Merge branch 'pci/virtualization'

  - Fix erroneous intel-iommu dependency on CONFIG_AMD_IOMMU (Bjorn
    Helgaas)

  - Move pci_prg_resp_pasid_required() to CONFIG_PCI_PRI (Bjorn Helgaas)

  - Allow VFs to use PRI (the PF PRI is shared by the VFs, but the code
    previously didn't recognize that) (Kuppuswamy Sathyanarayanan)

  - Allow VFs to use PASID (the PF PASID capability is shared by the VFs,
    but the code previously didn't recognize that) (Kuppuswamy
    Sathyanarayanan)

  - Disconnect PF and VF ATS enablement, since ATS in PFs and associated
    VFs can be enabled independently (Kuppuswamy Sathyanarayanan)

  - Cache PRI and PASID capability offsets (Kuppuswamy Sathyanarayanan)

  - Cache the PRI PRG Response PASID Required bit (Bjorn Helgaas)

  - Consolidate ATS declarations in linux/pci-ats.h (Krzysztof Wilczynski)

  - Remove unused PRI and PASID stubs (Bjorn Helgaas)

  - Removed unnecessary EXPORT_SYMBOL_GPL() from ATS, PRI, and PASID
    interfaces that are only used by built-in IOMMU drivers (Bjorn Helgaas)

  - Hide PRI and PASID state restoration functions used only inside the PCI
    core (Bjorn Helgaas)

  - Fix the UPDCR register address in the Intel ACS quirk (Steffen
    Liebergeld)

  - Add a DMA alias quirk for the Intel VCA NTB (Slawomir Pawlowski)

  - Serialize sysfs sriov_numvfs reads vs writes (Pierre Crégut)

  - Update Cavium ACS quirk for ThunderX2 and ThunderX3 (George Cherian)

  - Unify ACS quirk implementations (Bjorn Helgaas)

* pci/virtualization:
  PCI: Unify ACS quirk desired vs provided checking
  PCI: Make ACS quirk implementations more uniform
  PCI: Apply Cavium ACS quirk to ThunderX2 and ThunderX3
  PCI/IOV: Serialize sysfs sriov_numvfs reads vs writes
  PCI: Add DMA alias quirk for Intel VCA NTB
  PCI: Fix Intel ACS quirk UPDCR register address
  PCI/ATS: Make pci_restore_pri_state(), pci_restore_pasid_state() private
  PCI/ATS: Remove unnecessary EXPORT_SYMBOL_GPL()
  PCI/ATS: Remove unused PRI and PASID stubs
  PCI/ATS: Consolidate ATS declarations in linux/pci-ats.h
  PCI/ATS: Cache PRI PRG Response PASID Required bit
  PCI/ATS: Cache PASID Capability offset
  PCI/ATS: Cache PRI Capability offset
  PCI/ATS: Disable PF/VF ATS service independently
  PCI/ATS: Handle sharing of PF PASID Capability with all VFs
  PCI/ATS: Handle sharing of PF PRI Capability with all VFs
  PCI/ATS: Move pci_prg_resp_pasid_required() to CONFIG_PCI_PRI
  iommu/vt-d: Select PCI_PRI for INTEL_IOMMU_SVM

1 files changed, 106 insertions, 46 deletions

diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 5d5e6f2f7837..a241a09a0211 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -4081,6 +4081,40 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2260, quirk_mic_x200_dma_alias);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2264, quirk_mic_x200_dma_alias);
 
 /*
+ * Intel Visual Compute Accelerator (VCA) is a family of PCIe add-in devices
+ * exposing computational units via Non Transparent Bridges (NTB, PEX 87xx).
+ *
+ * Similarly to MIC x200, we need to add DMA aliases to allow buffer access
+ * when IOMMU is enabled.  These aliases allow computational unit access to
+ * host memory.  These aliases mark the whole VCA device as one IOMMU
+ * group.
+ *
+ * All possible slot numbers (0x20) are used, since we are unable to tell
+ * what slot is used on other side.  This quirk is intended for both host
+ * and computational unit sides.  The VCA devices have up to five functions
+ * (four for DMA channels and one additional).
+ */
+static void quirk_pex_vca_alias(struct pci_dev *pdev)
+{
+	const unsigned int num_pci_slots = 0x20;
+	unsigned int slot;
+
+	for (slot = 0; slot < num_pci_slots; slot++) {
+		pci_add_dma_alias(pdev, PCI_DEVFN(slot, 0x0));
+		pci_add_dma_alias(pdev, PCI_DEVFN(slot, 0x1));
+		pci_add_dma_alias(pdev, PCI_DEVFN(slot, 0x2));
+		pci_add_dma_alias(pdev, PCI_DEVFN(slot, 0x3));
+		pci_add_dma_alias(pdev, PCI_DEVFN(slot, 0x4));
+	}
+}
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2954, quirk_pex_vca_alias);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2955, quirk_pex_vca_alias);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2956, quirk_pex_vca_alias);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2958, quirk_pex_vca_alias);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2959, quirk_pex_vca_alias);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x295A, quirk_pex_vca_alias);
+
+/*
  * The IOMMU and interrupt controller on Broadcom Vulcan/Cavium ThunderX2 are
  * associated not at the root bus, but at a bridge below. This quirk avoids
  * generating invalid DMA aliases.
@@ -4263,6 +4297,24 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_CHELSIO, PCI_ANY_ID,
 			 quirk_chelsio_T5_disable_root_port_attributes);
 
 /*
+ * pci_acs_ctrl_enabled - compare desired ACS controls with those provided
+ *			  by a device
+ * @acs_ctrl_req: Bitmask of desired ACS controls
+ * @acs_ctrl_ena: Bitmask of ACS controls enabled or provided implicitly by
+ *		  the hardware design
+ *
+ * Return 1 if all ACS controls in the @acs_ctrl_req bitmask are included
+ * in @acs_ctrl_ena, i.e., the device provides all the access controls the
+ * caller desires.  Return 0 otherwise.
+ */
+static int pci_acs_ctrl_enabled(u16 acs_ctrl_req, u16 acs_ctrl_ena)
+{
+	if ((acs_ctrl_req & acs_ctrl_ena) == acs_ctrl_req)
+		return 1;
+	return 0;
+}
+
+/*
  * AMD has indicated that the devices below do not support peer-to-peer
  * in any system where they are found in the southbridge with an AMD
  * IOMMU in the system.  Multifunction devices that do not support
@@ -4305,7 +4357,7 @@ static int pci_quirk_amd_sb_acs(struct pci_dev *dev, u16 acs_flags)
 	/* Filter out flags not applicable to multifunction */
 	acs_flags &= (PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_EC | PCI_ACS_DT);
 
-	return acs_flags & ~(PCI_ACS_RR | PCI_ACS_CR) ? 0 : 1;
+	return pci_acs_ctrl_enabled(acs_flags, PCI_ACS_RR | PCI_ACS_CR);
 #else
 	return -ENODEV;
 #endif
@@ -4313,33 +4365,38 @@ static int pci_quirk_amd_sb_acs(struct pci_dev *dev, u16 acs_flags)
 
 static bool pci_quirk_cavium_acs_match(struct pci_dev *dev)
 {
+	if (!pci_is_pcie(dev) || pci_pcie_type(dev) != PCI_EXP_TYPE_ROOT_PORT)
+		return false;
+
+	switch (dev->device) {
 	/*
-	 * Effectively selects all downstream ports for whole ThunderX 1
-	 * family by 0xf800 mask (which represents 8 SoCs), while the lower
-	 * bits of device ID are used to indicate which subdevice is used
-	 * within the SoC.
+	 * Effectively selects all downstream ports for whole ThunderX1
+	 * (which represents 8 SoCs).
 	 */
-	return (pci_is_pcie(dev) &&
-		(pci_pcie_type(dev) == PCI_EXP_TYPE_ROOT_PORT) &&
-		((dev->device & 0xf800) == 0xa000));
+	case 0xa000 ... 0xa7ff: /* ThunderX1 */
+	case 0xaf84:  /* ThunderX2 */
+	case 0xb884:  /* ThunderX3 */
+		return true;
+	default:
+		return false;
+	}
 }
 
 static int pci_quirk_cavium_acs(struct pci_dev *dev, u16 acs_flags)
 {
+	if (!pci_quirk_cavium_acs_match(dev))
+		return -ENOTTY;
+
 	/*
-	 * Cavium root ports don't advertise an ACS capability.  However,
+	 * Cavium Root Ports don't advertise an ACS capability.  However,
 	 * the RTL internally implements similar protection as if ACS had
-	 * Request Redirection, Completion Redirection, Source Validation,
+	 * Source Validation, Request Redirection, Completion Redirection,
 	 * and Upstream Forwarding features enabled.  Assert that the
 	 * hardware implements and enables equivalent ACS functionality for
 	 * these flags.
 	 */
-	acs_flags &= ~(PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_SV | PCI_ACS_UF);
-
-	if (!pci_quirk_cavium_acs_match(dev))
-		return -ENOTTY;
-
-	return acs_flags ? 0 : 1;
+	return pci_acs_ctrl_enabled(acs_flags,
+		PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
 }
 
 static int pci_quirk_xgene_acs(struct pci_dev *dev, u16 acs_flags)
@@ -4349,13 +4406,12 @@ static int pci_quirk_xgene_acs(struct pci_dev *dev, u16 acs_flags)
 	 * transactions with others, allowing masking out these bits as if they
 	 * were unimplemented in the ACS capability.
 	 */
-	acs_flags &= ~(PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
-
-	return acs_flags ? 0 : 1;
+	return pci_acs_ctrl_enabled(acs_flags,
+		PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
 }
 
 /*
- * Many Intel PCH root ports do provide ACS-like features to disable peer
+ * Many Intel PCH Root Ports do provide ACS-like features to disable peer
  * transactions and validate bus numbers in requests, but do not provide an
  * actual PCIe ACS capability.  This is the list of device IDs known to fall
  * into that category as provided by Intel in Red Hat bugzilla 1037684.
@@ -4403,37 +4459,32 @@ static bool pci_quirk_intel_pch_acs_match(struct pci_dev *dev)
 	return false;
 }
 
-#define INTEL_PCH_ACS_FLAGS (PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF | PCI_ACS_SV)
-
 static int pci_quirk_intel_pch_acs(struct pci_dev *dev, u16 acs_flags)
 {
-	u16 flags = dev->dev_flags & PCI_DEV_FLAGS_ACS_ENABLED_QUIRK ?
-		    INTEL_PCH_ACS_FLAGS : 0;
-
 	if (!pci_quirk_intel_pch_acs_match(dev))
 		return -ENOTTY;
 
-	return acs_flags & ~flags ? 0 : 1;
+	if (dev->dev_flags & PCI_DEV_FLAGS_ACS_ENABLED_QUIRK)
+		return pci_acs_ctrl_enabled(acs_flags,
+			PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
+
+	return pci_acs_ctrl_enabled(acs_flags, 0);
 }
 
 /*
- * These QCOM root ports do provide ACS-like features to disable peer
+ * These QCOM Root Ports do provide ACS-like features to disable peer
  * transactions and validate bus numbers in requests, but do not provide an
  * actual PCIe ACS capability.  Hardware supports source validation but it
  * will report the issue as Completer Abort instead of ACS Violation.
- * Hardware doesn't support peer-to-peer and each root port is a root
- * complex with unique segment numbers.  It is not possible for one root
- * port to pass traffic to another root port.  All PCIe transactions are
- * terminated inside the root port.
+ * Hardware doesn't support peer-to-peer and each Root Port is a Root
+ * Complex with unique segment numbers.  It is not possible for one Root
+ * Port to pass traffic to another Root Port.  All PCIe transactions are
+ * terminated inside the Root Port.
  */
 static int pci_quirk_qcom_rp_acs(struct pci_dev *dev, u16 acs_flags)
 {
-	u16 flags = (PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF | PCI_ACS_SV);
-	int ret = acs_flags & ~flags ? 0 : 1;
-
-	pci_info(dev, "Using QCOM ACS Quirk (%d)\n", ret);
-
-	return ret;
+	return pci_acs_ctrl_enabled(acs_flags,
+		PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
 }
 
 static int pci_quirk_al_acs(struct pci_dev *dev, u16 acs_flags)
@@ -4534,7 +4585,7 @@ static int pci_quirk_intel_spt_pch_acs(struct pci_dev *dev, u16 acs_flags)
 
 	pci_read_config_dword(dev, pos + INTEL_SPT_ACS_CTRL, &ctrl);
 
-	return acs_flags & ~ctrl ? 0 : 1;
+	return pci_acs_ctrl_enabled(acs_flags, ctrl);
 }
 
 static int pci_quirk_mf_endpoint_acs(struct pci_dev *dev, u16 acs_flags)
@@ -4548,10 +4599,9 @@ static int pci_quirk_mf_endpoint_acs(struct pci_dev *dev, u16 acs_flags)
 	 * perform peer-to-peer with other functions, allowing us to mask out
 	 * these bits as if they were unimplemented in the ACS capability.
 	 */
-	acs_flags &= ~(PCI_ACS_SV | PCI_ACS_TB | PCI_ACS_RR |
-		       PCI_ACS_CR | PCI_ACS_UF | PCI_ACS_DT);
-
-	return acs_flags ? 0 : 1;
+	return pci_acs_ctrl_enabled(acs_flags,
+		PCI_ACS_SV | PCI_ACS_TB | PCI_ACS_RR |
+		PCI_ACS_CR | PCI_ACS_UF | PCI_ACS_DT);
 }
 
 static int pci_quirk_brcm_acs(struct pci_dev *dev, u16 acs_flags)
@@ -4562,9 +4612,8 @@ static int pci_quirk_brcm_acs(struct pci_dev *dev, u16 acs_flags)
 	 * Allow each Root Port to be in a separate IOMMU group by masking
 	 * SV/RR/CR/UF bits.
 	 */
-	acs_flags &= ~(PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
-
-	return acs_flags ? 0 : 1;
+	return pci_acs_ctrl_enabled(acs_flags,
+		PCI_ACS_SV | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF);
 }
 
 static const struct pci_dev_acs_enabled {
@@ -4666,6 +4715,17 @@ static const struct pci_dev_acs_enabled {
 	{ 0 }
 };
 
+/*
+ * pci_dev_specific_acs_enabled - check whether device provides ACS controls
+ * @dev:	PCI device
+ * @acs_flags:	Bitmask of desired ACS controls
+ *
+ * Returns:
+ *   -ENOTTY:	No quirk applies to this device; we can't tell whether the
+ *		device provides the desired controls
+ *   0:		Device does not provide all the desired controls
+ *   >0:	Device provides all the controls in @acs_flags
+ */
 int pci_dev_specific_acs_enabled(struct pci_dev *dev, u16 acs_flags)
 {
 	const struct pci_dev_acs_enabled *i;
@@ -4706,7 +4766,7 @@ int pci_dev_specific_acs_enabled(struct pci_dev *dev, u16 acs_flags)
 #define INTEL_BSPR_REG_BPPD  (1 << 9)
 
 /* Upstream Peer Decode Configuration Register */
-#define INTEL_UPDCR_REG 0x1114
+#define INTEL_UPDCR_REG 0x1014
 /* 5:0 Peer Decode Enable bits */
 #define INTEL_UPDCR_REG_MASK 0x3f