diff options
Diffstat (limited to 'drivers/dma/ioat')
| -rw-r--r-- | drivers/dma/ioat/Makefile | 2 | ||||
| -rw-r--r-- | drivers/dma/ioat/dca.c | 684 | ||||
| -rw-r--r-- | drivers/dma/ioat/dma.c | 1139 | ||||
| -rw-r--r-- | drivers/dma/ioat/dma.h | 306 | ||||
| -rw-r--r-- | drivers/dma/ioat/dma_v2.c | 881 | ||||
| -rw-r--r-- | drivers/dma/ioat/dma_v2.h | 146 | ||||
| -rw-r--r-- | drivers/dma/ioat/hw.h | 73 | ||||
| -rw-r--r-- | drivers/dma/ioat/pci.c | 179 | ||||
| -rw-r--r-- | drivers/dma/ioat/registers.h | 230 |
9 files changed, 3640 insertions, 0 deletions
diff --git a/drivers/dma/ioat/Makefile b/drivers/dma/ioat/Makefile new file mode 100644 index 00000000000..205a639e84d --- /dev/null +++ b/drivers/dma/ioat/Makefile @@ -0,0 +1,2 @@ +obj-$(CONFIG_INTEL_IOATDMA) += ioatdma.o +ioatdma-objs := pci.o dma.o dma_v2.o dca.o diff --git a/drivers/dma/ioat/dca.c b/drivers/dma/ioat/dca.c new file mode 100644 index 00000000000..69d02615c4d --- /dev/null +++ b/drivers/dma/ioat/dca.c @@ -0,0 +1,684 @@ +/* + * Intel I/OAT DMA Linux driver + * Copyright(c) 2007 - 2009 Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + * + * The full GNU General Public License is included in this distribution in + * the file called "COPYING". + * + */ + +#include <linux/kernel.h> +#include <linux/pci.h> +#include <linux/smp.h> +#include <linux/interrupt.h> +#include <linux/dca.h> + +/* either a kernel change is needed, or we need something like this in kernel */ +#ifndef CONFIG_SMP +#include <asm/smp.h> +#undef cpu_physical_id +#define cpu_physical_id(cpu) (cpuid_ebx(1) >> 24) +#endif + +#include "dma.h" +#include "registers.h" + +/* + * Bit 7 of a tag map entry is the "valid" bit, if it is set then bits 0:6 + * contain the bit number of the APIC ID to map into the DCA tag. If the valid + * bit is not set, then the value must be 0 or 1 and defines the bit in the tag. + */ +#define DCA_TAG_MAP_VALID 0x80 + +#define DCA3_TAG_MAP_BIT_TO_INV 0x80 +#define DCA3_TAG_MAP_BIT_TO_SEL 0x40 +#define DCA3_TAG_MAP_LITERAL_VAL 0x1 + +#define DCA_TAG_MAP_MASK 0xDF + +/* expected tag map bytes for I/OAT ver.2 */ +#define DCA2_TAG_MAP_BYTE0 0x80 +#define DCA2_TAG_MAP_BYTE1 0x0 +#define DCA2_TAG_MAP_BYTE2 0x81 +#define DCA2_TAG_MAP_BYTE3 0x82 +#define DCA2_TAG_MAP_BYTE4 0x82 + +/* verify if tag map matches expected values */ +static inline int dca2_tag_map_valid(u8 *tag_map) +{ + return ((tag_map[0] == DCA2_TAG_MAP_BYTE0) && + (tag_map[1] == DCA2_TAG_MAP_BYTE1) && + (tag_map[2] == DCA2_TAG_MAP_BYTE2) && + (tag_map[3] == DCA2_TAG_MAP_BYTE3) && + (tag_map[4] == DCA2_TAG_MAP_BYTE4)); +} + +/* + * "Legacy" DCA systems do not implement the DCA register set in the + * I/OAT device. Software needs direct support for their tag mappings. + */ + +#define APICID_BIT(x) (DCA_TAG_MAP_VALID | (x)) +#define IOAT_TAG_MAP_LEN 8 + +static u8 ioat_tag_map_BNB[IOAT_TAG_MAP_LEN] = { + 1, APICID_BIT(1), APICID_BIT(2), APICID_BIT(2), }; +static u8 ioat_tag_map_SCNB[IOAT_TAG_MAP_LEN] = { + 1, APICID_BIT(1), APICID_BIT(2), APICID_BIT(2), }; +static u8 ioat_tag_map_CNB[IOAT_TAG_MAP_LEN] = { + 1, APICID_BIT(1), APICID_BIT(3), APICID_BIT(4), APICID_BIT(2), }; +static u8 ioat_tag_map_UNISYS[IOAT_TAG_MAP_LEN] = { 0 }; + +/* pack PCI B/D/F into a u16 */ +static inline u16 dcaid_from_pcidev(struct pci_dev *pci) +{ + return (pci->bus->number << 8) | pci->devfn; +} + +static int dca_enabled_in_bios(struct pci_dev *pdev) +{ + /* CPUID level 9 returns DCA configuration */ + /* Bit 0 indicates DCA enabled by the BIOS */ + unsigned long cpuid_level_9; + int res; + + cpuid_level_9 = cpuid_eax(9); + res = test_bit(0, &cpuid_level_9); + if (!res) + dev_err(&pdev->dev, "DCA is disabled in BIOS\n"); + + return res; +} + +static int system_has_dca_enabled(struct pci_dev *pdev) +{ + if (boot_cpu_has(X86_FEATURE_DCA)) + return dca_enabled_in_bios(pdev); + + dev_err(&pdev->dev, "boot cpu doesn't have X86_FEATURE_DCA\n"); + return 0; +} + +struct ioat_dca_slot { + struct pci_dev *pdev; /* requester device */ + u16 rid; /* requester id, as used by IOAT */ +}; + +#define IOAT_DCA_MAX_REQ 6 +#define IOAT3_DCA_MAX_REQ 2 + +struct ioat_dca_priv { + void __iomem *iobase; + void __iomem *dca_base; + int max_requesters; + int requester_count; + u8 tag_map[IOAT_TAG_MAP_LEN]; + struct ioat_dca_slot req_slots[0]; +}; + +/* 5000 series chipset DCA Port Requester ID Table Entry Format + * [15:8] PCI-Express Bus Number + * [7:3] PCI-Express Device Number + * [2:0] PCI-Express Function Number + * + * 5000 series chipset DCA control register format + * [7:1] Reserved (0) + * [0] Ignore Function Number + */ + +static int ioat_dca_add_requester(struct dca_provider *dca, struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + u16 id; + + /* This implementation only supports PCI-Express */ + if (dev->bus != &pci_bus_type) + return -ENODEV; + pdev = to_pci_dev(dev); + id = dcaid_from_pcidev(pdev); + + if (ioatdca->requester_count == ioatdca->max_requesters) + return -ENODEV; + + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == NULL) { + /* found an empty slot */ + ioatdca->requester_count++; + ioatdca->req_slots[i].pdev = pdev; + ioatdca->req_slots[i].rid = id; + writew(id, ioatdca->dca_base + (i * 4)); + /* make sure the ignore function bit is off */ + writeb(0, ioatdca->dca_base + (i * 4) + 2); + return i; + } + } + /* Error, ioatdma->requester_count is out of whack */ + return -EFAULT; +} + +static int ioat_dca_remove_requester(struct dca_provider *dca, + struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + + /* This implementation only supports PCI-Express */ + if (dev->bus != &pci_bus_type) + return -ENODEV; + pdev = to_pci_dev(dev); + + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == pdev) { + writew(0, ioatdca->dca_base + (i * 4)); + ioatdca->req_slots[i].pdev = NULL; + ioatdca->req_slots[i].rid = 0; + ioatdca->requester_count--; + return i; + } + } + return -ENODEV; +} + +static u8 ioat_dca_get_tag(struct dca_provider *dca, + struct device *dev, + int cpu) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + int i, apic_id, bit, value; + u8 entry, tag; + + tag = 0; + apic_id = cpu_physical_id(cpu); + + for (i = 0; i < IOAT_TAG_MAP_LEN; i++) { + entry = ioatdca->tag_map[i]; + if (entry & DCA_TAG_MAP_VALID) { + bit = entry & ~DCA_TAG_MAP_VALID; + value = (apic_id & (1 << bit)) ? 1 : 0; + } else { + value = entry ? 1 : 0; + } + tag |= (value << i); + } + return tag; +} + +static int ioat_dca_dev_managed(struct dca_provider *dca, + struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + + pdev = to_pci_dev(dev); + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == pdev) + return 1; + } + return 0; +} + +static struct dca_ops ioat_dca_ops = { + .add_requester = ioat_dca_add_requester, + .remove_requester = ioat_dca_remove_requester, + .get_tag = ioat_dca_get_tag, + .dev_managed = ioat_dca_dev_managed, +}; + + +struct dca_provider * __devinit +ioat_dca_init(struct pci_dev *pdev, void __iomem *iobase) +{ + struct dca_provider *dca; + struct ioat_dca_priv *ioatdca; + u8 *tag_map = NULL; + int i; + int err; + u8 version; + u8 max_requesters; + + if (!system_has_dca_enabled(pdev)) + return NULL; + + /* I/OAT v1 systems must have a known tag_map to support DCA */ + switch (pdev->vendor) { + case PCI_VENDOR_ID_INTEL: + switch (pdev->device) { + case PCI_DEVICE_ID_INTEL_IOAT: + tag_map = ioat_tag_map_BNB; + break; + case PCI_DEVICE_ID_INTEL_IOAT_CNB: + tag_map = ioat_tag_map_CNB; + break; + case PCI_DEVICE_ID_INTEL_IOAT_SCNB: + tag_map = ioat_tag_map_SCNB; + break; + } + break; + case PCI_VENDOR_ID_UNISYS: + switch (pdev->device) { + case PCI_DEVICE_ID_UNISYS_DMA_DIRECTOR: + tag_map = ioat_tag_map_UNISYS; + break; + } + break; + } + if (tag_map == NULL) + return NULL; + + version = readb(iobase + IOAT_VER_OFFSET); + if (version == IOAT_VER_3_0) + max_requesters = IOAT3_DCA_MAX_REQ; + else + max_requesters = IOAT_DCA_MAX_REQ; + + dca = alloc_dca_provider(&ioat_dca_ops, + sizeof(*ioatdca) + + (sizeof(struct ioat_dca_slot) * max_requesters)); + if (!dca) + return NULL; + + ioatdca = dca_priv(dca); + ioatdca->max_requesters = max_requesters; + ioatdca->dca_base = iobase + 0x54; + + /* copy over the APIC ID to DCA tag mapping */ + for (i = 0; i < IOAT_TAG_MAP_LEN; i++) + ioatdca->tag_map[i] = tag_map[i]; + + err = register_dca_provider(dca, &pdev->dev); + if (err) { + free_dca_provider(dca); + return NULL; + } + + return dca; +} + + +static int ioat2_dca_add_requester(struct dca_provider *dca, struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + u16 id; + u16 global_req_table; + + /* This implementation only supports PCI-Express */ + if (dev->bus != &pci_bus_type) + return -ENODEV; + pdev = to_pci_dev(dev); + id = dcaid_from_pcidev(pdev); + + if (ioatdca->requester_count == ioatdca->max_requesters) + return -ENODEV; + + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == NULL) { + /* found an empty slot */ + ioatdca->requester_count++; + ioatdca->req_slots[i].pdev = pdev; + ioatdca->req_slots[i].rid = id; + global_req_table = + readw(ioatdca->dca_base + IOAT_DCA_GREQID_OFFSET); + writel(id | IOAT_DCA_GREQID_VALID, + ioatdca->iobase + global_req_table + (i * 4)); + return i; + } + } + /* Error, ioatdma->requester_count is out of whack */ + return -EFAULT; +} + +static int ioat2_dca_remove_requester(struct dca_provider *dca, + struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + u16 global_req_table; + + /* This implementation only supports PCI-Express */ + if (dev->bus != &pci_bus_type) + return -ENODEV; + pdev = to_pci_dev(dev); + + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == pdev) { + global_req_table = + readw(ioatdca->dca_base + IOAT_DCA_GREQID_OFFSET); + writel(0, ioatdca->iobase + global_req_table + (i * 4)); + ioatdca->req_slots[i].pdev = NULL; + ioatdca->req_slots[i].rid = 0; + ioatdca->requester_count--; + return i; + } + } + return -ENODEV; +} + +static u8 ioat2_dca_get_tag(struct dca_provider *dca, + struct device *dev, + int cpu) +{ + u8 tag; + + tag = ioat_dca_get_tag(dca, dev, cpu); + tag = (~tag) & 0x1F; + return tag; +} + +static struct dca_ops ioat2_dca_ops = { + .add_requester = ioat2_dca_add_requester, + .remove_requester = ioat2_dca_remove_requester, + .get_tag = ioat2_dca_get_tag, + .dev_managed = ioat_dca_dev_managed, +}; + +static int ioat2_dca_count_dca_slots(void __iomem *iobase, u16 dca_offset) +{ + int slots = 0; + u32 req; + u16 global_req_table; + + global_req_table = readw(iobase + dca_offset + IOAT_DCA_GREQID_OFFSET); + if (global_req_table == 0) + return 0; + do { + req = readl(iobase + global_req_table + (slots * sizeof(u32))); + slots++; + } while ((req & IOAT_DCA_GREQID_LASTID) == 0); + + return slots; +} + +struct dca_provider * __devinit +ioat2_dca_init(struct pci_dev *pdev, void __iomem *iobase) +{ + struct dca_provider *dca; + struct ioat_dca_priv *ioatdca; + int slots; + int i; + int err; + u32 tag_map; + u16 dca_offset; + u16 csi_fsb_control; + u16 pcie_control; + u8 bit; + + if (!system_has_dca_enabled(pdev)) + return NULL; + + dca_offset = readw(iobase + IOAT_DCAOFFSET_OFFSET); + if (dca_offset == 0) + return NULL; + + slots = ioat2_dca_count_dca_slots(iobase, dca_offset); + if (slots == 0) + return NULL; + + dca = alloc_dca_provider(&ioat2_dca_ops, + sizeof(*ioatdca) + + (sizeof(struct ioat_dca_slot) * slots)); + if (!dca) + return NULL; + + ioatdca = dca_priv(dca); + ioatdca->iobase = iobase; + ioatdca->dca_base = iobase + dca_offset; + ioatdca->max_requesters = slots; + + /* some bios might not know to turn these on */ + csi_fsb_control = readw(ioatdca->dca_base + IOAT_FSB_CAP_ENABLE_OFFSET); + if ((csi_fsb_control & IOAT_FSB_CAP_ENABLE_PREFETCH) == 0) { + csi_fsb_control |= IOAT_FSB_CAP_ENABLE_PREFETCH; + writew(csi_fsb_control, + ioatdca->dca_base + IOAT_FSB_CAP_ENABLE_OFFSET); + } + pcie_control = readw(ioatdca->dca_base + IOAT_PCI_CAP_ENABLE_OFFSET); + if ((pcie_control & IOAT_PCI_CAP_ENABLE_MEMWR) == 0) { + pcie_control |= IOAT_PCI_CAP_ENABLE_MEMWR; + writew(pcie_control, + ioatdca->dca_base + IOAT_PCI_CAP_ENABLE_OFFSET); + } + + + /* TODO version, compatibility and configuration checks */ + + /* copy out the APIC to DCA tag map */ + tag_map = readl(ioatdca->dca_base + IOAT_APICID_TAG_MAP_OFFSET); + for (i = 0; i < 5; i++) { + bit = (tag_map >> (4 * i)) & 0x0f; + if (bit < 8) + ioatdca->tag_map[i] = bit | DCA_TAG_MAP_VALID; + else + ioatdca->tag_map[i] = 0; + } + + if (!dca2_tag_map_valid(ioatdca->tag_map)) { + dev_err(&pdev->dev, "APICID_TAG_MAP set incorrectly by BIOS, " + "disabling DCA\n"); + free_dca_provider(dca); + return NULL; + } + + err = register_dca_provider(dca, &pdev->dev); + if (err) { + free_dca_provider(dca); + return NULL; + } + + return dca; +} + +static int ioat3_dca_add_requester(struct dca_provider *dca, struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + u16 id; + u16 global_req_table; + + /* This implementation only supports PCI-Express */ + if (dev->bus != &pci_bus_type) + return -ENODEV; + pdev = to_pci_dev(dev); + id = dcaid_from_pcidev(pdev); + + if (ioatdca->requester_count == ioatdca->max_requesters) + return -ENODEV; + + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == NULL) { + /* found an empty slot */ + ioatdca->requester_count++; + ioatdca->req_slots[i].pdev = pdev; + ioatdca->req_slots[i].rid = id; + global_req_table = + readw(ioatdca->dca_base + IOAT3_DCA_GREQID_OFFSET); + writel(id | IOAT_DCA_GREQID_VALID, + ioatdca->iobase + global_req_table + (i * 4)); + return i; + } + } + /* Error, ioatdma->requester_count is out of whack */ + return -EFAULT; +} + +static int ioat3_dca_remove_requester(struct dca_provider *dca, + struct device *dev) +{ + struct ioat_dca_priv *ioatdca = dca_priv(dca); + struct pci_dev *pdev; + int i; + u16 global_req_table; + + /* This implementation only supports PCI-Express */ + if (dev->bus != &pci_bus_type) + return -ENODEV; + pdev = to_pci_dev(dev); + + for (i = 0; i < ioatdca->max_requesters; i++) { + if (ioatdca->req_slots[i].pdev == pdev) { + global_req_table = + readw(ioatdca->dca_base + IOAT3_DCA_GREQID_OFFSET); + writel(0, ioatdca->iobase + global_req_table + (i * 4)); + ioatdca->req_slots[i].pdev = NULL; + ioatdca->req_slots[i].rid = 0; + ioatdca->requester_count--; + return i; + } + } + return -ENODEV; +} + +static u8 ioat3_dca_get_tag(struct dca_provider *dca, + struct device *dev, + int cpu) +{ + u8 tag; + + struct ioat_dca_priv *ioatdca = dca_priv(dca); + int i, apic_id, bit, value; + u8 entry; + + tag = 0; + apic_id = cpu_physical_id(cpu); + + for (i = 0; i < IOAT_TAG_MAP_LEN; i++) { + entry = ioatdca->tag_map[i]; + if (entry & DCA3_TAG_MAP_BIT_TO_SEL) { + bit = entry & + ~(DCA3_TAG_MAP_BIT_TO_SEL | DCA3_TAG_MAP_BIT_TO_INV); + value = (apic_id & (1 << bit)) ? 1 : 0; + } else if (entry & DCA3_TAG_MAP_BIT_TO_INV) { + bit = entry & ~DCA3_TAG_MAP_BIT_TO_INV; + value = (apic_id & (1 << bit)) ? 0 : 1; + } else { + value = (entry & DCA3_TAG_MAP_LITERAL_VAL) ? 1 : 0; + } + tag |= (value << i); + } + + return tag; +} + +static struct dca_ops ioat3_dca_ops = { + .add_requester = ioat3_dca_add_requester, + .remove_requester = ioat3_dca_remove_requester, + .get_tag = ioat3_dca_get_tag, + .dev_managed = ioat_dca_dev_managed, +}; + +static int ioat3_dca_count_dca_slots(void *iobase, u16 dca_offset) +{ + int slots = 0; + u32 req; + u16 global_req_table; + + global_req_table = readw(iobase + dca_offset + IOAT3_DCA_GREQID_OFFSET); + if (global_req_table == 0) + return 0; + + do { + req = readl(iobase + global_req_table + (slots * sizeof(u32))); + slots++; + } while ((req & IOAT_DCA_GREQID_LASTID) == 0); + + return slots; +} + +struct dca_provider * __devinit +ioat3_dca_init(struct pci_dev *pdev, void __iomem *iobase) +{ + struct dca_provider *dca; + struct ioat_dca_priv *ioatdca; + int slots; + int i; + int err; + u16 dca_offset; + u16 csi_fsb_control; + u16 pcie_control; + u8 bit; + + union { + u64 full; + struct { + u32 low; + u32 high; + }; + } tag_map; + + if (!system_has_dca_enabled(pdev)) + return NULL; + + dca_offset = readw(iobase + IOAT_DCAOFFSET_OFFSET); + if (dca_offset == 0) + return NULL; + + slots = ioat3_dca_count_dca_slots(iobase, dca_offset); + if (slots == 0) + return NULL; + + dca = alloc_dca_provider(&ioat3_dca_ops, + sizeof(*ioatdca) + + (sizeof(struct ioat_dca_slot) * slots)); + if (!dca) + return NULL; + + ioatdca = dca_priv(dca); + ioatdca->iobase = iobase; + ioatdca->dca_base = iobase + dca_offset; + ioatdca->max_requesters = slots; + + /* some bios might not know to turn these on */ + csi_fsb_control = readw(ioatdca->dca_base + IOAT3_CSI_CONTROL_OFFSET); + if ((csi_fsb_control & IOAT3_CSI_CONTROL_PREFETCH) == 0) { + csi_fsb_control |= IOAT3_CSI_CONTROL_PREFETCH; + writew(csi_fsb_control, + ioatdca->dca_base + IOAT3_CSI_CONTROL_OFFSET); + } + pcie_control = readw(ioatdca->dca_base + IOAT3_PCI_CONTROL_OFFSET); + if ((pcie_control & IOAT3_PCI_CONTROL_MEMWR) == 0) { + pcie_control |= IOAT3_PCI_CONTROL_MEMWR; + writew(pcie_control, + ioatdca->dca_base + IOAT3_PCI_CONTROL_OFFSET); + } + + + /* TODO version, compatibility and configuration checks */ + + /* copy out the APIC to DCA tag map */ + tag_map.low = + readl(ioatdca->dca_base + IOAT3_APICID_TAG_MAP_OFFSET_LOW); + tag_map.high = + readl(ioatdca->dca_base + IOAT3_APICID_TAG_MAP_OFFSET_HIGH); + for (i = 0; i < 8; i++) { + bit = tag_map.full >> (8 * i); + ioatdca->tag_map[i] = bit & DCA_TAG_MAP_MASK; + } + + err = register_dca_provider(dca, &pdev->dev); + if (err) { + free_dca_provider(dca); + return NULL; + } + + return dca; +} diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c new file mode 100644 index 00000000000..17a518d0386 --- /dev/null +++ b/drivers/dma/ioat/dma.c @@ -0,0 +1,1139 @@ +/* + * Intel I/OAT DMA Linux driver + * Copyright(c) 2004 - 2009 Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + * + * The full GNU General Public License is included in this distribution in + * the file called "COPYING". + * + */ + +/* + * This driver supports an Intel I/OAT DMA engine, which does asynchronous + * copy operations. + */ + +#include <linux/init.h> +#include <linux/module.h> +#include <linux/pci.h> +#include <linux/interrupt.h> +#include <linux/dmaengine.h> +#include <linux/delay.h> +#include <linux/dma-mapping.h> +#include <linux/workqueue.h> +#include <linux/i7300_idle.h> +#include "dma.h" +#include "registers.h" +#include "hw.h" + +int ioat_pending_level = 4; +module_param(ioat_pending_level, int, 0644); +MODULE_PARM_DESC(ioat_pending_level, + "high-water mark for pushing ioat descriptors (default: 4)"); + +/* internal functions */ +static void ioat1_cleanup(struct ioat_dma_chan *ioat); +static void ioat1_dma_start_null_desc(struct ioat_dma_chan *ioat); + +/** + * ioat_dma_do_interrupt - handler used for single vector interrupt mode + * @irq: interrupt id + * @data: interrupt data + */ +static irqreturn_t ioat_dma_do_interrupt(int irq, void *data) +{ + struct ioatdma_device *instance = data; + struct ioat_chan_common *chan; + unsigned long attnstatus; + int bit; + u8 intrctrl; + + intrctrl = readb(instance->reg_base + IOAT_INTRCTRL_OFFSET); + + if (!(intrctrl & IOAT_INTRCTRL_MASTER_INT_EN)) + return IRQ_NONE; + + if (!(intrctrl & IOAT_INTRCTRL_INT_STATUS)) { + writeb(intrctrl, instance->reg_base + IOAT_INTRCTRL_OFFSET); + return IRQ_NONE; + } + + attnstatus = readl(instance->reg_base + IOAT_ATTNSTATUS_OFFSET); + for_each_bit(bit, &attnstatus, BITS_PER_LONG) { + chan = ioat_chan_by_index(instance, bit); + tasklet_schedule(&chan->cleanup_task); + } + + writeb(intrctrl, instance->reg_base + IOAT_INTRCTRL_OFFSET); + return IRQ_HANDLED; +} + +/** + * ioat_dma_do_interrupt_msix - handler used for vector-per-channel interrupt mode + * @irq: interrupt id + * @data: interrupt data + */ +static irqreturn_t ioat_dma_do_interrupt_msix(int irq, void *data) +{ + struct ioat_chan_common *chan = data; + + tasklet_schedule(&chan->cleanup_task); + + return IRQ_HANDLED; +} + +static void ioat1_cleanup_tasklet(unsigned long data); + +/* common channel initialization */ +void ioat_init_channel(struct ioatdma_device *device, + struct ioat_chan_common *chan, int idx, + void (*timer_fn)(unsigned long), + void (*tasklet)(unsigned long), + unsigned long ioat) +{ + struct dma_device *dma = &device->common; + + chan->device = device; + chan->reg_base = device->reg_base + (0x80 * (idx + 1)); + spin_lock_init(&chan->cleanup_lock); + chan->common.device = dma; + list_add_tail(&chan->common.device_node, &dma->channels); + device->idx[idx] = chan; + init_timer(&chan->timer); + chan->timer.function = timer_fn; + chan->timer.data = ioat; + tasklet_init(&chan->cleanup_task, tasklet, ioat); + tasklet_disable(&chan->cleanup_task); +} + +static void ioat1_timer_event(unsigned long data); + +/** + * ioat1_dma_enumerate_channels - find and initialize the device's channels + * @device: the device to be enumerated + */ +static int ioat1_enumerate_channels(struct ioatdma_device *device) +{ + u8 xfercap_scale; + u32 xfercap; + int i; + struct ioat_dma_chan *ioat; + struct device *dev = &device->pdev->dev; + struct dma_device *dma = &device->common; + + INIT_LIST_HEAD(&dma->channels); + dma->chancnt = readb(device->reg_base + IOAT_CHANCNT_OFFSET); + dma->chancnt &= 0x1f; /* bits [4:0] valid */ + if (dma->chancnt > ARRAY_SIZE(device->idx)) { + dev_warn(dev, "(%d) exceeds max supported channels (%zu)\n", + dma->chancnt, ARRAY_SIZE(device->idx)); + dma->chancnt = ARRAY_SIZE(device->idx); + } + xfercap_scale = readb(device->reg_base + IOAT_XFERCAP_OFFSET); + xfercap_scale &= 0x1f; /* bits [4:0] valid */ + xfercap = (xfercap_scale == 0 ? -1 : (1UL << xfercap_scale)); + dev_dbg(dev, "%s: xfercap = %d\n", __func__, xfercap); + +#ifdef CONFIG_I7300_IDLE_IOAT_CHANNEL + if (i7300_idle_platform_probe(NULL, NULL, 1) == 0) + dma->chancnt--; +#endif + for (i = 0; i < dma->chancnt; i++) { + ioat = devm_kzalloc(dev, sizeof(*ioat), GFP_KERNEL); + if (!ioat) + break; + + ioat_init_channel(device, &ioat->base, i, + ioat1_timer_event, + ioat1_cleanup_tasklet, + (unsigned long) ioat); + ioat->xfercap = xfercap; + spin_lock_init(&ioat->desc_lock); + INIT_LIST_HEAD(&ioat->free_desc); + INIT_LIST_HEAD(&ioat->used_desc); + } + dma->chancnt = i; + return i; +} + +/** + * ioat_dma_memcpy_issue_pending - push potentially unrecognized appended + * descriptors to hw + * @chan: DMA channel handle + */ +static inline void +__ioat1_dma_memcpy_issue_pending(struct ioat_dma_chan *ioat) +{ + void __iomem *reg_base = ioat->base.reg_base; + + dev_dbg(to_dev(&ioat->base), "%s: pending: %d\n", + __func__, ioat->pending); + ioat->pending = 0; + writeb(IOAT_CHANCMD_APPEND, reg_base + IOAT1_CHANCMD_OFFSET); +} + +static void ioat1_dma_memcpy_issue_pending(struct dma_chan *chan) +{ + struct ioat_dma_chan *ioat = to_ioat_chan(chan); + + if (ioat->pending > 0) { + spin_lock_bh(&ioat->desc_lock); + __ioat1_dma_memcpy_issue_pending(ioat); + spin_unlock_bh(&ioat->desc_lock); + } +} + +/** + * ioat1_reset_channel - restart a channel + * @ioat: IOAT DMA channel handle + */ +static void ioat1_reset_channel(struct ioat_dma_chan *ioat) +{ + struct ioat_chan_common *chan = &ioat->base; + void __iomem *reg_base = chan->reg_base; + u32 chansts, chanerr; + + dev_warn(to_dev(chan), "reset\n"); + chanerr = readl(reg_base + IOAT_CHANERR_OFFSET); + chansts = *chan->completion & IOAT_CHANSTS_STATUS; + if (chanerr) { + dev_err(to_dev(chan), + "chan%d, CHANSTS = 0x%08x CHANERR = 0x%04x, clearing\n", + chan_num(chan), chansts, chanerr); + writel(chanerr, reg_base + IOAT_CHANERR_OFFSET); + } + + /* + * whack it upside the head with a reset + * and wait for things to settle out. + * force the pending count to a really big negative + * to make sure no one forces an issue_pending + * while we're waiting. + */ + + ioat->pending = INT_MIN; + writeb(IOAT_CHANCMD_RESET, + reg_base + IOAT_CHANCMD_OFFSET(chan->device->version)); + set_bit(IOAT_RESET_PENDING, &chan->state); + mod_timer(&chan->timer, jiffies + RESET_DELAY); +} + +static dma_cookie_t ioat1_tx_submit(struct dma_async_tx_descriptor *tx) +{ + struct dma_chan *c = tx->chan; + struct ioat_dma_chan *ioat = to_ioat_chan(c); + struct ioat_desc_sw *desc = tx_to_ioat_desc(tx); + struct ioat_chan_common *chan = &ioat->base; + struct ioat_desc_sw *first; + struct ioat_desc_sw *chain_tail; + dma_cookie_t cookie; + + spin_lock_bh(&ioat->desc_lock); + /* cookie incr and addition to used_list must be atomic */ + cookie = c->cookie; + cookie++; + if (cookie < 0) + cookie = 1; + c->cookie = cookie; + tx->cookie = cookie; + dev_dbg(to_dev(&ioat->base), "%s: cookie: %d\n", __func__, cookie); + + /* write address into NextDescriptor field of last desc in chain */ + first = to_ioat_desc(tx->tx_list.next); + chain_tail = to_ioat_desc(ioat->used_desc.prev); + /* make descriptor updates globally visible before chaining */ + wmb(); + chain_tail->hw->next = first->txd.phys; + list_splice_tail_init(&tx->tx_list, &ioat->used_desc); + dump_desc_dbg(ioat, chain_tail); + dump_desc_dbg(ioat, first); + + if (!test_and_set_bit(IOAT_COMPLETION_PENDING, &chan->state)) + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + + ioat->pending += desc->hw->tx_cnt; + if (ioat->pending >= ioat_pending_level) + __ioat1_dma_memcpy_issue_pending(ioat); + spin_unlock_bh(&ioat->desc_lock); + + return cookie; +} + +/** + * ioat_dma_alloc_descriptor - allocate and return a sw and hw descriptor pair + * @ioat: the channel supplying the memory pool for the descriptors + * @flags: allocation flags + */ +static struct ioat_desc_sw * +ioat_dma_alloc_descriptor(struct ioat_dma_chan *ioat, gfp_t flags) +{ + struct ioat_dma_descriptor *desc; + struct ioat_desc_sw *desc_sw; + struct ioatdma_device *ioatdma_device; + dma_addr_t phys; + + ioatdma_device = ioat->base.device; + desc = pci_pool_alloc(ioatdma_device->dma_pool, flags, &phys); + if (unlikely(!desc)) + return NULL; + + desc_sw = kzalloc(sizeof(*desc_sw), flags); + if (unlikely(!desc_sw)) { + pci_pool_free(ioatdma_device->dma_pool, desc, phys); + return NULL; + } + + memset(desc, 0, sizeof(*desc)); + + dma_async_tx_descriptor_init(&desc_sw->txd, &ioat->base.common); + desc_sw->txd.tx_submit = ioat1_tx_submit; + desc_sw->hw = desc; + desc_sw->txd.phys = phys; + set_desc_id(desc_sw, -1); + + return desc_sw; +} + +static int ioat_initial_desc_count = 256; +module_param(ioat_initial_desc_count, int, 0644); +MODULE_PARM_DESC(ioat_initial_desc_count, + "ioat1: initial descriptors per channel (default: 256)"); +/** + * ioat1_dma_alloc_chan_resources - returns the number of allocated descriptors + * @chan: the channel to be filled out + */ +static int ioat1_dma_alloc_chan_resources(struct dma_chan *c) +{ + struct ioat_dma_chan *ioat = to_ioat_chan(c); + struct ioat_chan_common *chan = &ioat->base; + struct ioat_desc_sw *desc; + u32 chanerr; + int i; + LIST_HEAD(tmp_list); + + /* have we already been set up? */ + if (!list_empty(&ioat->free_desc)) + return ioat->desccount; + + /* Setup register to interrupt and write completion status on error */ + writew(IOAT_CHANCTRL_RUN, chan->reg_base + IOAT_CHANCTRL_OFFSET); + + chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET); + if (chanerr) { + dev_err(to_dev(chan), "CHANERR = %x, clearing\n", chanerr); + writel(chanerr, chan->reg_base + IOAT_CHANERR_OFFSET); + } + + /* Allocate descriptors */ + for (i = 0; i < ioat_initial_desc_count; i++) { + desc = ioat_dma_alloc_descriptor(ioat, GFP_KERNEL); + if (!desc) { + dev_err(to_dev(chan), "Only %d initial descriptors\n", i); + break; + } + set_desc_id(desc, i); + list_add_tail(&desc->node, &tmp_list); + } + spin_lock_bh(&ioat->desc_lock); + ioat->desccount = i; + list_splice(&tmp_list, &ioat->free_desc); + spin_unlock_bh(&ioat->desc_lock); + + /* allocate a completion writeback area */ + /* doing 2 32bit writes to mmio since 1 64b write doesn't work */ + chan->completion = pci_pool_alloc(chan->device->completion_pool, + GFP_KERNEL, &chan->completion_dma); + memset(chan->completion, 0, sizeof(*chan->completion)); + writel(((u64) chan->completion_dma) & 0x00000000FFFFFFFF, + chan->reg_base + IOAT_CHANCMP_OFFSET_LOW); + writel(((u64) chan->completion_dma) >> 32, + chan->reg_base + IOAT_CHANCMP_OFFSET_HIGH); + + tasklet_enable(&chan->cleanup_task); + ioat1_dma_start_null_desc(ioat); /* give chain to dma device */ + dev_dbg(to_dev(chan), "%s: allocated %d descriptors\n", + __func__, ioat->desccount); + return ioat->desccount; +} + +/** + * ioat1_dma_free_chan_resources - release all the descriptors + * @chan: the channel to be cleaned + */ +static void ioat1_dma_free_chan_resources(struct dma_chan *c) +{ + struct ioat_dma_chan *ioat = to_ioat_chan(c); + struct ioat_chan_common *chan = &ioat->base; + struct ioatdma_device *ioatdma_device = chan->device; + struct ioat_desc_sw *desc, *_desc; + int in_use_descs = 0; + + /* Before freeing channel resources first check + * if they have been previously allocated for this channel. + */ + if (ioat->desccount == 0) + return; + + tasklet_disable(&chan->cleanup_task); + del_timer_sync(&chan->timer); + ioat1_cleanup(ioat); + + /* Delay 100ms after reset to allow internal DMA logic to quiesce + * before removing DMA descriptor resources. + */ + writeb(IOAT_CHANCMD_RESET, + chan->reg_base + IOAT_CHANCMD_OFFSET(chan->device->version)); + mdelay(100); + + spin_lock_bh(&ioat->desc_lock); + list_for_each_entry_safe(desc, _desc, &ioat->used_desc, node) { + dev_dbg(to_dev(chan), "%s: freeing %d from used list\n", + __func__, desc_id(desc)); + dump_desc_dbg(ioat, desc); + in_use_descs++; + list_del(&desc->node); + pci_pool_free(ioatdma_device->dma_pool, desc->hw, + desc->txd.phys); + kfree(desc); + } + list_for_each_entry_safe(desc, _desc, + &ioat->free_desc, node) { + list_del(&desc->node); + pci_pool_free(ioatdma_device->dma_pool, desc->hw, + desc->txd.phys); + kfree(desc); + } + spin_unlock_bh(&ioat->desc_lock); + + pci_pool_free(ioatdma_device->completion_pool, + chan->completion, + chan->completion_dma); + + /* one is ok since we left it on there on purpose */ + if (in_use_descs > 1) + dev_err(to_dev(chan), "Freeing %d in use descriptors!\n", + in_use_descs - 1); + + chan->last_completion = 0; + chan->completion_dma = 0; + ioat->pending = 0; + ioat->desccount = 0; +} + +/** + * ioat1_dma_get_next_descriptor - return the next available descriptor + * @ioat: IOAT DMA channel handle + * + * Gets the next descriptor from the chain, and must be called with the + * channel's desc_lock held. Allocates more descriptors if the channel + * has run out. + */ +static struct ioat_desc_sw * +ioat1_dma_get_next_descriptor(struct ioat_dma_chan *ioat) +{ + struct ioat_desc_sw *new; + + if (!list_empty(&ioat->free_desc)) { + new = to_ioat_desc(ioat->free_desc.next); + list_del(&new->node); + } else { + /* try to get another desc */ + new = ioat_dma_alloc_descriptor(ioat, GFP_ATOMIC); + if (!new) { + dev_err(to_dev(&ioat->base), "alloc failed\n"); + return NULL; + } + } + dev_dbg(to_dev(&ioat->base), "%s: allocated: %d\n", + __func__, desc_id(new)); + prefetch(new->hw); + return new; +} + +static struct dma_async_tx_descriptor * +ioat1_dma_prep_memcpy(struct dma_chan *c, dma_addr_t dma_dest, + dma_addr_t dma_src, size_t len, unsigned long flags) +{ + struct ioat_dma_chan *ioat = to_ioat_chan(c); + struct ioat_desc_sw *desc; + size_t copy; + LIST_HEAD(chain); + dma_addr_t src = dma_src; + dma_addr_t dest = dma_dest; + size_t total_len = len; + struct ioat_dma_descriptor *hw = NULL; + int tx_cnt = 0; + + spin_lock_bh(&ioat->desc_lock); + desc = ioat1_dma_get_next_descriptor(ioat); + do { + if (!desc) + break; + + tx_cnt++; + copy = min_t(size_t, len, ioat->xfercap); + + hw = desc->hw; + hw->size = copy; + hw->ctl = 0; + hw->src_addr = src; + hw->dst_addr = dest; + + list_add_tail(&desc->node, &chain); + + len -= copy; + dest += copy; + src += copy; + if (len) { + struct ioat_desc_sw *next; + + async_tx_ack(&desc->txd); + next = ioat1_dma_get_next_descriptor(ioat); + hw->next = next ? next->txd.phys : 0; + dump_desc_dbg(ioat, desc); + desc = next; + } else + hw->next = 0; + } while (len); + + if (!desc) { + struct ioat_chan_common *chan = &ioat->base; + + dev_err(to_dev(chan), + "chan%d - get_next_desc failed\n", chan_num(chan)); + list_splice(&chain, &ioat->free_desc); + spin_unlock_bh(&ioat->desc_lock); + return NULL; + } + spin_unlock_bh(&ioat->desc_lock); + + desc->txd.flags = flags; + desc->len = total_len; + list_splice(&chain, &desc->txd.tx_list); + hw->ctl_f.int_en = !!(flags & DMA_PREP_INTERRUPT); + hw->ctl_f.compl_write = 1; + hw->tx_cnt = tx_cnt; + dump_desc_dbg(ioat, desc); + + return &desc->txd; +} + +static void ioat1_cleanup_tasklet(unsigned long data) +{ + struct ioat_dma_chan *chan = (void *)data; + + ioat1_cleanup(chan); + writew(IOAT_CHANCTRL_RUN, chan->base.reg_base + IOAT_CHANCTRL_OFFSET); +} + +static void ioat_unmap(struct pci_dev *pdev, dma_addr_t addr, size_t len, + int direction, enum dma_ctrl_flags flags, bool dst) +{ + if ((dst && (flags & DMA_COMPL_DEST_UNMAP_SINGLE)) || + (!dst && (flags & DMA_COMPL_SRC_UNMAP_SINGLE))) + pci_unmap_single(pdev, addr, len, direction); + else + pci_unmap_page(pdev, addr, len, direction); +} + + +void ioat_dma_unmap(struct ioat_chan_common *chan, enum dma_ctrl_flags flags, + size_t len, struct ioat_dma_descriptor *hw) +{ + struct pci_dev *pdev = chan->device->pdev; + size_t offset = len - hw->size; + + if (!(flags & DMA_COMPL_SKIP_DEST_UNMAP)) + ioat_unmap(pdev, hw->dst_addr - offset, len, + PCI_DMA_FROMDEVICE, flags, 1); + + if (!(flags & DMA_COMPL_SKIP_SRC_UNMAP)) + ioat_unmap(pdev, hw->src_addr - offset, len, + PCI_DMA_TODEVICE, flags, 0); +} + +unsigned long ioat_get_current_completion(struct ioat_chan_common *chan) +{ + unsigned long phys_complete; + u64 completion; + + completion = *chan->completion; + phys_complete = ioat_chansts_to_addr(completion); + + dev_dbg(to_dev(chan), "%s: phys_complete: %#llx\n", __func__, + (unsigned long long) phys_complete); + + if (is_ioat_halted(completion)) { + u32 chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET); + dev_err(to_dev(chan), "Channel halted, chanerr = %x\n", + chanerr); + + /* TODO do something to salvage the situation */ + } + + return phys_complete; +} + +bool ioat_cleanup_preamble(struct ioat_chan_common *chan, + unsigned long *phys_complete) +{ + *phys_complete = ioat_get_current_completion(chan); + if (*phys_complete == chan->last_completion) + return false; + clear_bit(IOAT_COMPLETION_ACK, &chan->state); + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + + return true; +} + +static void __cleanup(struct ioat_dma_chan *ioat, unsigned long phys_complete) +{ + struct ioat_chan_common *chan = &ioat->base; + struct list_head *_desc, *n; + struct dma_async_tx_descriptor *tx; + + dev_dbg(to_dev(chan), "%s: phys_complete: %lx\n", + __func__, phys_complete); + list_for_each_safe(_desc, n, &ioat->used_desc) { + struct ioat_desc_sw *desc; + + prefetch(n); + desc = list_entry(_desc, typeof(*desc), node); + tx = &desc->txd; + /* + * Incoming DMA requests may use multiple descriptors, + * due to exceeding xfercap, perhaps. If so, only the + * last one will have a cookie, and require unmapping. + */ + dump_desc_dbg(ioat, desc); + if (tx->cookie) { + chan->completed_cookie = tx->cookie; + tx->cookie = 0; + ioat_dma_unmap(chan, tx->flags, desc->len, desc->hw); + if (tx->callback) { + tx->callback(tx->callback_param); + tx->callback = NULL; + } + } + + if (tx->phys != phys_complete) { + /* + * a completed entry, but not the last, so clean + * up if the client is done with the descriptor + */ + if (async_tx_test_ack(tx)) + list_move_tail(&desc->node, &ioat->free_desc); + } else { + /* + * last used desc. Do not remove, so we can + * append from it. + */ + + /* if nothing else is pending, cancel the + * completion timeout + */ + if (n == &ioat->used_desc) { + dev_dbg(to_dev(chan), + "%s cancel completion timeout\n", + __func__); + clear_bit(IOAT_COMPLETION_PENDING, &chan->state); + } + + /* TODO check status bits? */ + break; + } + } + + chan->last_completion = phys_complete; +} + +/** + * ioat1_cleanup - cleanup up finished descriptors + * @chan: ioat channel to be cleaned up + * + * To prevent lock contention we defer cleanup when the locks are + * contended with a terminal timeout that forces cleanup and catches + * completion notification errors. + */ +static void ioat1_cleanup(struct ioat_dma_chan *ioat) +{ + struct ioat_chan_common *chan = &ioat->base; + unsigned long phys_complete; + + prefetch(chan->completion); + + if (!spin_trylock_bh(&chan->cleanup_lock)) + return; + + if (!ioat_cleanup_preamble(chan, &phys_complete)) { + spin_unlock_bh(&chan->cleanup_lock); + return; + } + + if (!spin_trylock_bh(&ioat->desc_lock)) { + spin_unlock_bh(&chan->cleanup_lock); + return; + } + + __cleanup(ioat, phys_complete); + + spin_unlock_bh(&ioat->desc_lock); + spin_unlock_bh(&chan->cleanup_lock); +} + +static void ioat1_timer_event(unsigned long data) +{ + struct ioat_dma_chan *ioat = (void *) data; + struct ioat_chan_common *chan = &ioat->base; + + dev_dbg(to_dev(chan), "%s: state: %lx\n", __func__, chan->state); + + spin_lock_bh(&chan->cleanup_lock); + if (test_and_clear_bit(IOAT_RESET_PENDING, &chan->state)) { + struct ioat_desc_sw *desc; + + spin_lock_bh(&ioat->desc_lock); + + /* restart active descriptors */ + desc = to_ioat_desc(ioat->used_desc.prev); + ioat_set_chainaddr(ioat, desc->txd.phys); + ioat_start(chan); + + ioat->pending = 0; + set_bit(IOAT_COMPLETION_PENDING, &chan->state); + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + spin_unlock_bh(&ioat->desc_lock); + } else if (test_bit(IOAT_COMPLETION_PENDING, &chan->state)) { + unsigned long phys_complete; + + spin_lock_bh(&ioat->desc_lock); + /* if we haven't made progress and we have already + * acknowledged a pending completion once, then be more + * forceful with a restart + */ + if (ioat_cleanup_preamble(chan, &phys_complete)) + __cleanup(ioat, phys_complete); + else if (test_bit(IOAT_COMPLETION_ACK, &chan->state)) + ioat1_reset_channel(ioat); + else { + u64 status = ioat_chansts(chan); + + /* manually update the last completion address */ + if (ioat_chansts_to_addr(status) != 0) + *chan->completion = status; + + set_bit(IOAT_COMPLETION_ACK, &chan->state); + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + } + spin_unlock_bh(&ioat->desc_lock); + } + spin_unlock_bh(&chan->cleanup_lock); +} + +static enum dma_status +ioat1_dma_is_complete(struct dma_chan *c, dma_cookie_t cookie, + dma_cookie_t *done, dma_cookie_t *used) +{ + struct ioat_dma_chan *ioat = to_ioat_chan(c); + + if (ioat_is_complete(c, cookie, done, used) == DMA_SUCCESS) + return DMA_SUCCESS; + + ioat1_cleanup(ioat); + + return ioat_is_complete(c, cookie, done, used); +} + +static void ioat1_dma_start_null_desc(struct ioat_dma_chan *ioat) +{ + struct ioat_chan_common *chan = &ioat->base; + struct ioat_desc_sw *desc; + struct ioat_dma_descriptor *hw; + + spin_lock_bh(&ioat->desc_lock); + + desc = ioat1_dma_get_next_descriptor(ioat); + + if (!desc) { + dev_err(to_dev(chan), + "Unable to start null desc - get next desc failed\n"); + spin_unlock_bh(&ioat->desc_lock); + return; + } + + hw = desc->hw; + hw->ctl = 0; + hw->ctl_f.null = 1; + hw->ctl_f.int_en = 1; + hw->ctl_f.compl_write = 1; + /* set size to non-zero value (channel returns error when size is 0) */ + hw->size = NULL_DESC_BUFFER_SIZE; + hw->src_addr = 0; + hw->dst_addr = 0; + async_tx_ack(&desc->txd); + hw->next = 0; + list_add_tail(&desc->node, &ioat->used_desc); + dump_desc_dbg(ioat, desc); + + ioat_set_chainaddr(ioat, desc->txd.phys); + ioat_start(chan); + spin_unlock_bh(&ioat->desc_lock); +} + +/* + * Perform a IOAT transaction to verify the HW works. + */ +#define IOAT_TEST_SIZE 2000 + +static void __devinit ioat_dma_test_callback(void *dma_async_param) +{ + struct completion *cmp = dma_async_param; + + complete(cmp); +} + +/** + * ioat_dma_self_test - Perform a IOAT transaction to verify the HW works. + * @device: device to be tested + */ +static int __devinit ioat_dma_self_test(struct ioatdma_device *device) +{ + int i; + u8 *src; + u8 *dest; + struct dma_device *dma = &device->common; + struct device *dev = &device->pdev->dev; + struct dma_chan *dma_chan; + struct dma_async_tx_descriptor *tx; + dma_addr_t dma_dest, dma_src; + dma_cookie_t cookie; + int err = 0; + struct completion cmp; + unsigned long tmo; + unsigned long flags; + + src = kzalloc(sizeof(u8) * IOAT_TEST_SIZE, GFP_KERNEL); + if (!src) + return -ENOMEM; + dest = kzalloc(sizeof(u8) * IOAT_TEST_SIZE, GFP_KERNEL); + if (!dest) { + kfree(src); + return -ENOMEM; + } + + /* Fill in src buffer */ + for (i = 0; i < IOAT_TEST_SIZE; i++) + src[i] = (u8)i; + + /* Start copy, using first DMA channel */ + dma_chan = container_of(dma->channels.next, struct dma_chan, + device_node); + if (dma->device_alloc_chan_resources(dma_chan) < 1) { + dev_err(dev, "selftest cannot allocate chan resource\n"); + err = -ENODEV; + goto out; + } + + dma_src = dma_map_single(dev, src, IOAT_TEST_SIZE, DMA_TO_DEVICE); + dma_dest = dma_map_single(dev, dest, IOAT_TEST_SIZE, DMA_FROM_DEVICE); + flags = DMA_COMPL_SRC_UNMAP_SINGLE | DMA_COMPL_DEST_UNMAP_SINGLE | + DMA_PREP_INTERRUPT; + tx = device->common.device_prep_dma_memcpy(dma_chan, dma_dest, dma_src, + IOAT_TEST_SIZE, flags); + if (!tx) { + dev_err(dev, "Self-test prep failed, disabling\n"); + err = -ENODEV; + goto free_resources; + } + + async_tx_ack(tx); + init_completion(&cmp); + tx->callback = ioat_dma_test_callback; + tx->callback_param = &cmp; + cookie = tx->tx_submit(tx); + if (cookie < 0) { + dev_err(dev, "Self-test setup failed, disabling\n"); + err = -ENODEV; + goto free_resources; + } + dma->device_issue_pending(dma_chan); + + tmo = wait_for_completion_timeout(&cmp, msecs_to_jiffies(3000)); + + if (tmo == 0 || + dma->device_is_tx_complete(dma_chan, cookie, NULL, NULL) + != DMA_SUCCESS) { + dev_err(dev, "Self-test copy timed out, disabling\n"); + err = -ENODEV; + goto free_resources; + } + if (memcmp(src, dest, IOAT_TEST_SIZE)) { + dev_err(dev, "Self-test copy failed compare, disabling\n"); + err = -ENODEV; + goto free_resources; + } + +free_resources: + dma->device_free_chan_resources(dma_chan); +out: + kfree(src); + kfree(dest); + return err; +} + +static char ioat_interrupt_style[32] = "msix"; +module_param_string(ioat_interrupt_style, ioat_interrupt_style, + sizeof(ioat_interrupt_style), 0644); +MODULE_PARM_DESC(ioat_interrupt_style, + "set ioat interrupt style: msix (default), " + "msix-single-vector, msi, intx)"); + +/** + * ioat_dma_setup_interrupts - setup interrupt handler + * @device: ioat device + */ +static int ioat_dma_setup_interrupts(struct ioatdma_device *device) +{ + struct ioat_chan_common *chan; + struct pci_dev *pdev = device->pdev; + struct device *dev = &pdev->dev; + struct msix_entry *msix; + int i, j, msixcnt; + int err = -EINVAL; + u8 intrctrl = 0; + + if (!strcmp(ioat_interrupt_style, "msix")) + goto msix; + if (!strcmp(ioat_interrupt_style, "msix-single-vector")) + goto msix_single_vector; + if (!strcmp(ioat_interrupt_style, "msi")) + goto msi; + if (!strcmp(ioat_interrupt_style, "intx")) + goto intx; + dev_err(dev, "invalid ioat_interrupt_style %s\n", ioat_interrupt_style); + goto err_no_irq; + +msix: + /* The number of MSI-X vectors should equal the number of channels */ + msixcnt = device->common.chancnt; + for (i = 0; i < msixcnt; i++) + device->msix_entries[i].entry = i; + + err = pci_enable_msix(pdev, device->msix_entries, msixcnt); + if (err < 0) + goto msi; + if (err > 0) + goto msix_single_vector; + + for (i = 0; i < msixcnt; i++) { + msix = &device->msix_entries[i]; + chan = ioat_chan_by_index(device, i); + err = devm_request_irq(dev, msix->vector, + ioat_dma_do_interrupt_msix, 0, + "ioat-msix", chan); + if (err) { + for (j = 0; j < i; j++) { + msix = &device->msix_entries[j]; + chan = ioat_chan_by_index(device, j); + devm_free_irq(dev, msix->vector, chan); + } + goto msix_single_vector; + } + } + intrctrl |= IOAT_INTRCTRL_MSIX_VECTOR_CONTROL; + goto done; + +msix_single_vector: + msix = &device->msix_entries[0]; + msix->entry = 0; + err = pci_enable_msix(pdev, device->msix_entries, 1); + if (err) + goto msi; + + err = devm_request_irq(dev, msix->vector, ioat_dma_do_interrupt, 0, + "ioat-msix", device); + if (err) { + pci_disable_msix(pdev); + goto msi; + } + goto done; + +msi: + err = pci_enable_msi(pdev); + if (err) + goto intx; + + err = devm_request_irq(dev, pdev->irq, ioat_dma_do_interrupt, 0, + "ioat-msi", device); + if (err) { + pci_disable_msi(pdev); + goto intx; + } + goto done; + +intx: + err = devm_request_irq(dev, pdev->irq, ioat_dma_do_interrupt, + IRQF_SHARED, "ioat-intx", device); + if (err) + goto err_no_irq; + +done: + if (device->intr_quirk) + device->intr_quirk(device); + intrctrl |= IOAT_INTRCTRL_MASTER_INT_EN; + writeb(intrctrl, device->reg_base + IOAT_INTRCTRL_OFFSET); + return 0; + +err_no_irq: + /* Disable all interrupt generation */ + writeb(0, device->reg_base + IOAT_INTRCTRL_OFFSET); + dev_err(dev, "no usable interrupts\n"); + return err; +} + +static void ioat_disable_interrupts(struct ioatdma_device *device) +{ + /* Disable all interrupt generation */ + writeb(0, device->reg_base + IOAT_INTRCTRL_OFFSET); +} + +int __devinit ioat_probe(struct ioatdma_device *device) +{ + int err = -ENODEV; + struct dma_device *dma = &device->common; + struct pci_dev *pdev = device->pdev; + struct device *dev = &pdev->dev; + + /* DMA coherent memory pool for DMA descriptor allocations */ + device->dma_pool = pci_pool_create("dma_desc_pool", pdev, + sizeof(struct ioat_dma_descriptor), + 64, 0); + if (!device->dma_pool) { + err = -ENOMEM; + goto err_dma_pool; + } + + device->completion_pool = pci_pool_create("completion_pool", pdev, + sizeof(u64), SMP_CACHE_BYTES, + SMP_CACHE_BYTES); + + if (!device->completion_pool) { + err = -ENOMEM; + goto err_completion_pool; + } + + device->enumerate_channels(device); + + dma_cap_set(DMA_MEMCPY, dma->cap_mask); + dma->dev = &pdev->dev; + + dev_err(dev, "Intel(R) I/OAT DMA Engine found," + " %d channels, device version 0x%02x, driver version %s\n", + dma->chancnt, device->version, IOAT_DMA_VERSION); + + if (!dma->chancnt) { + dev_err(dev, "Intel(R) I/OAT DMA Engine problem found: " + "zero channels detected\n"); + goto err_setup_interrupts; + } + + err = ioat_dma_setup_interrupts(device); + if (err) + goto err_setup_interrupts; + + err = ioat_dma_self_test(device); + if (err) + goto err_self_test; + + return 0; + +err_self_test: + ioat_disable_interrupts(device); +err_setup_interrupts: + pci_pool_destroy(device->completion_pool); +err_completion_pool: + pci_pool_destroy(device->dma_pool); +err_dma_pool: + return err; +} + +int __devinit ioat_register(struct ioatdma_device *device) +{ + int err = dma_async_device_register(&device->common); + + if (err) { + ioat_disable_interrupts(device); + pci_pool_destroy(device->completion_pool); + pci_pool_destroy(device->dma_pool); + } + + return err; +} + +/* ioat1_intr_quirk - fix up dma ctrl register to enable / disable msi */ +static void ioat1_intr_quirk(struct ioatdma_device *device) +{ + struct pci_dev *pdev = device->pdev; + u32 dmactrl; + + pci_read_config_dword(pdev, IOAT_PCI_DMACTRL_OFFSET, &dmactrl); + if (pdev->msi_enabled) + dmactrl |= IOAT_PCI_DMACTRL_MSI_EN; + else + dmactrl &= ~IOAT_PCI_DMACTRL_MSI_EN; + pci_write_config_dword(pdev, IOAT_PCI_DMACTRL_OFFSET, dmactrl); +} + +int __devinit ioat1_dma_probe(struct ioatdma_device *device, int dca) +{ + struct pci_dev *pdev = device->pdev; + struct dma_device *dma; + int err; + + device->intr_quirk = ioat1_intr_quirk; + device->enumerate_channels = ioat1_enumerate_channels; + dma = &device->common; + dma->device_prep_dma_memcpy = ioat1_dma_prep_memcpy; + dma->device_issue_pending = ioat1_dma_memcpy_issue_pending; + dma->device_alloc_chan_resources = ioat1_dma_alloc_chan_resources; + dma->device_free_chan_resources = ioat1_dma_free_chan_resources; + dma->device_is_tx_complete = ioat1_dma_is_complete; + + err = ioat_probe(device); + if (err) + return err; + ioat_set_tcp_copy_break(4096); + err = ioat_register(device); + if (err) + return err; + if (dca) + device->dca = ioat_dca_init(pdev, device->reg_base); + + return err; +} + +void __devexit ioat_dma_remove(struct ioatdma_device *device) +{ + struct dma_device *dma = &device->common; + + ioat_disable_interrupts(device); + + dma_async_device_unregister(dma); + + pci_pool_destroy(device->dma_pool); + pci_pool_destroy(device->completion_pool); + + INIT_LIST_HEAD(&dma->channels); +} diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h new file mode 100644 index 00000000000..d9d6a7e3cd7 --- /dev/null +++ b/drivers/dma/ioat/dma.h @@ -0,0 +1,306 @@ +/* + * Copyright(c) 2004 - 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ +#ifndef IOATDMA_H +#define IOATDMA_H + +#include <linux/dmaengine.h> +#include "hw.h" +#include "registers.h" +#include <linux/init.h> +#include <linux/dmapool.h> +#include <linux/cache.h> +#include <linux/pci_ids.h> +#include <net/tcp.h> + +#define IOAT_DMA_VERSION "3.64" + +#define IOAT_LOW_COMPLETION_MASK 0xffffffc0 +#define IOAT_DMA_DCA_ANY_CPU ~0 + +#define to_ioatdma_device(dev) container_of(dev, struct ioatdma_device, common) +#define to_ioat_desc(lh) container_of(lh, struct ioat_desc_sw, node) +#define tx_to_ioat_desc(tx) container_of(tx, struct ioat_desc_sw, txd) +#define to_dev(ioat_chan) (&(ioat_chan)->device->pdev->dev) + +#define chan_num(ch) ((int)((ch)->reg_base - (ch)->device->reg_base) / 0x80) + +/* + * workaround for IOAT ver.3.0 null descriptor issue + * (channel returns error when size is 0) + */ +#define NULL_DESC_BUFFER_SIZE 1 + +/** + * struct ioatdma_device - internal representation of a IOAT device + * @pdev: PCI-Express device + * @reg_base: MMIO register space base address + * @dma_pool: for allocating DMA descriptors + * @common: embedded struct dma_device + * @version: version of ioatdma device + * @msix_entries: irq handlers + * @idx: per channel data + * @dca: direct cache access context + * @intr_quirk: interrupt setup quirk (for ioat_v1 devices) + * @enumerate_channels: hw version specific channel enumeration + */ + +struct ioatdma_device { + struct pci_dev *pdev; + void __iomem *reg_base; + struct pci_pool *dma_pool; + struct pci_pool *completion_pool; + struct dma_device common; + u8 version; + struct msix_entry msix_entries[4]; + struct ioat_chan_common *idx[4]; + struct dca_provider *dca; + void (*intr_quirk)(struct ioatdma_device *device); + int (*enumerate_channels)(struct ioatdma_device *device); +}; + +struct ioat_chan_common { + struct dma_chan common; + void __iomem *reg_base; + unsigned long last_completion; + spinlock_t cleanup_lock; + dma_cookie_t completed_cookie; + unsigned long state; + #define IOAT_COMPLETION_PENDING 0 + #define IOAT_COMPLETION_ACK 1 + #define IOAT_RESET_PENDING 2 + struct timer_list timer; + #define COMPLETION_TIMEOUT msecs_to_jiffies(100) + #define IDLE_TIMEOUT msecs_to_jiffies(2000) + #define RESET_DELAY msecs_to_jiffies(100) + struct ioatdma_device *device; + dma_addr_t completion_dma; + u64 *completion; + struct tasklet_struct cleanup_task; +}; + + +/** + * struct ioat_dma_chan - internal representation of a DMA channel + */ +struct ioat_dma_chan { + struct ioat_chan_common base; + + size_t xfercap; /* XFERCAP register value expanded out */ + + spinlock_t desc_lock; + struct list_head free_desc; + struct list_head used_desc; + + int pending; + u16 desccount; +}; + +static inline struct ioat_chan_common *to_chan_common(struct dma_chan *c) +{ + return container_of(c, struct ioat_chan_common, common); +} + +static inline struct ioat_dma_chan *to_ioat_chan(struct dma_chan *c) +{ + struct ioat_chan_common *chan = to_chan_common(c); + + return container_of(chan, struct ioat_dma_chan, base); +} + +/** + * ioat_is_complete - poll the status of an ioat transaction + * @c: channel handle + * @cookie: transaction identifier + * @done: if set, updated with last completed transaction + * @used: if set, updated with last used transaction + */ +static inline enum dma_status +ioat_is_complete(struct dma_chan *c, dma_cookie_t cookie, + dma_cookie_t *done, dma_cookie_t *used) +{ + struct ioat_chan_common *chan = to_chan_common(c); + dma_cookie_t last_used; + dma_cookie_t last_complete; + + last_used = c->cookie; + last_complete = chan->completed_cookie; + + if (done) + *done = last_complete; + if (used) + *used = last_used; + + return dma_async_is_complete(cookie, last_complete, last_used); +} + +/* wrapper around hardware descriptor format + additional software fields */ + +/** + * struct ioat_desc_sw - wrapper around hardware descriptor + * @hw: hardware DMA descriptor + * @node: this descriptor will either be on the free list, + * or attached to a transaction list (async_tx.tx_list) + * @txd: the generic software descriptor for all engines + * @id: identifier for debug + */ +struct ioat_desc_sw { + struct ioat_dma_descriptor *hw; + struct list_head node; + size_t len; + struct dma_async_tx_descriptor txd; + #ifdef DEBUG + int id; + #endif +}; + +#ifdef DEBUG +#define set_desc_id(desc, i) ((desc)->id = (i)) +#define desc_id(desc) ((desc)->id) +#else +#define set_desc_id(desc, i) +#define desc_id(desc) (0) +#endif + +static inline void +__dump_desc_dbg(struct ioat_chan_common *chan, struct ioat_dma_descriptor *hw, + struct dma_async_tx_descriptor *tx, int id) +{ + struct device *dev = to_dev(chan); + + dev_dbg(dev, "desc[%d]: (%#llx->%#llx) cookie: %d flags: %#x" + " ctl: %#x (op: %d int_en: %d compl: %d)\n", id, + (unsigned long long) tx->phys, + (unsigned long long) hw->next, tx->cookie, tx->flags, + hw->ctl, hw->ctl_f.op, hw->ctl_f.int_en, hw->ctl_f.compl_write); +} + +#define dump_desc_dbg(c, d) \ + ({ if (d) __dump_desc_dbg(&c->base, d->hw, &d->txd, desc_id(d)); 0; }) + +static inline void ioat_set_tcp_copy_break(unsigned long copybreak) +{ + #ifdef CONFIG_NET_DMA + sysctl_tcp_dma_copybreak = copybreak; + #endif +} + +static inline struct ioat_chan_common * +ioat_chan_by_index(struct ioatdma_device *device, int index) +{ + return device->idx[index]; +} + +static inline u64 ioat_chansts(struct ioat_chan_common *chan) +{ + u8 ver = chan->device->version; + u64 status; + u32 status_lo; + + /* We need to read the low address first as this causes the + * chipset to latch the upper bits for the subsequent read + */ + status_lo = readl(chan->reg_base + IOAT_CHANSTS_OFFSET_LOW(ver)); + status = readl(chan->reg_base + IOAT_CHANSTS_OFFSET_HIGH(ver)); + status <<= 32; + status |= status_lo; + + return status; +} + +static inline void ioat_start(struct ioat_chan_common *chan) +{ + u8 ver = chan->device->version; + + writeb(IOAT_CHANCMD_START, chan->reg_base + IOAT_CHANCMD_OFFSET(ver)); +} + +static inline u64 ioat_chansts_to_addr(u64 status) +{ + return status & IOAT_CHANSTS_COMPLETED_DESCRIPTOR_ADDR; +} + +static inline u32 ioat_chanerr(struct ioat_chan_common *chan) +{ + return readl(chan->reg_base + IOAT_CHANERR_OFFSET); +} + +static inline void ioat_suspend(struct ioat_chan_common *chan) +{ + u8 ver = chan->device->version; + + writeb(IOAT_CHANCMD_SUSPEND, chan->reg_base + IOAT_CHANCMD_OFFSET(ver)); +} + +static inline void ioat_set_chainaddr(struct ioat_dma_chan *ioat, u64 addr) +{ + struct ioat_chan_common *chan = &ioat->base; + + writel(addr & 0x00000000FFFFFFFF, + chan->reg_base + IOAT1_CHAINADDR_OFFSET_LOW); + writel(addr >> 32, + chan->reg_base + IOAT1_CHAINADDR_OFFSET_HIGH); +} + +static inline bool is_ioat_active(unsigned long status) +{ + return ((status & IOAT_CHANSTS_STATUS) == IOAT_CHANSTS_ACTIVE); +} + +static inline bool is_ioat_idle(unsigned long status) +{ + return ((status & IOAT_CHANSTS_STATUS) == IOAT_CHANSTS_DONE); +} + +static inline bool is_ioat_halted(unsigned long status) +{ + return ((status & IOAT_CHANSTS_STATUS) == IOAT_CHANSTS_HALTED); +} + +static inline bool is_ioat_suspended(unsigned long status) +{ + return ((status & IOAT_CHANSTS_STATUS) == IOAT_CHANSTS_SUSPENDED); +} + +/* channel was fatally programmed */ +static inline bool is_ioat_bug(unsigned long err) +{ + return !!(err & (IOAT_CHANERR_SRC_ADDR_ERR|IOAT_CHANERR_DEST_ADDR_ERR| + IOAT_CHANERR_NEXT_ADDR_ERR|IOAT_CHANERR_CONTROL_ERR| + IOAT_CHANERR_LENGTH_ERR)); +} + +int __devinit ioat_probe(struct ioatdma_device *device); +int __devinit ioat_register(struct ioatdma_device *device); +int __devinit ioat1_dma_probe(struct ioatdma_device *dev, int dca); +void __devexit ioat_dma_remove(struct ioatdma_device *device); +struct dca_provider * __devinit ioat_dca_init(struct pci_dev *pdev, + void __iomem *iobase); +unsigned long ioat_get_current_completion(struct ioat_chan_common *chan); +void ioat_init_channel(struct ioatdma_device *device, + struct ioat_chan_common *chan, int idx, + void (*timer_fn)(unsigned long), + void (*tasklet)(unsigned long), + unsigned long ioat); +void ioat_dma_unmap(struct ioat_chan_common *chan, enum dma_ctrl_flags flags, + size_t len, struct ioat_dma_descriptor *hw); +bool ioat_cleanup_preamble(struct ioat_chan_common *chan, + unsigned long *phys_complete); +#endif /* IOATDMA_H */ diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c new file mode 100644 index 00000000000..460b7730133 --- /dev/null +++ b/drivers/dma/ioat/dma_v2.c @@ -0,0 +1,881 @@ +/* + * Intel I/OAT DMA Linux driver + * Copyright(c) 2004 - 2009 Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + * + * The full GNU General Public License is included in this distribution in + * the file called "COPYING". + * + */ + +/* + * This driver supports an Intel I/OAT DMA engine (versions >= 2), which + * does asynchronous data movement and checksumming operations. + */ + +#include <linux/init.h> +#include <linux/module.h> +#include <linux/pci.h> +#include <linux/interrupt.h> +#include <linux/dmaengine.h> +#include <linux/delay.h> +#include <linux/dma-mapping.h> +#include <linux/workqueue.h> +#include <linux/i7300_idle.h> +#include "dma.h" +#include "dma_v2.h" +#include "registers.h" +#include "hw.h" + +static int ioat_ring_alloc_order = 8; +module_param(ioat_ring_alloc_order, int, 0644); +MODULE_PARM_DESC(ioat_ring_alloc_order, + "ioat2+: allocate 2^n descriptors per channel (default: n=8)"); +static int ioat_ring_max_alloc_order = IOAT_MAX_ORDER; +module_param(ioat_ring_max_alloc_order, int, 0644); +MODULE_PARM_DESC(ioat_ring_max_alloc_order, + "ioat2+: upper limit for dynamic ring resizing (default: n=16)"); + +static void __ioat2_issue_pending(struct ioat2_dma_chan *ioat) +{ + void * __iomem reg_base = ioat->base.reg_base; + + ioat->pending = 0; + ioat->dmacount += ioat2_ring_pending(ioat);; + ioat->issued = ioat->head; + /* make descriptor updates globally visible before notifying channel */ + wmb(); + writew(ioat->dmacount, reg_base + IOAT_CHAN_DMACOUNT_OFFSET); + dev_dbg(to_dev(&ioat->base), + "%s: head: %#x tail: %#x issued: %#x count: %#x\n", + __func__, ioat->head, ioat->tail, ioat->issued, ioat->dmacount); +} + +static void ioat2_issue_pending(struct dma_chan *chan) +{ + struct ioat2_dma_chan *ioat = to_ioat2_chan(chan); + + spin_lock_bh(&ioat->ring_lock); + if (ioat->pending == 1) + __ioat2_issue_pending(ioat); + spin_unlock_bh(&ioat->ring_lock); +} + +/** + * ioat2_update_pending - log pending descriptors + * @ioat: ioat2+ channel + * + * set pending to '1' unless pending is already set to '2', pending == 2 + * indicates that submission is temporarily blocked due to an in-flight + * reset. If we are already above the ioat_pending_level threshold then + * just issue pending. + * + * called with ring_lock held + */ +static void ioat2_update_pending(struct ioat2_dma_chan *ioat) +{ + if (unlikely(ioat->pending == 2)) + return; + else if (ioat2_ring_pending(ioat) > ioat_pending_level) + __ioat2_issue_pending(ioat); + else + ioat->pending = 1; +} + +static void __ioat2_start_null_desc(struct ioat2_dma_chan *ioat) +{ + struct ioat_ring_ent *desc; + struct ioat_dma_descriptor *hw; + int idx; + + if (ioat2_ring_space(ioat) < 1) { + dev_err(to_dev(&ioat->base), + "Unable to start null desc - ring full\n"); + return; + } + + dev_dbg(to_dev(&ioat->base), "%s: head: %#x tail: %#x issued: %#x\n", + __func__, ioat->head, ioat->tail, ioat->issued); + idx = ioat2_desc_alloc(ioat, 1); + desc = ioat2_get_ring_ent(ioat, idx); + + hw = desc->hw; + hw->ctl = 0; + hw->ctl_f.null = 1; + hw->ctl_f.int_en = 1; + hw->ctl_f.compl_write = 1; + /* set size to non-zero value (channel returns error when size is 0) */ + hw->size = NULL_DESC_BUFFER_SIZE; + hw->src_addr = 0; + hw->dst_addr = 0; + async_tx_ack(&desc->txd); + ioat2_set_chainaddr(ioat, desc->txd.phys); + dump_desc_dbg(ioat, desc); + __ioat2_issue_pending(ioat); +} + +static void ioat2_start_null_desc(struct ioat2_dma_chan *ioat) +{ + spin_lock_bh(&ioat->ring_lock); + __ioat2_start_null_desc(ioat); + spin_unlock_bh(&ioat->ring_lock); +} + +static void __cleanup(struct ioat2_dma_chan *ioat, unsigned long phys_complete) +{ + struct ioat_chan_common *chan = &ioat->base; + struct dma_async_tx_descriptor *tx; + struct ioat_ring_ent *desc; + bool seen_current = false; + u16 active; + int i; + + dev_dbg(to_dev(chan), "%s: head: %#x tail: %#x issued: %#x\n", + __func__, ioat->head, ioat->tail, ioat->issued); + + active = ioat2_ring_active(ioat); + for (i = 0; i < active && !seen_current; i++) { + prefetch(ioat2_get_ring_ent(ioat, ioat->tail + i + 1)); + desc = ioat2_get_ring_ent(ioat, ioat->tail + i); + tx = &desc->txd; + dump_desc_dbg(ioat, desc); + if (tx->cookie) { + ioat_dma_unmap(chan, tx->flags, desc->len, desc->hw); + chan->completed_cookie = tx->cookie; + tx->cookie = 0; + if (tx->callback) { + tx->callback(tx->callback_param); + tx->callback = NULL; + } + } + + if (tx->phys == phys_complete) + seen_current = true; + } + ioat->tail += i; + BUG_ON(!seen_current); /* no active descs have written a completion? */ + + chan->last_completion = phys_complete; + if (ioat->head == ioat->tail) { + dev_dbg(to_dev(chan), "%s: cancel completion timeout\n", + __func__); + clear_bit(IOAT_COMPLETION_PENDING, &chan->state); + mod_timer(&chan->timer, jiffies + IDLE_TIMEOUT); + } +} + +/** + * ioat2_cleanup - clean finished descriptors (advance tail pointer) + * @chan: ioat channel to be cleaned up + */ +static void ioat2_cleanup(struct ioat2_dma_chan *ioat) +{ + struct ioat_chan_common *chan = &ioat->base; + unsigned long phys_complete; + + prefetch(chan->completion); + + if (!spin_trylock_bh(&chan->cleanup_lock)) + return; + + if (!ioat_cleanup_preamble(chan, &phys_complete)) { + spin_unlock_bh(&chan->cleanup_lock); + return; + } + + if (!spin_trylock_bh(&ioat->ring_lock)) { + spin_unlock_bh(&chan->cleanup_lock); + return; + } + + __cleanup(ioat, phys_complete); + + spin_unlock_bh(&ioat->ring_lock); + spin_unlock_bh(&chan->cleanup_lock); +} + +static void ioat2_cleanup_tasklet(unsigned long data) +{ + struct ioat2_dma_chan *ioat = (void *) data; + + ioat2_cleanup(ioat); + writew(IOAT_CHANCTRL_RUN, ioat->base.reg_base + IOAT_CHANCTRL_OFFSET); +} + +static void __restart_chan(struct ioat2_dma_chan *ioat) +{ + struct ioat_chan_common *chan = &ioat->base; + + /* set the tail to be re-issued */ + ioat->issued = ioat->tail; + ioat->dmacount = 0; + set_bit(IOAT_COMPLETION_PENDING, &chan->state); + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + + dev_dbg(to_dev(chan), + "%s: head: %#x tail: %#x issued: %#x count: %#x\n", + __func__, ioat->head, ioat->tail, ioat->issued, ioat->dmacount); + + if (ioat2_ring_pending(ioat)) { + struct ioat_ring_ent *desc; + + desc = ioat2_get_ring_ent(ioat, ioat->tail); + ioat2_set_chainaddr(ioat, desc->txd.phys); + __ioat2_issue_pending(ioat); + } else + __ioat2_start_null_desc(ioat); +} + +static void ioat2_restart_channel(struct ioat2_dma_chan *ioat) +{ + struct ioat_chan_common *chan = &ioat->base; + unsigned long phys_complete; + u32 status; + + status = ioat_chansts(chan); + if (is_ioat_active(status) || is_ioat_idle(status)) + ioat_suspend(chan); + while (is_ioat_active(status) || is_ioat_idle(status)) { + status = ioat_chansts(chan); + cpu_relax(); + } + + if (ioat_cleanup_preamble(chan, &phys_complete)) + __cleanup(ioat, phys_complete); + + __restart_chan(ioat); +} + +static bool reshape_ring(struct ioat2_dma_chan *ioat, int order); + +static void ioat2_timer_event(unsigned long data) +{ + struct ioat2_dma_chan *ioat = (void *) data; + struct ioat_chan_common *chan = &ioat->base; + + spin_lock_bh(&chan->cleanup_lock); + if (test_bit(IOAT_COMPLETION_PENDING, &chan->state)) { + unsigned long phys_complete; + u64 status; + + spin_lock_bh(&ioat->ring_lock); + status = ioat_chansts(chan); + + /* when halted due to errors check for channel + * programming errors before advancing the completion state + */ + if (is_ioat_halted(status)) { + u32 chanerr; + + chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET); + BUG_ON(is_ioat_bug(chanerr)); + } + + /* if we haven't made progress and we have already + * acknowledged a pending completion once, then be more + * forceful with a restart + */ + if (ioat_cleanup_preamble(chan, &phys_complete)) + __cleanup(ioat, phys_complete); + else if (test_bit(IOAT_COMPLETION_ACK, &chan->state)) + ioat2_restart_channel(ioat); + else { + set_bit(IOAT_COMPLETION_ACK, &chan->state); + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + } + spin_unlock_bh(&ioat->ring_lock); + } else { + u16 active; + + /* if the ring is idle, empty, and oversized try to step + * down the size + */ + spin_lock_bh(&ioat->ring_lock); + active = ioat2_ring_active(ioat); + if (active == 0 && ioat->alloc_order > ioat_get_alloc_order()) + reshape_ring(ioat, ioat->alloc_order-1); + spin_unlock_bh(&ioat->ring_lock); + + /* keep shrinking until we get back to our minimum + * default size + */ + if (ioat->alloc_order > ioat_get_alloc_order()) + mod_timer(&chan->timer, jiffies + IDLE_TIMEOUT); + } + spin_unlock_bh(&chan->cleanup_lock); +} + +/** + * ioat2_enumerate_channels - find and initialize the device's channels + * @device: the device to be enumerated + */ +static int ioat2_enumerate_channels(struct ioatdma_device *device) +{ + struct ioat2_dma_chan *ioat; + struct device *dev = &device->pdev->dev; + struct dma_device *dma = &device->common; + u8 xfercap_log; + int i; + + INIT_LIST_HEAD(&dma->channels); + dma->chancnt = readb(device->reg_base + IOAT_CHANCNT_OFFSET); + dma->chancnt &= 0x1f; /* bits [4:0] valid */ + if (dma->chancnt > ARRAY_SIZE(device->idx)) { + dev_warn(dev, "(%d) exceeds max supported channels (%zu)\n", + dma->chancnt, ARRAY_SIZE(device->idx)); + dma->chancnt = ARRAY_SIZE(device->idx); + } + xfercap_log = readb(device->reg_base + IOAT_XFERCAP_OFFSET); + xfercap_log &= 0x1f; /* bits [4:0] valid */ + if (xfercap_log == 0) + return 0; + dev_dbg(dev, "%s: xfercap = %d\n", __func__, 1 << xfercap_log); + + /* FIXME which i/oat version is i7300? */ +#ifdef CONFIG_I7300_IDLE_IOAT_CHANNEL + if (i7300_idle_platform_probe(NULL, NULL, 1) == 0) + dma->chancnt--; +#endif + for (i = 0; i < dma->chancnt; i++) { + ioat = devm_kzalloc(dev, sizeof(*ioat), GFP_KERNEL); + if (!ioat) + break; + + ioat_init_channel(device, &ioat->base, i, + ioat2_timer_event, + ioat2_cleanup_tasklet, + (unsigned long) ioat); + ioat->xfercap_log = xfercap_log; + spin_lock_init(&ioat->ring_lock); + } + dma->chancnt = i; + return i; +} + +static dma_cookie_t ioat2_tx_submit_unlock(struct dma_async_tx_descriptor *tx) +{ + struct dma_chan *c = tx->chan; + struct ioat2_dma_chan *ioat = to_ioat2_chan(c); + struct ioat_chan_common *chan = &ioat->base; + dma_cookie_t cookie = c->cookie; + + cookie++; + if (cookie < 0) + cookie = 1; + tx->cookie = cookie; + c->cookie = cookie; + dev_dbg(to_dev(&ioat->base), "%s: cookie: %d\n", __func__, cookie); + + if (!test_and_set_bit(IOAT_COMPLETION_PENDING, &chan->state)) + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + ioat2_update_pending(ioat); + spin_unlock_bh(&ioat->ring_lock); + + return cookie; +} + +static struct ioat_ring_ent *ioat2_alloc_ring_ent(struct dma_chan *chan, gfp_t flags) +{ + struct ioat_dma_descriptor *hw; + struct ioat_ring_ent *desc; + struct ioatdma_device *dma; + dma_addr_t phys; + + dma = to_ioatdma_device(chan->device); + hw = pci_pool_alloc(dma->dma_pool, flags, &phys); + if (!hw) + return NULL; + memset(hw, 0, sizeof(*hw)); + + desc = kzalloc(sizeof(*desc), flags); + if (!desc) { + pci_pool_free(dma->dma_pool, hw, phys); + return NULL; + } + + dma_async_tx_descriptor_init(&desc->txd, chan); + desc->txd.tx_submit = ioat2_tx_submit_unlock; + desc->hw = hw; + desc->txd.phys = phys; + return desc; +} + +static void ioat2_free_ring_ent(struct ioat_ring_ent *desc, struct dma_chan *chan) +{ + struct ioatdma_device *dma; + + dma = to_ioatdma_device(chan->device); + pci_pool_free(dma->dma_pool, desc->hw, desc->txd.phys); + kfree(desc); +} + +static struct ioat_ring_ent **ioat2_alloc_ring(struct dma_chan *c, int order, gfp_t flags) +{ + struct ioat_ring_ent **ring; + int descs = 1 << order; + int i; + + if (order > ioat_get_max_alloc_order()) + return NULL; + + /* allocate the array to hold the software ring */ + ring = kcalloc(descs, sizeof(*ring), flags); + if (!ring) + return NULL; + for (i = 0; i < descs; i++) { + ring[i] = ioat2_alloc_ring_ent(c, flags); + if (!ring[i]) { + while (i--) + ioat2_free_ring_ent(ring[i], c); + kfree(ring); + return NULL; + } + set_desc_id(ring[i], i); + } + + /* link descs */ + for (i = 0; i < descs-1; i++) { + struct ioat_ring_ent *next = ring[i+1]; + struct ioat_dma_descriptor *hw = ring[i]->hw; + + hw->next = next->txd.phys; + } + ring[i]->hw->next = ring[0]->txd.phys; + + return ring; +} + +/* ioat2_alloc_chan_resources - allocate/initialize ioat2 descriptor ring + * @chan: channel to be initialized + */ +static int ioat2_alloc_chan_resources(struct dma_chan *c) +{ + struct ioat2_dma_chan *ioat = to_ioat2_chan(c); + struct ioat_chan_common *chan = &ioat->base; + struct ioat_ring_ent **ring; + u32 chanerr; + int order; + + /* have we already been set up? */ + if (ioat->ring) + return 1 << ioat->alloc_order; + + /* Setup register to interrupt and write completion status on error */ + writew(IOAT_CHANCTRL_RUN, chan->reg_base + IOAT_CHANCTRL_OFFSET); + + chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET); + if (chanerr) { + dev_err(to_dev(chan), "CHANERR = %x, clearing\n", chanerr); + writel(chanerr, chan->reg_base + IOAT_CHANERR_OFFSET); + } + + /* allocate a completion writeback area */ + /* doing 2 32bit writes to mmio since 1 64b write doesn't work */ + chan->completion = pci_pool_alloc(chan->device->completion_pool, + GFP_KERNEL, &chan->completion_dma); + if (!chan->completion) + return -ENOMEM; + + memset(chan->completion, 0, sizeof(*chan->completion)); + writel(((u64) chan->completion_dma) & 0x00000000FFFFFFFF, + chan->reg_base + IOAT_CHANCMP_OFFSET_LOW); + writel(((u64) chan->completion_dma) >> 32, + chan->reg_base + IOAT_CHANCMP_OFFSET_HIGH); + + order = ioat_get_alloc_order(); + ring = ioat2_alloc_ring(c, order, GFP_KERNEL); + if (!ring) + return -ENOMEM; + + spin_lock_bh(&ioat->ring_lock); + ioat->ring = ring; + ioat->head = 0; + ioat->issued = 0; + ioat->tail = 0; + ioat->pending = 0; + ioat->alloc_order = order; + spin_unlock_bh(&ioat->ring_lock); + + tasklet_enable(&chan->cleanup_task); + ioat2_start_null_desc(ioat); + + return 1 << ioat->alloc_order; +} + +static bool reshape_ring(struct ioat2_dma_chan *ioat, int order) +{ + /* reshape differs from normal ring allocation in that we want + * to allocate a new software ring while only + * extending/truncating the hardware ring + */ + struct ioat_chan_common *chan = &ioat->base; + struct dma_chan *c = &chan->common; + const u16 curr_size = ioat2_ring_mask(ioat) + 1; + const u16 active = ioat2_ring_active(ioat); + const u16 new_size = 1 << order; + struct ioat_ring_ent **ring; + u16 i; + + if (order > ioat_get_max_alloc_order()) + return false; + + /* double check that we have at least 1 free descriptor */ + if (active == curr_size) + return false; + + /* when shrinking, verify that we can hold the current active + * set in the new ring + */ + if (active >= new_size) + return false; + + /* allocate the array to hold the software ring */ + ring = kcalloc(new_size, sizeof(*ring), GFP_NOWAIT); + if (!ring) + return false; + + /* allocate/trim descriptors as needed */ + if (new_size > curr_size) { + /* copy current descriptors to the new ring */ + for (i = 0; i < curr_size; i++) { + u16 curr_idx = (ioat->tail+i) & (curr_size-1); + u16 new_idx = (ioat->tail+i) & (new_size-1); + + ring[new_idx] = ioat->ring[curr_idx]; + set_desc_id(ring[new_idx], new_idx); + } + + /* add new descriptors to the ring */ + for (i = curr_size; i < new_size; i++) { + u16 new_idx = (ioat->tail+i) & (new_size-1); + + ring[new_idx] = ioat2_alloc_ring_ent(c, GFP_NOWAIT); + if (!ring[new_idx]) { + while (i--) { + u16 new_idx = (ioat->tail+i) & (new_size-1); + + ioat2_free_ring_ent(ring[new_idx], c); + } + kfree(ring); + return false; + } + set_desc_id(ring[new_idx], new_idx); + } + + /* hw link new descriptors */ + for (i = curr_size-1; i < new_size; i++) { + u16 new_idx = (ioat->tail+i) & (new_size-1); + struct ioat_ring_ent *next = ring[(new_idx+1) & (new_size-1)]; + struct ioat_dma_descriptor *hw = ring[new_idx]->hw; + + hw->next = next->txd.phys; + } + } else { + struct ioat_dma_descriptor *hw; + struct ioat_ring_ent *next; + + /* copy current descriptors to the new ring, dropping the + * removed descriptors + */ + for (i = 0; i < new_size; i++) { + u16 curr_idx = (ioat->tail+i) & (curr_size-1); + u16 new_idx = (ioat->tail+i) & (new_size-1); + + ring[new_idx] = ioat->ring[curr_idx]; + set_desc_id(ring[new_idx], new_idx); + } + + /* free deleted descriptors */ + for (i = new_size; i < curr_size; i++) { + struct ioat_ring_ent *ent; + + ent = ioat2_get_ring_ent(ioat, ioat->tail+i); + ioat2_free_ring_ent(ent, c); + } + + /* fix up hardware ring */ + hw = ring[(ioat->tail+new_size-1) & (new_size-1)]->hw; + next = ring[(ioat->tail+new_size) & (new_size-1)]; + hw->next = next->txd.phys; + } + + dev_dbg(to_dev(chan), "%s: allocated %d descriptors\n", + __func__, new_size); + + kfree(ioat->ring); + ioat->ring = ring; + ioat->alloc_order = order; + + return true; +} + +/** + * ioat2_alloc_and_lock - common descriptor alloc boilerplate for ioat2,3 ops + * @idx: gets starting descriptor index on successful allocation + * @ioat: ioat2,3 channel (ring) to operate on + * @num_descs: allocation length + */ +static int ioat2_alloc_and_lock(u16 *idx, struct ioat2_dma_chan *ioat, int num_descs) +{ + struct ioat_chan_common *chan = &ioat->base; + + spin_lock_bh(&ioat->ring_lock); + /* never allow the last descriptor to be consumed, we need at + * least one free at all times to allow for on-the-fly ring + * resizing. + */ + while (unlikely(ioat2_ring_space(ioat) <= num_descs)) { + if (reshape_ring(ioat, ioat->alloc_order + 1) && + ioat2_ring_space(ioat) > num_descs) + break; + + if (printk_ratelimit()) + dev_dbg(to_dev(chan), + "%s: ring full! num_descs: %d (%x:%x:%x)\n", + __func__, num_descs, ioat->head, ioat->tail, + ioat->issued); + spin_unlock_bh(&ioat->ring_lock); + + /* progress reclaim in the allocation failure case we + * may be called under bh_disabled so we need to trigger + * the timer event directly + */ + spin_lock_bh(&chan->cleanup_lock); + if (jiffies > chan->timer.expires && + timer_pending(&chan->timer)) { + mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT); + spin_unlock_bh(&chan->cleanup_lock); + ioat2_timer_event((unsigned long) ioat); + } else + spin_unlock_bh(&chan->cleanup_lock); + return -ENOMEM; + } + + dev_dbg(to_dev(chan), "%s: num_descs: %d (%x:%x:%x)\n", + __func__, num_descs, ioat->head, ioat->tail, ioat->issued); + + *idx = ioat2_desc_alloc(ioat, num_descs); + return 0; /* with ioat->ring_lock held */ +} + +static struct dma_async_tx_descriptor * +ioat2_dma_prep_memcpy_lock(struct dma_chan *c, dma_addr_t dma_dest, + dma_addr_t dma_src, size_t len, unsigned long flags) +{ + struct ioat2_dma_chan *ioat = to_ioat2_chan(c); + struct ioat_dma_descriptor *hw; + struct ioat_ring_ent *desc; + dma_addr_t dst = dma_dest; + dma_addr_t src = dma_src; + size_t total_len = len; + int num_descs; + u16 idx; + int i; + + num_descs = ioat2_xferlen_to_descs(ioat, len); + if (likely(num_descs) && + ioat2_alloc_and_lock(&idx, ioat, num_descs) == 0) + /* pass */; + else + return NULL; + for (i = 0; i < num_descs; i++) { + size_t copy = min_t(size_t, len, 1 << ioat->xfercap_log); + + desc = ioat2_get_ring_ent(ioat, idx + i); + hw = desc->hw; + + hw->size = copy; + hw->ctl = 0; + hw->src_addr = src; + hw->dst_addr = dst; + + len -= copy; + dst += copy; + src += copy; + dump_desc_dbg(ioat, desc); + } + + desc->txd.flags = flags; + desc->len = total_len; + hw->ctl_f.int_en = !!(flags & DMA_PREP_INTERRUPT); + hw->ctl_f.compl_write = 1; + dump_desc_dbg(ioat, desc); + /* we leave the channel locked to ensure in order submission */ + + return &desc->txd; +} + +/** + * ioat2_free_chan_resources - release all the descriptors + * @chan: the channel to be cleaned + */ +static void ioat2_free_chan_resources(struct dma_chan *c) +{ + struct ioat2_dma_chan *ioat = to_ioat2_chan(c); + struct ioat_chan_common *chan = &ioat->base; + struct ioatdma_device *ioatdma_device = chan->device; + struct ioat_ring_ent *desc; + const u16 total_descs = 1 << ioat->alloc_order; + int descs; + int i; + + /* Before freeing channel resources first check + * if they have been previously allocated for this channel. + */ + if (!ioat->ring) + return; + + tasklet_disable(&chan->cleanup_task); + del_timer_sync(&chan->timer); + ioat2_cleanup(ioat); + + /* Delay 100ms after reset to allow internal DMA logic to quiesce + * before removing DMA descriptor resources. + */ + writeb(IOAT_CHANCMD_RESET, + chan->reg_base + IOAT_CHANCMD_OFFSET(chan->device->version)); + mdelay(100); + + spin_lock_bh(&ioat->ring_lock); + descs = ioat2_ring_space(ioat); + dev_dbg(to_dev(chan), "freeing %d idle descriptors\n", descs); + for (i = 0; i < descs; i++) { + desc = ioat2_get_ring_ent(ioat, ioat->head + i); + ioat2_free_ring_ent(desc, c); + } + + if (descs < total_descs) + dev_err(to_dev(chan), "Freeing %d in use descriptors!\n", + total_descs - descs); + + for (i = 0; i < total_descs - descs; i++) { + desc = ioat2_get_ring_ent(ioat, ioat->tail + i); + dump_desc_dbg(ioat, desc); + ioat2_free_ring_ent(desc, c); + } + + kfree(ioat->ring); + ioat->ring = NULL; + ioat->alloc_order = 0; + pci_pool_free(ioatdma_device->completion_pool, + chan->completion, + chan->completion_dma); + spin_unlock_bh(&ioat->ring_lock); + + chan->last_completion = 0; + chan->completion_dma = 0; + ioat->pending = 0; + ioat->dmacount = 0; +} + +static enum dma_status +ioat2_is_complete(struct dma_chan *c, dma_cookie_t cookie, + dma_cookie_t *done, dma_cookie_t *used) +{ + struct ioat2_dma_chan *ioat = to_ioat2_chan(c); + + if (ioat_is_complete(c, cookie, done, used) == DMA_SUCCESS) + return DMA_SUCCESS; + + ioat2_cleanup(ioat); + + return ioat_is_complete(c, cookie, done, used); +} + +int __devinit ioat2_dma_probe(struct ioatdma_device *device, int dca) +{ + struct pci_dev *pdev = device->pdev; + struct dma_device *dma; + struct dma_chan *c; + struct ioat_chan_common *chan; + int err; + + device->enumerate_channels = ioat2_enumerate_channels; + dma = &device->common; + dma->device_prep_dma_memcpy = ioat2_dma_prep_memcpy_lock; + dma->device_issue_pending = ioat2_issue_pending; + dma->device_alloc_chan_resources = ioat2_alloc_chan_resources; + dma->device_free_chan_resources = ioat2_free_chan_resources; + dma->device_is_tx_complete = ioat2_is_complete; + + err = ioat_probe(device); + if (err) + return err; + ioat_set_tcp_copy_break(2048); + + list_for_each_entry(c, &dma->channels, device_node) { + chan = to_chan_common(c); + writel(IOAT_DCACTRL_CMPL_WRITE_ENABLE | IOAT_DMA_DCA_ANY_CPU, + chan->reg_base + IOAT_DCACTRL_OFFSET); + } + + err = ioat_register(device); + if (err) + return err; + if (dca) + device->dca = ioat2_dca_init(pdev, device->reg_base); + + return err; +} + +int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca) +{ + struct pci_dev *pdev = device->pdev; + struct dma_device *dma; + struct dma_chan *c; + struct ioat_chan_common *chan; + int err; + u16 dev_id; + + device->enumerate_channels = ioat2_enumerate_channels; + dma = &device->common; + dma->device_prep_dma_memcpy = ioat2_dma_prep_memcpy_lock; + dma->device_issue_pending = ioat2_issue_pending; + dma->device_alloc_chan_resources = ioat2_alloc_chan_resources; + dma->device_free_chan_resources = ioat2_free_chan_resources; + dma->device_is_tx_complete = ioat2_is_complete; + + /* -= IOAT ver.3 workarounds =- */ + /* Write CHANERRMSK_INT with 3E07h to mask out the errors + * that can cause stability issues for IOAT ver.3 + */ + pci_write_config_dword(pdev, IOAT_PCI_CHANERRMASK_INT_OFFSET, 0x3e07); + + /* Clear DMAUNCERRSTS Cfg-Reg Parity Error status bit + * (workaround for spurious config parity error after restart) + */ + pci_read_config_word(pdev, IOAT_PCI_DEVICE_ID_OFFSET, &dev_id); + if (dev_id == PCI_DEVICE_ID_INTEL_IOAT_TBG0) + pci_write_config_dword(pdev, IOAT_PCI_DMAUNCERRSTS_OFFSET, 0x10); + + err = ioat_probe(device); + if (err) + return err; + ioat_set_tcp_copy_break(262144); + + list_for_each_entry(c, &dma->channels, device_node) { + chan = to_chan_common(c); + writel(IOAT_DMA_DCA_ANY_CPU, + chan->reg_base + IOAT_DCACTRL_OFFSET); + } + + err = ioat_register(device); + if (err) + return err; + if (dca) + device->dca = ioat3_dca_init(pdev, device->reg_base); + + return err; +} diff --git a/drivers/dma/ioat/dma_v2.h b/drivers/dma/ioat/dma_v2.h new file mode 100644 index 00000000000..9baa3d6065f --- /dev/null +++ b/drivers/dma/ioat/dma_v2.h @@ -0,0 +1,146 @@ +/* + * Copyright(c) 2004 - 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ +#ifndef IOATDMA_V2_H +#define IOATDMA_V2_H + +#include <linux/dmaengine.h> +#include "dma.h" +#include "hw.h" + + +extern int ioat_pending_level; + +/* + * workaround for IOAT ver.3.0 null descriptor issue + * (channel returns error when size is 0) + */ +#define NULL_DESC_BUFFER_SIZE 1 + +#define IOAT_MAX_ORDER 16 +#define ioat_get_alloc_order() \ + (min(ioat_ring_alloc_order, IOAT_MAX_ORDER)) +#define ioat_get_max_alloc_order() \ + (min(ioat_ring_max_alloc_order, IOAT_MAX_ORDER)) + +/* struct ioat2_dma_chan - ioat v2 / v3 channel attributes + * @base: common ioat channel parameters + * @xfercap_log; log2 of channel max transfer length (for fast division) + * @head: allocated index + * @issued: hardware notification point + * @tail: cleanup index + * @pending: lock free indicator for issued != head + * @dmacount: identical to 'head' except for occasionally resetting to zero + * @alloc_order: log2 of the number of allocated descriptors + * @ring: software ring buffer implementation of hardware ring + * @ring_lock: protects ring attributes + */ +struct ioat2_dma_chan { + struct ioat_chan_common base; + size_t xfercap_log; + u16 head; + u16 issued; + u16 tail; + u16 dmacount; + u16 alloc_order; + int pending; + struct ioat_ring_ent **ring; + spinlock_t ring_lock; +}; + +static inline struct ioat2_dma_chan *to_ioat2_chan(struct dma_chan *c) +{ + struct ioat_chan_common *chan = to_chan_common(c); + + return container_of(chan, struct ioat2_dma_chan, base); +} + +static inline u16 ioat2_ring_mask(struct ioat2_dma_chan *ioat) +{ + return (1 << ioat->alloc_order) - 1; +} + +/* count of descriptors in flight with the engine */ +static inline u16 ioat2_ring_active(struct ioat2_dma_chan *ioat) +{ + return (ioat->head - ioat->tail) & ioat2_ring_mask(ioat); +} + +/* count of descriptors pending submission to hardware */ +static inline u16 ioat2_ring_pending(struct ioat2_dma_chan *ioat) +{ + return (ioat->head - ioat->issued) & ioat2_ring_mask(ioat); +} + +static inline u16 ioat2_ring_space(struct ioat2_dma_chan *ioat) +{ + u16 num_descs = ioat2_ring_mask(ioat) + 1; + u16 active = ioat2_ring_active(ioat); + + BUG_ON(active > num_descs); + + return num_descs - active; +} + +/* assumes caller already checked space */ +static inline u16 ioat2_desc_alloc(struct ioat2_dma_chan *ioat, u16 len) +{ + ioat->head += len; + return ioat->head - len; +} + +static inline u16 ioat2_xferlen_to_descs(struct ioat2_dma_chan *ioat, size_t len) +{ + u16 num_descs = len >> ioat->xfercap_log; + + num_descs += !!(len & ((1 << ioat->xfercap_log) - 1)); + return num_descs; +} + +struct ioat_ring_ent { + struct ioat_dma_descriptor *hw; + struct dma_async_tx_descriptor txd; + size_t len; + #ifdef DEBUG + int id; + #endif +}; + +static inline struct ioat_ring_ent * +ioat2_get_ring_ent(struct ioat2_dma_chan *ioat, u16 idx) +{ + return ioat->ring[idx & ioat2_ring_mask(ioat)]; +} + +static inline void ioat2_set_chainaddr(struct ioat2_dma_chan *ioat, u64 addr) +{ + struct ioat_chan_common *chan = &ioat->base; + + writel(addr & 0x00000000FFFFFFFF, + chan->reg_base + IOAT2_CHAINADDR_OFFSET_LOW); + writel(addr >> 32, + chan->reg_base + IOAT2_CHAINADDR_OFFSET_HIGH); +} + +int __devinit ioat2_dma_probe(struct ioatdma_device *dev, int dca); +int __devinit ioat3_dma_probe(struct ioatdma_device *dev, int dca); +struct dca_provider * __devinit ioat2_dca_init(struct pci_dev *pdev, void __iomem *iobase); +struct dca_provider * __devinit ioat3_dca_init(struct pci_dev *pdev, void __iomem *iobase); +#endif /* IOATDMA_V2_H */ diff --git a/drivers/dma/ioat/hw.h b/drivers/dma/ioat/hw.h new file mode 100644 index 00000000000..7481fb13ce0 --- /dev/null +++ b/drivers/dma/ioat/hw.h @@ -0,0 +1,73 @@ +/* + * Copyright(c) 2004 - 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ +#ifndef _IOAT_HW_H_ +#define _IOAT_HW_H_ + +/* PCI Configuration Space Values */ +#define IOAT_PCI_VID 0x8086 +#define IOAT_MMIO_BAR 0 + +/* CB device ID's */ +#define IOAT_PCI_DID_5000 0x1A38 +#define IOAT_PCI_DID_CNB 0x360B +#define IOAT_PCI_DID_SCNB 0x65FF +#define IOAT_PCI_DID_SNB 0x402F + +#define IOAT_PCI_RID 0x00 +#define IOAT_PCI_SVID 0x8086 +#define IOAT_PCI_SID 0x8086 +#define IOAT_VER_1_2 0x12 /* Version 1.2 */ +#define IOAT_VER_2_0 0x20 /* Version 2.0 */ +#define IOAT_VER_3_0 0x30 /* Version 3.0 */ + +struct ioat_dma_descriptor { + uint32_t size; + union { + uint32_t ctl; + struct { + unsigned int int_en:1; + unsigned int src_snoop_dis:1; + unsigned int dest_snoop_dis:1; + unsigned int compl_write:1; + unsigned int fence:1; + unsigned int null:1; + unsigned int src_brk:1; + unsigned int dest_brk:1; + unsigned int bundle:1; + unsigned int dest_dca:1; + unsigned int hint:1; + unsigned int rsvd2:13; + unsigned int op:8; + } ctl_f; + }; + uint64_t src_addr; + uint64_t dst_addr; + uint64_t next; + uint64_t rsv1; + uint64_t rsv2; + /* store some driver data in an unused portion of the descriptor */ + union { + uint64_t user1; + uint64_t tx_cnt; + }; + uint64_t user2; +}; +#endif diff --git a/drivers/dma/ioat/pci.c b/drivers/dma/ioat/pci.c new file mode 100644 index 00000000000..c4e43226925 --- /dev/null +++ b/drivers/dma/ioat/pci.c @@ -0,0 +1,179 @@ +/* + * Intel I/OAT DMA Linux driver + * Copyright(c) 2007 - 2009 Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + * + * The full GNU General Public License is included in this distribution in + * the file called "COPYING". + * + */ + +/* + * This driver supports an Intel I/OAT DMA engine, which does asynchronous + * copy operations. + */ + +#include <linux/init.h> +#include <linux/module.h> +#include <linux/pci.h> +#include <linux/interrupt.h> +#include <linux/dca.h> +#include "dma.h" +#include "dma_v2.h" +#include "registers.h" +#include "hw.h" + +MODULE_VERSION(IOAT_DMA_VERSION); +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Intel Corporation"); + +static struct pci_device_id ioat_pci_tbl[] = { + /* I/OAT v1 platforms */ + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_CNB) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_SCNB) }, + { PCI_DEVICE(PCI_VENDOR_ID_UNISYS, PCI_DEVICE_ID_UNISYS_DMA_DIRECTOR) }, + + /* I/OAT v2 platforms */ + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_SNB) }, + + /* I/OAT v3 platforms */ + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG0) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG1) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG2) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG3) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG4) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG5) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG6) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IOAT_TBG7) }, + { 0, } +}; + +static int __devinit ioat_pci_probe(struct pci_dev *pdev, + const struct pci_device_id *id); +static void __devexit ioat_remove(struct pci_dev *pdev); + +static int ioat_dca_enabled = 1; +module_param(ioat_dca_enabled, int, 0644); +MODULE_PARM_DESC(ioat_dca_enabled, "control support of dca service (default: 1)"); + +#define DRV_NAME "ioatdma" + +static struct pci_driver ioat_pci_driver = { + .name = DRV_NAME, + .id_table = ioat_pci_tbl, + .probe = ioat_pci_probe, + .remove = __devexit_p(ioat_remove), +}; + +static struct ioatdma_device * +alloc_ioatdma(struct pci_dev *pdev, void __iomem *iobase) +{ + struct device *dev = &pdev->dev; + struct ioatdma_device *d = devm_kzalloc(dev, sizeof(*d), GFP_KERNEL); + + if (!d) + return NULL; + d->pdev = pdev; + d->reg_base = iobase; + return d; +} + +static int __devinit ioat_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + void __iomem * const *iomap; + struct device *dev = &pdev->dev; + struct ioatdma_device *device; + int err; + + err = pcim_enable_device(pdev); + if (err) + return err; + + err = pcim_iomap_regions(pdev, 1 << IOAT_MMIO_BAR, DRV_NAME); + if (err) + return err; + iomap = pcim_iomap_table(pdev); + if (!iomap) + return -ENOMEM; + + err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64)); + if (err) + err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); + if (err) + return err; + + err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64)); + if (err) + err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32)); + if (err) + return err; + + device = devm_kzalloc(dev, sizeof(*device), GFP_KERNEL); + if (!device) + return -ENOMEM; + + pci_set_master(pdev); + + device = alloc_ioatdma(pdev, iomap[IOAT_MMIO_BAR]); + if (!device) + return -ENOMEM; + pci_set_drvdata(pdev, device); + + device->version = readb(device->reg_base + IOAT_VER_OFFSET); + if (device->version == IOAT_VER_1_2) + err = ioat1_dma_probe(device, ioat_dca_enabled); + else if (device->version == IOAT_VER_2_0) + err = ioat2_dma_probe(device, ioat_dca_enabled); + else if (device->version >= IOAT_VER_3_0) + err = ioat3_dma_probe(device, ioat_dca_enabled); + else + return -ENODEV; + + if (err) { + dev_err(dev, "Intel(R) I/OAT DMA Engine init failed\n"); + return -ENODEV; + } + + return 0; +} + +static void __devexit ioat_remove(struct pci_dev *pdev) +{ + struct ioatdma_device *device = pci_get_drvdata(pdev); + + if (!device) + return; + + dev_err(&pdev->dev, "Removing dma and dca services\n"); + if (device->dca) { + unregister_dca_provider(device->dca); + free_dca_provider(device->dca); + device->dca = NULL; + } + ioat_dma_remove(device); +} + +static int __init ioat_init_module(void) +{ + return pci_register_driver(&ioat_pci_driver); +} +module_init(ioat_init_module); + +static void __exit ioat_exit_module(void) +{ + pci_unregister_driver(&ioat_pci_driver); +} +module_exit(ioat_exit_module); diff --git a/drivers/dma/ioat/registers.h b/drivers/dma/ioat/registers.h new file mode 100644 index 00000000000..e4334a19538 --- /dev/null +++ b/drivers/dma/ioat/registers.h @@ -0,0 +1,230 @@ +/* + * Copyright(c) 2004 - 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution in the + * file called COPYING. + */ +#ifndef _IOAT_REGISTERS_H_ +#define _IOAT_REGISTERS_H_ + +#define IOAT_PCI_DMACTRL_OFFSET 0x48 +#define IOAT_PCI_DMACTRL_DMA_EN 0x00000001 +#define IOAT_PCI_DMACTRL_MSI_EN 0x00000002 + +#define IOAT_PCI_DEVICE_ID_OFFSET 0x02 +#define IOAT_PCI_DMAUNCERRSTS_OFFSET 0x148 +#define IOAT_PCI_CHANERRMASK_INT_OFFSET 0x184 + +/* MMIO Device Registers */ +#define IOAT_CHANCNT_OFFSET 0x00 /* 8-bit */ + +#define IOAT_XFERCAP_OFFSET 0x01 /* 8-bit */ +#define IOAT_XFERCAP_4KB 12 +#define IOAT_XFERCAP_8KB 13 +#define IOAT_XFERCAP_16KB 14 +#define IOAT_XFERCAP_32KB 15 +#define IOAT_XFERCAP_32GB 0 + +#define IOAT_GENCTRL_OFFSET 0x02 /* 8-bit */ +#define IOAT_GENCTRL_DEBUG_EN 0x01 + +#define IOAT_INTRCTRL_OFFSET 0x03 /* 8-bit */ +#define IOAT_INTRCTRL_MASTER_INT_EN 0x01 /* Master Interrupt Enable */ +#define IOAT_INTRCTRL_INT_STATUS 0x02 /* ATTNSTATUS -or- Channel Int */ +#define IOAT_INTRCTRL_INT 0x04 /* INT_STATUS -and- MASTER_INT_EN */ +#define IOAT_INTRCTRL_MSIX_VECTOR_CONTROL 0x08 /* Enable all MSI-X vectors */ + +#define IOAT_ATTNSTATUS_OFFSET 0x04 /* Each bit is a channel */ + +#define IOAT_VER_OFFSET 0x08 /* 8-bit */ +#define IOAT_VER_MAJOR_MASK 0xF0 +#define IOAT_VER_MINOR_MASK 0x0F +#define GET_IOAT_VER_MAJOR(x) (((x) & IOAT_VER_MAJOR_MASK) >> 4) +#define GET_IOAT_VER_MINOR(x) ((x) & IOAT_VER_MINOR_MASK) + +#define IOAT_PERPORTOFFSET_OFFSET 0x0A /* 16-bit */ + +#define IOAT_INTRDELAY_OFFSET 0x0C /* 16-bit */ +#define IOAT_INTRDELAY_INT_DELAY_MASK 0x3FFF /* Interrupt Delay Time */ +#define IOAT_INTRDELAY_COALESE_SUPPORT 0x8000 /* Interrupt Coalescing Supported */ + +#define IOAT_DEVICE_STATUS_OFFSET 0x0E /* 16-bit */ +#define IOAT_DEVICE_STATUS_DEGRADED_MODE 0x0001 + +#define IOAT_CHANNEL_MMIO_SIZE 0x80 /* Each Channel MMIO space is this size */ + +/* DMA Channel Registers */ +#define IOAT_CHANCTRL_OFFSET 0x00 /* 16-bit Channel Control Register */ +#define IOAT_CHANCTRL_CHANNEL_PRIORITY_MASK 0xF000 +#define IOAT_CHANCTRL_CHANNEL_IN_USE 0x0100 +#define IOAT_CHANCTRL_DESCRIPTOR_ADDR_SNOOP_CONTROL 0x0020 +#define IOAT_CHANCTRL_ERR_INT_EN 0x0010 +#define IOAT_CHANCTRL_ANY_ERR_ABORT_EN 0x0008 +#define IOAT_CHANCTRL_ERR_COMPLETION_EN 0x0004 +#define IOAT_CHANCTRL_INT_REARM 0x0001 +#define IOAT_CHANCTRL_RUN (IOAT_CHANCTRL_INT_REARM |\ + IOAT_CHANCTRL_ERR_COMPLETION_EN |\ + IOAT_CHANCTRL_ANY_ERR_ABORT_EN |\ + IOAT_CHANCTRL_ERR_INT_EN) + +#define IOAT_DMA_COMP_OFFSET 0x02 /* 16-bit DMA channel compatibility */ +#define IOAT_DMA_COMP_V1 0x0001 /* Compatibility with DMA version 1 */ +#define IOAT_DMA_COMP_V2 0x0002 /* Compatibility with DMA version 2 */ + + +#define IOAT1_CHANSTS_OFFSET 0x04 /* 64-bit Channel Status Register */ +#define IOAT2_CHANSTS_OFFSET 0x08 /* 64-bit Channel Status Register */ +#define IOAT_CHANSTS_OFFSET(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHANSTS_OFFSET : IOAT2_CHANSTS_OFFSET) +#define IOAT1_CHANSTS_OFFSET_LOW 0x04 +#define IOAT2_CHANSTS_OFFSET_LOW 0x08 +#define IOAT_CHANSTS_OFFSET_LOW(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHANSTS_OFFSET_LOW : IOAT2_CHANSTS_OFFSET_LOW) +#define IOAT1_CHANSTS_OFFSET_HIGH 0x08 +#define IOAT2_CHANSTS_OFFSET_HIGH 0x0C +#define IOAT_CHANSTS_OFFSET_HIGH(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHANSTS_OFFSET_HIGH : IOAT2_CHANSTS_OFFSET_HIGH) +#define IOAT_CHANSTS_COMPLETED_DESCRIPTOR_ADDR (~0x3fULL) +#define IOAT_CHANSTS_SOFT_ERR 0x10ULL +#define IOAT_CHANSTS_UNAFFILIATED_ERR 0x8ULL +#define IOAT_CHANSTS_STATUS 0x7ULL +#define IOAT_CHANSTS_ACTIVE 0x0 +#define IOAT_CHANSTS_DONE 0x1 +#define IOAT_CHANSTS_SUSPENDED 0x2 +#define IOAT_CHANSTS_HALTED 0x3 + + + +#define IOAT_CHAN_DMACOUNT_OFFSET 0x06 /* 16-bit DMA Count register */ + +#define IOAT_DCACTRL_OFFSET 0x30 /* 32 bit Direct Cache Access Control Register */ +#define IOAT_DCACTRL_CMPL_WRITE_ENABLE 0x10000 +#define IOAT_DCACTRL_TARGET_CPU_MASK 0xFFFF /* APIC ID */ + +/* CB DCA Memory Space Registers */ +#define IOAT_DCAOFFSET_OFFSET 0x14 +/* CB_BAR + IOAT_DCAOFFSET value */ +#define IOAT_DCA_VER_OFFSET 0x00 +#define IOAT_DCA_VER_MAJOR_MASK 0xF0 +#define IOAT_DCA_VER_MINOR_MASK 0x0F + +#define IOAT_DCA_COMP_OFFSET 0x02 +#define IOAT_DCA_COMP_V1 0x1 + +#define IOAT_FSB_CAPABILITY_OFFSET 0x04 +#define IOAT_FSB_CAPABILITY_PREFETCH 0x1 + +#define IOAT_PCI_CAPABILITY_OFFSET 0x06 +#define IOAT_PCI_CAPABILITY_MEMWR 0x1 + +#define IOAT_FSB_CAP_ENABLE_OFFSET 0x08 +#define IOAT_FSB_CAP_ENABLE_PREFETCH 0x1 + +#define IOAT_PCI_CAP_ENABLE_OFFSET 0x0A +#define IOAT_PCI_CAP_ENABLE_MEMWR 0x1 + +#define IOAT_APICID_TAG_MAP_OFFSET 0x0C +#define IOAT_APICID_TAG_MAP_TAG0 0x0000000F +#define IOAT_APICID_TAG_MAP_TAG0_SHIFT 0 +#define IOAT_APICID_TAG_MAP_TAG1 0x000000F0 +#define IOAT_APICID_TAG_MAP_TAG1_SHIFT 4 +#define IOAT_APICID_TAG_MAP_TAG2 0x00000F00 +#define IOAT_APICID_TAG_MAP_TAG2_SHIFT 8 +#define IOAT_APICID_TAG_MAP_TAG3 0x0000F000 +#define IOAT_APICID_TAG_MAP_TAG3_SHIFT 12 +#define IOAT_APICID_TAG_MAP_TAG4 0x000F0000 +#define IOAT_APICID_TAG_MAP_TAG4_SHIFT 16 +#define IOAT_APICID_TAG_CB2_VALID 0x8080808080 + +#define IOAT_DCA_GREQID_OFFSET 0x10 +#define IOAT_DCA_GREQID_SIZE 0x04 +#define IOAT_DCA_GREQID_MASK 0xFFFF +#define IOAT_DCA_GREQID_IGNOREFUN 0x10000000 +#define IOAT_DCA_GREQID_VALID 0x20000000 +#define IOAT_DCA_GREQID_LASTID 0x80000000 + +#define IOAT3_CSI_CAPABILITY_OFFSET 0x08 +#define IOAT3_CSI_CAPABILITY_PREFETCH 0x1 + +#define IOAT3_PCI_CAPABILITY_OFFSET 0x0A +#define IOAT3_PCI_CAPABILITY_MEMWR 0x1 + +#define IOAT3_CSI_CONTROL_OFFSET 0x0C +#define IOAT3_CSI_CONTROL_PREFETCH 0x1 + +#define IOAT3_PCI_CONTROL_OFFSET 0x0E +#define IOAT3_PCI_CONTROL_MEMWR 0x1 + +#define IOAT3_APICID_TAG_MAP_OFFSET 0x10 +#define IOAT3_APICID_TAG_MAP_OFFSET_LOW 0x10 +#define IOAT3_APICID_TAG_MAP_OFFSET_HIGH 0x14 + +#define IOAT3_DCA_GREQID_OFFSET 0x02 + +#define IOAT1_CHAINADDR_OFFSET 0x0C /* 64-bit Descriptor Chain Address Register */ +#define IOAT2_CHAINADDR_OFFSET 0x10 /* 64-bit Descriptor Chain Address Register */ +#define IOAT_CHAINADDR_OFFSET(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHAINADDR_OFFSET : IOAT2_CHAINADDR_OFFSET) +#define IOAT1_CHAINADDR_OFFSET_LOW 0x0C +#define IOAT2_CHAINADDR_OFFSET_LOW 0x10 +#define IOAT_CHAINADDR_OFFSET_LOW(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHAINADDR_OFFSET_LOW : IOAT2_CHAINADDR_OFFSET_LOW) +#define IOAT1_CHAINADDR_OFFSET_HIGH 0x10 +#define IOAT2_CHAINADDR_OFFSET_HIGH 0x14 +#define IOAT_CHAINADDR_OFFSET_HIGH(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHAINADDR_OFFSET_HIGH : IOAT2_CHAINADDR_OFFSET_HIGH) + +#define IOAT1_CHANCMD_OFFSET 0x14 /* 8-bit DMA Channel Command Register */ +#define IOAT2_CHANCMD_OFFSET 0x04 /* 8-bit DMA Channel Command Register */ +#define IOAT_CHANCMD_OFFSET(ver) ((ver) < IOAT_VER_2_0 \ + ? IOAT1_CHANCMD_OFFSET : IOAT2_CHANCMD_OFFSET) +#define IOAT_CHANCMD_RESET 0x20 +#define IOAT_CHANCMD_RESUME 0x10 +#define IOAT_CHANCMD_ABORT 0x08 +#define IOAT_CHANCMD_SUSPEND 0x04 +#define IOAT_CHANCMD_APPEND 0x02 +#define IOAT_CHANCMD_START 0x01 + +#define IOAT_CHANCMP_OFFSET 0x18 /* 64-bit Channel Completion Address Register */ +#define IOAT_CHANCMP_OFFSET_LOW 0x18 +#define IOAT_CHANCMP_OFFSET_HIGH 0x1C + +#define IOAT_CDAR_OFFSET 0x20 /* 64-bit Current Descriptor Address Register */ +#define IOAT_CDAR_OFFSET_LOW 0x20 +#define IOAT_CDAR_OFFSET_HIGH 0x24 + +#define IOAT_CHANERR_OFFSET 0x28 /* 32-bit Channel Error Register */ +#define IOAT_CHANERR_SRC_ADDR_ERR 0x0001 +#define IOAT_CHANERR_DEST_ADDR_ERR 0x0002 +#define IOAT_CHANERR_NEXT_ADDR_ERR 0x0004 +#define IOAT_CHANERR_NEXT_DESC_ALIGN_ERR 0x0008 +#define IOAT_CHANERR_CHAIN_ADDR_VALUE_ERR 0x0010 +#define IOAT_CHANERR_CHANCMD_ERR 0x0020 +#define IOAT_CHANERR_CHIPSET_UNCORRECTABLE_DATA_INTEGRITY_ERR 0x0040 +#define IOAT_CHANERR_DMA_UNCORRECTABLE_DATA_INTEGRITY_ERR 0x0080 +#define IOAT_CHANERR_READ_DATA_ERR 0x0100 +#define IOAT_CHANERR_WRITE_DATA_ERR 0x0200 +#define IOAT_CHANERR_CONTROL_ERR 0x0400 +#define IOAT_CHANERR_LENGTH_ERR 0x0800 +#define IOAT_CHANERR_COMPLETION_ADDR_ERR 0x1000 +#define IOAT_CHANERR_INT_CONFIGURATION_ERR 0x2000 +#define IOAT_CHANERR_SOFT_ERR 0x4000 +#define IOAT_CHANERR_UNAFFILIATED_ERR 0x8000 + +#define IOAT_CHANERR_MASK_OFFSET 0x2C /* 32-bit Channel Error Register */ + +#endif /* _IOAT_REGISTERS_H_ */ |