Thanks. very good learning material.
On Tue, Nov 30, 2010 at 1:16 AM, Michael S. Tsirkin <mst@xxxxxxxxxx> wrote:
> This is the userspace part of the tool: it includes a bunch of stubs for
> linux APIs, somewhat simular to linuxsched. This makes it possible to
> recompile the ring code in userspace.
>
> A small test example is implemented combining this with vhost_test
> module.
>
> Signed-off-by: Michael S. Tsirkin <mst@xxxxxxxxxx>
>
> ---
>
> diff --git a/tools/virtio/Makefile b/tools/virtio/Makefile
> new file mode 100644
> index 0000000..d1d442e
> --- /dev/null
> +++ b/tools/virtio/Makefile
> @@ -0,0 +1,12 @@
> +all: test mod
> +test: virtio_test
> +virtio_test: virtio_ring.o virtio_test.o
> +CFLAGS += -g -O2 -Wall -I. -I ../../usr/include/ -Wno-pointer-sign -fno-strict-overflow -MMD
> +vpath %.c ../../drivers/virtio
> +mod:
> + ${MAKE} -C `pwd`/../.. M=`pwd`/vhost_test
> +.PHONY: all test mod clean
> +clean:
> + ${RM} *.o vhost_test/*.o vhost_test/.*.cmd \
> + vhost_test/Module.symvers vhost_test/modules.order *.d
> +-include *.d
> diff --git a/tools/virtio/linux/device.h b/tools/virtio/linux/device.h
> new file mode 100644
> index 0000000..4ad7e1d
> --- /dev/null
> +++ b/tools/virtio/linux/device.h
> @@ -0,0 +1,2 @@
> +#ifndef LINUX_DEVICE_H
> +#endif
> diff --git a/tools/virtio/linux/slab.h b/tools/virtio/linux/slab.h
> new file mode 100644
> index 0000000..81baeac
> --- /dev/null
> +++ b/tools/virtio/linux/slab.h
> @@ -0,0 +1,2 @@
> +#ifndef LINUX_SLAB_H
> +#endif
> diff --git a/tools/virtio/linux/virtio.h b/tools/virtio/linux/virtio.h
> new file mode 100644
> index 0000000..669bcdd
> --- /dev/null
> +++ b/tools/virtio/linux/virtio.h
> @@ -0,0 +1,223 @@
> +#ifndef LINUX_VIRTIO_H
> +#define LINUX_VIRTIO_H
> +
> +#include <stdbool.h>
> +#include <stdlib.h>
> +#include <stddef.h>
> +#include <stdio.h>
> +#include <string.h>
> +#include <assert.h>
> +
> +#include <linux/types.h>
> +#include <errno.h>
> +
> +typedef unsigned long long dma_addr_t;
> +
> +struct scatterlist {
> + unsigned long page_link;
> + unsigned int offset;
> + unsigned int length;
> + dma_addr_t dma_address;
> +};
> +
> +struct page {
> + unsigned long long dummy;
> +};
> +
> +#define BUG_ON(__BUG_ON_cond) assert(!(__BUG_ON_cond))
> +
> +/* Physical == Virtual */
> +#define virt_to_phys(p) ((unsigned long)p)
> +#define phys_to_virt(a) ((void *)(unsigned long)(a))
> +/* Page address: Virtual / 4K */
> +#define virt_to_page(p) ((struct page*)((virt_to_phys(p) / 4096) * \
> + sizeof(struct page)))
> +#define offset_in_page(p) (((unsigned long)p) % 4096)
> +#define sg_phys(sg) ((sg->page_link & ~0x3) / sizeof(struct page) * 4096 + \
> + sg->offset)
> +static inline void sg_mark_end(struct scatterlist *sg)
> +{
> + /*
> + * Set termination bit, clear potential chain bit
> + */
> + sg->page_link |= 0x02;
> + sg->page_link &= ~0x01;
> +}
> +static inline void sg_init_table(struct scatterlist *sgl, unsigned int nents)
> +{
> + memset(sgl, 0, sizeof(*sgl) * nents);
> + sg_mark_end(&sgl[nents - 1]);
> +}
> +static inline void sg_assign_page(struct scatterlist *sg, struct page *page)
> +{
> + unsigned long page_link = sg->page_link & 0x3;
> +
> + /*
> + * In order for the low bit stealing approach to work, pages
> + * must be aligned at a 32-bit boundary as a minimum.
> + */
> + BUG_ON((unsigned long) page & 0x03);
> + sg->page_link = page_link | (unsigned long) page;
> +}
> +
> +static inline void sg_set_page(struct scatterlist *sg, struct page *page,
> + unsigned int len, unsigned int offset)
> +{
> + sg_assign_page(sg, page);
> + sg->offset = offset;
> + sg->length = len;
> +}
> +
> +static inline void sg_set_buf(struct scatterlist *sg, const void *buf,
> + unsigned int buflen)
> +{
> + sg_set_page(sg, virt_to_page(buf), buflen, offset_in_page(buf));
> +}
> +
> +static inline void sg_init_one(struct scatterlist *sg, const void *buf, unsigned int buflen)
> +{
> + sg_init_table(sg, 1);
> + sg_set_buf(sg, buf, buflen);
> +}
> +
> +typedef __u16 u16;
> +
> +typedef enum {
> + GFP_KERNEL,
> + GFP_ATOMIC,
> +} gfp_t;
> +typedef enum {
> + IRQ_NONE,
> + IRQ_HANDLED
> +} irqreturn_t;
> +
> +static inline void *kmalloc(size_t s, gfp_t gfp)
> +{
> + return malloc(s);
> +}
> +
> +static inline void kfree(void *p)
> +{
> + free(p);
> +}
> +
> +#define container_of(ptr, type, member) ({ \
> + const typeof( ((type *)0)->member ) *__mptr = (ptr); \
> + (type *)( (char *)__mptr - offsetof(type,member) );})
> +
> +#define uninitialized_var(x) x = x
> +
> +# ifndef likely
> +# define likely(x) (__builtin_expect(!!(x), 1))
> +# endif
> +# ifndef unlikely
> +# define unlikely(x) (__builtin_expect(!!(x), 0))
> +# endif
> +
> +#define pr_err(format, ...) fprintf (stderr, format, ## __VA_ARGS__)
> +#ifdef DEBUG
> +#define pr_debug(format, ...) fprintf (stderr, format, ## __VA_ARGS__)
> +#else
> +#define pr_debug(format, ...) do {} while (0)
> +#endif
> +#define dev_err(dev, format, ...) fprintf (stderr, format, ## __VA_ARGS__)
> +#define dev_warn(dev, format, ...) fprintf (stderr, format, ## __VA_ARGS__)
> +
> +/* TODO: empty stubs for now. Broken but enough for virtio_ring.c */
> +#define list_add_tail(a, b) do {} while (0)
> +#define list_del(a) do {} while (0)
> +
> +#define BIT_WORD(nr) ((nr) / BITS_PER_LONG)
> +#define BITS_PER_BYTE 8
> +#define BITS_PER_LONG (sizeof(long) * BITS_PER_BYTE)
> +#define BIT_MASK(nr) (1UL << ((nr) % BITS_PER_LONG))
> +/* TODO: Not atomic as it should be:
> + * we don't use this for anything important. */
> +static inline void clear_bit(int nr, volatile unsigned long *addr)
> +{
> + unsigned long mask = BIT_MASK(nr);
> + unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
> +
> + *p &= ~mask;
> +}
> +
> +static inline int test_bit(int nr, const volatile unsigned long *addr)
> +{
> + return 1UL & (addr[BIT_WORD(nr)] >> (nr & (BITS_PER_LONG-1)));
> +}
> +
> +/* The only feature we care to support */
> +#define virtio_has_feature(dev, feature) \
> + test_bit((feature), (dev)->features)
> +/* end of stubs */
> +
> +struct virtio_device {
> + void *dev;
> + unsigned long features[1];
> +};
> +
> +struct virtqueue {
> + /* TODO: commented as list macros are empty stubs for now.
> + * Broken but enough for virtio_ring.c
> + * struct list_head list; */
> + void (*callback)(struct virtqueue *vq);
> + const char *name;
> + struct virtio_device *vdev;
> + void *priv;
> +};
> +
> +#define EXPORT_SYMBOL_GPL(__EXPORT_SYMBOL_GPL_name) \
> + void __EXPORT_SYMBOL_GPL##__EXPORT_SYMBOL_GPL_name() { \
> +}
> +#define MODULE_LICENSE(__MODULE_LICENSE_value) \
> + const char *__MODULE_LICENSE_name = __MODULE_LICENSE_value
> +
> +#define CONFIG_SMP
> +
> +#if defined(__i386__) || defined(__x86_64__)
> +#define barrier() asm volatile("" ::: "memory")
> +#define mb() __sync_synchronize()
> +
> +#define smp_mb() mb()
> +# define smp_rmb() barrier()
> +# define smp_wmb() barrier()
> +#else
> +#error Please fill in barrier macros
> +#endif
> +
> +/* Interfaces exported by virtio_ring. */
> +int virtqueue_add_buf_gfp(struct virtqueue *vq,
> + struct scatterlist sg[],
> + unsigned int out_num,
> + unsigned int in_num,
> + void *data,
> + gfp_t gfp);
> +
> +static inline int virtqueue_add_buf(struct virtqueue *vq,
> + struct scatterlist sg[],
> + unsigned int out_num,
> + unsigned int in_num,
> + void *data)
> +{
> + return virtqueue_add_buf_gfp(vq, sg, out_num, in_num, data, GFP_ATOMIC);
> +}
> +
> +void virtqueue_kick(struct virtqueue *vq);
> +
> +void *virtqueue_get_buf(struct virtqueue *vq, unsigned int *len);
> +
> +void virtqueue_disable_cb(struct virtqueue *vq);
> +
> +bool virtqueue_enable_cb(struct virtqueue *vq);
> +
> +void *virtqueue_detach_unused_buf(struct virtqueue *vq);
> +struct virtqueue *vring_new_virtqueue(unsigned int num,
> + unsigned int vring_align,
> + struct virtio_device *vdev,
> + void *pages,
> + void (*notify)(struct virtqueue *vq),
> + void (*callback)(struct virtqueue *vq),
> + const char *name);
> +void vring_del_virtqueue(struct virtqueue *vq);
> +
> +#endif
> diff --git a/tools/virtio/vhost_test/Makefile b/tools/virtio/vhost_test/Makefile
> new file mode 100644
> index 0000000..a1d35b8
> --- /dev/null
> +++ b/tools/virtio/vhost_test/Makefile
> @@ -0,0 +1,2 @@
> +obj-m += vhost_test.o
> +EXTRA_CFLAGS += -Idrivers/vhost
> diff --git a/tools/virtio/vhost_test/vhost_test.c b/tools/virtio/vhost_test/vhost_test.c
> new file mode 100644
> index 0000000..1873518
> --- /dev/null
> +++ b/tools/virtio/vhost_test/vhost_test.c
> @@ -0,0 +1 @@
> +#include "test.c"
> diff --git a/tools/virtio/virtio_test.c b/tools/virtio/virtio_test.c
> new file mode 100644
> index 0000000..808ae86
> --- /dev/null
> +++ b/tools/virtio/virtio_test.c
> @@ -0,0 +1,248 @@
> +#define _GNU_SOURCE
> +#include <getopt.h>
> +#include <string.h>
> +#include <poll.h>
> +#include <sys/eventfd.h>
> +#include <stdlib.h>
> +#include <assert.h>
> +#include <unistd.h>
> +#include <sys/ioctl.h>
> +#include <sys/stat.h>
> +#include <sys/types.h>
> +#include <fcntl.h>
> +#include <linux/vhost.h>
> +#include <linux/virtio.h>
> +#include <linux/virtio_ring.h>
> +#include "../../drivers/vhost/test.h"
> +
> +struct vq_info {
> + int kick;
> + int call;
> + int num;
> + int idx;
> + void *ring;
> + /* copy used for control */
> + struct vring vring;
> + struct virtqueue *vq;
> +};
> +
> +struct vdev_info {
> + struct virtio_device vdev;
> + int control;
> + struct pollfd fds[1];
> + struct vq_info vqs[1];
> + int nvqs;
> + void *buf;
> + size_t buf_size;
> + struct vhost_memory *mem;
> +};
> +
> +void vq_notify(struct virtqueue *vq)
> +{
> + struct vq_info *info = vq->priv;
> + unsigned long long v = 1;
> + int r;
> + r = write(info->kick, &v, sizeof v);
> + assert(r == sizeof v);
> +}
> +
> +void vq_callback(struct virtqueue *vq)
> +{
> +}
> +
> +
> +void vhost_vq_setup(struct vdev_info *dev, struct vq_info *info)
> +{
> + struct vhost_vring_state state = { .index = info->idx };
> + struct vhost_vring_file file = { .index = info->idx };
> + unsigned long long features = dev->vdev.features[0];
> + struct vhost_vring_addr addr = {
> + .index = info->idx,
> + .desc_user_addr = (uint64_t)(unsigned long)info->vring.desc,
> + .avail_user_addr = (uint64_t)(unsigned long)info->vring.avail,
> + .used_user_addr = (uint64_t)(unsigned long)info->vring.used,
> + };
> + int r;
> + r = ioctl(dev->control, VHOST_SET_FEATURES, &features);
> + assert(r >= 0);
> + state.num = info->vring.num;
> + r = ioctl(dev->control, VHOST_SET_VRING_NUM, &state);
> + assert(r >= 0);
> + state.num = 0;
> + r = ioctl(dev->control, VHOST_SET_VRING_BASE, &state);
> + assert(r >= 0);
> + r = ioctl(dev->control, VHOST_SET_VRING_ADDR, &addr);
> + assert(r >= 0);
> + file.fd = info->kick;
> + r = ioctl(dev->control, VHOST_SET_VRING_KICK, &file);
> + assert(r >= 0);
> + file.fd = info->call;
> + r = ioctl(dev->control, VHOST_SET_VRING_CALL, &file);
> + assert(r >= 0);
> +}
> +
> +static void vq_info_add(struct vdev_info *dev, int num)
> +{
> + struct vq_info *info = &dev->vqs[dev->nvqs];
> + int r;
> + info->idx = dev->nvqs;
> + info->kick = eventfd(0, EFD_NONBLOCK);
> + info->call = eventfd(0, EFD_NONBLOCK);
> + r = posix_memalign(&info->ring, 4096, vring_size(num, 4096));
> + assert(r >= 0);
> + memset(info->ring, 0, vring_size(num, 4096));
> + vring_init(&info->vring, num, info->ring, 4096);
> + info->vq = vring_new_virtqueue(info->vring.num, 4096, &dev->vdev, info->ring,
> + vq_notify, vq_callback, "test");
> + assert(info->vq);
> + info->vq->priv = info;
> + vhost_vq_setup(dev, info);
> + dev->fds[info->idx].fd = info->call;
> + dev->fds[info->idx].events = POLLIN;
> + dev->nvqs++;
> +}
> +
> +static void vdev_info_init(struct vdev_info* dev, unsigned long long features)
> +{
> + int r;
> + memset(dev, 0, sizeof *dev);
> + dev->vdev.features[0] = features;
> + dev->vdev.features[1] = features >> 32;
> + dev->buf_size = 1024;
> + dev->buf = malloc(dev->buf_size);
> + assert(dev->buf);
> + dev->control = open("/dev/vhost-test", O_RDWR);
> + assert(dev->control >= 0);
> + r = ioctl(dev->control, VHOST_SET_OWNER, NULL);
> + assert(r >= 0);
> + dev->mem = malloc(offsetof(struct vhost_memory, regions) +
> + sizeof dev->mem->regions[0]);
> + assert(dev->mem);
> + memset(dev->mem, 0, offsetof(struct vhost_memory, regions) +
> + sizeof dev->mem->regions[0]);
> + dev->mem->nregions = 1;
> + dev->mem->regions[0].guest_phys_addr = (long)dev->buf;
> + dev->mem->regions[0].userspace_addr = (long)dev->buf;
> + dev->mem->regions[0].memory_size = dev->buf_size;
> + r = ioctl(dev->control, VHOST_SET_MEM_TABLE, dev->mem);
> + assert(r >= 0);
> +}
> +
> +/* TODO: this is pretty bad: we get a cache line bounce
> + * for the wait queue on poll and another one on read,
> + * plus the read which is there just to clear the
> + * current state. */
> +static void wait_for_interrupt(struct vdev_info *dev)
> +{
> + int i;
> + unsigned long long val;
> + poll(dev->fds, dev->nvqs, -1);
> + for (i = 0; i < dev->nvqs; ++i)
> + if (dev->fds[i].revents & POLLIN) {
> + read(dev->fds[i].fd, &val, sizeof val);
> + }
> +}
> +
> +static void run_test(struct vdev_info *dev, struct vq_info *vq, int bufs)
> +{
> + struct scatterlist sl;
> + long started = 0, completed = 0;
> + long completed_before;
> + int r, test = 1;
> + unsigned len;
> + long long spurious = 0;
> + r = ioctl(dev->control, VHOST_TEST_RUN, &test);
> + assert(r >= 0);
> + for (;;) {
> + virtqueue_disable_cb(vq->vq);
> + completed_before = completed;
> + do {
> + if (started < bufs) {
> + sg_init_one(&sl, dev->buf, dev->buf_size);
> + r = virtqueue_add_buf(vq->vq, &sl, 1, 0,
> + dev->buf + started);
> + if (likely(r >= 0)) {
> + ++started;
> + virtqueue_kick(vq->vq);
> + }
> + } else
> + r = -1;
> +
> + /* Flush out completed bufs if any */
> + if (virtqueue_get_buf(vq->vq, &len)) {
> + ++completed;
> + r = 0;
> + }
> +
> + } while (r >= 0);
> + if (completed == completed_before)
> + ++spurious;
> + assert(completed <= bufs);
> + assert(started <= bufs);
> + if (completed == bufs)
> + break;
> + if (virtqueue_enable_cb(vq->vq)) {
> + wait_for_interrupt(dev);
> + }
> + }
> + test = 0;
> + r = ioctl(dev->control, VHOST_TEST_RUN, &test);
> + assert(r >= 0);
> + fprintf(stderr, "spurious wakeus: 0x%llx\n", spurious);
> +}
> +
> +const char optstring[] = "h";
> +const struct option longopts[] = {
> + {
> + .name = "help",
> + .val = 'h',
> + },
> + {
> + .name = "indirect",
> + .val = 'I',
> + },
> + {
> + .name = "no-indirect",
> + .val = 'i',
> + },
> + {
> + }
> +};
> +
> +static void help()
> +{
> + fprintf(stderr, "Usage: virtio_test [--help] [--no-indirect]\n");
> +}
> +
> +int main(int argc, char **argv)
> +{
> + struct vdev_info dev;
> + unsigned long long features = 1ULL << VIRTIO_RING_F_INDIRECT_DESC;
> + int o;
> +
> + for (;;) {
> + o = getopt_long(argc, argv, optstring, longopts, NULL);
> + switch (o) {
> + case -1:
> + goto done;
> + case '?':
> + help();
> + exit(2);
> + case 'h':
> + help();
> + goto done;
> + case 'i':
> + features &= ~(1ULL << VIRTIO_RING_F_INDIRECT_DESC);
> + break;
> + default:
> + assert(0);
> + break;
> + }
> + }
> +
> +done:
> + vdev_info_init(&dev, features);
> + vq_info_add(&dev, 256);
> + run_test(&dev, &dev.vqs[0], 0x100000);
> + return 0;
> +}
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
Regards,
Zhi Yong Wu
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
