diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2026-07-31 16:14:19 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2026-07-31 16:14:19 -0700 |
| commit | 5d0c32d6ec00cf155d2263b82ec255faa3888c9f (patch) | |
| tree | 7ce5ab4c1252a3af6d96dd53c90ffb4219b854de /io_uring/net.c | |
| parent | 680d49d84cb83ae5836317c00d0098aee47357d9 (diff) | |
| parent | bc0e8faf90e776a2f1f3967a04e8091e6bdb4977 (diff) | |
Merge tag 'io_uring-7.2-20260731' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring fixes from Jens Axboe:
- Fix for a bug in how length caps are handled in multishot, and along
with it, a generic fix for avoiding these kinds of conversion issues
in the future.
- Ensure that task restrictions are always preserved across exec.
- Revert of the io_uring controlled epoll restriction, which disallowed
nested contexts. Turns out that libuv is already using it like that,
so we cannot simply remove it, sadly.
- Fix for a reference leak in the zcrx code.
* tag 'io_uring-7.2-20260731' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
io_uring: preserve task restrictions across exec
io_uring/zcrx: don't clear master_ctx from the import path
Revert "io_uring/epoll: disallow adding an epoll file to an epoll context"
io_uring/kbuf: cap buffer selection length at MAX_RW_COUNT
io_uring/net: initialize mshot_len for send
Diffstat (limited to 'io_uring/net.c')
| -rw-r--r-- | io_uring/net.c | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/io_uring/net.c b/io_uring/net.c index 00a7df803b99..a74d15f7b7d2 100644 --- a/io_uring/net.c +++ b/io_uring/net.c @@ -445,6 +445,7 @@ int io_sendmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) req->flags |= REQ_F_NOWAIT; if (req->flags & REQ_F_BUFFER_SELECT) sr->buf_group = req->buf_index; + sr->mshot_total_len = sr->mshot_len = 0; if (sr->flags & IORING_RECVSEND_BUNDLE) { if (req->opcode == IORING_OP_SENDMSG) return -EINVAL; |