linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Joanne Koong <joannelkoong@gmail.com>
To: miklos@szeredi.hu, axboe@kernel.dk
Cc: bschubert@ddn.com, asml.silence@gmail.com,
	io-uring@vger.kernel.org, csander@purestorage.com,
	xiaobing.li@samsung.com, linux-fsdevel@vger.kernel.org
Subject: [PATCH v1 04/30] io_uring/kbuf: add mmap support for kernel-managed buffer rings
Date: Tue,  2 Dec 2025 16:34:59 -0800	[thread overview]
Message-ID: <20251203003526.2889477-5-joannelkoong@gmail.com> (raw)
In-Reply-To: <20251203003526.2889477-1-joannelkoong@gmail.com>

Add support for mmapping kernel-managed buffer rings (kmbuf) to
userspace, allowing applications to access the kernel-allocated buffers.

Similar to application-provided buffer rings (pbuf), kmbuf rings use the
buffer group ID encoded in the mmap offset to identify which buffer ring
to map. The implementation follows the same pattern as pbuf rings.

New mmap offset constants are introduced:
  - IORING_OFF_KMBUF_RING (0x88000000): Base offset for kmbuf mappings
  - IORING_OFF_KMBUF_SHIFT (16): Shift value to encode buffer group ID

The mmap offset is calculated during registration, encoding the bgid
shifted by IORING_OFF_KMBUF_SHIFT. The io_buf_get_region() helper
retrieves the appropriate region.

This allows userspace to mmap the kernel-allocated buffer region and
access the buffers directly.

Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
---
 include/uapi/linux/io_uring.h |  2 ++
 io_uring/kbuf.c               | 11 +++++++++--
 io_uring/kbuf.h               |  5 +++--
 io_uring/memmap.c             |  5 ++++-
 4 files changed, 18 insertions(+), 5 deletions(-)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 589755a4e2b4..96e936503ef6 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -533,6 +533,8 @@ struct io_uring_cqe {
 #define IORING_OFF_SQES			0x10000000ULL
 #define IORING_OFF_PBUF_RING		0x80000000ULL
 #define IORING_OFF_PBUF_SHIFT		16
+#define IORING_OFF_KMBUF_RING		0x88000000ULL
+#define IORING_OFF_KMBUF_SHIFT		16
 #define IORING_OFF_MMAP_MASK		0xf8000000ULL
 
 /*
diff --git a/io_uring/kbuf.c b/io_uring/kbuf.c
index 1668718ac8fd..619bba43dda3 100644
--- a/io_uring/kbuf.c
+++ b/io_uring/kbuf.c
@@ -766,16 +766,23 @@ int io_register_pbuf_status(struct io_ring_ctx *ctx, void __user *arg)
 	return 0;
 }
 
-struct io_mapped_region *io_pbuf_get_region(struct io_ring_ctx *ctx,
-					    unsigned int bgid)
+struct io_mapped_region *io_buf_get_region(struct io_ring_ctx *ctx,
+					   unsigned int bgid,
+					   bool kernel_managed)
 {
 	struct io_buffer_list *bl;
+	bool is_kernel_managed;
 
 	lockdep_assert_held(&ctx->mmap_lock);
 
 	bl = xa_load(&ctx->io_bl_xa, bgid);
 	if (!bl || !(bl->flags & IOBL_BUF_RING))
 		return NULL;
+
+	is_kernel_managed = !!(bl->flags & IOBL_KERNEL_MANAGED);
+	if (is_kernel_managed != kernel_managed)
+		return NULL;
+
 	return &bl->region;
 }
 
diff --git a/io_uring/kbuf.h b/io_uring/kbuf.h
index 62c80a1ebf03..11d165888b8e 100644
--- a/io_uring/kbuf.h
+++ b/io_uring/kbuf.h
@@ -88,8 +88,9 @@ unsigned int __io_put_kbufs(struct io_kiocb *req, struct io_buffer_list *bl,
 bool io_kbuf_commit(struct io_kiocb *req,
 		    struct io_buffer_list *bl, int len, int nr);
 
-struct io_mapped_region *io_pbuf_get_region(struct io_ring_ctx *ctx,
-					    unsigned int bgid);
+struct io_mapped_region *io_buf_get_region(struct io_ring_ctx *ctx,
+					   unsigned int bgid,
+					   bool kernel_managed);
 
 static inline bool io_kbuf_recycle_ring(struct io_kiocb *req,
 					struct io_buffer_list *bl)
diff --git a/io_uring/memmap.c b/io_uring/memmap.c
index a46b027882f8..1832ef923e99 100644
--- a/io_uring/memmap.c
+++ b/io_uring/memmap.c
@@ -357,7 +357,10 @@ static struct io_mapped_region *io_mmap_get_region(struct io_ring_ctx *ctx,
 		return &ctx->sq_region;
 	case IORING_OFF_PBUF_RING:
 		id = (offset & ~IORING_OFF_MMAP_MASK) >> IORING_OFF_PBUF_SHIFT;
-		return io_pbuf_get_region(ctx, id);
+		return io_buf_get_region(ctx, id, false);
+	case IORING_OFF_KMBUF_RING:
+		id = (offset & ~IORING_OFF_MMAP_MASK) >> IORING_OFF_KMBUF_SHIFT;
+		return io_buf_get_region(ctx, id, true);
 	case IORING_MAP_OFF_PARAM_REGION:
 		return &ctx->param_region;
 	case IORING_MAP_OFF_ZCRX_REGION:
-- 
2.47.3


  parent reply	other threads:[~2025-12-03  0:36 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-12-03  0:34 [PATCH v1 00/30] fuse/io-uring: add kernel-managed buffer rings and zero-copy Joanne Koong
2025-12-03  0:34 ` [PATCH v1 01/30] io_uring/kbuf: refactor io_buf_pbuf_register() logic into generic helpers Joanne Koong
2025-12-03  0:34 ` [PATCH v1 02/30] io_uring/kbuf: rename io_unregister_pbuf_ring() to io_unregister_buf_ring() Joanne Koong
2025-12-03  0:34 ` [PATCH v1 03/30] io_uring/kbuf: add support for kernel-managed buffer rings Joanne Koong
2025-12-03  0:34 ` Joanne Koong [this message]
2025-12-03  0:35 ` [PATCH v1 05/30] io_uring/kbuf: support kernel-managed buffer rings in buffer selection Joanne Koong
2025-12-03  0:35 ` [PATCH v1 06/30] io_uring/kbuf: add buffer ring pinning/unpinning Joanne Koong
2025-12-03  4:13   ` Caleb Sander Mateos
2025-12-04 18:41     ` Joanne Koong
2025-12-03  0:35 ` [PATCH v1 07/30] io_uring/rsrc: add fixed buffer table pinning/unpinning Joanne Koong
2025-12-03  4:49   ` Caleb Sander Mateos
2025-12-03 22:52     ` Joanne Koong
2025-12-04  1:24       ` Caleb Sander Mateos
2025-12-04 20:07         ` Joanne Koong
2025-12-10  3:35           ` Caleb Sander Mateos
2025-12-13  6:07             ` Joanne Koong
2025-12-03  0:35 ` [PATCH v1 08/30] io_uring/kbuf: add recycling for pinned kernel managed buffer rings Joanne Koong
2025-12-03  0:35 ` [PATCH v1 09/30] io_uring: add io_uring_cmd_import_fixed_index() Joanne Koong
2025-12-03 21:43   ` Caleb Sander Mateos
2025-12-04 18:56     ` Joanne Koong
2025-12-05 16:56       ` Caleb Sander Mateos
2025-12-05 23:28         ` Joanne Koong
2025-12-11  2:57           ` Caleb Sander Mateos
2025-12-03  0:35 ` [PATCH v1 10/30] io_uring/kbuf: add io_uring_is_kmbuf_ring() Joanne Koong
2025-12-03  0:35 ` [PATCH v1 11/30] io_uring/kbuf: return buffer id in buffer selection Joanne Koong
2025-12-03 21:53   ` Caleb Sander Mateos
2025-12-04 19:22     ` Joanne Koong
2025-12-04 21:57       ` Caleb Sander Mateos
2025-12-03  0:35 ` [PATCH v1 12/30] io_uring/kbuf: export io_ring_buffer_select() Joanne Koong
2025-12-03  0:35 ` [PATCH v1 13/30] io_uring/cmd: set selected buffer index in __io_uring_cmd_done() Joanne Koong
2025-12-03  0:35 ` [PATCH v1 14/30] io_uring: add release callback for ring death Joanne Koong
2025-12-03 22:25   ` Caleb Sander Mateos
2025-12-03 22:54     ` Joanne Koong
2025-12-03  0:35 ` [PATCH v1 15/30] fuse: refactor io-uring logic for getting next fuse request Joanne Koong
2025-12-03  0:35 ` [PATCH v1 16/30] fuse: refactor io-uring header copying to ring Joanne Koong
2025-12-03  0:35 ` [PATCH v1 17/30] fuse: refactor io-uring header copying from ring Joanne Koong
2025-12-03  0:35 ` [PATCH v1 18/30] fuse: use enum types for header copying Joanne Koong
2025-12-03  0:35 ` [PATCH v1 19/30] fuse: refactor setting up copy state for payload copying Joanne Koong
2025-12-03  0:35 ` [PATCH v1 20/30] fuse: support buffer copying for kernel addresses Joanne Koong
2025-12-03  0:35 ` [PATCH v1 21/30] fuse: add io-uring kernel-managed buffer ring Joanne Koong
2025-12-03  0:35 ` [PATCH v1 22/30] io_uring/rsrc: refactor io_buffer_register_bvec()/io_buffer_unregister_bvec() Joanne Koong
2025-12-07  8:33   ` Caleb Sander Mateos
2025-12-13  5:11     ` Joanne Koong
2025-12-03  0:35 ` [PATCH v1 23/30] io_uring/rsrc: split io_buffer_register_request() logic Joanne Koong
2025-12-07  8:41   ` Caleb Sander Mateos
2025-12-13  5:24     ` Joanne Koong
2025-12-03  0:35 ` [PATCH v1 24/30] io_uring/rsrc: Allow buffer release callback to be optional Joanne Koong
2025-12-07  8:42   ` Caleb Sander Mateos
2025-12-03  0:35 ` [PATCH v1 25/30] io_uring/rsrc: add io_buffer_register_bvec() Joanne Koong
2025-12-03  0:35 ` [PATCH v1 26/30] io_uring/rsrc: export io_buffer_unregister Joanne Koong
2025-12-03  0:35 ` [PATCH v1 27/30] fuse: rename fuse_set_zero_arg0() to fuse_zero_in_arg0() Joanne Koong
2025-12-03  0:35 ` [PATCH v1 28/30] fuse: enforce op header for every payload reply Joanne Koong
2025-12-03  0:35 ` [PATCH v1 29/30] fuse: add zero-copy over io-uring Joanne Koong
2025-12-03  0:35 ` [PATCH v1 30/30] docs: fuse: add io-uring bufring and zero-copy documentation Joanne Koong
2025-12-13  7:52   ` Askar Safin
2025-12-13  9:14 ` [PATCH v1 00/30] fuse/io-uring: add kernel-managed buffer rings and zero-copy Askar Safin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20251203003526.2889477-5-joannelkoong@gmail.com \
    --to=joannelkoong@gmail.com \
    --cc=asml.silence@gmail.com \
    --cc=axboe@kernel.dk \
    --cc=bschubert@ddn.com \
    --cc=csander@purestorage.com \
    --cc=io-uring@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=xiaobing.li@samsung.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).