The SQ and CQ ring heads are read twice - once for verifying that it's
within bounds, and once inside the loops copying SQE and CQE entries.
This is technically incorrect, in case the values could get modified
in between verifying them and using them in the copy loop. While this
won't lead to anything truly nefarious, it may cause longer loop times
for the copies than expected.

Read the ring head values once, and use the verified value in the copy
loops.

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

It can be a bit hard to tell which parts of io_register_resize_rings()
are operating on shared memory, and which ones are not. And anything
reading or writing to those regions should really use the read/write
once primitives.

Hence add those, ensuring sanity in how this memory is accessed, and
helping document the shared nature of it.

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

Normally the kernel would not expect an application to modify any of
the data shared with the kernel during a resize operation, but of
course the kernel cannot always assume good intent on behalf of the
application.

As part of resizing the rings, existing SQEs and CQEs are copied over
to the new storage. Resizing uses the masks in the newly allocated
shared storage to index the arrays, however it's possible that malicious
userspace could modify these after they have been sanity checked.

Use the validated and locally stored CQ and SQ ring sizing for masking
to ensure the values are both stable and valid.

Fixes: 79cfe9e59c2a ("io_uring/register: add IORING_REGISTER_RESIZE_RINGS")
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

With DEFER_TASKRUN, we know the ring can't be both waited upon and
resized at the same time. This is important for CQ resizing. Allowing SQ
ring resizing is more trivial, but isn't the interesting use case. Hence
limit ring resizing in general to DEFER_TASKRUN only for now. This isn't
a huge problem as CQ ring resizing is generally the most useful on
networking type of workloads where it can be hard to size the ring
appropriately upfront, and those should be using DEFER_TASKRUN for
better performance.

Fixes: 79cfe9e59c2a ("io_uring/register: add IORING_REGISTER_RESIZE_RINGS")
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Syz reports:

BUG: KCSAN: data-race in __se_sys_io_uring_register / io_sqe_files_register

read-write to 0xffff8881021940b8 of 4 bytes by task 5923 on cpu 1:
 io_sqe_files_register+0x2c4/0x3b0 io_uring/rsrc.c:713
 __io_uring_register io_uring/register.c:403 [inline]
 __do_sys_io_uring_register io_uring/register.c:611 [inline]
 __se_sys_io_uring_register+0x8d0/0x1280 io_uring/register.c:591
 __x64_sys_io_uring_register+0x55/0x70 io_uring/register.c:591
 x64_sys_call+0x202/0x2d60 arch/x86/include/generated/asm/syscalls_64.h:428
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xc9/0x1c0 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

read to 0xffff8881021940b8 of 4 bytes by task 5924 on cpu 0:
 __do_sys_io_uring_register io_uring/register.c:613 [inline]
 __se_sys_io_uring_register+0xe4a/0x1280 io_uring/register.c:591
 __x64_sys_io_uring_register+0x55/0x70 io_uring/register.c:591
 x64_sys_call+0x202/0x2d60 arch/x86/include/generated/asm/syscalls_64.h:428
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xc9/0x1c0 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

Which should be due to reading the table size after unlock. We don't
care much as it's just to print it in trace, but we might as well do it
under the lock.

Reported-by: syzbot+5a486fef3de40e0d8c76@syzkaller.appspotmail.com
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8233af2886a37b57f79e444e3db88fcfda1817ac.1731942203.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Now we've got a more generic region registration API, place
IORING_ENTER_EXT_ARG_REG and re-enable it.

First, the user has to register a region with the
IORING_MEM_REGION_REG_WAIT_ARG flag set. It can only be done for a
ring in a disabled state, aka IORING_SETUP_R_DISABLED, to avoid races
with already running waiters. With that we should have stable constant
values for ctx->cq_wait_{size,arg} in io_get_ext_arg_reg() and hence no
READ_ONCE required.

The other API difference is that we're now passing byte offsets instead
of indexes. The user _must_ align all offsets / pointers to the native
word size, failing to do so might but not necessarily has to lead to a
failure usually returned as -EFAULT. liburing will be hiding this
details from users.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/81822c1b4ffbe8ad391b4f9ad1564def0d26d990.1731689588.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Regions will serve multiple purposes. First, with it we can decouple
ring/etc. object creation from registration / mapping of the memory they
will be placed in. We already have hacks that allow to put both SQ and
CQ into the same huge page, in the future we should be able to:

region = create_region(io_ring);
create_pbuf_ring(io_uring, region, offset=0);
create_pbuf_ring(io_uring, region, offset=N);

The second use case is efficiently passing parameters. The following
patch enables back on top of regions IORING_ENTER_EXT_ARG_REG, which
optimises wait arguments. It'll also be useful for request arguments
replacing iovecs, msghdr, etc. pointers. Eventually it would also be
handy for BPF as well if it comes to fruition.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0798cf3a14fad19cfc96fc9feca5f3e11481691d.1731689588.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Disable wait argument registration as it'll be replaced with a more
generic feature. We'll still need IORING_ENTER_EXT_ARG_REG parsing
in a few commits so leave it be.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/70b1d1d218c41ba77a76d1789c8641dab0b0563e.1731689588.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

For files, there's nr_user_files/file_table/file_data, and buffers have
nr_user_bufs/user_bufs/buf_data. There's no reason why file_table and
file_data can't be the same thing, and ditto for the buffer side. That
gets rid of more io_ring_ctx state that's in two spots rather than just
being in one spot, as it should be. Put all the registered file data in
one locations, and ditto on the buffer front.

This also avoids having both io_rsrc_data->nodes being an allocated
array, and ->user_bufs[] or ->file_table.nodes. There's no reason to
have this information duplicated. Keep it in one spot, io_rsrc_data,
along with how many resources are available.

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

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>