#include <sys/socket.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
+#include <stdbool.h>
#include <rte_log.h>
#include <rte_pci.h>
#include <rte_eal_memconfig.h>
#include <rte_malloc.h>
-#include <eal_private.h>
+#include <rte_vfio.h>
#include "eal_filesystem.h"
#include "eal_pci_init.h"
-#include "eal_vfio.h"
+#include "eal_private.h"
/**
* @file
* This file is only compiled if CONFIG_RTE_EAL_VFIO is set to "y".
*/
-#ifdef VFIO_PRESENT
+#ifdef RTE_EAL_VFIO
#define PAGE_SIZE (sysconf(_SC_PAGESIZE))
#define PAGE_MASK (~(PAGE_SIZE - 1))
};
EAL_REGISTER_TAILQ(rte_vfio_tailq)
-#define VFIO_DIR "/dev/vfio"
-#define VFIO_CONTAINER_PATH "/dev/vfio/vfio"
-#define VFIO_GROUP_FMT "/dev/vfio/%u"
-#define VFIO_NOIOMMU_GROUP_FMT "/dev/vfio/noiommu-%u"
-#define VFIO_GET_REGION_ADDR(x) ((uint64_t) x << 40ULL)
-#define VFIO_GET_REGION_IDX(x) (x >> 40)
-
-/* per-process VFIO config */
-static struct vfio_config vfio_cfg;
-
-/* DMA mapping function prototype.
- * Takes VFIO container fd as a parameter.
- * Returns 0 on success, -1 on error.
- * */
-typedef int (*vfio_dma_func_t)(int);
-
-struct vfio_iommu_type {
- int type_id;
- const char *name;
- vfio_dma_func_t dma_map_func;
-};
-
-static int vfio_type1_dma_map(int);
-static int vfio_noiommu_dma_map(int);
-
-/* IOMMU types we support */
-static const struct vfio_iommu_type iommu_types[] = {
- /* x86 IOMMU, otherwise known as type 1 */
- { RTE_VFIO_TYPE1, "Type 1", &vfio_type1_dma_map},
- /* IOMMU-less mode */
- { RTE_VFIO_NOIOMMU, "No-IOMMU", &vfio_noiommu_dma_map},
-};
-
-int
-vfio_type1_dma_map(int vfio_container_fd)
-{
- const struct rte_memseg *ms = rte_eal_get_physmem_layout();
- int i, ret;
-
- /* map all DPDK segments for DMA. use 1:1 PA to IOVA mapping */
- for (i = 0; i < RTE_MAX_MEMSEG; i++) {
- struct vfio_iommu_type1_dma_map dma_map;
-
- if (ms[i].addr == NULL)
- break;
-
- memset(&dma_map, 0, sizeof(dma_map));
- dma_map.argsz = sizeof(struct vfio_iommu_type1_dma_map);
- dma_map.vaddr = ms[i].addr_64;
- dma_map.size = ms[i].len;
- dma_map.iova = ms[i].phys_addr;
- dma_map.flags = VFIO_DMA_MAP_FLAG_READ | VFIO_DMA_MAP_FLAG_WRITE;
-
- ret = ioctl(vfio_container_fd, VFIO_IOMMU_MAP_DMA, &dma_map);
-
- if (ret) {
- RTE_LOG(ERR, EAL, " cannot set up DMA remapping, "
- "error %i (%s)\n", errno, strerror(errno));
- return -1;
- }
- }
-
- return 0;
-}
-
-int
-vfio_noiommu_dma_map(int __rte_unused vfio_container_fd)
-{
- /* No-IOMMU mode does not need DMA mapping */
- return 0;
-}
-
int
pci_vfio_read_config(const struct rte_intr_handle *intr_handle,
void *buf, size_t len, off_t offs)
/* get PCI BAR number where MSI-X interrupts are */
static int
-pci_vfio_get_msix_bar(int fd, int *msix_bar, uint32_t *msix_table_offset,
- uint32_t *msix_table_size)
+pci_vfio_get_msix_bar(int fd, struct pci_msix_table *msix_table)
{
int ret;
uint32_t reg;
return -1;
}
- *msix_bar = reg & RTE_PCI_MSIX_TABLE_BIR;
- *msix_table_offset = reg & RTE_PCI_MSIX_TABLE_OFFSET;
- *msix_table_size = 16 * (1 + (flags & RTE_PCI_MSIX_FLAGS_QSIZE));
+ msix_table->bar_index = reg & RTE_PCI_MSIX_TABLE_BIR;
+ msix_table->offset = reg & RTE_PCI_MSIX_TABLE_OFFSET;
+ msix_table->size =
+ 16 * (1 + (flags & RTE_PCI_MSIX_FLAGS_QSIZE));
return 0;
}
/* set PCI bus mastering */
static int
-pci_vfio_set_bus_master(int dev_fd)
+pci_vfio_set_bus_master(int dev_fd, bool op)
{
uint16_t reg;
int ret;
return -1;
}
- /* set the master bit */
- reg |= PCI_COMMAND_MASTER;
+ if (op)
+ /* set the master bit */
+ reg |= PCI_COMMAND_MASTER;
+ else
+ reg &= ~(PCI_COMMAND_MASTER);
ret = pwrite64(dev_fd, ®, sizeof(reg),
VFIO_GET_REGION_ADDR(VFIO_PCI_CONFIG_REGION_INDEX) +
return 0;
}
-/* pick IOMMU type. returns a pointer to vfio_iommu_type or NULL for error */
-static const struct vfio_iommu_type *
-pci_vfio_set_iommu_type(int vfio_container_fd) {
- unsigned idx;
- for (idx = 0; idx < RTE_DIM(iommu_types); idx++) {
- const struct vfio_iommu_type *t = &iommu_types[idx];
-
- int ret = ioctl(vfio_container_fd, VFIO_SET_IOMMU,
- t->type_id);
- if (!ret) {
- RTE_LOG(NOTICE, EAL, " using IOMMU type %d (%s)\n",
- t->type_id, t->name);
- return t;
- }
- /* not an error, there may be more supported IOMMU types */
- RTE_LOG(DEBUG, EAL, " set IOMMU type %d (%s) failed, "
- "error %i (%s)\n", t->type_id, t->name, errno,
- strerror(errno));
- }
- /* if we didn't find a suitable IOMMU type, fail */
- return NULL;
-}
-
-/* check if we have any supported extensions */
-static int
-pci_vfio_has_supported_extensions(int vfio_container_fd) {
- int ret;
- unsigned idx, n_extensions = 0;
- for (idx = 0; idx < RTE_DIM(iommu_types); idx++) {
- const struct vfio_iommu_type *t = &iommu_types[idx];
-
- ret = ioctl(vfio_container_fd, VFIO_CHECK_EXTENSION,
- t->type_id);
- if (ret < 0) {
- RTE_LOG(ERR, EAL, " could not get IOMMU type, "
- "error %i (%s)\n", errno,
- strerror(errno));
- close(vfio_container_fd);
- return -1;
- } else if (ret == 1) {
- /* we found a supported extension */
- n_extensions++;
- }
- RTE_LOG(DEBUG, EAL, " IOMMU type %d (%s) is %s\n",
- t->type_id, t->name,
- ret ? "supported" : "not supported");
- }
-
- /* if we didn't find any supported IOMMU types, fail */
- if (!n_extensions) {
- close(vfio_container_fd);
- return -1;
- }
-
- return 0;
-}
-
/* set up interrupt support (but not enable interrupts) */
static int
pci_vfio_setup_interrupts(struct rte_pci_device *dev, int vfio_dev_fd)
intr_idx = VFIO_PCI_NUM_IRQS;
/* get interrupt type from internal config (MSI-X by default, can be
- * overriden from the command line
+ * overridden from the command line
*/
switch (internal_config.vfio_intr_mode) {
case RTE_INTR_MODE_MSIX:
return -1;
}
-/* open container fd or get an existing one */
-int
-pci_vfio_get_container_fd(void)
+static int
+pci_vfio_is_ioport_bar(int vfio_dev_fd, int bar_index)
{
- int ret, vfio_container_fd;
-
- /* if we're in a primary process, try to open the container */
- if (internal_config.process_type == RTE_PROC_PRIMARY) {
- vfio_container_fd = open(VFIO_CONTAINER_PATH, O_RDWR);
- if (vfio_container_fd < 0) {
- RTE_LOG(ERR, EAL, " cannot open VFIO container, "
- "error %i (%s)\n", errno, strerror(errno));
- return -1;
- }
-
- /* check VFIO API version */
- ret = ioctl(vfio_container_fd, VFIO_GET_API_VERSION);
- if (ret != VFIO_API_VERSION) {
- if (ret < 0)
- RTE_LOG(ERR, EAL, " could not get VFIO API version, "
- "error %i (%s)\n", errno, strerror(errno));
- else
- RTE_LOG(ERR, EAL, " unsupported VFIO API version!\n");
- close(vfio_container_fd);
- return -1;
- }
-
- ret = pci_vfio_has_supported_extensions(vfio_container_fd);
- if (ret) {
- RTE_LOG(ERR, EAL, " no supported IOMMU "
- "extensions found!\n");
- return -1;
- }
-
- return vfio_container_fd;
- } else {
- /*
- * if we're in a secondary process, request container fd from the
- * primary process via our socket
- */
- int socket_fd;
+ uint32_t ioport_bar;
+ int ret;
- socket_fd = vfio_mp_sync_connect_to_primary();
- if (socket_fd < 0) {
- RTE_LOG(ERR, EAL, " cannot connect to primary process!\n");
- return -1;
- }
- if (vfio_mp_sync_send_request(socket_fd, SOCKET_REQ_CONTAINER) < 0) {
- RTE_LOG(ERR, EAL, " cannot request container fd!\n");
- close(socket_fd);
- return -1;
- }
- vfio_container_fd = vfio_mp_sync_receive_fd(socket_fd);
- if (vfio_container_fd < 0) {
- RTE_LOG(ERR, EAL, " cannot get container fd!\n");
- close(socket_fd);
- return -1;
- }
- close(socket_fd);
- return vfio_container_fd;
+ ret = pread64(vfio_dev_fd, &ioport_bar, sizeof(ioport_bar),
+ VFIO_GET_REGION_ADDR(VFIO_PCI_CONFIG_REGION_INDEX)
+ + PCI_BASE_ADDRESS_0 + bar_index*4);
+ if (ret != sizeof(ioport_bar)) {
+ RTE_LOG(ERR, EAL, "Cannot read command (%x) from config space!\n",
+ PCI_BASE_ADDRESS_0 + bar_index*4);
+ return -1;
}
- return -1;
+ return (ioport_bar & PCI_BASE_ADDRESS_SPACE_IO) != 0;
}
-/* open group fd or get an existing one */
-int
-pci_vfio_get_group_fd(int iommu_group_no)
+static int
+pci_vfio_setup_device(struct rte_pci_device *dev, int vfio_dev_fd)
{
- int i;
- int vfio_group_fd;
- char filename[PATH_MAX];
-
- /* check if we already have the group descriptor open */
- for (i = 0; i < vfio_cfg.vfio_group_idx; i++)
- if (vfio_cfg.vfio_groups[i].group_no == iommu_group_no)
- return vfio_cfg.vfio_groups[i].fd;
-
- /* if primary, try to open the group */
- if (internal_config.process_type == RTE_PROC_PRIMARY) {
- /* try regular group format */
- snprintf(filename, sizeof(filename),
- VFIO_GROUP_FMT, iommu_group_no);
- vfio_group_fd = open(filename, O_RDWR);
- if (vfio_group_fd < 0) {
- /* if file not found, it's not an error */
- if (errno != ENOENT) {
- RTE_LOG(ERR, EAL, "Cannot open %s: %s\n", filename,
- strerror(errno));
- return -1;
- }
-
- /* special case: try no-IOMMU path as well */
- snprintf(filename, sizeof(filename),
- VFIO_NOIOMMU_GROUP_FMT, iommu_group_no);
- vfio_group_fd = open(filename, O_RDWR);
- if (vfio_group_fd < 0) {
- if (errno != ENOENT) {
- RTE_LOG(ERR, EAL, "Cannot open %s: %s\n", filename,
- strerror(errno));
- return -1;
- }
- return 0;
- }
- /* noiommu group found */
- }
+ if (pci_vfio_setup_interrupts(dev, vfio_dev_fd) != 0) {
+ RTE_LOG(ERR, EAL, "Error setting up interrupts!\n");
+ return -1;
+ }
- /* if the fd is valid, create a new group for it */
- if (vfio_cfg.vfio_group_idx == VFIO_MAX_GROUPS) {
- RTE_LOG(ERR, EAL, "Maximum number of VFIO groups reached!\n");
- close(vfio_group_fd);
- return -1;
- }
- vfio_cfg.vfio_groups[vfio_cfg.vfio_group_idx].group_no = iommu_group_no;
- vfio_cfg.vfio_groups[vfio_cfg.vfio_group_idx].fd = vfio_group_fd;
- return vfio_group_fd;
+ /* set bus mastering for the device */
+ if (pci_vfio_set_bus_master(vfio_dev_fd, true)) {
+ RTE_LOG(ERR, EAL, "Cannot set up bus mastering!\n");
+ return -1;
}
- /* if we're in a secondary process, request group fd from the primary
- * process via our socket
- */
- else {
- int socket_fd, ret;
- socket_fd = vfio_mp_sync_connect_to_primary();
+ /* Reset the device */
+ ioctl(vfio_dev_fd, VFIO_DEVICE_RESET);
- if (socket_fd < 0) {
- RTE_LOG(ERR, EAL, " cannot connect to primary process!\n");
- return -1;
- }
- if (vfio_mp_sync_send_request(socket_fd, SOCKET_REQ_GROUP) < 0) {
- RTE_LOG(ERR, EAL, " cannot request container fd!\n");
- close(socket_fd);
- return -1;
- }
- if (vfio_mp_sync_send_request(socket_fd, iommu_group_no) < 0) {
- RTE_LOG(ERR, EAL, " cannot send group number!\n");
- close(socket_fd);
- return -1;
- }
- ret = vfio_mp_sync_receive_request(socket_fd);
- switch (ret) {
- case SOCKET_NO_FD:
- close(socket_fd);
- return 0;
- case SOCKET_OK:
- vfio_group_fd = vfio_mp_sync_receive_fd(socket_fd);
- /* if we got the fd, return it */
- if (vfio_group_fd > 0) {
- close(socket_fd);
- return vfio_group_fd;
- }
- /* fall-through on error */
- default:
- RTE_LOG(ERR, EAL, " cannot get container fd!\n");
- close(socket_fd);
- return -1;
- }
- }
- return -1;
+ return 0;
}
-/* parse IOMMU group number for a PCI device
- * returns 1 on success, -1 for errors, 0 for non-existent group
- */
static int
-pci_vfio_get_group_no(const char *pci_addr, int *iommu_group_no)
+pci_vfio_mmap_bar(int vfio_dev_fd, struct mapped_pci_resource *vfio_res,
+ int bar_index, int additional_flags)
{
- char linkname[PATH_MAX];
- char filename[PATH_MAX];
- char *tok[16], *group_tok, *end;
- int ret;
-
- memset(linkname, 0, sizeof(linkname));
- memset(filename, 0, sizeof(filename));
-
- /* try to find out IOMMU group for this device */
- snprintf(linkname, sizeof(linkname),
- "%s/%s/iommu_group", pci_get_sysfs_path(), pci_addr);
-
- ret = readlink(linkname, filename, sizeof(filename));
-
- /* if the link doesn't exist, no VFIO for us */
- if (ret < 0)
+ struct memreg {
+ unsigned long offset, size;
+ } memreg[2] = {};
+ void *bar_addr;
+ struct pci_msix_table *msix_table = &vfio_res->msix_table;
+ struct pci_map *bar = &vfio_res->maps[bar_index];
+
+ if (bar->size == 0)
+ /* Skip this BAR */
return 0;
- ret = rte_strsplit(filename, sizeof(filename),
- tok, RTE_DIM(tok), '/');
+ if (msix_table->bar_index == bar_index) {
+ /*
+ * VFIO will not let us map the MSI-X table,
+ * but we can map around it.
+ */
+ uint32_t table_start = msix_table->offset;
+ uint32_t table_end = table_start + msix_table->size;
+ table_end = (table_end + ~PAGE_MASK) & PAGE_MASK;
+ table_start &= PAGE_MASK;
+
+ if (table_start == 0 && table_end >= bar->size) {
+ /* Cannot map this BAR */
+ RTE_LOG(DEBUG, EAL, "Skipping BAR%d\n", bar_index);
+ bar->size = 0;
+ bar->addr = 0;
+ return 0;
+ }
- if (ret <= 0) {
- RTE_LOG(ERR, EAL, " %s cannot get IOMMU group\n", pci_addr);
- return -1;
+ memreg[0].offset = bar->offset;
+ memreg[0].size = table_start;
+ memreg[1].offset = bar->offset + table_end;
+ memreg[1].size = bar->size - table_end;
+
+ RTE_LOG(DEBUG, EAL,
+ "Trying to map BAR%d that contains the MSI-X "
+ "table. Trying offsets: "
+ "0x%04lx:0x%04lx, 0x%04lx:0x%04lx\n", bar_index,
+ memreg[0].offset, memreg[0].size,
+ memreg[1].offset, memreg[1].size);
+ } else {
+ memreg[0].offset = bar->offset;
+ memreg[0].size = bar->size;
}
- /* IOMMU group is always the last token */
- errno = 0;
- group_tok = tok[ret - 1];
- end = group_tok;
- *iommu_group_no = strtol(group_tok, &end, 10);
- if ((end != group_tok && *end != '\0') || errno != 0) {
- RTE_LOG(ERR, EAL, " %s error parsing IOMMU number!\n", pci_addr);
+ /* reserve the address using an inaccessible mapping */
+ bar_addr = mmap(bar->addr, bar->size, 0, MAP_PRIVATE |
+ MAP_ANONYMOUS | additional_flags, -1, 0);
+ if (bar_addr != MAP_FAILED) {
+ void *map_addr = NULL;
+ if (memreg[0].size) {
+ /* actual map of first part */
+ map_addr = pci_map_resource(bar_addr, vfio_dev_fd,
+ memreg[0].offset,
+ memreg[0].size,
+ MAP_FIXED);
+ }
+
+ /* if there's a second part, try to map it */
+ if (map_addr != MAP_FAILED
+ && memreg[1].offset && memreg[1].size) {
+ void *second_addr = RTE_PTR_ADD(bar_addr,
+ memreg[1].offset -
+ (uintptr_t)bar->offset);
+ map_addr = pci_map_resource(second_addr,
+ vfio_dev_fd,
+ memreg[1].offset,
+ memreg[1].size,
+ MAP_FIXED);
+ }
+
+ if (map_addr == MAP_FAILED || !map_addr) {
+ munmap(bar_addr, bar->size);
+ bar_addr = MAP_FAILED;
+ RTE_LOG(ERR, EAL, "Failed to map pci BAR%d\n",
+ bar_index);
+ return -1;
+ }
+ } else {
+ RTE_LOG(ERR, EAL,
+ "Failed to create inaccessible mapping for BAR%d\n",
+ bar_index);
return -1;
}
- return 1;
-}
-
-static void
-clear_current_group(void)
-{
- vfio_cfg.vfio_groups[vfio_cfg.vfio_group_idx].group_no = 0;
- vfio_cfg.vfio_groups[vfio_cfg.vfio_group_idx].fd = -1;
+ bar->addr = bar_addr;
+ return 0;
}
-
-/*
- * map the PCI resources of a PCI device in virtual memory (VFIO version).
- * primary and secondary processes follow almost exactly the same path
- */
-int
-pci_vfio_map_resource(struct rte_pci_device *dev)
+static int
+pci_vfio_map_resource_primary(struct rte_pci_device *dev)
{
- struct vfio_group_status group_status = {
- .argsz = sizeof(group_status)
- };
struct vfio_device_info device_info = { .argsz = sizeof(device_info) };
- int vfio_group_fd, vfio_dev_fd;
- int iommu_group_no;
char pci_addr[PATH_MAX] = {0};
+ int vfio_dev_fd;
struct rte_pci_addr *loc = &dev->addr;
- int i, ret, msix_bar;
+ int i, ret;
struct mapped_pci_resource *vfio_res = NULL;
- struct mapped_pci_res_list *vfio_res_list = RTE_TAILQ_CAST(rte_vfio_tailq.head, mapped_pci_res_list);
+ struct mapped_pci_res_list *vfio_res_list =
+ RTE_TAILQ_CAST(rte_vfio_tailq.head, mapped_pci_res_list);
struct pci_map *maps;
- uint32_t msix_table_offset = 0;
- uint32_t msix_table_size = 0;
- uint32_t ioport_bar;
dev->intr_handle.fd = -1;
dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
snprintf(pci_addr, sizeof(pci_addr), PCI_PRI_FMT,
loc->domain, loc->bus, loc->devid, loc->function);
- /* get group number */
- ret = pci_vfio_get_group_no(pci_addr, &iommu_group_no);
- if (ret == 0) {
- RTE_LOG(WARNING, EAL, " %s not managed by VFIO driver, skipping\n",
- pci_addr);
- return 1;
- }
+ ret = vfio_setup_device(pci_get_sysfs_path(), pci_addr,
+ &vfio_dev_fd, &device_info);
+ if (ret)
+ return ret;
- /* if negative, something failed */
- if (ret < 0)
- return -1;
-
- /* get the actual group fd */
- vfio_group_fd = pci_vfio_get_group_fd(iommu_group_no);
- if (vfio_group_fd < 0)
- return -1;
+ /* allocate vfio_res and get region info */
+ vfio_res = rte_zmalloc("VFIO_RES", sizeof(*vfio_res), 0);
+ if (vfio_res == NULL) {
+ RTE_LOG(ERR, EAL,
+ "%s(): cannot store uio mmap details\n", __func__);
+ goto err_vfio_dev_fd;
+ }
+ memcpy(&vfio_res->pci_addr, &dev->addr, sizeof(vfio_res->pci_addr));
- /* store group fd */
- vfio_cfg.vfio_groups[vfio_cfg.vfio_group_idx].group_no = iommu_group_no;
- vfio_cfg.vfio_groups[vfio_cfg.vfio_group_idx].fd = vfio_group_fd;
+ /* get number of registers (up to BAR5) */
+ vfio_res->nb_maps = RTE_MIN((int) device_info.num_regions,
+ VFIO_PCI_BAR5_REGION_INDEX + 1);
- /* if group_fd == 0, that means the device isn't managed by VFIO */
- if (vfio_group_fd == 0) {
- RTE_LOG(WARNING, EAL, " %s not managed by VFIO driver, skipping\n",
- pci_addr);
- /* we store 0 as group fd to distinguish between existing but
- * unbound VFIO groups, and groups that don't exist at all.
- */
- vfio_cfg.vfio_group_idx++;
- return 1;
- }
+ /* map BARs */
+ maps = vfio_res->maps;
- /*
- * at this point, we know at least one port on this device is bound to VFIO,
- * so we can proceed to try and set this particular port up
+ vfio_res->msix_table.bar_index = -1;
+ /* get MSI-X BAR, if any (we have to know where it is because we can't
+ * easily mmap it when using VFIO)
*/
-
- /* check if the group is viable */
- ret = ioctl(vfio_group_fd, VFIO_GROUP_GET_STATUS, &group_status);
- if (ret) {
- RTE_LOG(ERR, EAL, " %s cannot get group status, "
- "error %i (%s)\n", pci_addr, errno, strerror(errno));
- close(vfio_group_fd);
- clear_current_group();
- return -1;
- } else if (!(group_status.flags & VFIO_GROUP_FLAGS_VIABLE)) {
- RTE_LOG(ERR, EAL, " %s VFIO group is not viable!\n", pci_addr);
- close(vfio_group_fd);
- clear_current_group();
- return -1;
+ ret = pci_vfio_get_msix_bar(vfio_dev_fd, &vfio_res->msix_table);
+ if (ret < 0) {
+ RTE_LOG(ERR, EAL, " %s cannot get MSI-X BAR number!\n",
+ pci_addr);
+ goto err_vfio_dev_fd;
}
- /*
- * at this point, we know that this group is viable (meaning, all devices
- * are either bound to VFIO or not bound to anything)
- */
+ for (i = 0; i < (int) vfio_res->nb_maps; i++) {
+ struct vfio_region_info reg = { .argsz = sizeof(reg) };
+ void *bar_addr;
- /* check if group does not have a container yet */
- if (!(group_status.flags & VFIO_GROUP_FLAGS_CONTAINER_SET)) {
+ reg.index = i;
- /* add group to a container */
- ret = ioctl(vfio_group_fd, VFIO_GROUP_SET_CONTAINER,
- &vfio_cfg.vfio_container_fd);
+ ret = ioctl(vfio_dev_fd, VFIO_DEVICE_GET_REGION_INFO, ®);
if (ret) {
- RTE_LOG(ERR, EAL, " %s cannot add VFIO group to container, "
+ RTE_LOG(ERR, EAL, " %s cannot get device region info "
"error %i (%s)\n", pci_addr, errno, strerror(errno));
- close(vfio_group_fd);
- clear_current_group();
- return -1;
+ goto err_vfio_res;
}
- /*
- * at this point we know that this group has been successfully
- * initialized, so we increment vfio_group_idx to indicate that we can
- * add new groups.
- */
- vfio_cfg.vfio_group_idx++;
- }
- /*
- * pick an IOMMU type and set up DMA mappings for container
- *
- * needs to be done only once, only when at least one group is assigned to
- * a container and only in primary process
- */
- if (internal_config.process_type == RTE_PROC_PRIMARY &&
- vfio_cfg.vfio_container_has_dma == 0) {
- /* select an IOMMU type which we will be using */
- const struct vfio_iommu_type *t =
- pci_vfio_set_iommu_type(vfio_cfg.vfio_container_fd);
- if (!t) {
- RTE_LOG(ERR, EAL, " %s failed to select IOMMU type\n", pci_addr);
- return -1;
+ /* chk for io port region */
+ ret = pci_vfio_is_ioport_bar(vfio_dev_fd, i);
+ if (ret < 0)
+ goto err_vfio_res;
+ else if (ret) {
+ RTE_LOG(INFO, EAL, "Ignore mapping IO port bar(%d)\n",
+ i);
+ continue;
}
- ret = t->dma_map_func(vfio_cfg.vfio_container_fd);
- if (ret) {
- RTE_LOG(ERR, EAL, " %s DMA remapping failed, "
- "error %i (%s)\n", pci_addr, errno, strerror(errno));
- return -1;
+
+ /* skip non-mmapable BARs */
+ if ((reg.flags & VFIO_REGION_INFO_FLAG_MMAP) == 0)
+ continue;
+
+ /* try mapping somewhere close to the end of hugepages */
+ if (pci_map_addr == NULL)
+ pci_map_addr = pci_find_max_end_va();
+
+ bar_addr = pci_map_addr;
+ pci_map_addr = RTE_PTR_ADD(bar_addr, (size_t) reg.size);
+
+ maps[i].addr = bar_addr;
+ maps[i].offset = reg.offset;
+ maps[i].size = reg.size;
+ maps[i].path = NULL; /* vfio doesn't have per-resource paths */
+
+ ret = pci_vfio_mmap_bar(vfio_dev_fd, vfio_res, i, 0);
+ if (ret < 0) {
+ RTE_LOG(ERR, EAL, " %s mapping BAR%i failed: %s\n",
+ pci_addr, i, strerror(errno));
+ goto err_vfio_res;
}
- vfio_cfg.vfio_container_has_dma = 1;
- }
- /* get a file descriptor for the device */
- vfio_dev_fd = ioctl(vfio_group_fd, VFIO_GROUP_GET_DEVICE_FD, pci_addr);
- if (vfio_dev_fd < 0) {
- /* if we cannot get a device fd, this simply means that this
- * particular port is not bound to VFIO
- */
- RTE_LOG(WARNING, EAL, " %s not managed by VFIO driver, skipping\n",
- pci_addr);
- return 1;
+ dev->mem_resource[i].addr = maps[i].addr;
}
- /* test and setup the device */
- ret = ioctl(vfio_dev_fd, VFIO_DEVICE_GET_INFO, &device_info);
- if (ret) {
- RTE_LOG(ERR, EAL, " %s cannot get device info, "
- "error %i (%s)\n", pci_addr, errno, strerror(errno));
- close(vfio_dev_fd);
- return -1;
+ if (pci_vfio_setup_device(dev, vfio_dev_fd) < 0) {
+ RTE_LOG(ERR, EAL, " %s setup device failed\n", pci_addr);
+ goto err_vfio_res;
}
- /* get MSI-X BAR, if any (we have to know where it is because we can't
- * easily mmap it when using VFIO) */
- msix_bar = -1;
- ret = pci_vfio_get_msix_bar(vfio_dev_fd, &msix_bar,
- &msix_table_offset, &msix_table_size);
- if (ret < 0) {
- RTE_LOG(ERR, EAL, " %s cannot get MSI-X BAR number!\n", pci_addr);
- close(vfio_dev_fd);
- return -1;
- }
+ TAILQ_INSERT_TAIL(vfio_res_list, vfio_res, next);
- /* if we're in a primary process, allocate vfio_res and get region info */
- if (internal_config.process_type == RTE_PROC_PRIMARY) {
- vfio_res = rte_zmalloc("VFIO_RES", sizeof(*vfio_res), 0);
- if (vfio_res == NULL) {
- RTE_LOG(ERR, EAL,
- "%s(): cannot store uio mmap details\n", __func__);
- close(vfio_dev_fd);
- return -1;
- }
- memcpy(&vfio_res->pci_addr, &dev->addr, sizeof(vfio_res->pci_addr));
+ return 0;
+err_vfio_res:
+ rte_free(vfio_res);
+err_vfio_dev_fd:
+ close(vfio_dev_fd);
+ return -1;
+}
- /* get number of registers (up to BAR5) */
- vfio_res->nb_maps = RTE_MIN((int) device_info.num_regions,
- VFIO_PCI_BAR5_REGION_INDEX + 1);
- } else {
- /* if we're in a secondary process, just find our tailq entry */
- TAILQ_FOREACH(vfio_res, vfio_res_list, next) {
- if (memcmp(&vfio_res->pci_addr, &dev->addr, sizeof(dev->addr)))
- continue;
- break;
- }
- /* if we haven't found our tailq entry, something's wrong */
- if (vfio_res == NULL) {
- RTE_LOG(ERR, EAL, " %s cannot find TAILQ entry for PCI device!\n",
- pci_addr);
- close(vfio_dev_fd);
- return -1;
- }
+static int
+pci_vfio_map_resource_secondary(struct rte_pci_device *dev)
+{
+ struct vfio_device_info device_info = { .argsz = sizeof(device_info) };
+ char pci_addr[PATH_MAX] = {0};
+ int vfio_dev_fd;
+ struct rte_pci_addr *loc = &dev->addr;
+ int i, ret;
+ struct mapped_pci_resource *vfio_res = NULL;
+ struct mapped_pci_res_list *vfio_res_list =
+ RTE_TAILQ_CAST(rte_vfio_tailq.head, mapped_pci_res_list);
+
+ struct pci_map *maps;
+
+ dev->intr_handle.fd = -1;
+ dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
+
+ /* store PCI address string */
+ snprintf(pci_addr, sizeof(pci_addr), PCI_PRI_FMT,
+ loc->domain, loc->bus, loc->devid, loc->function);
+
+ ret = vfio_setup_device(pci_get_sysfs_path(), pci_addr,
+ &vfio_dev_fd, &device_info);
+ if (ret)
+ return ret;
+
+ /* if we're in a secondary process, just find our tailq entry */
+ TAILQ_FOREACH(vfio_res, vfio_res_list, next) {
+ if (rte_eal_compare_pci_addr(&vfio_res->pci_addr,
+ &dev->addr))
+ continue;
+ break;
+ }
+ /* if we haven't found our tailq entry, something's wrong */
+ if (vfio_res == NULL) {
+ RTE_LOG(ERR, EAL, " %s cannot find TAILQ entry for PCI device!\n",
+ pci_addr);
+ goto err_vfio_dev_fd;
}
/* map BARs */
maps = vfio_res->maps;
for (i = 0; i < (int) vfio_res->nb_maps; i++) {
- struct vfio_region_info reg = { .argsz = sizeof(reg) };
- void *bar_addr;
- struct memreg {
- unsigned long offset, size;
- } memreg[2] = {};
-
- reg.index = i;
-
- ret = ioctl(vfio_dev_fd, VFIO_DEVICE_GET_REGION_INFO, ®);
-
- if (ret) {
- RTE_LOG(ERR, EAL, " %s cannot get device region info "
- "error %i (%s)\n", pci_addr, errno, strerror(errno));
- close(vfio_dev_fd);
- if (internal_config.process_type == RTE_PROC_PRIMARY)
- rte_free(vfio_res);
- return -1;
+ ret = pci_vfio_mmap_bar(vfio_dev_fd, vfio_res, i, MAP_FIXED);
+ if (ret < 0) {
+ RTE_LOG(ERR, EAL, " %s mapping BAR%i failed: %s\n",
+ pci_addr, i, strerror(errno));
+ goto err_vfio_dev_fd;
}
- /* chk for io port region */
- ret = pread64(vfio_dev_fd, &ioport_bar, sizeof(ioport_bar),
- VFIO_GET_REGION_ADDR(VFIO_PCI_CONFIG_REGION_INDEX)
- + PCI_BASE_ADDRESS_0 + i*4);
-
- if (ret != sizeof(ioport_bar)) {
- RTE_LOG(ERR, EAL,
- "Cannot read command (%x) from config space!\n",
- PCI_BASE_ADDRESS_0 + i*4);
- return -1;
- }
+ dev->mem_resource[i].addr = maps[i].addr;
+ }
- if (ioport_bar & PCI_BASE_ADDRESS_SPACE_IO) {
- RTE_LOG(INFO, EAL,
- "Ignore mapping IO port bar(%d) addr: %x\n",
- i, ioport_bar);
- continue;
- }
+ return 0;
+err_vfio_dev_fd:
+ close(vfio_dev_fd);
+ return -1;
+}
- /* skip non-mmapable BARs */
- if ((reg.flags & VFIO_REGION_INFO_FLAG_MMAP) == 0)
- continue;
+/*
+ * map the PCI resources of a PCI device in virtual memory (VFIO version).
+ * primary and secondary processes follow almost exactly the same path
+ */
+int
+pci_vfio_map_resource(struct rte_pci_device *dev)
+{
+ if (internal_config.process_type == RTE_PROC_PRIMARY)
+ return pci_vfio_map_resource_primary(dev);
+ else
+ return pci_vfio_map_resource_secondary(dev);
+}
- if (i == msix_bar) {
- /*
- * VFIO will not let us map the MSI-X table,
- * but we can map around it.
- */
- uint32_t table_start = msix_table_offset;
- uint32_t table_end = table_start + msix_table_size;
- table_end = (table_end + ~PAGE_MASK) & PAGE_MASK;
- table_start &= PAGE_MASK;
-
- if (table_start == 0 && table_end >= reg.size) {
- /* Cannot map this BAR */
- RTE_LOG(DEBUG, EAL, "Skipping BAR %d\n", i);
- continue;
- } else {
- memreg[0].offset = reg.offset;
- memreg[0].size = table_start;
- memreg[1].offset = table_end;
- memreg[1].size = reg.size - table_end;
-
- RTE_LOG(DEBUG, EAL,
- "Trying to map BAR %d that contains the MSI-X "
- "table. Trying offsets: "
- "0x%04lx:0x%04lx, 0x%04lx:0x%04lx\n", i,
- memreg[0].offset, memreg[0].size,
- memreg[1].offset, memreg[1].size);
- }
- } else {
- memreg[0].offset = reg.offset;
- memreg[0].size = reg.size;
- }
+int
+pci_vfio_unmap_resource(struct rte_pci_device *dev)
+{
+ char pci_addr[PATH_MAX] = {0};
+ struct rte_pci_addr *loc = &dev->addr;
+ int i, ret;
+ struct mapped_pci_resource *vfio_res = NULL;
+ struct mapped_pci_res_list *vfio_res_list;
- /* try to figure out an address */
- if (internal_config.process_type == RTE_PROC_PRIMARY) {
- /* try mapping somewhere close to the end of hugepages */
- if (pci_map_addr == NULL)
- pci_map_addr = pci_find_max_end_va();
+ struct pci_map *maps;
- bar_addr = pci_map_addr;
- pci_map_addr = RTE_PTR_ADD(bar_addr, (size_t) reg.size);
- } else {
- bar_addr = maps[i].addr;
- }
+ /* store PCI address string */
+ snprintf(pci_addr, sizeof(pci_addr), PCI_PRI_FMT,
+ loc->domain, loc->bus, loc->devid, loc->function);
- /* reserve the address using an inaccessible mapping */
- bar_addr = mmap(bar_addr, reg.size, 0, MAP_PRIVATE |
- MAP_ANONYMOUS, -1, 0);
- if (bar_addr != MAP_FAILED) {
- void *map_addr = NULL;
- if (memreg[0].size) {
- /* actual map of first part */
- map_addr = pci_map_resource(bar_addr, vfio_dev_fd,
- memreg[0].offset,
- memreg[0].size,
- MAP_FIXED);
- }
- /* if there's a second part, try to map it */
- if (map_addr != MAP_FAILED
- && memreg[1].offset && memreg[1].size) {
- void *second_addr = RTE_PTR_ADD(bar_addr, memreg[1].offset);
- map_addr = pci_map_resource(second_addr,
- vfio_dev_fd, memreg[1].offset,
- memreg[1].size,
- MAP_FIXED);
- }
+ if (close(dev->intr_handle.fd) < 0) {
+ RTE_LOG(INFO, EAL, "Error when closing eventfd file descriptor for %s\n",
+ pci_addr);
+ return -1;
+ }
- if (map_addr == MAP_FAILED || !map_addr) {
- munmap(bar_addr, reg.size);
- bar_addr = MAP_FAILED;
- }
- }
+ if (pci_vfio_set_bus_master(dev->intr_handle.vfio_dev_fd, false)) {
+ RTE_LOG(ERR, EAL, " %s cannot unset bus mastering for PCI device!\n",
+ pci_addr);
+ return -1;
+ }
- if (bar_addr == MAP_FAILED ||
- (internal_config.process_type == RTE_PROC_SECONDARY &&
- bar_addr != maps[i].addr)) {
- RTE_LOG(ERR, EAL, " %s mapping BAR%i failed: %s\n", pci_addr, i,
- strerror(errno));
- close(vfio_dev_fd);
- if (internal_config.process_type == RTE_PROC_PRIMARY)
- rte_free(vfio_res);
- return -1;
- }
+ ret = vfio_release_device(pci_get_sysfs_path(), pci_addr,
+ dev->intr_handle.vfio_dev_fd);
+ if (ret < 0) {
+ RTE_LOG(ERR, EAL,
+ "%s(): cannot release device\n", __func__);
+ return ret;
+ }
- maps[i].addr = bar_addr;
- maps[i].offset = reg.offset;
- maps[i].size = reg.size;
- maps[i].path = NULL; /* vfio doesn't have per-resource paths */
- dev->mem_resource[i].addr = bar_addr;
+ vfio_res_list = RTE_TAILQ_CAST(rte_vfio_tailq.head, mapped_pci_res_list);
+ /* Get vfio_res */
+ TAILQ_FOREACH(vfio_res, vfio_res_list, next) {
+ if (memcmp(&vfio_res->pci_addr, &dev->addr, sizeof(dev->addr)))
+ continue;
+ break;
+ }
+ /* if we haven't found our tailq entry, something's wrong */
+ if (vfio_res == NULL) {
+ RTE_LOG(ERR, EAL, " %s cannot find TAILQ entry for PCI device!\n",
+ pci_addr);
+ return -1;
}
- /* if secondary process, do not set up interrupts */
- if (internal_config.process_type == RTE_PROC_PRIMARY) {
- if (pci_vfio_setup_interrupts(dev, vfio_dev_fd) != 0) {
- RTE_LOG(ERR, EAL, " %s error setting up interrupts!\n", pci_addr);
- close(vfio_dev_fd);
- rte_free(vfio_res);
- return -1;
- }
+ /* unmap BARs */
+ maps = vfio_res->maps;
- /* set bus mastering for the device */
- if (pci_vfio_set_bus_master(vfio_dev_fd)) {
- RTE_LOG(ERR, EAL, " %s cannot set up bus mastering!\n", pci_addr);
- close(vfio_dev_fd);
- rte_free(vfio_res);
- return -1;
- }
+ RTE_LOG(INFO, EAL, "Releasing pci mapped resource for %s\n",
+ pci_addr);
+ for (i = 0; i < (int) vfio_res->nb_maps; i++) {
- /* Reset the device */
- ioctl(vfio_dev_fd, VFIO_DEVICE_RESET);
+ /*
+ * We do not need to be aware of MSI-X table BAR mappings as
+ * when mapping. Just using current maps array is enough
+ */
+ if (maps[i].addr) {
+ RTE_LOG(INFO, EAL, "Calling pci_unmap_resource for %s at %p\n",
+ pci_addr, maps[i].addr);
+ pci_unmap_resource(maps[i].addr, maps[i].size);
+ }
}
- if (internal_config.process_type == RTE_PROC_PRIMARY)
- TAILQ_INSERT_TAIL(vfio_res_list, vfio_res, next);
+ TAILQ_REMOVE(vfio_res_list, vfio_res, next);
return 0;
}
return -1;
}
-int
-pci_vfio_enable(void)
-{
- /* initialize group list */
- int i;
- int vfio_available;
-
- for (i = 0; i < VFIO_MAX_GROUPS; i++) {
- vfio_cfg.vfio_groups[i].fd = -1;
- vfio_cfg.vfio_groups[i].group_no = -1;
- }
-
- /* inform the user that we are probing for VFIO */
- RTE_LOG(INFO, EAL, "Probing VFIO support...\n");
-
- /* check if vfio-pci module is loaded */
- vfio_available = rte_eal_check_module("vfio_pci");
-
- /* return error directly */
- if (vfio_available == -1) {
- RTE_LOG(INFO, EAL, "Could not get loaded module details!\n");
- return -1;
- }
-
- /* return 0 if VFIO modules not loaded */
- if (vfio_available == 0) {
- RTE_LOG(DEBUG, EAL, "VFIO modules not loaded, "
- "skipping VFIO support...\n");
- return 0;
- }
-
- vfio_cfg.vfio_container_fd = pci_vfio_get_container_fd();
-
- /* check if we have VFIO driver enabled */
- if (vfio_cfg.vfio_container_fd != -1) {
- RTE_LOG(NOTICE, EAL, "VFIO support initialized\n");
- vfio_cfg.vfio_enabled = 1;
- } else {
- RTE_LOG(NOTICE, EAL, "VFIO support could not be initialized\n");
- }
-
- return 0;
-}
-
int
pci_vfio_is_enabled(void)
{
- return vfio_cfg.vfio_enabled;
+ return vfio_is_enabled("vfio_pci");
}
#endif