Merge tag 'dma-mapping-6.12-2024-09-19' of git://git.infradead.org/users/hch/dma-mapping

Pull dma-mapping updates from Christoph Hellwig:

 - support DMA zones for arm64 systems where memory starts at > 4GB
   (Baruch Siach, Catalin Marinas)

 - support direct calls into dma-iommu and thus obsolete dma_map_ops for
   many common configurations (Leon Romanovsky)

 - add DMA-API tracing (Sean Anderson)

 - remove the not very useful return value from various dma_set_* APIs
   (Christoph Hellwig)

 - misc cleanups and minor optimizations (Chen Y, Yosry Ahmed, Christoph
   Hellwig)

* tag 'dma-mapping-6.12-2024-09-19' of git://git.infradead.org/users/hch/dma-mapping:
  dma-mapping: reflow dma_supported
  dma-mapping: reliably inform about DMA support for IOMMU
  dma-mapping: add tracing for dma-mapping API calls
  dma-mapping: use IOMMU DMA calls for common alloc/free page calls
  dma-direct: optimize page freeing when it is not addressable
  dma-mapping: clearly mark DMA ops as an architecture feature
  vdpa_sim: don't select DMA_OPS
  arm64: mm: keep low RAM dma zone
  dma-mapping: don't return errors from dma_set_max_seg_size
  dma-mapping: don't return errors from dma_set_seg_boundary
  dma-mapping: don't return errors from dma_set_min_align_mask
  scsi: check that busses support the DMA API before setting dma parameters
  arm64: mm: fix DMA zone when dma-ranges is missing
  dma-mapping: direct calls for dma-iommu
  dma-mapping: call ->unmap_page and ->unmap_sg unconditionally
  arm64: support DMA zone above 4GB
  dma-mapping: replace zone_dma_bits by zone_dma_limit
  dma-mapping: use bit masking to check VM_DMA_COHERENT

20 files changed, 76 insertions, 125 deletions

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