_DMA_OPS
 	select VHOST_IOTLB
 	select IOMMU_IOVA
 	help
diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig
index d5989871dd5d..f7d6f47971fd 100644
--- a/drivers/xen/Kconfig
+++ b/drivers/xen/Kconfig
@@ -177,8 +177,8 @@ config XEN_GRANT_DMA_ALLOC
 
 config SWIOTLB_XEN
 	def_bool y
+	depends on ARCH_HAS_DMA_OPS
 	depends on XEN_PV || ARM || ARM64
-	select DMA_OPS
 	select SWIOTLB
 
 config XEN_PCI_STUB
@@ -348,10 +348,10 @@ config XEN_GRANT_DMA_IOMMU
 
 config XEN_GRANT_DMA_OPS
 	bool
-	select DMA_OPS
 
 config XEN_VIRTIO
 	bool "Xen virtio support"
+	depends on ARCH_HAS_DMA_OPS
 	depends on VIRTIO
 	select XEN_GRANT_DMA_OPS
 	select XEN_GRANT_DMA_IOMMU if OF
diff --git a/include/linux/device.h b/include/linux/device.h
index 34eb20f5966f..b4bde8d22697 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -707,6 +707,8 @@ struct device_physical_location {
  *		for dma allocations.  This flag is managed by the dma ops
  *		instance from ->dma_supported.
  * @dma_skip_sync: DMA sync operations can be skipped for coherent buffers.
+ * @dma_iommu: Device is using default IOMMU implementation for DMA and
+ *		doesn't rely on dma_ops structure.
  *
  * At the lowest level, every device in a Linux system is represented by an
  * instance of struct device. The device structure contains the information
@@ -748,7 +750,7 @@ struct device {
 	struct dev_pin_info	*pins;
 #endif
 	struct dev_msi_info	msi;
-#ifdef CONFIG_DMA_OPS
+#ifdef CONFIG_ARCH_HAS_DMA_OPS
 	const struct dma_map_ops *dma_ops;
 #endif
 	u64		*dma_mask;	/* dma mask (if dma'able device) */
@@ -822,6 +824,9 @@ struct device {
 #ifdef CONFIG_DMA_NEED_SYNC
 	bool			dma_skip_sync:1;
 #endif
+#ifdef CONFIG_IOMMU_DMA
+	bool			dma_iommu:1;
+#endif
 };
 
 /**
diff --git a/include/linux/dma-direct.h b/include/linux/dma-direct.h
index edbe13d00776..d7e30d4f7503 100644
--- a/include/linux/dma-direct.h
+++ b/include/linux/dma-direct.h
@@ -12,7 +12,7 @@
 #include <linux/mem_encrypt.h>
 #include <linux/swiotlb.h>
 
-extern unsigned int zone_dma_bits;
+extern u64 zone_dma_limit;
 
 /*
  * Record the mapping of CPU physical to DMA addresses for a given region.
diff --git a/include/linux/dma-map-ops.h b/include/linux/dma-map-ops.h
index 02a1c825896b..9668ddf3696e 100644
--- a/include/linux/dma-map-ops.h
+++ b/include/linux/dma-map-ops.h
@@ -13,20 +13,7 @@
 struct cma;
 struct iommu_ops;
 
-/*
- * Values for struct dma_map_ops.flags:
- *
- * DMA_F_PCI_P2PDMA_SUPPORTED: Indicates the dma_map_ops implementation can
- * handle PCI P2PDMA pages in the map_sg/unmap_sg operation.
- * DMA_F_CAN_SKIP_SYNC: DMA sync operations can be skipped if the device is
- * coherent and it's not an SWIOTLB buffer.
- */
-#define DMA_F_PCI_P2PDMA_SUPPORTED     (1 << 0)
-#define DMA_F_CAN_SKIP_SYNC            (1 << 1)
-
 struct dma_map_ops {
-	unsigned int flags;
-
 	void *(*alloc)(struct device *dev, size_t size,
 			dma_addr_t *dma_handle, gfp_t gfp,
 			unsigned long attrs);
@@ -88,7 +75,7 @@ struct dma_map_ops {
 	unsigned long (*get_merge_boundary)(struct device *dev);
 };
 
-#ifdef CONFIG_DMA_OPS
+#ifdef CONFIG_ARCH_HAS_DMA_OPS
 #include <asm/dma-mapping.h>
 
 static inline const struct dma_map_ops *get_dma_ops(struct device *dev)
@@ -103,7 +90,7 @@ static inline void set_dma_ops(struct device *dev,
 {
 	dev->dma_ops = dma_ops;
 }
-#else /* CONFIG_DMA_OPS */
+#else /* CONFIG_ARCH_HAS_DMA_OPS */
 static inline const struct dma_map_ops *get_dma_ops(struct device *dev)
 {
 	return NULL;
@@ -112,7 +99,7 @@ static inline void set_dma_ops(struct device *dev,
 			       const struct dma_map_ops *dma_ops)
 {
 }
-#endif /* CONFIG_DMA_OPS */
+#endif /* CONFIG_ARCH_HAS_DMA_OPS */
 
 #ifdef CONFIG_DMA_CMA
 extern struct cma *dma_contiguous_default_area;
diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index f693aafe221f..1524da363734 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -524,13 +524,11 @@ static inline unsigned int dma_get_max_seg_size(struct device *dev)
 	return SZ_64K;
 }
 
-static inline int dma_set_max_seg_size(struct device *dev, unsigned int size)
+static inline void dma_set_max_seg_size(struct device *dev, unsigned int size)
 {
-	if (dev->dma_parms) {
-		dev->dma_parms->max_segment_size = size;
-		return 0;
-	}
-	return -EIO;
+	if (WARN_ON_ONCE(!dev->dma_parms))
+		return;
+	dev->dma_parms->max_segment_size = size;
 }
 
 static inline unsigned long dma_get_seg_boundary(struct device *dev)
@@ -559,13 +557,11 @@ static inline unsigned long dma_get_seg_boundary_nr_pages(struct device *dev,
 	return (dma_get_seg_boundary(dev) >> page_shift) + 1;
 }
 
-static inline int dma_set_seg_boundary(struct device *dev, unsigned long mask)
+static inline void dma_set_seg_boundary(struct device *dev, unsigned long mask)
 {
-	if (dev->dma_parms) {
-		dev->dma_parms->segment_boundary_mask = mask;
-		return 0;
-	}
-	return -EIO;
+	if (WARN_ON_ONCE(!dev->dma_parms))
+		return;
+	dev->dma_parms->segment_boundary_mask = mask;
 }
 
 static inline unsigned int dma_get_min_align_mask(struct device *dev)
@@ -575,13 +571,12 @@ static inline unsigned int dma_get_min_align_mask(struct device *dev)
 	return 0;
 }
 
-static inline int dma_set_min_align_mask(struct device *dev,
+static inline void dma_set_min_align_mask(struct device *dev,
 		unsigned int min_align_mask)
 {
 	if (WARN_ON_ONCE(!dev->dma_parms))
-		return -EIO;
+		return;
 	dev->dma_parms->min_align_mask = min_align_mask;
-	return 0;
 }
 
 #ifndef dma_get_cache_alignment
diff --git a/include/linux/iommu-dma.h b/include/linux/iommu-dma.h
new file mode 100644
index 000000000000..1bb55ca1ab79
--- /dev/null
+++ b/include/linux/iommu-dma.h
@@ -0,0 +1,155 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2024, NVIDIA CORPORATION & AFFILIATES. All rights reserved
+ *
+ * DMA operations that map physical memory through IOMMU.
+ */
+#ifndef _LINUX_IOMMU_DMA_H
+#define _LINUX_IOMMU_DMA_H
+
+#include <linux/dma-direction.h>
+
+#ifdef CONFIG_IOMMU_DMA
+static inline bool use_dma_iommu(struct device *dev)
+{
+	return dev->dma_iommu;
+}
+dma_addr_t iommu_dma_map_page(struct device *dev, struct page *page,
+		unsigned long offset, size_t size, enum dma_data_direction dir,
+		unsigned long attrs);
+void iommu_dma_unmap_page(struct device *dev, dma_addr_t dma_handle,
+		size_t size, enum dma_data_direction dir, unsigned long attrs);
+int iommu_dma_map_sg(struct device *dev, struct scatterlist *sg, int nents,
+		enum dma_data_direction dir, unsigned long attrs);
+void iommu_dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nents,
+		enum dma_data_direction dir, unsigned long attrs);
+void *iommu_dma_alloc(struct device *dev, size_t size, dma_addr_t *handle,
+		gfp_t gfp, unsigned long attrs);
+int iommu_dma_mmap(struct device *dev, struct vm_area_struct *vma,
+		void *cpu_addr, dma_addr_t dma_addr, size_t size,
+		unsigned long attrs);
+int iommu_dma_get_sgtable(struct device *dev, struct sg_table *sgt,
+		void *cpu_addr, dma_addr_t dma_addr, size_t size,
+		unsigned long attrs);
+unsigned long iommu_dma_get_merge_boundary(struct device *dev);
+size_t iommu_dma_opt_mapping_size(void);
+size_t iommu_dma_max_mapping_size(struct device *dev);
+void iommu_dma_free(struct device *dev, size_t size, void *cpu_addr,
+		dma_addr_t handle, unsigned long attrs);
+dma_addr_t iommu_dma_map_resource(struct device *dev, phys_addr_t phys,
+		size_t size, enum dma_data_direction dir, unsigned long attrs);
+void iommu_dma_unmap_resource(struct device *dev, dma_addr_t handle,
+		size_t size, enum dma_data_direction dir, unsigned long attrs);
+struct sg_table *iommu_dma_alloc_noncontiguous(struct device *dev, size_t size,
+		enum dma_data_direction dir, gfp_t gfp, unsigned long attrs);
+void iommu_dma_free_noncontiguous(struct device *dev, size_t size,
+		struct sg_table *sgt, enum dma_data_direction dir);
+void iommu_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle,
+		size_t size, enum dma_data_direction dir);
+void iommu_dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle,
+		size_t size, enum dma_data_direction dir);
+void iommu_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sgl,
+		int nelems, enum dma_data_direction dir);
+void iommu_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sgl,
+		int nelems, enum dma_data_direction dir);
+#else
+static inline bool use_dma_iommu(struct device *dev)
+{
+	return false;
+}
+static inline dma_addr_t iommu_dma_map_page(struct device *dev,
+		struct page *page, unsigned long offset, size_t size,
+		enum dma_data_direction dir, unsigned long attrs)
+{
+	return DMA_MAPPING_ERROR;
+}
+static inline void iommu_dma_unmap_page(struct device *dev,
+		dma_addr_t dma_handle, size_t size, enum dma_data_direction dir,
+		unsigned long attrs)
+{
+}
+static inline int iommu_dma_map_sg(struct device *dev, struct scatterlist *sg,
+		int nents, enum dma_data_direction dir, unsigned long attrs)
+{
+	return -EINVAL;
+}
+static inline void iommu_dma_unmap_sg(struct device *dev,
+		struct scatterlist *sg, int nents, enum dma_data_direction dir,
+		unsigned long attrs)
+{
+}
+static inline void *iommu_dma_alloc(struct device *dev, size_t size,
+		dma_addr_t *handle, gfp_t gfp, unsigned long attrs)
+{
+	return NULL;
+}
+static inline int iommu_dma_mmap(struct device *dev, struct vm_area_struct *vma,
+		void *cpu_addr, dma_addr_t dma_addr, size_t size,
+		unsigned long attrs)
+{
+	return -EINVAL;
+}
+static inline int iommu_dma_get_sgtable(struct device *dev,
+		struct sg_table *sgt, void *cpu_addr, dma_addr_t dma_addr,
+		size_t size, unsigned long attrs)
+{
+	return -EINVAL;
+}
+static inline unsigned long iommu_dma_get_merge_boundary(struct device *dev)
+{
+	return 0;
+}
+static inline size_t iommu_dma_opt_mapping_size(void)
+{
+	return 0;
+}
+static inline size_t iommu_dma_max_mapping_size(struct device *dev)
+{
+	return 0;
+}
+static inline void iommu_dma_free(struct device *dev, size_t size,
+		void *cpu_addr, dma_addr_t handle, unsigned long attrs)
+{
+}
+static inline dma_addr_t iommu_dma_map_resource(struct device *dev,
+		phys_addr_t phys, size_t size, enum dma_data_direction dir,
+		unsigned long attrs)
+{
+	return DMA_MAPPING_ERROR;
+}
+static inline void iommu_dma_unmap_resource(struct device *dev,
+		dma_addr_t handle, size_t size, enum dma_data_direction dir,
+		unsigned long attrs)
+{
+}
+static inline struct sg_table *
+iommu_dma_alloc_noncontiguous(struct device *dev, size_t size,
+		enum dma_data_direction dir, gfp_t gfp, unsigned long attrs)
+{
+	return NULL;
+}
+static inline void iommu_dma_free_noncontiguous(struct device *dev, size_t size,
+		struct sg_table *sgt, enum dma_data_direction dir)
+{
+}
+static inline void iommu_dma_sync_single_for_cpu(struct device *dev,
+		dma_addr_t dma_handle, size_t size,
+		enum dma_data_direction dir)
+{
+}
+static inline void iommu_dma_sync_single_for_device(struct device *dev,
+		dma_addr_t dma_handle, size_t size, enum dma_data_direction dir)
+{
+}
+static inline void iommu_dma_sync_sg_for_cpu(struct device *dev,
+		struct scatterlist *sgl, int nelems,
+		enum dma_data_direction dir)
+{
+}
+static inline void iommu_dma_sync_sg_for_device(struct device *dev,
+		struct scatterlist *sgl, int nelems,
+		enum dma_data_direction dir)
+{
+}
+#endif /* CONFIG_IOMMU_DMA */
+#endif /* _LINUX_IOMMU_DMA_H */
diff --git a/include/trace/events/dma.h b/include/trace/events/dma.h
new file mode 100644
index 000000000000..f57f05331d73
--- /dev/null
+++ b/include/trace/events/dma.h
@@ -0,0 +1,341 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM dma
+
+#if !defined(_TRACE_DMA_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_DMA_H
+
+#include <linux/tracepoint.h>
+#include <linux/dma-direction.h>
+#include <linux/dma-mapping.h>
+#include <trace/events/mmflags.h>
+
+TRACE_DEFINE_ENUM(DMA_BIDIRECTIONAL);
+TRACE_DEFINE_ENUM(DMA_TO_DEVICE);
+TRACE_DEFINE_ENUM(DMA_FROM_DEVICE);
+TRACE_DEFINE_ENUM(DMA_NONE);
+
+#define decode_dma_data_direction(dir) \
+	__print_symbolic(dir, \
+		{ DMA_BIDIRECTIONAL, "BIDIRECTIONAL" }, \
+		{ DMA_TO_DEVICE, "TO_DEVICE" }, \
+		{ DMA_FROM_DEVICE, "FROM_DEVICE" }, \
+		{ DMA_NONE, "NONE" })
+
+#define decode_dma_attrs(attrs) \
+	__print_flags(attrs, "|", \
+		{ DMA_ATTR_WEAK_ORDERING, "WEAK_ORDERING" }, \
+		{ DMA_ATTR_WRITE_COMBINE, "WRITE_COMBINE" }, \
+		{ DMA_ATTR_NO_KERNEL_MAPPING, "NO_KERNEL_MAPPING" }, \
+		{ DMA_ATTR_SKIP_CPU_SYNC, "SKIP_CPU_SYNC" }, \
+		{ DMA_ATTR_FORCE_CONTIGUOUS, "FORCE_CONTIGUOUS" }, \
+		{ DMA_ATTR_ALLOC_SINGLE_PAGES, "ALLOC_SINGLE_PAGES" }, \
+		{ DMA_ATTR_NO_WARN, "NO_WARN" }, \
+		{ DMA_ATTR_PRIVILEGED, "PRIVILEGED" })
+
+DECLARE_EVENT_CLASS(dma_map,
+	TP_PROTO(struct device *dev, phys_addr_t phys_addr, dma_addr_t dma_addr,
+		 size_t size, enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, phys_addr, dma_addr, size, dir, attrs),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__field(u64, phys_addr)
+		__field(u64, dma_addr)
+		__field(size_t, size)
+		__field(enum dma_data_direction, dir)
+		__field(unsigned long, attrs)
+	),
+
+	TP_fast_assign(
+		__assign_str(device);
+		__entry->phys_addr = phys_addr;
+		__entry->dma_addr = dma_addr;
+		__entry->size = size;
+		__entry->dir = dir;
+		__entry->attrs = attrs;
+	),
+
+	TP_printk("%s dir=%s dma_addr=%llx size=%zu phys_addr=%llx attrs=%s",
+		__get_str(device),
+		decode_dma_data_direction(__entry->dir),
+		__entry->dma_addr,
+		__entry->size,
+		__entry->phys_addr,
+		decode_dma_attrs(__entry->attrs))
+);
+
+DEFINE_EVENT(dma_map, dma_map_page,
+	TP_PROTO(struct device *dev, phys_addr_t phys_addr, dma_addr_t dma_addr,
+		 size_t size, enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, phys_addr, dma_addr, size, dir, attrs));
+
+DEFINE_EVENT(dma_map, dma_map_resource,
+	TP_PROTO(struct device *dev, phys_addr_t phys_addr, dma_addr_t dma_addr,
+		 size_t size, enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, phys_addr, dma_addr, size, dir, attrs));
+
+DECLARE_EVENT_CLASS(dma_unmap,
+	TP_PROTO(struct device *dev, dma_addr_t addr, size_t size,
+		 enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, addr, size, dir, attrs),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__field(u64, addr)
+		__field(size_t, size)
+		__field(enum dma_data_direction, dir)
+		__field(unsigned long, attrs)
+	),
+
+	TP_fast_assign(
+		__assign_str(device);
+		__entry->addr = addr;
+		__entry->size = size;
+		__entry->dir = dir;
+		__entry->attrs = attrs;
+	),
+
+	TP_printk("%s dir=%s dma_addr=%llx size=%zu attrs=%s",
+		__get_str(device),
+		decode_dma_data_direction(__entry->dir),
+		__entry->addr,
+		__entry->size,
+		decode_dma_attrs(__entry->attrs))
+);
+
+DEFINE_EVENT(dma_unmap, dma_unmap_page,
+	TP_PROTO(struct device *dev, dma_addr_t addr, size_t size,
+		 enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, addr, size, dir, attrs));
+
+DEFINE_EVENT(dma_unmap, dma_unmap_resource,
+	TP_PROTO(struct device *dev, dma_addr_t addr, size_t size,
+		 enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, addr, size, dir, attrs));
+
+TRACE_EVENT(dma_alloc,
+	TP_PROTO(struct device *dev, void *virt_addr, dma_addr_t dma_addr,
+		 size_t size, gfp_t flags, unsigned long attrs),
+	TP_ARGS(dev, virt_addr, dma_addr, size, flags, attrs),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__field(u64, phys_addr)
+		__field(u64, dma_addr)
+		__field(size_t, size)
+		__field(gfp_t, flags)
+		__field(unsigned long, attrs)
+	),
+
+	TP_fast_assign(
+		__assign_str(device);
+		__entry->phys_addr = virt_to_phys(virt_addr);
+		__entry->dma_addr = dma_addr;
+		__entry->size = size;
+		__entry->flags = flags;
+		__entry->attrs = attrs;
+	),
+
+	TP_printk("%s dma_addr=%llx size=%zu phys_addr=%llx flags=%s attrs=%s",
+		__get_str(device),
+		__entry->dma_addr,
+		__entry->size,
+		__entry->phys_addr,
+		show_gfp_flags(__entry->flags),
+		decode_dma_attrs(__entry->attrs))
+);
+
+TRACE_EVENT(dma_free,
+	TP_PROTO(struct device *dev, void *virt_addr, dma_addr_t dma_addr,
+		 size_t size, unsigned long attrs),
+	TP_ARGS(dev, virt_addr, dma_addr, size, attrs),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__field(u64, phys_addr)
+		__field(u64, dma_addr)
+		__field(size_t, size)
+		__field(unsigned long, attrs)
+	),
+
+	TP_fast_assign(
+		__assign_str(device);
+		__entry->phys_addr = virt_to_phys(virt_addr);
+		__entry->dma_addr = dma_addr;
+		__entry->size = size;
+		__entry->attrs = attrs;
+	),
+
+	TP_printk("%s dma_addr=%llx size=%zu phys_addr=%llx attrs=%s",
+		__get_str(device),
+		__entry->dma_addr,
+		__entry->size,
+		__entry->phys_addr,
+		decode_dma_attrs(__entry->attrs))
+);
+
+TRACE_EVENT(dma_map_sg,
+	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+		 int ents, enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, sg, nents, ents, dir, attrs),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__dynamic_array(u64, phys_addrs, nents)
+		__dynamic_array(u64, dma_addrs, ents)
+		__dynamic_array(unsigned int, lengths, ents)
+		__field(enum dma_data_direction, dir)
+		__field(unsigned long, attrs)
+	),
+
+	TP_fast_assign(
+		int i;
+
+		__assign_str(device);
+		for (i = 0; i < nents; i++)
+			((u64 *)__get_dynamic_array(phys_addrs))[i] =
+				sg_phys(sg + i);
+		for (i = 0; i < ents; i++) {
+			((u64 *)__get_dynamic_array(dma_addrs))[i] =
+				sg_dma_address(sg + i);
+			((unsigned int *)__get_dynamic_array(lengths))[i] =
+				sg_dma_len(sg + i);
+		}
+		__entry->dir = dir;
+		__entry->attrs = attrs;
+	),
+
+	TP_printk("%s dir=%s dma_addrs=%s sizes=%s phys_addrs=%s attrs=%s",
+		__get_str(device),
+		decode_dma_data_direction(__entry->dir),
+		__print_array(__get_dynamic_array(dma_addrs),
+			      __get_dynamic_array_len(dma_addrs) /
+				sizeof(u64), sizeof(u64)),
+		__print_array(__get_dynamic_array(lengths),
+			      __get_dynamic_array_len(lengths) /
+				sizeof(unsigned int), sizeof(unsigned int)),
+		__print_array(__get_dynamic_array(phys_addrs),
+			      __get_dynamic_array_len(phys_addrs) /
+				sizeof(u64), sizeof(u64)),
+		decode_dma_attrs(__entry->attrs))
+);
+
+TRACE_EVENT(dma_unmap_sg,
+	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+		 enum dma_data_direction dir, unsigned long attrs),
+	TP_ARGS(dev, sg, nents, dir, attrs),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__dynamic_array(u64, addrs, nents)
+		__field(enum dma_data_direction, dir)
+		__field(unsigned long, attrs)
+	),
+
+	TP_fast_assign(
+		int i;
+
+		__assign_str(device);
+		for (i = 0; i < nents; i++)
+			((u64 *)__get_dynamic_array(addrs))[i] =
+				sg_phys(sg + i);
+		__entry->dir = dir;
+		__entry->attrs = attrs;
+	),
+
+	TP_printk("%s dir=%s phys_addrs=%s attrs=%s",
+		__get_str(device),
+		decode_dma_data_direction(__entry->dir),
+		__print_array(__get_dynamic_array(addrs),
+			      __get_dynamic_array_len(addrs) /
+				sizeof(u64), sizeof(u64)),
+		decode_dma_attrs(__entry->attrs))
+);
+
+DECLARE_EVENT_CLASS(dma_sync_single,
+	TP_PROTO(struct device *dev, dma_addr_t dma_addr, size_t size,
+		 enum dma_data_direction dir),
+	TP_ARGS(dev, dma_addr, size, dir),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__field(u64, dma_addr)
+		__field(size_t, size)
+		__field(enum dma_data_direction, dir)
+	),
+
+	TP_fast_assign(
+		__assign_str(device);
+		__entry->dma_addr = dma_addr;
+		__entry->size = size;
+		__entry->dir = dir;
+	),
+
+	TP_printk("%s dir=%s dma_addr=%llx size=%zu",
+		__get_str(device),
+		decode_dma_data_direction(__entry->dir),
+		__entry->dma_addr,
+		__entry->size)
+);
+
+DEFINE_EVENT(dma_sync_single, dma_sync_single_for_cpu,
+	TP_PROTO(struct device *dev, dma_addr_t dma_addr, size_t size,
+		 enum dma_data_direction dir),
+	TP_ARGS(dev, dma_addr, size, dir));
+
+DEFINE_EVENT(dma_sync_single, dma_sync_single_for_device,
+	TP_PROTO(struct device *dev, dma_addr_t dma_addr, size_t size,
+		 enum dma_data_direction dir),
+	TP_ARGS(dev, dma_addr, size, dir));
+
+DECLARE_EVENT_CLASS(dma_sync_sg,
+	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+		 enum dma_data_direction dir),
+	TP_ARGS(dev, sg, nents, dir),
+
+	TP_STRUCT__entry(
+		__string(device, dev_name(dev))
+		__dynamic_array(u64, dma_addrs, nents)
+		__dynamic_array(unsigned int, lengths, nents)
+		__field(enum dma_data_direction, dir)
+	),
+
+	TP_fast_assign(
+		int i;
+
+		__assign_str(device);
+		for (i = 0; i < nents; i++) {
+			((u64 *)__get_dynamic_array(dma_addrs))[i] =
+				sg_dma_address(sg + i);
+			((unsigned int *)__get_dynamic_array(lengths))[i] =
+				sg_dma_len(sg + i);
+		}
+		__entry->dir = dir;
+	),
+
+	TP_printk("%s dir=%s dma_addrs=%s sizes=%s",
+		__get_str(device),
+		decode_dma_data_direction(__entry->dir),
+		__print_array(__get_dynamic_array(dma_addrs),
+			      __get_dynamic_array_len(dma_addrs) /
+				sizeof(u64), sizeof(u64)),
+		__print_array(__get_dynamic_array(lengths),
+			      __get_dynamic_array_len(lengths) /
+				sizeof(unsigned int), sizeof(unsigned int)))
+);
+
+DEFINE_EVENT(dma_sync_sg, dma_sync_sg_for_cpu,
+	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+		 enum dma_data_direction dir),
+	TP_ARGS(dev, sg, nents, dir));
+
+DEFINE_EVENT(dma_sync_sg, dma_sync_sg_for_device,
+	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+		 enum dma_data_direction dir),
+	TP_ARGS(dev, sg, nents, dir));
+
+#endif /*  _TRACE_DMA_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>
diff --git a/kernel/dma/Kconfig b/kernel/dma/Kconfig
index c06e56be0ca1..4c0dcd909121 100644
--- a/kernel/dma/Kconfig
+++ b/kernel/dma/Kconfig
@@ -8,8 +8,7 @@ config HAS_DMA
 	depends on !NO_DMA
 	default y
 
-config DMA_OPS
-	depends on HAS_DMA
+config DMA_OPS_HELPERS
 	bool
 
 #
@@ -109,8 +108,8 @@ config DMA_BOUNCE_UNALIGNED_KMALLOC
 
 config DMA_NEED_SYNC
 	def_bool ARCH_HAS_SYNC_DMA_FOR_DEVICE || ARCH_HAS_SYNC_DMA_FOR_CPU || \
-		 ARCH_HAS_SYNC_DMA_FOR_CPU_ALL || DMA_API_DEBUG || DMA_OPS || \
-		 SWIOTLB
+		 ARCH_HAS_SYNC_DMA_FOR_CPU_ALL || DMA_API_DEBUG || \
+		 ARCH_HAS_DMA_OPS || SWIOTLB
 
 config DMA_RESTRICTED_POOL
 	bool "DMA Restricted Pool"
diff --git a/kernel/dma/Makefile b/kernel/dma/Makefile
index 21926e46ef4f..6977033444a3 100644
--- a/kernel/dma/Makefile
+++ b/kernel/dma/Makefile
@@ -1,8 +1,8 @@
 # SPDX-License-Identifier: GPL-2.0
 
 obj-$(CONFIG_HAS_DMA)			+= mapping.o direct.o
-obj-$(CONFIG_DMA_OPS)			+= ops_helpers.o
-obj-$(CONFIG_DMA_OPS)			+= dummy.o
+obj-$(CONFIG_DMA_OPS_HELPERS)		+= ops_helpers.o
+obj-$(CONFIG_ARCH_HAS_DMA_OPS)		+= dummy.o
 obj-$(CONFIG_DMA_CMA)			+= contiguous.o
 obj-$(CONFIG_DMA_DECLARE_COHERENT)	+= coherent.o
 obj-$(CONFIG_DMA_API_DEBUG)		+= debug.o
diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c
index 4480a3cd92e0..5b4e6d3bf7bc 100644
--- a/kernel/dma/direct.c
+++ b/kernel/dma/direct.c
@@ -20,7 +20,7 @@
  * it for entirely different regions. In that case the arch code needs to
  * override the variable below for dma-direct to work properly.
  */
-unsigned int zone_dma_bits __ro_after_init = 24;
+u64 zone_dma_limit __ro_after_init = DMA_BIT_MASK(24);
 
 static inline dma_addr_t phys_to_dma_direct(struct device *dev,
 		phys_addr_t phys)
@@ -59,7 +59,7 @@ static gfp_t dma_direct_optimal_gfp_mask(struct device *dev, u64 *phys_limit)
 	 * zones.
 	 */
 	*phys_limit = dma_to_phys(dev, dma_limit);
-	if (*phys_limit <= DMA_BIT_MASK(zone_dma_bits))
+	if (*phys_limit <= zone_dma_limit)
 		return GFP_DMA;
 	if (*phys_limit <= DMA_BIT_MASK(32))
 		return GFP_DMA32;
@@ -140,7 +140,7 @@ again:
 	if (!page)
 		page = alloc_pages_node(node, gfp, get_order(size));
 	if (page && !dma_coherent_ok(dev, page_to_phys(page), size)) {
-		dma_free_contiguous(dev, page, size);
+		__free_pages(page, get_order(size));
 		page = NULL;
 
 		if (IS_ENABLED(CONFIG_ZONE_DMA32) &&
@@ -580,7 +580,7 @@ int dma_direct_supported(struct device *dev, u64 mask)
 	 * part of the check.
 	 */
 	if (IS_ENABLED(CONFIG_ZONE_DMA))
-		min_mask = min_t(u64, min_mask, DMA_BIT_MASK(zone_dma_bits));
+		min_mask = min_t(u64, min_mask, zone_dma_limit);
 	return mask >= phys_to_dma_unencrypted(dev, min_mask);
 }
 
diff --git a/kernel/dma/dummy.c b/kernel/dma/dummy.c
index b492d59ac77e..92de80e5b057 100644
--- a/kernel/dma/dummy.c
+++ b/kernel/dma/dummy.c
@@ -17,6 +17,15 @@ static dma_addr_t dma_dummy_map_page(struct device *dev, struct page *page,
 {
 	return DMA_MAPPING_ERROR;
 }
+static void dma_dummy_unmap_page(struct device *dev, dma_addr_t dma_handle,
+		size_t size, enum dma_data_direction dir, unsigned long attrs)
+{
+	/*
+	 * Dummy ops doesn't support map_page, so unmap_page should never be
+	 * called.
+	 */
+	WARN_ON_ONCE(true);
+}
 
 static int dma_dummy_map_sg(struct device *dev, struct scatterlist *sgl,
 		int nelems, enum dma_data_direction dir,
@@ -25,6 +34,16 @@ static int dma_dummy_map_sg(struct device *dev, struct scatterlist *sgl,
 	return -EINVAL;
 }
 
+static void dma_dummy_unmap_sg(struct device *dev, struct scatterlist *sgl,
+		int nelems, enum dma_data_direction dir,
+		unsigned long attrs)
+{
+	/*
+	 * Dummy ops doesn't support map_sg, so unmap_sg should never be called.
+	 */
+	WARN_ON_ONCE(true);
+}
+
 static int dma_dummy_supported(struct device *hwdev, u64 mask)
 {
 	return 0;
@@ -33,6 +52,8 @@ static int dma_dummy_supported(struct device *hwdev, u64 mask)
 const struct dma_map_ops dma_dummy_ops = {
 	.mmap                   = dma_dummy_mmap,
 	.map_page               = dma_dummy_map_page,
+	.unmap_page             = dma_dummy_unmap_page,
 	.map_sg                 = dma_dummy_map_sg,
+	.unmap_sg               = dma_dummy_unmap_sg,
 	.dma_supported          = dma_dummy_supported,
 };
diff --git a/kernel/dma/mapping.c b/kernel/dma/mapping.c
index b1c18058d55f..b839683da0ba 100644
--- a/kernel/dma/mapping.c
+++ b/kernel/dma/mapping.c
@@ -10,6 +10,7 @@
 #include <linux/dma-map-ops.h>
 #include <linux/export.h>
 #include <linux/gfp.h>
+#include <linux/iommu-dma.h>
 #include <linux/kmsan.h>
 #include <linux/of_device.h>
 #include <linux/slab.h>
@@ -17,6 +18,9 @@
 #include "debug.h"
 #include "direct.h"
 
+#define CREATE_TRACE_POINTS
+#include <trace/events/dma.h>
+
 #if defined(CONFIG_ARCH_HAS_SYNC_DMA_FOR_DEVICE) || \
 	defined(CONFIG_ARCH_HAS_SYNC_DMA_FOR_CPU) || \
 	defined(CONFIG_ARCH_HAS_SYNC_DMA_FOR_CPU_ALL)
@@ -116,8 +120,12 @@ EXPORT_SYMBOL(dmam_alloc_attrs);
 static bool dma_go_direct(struct device *dev, dma_addr_t mask,
 		const struct dma_map_ops *ops)
 {
+	if (use_dma_iommu(dev))
+		return false;
+
 	if (likely(!ops))
 		return true;
+
 #ifdef CONFIG_DMA_OPS_BYPASS
 	if (dev->dma_ops_bypass)
 		return min_not_zero(mask, dev->bus_dma_limit) >=
@@ -159,9 +167,13 @@ dma_addr_t dma_map_page_attrs(struct device *dev, struct page *page,
 	if (dma_map_direct(dev, ops) ||
 	    arch_dma_map_page_direct(dev, page_to_phys(page) + offset + size))
 		addr = dma_direct_map_page(dev, page, offset, size, dir, attrs);
+	else if (use_dma_iommu(dev))
+		addr = iommu_dma_map_page(dev, page, offset, size, dir, attrs);
 	else
 		addr = ops->map_page(dev, page, offset, size, dir, attrs);
 	kmsan_handle_dma(page, offset, size, dir);
+	trace_dma_map_page(dev, page_to_phys(page) + offset, addr, size, dir,
+			   attrs);
 	debug_dma_map_page(dev, page, offset, size, dir, addr, attrs);
 
 	return addr;
@@ -177,8 +189,11 @@ void dma_unmap_page_attrs(struct device *dev, dma_addr_t addr, size_t size,
 	if (dma_map_direct(dev, ops) ||
 	    arch_dma_unmap_page_direct(dev, addr + size))
 		dma_direct_unmap_page(dev, addr, size, dir, attrs);
-	else if (ops->unmap_page)
+	else if (use_dma_iommu(dev))
+		iommu_dma_unmap_page(dev, addr, size, dir, attrs);
+	else
 		ops->unmap_page(dev, addr, size, dir, attrs);
+	trace_dma_unmap_page(dev, addr, size, dir, attrs);
 	debug_dma_unmap_page(dev, addr, size, dir);
 }
 EXPORT_SYMBOL(dma_unmap_page_attrs);
@@ -197,11 +212,14 @@ static int __dma_map_sg_attrs(struct device *dev, struct scatterlist *sg,
 	if (dma_map_direct(dev, ops) ||
 	    arch_dma_map_sg_direct(dev, sg, nents))
 		ents = dma_direct_map_sg(dev, sg, nents, dir, attrs);
+	else if (use_dma_iommu(dev))
+		ents = iommu_dma_map_sg(dev, sg, nents, dir, attrs);
 	else
 		ents = ops->map_sg(dev, sg, nents, dir, attrs);
 
 	if (ents > 0) {
 		kmsan_handle_dma_sg(sg, nents, dir);
+		trace_dma_map_sg(dev, sg, nents, ents, dir, attrs);
 		debug_dma_map_sg(dev, sg, nents, ents, dir, attrs);
 	} else if (WARN_ON_ONCE(ents != -EINVAL && ents != -ENOMEM &&
 				ents != -EIO && ents != -EREMOTEIO)) {
@@ -287,10 +305,13 @@ void dma_unmap_sg_attrs(struct device *dev, struct scatterlist *sg,
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
 	BUG_ON(!valid_dma_direction(dir));
+	trace_dma_unmap_sg(dev, sg, nents, dir, attrs);
 	debug_dma_unmap_sg(dev, sg, nents, dir);
 	if (dma_map_direct(dev, ops) ||
 	    arch_dma_unmap_sg_direct(dev, sg, nents))
 		dma_direct_unmap_sg(dev, sg, nents, dir, attrs);
+	else if (use_dma_iommu(dev))
+		iommu_dma_unmap_sg(dev, sg, nents, dir, attrs);
 	else if (ops->unmap_sg)
 		ops->unmap_sg(dev, sg, nents, dir, attrs);
 }
@@ -309,9 +330,12 @@ dma_addr_t dma_map_resource(struct device *dev, phys_addr_t phys_addr,
 
 	if (dma_map_direct(dev, ops))
 		addr = dma_direct_map_resource(dev, phys_addr, size, dir, attrs);
+	else if (use_dma_iommu(dev))
+		addr = iommu_dma_map_resource(dev, phys_addr, size, dir, attrs);
 	else if (ops->map_resource)
 		addr = ops->map_resource(dev, phys_addr, size, dir, attrs);
 
+	trace_dma_map_resource(dev, phys_addr, addr, size, dir, attrs);
 	debug_dma_map_resource(dev, phys_addr, size, dir, addr, attrs);
 	return addr;
 }
@@ -323,8 +347,13 @@ void dma_unmap_resource(struct device *dev, dma_addr_t addr, size_t size,
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
 	BUG_ON(!valid_dma_direction(dir));
-	if (!dma_map_direct(dev, ops) && ops->unmap_resource)
+	if (dma_map_direct(dev, ops))
+		; /* nothing to do: uncached and no swiotlb */
+	else if (use_dma_iommu(dev))
+		iommu_dma_unmap_resource(dev, addr, size, dir, attrs);
+	else if (ops->unmap_resource)
 		ops->unmap_resource(dev, addr, size, dir, attrs);
+	trace_dma_unmap_resource(dev, addr, size, dir, attrs);
 	debug_dma_unmap_resource(dev, addr, size, dir);
 }
 EXPORT_SYMBOL(dma_unmap_resource);
@@ -338,8 +367,11 @@ void __dma_sync_single_for_cpu(struct device *dev, dma_addr_t addr, size_t size,
 	BUG_ON(!valid_dma_direction(dir));
 	if (dma_map_direct(dev, ops))
 		dma_direct_sync_single_for_cpu(dev, addr, size, dir);
+	else if (use_dma_iommu(dev))
+		iommu_dma_sync_single_for_cpu(dev, addr, size, dir);
 	else if (ops->sync_single_for_cpu)
 		ops->sync_single_for_cpu(dev, addr, size, dir);
+	trace_dma_sync_single_for_cpu(dev, addr, size, dir);
 	debug_dma_sync_single_for_cpu(dev, addr, size, dir);
 }
 EXPORT_SYMBOL(__dma_sync_single_for_cpu);
@@ -352,8 +384,11 @@ void __dma_sync_single_for_device(struct device *dev, dma_addr_t addr,
 	BUG_ON(!valid_dma_direction(dir));
 	if (dma_map_direct(dev, ops))
 		dma_direct_sync_single_for_device(dev, addr, size, dir);
+	else if (use_dma_iommu(dev))
+		iommu_dma_sync_single_for_device(dev, addr, size, dir);
 	else if (ops->sync_single_for_device)
 		ops->sync_single_for_device(dev, addr, size, dir);
+	trace_dma_sync_single_for_device(dev, addr, size, dir);
 	debug_dma_sync_single_for_device(dev, addr, size, dir);
 }
 EXPORT_SYMBOL(__dma_sync_single_for_device);
@@ -366,8 +401,11 @@ void __dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg,
 	BUG_ON(!valid_dma_direction(dir));
 	if (dma_map_direct(dev, ops))
 		dma_direct_sync_sg_for_cpu(dev, sg, nelems, dir);
+	else if (use_dma_iommu(dev))
+		iommu_dma_sync_sg_for_cpu(dev, sg, nelems, dir);
 	else if (ops->sync_sg_for_cpu)
 		ops->sync_sg_for_cpu(dev, sg, nelems, dir);
+	trace_dma_sync_sg_for_cpu(dev, sg, nelems, dir);
 	debug_dma_sync_sg_for_cpu(dev, sg, nelems, dir);
 }
 EXPORT_SYMBOL(__dma_sync_sg_for_cpu);
@@ -380,8 +418,11 @@ void __dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg,
 	BUG_ON(!valid_dma_direction(dir));
 	if (dma_map_direct(dev, ops))
 		dma_direct_sync_sg_for_device(dev, sg, nelems, dir);
+	else if (use_dma_iommu(dev))
+		iommu_dma_sync_sg_for_device(dev, sg, nelems, dir);
 	else if (ops->sync_sg_for_device)
 		ops->sync_sg_for_device(dev, sg, nelems, dir);
+	trace_dma_sync_sg_for_device(dev, sg, nelems, dir);
 	debug_dma_sync_sg_for_device(dev, sg, nelems, dir);
 }
 EXPORT_SYMBOL(__dma_sync_sg_for_device);
@@ -405,7 +446,7 @@ static void dma_setup_need_sync(struct device *dev)
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
-	if (dma_map_direct(dev, ops) || (ops->flags & DMA_F_CAN_SKIP_SYNC))
+	if (dma_map_direct(dev, ops) || use_dma_iommu(dev))
 		/*
 		 * dma_skip_sync will be reset to %false on first SWIOTLB buffer
 		 * mapping, if any. During the device initialization, it's
@@ -446,6 +487,9 @@ int dma_get_sgtable_attrs(struct device *dev, struct sg_table *sgt,
 	if (dma_alloc_direct(dev, ops))
 		return dma_direct_get_sgtable(dev, sgt, cpu_addr, dma_addr,
 				size, attrs);
+	if (use_dma_iommu(dev))
+		return iommu_dma_get_sgtable(dev, sgt, cpu_addr, dma_addr,
+				size, attrs);
 	if (!ops->get_sgtable)
 		return -ENXIO;
 	return ops->get_sgtable(dev, sgt, cpu_addr, dma_addr, size, attrs);
@@ -482,6 +526,8 @@ bool dma_can_mmap(struct device *dev)
 
 	if (dma_alloc_direct(dev, ops))
 		return dma_direct_can_mmap(dev);
+	if (use_dma_iommu(dev))
+		return true;
 	return ops->mmap != NULL;
 }
 EXPORT_SYMBOL_GPL(dma_can_mmap);
@@ -508,6 +554,9 @@ int dma_mmap_attrs(struct device *dev, struct vm_area_struct *vma,
 	if (dma_alloc_direct(dev, ops))
 		return dma_direct_mmap(dev, vma, cpu_addr, dma_addr, size,
 				attrs);
+	if (use_dma_iommu(dev))
+		return iommu_dma_mmap(dev, vma, cpu_addr, dma_addr, size,
+				      attrs);
 	if (!ops->mmap)
 		return -ENXIO;
 	return ops->mmap(dev, vma, cpu_addr, dma_addr, size, attrs);
@@ -559,11 +608,14 @@ void *dma_alloc_attrs(struct device *dev, size_t size, dma_addr_t *dma_handle,
 
 	if (dma_alloc_direct(dev, ops))
 		cpu_addr = dma_direct_alloc(dev, size, dma_handle, flag, attrs);
+	else if (use_dma_iommu(dev))
+		cpu_addr = iommu_dma_alloc(dev, size, dma_handle, flag, attrs);
 	else if (ops->alloc)
 		cpu_addr = ops->alloc(dev, size, dma_handle, flag, attrs);
 	else
 		return NULL;
 
+	trace_dma_alloc(dev, cpu_addr, *dma_handle, size, flag, attrs);
 	debug_dma_alloc_coherent(dev, size, *dma_handle, cpu_addr, attrs);
 	return cpu_addr;
 }
@@ -588,9 +640,12 @@ void dma_free_attrs(struct device *dev, size_t size, void *cpu_addr,
 	if (!cpu_addr)
 		return;
 
+	trace_dma_free(dev, cpu_addr, dma_handle, size, attrs);
 	debug_dma_free_coherent(dev, size, cpu_addr, dma_handle);
 	if (dma_alloc_direct(dev, ops))
 		dma_direct_free(dev, size, cpu_addr, dma_handle, attrs);
+	else if (use_dma_iommu(dev))
+		iommu_dma_free(dev, size, cpu_addr, dma_handle, attrs);
 	else if (ops->free)
 		ops->free(dev, size, cpu_addr, dma_handle, attrs);
 }
@@ -611,6 +666,8 @@ static struct page *__dma_alloc_pages(struct device *dev, size_t size,
 	size = PAGE_ALIGN(size);
 	if (dma_alloc_direct(dev, ops))
 		return dma_direct_alloc_pages(dev, size, dma_handle, dir, gfp);
+	if (use_dma_iommu(dev))
+		return dma_common_alloc_pages(dev, size, dma_handle, dir, gfp);
 	if (!ops->alloc_pages_op)
 		return NULL;
 	return ops->alloc_pages_op(dev, size, dma_handle, dir, gfp);
@@ -621,8 +678,11 @@ struct page *dma_alloc_pages(struct device *dev, size_t size,
 {
 	struct page *page = __dma_alloc_pages(dev, size, dma_handle, dir, gfp);
 
-	if (page)
+	if (page) {
+		trace_dma_map_page(dev, page_to_phys(page), *dma_handle, size,
+				   dir, 0);
 		debug_dma_map_page(dev, page, 0, size, dir, *dma_handle, 0);
+	}
 	return page;
 }
 EXPORT_SYMBOL_GPL(dma_alloc_pages);
@@ -635,6 +695,8 @@ static void __dma_free_pages(struct device *dev, size_t size, struct page *page,
 	size = PAGE_ALIGN(size);
 	if (dma_alloc_direct(dev, ops))
 		dma_direct_free_pages(dev, size, page, dma_handle, dir);
+	else if (use_dma_iommu(dev))
+		dma_common_free_pages(dev, size, page, dma_handle, dir);
 	else if (ops->free_pages)
 		ops->free_pages(dev, size, page, dma_handle, dir);
 }
@@ -642,6 +704,7 @@ static void __dma_free_pages(struct device *dev, size_t size, struct page *page,
 void dma_free_pages(struct device *dev, size_t size, struct page *page,
 		dma_addr_t dma_handle, enum dma_data_direction dir)
 {
+	trace_dma_unmap_page(dev, dma_handle, size, dir, 0);
 	debug_dma_unmap_page(dev, dma_handle, size, dir);
 	__dma_free_pages(dev, size, page, dma_handle, dir);
 }
@@ -697,11 +760,14 @@ struct sg_table *dma_alloc_noncontiguous(struct device *dev, size_t size,
 
 	if (ops && ops->alloc_noncontiguous)
 		sgt = ops->alloc_noncontiguous(dev, size, dir, gfp, attrs);
+	else if (use_dma_iommu(dev))
+		sgt = iommu_dma_alloc_noncontiguous(dev, size, dir, gfp, attrs);
 	else
 		sgt = alloc_single_sgt(dev, size, dir, gfp);
 
 	if (sgt) {
 		sgt->nents = 1;
+		trace_dma_map_sg(dev, sgt->sgl, sgt->orig_nents, 1, dir, attrs);
 		debug_dma_map_sg(dev, sgt->sgl, sgt->orig_nents, 1, dir, attrs);
 	}
 	return sgt;
@@ -722,9 +788,12 @@ void dma_free_noncontiguous(struct device *dev, size_t size,
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
+	trace_dma_unmap_sg(dev, sgt->sgl, sgt->orig_nents, dir, 0);
 	debug_dma_unmap_sg(dev, sgt->sgl, sgt->orig_nents, dir);
 	if (ops && ops->free_noncontiguous)
 		ops->free_noncontiguous(dev, size, sgt, dir);
+	else if (use_dma_iommu(dev))
+		iommu_dma_free_noncontiguous(dev, size, sgt, dir);
 	else
 		free_single_sgt(dev, size, sgt, dir);
 }
@@ -772,32 +841,37 @@ static int dma_supported(struct device *dev, u64 mask)
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
+	if (use_dma_iommu(dev)) {
+		if (WARN_ON(ops))
+			return false;
+		return true;
+	}
+
 	/*
-	 * ->dma_supported sets the bypass flag, so we must always call
-	 * into the method here unless the device is truly direct mapped.
+	 * ->dma_supported sets and clears the bypass flag, so ignore it here
+	 * and always call into the method if there is one.
 	 */
-	if (!ops)
-		return dma_direct_supported(dev, mask);
-	if (!ops->dma_supported)
-		return 1;
-	return ops->dma_supported(dev, mask);
+	if (ops) {
+		if (!ops->dma_supported)
+			return true;
+		return ops->dma_supported(dev, mask);
+	}
+
+	return dma_direct_supported(dev, mask);
 }
 
 bool dma_pci_p2pdma_supported(struct device *dev)
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
-	/* if ops is not set, dma direct will be used which supports P2PDMA */
-	if (!ops)
-		return true;
-
 	/*
 	 * Note: dma_ops_bypass is not checked here because P2PDMA should
 	 * not be used with dma mapping ops that do not have support even
 	 * if the specific device is bypassing them.
 	 */
 
-	return ops->flags & DMA_F_PCI_P2PDMA_SUPPORTED;
+	/* if ops is not set, dma direct and default IOMMU support P2PDMA */
+	return !ops;
 }
 EXPORT_SYMBOL_GPL(dma_pci_p2pdma_supported);
 
@@ -865,6 +939,8 @@ size_t dma_max_mapping_size(struct device *dev)
 
 	if (dma_map_direct(dev, ops))
 		size = dma_direct_max_mapping_size(dev);
+	else if (use_dma_iommu(dev))
+		size = iommu_dma_max_mapping_size(dev);
 	else if (ops && ops->max_mapping_size)
 		size = ops->max_mapping_size(dev);
 
@@ -877,7 +953,9 @@ size_t dma_opt_mapping_size(struct device *dev)
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 	size_t size = SIZE_MAX;
 
-	if (ops && ops->opt_mapping_size)
+	if (use_dma_iommu(dev))
+		size = iommu_dma_opt_mapping_size();
+	else if (ops && ops->opt_mapping_size)
 		size = ops->opt_mapping_size();
 
 	return min(dma_max_mapping_size(dev), size);
@@ -888,6 +966,9 @@ unsigned long dma_get_merge_boundary(struct device *dev)
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
+	if (use_dma_iommu(dev))
+		return iommu_dma_get_merge_boundary(dev);
+
 	if (!ops || !ops->get_merge_boundary)
 		return 0;	/* can't merge */
 
diff --git a/kernel/dma/ops_helpers.c b/kernel/dma/ops_helpers.c
index af4a6ef48ce0..9afd569eadb9 100644
--- a/kernel/dma/ops_helpers.c
+++ b/kernel/dma/ops_helpers.c
@@ -4,6 +4,7 @@
  * the allocated memory contains normal pages in the direct kernel mapping.
  */
 #include <linux/dma-map-ops.h>
+#include <linux/iommu-dma.h>
 
 static struct page *dma_common_vaddr_to_page(void *cpu_addr)
 {
@@ -70,8 +71,12 @@ struct page *dma_common_alloc_pages(struct device *dev, size_t size,
 	if (!page)
 		return NULL;
 
-	*dma_handle = ops->map_page(dev, page, 0, size, dir,
-				    DMA_ATTR_SKIP_CPU_SYNC);
+	if (use_dma_iommu(dev))
+		*dma_handle = iommu_dma_map_page(dev, page, 0, size, dir,
+						 DMA_ATTR_SKIP_CPU_SYNC);
+	else
+		*dma_handle = ops->map_page(dev, page, 0, size, dir,
+					    DMA_ATTR_SKIP_CPU_SYNC);
 	if (*dma_handle == DMA_MAPPING_ERROR) {
 		dma_free_contiguous(dev, page, size);
 		return NULL;
@@ -86,7 +91,10 @@ void dma_common_free_pages(struct device *dev, size_t size, struct page *page,
 {
 	const struct dma_map_ops *ops = get_dma_ops(dev);
 
-	if (ops->unmap_page)
+	if (use_dma_iommu(dev))
+		iommu_dma_unmap_page(dev, dma_handle, size, dir,
+				     DMA_ATTR_SKIP_CPU_SYNC);
+	else if (ops->unmap_page)
 		ops->unmap_page(dev, dma_handle, size, dir,
 				DMA_ATTR_SKIP_CPU_SYNC);
 	dma_free_contiguous(dev, page, size);
diff --git a/kernel/dma/pool.c b/kernel/dma/pool.c
index d10613eb0f63..7b04f7575796 100644
--- a/kernel/dma/pool.c
+++ b/kernel/dma/pool.c
@@ -70,9 +70,9 @@ static bool cma_in_zone(gfp_t gfp)
 	/* CMA can't cross zone boundaries, see cma_activate_area() */
 	end = cma_get_base(cma) + size - 1;
 	if (IS_ENABLED(CONFIG_ZONE_DMA) && (gfp & GFP_DMA))
-		return end <= DMA_BIT_MASK(zone_dma_bits);
+		return end <= zone_dma_limit;
 	if (IS_ENABLED(CONFIG_ZONE_DMA32) && (gfp & GFP_DMA32))
-		return end <= DMA_BIT_MASK(32);
+		return end <= max(DMA_BIT_MASK(32), zone_dma_limit);
 	return true;
 }
 
diff --git a/kernel/dma/remap.c b/kernel/dma/remap.c
index 27596f3b4aef..9e2afad1c615 100644
--- a/kernel/dma/remap.c
+++ b/kernel/dma/remap.c
@@ -10,8 +10,10 @@ struct page **dma_common_find_pages(void *cpu_addr)
 {
 	struct vm_struct *area = find_vm_area(cpu_addr);
 
-	if (!area || area->flags != VM_DMA_COHERENT)
+	if (!area || !(area->flags & VM_DMA_COHERENT))
 		return NULL;
+	WARN(area->flags != VM_DMA_COHERENT,
+	     "unexpected flags in area: %p\n", cpu_addr);
 	return area->pages;
 }
 
@@ -61,7 +63,7 @@ void dma_common_free_remap(void *cpu_addr, size_t size)
 {
 	struct vm_struct *area = find_vm_area(cpu_addr);
 
-	if (!area || area->flags != VM_DMA_COHERENT) {
+	if (!area || !(area->flags & VM_DMA_COHERENT)) {
 		WARN(1, "trying to free invalid coherent area: %p\n", cpu_addr);
 		return;
 	}
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index df68d29740a0..abcf3fa63a56 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -450,9 +450,9 @@ int swiotlb_init_late(size_t size, gfp_t gfp_mask,
 	if (!remap)
 		io_tlb_default_mem.can_grow = true;
 	if (IS_ENABLED(CONFIG_ZONE_DMA) && (gfp_mask & __GFP_DMA))
-		io_tlb_default_mem.phys_limit = DMA_BIT_MASK(zone_dma_bits);
+		io_tlb_default_mem.phys_limit = zone_dma_limit;
 	else if (IS_ENABLED(CONFIG_ZONE_DMA32) && (gfp_mask & __GFP_DMA32))
-		io_tlb_default_mem.phys_limit = DMA_BIT_MASK(32);
+		io_tlb_default_mem.phys_limit = max(DMA_BIT_MASK(32), zone_dma_limit);
 	else
 		io_tlb_default_mem.phys_limit = virt_to_phys(high_memory - 1);
 #endif
@@ -629,7 +629,7 @@ static struct page *swiotlb_alloc_tlb(struct device *dev, size_t bytes,
 	}
 
 	gfp &= ~GFP_ZONEMASK;
-	if (phys_limit <= DMA_BIT_MASK(zone_dma_bits))
+	if (phys_limit <= zone_dma_limit)
 		gfp |= __GFP_DMA;
 	else if (phys_limit <= DMA_BIT_MASK(32))
 		gfp |= __GFP_DMA32;