* Re: [PATCHv5 2/4] Add support IORING_SETUP_SQE_MIXED [not found] ` <20251021213329.784558-3-kbusch@meta.com> @ 2025-10-21 22:08 ` Jens Axboe 0 siblings, 0 replies; 3+ messages in thread From: Jens Axboe @ 2025-10-21 22:08 UTC (permalink / raw) To: Keith Busch, csander, io-uring; +Cc: Keith Busch (Adding right list) On 10/21/25 3:33 PM, Keith Busch wrote: > +/* > + * Return a 128B sqe to fill. Applications must later call io_uring_submit() > + * when it's ready to tell the kernel about it. The caller may call this > + * function multiple times before calling io_uring_submit(). > + * > + * Returns a vacant 128B sqe, or NULL if we're full. If the current tail is the > + * last entry in the ring, this function will insert a nop + skip complete such > + * that the 128b entry wraps back to the beginning of the queue for a > + * contiguous big sq entry. It's up to the caller to use a 128b opcode in order > + * for the kernel to know how to advance its sq head pointer. > + */ > +IOURINGINLINE struct io_uring_sqe *io_uring_get_sqe128_mixed(struct io_uring *ring) > + LIBURING_NOEXCEPT > +{ I would probably just name this io_uring_get_sqe128() and have it work for both MIXED and SQE128. That would make for a cleaner API for the application. > + struct io_uring_sq *sq = &ring->sq; > + unsigned head = io_uring_load_sq_head(ring), tail = sq->sqe_tail; > + struct io_uring_sqe *sqe; > + > + if (!(ring->flags & IORING_SETUP_SQE_MIXED)) > + return NULL; > + > + if (((tail + 1) & sq->ring_mask) == 0) { > + if ((tail + 2) - head >= sq->ring_entries) > + return NULL; > + > + sqe = _io_uring_get_sqe(ring); > + io_uring_prep_nop(sqe); > + sqe->flags |= IOSQE_CQE_SKIP_SUCCESS; > + tail = sq->sqe_tail; > + } else if ((tail + 1) - head >= sq->ring_entries) { > + return NULL; > + } > + > + sqe = &sq->sqes[tail & sq->ring_mask]; > + sq->sqe_tail = tail + 2; > + io_uring_initialize_sqe(sqe); > + > + return sqe; > +} Spurious newline before turn, just group them. -- Jens Axboe ^ permalink raw reply [flat|nested] 3+ messages in thread
[parent not found: <20251021213329.784558-5-kbusch@meta.com>]
* Re: [PATCHv5 4/4] Add mixed sqe test for uring commands [not found] ` <20251021213329.784558-5-kbusch@meta.com> @ 2025-10-21 22:09 ` Jens Axboe 0 siblings, 0 replies; 3+ messages in thread From: Jens Axboe @ 2025-10-21 22:09 UTC (permalink / raw) To: Keith Busch, csander, io-uring; +Cc: Keith Busch > + ret = io_uring_wait_cqe(ring, &cqe); > + if (ret < 0) > + fprintf(stderr, "wait completion %d\n", ret); > + else if (cqe->user_data != seq) > + fprintf(stderr, "Unexpected user_data: %ld\n", (long) cqe->user_data); > + else { > + io_uring_cqe_seen(ring, cqe); > + return T_EXIT_PASS; > + } > + return T_EXIT_FAIL; All braces if one has braces. In a few different spots. Outside of those little nits and the previous comment, I think this is looking fine. -- Jens Axboe ^ permalink raw reply [flat|nested] 3+ messages in thread
* [PATCHv5 0/4] liburing: support for mix sized sqe's @ 2025-10-13 18:00 Keith Busch 2025-10-13 18:00 ` [PATCHv5 4/4] Add mixed sqe test for uring commands Keith Busch 0 siblings, 1 reply; 3+ messages in thread From: Keith Busch @ 2025-10-13 18:00 UTC (permalink / raw) To: io-uring, axboe, csander; +Cc: Keith Busch From: Keith Busch <kbusch@kernel.org> Changes from v3 and v4: - Removed the old artifacts leaked from previous versions - A prep patch introducing a helper for uring_cmd Keith Busch (4): liburing: provide uring_cmd prep function Add support IORING_SETUP_SQE_MIXED Add nop testing for IORING_SETUP_SQE_MIXED Add mixed sqe test for uring commands src/include/liburing.h | 82 ++++++++++++++++++- src/include/liburing/io_uring.h | 8 ++ test/Makefile | 3 + test/io_uring_passthrough.c | 14 +--- test/sqe-mixed-bad-wrap.c | 87 ++++++++++++++++++++ test/sqe-mixed-nop.c | 82 +++++++++++++++++++ test/sqe-mixed-uring_cmd.c | 140 ++++++++++++++++++++++++++++++++ 7 files changed, 402 insertions(+), 14 deletions(-) create mode 100644 test/sqe-mixed-bad-wrap.c create mode 100644 test/sqe-mixed-nop.c create mode 100644 test/sqe-mixed-uring_cmd.c -- 2.47.3 ^ permalink raw reply [flat|nested] 3+ messages in thread
* [PATCHv5 4/4] Add mixed sqe test for uring commands 2025-10-13 18:00 [PATCHv5 0/4] liburing: support for mix sized sqe's Keith Busch @ 2025-10-13 18:00 ` Keith Busch 0 siblings, 0 replies; 3+ messages in thread From: Keith Busch @ 2025-10-13 18:00 UTC (permalink / raw) To: io-uring, axboe, csander; +Cc: Keith Busch From: Keith Busch <kbusch@kernel.org> Signed-off-by: Keith Busch <kbusch@kernel.org> --- test/Makefile | 1 + test/sqe-mixed-uring_cmd.c | 140 +++++++++++++++++++++++++++++++++++++ 2 files changed, 141 insertions(+) diff --git a/test/Makefile b/test/Makefile index 2c250c81..2b2e3967 100644 --- a/test/Makefile +++ b/test/Makefile @@ -236,6 +236,7 @@ test_srcs := \ sq-space_left.c \ sqe-mixed-nop.c \ sqe-mixed-bad-wrap.c \ + sqe-mixed-uring_cmd.c \ sqwait.c \ stdout.c \ submit-and-wait.c \ diff --git a/test/sqe-mixed-uring_cmd.c b/test/sqe-mixed-uring_cmd.c index e69de29b..7ac5f4ab 100644 --- a/test/sqe-mixed-uring_cmd.c +++ b/test/sqe-mixed-uring_cmd.c @@ -0,0 +1,140 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Description: mixed sqes utilizing basic nop and io_uring passthrough commands + */ +#include <errno.h> +#include <stdio.h> +#include <string.h> + +#include "helpers.h" +#include "liburing.h" +#include "nvme.h" + +#define len 0x1000 +static unsigned char buf[len]; +static int seq; + +static int test_single_nop(struct io_uring *ring) +{ + struct io_uring_cqe *cqe; + struct io_uring_sqe *sqe; + int ret; + + sqe = io_uring_get_sqe(ring); + if (!sqe) { + fprintf(stderr, "get sqe failed\n"); + return T_EXIT_FAIL; + } + + io_uring_prep_nop(sqe); + sqe->user_data = ++seq; + + ret = io_uring_submit(ring); + if (ret <= 0) { + fprintf(stderr, "sqe submit failed: %d\n", ret); + return T_EXIT_FAIL; + } + + ret = io_uring_wait_cqe(ring, &cqe); + if (ret < 0) + fprintf(stderr, "wait completion %d\n", ret); + else if (cqe->user_data != seq) + fprintf(stderr, "Unexpected user_data: %ld\n", (long) cqe->user_data); + else { + io_uring_cqe_seen(ring, cqe); + return T_EXIT_PASS; + } + return T_EXIT_FAIL; +} + +static int test_single_nvme_read(struct io_uring *ring, int fd) +{ + struct nvme_uring_cmd *cmd; + struct io_uring_cqe *cqe; + struct io_uring_sqe *sqe; + int ret; + + sqe = io_uring_get_sqe128_mixed(ring); + if (!sqe) { + fprintf(stderr, "get sqe failed\n"); + return T_EXIT_FAIL; + } + + io_uring_prep_uring_cmd128(sqe, NVME_URING_CMD_IO, fd); + sqe->user_data = ++seq; + + cmd = (struct nvme_uring_cmd *)sqe->cmd; + memset(cmd, 0, sizeof(struct nvme_uring_cmd)); + cmd->opcode = nvme_cmd_read; + cmd->cdw12 = (len >> lba_shift) - 1; + cmd->addr = (__u64)(uintptr_t)buf; + cmd->data_len = len; + cmd->nsid = nsid; + + ret = io_uring_submit(ring); + if (ret <= 0) { + fprintf(stderr, "sqe submit failed: %d\n", ret); + return T_EXIT_FAIL; + } + + ret = io_uring_wait_cqe(ring, &cqe); + if (ret < 0) + fprintf(stderr, "wait completion %d\n", ret); + else if (cqe->res != 0) + fprintf(stderr, "cqe res %d, wanted 0\n", cqe->res); + else if (cqe->user_data != seq) + fprintf(stderr, "Unexpected user_data: %ld\n", (long) cqe->user_data); + else { + io_uring_cqe_seen(ring, cqe); + return T_EXIT_PASS; + } + return T_EXIT_FAIL; +} + +int main(int argc, char *argv[]) +{ + struct io_uring ring; + int fd, ret, i; + + if (argc < 2) + return T_EXIT_SKIP; + + ret = nvme_get_info(argv[1]); + if (ret) + return T_EXIT_SKIP; + + fd = open(argv[1], O_RDONLY); + if (fd < 0) { + if (errno == EACCES || errno == EPERM) + return T_EXIT_SKIP; + perror("file open"); + return T_EXIT_FAIL; + } + + ret = io_uring_queue_init(8, &ring, + IORING_SETUP_CQE_MIXED | IORING_SETUP_SQE_MIXED); + if (ret) { + if (ret == -EINVAL) + ret = T_EXIT_SKIP; + else { + fprintf(stderr, "ring setup failed: %d\n", ret); + ret = T_EXIT_FAIL; + } + goto close; + } + + for (i = 0; i < 32; i++) { + if (i & 1) + ret = test_single_nvme_read(&ring, fd); + else + ret = test_single_nop(&ring); + + if (ret) + break; + } + + io_uring_queue_exit(&ring); +close: + close(fd); + return ret; +} -- 2.47.3 ^ permalink raw reply related [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-10-21 22:09 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <20251021213329.784558-1-kbusch@meta.com>
[not found] ` <20251021213329.784558-3-kbusch@meta.com>
2025-10-21 22:08 ` [PATCHv5 2/4] Add support IORING_SETUP_SQE_MIXED Jens Axboe
[not found] ` <20251021213329.784558-5-kbusch@meta.com>
2025-10-21 22:09 ` [PATCHv5 4/4] Add mixed sqe test for uring commands Jens Axboe
2025-10-13 18:00 [PATCHv5 0/4] liburing: support for mix sized sqe's Keith Busch
2025-10-13 18:00 ` [PATCHv5 4/4] Add mixed sqe test for uring commands Keith Busch
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).