io_uring: add support for fixed wait regions

Generally applications have 1 or a few waits of waiting, yet they pass
in a struct io_uring_getevents_arg every time. This needs to get copied
and, in turn, the timeout value needs to get copied.

Rather than do this for every invocation, allow the application to
register a fixed set of wait regions that can simply be indexed when
asking the kernel to wait on events.

At ring setup time, the application can register a number of these wait
regions and initialize region/index 0 upfront:

	struct io_uring_reg_wait *reg;

	reg = io_uring_setup_reg_wait(ring, nr_regions, &ret);

	/* set timeout and mark as set, sigmask/sigmask_sz as needed */
	reg->ts.tv_sec = 0;
	reg->ts.tv_nsec = 100000;
	reg->flags = IORING_REG_WAIT_TS;

where nr_regions >= 1 && nr_regions <= PAGE_SIZE / sizeof(*reg). The
above initializes index 0, but 63 other regions can be initialized,
if needed. Now, instead of doing:

	struct __kernel_timespec timeout = { .tv_nsec = 100000, };

	io_uring_submit_and_wait_timeout(ring, &cqe, nr, &t, NULL);

to wait for events for each submit_and_wait, or just wait, operation, it
can just reference the above region at offset 0 and do:

	io_uring_submit_and_wait_reg(ring, &cqe, nr, 0);

to achieve the same goal of waiting 100usec without needing to copy
both struct io_uring_getevents_arg (24b) and struct __kernel_timeout
(16b) for each invocation. Struct io_uring_reg_wait looks as follows:

struct io_uring_reg_wait {
	struct __kernel_timespec	ts;
	__u32				min_wait_usec;
	__u32				flags;
	__u64				sigmask;
	__u32				sigmask_sz;
	__u32				pad[3];
	__u64				pad2[2];
};

embedding the timeout itself in the region, rather than passing it as
a pointer as well. Note that the signal mask is still passed as a
pointer, both for compatability reasons, but also because there doesn't
seem to be a lot of high frequency waits scenarios that involve setting
and resetting the signal mask for each wait.

The application is free to modify any region before a wait call, or it
can use keep multiple regions with different settings to avoid needing to
modify the same one for wait calls. Up to a page size of regions is mapped
by default, allowing PAGE_SIZE / 64 available regions for use.

The registered region must fit within a page. On a 4kb page size system,
that allows for 64 wait regions if a full page is used, as the size of
struct io_uring_reg_wait is 64b. The region registered must be aligned
to io_uring_reg_wait in size. It's valid to register less than 64
entries.

In network performance testing with zero-copy, this reduced the time
spent waiting on the TX side from 3.12% to 0.3% and the RX side from 4.4%
to 0.3%.

Wait regions are fixed for the lifetime of the ring - once registered,
they are persistent until the ring is torn down. The regions support
minimum wait timeout as well as the regular waits.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

1 files changed, 82 insertions, 0 deletions

diff --git a/io_uring/register.c b/io_uring/register.c
index fc6c94d694b2..1eb686eaa310 100644
--- a/io_uring/register.c
+++ b/io_uring/register.c
@@ -570,6 +570,82 @@ out:
 	return ret;
 }
 
+void io_unregister_cqwait_reg(struct io_ring_ctx *ctx)
+{
+	unsigned short npages = 1;
+
+	if (!ctx->cq_wait_page)
+		return;
+
+	io_pages_unmap(ctx->cq_wait_arg, &ctx->cq_wait_page, &npages, true);
+	ctx->cq_wait_arg = NULL;
+	if (ctx->user)
+		__io_unaccount_mem(ctx->user, 1);
+}
+
+/*
+ * Register a page holding N entries of struct io_uring_reg_wait, which can
+ * be used via io_uring_enter(2) if IORING_GETEVENTS_EXT_ARG_REG is set.
+ * If that is set with IORING_GETEVENTS_EXT_ARG, then instead of passing
+ * in a pointer for a struct io_uring_getevents_arg, an index into this
+ * registered array is passed, avoiding two (arg + timeout) copies per
+ * invocation.
+ */
+static int io_register_cqwait_reg(struct io_ring_ctx *ctx, void __user *uarg)
+{
+	struct io_uring_cqwait_reg_arg arg;
+	struct io_uring_reg_wait *reg;
+	struct page **pages;
+	unsigned long len;
+	int nr_pages, poff;
+	int ret;
+
+	if (ctx->cq_wait_page || ctx->cq_wait_arg)
+		return -EBUSY;
+	if (copy_from_user(&arg, uarg, sizeof(arg)))
+		return -EFAULT;
+	if (!arg.nr_entries || arg.flags)
+		return -EINVAL;
+	if (arg.struct_size != sizeof(*reg))
+		return -EINVAL;
+	if (check_mul_overflow(arg.struct_size, arg.nr_entries, &len))
+		return -EOVERFLOW;
+	if (len > PAGE_SIZE)
+		return -EINVAL;
+	/* offset + len must fit within a page, and must be reg_wait aligned */
+	poff = arg.user_addr & ~PAGE_MASK;
+	if (len + poff > PAGE_SIZE)
+		return -EINVAL;
+	if (poff % arg.struct_size)
+		return -EINVAL;
+
+	pages = io_pin_pages(arg.user_addr, len, &nr_pages);
+	if (IS_ERR(pages))
+		return PTR_ERR(pages);
+	ret = -EINVAL;
+	if (nr_pages != 1)
+		goto out_free;
+	if (ctx->user) {
+		ret = __io_account_mem(ctx->user, 1);
+		if (ret)
+			goto out_free;
+	}
+
+	reg = vmap(pages, 1, VM_MAP, PAGE_KERNEL);
+	if (reg) {
+		ctx->cq_wait_index = arg.nr_entries - 1;
+		WRITE_ONCE(ctx->cq_wait_page, pages);
+		WRITE_ONCE(ctx->cq_wait_arg, (void *) reg + poff);
+		return 0;
+	}
+	ret = -ENOMEM;
+	if (ctx->user)
+		__io_unaccount_mem(ctx->user, 1);
+out_free:
+	io_pages_free(&pages, nr_pages);
+	return ret;
+}
+
 static int __io_uring_register(struct io_ring_ctx *ctx, unsigned opcode,
 			       void __user *arg, unsigned nr_args)
 	__releases(ctx->uring_lock)
@@ -764,6 +840,12 @@ static int __io_uring_register(struct io_ring_ctx *ctx, unsigned opcode,
 			break;
 		ret = io_register_resize_rings(ctx, arg);
 		break;
+	case IORING_REGISTER_CQWAIT_REG:
+		ret = -EINVAL;
+		if (!arg || nr_args != 1)
+			break;
+		ret = io_register_cqwait_reg(ctx, arg);
+		break;
 	default:
 		ret = -EINVAL;
 		break;