Merge branch 'pci/endpoint'

- Free all previously requested IRQs in epf_ntb_db_bar_init_msi_doorbell()
  error path (Koichiro Den)

- Free doorbell IRQ in pci-epf-test only if it has actually been requested
  (Koichiro Den)

- Discard pointer to doorbell message array after freeing it in
  pci_epf_alloc_doorbell() error path (Koichiro Den)

- Advertise dynamic inbound mapping support in pci-epf-test and update host
  pci_endpoint_test to skip doorbell testing if not advertised by endpoint
  (Koichiro Den)

- Constify configfs item and group operations (Christophe JAILLET)

- Use array_index_nospec() on configfs MW show/store attributes (Koichiro
  Den)

- Return -ERANGE (not -EINVAL) for configfs out-of-range MW index (Koichiro
  Den)

- Return 0, not remaining timeout, when MHI eDMA ops complete so
  mhi_ep_ring_add_element() doesn't interpret non-zero as failure (Daniel
  Hodges)

- Remove vntb and ntb duplicate resource teardown that leads to oops when
  .allow_link() fails or .drop_link() is called (Koichiro Den)

- Disable vntb delayed work before clearing BAR mappings and doorbells to
  avoid oops caused by doing the work after resources have been torn down
  (Koichiro Den)

- Fix pci_epf_add_vepf() kernel-doc typo (Alok Tiwari)

- Propagate pci_epf_create() errors to pci_epf_make() callers (Alok Tiwari)

- Remove redundant BAR_RESERVED annotation for the high order part of a
  64-bit BAR (Niklas Cassel)

- Add a way to describe reserved subregions within BARs, e.g.,
  platform-owned fixed register windows, and use it for the RK3588 BAR4 DMA
  ctrl window (Koichiro Den)

- Add BAR_DISABLED for BARs that will never be available to an EPF driver,
  and change some BAR_RESERVED annotations to BAR_DISABLED (Niklas Cassel)

- Disable BARs in common code instead of in each glue driver (Niklas
  Cassel)

- Advertise reserved BARs in Capabilities so host-side drivers can skip
  them (Niklas Cassel)

- Skip reserved BARs in selftests (Niklas Cassel)

- Improve error messages and include device name when available (Manivannan
  Sadhasivam)

- Add NTB .get_dma_dev() callback for cases where DMA API requires a
  different device, e.g., vNTB devices (Koichiro Den)

- Return -EINVAL, not -ENOSPC, if endpoint test determines the subrange
  size is too small (Koichiro Den)

- Add reserved region types for MSI-X Table and PBA so Endpoint controllers
  can them as describe hardware-owned regions in a BAR_RESERVED BAR
  (Manikanta Maddireddy)

- Make Tegra194/234 BAR0 programmable and remove 1MB size limit (Manikanta
  Maddireddy)

- Expose Tegra BAR2 (MSI-X) and BAR4 (DMA) as 64-bit BAR_RESERVED
  (Manikanta Maddireddy)

- Add Tegra194 and Tegra234 device table entries to pci_endpoint_test
  (Manikanta Maddireddy)

- Skip the BAR subrange selftest if there are not enough inbound window
  resources to run the test (Christian Bruel)

* pci/endpoint:
  selftests: pci_endpoint: Skip BAR subrange test on -ENOSPC
  misc: pci_endpoint_test: Add Tegra194 and Tegra234 device table entries
  PCI: tegra194: Expose BAR2 (MSI-X) and BAR4 (DMA) as 64-bit BAR_RESERVED
  PCI: tegra194: Make BAR0 programmable and remove 1MB size limit
  PCI: endpoint: Add reserved region type for MSI-X Table and PBA
  misc: pci_endpoint_test: Use -EINVAL for small subrange size
  PCI: endpoint: pci-epf-vntb: Implement .get_dma_dev()
  NTB: ntb_transport: Use ntb_get_dma_dev() for DMA buffers
  NTB: core: Add .get_dma_dev() callback to ntb_dev_ops
  PCI: endpoint: Improve error messages
  PCI: endpoint: Print the EPF name in the error log of pci_epf_make()
  selftests: pci_endpoint: Skip reserved BARs
  misc: pci_endpoint_test: Give reserved BARs a distinct error code
  PCI: endpoint: pci-epf-test: Advertise reserved BARs
  PCI: dwc: Disable BARs in common code instead of in each glue driver
  PCI: dwc: Replace certain BAR_RESERVED with BAR_DISABLED in glue drivers
  PCI: endpoint: Introduce pci_epc_bar_type BAR_DISABLED
  PCI: dw-rockchip: Describe RK3588 BAR4 DMA ctrl window
  PCI: endpoint: Describe reserved subregions within BARs
  PCI: endpoint: Allow only_64bit on BAR_RESERVED
  PCI: endpoint: Do not mark the BAR succeeding a 64-bit BAR as BAR_RESERVED
  PCI: endpoint: Propagate error from pci_epf_create()
  PCI: endpoint: Fix typo in pci_epf_add_vepf() kernel-doc
  PCI: endpoint: pci-epf-vntb: Stop cmd_handler work in epf_ntb_epc_cleanup
  PCI: endpoint: pci-epf-ntb: Remove duplicate resource teardown
  PCI: endpoint: pci-epf-vntb: Remove duplicate resource teardown
  PCI: epf-mhi: Return 0, not remaining timeout, when eDMA ops complete
  PCI: endpoint: pci-epf-vntb: Return -ERANGE for out-of-range MW index
  PCI: endpoint: pci-epf-vntb: Use array_index_nospec() on mws_size[] access
  PCI: endpoint: Constify struct configfs_item_operations and configfs_group_operations
  selftests: pci_endpoint: Skip doorbell test when unsupported
  misc: pci_endpoint_test: Gate doorbell test on dynamic inbound mapping
  PCI: endpoint: pci-epf-test: Advertise dynamic inbound mapping support
  PCI: endpoint: pci-ep-msi: Fix error unwind and prevent double alloc
  PCI: endpoint: pci-epf-test: Don't free doorbell IRQ unless requested
  PCI: endpoint: pci-epf-vntb: Fix MSI doorbell IRQ unwind

2 files changed, 66 insertions, 8 deletions

diff --git a/include/linux/ntb.h b/include/linux/ntb.h
index 8ff9d663096b..879c3e89e026 100644
--- a/include/linux/ntb.h
+++ b/include/linux/ntb.h
@@ -256,6 +256,7 @@ static inline int ntb_ctx_ops_is_valid(const struct ntb_ctx_ops *ops)
  * @msg_clear_mask:	See ntb_msg_clear_mask().
  * @msg_read:		See ntb_msg_read().
  * @peer_msg_write:	See ntb_peer_msg_write().
+ * @get_dma_dev:	See ntb_get_dma_dev().
  */
 struct ntb_dev_ops {
 	int (*port_number)(struct ntb_dev *ntb);
@@ -329,6 +330,7 @@ struct ntb_dev_ops {
 	int (*msg_clear_mask)(struct ntb_dev *ntb, u64 mask_bits);
 	u32 (*msg_read)(struct ntb_dev *ntb, int *pidx, int midx);
 	int (*peer_msg_write)(struct ntb_dev *ntb, int pidx, int midx, u32 msg);
+	struct device *(*get_dma_dev)(struct ntb_dev *ntb);
 };
 
 static inline int ntb_dev_ops_is_valid(const struct ntb_dev_ops *ops)
@@ -391,6 +393,8 @@ static inline int ntb_dev_ops_is_valid(const struct ntb_dev_ops *ops)
 		/* !ops->msg_clear_mask == !ops->msg_count	&& */
 		!ops->msg_read == !ops->msg_count		&&
 		!ops->peer_msg_write == !ops->msg_count		&&
+
+		/* ops->get_dma_dev is optional */
 		1;
 }
 
@@ -1564,6 +1568,26 @@ static inline int ntb_peer_msg_write(struct ntb_dev *ntb, int pidx, int midx,
 }
 
 /**
+ * ntb_get_dma_dev() - get the device to use for DMA allocations/mappings
+ * @ntb:	NTB device context.
+ *
+ * Return a struct device suitable for DMA API allocations and mappings.
+ * This is typically the parent of the NTB device, but may be overridden by a
+ * driver by implementing .get_dma_dev().
+ *
+ * Drivers that implement .get_dma_dev() must return a non-NULL pointer.
+ *
+ * Return: device pointer to use for DMA operations.
+ */
+static inline struct device *ntb_get_dma_dev(struct ntb_dev *ntb)
+{
+	if (!ntb->ops->get_dma_dev)
+		return ntb->dev.parent;
+
+	return ntb->ops->get_dma_dev(ntb);
+}
+
+/**
  * ntb_peer_resource_idx() - get a resource index for a given peer idx
  * @ntb:	NTB device context.
  * @pidx:	Peer port index.
diff --git a/include/linux/pci-epc.h b/include/linux/pci-epc.h
index c021c7af175f..1eca1264815b 100644
--- a/include/linux/pci-epc.h
+++ b/include/linux/pci-epc.h
@@ -191,13 +191,49 @@ struct pci_epc {
  * @BAR_RESIZABLE: The BAR implements the PCI-SIG Resizable BAR Capability.
  *		   NOTE: An EPC driver can currently only set a single supported
  *		   size.
- * @BAR_RESERVED: The BAR should not be touched by an EPF driver.
+ * @BAR_RESERVED: Used for HW-backed BARs (e.g. MSI-X table, DMA regs). The BAR
+ *		  should not be disabled by an EPC driver. The BAR should not be
+ *		  reprogrammed by an EPF driver. An EPF driver is allowed to
+ *		  disable the BAR if absolutely necessary. (However, right now
+ *		  there is no EPC operation to disable a BAR that has not been
+ *		  programmed using pci_epc_set_bar().)
+ * @BAR_DISABLED: The BAR should be disabled by an EPC driver. The BAR will be
+ *		  unavailable to an EPF driver.
  */
 enum pci_epc_bar_type {
 	BAR_PROGRAMMABLE = 0,
 	BAR_FIXED,
 	BAR_RESIZABLE,
 	BAR_RESERVED,
+	BAR_DISABLED,
+};
+
+/**
+ * enum pci_epc_bar_rsvd_region_type - type of a fixed subregion behind a BAR
+ * @PCI_EPC_BAR_RSVD_DMA_CTRL_MMIO: Integrated DMA controller MMIO window
+ * @PCI_EPC_BAR_RSVD_MSIX_TBL_RAM: MSI-X table structure
+ * @PCI_EPC_BAR_RSVD_MSIX_PBA_RAM: MSI-X PBA structure
+ *
+ * BARs marked BAR_RESERVED are owned by the SoC/EPC hardware and must not be
+ * reprogrammed by EPF drivers. Some of them still expose fixed subregions that
+ * EPFs may want to reference (e.g. embedded doorbell fallback).
+ */
+enum pci_epc_bar_rsvd_region_type {
+	PCI_EPC_BAR_RSVD_DMA_CTRL_MMIO = 0,
+	PCI_EPC_BAR_RSVD_MSIX_TBL_RAM,
+	PCI_EPC_BAR_RSVD_MSIX_PBA_RAM,
+};
+
+/**
+ * struct pci_epc_bar_rsvd_region - fixed subregion behind a BAR
+ * @type: reserved region type
+ * @offset: offset within the BAR aperture
+ * @size: size of the reserved region
+ */
+struct pci_epc_bar_rsvd_region {
+	enum pci_epc_bar_rsvd_region_type	type;
+	resource_size_t				offset;
+	resource_size_t				size;
 };
 
 /**
@@ -206,18 +242,16 @@ enum pci_epc_bar_type {
  * @fixed_size: the fixed size, only applicable if type is BAR_FIXED_MASK.
  * @only_64bit: if true, an EPF driver is not allowed to choose if this BAR
  *		should be configured as 32-bit or 64-bit, the EPF driver must
- *		configure this BAR as 64-bit. Additionally, the BAR succeeding
- *		this BAR must be set to type BAR_RESERVED.
- *
- *		only_64bit should not be set on a BAR of type BAR_RESERVED.
- *		(If BARx is a 64-bit BAR that an EPF driver is not allowed to
- *		touch, then both BARx and BARx+1 must be set to type
- *		BAR_RESERVED.)
+ *		configure this BAR as 64-bit.
+ * @nr_rsvd_regions: number of fixed subregions described for BAR_RESERVED
+ * @rsvd_regions: fixed subregions behind BAR_RESERVED
  */
 struct pci_epc_bar_desc {
 	enum pci_epc_bar_type type;
 	u64 fixed_size;
 	bool only_64bit;
+	u8 nr_rsvd_regions;
+	const struct pci_epc_bar_rsvd_region *rsvd_regions;
 };
 
 /**