io_uring/register: limit ring resizing to DEFER_TASKRUN

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>

diff --git a/io_uring/register.c b/io_uring/register.c
index 1e99c783abdf6e0bd066e6f0da873f316e5bf130..fdd44914c39c7c7457b1775f03bdb5bb1ed86b0a 100644 (file)
--- a/io_uring/register.c
+++ b/io_uring/register.c
@@ -414,6 +414,9 @@ static int io_register_resize_rings(struct io_ring_ctx *ctx, void __user *arg)
        if (ctx->flags & IORING_SETUP_SINGLE_ISSUER &&
            current != ctx->submitter_task)
                return -EEXIST;
+       /* limited to DEFER_TASKRUN for now */
+       if (!(ctx->flags & IORING_SETUP_DEFER_TASKRUN))
+               return -EINVAL;
        if (copy_from_user(&p, arg, sizeof(p)))
                return -EFAULT;
        if (p.flags & ~RESIZE_FLAGS)