Hi Linus,
On top of everything else, this adds support for passthrough for
io_uring. The initial feature for this is NVMe passthrough support,
which allows non-filesystem based IO commands and admin commands. To
support this, io_uring grows support for SQE and CQE members that are
twice as big, allowing to pass in a full NVMe command without having to
copy data around. And to complete with more than just a single 32-bit
value as the output.
This will cause a merge conflict as well, with the provided buffer
change from the core branch, and adding CQE32 support for NOP in this
branch. Resolution:
diff --cc fs/io_uring.c
index 1015dd49e7e5,c5a476e6c068..395d3a921b53
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@@ -4997,17 -5169,11 +5256,23 @@@ static int io_nop_prep(struct io_kiocb
*/
static int io_nop(struct io_kiocb *req, unsigned int issue_flags)
{
++ unsigned int cflags;
+ void __user *buf;
+
+ if (req->flags & REQ_F_BUFFER_SELECT) {
+ size_t len = 1;
+
+ buf = io_buffer_select(req, &len, issue_flags);
+ if (!buf)
+ return -ENOBUFS;
+ }
+
- __io_req_complete(req, issue_flags, 0, io_put_kbuf(req, issue_flags));
++ cflags = io_put_kbuf(req, issue_flags);
+ if (!(req->ctx->flags & IORING_SETUP_CQE32))
- __io_req_complete(req, issue_flags, 0, 0);
++ __io_req_complete(req, issue_flags, 0, cflags);
+ else
- __io_req_complete32(req, issue_flags, 0, 0, req->nop.extra1,
- req->nop.extra2);
++ __io_req_complete32(req, issue_flags, 0, cflags,
++ req->nop.extra1, req->nop.extra2);
return 0;
}
Please pull!
The following changes since commit 7ccba24d3bc084d891def1a6fea504e4cb327a8c:
io_uring: don't clear req->kbuf when buffer selection is done (2022-05-09 06:29:06 -0600)
are available in the Git repository at:
git://git.kernel.dk/linux-block.git tags/for-5.19/io_uring-passthrough-2022-05-22
for you to fetch changes up to 3fe07bcd800d6e5e4e4263ca2564d69095c157bf:
io_uring: cleanup handling of the two task_work lists (2022-05-21 09:17:05 -0600)
----------------------------------------------------------------
for-5.19/io_uring-passthrough-2022-05-22
----------------------------------------------------------------
Anuj Gupta (1):
nvme: add vectored-io support for uring-cmd
Christoph Hellwig (1):
nvme: refactor nvme_submit_user_cmd()
Jens Axboe (6):
Merge branch 'for-5.19/io_uring' into for-5.19/io_uring-passthrough
Merge branch 'for-5.19/io_uring-socket' into for-5.19/io_uring-passthrough
io_uring: add support for 128-byte SQEs
fs,io_uring: add infrastructure for uring-cmd
block: wire-up support for passthrough plugging
io_uring: cleanup handling of the two task_work lists
Kanchan Joshi (3):
nvme: wire-up uring-cmd support for io-passthru on char-device.
nvme: helper for uring-passthrough checks
nvme: enable uring-passthrough for admin commands
Ming Lei (1):
blk-mq: fix passthrough plugging
Stefan Roesch (12):
io_uring: support CQE32 in io_uring_cqe
io_uring: store add. return values for CQE32
io_uring: change ring size calculation for CQE32
io_uring: add CQE32 setup processing
io_uring: add CQE32 completion processing
io_uring: modify io_get_cqe for CQE32
io_uring: flush completions for CQE32
io_uring: overflow processing for CQE32
io_uring: add tracing for additional CQE32 fields
io_uring: support CQE32 in /proc info
io_uring: enable CQE32
io_uring: support CQE32 for nop operation
block/blk-mq.c | 109 ++++----
drivers/nvme/host/core.c | 2 +
drivers/nvme/host/ioctl.c | 278 +++++++++++++++++++-
drivers/nvme/host/multipath.c | 1 +
drivers/nvme/host/nvme.h | 5 +
fs/io_uring.c | 444 ++++++++++++++++++++++++++------
include/linux/fs.h | 2 +
include/linux/io_uring.h | 33 +++
include/trace/events/io_uring.h | 18 +-
include/uapi/linux/io_uring.h | 24 +-
include/uapi/linux/nvme_ioctl.h | 28 ++
11 files changed, 806 insertions(+), 138 deletions(-)
--
Jens Axboe
