io-uring: Make statx api stable
v3 cover io-uring: Make statx api stable
One of the key architectual tenets of io-uring is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
Patches:
Patch 1: fs: replace const char* parameter in vfs_statx and do_statx with
struct filename
Create filename object outside of do_statx and vfs_statx, so io-uring can create the filename object during the prepare phase
Patch 2: io-uring: Copy path name during prepare stage for statx
Create and store filename object during prepare phase
There is also a patch for the liburing libray to add a new test case. This patch makes sure that the api is stable.
"liburing: add test for stable statx api"
The patch has been tested with the liburing test suite and fstests.
Changes:
V2: don't check name in vfs_fstatat
V3: don't check name in statx syscall
Stefan Roesch (2):
fs: replace const char* parameter in vfs_statx and do_statx with
struct filename
io-uring: Copy path name during prepare stage for statx
fs/internal.h | 4 +++- fs/io_uring.c | 22 ++++++++++++++++++++-- fs/stat.c | 47 ++++++++++++++++++++++++++++++++++------------- 3 files changed, 57 insertions(+), 16 deletions(-)
base-commit: 705d84a366cfccda1e7aec1113a5399cd2ffee7d
Signature · 1 line
2.30.2
PATCH v3 1/2 fs: replace const char* parameter in vfs_statx and do_statx with struct filename
This replaces the const char* __user filename parameter in the two functions do_statx and vfs_statx with a struct filename *. In addition to be able to correctly construct a filename object a new helper function getname_statx_lookup_flags is introduced. The function makes sure that do_statx and vfs_statx is invoked with the correct lookup flags.
--- fs/internal.h | 4 +++- fs/stat.c | 47 ++++++++++++++++++++++++++++++++++------------- 2 files changed, 37 insertions(+), 14 deletions(-)
fs/internal.h +3 -1
@@ -184,7 +184,9 @@ int sb_init_dio_done_wq(struct super_block *sb);/* * fs/stat.c: */ -int do_statx (int dfd,const char __user *filename,unsigned flags, + +int getname_statx_lookup_flags (int flags); +int do_statx (int dfd,struct filename *filename,unsigned int flags,unsigned int mask,struct statx __user *buffer);/*
fs/stat.c +34 -13
84 lines not on this page · raw message
Signature · 1 line
2.30.2
PATCH v3 2/2 io-uring: Copy path name during prepare stage for statx
In reply to Stefan RoeschOne of the key architectual tenets is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
This changes replaces the const char * filename pointer in the io_statx structure with a struct filename *. In addition it also creates the filename object during the prepare phase.
With this change, the opcode also needs to invoke cleanup, so the filename object gets freed after processing the request.
--- fs/io_uring.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-)
fs/io_uring.c +20 -2
@@ -642,7 +642,7 @@ struct io_statx {int dfd;unsigned int mask;unsigned int flags; -const char __user *filename; +struct filename *filename;struct statx __user *buffer; };
@@ -4721,6 +4721,8 @@ static int io_fadvise(struct io_kiocb *req, unsigned int issue_flags)static int io_statx_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) { +const char __user *path; +if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL ))return -EINVAL ;if (sqe->ioprio || sqe->buf_index || sqe->splice_fd_in)
@@ -4730,10 +4732,22 @@ static int io_statx_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) req->statx.dfd =READ_ONCE (sqe->fd); req->statx.mask =READ_ONCE (sqe->len); - req->statx.filename =u64_to_user_ptr (READ_ONCE (sqe->addr)); + path =u64_to_user_ptr (READ_ONCE (sqe->addr)); req->statx.buffer =u64_to_user_ptr (READ_ONCE (sqe->addr2)); req->statx.flags =READ_ONCE (sqe->statx_flags); + req->statx.filename =getname_flags (path, +getname_statx_lookup_flags (req->statx.flags), +NULL ); + +if (IS_ERR (req->statx.filename)) { +int ret =PTR_ERR (req->statx.filename); + + req->statx.filename =NULL ; +return ret; + } + + req->flags |=REQ_F_NEED_CLEANUP ;return 0 ; }
@@ -6708,6 +6722,10 @@ static void io_clean_op(struct io_kiocb *req)putname (req->hardlink.oldpath);putname (req->hardlink.newpath);break ; +case IORING_OP_STATX : +if (req->statx.filename) +putname (req->statx.filename); +break ; } }if ((req->flags &REQ_F_POLLED ) && req->apoll) {
Signature · 1 line
2.30.2
Jens Axboe
maintainerQuoted 15 lines · Stefan Roesch
One of the key architectual tenets of io-uring is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
Patches:
Patch 1: fs: replace const char* parameter invfs_statxanddo_statxwith
struct filename
Create filename object outside ofdo_statxandvfs_statx, so io-uring can create the filename object during the prepare phasePatch 2: io-uring: Copy path name during prepare stage for statx
Create and store filename object during prepare phase
Quoted 5 lines
There is also a patch for the liburing libray to add a new test case. This patch makes sure that the api is stable.
"liburing: add test for stable statx api"The patch has been tested with the liburing test suite and fstests.
Al, are you happy with this version?
Signature · 1 line
Jens Axboe
Jens Axboe
maintainerQuoted 16 lines · Jens Axboe
Quoted 15 lines · Stefan Roesch
One of the key architectual tenets of io-uring is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
Patches:
Patch 1: fs: replace const char* parameter invfs_statxanddo_statxwith
struct filename
Create filename object outside ofdo_statxandvfs_statx, so io-uring can create the filename object during the prepare phasePatch 2: io-uring: Copy path name during prepare stage for statx
Create and store filename object during prepare phase
Quoted 7 lines
Quoted 5 lines
There is also a patch for the liburing libray to add a new test case. This patch makes sure that the api is stable.
"liburing: add test for stable statx api"The patch has been tested with the liburing test suite and fstests.
Al, are you happy with this version?
I have staged this one for 5.18, it's in for-5.18/io_uring-statx. It will be sent separately from the general io_uring fixes/updates.
Signature · 1 line
Jens Axboe
Quoted 5 lines · Stefan Roesch
One of the key architectual tenets of io-uring is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
Quoted 7 lines
Patches:
Patch 1: fs: replace const char* parameter invfs_statxanddo_statxwith
struct filename
Create filename object outside ofdo_statxandvfs_statx, so io-uring can create the filename object during the prepare phase[...]
Applied, thanks!
[1/2] fs: replace const char* parameter in vfs_statx and do_statx with struct filename
commit: 30512d54fae354a2359a740b75a1451b68aa3807
[2/2] io-uring: Copy path name during prepare stage for statx
commit: 1e0561928e3ab5018615403a2a1293e1e44ee03eBest regards,
Signature · 1 line
Jens Axboe
Hi,
Quoted 12 lines · Jens Axboe
Quoted 11 lines · Stefan Roesch
One of the key architectual tenets of io-uring is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
Patches:
Patch 1: fs: replace const char* parameter invfs_statxanddo_statxwith
struct filename
Create filename object outside ofdo_statxandvfs_statx, so io-uring can create the filename object during the prepare phase
Quoted 7 lines
Quoted 1 line
[...]
Applied, thanks!
[1/2] fs: replace const char* parameter in
vfs_statxanddo_statxwith struct filenamecommit: 30512d54fae354a2359a740b75a1451b68aa3807 [2/2] io-uring: Copy path name during prepare stage for statx commit: 1e0561928e3ab5018615403a2a1293e1e44ee03e
Those 2 commits landed in todays Linux next-20220223. They affect userspace in a way that breaks systemd opration:
...
Freeing unused kernel image (initmem) memory: 1024K Run /sbin/init as init process
systemd[1]: System time before build time, advancing clock. systemd[1]: Cannot be run in a chroot() environment. systemd[1]: Freezing execution.
Reverting them on top of next-20220223 fixes the boot issue. Btw, those patches are not bisectable. The code at 30512d54fae354a2359a740b75a1451b68aa3807 doesn't compile.
Best regards
Signature · 2 lines
Marek Szyprowski, PhD Samsung R&D Institute Poland
Jens Axboe
maintainerQuoted 34 lines · Marek Szyprowski
Hi,
Quoted 12 lines · Jens Axboe
Quoted 11 lines · Stefan Roesch
One of the key architectual tenets of io-uring is to keep the parameters for io-uring stable. After the call has been submitted, its value can be changed. Unfortunaltely this is not the case for the current statx implementation.
Patches:
Patch 1: fs: replace const char* parameter invfs_statxanddo_statxwith
struct filename
Create filename object outside ofdo_statxandvfs_statx, so io-uring can create the filename object during the prepare phaseQuoted 7 lines
Quoted 1 line
[...]
Applied, thanks!
[1/2] fs: replace const char* parameter in
vfs_statxanddo_statxwith struct filenamecommit: 30512d54fae354a2359a740b75a1451b68aa3807 [2/2] io-uring: Copy path name during prepare stage for statx commit: 1e0561928e3ab5018615403a2a1293e1e44ee03eThose 2 commits landed in todays Linux next-20220223. They affect userspace in a way that breaks systemd opration:
...
Freeing unused kernel image (initmem) memory: 1024K Run
/sbin/initas init process
systemd[1]: System time before build time, advancing clock. systemd[1]: Cannot be run in achroot()environment. systemd[1]: Freezing execution.
Quoted 3 lines
Reverting them on top of next-20220223 fixes the boot issue. Btw, those patches are not bisectable. The code at
30512d54fae354a2359a740b75a1451b68aa3807doesn't compile.
Thanks, I'll drop them from for-next until we figure out what that is.
Signature · 1 line
Jens Axboe
Quoted 11 lines · Jens Axboe
Quoted 9 lines
Freeing unused kernel image (initmem) memory: 1024K Run
/sbin/initas init process
systemd[1]: System time before build time, advancing clock. systemd[1]: Cannot be run in achroot()environment. systemd[1]: Freezing execution.Reverting them on top of next-20220223 fixes the boot issue. Btw, those patches are not bisectable. The code at
30512d54fae354a2359a740b75a1451b68aa3807doesn't compile.Thanks, I'll drop them from for-next until we figure out what that is.
FYI, this breaks the boot on bare-metal as well.
Loading, please wait...
Starting version 245.4-4ubuntu3.15
Running in chroot, ignoring request.
Running in chroot, ignoring request.
Running in chroot, ignoring request.
Running in chroot, ignoring request.
Running in chroot, ignoring request.
/scripts/init-top/console_setup: line 90: can't create /dev/tty1: No such device or address /scripts/init-top/console_setup: line 126: can't open /dev/tty1: No such device or address
Begin: Loading essential drivers ... /init: line 155: modprobe: Permission denied /init: line 155: modprobe: Permission denied
/init: line 155: modprobe: Permission denied
/init: line 155: modprobe: Permission denied
/init: line 155: modprobe: Permission denied
/init: line 155: modprobe: Permission denied
/init: line 155: modprobe: Permission denied
/init: line 155: modprobe: Permission denied
Quoted 7 lines · Qian Cai
Quoted 6 lines · Jens Axboe
Quoted 6 lines
Freeing unused kernel image (initmem) memory: 1024K Run
/sbin/initas init process
systemd[1]: System time before build time, advancing clock. systemd[1]: Cannot be run in achroot()environment. systemd[1]: Freezing execution.
Quoted 7 lines
Quoted 5 lines
Quoted 3 lines
Reverting them on top of next-20220223 fixes the boot issue. Btw, those patches are not bisectable. The code at
30512d54fae354a2359a740b75a1451b68aa3807doesn't compile.Thanks, I'll drop them from for-next until we figure out what that is.
FYI, this breaks the boot on bare-metal as well.
It also broke my boot on a simple debian testing KVM guest as well. Reverting those two commits fixes my boot.
Jens, any chance we can include / ask for a bit more run time testing? What sort of tests get run ?
Luis