* Re: [PATCH v5 05/22] liveupdate: kho: when live update add KHO image during kexec load
From: Mike Rapoport @ 2025-11-10 12:47 UTC (permalink / raw)
To: Pasha Tatashin
Cc: pratyush, jasonmiu, graf, dmatlack, rientjes, corbet, rdunlap,
ilpo.jarvinen, kanie, ojeda, aliceryhl, masahiroy, akpm, tj,
yoann.congal, mmaurer, roman.gushchin, chenridong, axboe,
mark.rutland, jannh, vincent.guittot, hannes, dan.j.williams,
david, joel.granados, rostedt, anna.schumaker, song, zhangguopeng,
linux, linux-kernel, linux-doc, linux-mm, gregkh, tglx, mingo, bp,
dave.hansen, x86, hpa, rafael, dakr, bartosz.golaszewski,
cw00.choi, myungjoo.ham, yesanishhere, Jonathan.Cameron,
quic_zijuhu, aleksander.lobakin, ira.weiny, andriy.shevchenko,
leon, lukas, bhelgaas, wagi, djeffery, stuart.w.hayes, ptyadav,
lennart, brauner, linux-api, linux-fsdevel, saeedm, ajayachandra,
jgg, parav, leonro, witu, hughd, skhawaja, chrisl
In-Reply-To: <20251107210526.257742-6-pasha.tatashin@soleen.com>
On Fri, Nov 07, 2025 at 04:03:03PM -0500, Pasha Tatashin wrote:
> In case KHO is driven from within kernel via live update, finalize will
> always happen during reboot, so add the KHO image unconditionally.
>
> Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
> ---
> kernel/liveupdate/kexec_handover.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/liveupdate/kexec_handover.c b/kernel/liveupdate/kexec_handover.c
> index 9f0913e101be..b54ca665e005 100644
> --- a/kernel/liveupdate/kexec_handover.c
> +++ b/kernel/liveupdate/kexec_handover.c
> @@ -15,6 +15,7 @@
> #include <linux/kexec_handover.h>
> #include <linux/libfdt.h>
> #include <linux/list.h>
> +#include <linux/liveupdate.h>
> #include <linux/memblock.h>
> #include <linux/page-isolation.h>
> #include <linux/vmalloc.h>
> @@ -1489,7 +1490,7 @@ int kho_fill_kimage(struct kimage *image)
> int err = 0;
> struct kexec_buf scratch;
>
> - if (!kho_out.finalized)
> + if (!kho_out.finalized && !liveupdate_enabled())
> return 0;
This feels backwards, I don't think KHO should call liveupdate methods.
> image->kho.fdt = virt_to_phys(kho_out.fdt);
> --
> 2.51.2.1041.gc1ab5b90ca-goog
>
--
Sincerely yours,
Mike.
^ permalink raw reply
* Re: [PATCH v5 02/22] liveupdate: luo_core: integrate with KHO
From: Mike Rapoport @ 2025-11-10 13:00 UTC (permalink / raw)
To: Pasha Tatashin
Cc: pratyush, jasonmiu, graf, dmatlack, rientjes, corbet, rdunlap,
ilpo.jarvinen, kanie, ojeda, aliceryhl, masahiroy, akpm, tj,
yoann.congal, mmaurer, roman.gushchin, chenridong, axboe,
mark.rutland, jannh, vincent.guittot, hannes, dan.j.williams,
david, joel.granados, rostedt, anna.schumaker, song, zhangguopeng,
linux, linux-kernel, linux-doc, linux-mm, gregkh, tglx, mingo, bp,
dave.hansen, x86, hpa, rafael, dakr, bartosz.golaszewski,
cw00.choi, myungjoo.ham, yesanishhere, Jonathan.Cameron,
quic_zijuhu, aleksander.lobakin, ira.weiny, andriy.shevchenko,
leon, lukas, bhelgaas, wagi, djeffery, stuart.w.hayes, ptyadav,
lennart, brauner, linux-api, linux-fsdevel, saeedm, ajayachandra,
jgg, parav, leonro, witu, hughd, skhawaja, chrisl
In-Reply-To: <20251107210526.257742-3-pasha.tatashin@soleen.com>
On Fri, Nov 07, 2025 at 04:03:00PM -0500, Pasha Tatashin wrote:
> Integrate the LUO with the KHO framework to enable passing LUO state
> across a kexec reboot.
>
> When LUO is transitioned to a "prepared" state, it tells KHO to
> finalize, so all memory segments that were added to KHO preservation
> list are getting preserved. After "Prepared" state no new segments
> can be preserved. If LUO is canceled, it also tells KHO to cancel the
> serialization, and therefore, later LUO can go back into the prepared
> state.
>
> This patch introduces the following changes:
> - During the KHO finalization phase allocate FDT blob.
> - Populate this FDT with a LUO compatibility string ("luo-v1").
>
> LUO now depends on `CONFIG_KEXEC_HANDOVER`. The core state transition
> logic (`luo_do_*_calls`) remains unimplemented in this patch.
>
> Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
> ---
...
> @@ -69,7 +197,22 @@ early_param("liveupdate", early_liveupdate_param);
> */
> int liveupdate_reboot(void)
> {
> - return 0;
> + int err;
> +
> + if (!liveupdate_enabled())
> + return 0;
> +
> + err = kho_finalize();
kho_finalize() should be really called from kernel_kexec().
We avoided it because of the concern that memory allocations that late in
reboot could be an issue. But I looked at hibernate() and it does
allocations on reboot->hibernate path, so adding kho_finalize() as the
first step of kernel_kexec() seems fine.
And if we prioritize stateless memory tracking in KHO, it won't be a
concern at all.
> + if (err) {
> + pr_err("kho_finalize failed %d\n", err);
> + /*
> + * kho_finalize() may return libfdt errors, to aboid passing to
> + * userspace unknown errors, change this to EAGAIN.
> + */
> + err = -EAGAIN;
> + }
> +
> + return err;
> }
>
> /**
--
Sincerely yours,
Mike.
^ permalink raw reply
* Re: [PATCH v5 05/22] liveupdate: kho: when live update add KHO image during kexec load
From: Pasha Tatashin @ 2025-11-10 15:31 UTC (permalink / raw)
To: Mike Rapoport
Cc: pratyush, jasonmiu, graf, dmatlack, rientjes, corbet, rdunlap,
ilpo.jarvinen, kanie, ojeda, aliceryhl, masahiroy, akpm, tj,
yoann.congal, mmaurer, roman.gushchin, chenridong, axboe,
mark.rutland, jannh, vincent.guittot, hannes, dan.j.williams,
david, joel.granados, rostedt, anna.schumaker, song, zhangguopeng,
linux, linux-kernel, linux-doc, linux-mm, gregkh, tglx, mingo, bp,
dave.hansen, x86, hpa, rafael, dakr, bartosz.golaszewski,
cw00.choi, myungjoo.ham, yesanishhere, Jonathan.Cameron,
quic_zijuhu, aleksander.lobakin, ira.weiny, andriy.shevchenko,
leon, lukas, bhelgaas, wagi, djeffery, stuart.w.hayes, ptyadav,
lennart, brauner, linux-api, linux-fsdevel, saeedm, ajayachandra,
jgg, parav, leonro, witu, hughd, skhawaja, chrisl
In-Reply-To: <aRHe3syRvOrCYC9u@kernel.org>
On Mon, Nov 10, 2025 at 7:47 AM Mike Rapoport <rppt@kernel.org> wrote:
>
> On Fri, Nov 07, 2025 at 04:03:03PM -0500, Pasha Tatashin wrote:
> > In case KHO is driven from within kernel via live update, finalize will
> > always happen during reboot, so add the KHO image unconditionally.
> >
> > Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
> > ---
> > kernel/liveupdate/kexec_handover.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/kernel/liveupdate/kexec_handover.c b/kernel/liveupdate/kexec_handover.c
> > index 9f0913e101be..b54ca665e005 100644
> > --- a/kernel/liveupdate/kexec_handover.c
> > +++ b/kernel/liveupdate/kexec_handover.c
> > @@ -15,6 +15,7 @@
> > #include <linux/kexec_handover.h>
> > #include <linux/libfdt.h>
> > #include <linux/list.h>
> > +#include <linux/liveupdate.h>
> > #include <linux/memblock.h>
> > #include <linux/page-isolation.h>
> > #include <linux/vmalloc.h>
> > @@ -1489,7 +1490,7 @@ int kho_fill_kimage(struct kimage *image)
> > int err = 0;
> > struct kexec_buf scratch;
> >
> > - if (!kho_out.finalized)
> > + if (!kho_out.finalized && !liveupdate_enabled())
> > return 0;
>
> This feels backwards, I don't think KHO should call liveupdate methods.
It is backward, but it is a requirement until KHO becomes stateless.
LUO does not have dependencies on userspace state of when kexec is
loaded. In fact the next kernel must be loaded before the brownout as
it is an expensive operation. The sequence of events should:
1. Load the next kernel in memory
2. Preserve resources via LUO
3. Do Kexec reboot
Pasha
^ permalink raw reply
* Re: [PATCH v5 02/22] liveupdate: luo_core: integrate with KHO
From: Pasha Tatashin @ 2025-11-10 15:43 UTC (permalink / raw)
To: Mike Rapoport
Cc: pratyush, jasonmiu, graf, dmatlack, rientjes, corbet, rdunlap,
ilpo.jarvinen, kanie, ojeda, aliceryhl, masahiroy, akpm, tj,
yoann.congal, mmaurer, roman.gushchin, chenridong, axboe,
mark.rutland, jannh, vincent.guittot, hannes, dan.j.williams,
david, joel.granados, rostedt, anna.schumaker, song, zhangguopeng,
linux, linux-kernel, linux-doc, linux-mm, gregkh, tglx, mingo, bp,
dave.hansen, x86, hpa, rafael, dakr, bartosz.golaszewski,
cw00.choi, myungjoo.ham, yesanishhere, Jonathan.Cameron,
quic_zijuhu, aleksander.lobakin, ira.weiny, andriy.shevchenko,
leon, lukas, bhelgaas, wagi, djeffery, stuart.w.hayes, ptyadav,
lennart, brauner, linux-api, linux-fsdevel, saeedm, ajayachandra,
jgg, parav, leonro, witu, hughd, skhawaja, chrisl
In-Reply-To: <aRHiCxoJnEGmj17q@kernel.org>
>
> kho_finalize() should be really called from kernel_kexec().
>
> We avoided it because of the concern that memory allocations that late in
> reboot could be an issue. But I looked at hibernate() and it does
> allocations on reboot->hibernate path, so adding kho_finalize() as the
> first step of kernel_kexec() seems fine.
This isn't a regular reboot; it's a live update. The
liveupdate_reboot() is designed to be reversible and allows us to
return an error, undoing the freeze() operations via unfreeze() in
case of failure.
This is why this call is placed first in reboot(), before any
irreversible reboot notifiers or shutdown callbacks are performed. If
an allocation problem occurs in KHO, the error is simply reported back
to userspace, and the live update update is safely aborted.
> And if we prioritize stateless memory tracking in KHO, it won't be a
> concern at all.
We are prioritizing stateless KHO work ;-) +Jason Miu
Once KHO is stateless, the kho_finalize() is going to be removed.
>
> > + if (err) {
> > + pr_err("kho_finalize failed %d\n", err);
> > + /*
> > + * kho_finalize() may return libfdt errors, to aboid passing to
> > + * userspace unknown errors, change this to EAGAIN.
> > + */
> > + err = -EAGAIN;
> > + }
> > +
> > + return err;
> > }
> >
> > /**
>
> --
> Sincerely yours,
> Mike.
^ permalink raw reply
* Re: [PATCH v5 08/22] liveupdate: luo_file: implement file systems callbacks
From: Pratyush Yadav @ 2025-11-10 17:27 UTC (permalink / raw)
To: Pasha Tatashin
Cc: pratyush, jasonmiu, graf, rppt, dmatlack, rientjes, corbet,
rdunlap, ilpo.jarvinen, kanie, ojeda, aliceryhl, masahiroy, akpm,
tj, yoann.congal, mmaurer, roman.gushchin, chenridong, axboe,
mark.rutland, jannh, vincent.guittot, hannes, dan.j.williams,
david, joel.granados, rostedt, anna.schumaker, song, zhangguopeng,
linux, linux-kernel, linux-doc, linux-mm, gregkh, tglx, mingo, bp,
dave.hansen, x86, hpa, rafael, dakr, bartosz.golaszewski,
cw00.choi, myungjoo.ham, yesanishhere, Jonathan.Cameron,
quic_zijuhu, aleksander.lobakin, ira.weiny, andriy.shevchenko,
leon, lukas, bhelgaas, wagi, djeffery, stuart.w.hayes, lennart,
brauner, linux-api, linux-fsdevel, saeedm, ajayachandra, jgg,
parav, leonro, witu, hughd, skhawaja, chrisl
In-Reply-To: <20251107210526.257742-9-pasha.tatashin@soleen.com>
Hi Pasha,
Caught a small bug during some of my testing.
On Fri, Nov 07 2025, Pasha Tatashin wrote:
> This patch implements the core mechanism for managing preserved
> files throughout the live update lifecycle. It provides the logic to
> invoke the file handler callbacks (preserve, unpreserve, freeze,
> unfreeze, retrieve, and finish) at the appropriate stages.
>
> During the reboot phase, luo_file_freeze() serializes the final
> metadata for each file (handler compatible string, token, and data
> handle) into a memory region preserved by KHO. In the new kernel,
> luo_file_deserialize() reconstructs the in-memory file list from this
> data, preparing the session for retrieval.
>
> Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
[...]
> +int luo_preserve_file(struct luo_session *session, u64 token, int fd)
> +{
> + struct liveupdate_file_op_args args = {0};
> + struct liveupdate_file_handler *fh;
> + struct luo_file *luo_file;
> + struct file *file;
> + int err = -ENOENT;
> +
> + lockdep_assert_held(&session->mutex);
> +
> + if (luo_token_is_used(session, token))
> + return -EEXIST;
> +
> + file = fget(fd);
> + if (!file)
> + return -EBADF;
> +
> + err = luo_session_alloc_files_mem(session);
err gets set to 0 here...
> + if (err)
> + goto exit_err;
> +
> + if (session->count == LUO_FILE_MAX) {
> + err = -ENOSPC;
> + goto exit_err;
> + }
> +
> + list_for_each_entry(fh, &luo_file_handler_list, list) {
> + if (fh->ops->can_preserve(fh, file)) {
> + err = 0;
> + break;
> + }
> + }
... say no file handler can preserve this file ...
> +
> + /* err is still -ENOENT if no handler was found */
> + if (err)
... err is not ENOENT, but 0. So this function does not error but, but
goes ahead with fh == luo_file_handler_list (since end of list). This
causes an out-of-bounds access. It eventually causes a kernel fault and
panic.
You should drop the ENOENT at initialization time and set it right
before list_for_each_entry().
> + goto exit_err;
> +
> + luo_file = kzalloc(sizeof(*luo_file), GFP_KERNEL);
> + if (!luo_file) {
> + err = -ENOMEM;
> + goto exit_err;
> + }
> +
> + luo_file->file = file;
> + luo_file->fh = fh;
> + luo_file->token = token;
> + luo_file->retrieved = false;
> + mutex_init(&luo_file->mutex);
> +
> + args.handler = fh;
> + args.session = (struct liveupdate_session *)session;
> + args.file = file;
> + err = fh->ops->preserve(&args);
> + if (err) {
> + mutex_destroy(&luo_file->mutex);
> + kfree(luo_file);
> + goto exit_err;
> + } else {
> + luo_file->serialized_data = args.serialized_data;
> + list_add_tail(&luo_file->list, &session->files_list);
> + session->count++;
> + }
> +
> + return 0;
> +
> +exit_err:
> + fput(file);
> + luo_session_free_files_mem(session);
> +
> + return err;
> +}
[...]
--
Regards,
Pratyush Yadav
^ permalink raw reply
* Re: [PATCH v5 08/22] liveupdate: luo_file: implement file systems callbacks
From: Pasha Tatashin @ 2025-11-10 17:42 UTC (permalink / raw)
To: Pratyush Yadav
Cc: jasonmiu, graf, rppt, dmatlack, rientjes, corbet, rdunlap,
ilpo.jarvinen, kanie, ojeda, aliceryhl, masahiroy, akpm, tj,
yoann.congal, mmaurer, roman.gushchin, chenridong, axboe,
mark.rutland, jannh, vincent.guittot, hannes, dan.j.williams,
david, joel.granados, rostedt, anna.schumaker, song, zhangguopeng,
linux, linux-kernel, linux-doc, linux-mm, gregkh, tglx, mingo, bp,
dave.hansen, x86, hpa, rafael, dakr, bartosz.golaszewski,
cw00.choi, myungjoo.ham, yesanishhere, Jonathan.Cameron,
quic_zijuhu, aleksander.lobakin, ira.weiny, andriy.shevchenko,
leon, lukas, bhelgaas, wagi, djeffery, stuart.w.hayes, lennart,
brauner, linux-api, linux-fsdevel, saeedm, ajayachandra, jgg,
parav, leonro, witu, hughd, skhawaja, chrisl
In-Reply-To: <mafs0ms4tajcs.fsf@kernel.org>
On Mon, Nov 10, 2025 at 12:27 PM Pratyush Yadav <pratyush@kernel.org> wrote:
>
> Hi Pasha,
>
> Caught a small bug during some of my testing.
>
> On Fri, Nov 07 2025, Pasha Tatashin wrote:
>
> > This patch implements the core mechanism for managing preserved
> > files throughout the live update lifecycle. It provides the logic to
> > invoke the file handler callbacks (preserve, unpreserve, freeze,
> > unfreeze, retrieve, and finish) at the appropriate stages.
> >
> > During the reboot phase, luo_file_freeze() serializes the final
> > metadata for each file (handler compatible string, token, and data
> > handle) into a memory region preserved by KHO. In the new kernel,
> > luo_file_deserialize() reconstructs the in-memory file list from this
> > data, preparing the session for retrieval.
> >
> > Signed-off-by: Pasha Tatashin <pasha.tatashin@soleen.com>
> [...]
> > +int luo_preserve_file(struct luo_session *session, u64 token, int fd)
> > +{
> > + struct liveupdate_file_op_args args = {0};
> > + struct liveupdate_file_handler *fh;
> > + struct luo_file *luo_file;
> > + struct file *file;
> > + int err = -ENOENT;
> > +
> > + lockdep_assert_held(&session->mutex);
> > +
> > + if (luo_token_is_used(session, token))
> > + return -EEXIST;
> > +
> > + file = fget(fd);
> > + if (!file)
> > + return -EBADF;
> > +
> > + err = luo_session_alloc_files_mem(session);
>
> err gets set to 0 here...
>
> > + if (err)
> > + goto exit_err;
> > +
> > + if (session->count == LUO_FILE_MAX) {
> > + err = -ENOSPC;
> > + goto exit_err;
> > + }
> > +
> > + list_for_each_entry(fh, &luo_file_handler_list, list) {
> > + if (fh->ops->can_preserve(fh, file)) {
> > + err = 0;
> > + break;
> > + }
> > + }
>
> ... say no file handler can preserve this file ...
>
> > +
> > + /* err is still -ENOENT if no handler was found */
> > + if (err)
>
> ... err is not ENOENT, but 0. So this function does not error but, but
> goes ahead with fh == luo_file_handler_list (since end of list). This
> causes an out-of-bounds access. It eventually causes a kernel fault and
> panic.
>
> You should drop the ENOENT at initialization time and set it right
> before list_for_each_entry().
Right, thank you for reporting this. Should add it to self-tests,
where we try to preserve FD that does not have a file handler.
Pasha
>
> > + goto exit_err;
> > +
> > + luo_file = kzalloc(sizeof(*luo_file), GFP_KERNEL);
> > + if (!luo_file) {
> > + err = -ENOMEM;
> > + goto exit_err;
> > + }
> > +
> > + luo_file->file = file;
> > + luo_file->fh = fh;
> > + luo_file->token = token;
> > + luo_file->retrieved = false;
> > + mutex_init(&luo_file->mutex);
> > +
> > + args.handler = fh;
> > + args.session = (struct liveupdate_session *)session;
> > + args.file = file;
> > + err = fh->ops->preserve(&args);
> > + if (err) {
> > + mutex_destroy(&luo_file->mutex);
> > + kfree(luo_file);
> > + goto exit_err;
> > + } else {
> > + luo_file->serialized_data = args.serialized_data;
> > + list_add_tail(&luo_file->list, &session->files_list);
> > + session->count++;
> > + }
> > +
> > + return 0;
> > +
> > +exit_err:
> > + fput(file);
> > + luo_session_free_files_mem(session);
> > +
> > + return err;
> > +}
> [...]
>
> --
> Regards,
> Pratyush Yadav
^ permalink raw reply
* Re: RFC: Serial port DTR/RTS - O_NRESETDEV
From: Theodore Ts'o @ 2025-11-10 20:19 UTC (permalink / raw)
To: Maarten Brock
Cc: H. Peter Anvin, linux-serial@vger.kernel.org,
linux-api@vger.kernel.org, LKML
In-Reply-To: <AMBPR05MB11925DA076098B05E418BF64283CEA@AMBPR05MB11925.eurprd05.prod.outlook.com>
On Mon, Nov 10, 2025 at 10:06:02AM +0000, Maarten Brock wrote:
> I fully agree that you cannot expect users that wired something like RS485 Driver
> Enable or a microcontroller reset to RTS or DTR to write their own kernel driver.
> And you need to open the port to make the appropriate settings. But opening a
> port should not e.g. claim the RS485 bus and mess up whatever communication
> was going on there.
Again, the existing seral driver code *will* mess with RTS and DTR at
boot up because that's part of the autoconfiuration code, and that was
added because it was needed for some number of serial ports.
If that's going to "mess up" the RS485 bus, maybe we need accept that
RS-232 != RS-485 and have a different driver for the two. That's
going to be a lot simpler than trying to make the same code work for
both RS-232 and RS-485, and claiming that the existing RS-232 code is
"fundamentally buggy" when it interacts poorly with something that has
very different requirements than the historical RS-232 use cases.
- Ted
^ permalink raw reply
* Re: RFC: Serial port DTR/RTS - O_NRESETDEV
From: H. Peter Anvin @ 2025-11-10 21:05 UTC (permalink / raw)
To: Theodore Ts'o, Maarten Brock
Cc: linux-serial@vger.kernel.org, linux-api@vger.kernel.org, LKML
In-Reply-To: <20251110201933.GH2988753@mit.edu>
On November 10, 2025 12:19:33 PM PST, Theodore Ts'o <tytso@mit.edu> wrote:
>On Mon, Nov 10, 2025 at 10:06:02AM +0000, Maarten Brock wrote:
>> I fully agree that you cannot expect users that wired something like RS485 Driver
>> Enable or a microcontroller reset to RTS or DTR to write their own kernel driver.
>> And you need to open the port to make the appropriate settings. But opening a
>> port should not e.g. claim the RS485 bus and mess up whatever communication
>> was going on there.
>
>Again, the existing seral driver code *will* mess with RTS and DTR at
>boot up because that's part of the autoconfiuration code, and that was
>added because it was needed for some number of serial ports.
>
>If that's going to "mess up" the RS485 bus, maybe we need accept that
>RS-232 != RS-485 and have a different driver for the two. That's
>going to be a lot simpler than trying to make the same code work for
>both RS-232 and RS-485, and claiming that the existing RS-232 code is
>"fundamentally buggy" when it interacts poorly with something that has
>very different requirements than the historical RS-232 use cases.
>
> - Ted
I didn't say it was fundamentally buggy; if I did or implied it I apologize. It is most definitely not; however, in some cases it is undesired or undesirable *due to shifts in usage patterns.*
This isn't a bug at all but an enormous strength. That a 65-year-old standard — both hardware and software — designed for teletypes and dumb terminals can still be useful today is almost the definition of success. And, yes, some glitches in that process are going to be inevitable — like the mistake of retaining the termio Bxxx emumeration constants in the termios interface. But we deal with it by gradual evolution of interfaces.
One such example is RTS itself: the RS485 definition is, in fact, the originally intended meaning of RTS: it is a request to the DCE to negotiate transmission privilege and activate transmission mode over a half duplex channel, after which it asserts CTS. RTS/CTS flow control was a nonstandard adaption to allow for binary transparent flow control over full duplex links — it wasn't formally standardized until 1991, and the signal is formally named RTR when used that way. However, RTR and RTS share hardware in nearly all existing implementations, and share pins in the standard — so whether or not you are using RTR or RTS is a property of the DCE, not DTE, and needs to be configured into the DTE.
Requiring new drivers for the gajillion different hardware devices already supported and then having the problem of which drivers claim it isn't really any better of a solution; one could in fact argue it is *exactly* equivalent to being able to indicate to the driver what mode one wants it to operate in before it does its configuration.
The parport driver layer is kind of similar to this, in some ways, but in the tty layer that is mostly handled by line disciplines instead. (The parport hardware was generally abused on a much lower level, as a substitute for GPIOs, so even the notion of a byte stream wasn't there.)
*If* I'm reading the code correctly – which is a little complicated due to the sheer number of abstraction layers – hardware initialization is already deferred until first open, which would mean that disabling autoconfiguration (one of the features in TIOCSSERIAL) would again be a valid reason for wanting to be able to communicate with a device driver before requiring that it puts the underlying hardware in the state expected for operation *in the mode configured* (catch-22).
As I stated, this is inherently going to be a best effort. For some devices that may mean simply leaving the power on default in place (in this specific case, presumably, DTR# and RTS# deasserted.) However, once the state is already known to the kernel then there is no such issue for any hardware.
As far as naming is concerned: O_RAW is really suboptimal, as it has exactly the same implications as O_DIRECT (I/O goes straight to the device with no buffering.) I don't like the idea of abusing O_DIRECT at all; I only brought it up as a fallback alternative. I genuinely do believe that if we assign a new open flag it will find use cases outside the tty/serial port subsystems, and if there is anything Unix has done right it is to generalize interfaces as much as possible, case in point descriptors.
^ permalink raw reply
* Re: RFC: Serial port DTR/RTS - O_NRESETDEV
From: Theodore Ts'o @ 2025-11-11 3:51 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Maarten Brock, linux-serial@vger.kernel.org,
linux-api@vger.kernel.org, LKML
In-Reply-To: <0F8021E8-F288-4669-8195-9948844E36FD@zytor.com>
On Mon, Nov 10, 2025 at 01:05:55PM -0800, H. Peter Anvin wrote:
>
> The parport driver layer is kind of similar to this, in some ways,
> but in the tty layer that is mostly handled by line disciplines
> instead. (The parport hardware was generally abused on a much lower
> level, as a substitute for GPIOs, so even the notion of a byte
> stream wasn't there.)
I'm not an RS-485 expert, but over the years, I've heard all sorts of
"interesting" timing requirements. For example RTS can only be
dropped when the UART's shift register has completely drained. That's
not when the OS has sent the characters to the FIFO (which is why
tcdrain isn't quite what you want); it's not when the UART has sent
the transmit interrupt, but when the UART's shift register is *empty*.
Doing this via the termios interface is decidedly non-optimal[1].
[1] https://www.moxa.com.cn/getmedia/a07808dd-f3d7-473b-9a2f-93b5ce673bb1/moxa-the-secrets-of-rs-485-half-duplex-communication-tech-note-v1.0.pdf
It *can* be done, sure, with some strategic usleep()'s, but not
necessarily optimally, and it is decidedly hacky. From what I can
tell, it's actually not *that* different from treating RTS as a GPIO
pin, and really, this ought to be done in the kernel if you want
optimal control oer the timing of when RTS gets toggeled.
> *If* I'm reading the code correctly – which is a little complicated due to the sheer number of abstraction layers – hardware initialization is already deferred until first open, which would mean that disabling autoconfiguration (one of the features in TIOCSSERIAL) would again be a valid reason for wanting to be able to communicate with a device driver before requiring that it puts the underlying hardware in the state expected for operation *in the mode configured* (catch-22).
Well, part of the problem is that the PC's serial port predates PCI,
so there's no way we can tell whether or not there is a serial port at
a particular I/O port except by poking certain I/O ports, and seeing
if there is something that looks like a UART which responds.
Hopefully this won't accidentaly cause the NSA's thermite-charge
self-destruct charge from getting set off, and in practice, mainboard
designers who try to put things at the COM1/2/3/4 generally get weeded
out thanks to natural selection, hopefully without harming users along
the way. :-)
Worse, we also wanted to support serial cards that were at
non-standard ports, or using non-standard interrupts, and so that's
one of the reason hardware initialization is deferred until after we
have a chance to configure the serial device's I/O port and interrupt.
Now, we are in the 21st century, and just as we are trying to declare
32-bit i386 dead (to the disappointment of legacy hardware owners
everywhere who are whining about Debian dropping installer support for
i386), we could try to declare the ISA bus as dead, and no longer
being supported, and we could probably drop a bunch of this
complexity. We probably would need to support COM 1/2/3/4 ports since
these still tend to be hardwared without a way of detecting it their
existing via USB or PCI bus discovery mechanisms. And for those, we
would still need to do UART autoconfiguration. I suppose we could
just assume that all UART's are 16550A's, and that noone would
actually use older UART's type --- except (a) I bet there are some
cheap-skate embedded computer designers who decided to use a 8250
instead of 16550A for $REASONS$, and because of the extreme timing
requirements of some RS-485 use cases, I believe I have seen
recommendations to use setserial to actually force the UART into 8250
mode, to disable the TX FIFO --- all to satisfying the weird-shit RTS
timing requirements. (This is where I get my beief that RS-485 !=
RS-232, and if you want to do all of this weird-shit timing
requirements because you are trying to implemet a multi-node bus which
is half-duplex, maybe this should be done in the kernel, damn it!)
How about this? If you really don't want to open the device to
configure it not to assert DTR/RTS, perhaps we could specify on the
boot command line that one of TTYS[0123] has a different default line
discipline, and when tty's are opened when the RS-485 line displine,
the RS-485 rules apply, which would include not messing with DTR/RTS,
and if you want to play games with RTS getting dropped as soon as the
last bit has let the transmit shift register, we can get the kernel do
this in the line discpline, which arguably would be *way* more
reliable than playing GPIO-style timing games in userspace.
- Ted
^ permalink raw reply
* Re: RFC: Serial port DTR/RTS - O_NRESETDEV
From: H. Peter Anvin @ 2025-11-11 3:57 UTC (permalink / raw)
To: Theodore Ts'o
Cc: Maarten Brock, linux-serial@vger.kernel.org,
linux-api@vger.kernel.org, LKML
In-Reply-To: <20251111035143.GJ2988753@mit.edu>
On November 10, 2025 7:51:43 PM PST, Theodore Ts'o <tytso@mit.edu> wrote:
>On Mon, Nov 10, 2025 at 01:05:55PM -0800, H. Peter Anvin wrote:
>>
>> The parport driver layer is kind of similar to this, in some ways,
>> but in the tty layer that is mostly handled by line disciplines
>> instead. (The parport hardware was generally abused on a much lower
>> level, as a substitute for GPIOs, so even the notion of a byte
>> stream wasn't there.)
>
>I'm not an RS-485 expert, but over the years, I've heard all sorts of
>"interesting" timing requirements. For example RTS can only be
>dropped when the UART's shift register has completely drained. That's
>not when the OS has sent the characters to the FIFO (which is why
>tcdrain isn't quite what you want); it's not when the UART has sent
>the transmit interrupt, but when the UART's shift register is *empty*.
>Doing this via the termios interface is decidedly non-optimal[1].
>
>[1] https://www.moxa.com.cn/getmedia/a07808dd-f3d7-473b-9a2f-93b5ce673bb1/moxa-the-secrets-of-rs-485-half-duplex-communication-tech-note-v1.0.pdf
>
>It *can* be done, sure, with some strategic usleep()'s, but not
>necessarily optimally, and it is decidedly hacky. From what I can
>tell, it's actually not *that* different from treating RTS as a GPIO
>pin, and really, this ought to be done in the kernel if you want
>optimal control oer the timing of when RTS gets toggeled.
>
>> *If* I'm reading the code correctly – which is a little complicated due to the sheer number of abstraction layers – hardware initialization is already deferred until first open, which would mean that disabling autoconfiguration (one of the features in TIOCSSERIAL) would again be a valid reason for wanting to be able to communicate with a device driver before requiring that it puts the underlying hardware in the state expected for operation *in the mode configured* (catch-22).
>
>Well, part of the problem is that the PC's serial port predates PCI,
>so there's no way we can tell whether or not there is a serial port at
>a particular I/O port except by poking certain I/O ports, and seeing
>if there is something that looks like a UART which responds.
>Hopefully this won't accidentaly cause the NSA's thermite-charge
>self-destruct charge from getting set off, and in practice, mainboard
>designers who try to put things at the COM1/2/3/4 generally get weeded
>out thanks to natural selection, hopefully without harming users along
>the way. :-)
>
>Worse, we also wanted to support serial cards that were at
>non-standard ports, or using non-standard interrupts, and so that's
>one of the reason hardware initialization is deferred until after we
>have a chance to configure the serial device's I/O port and interrupt.
>
>Now, we are in the 21st century, and just as we are trying to declare
>32-bit i386 dead (to the disappointment of legacy hardware owners
>everywhere who are whining about Debian dropping installer support for
>i386), we could try to declare the ISA bus as dead, and no longer
>being supported, and we could probably drop a bunch of this
>complexity. We probably would need to support COM 1/2/3/4 ports since
>these still tend to be hardwared without a way of detecting it their
>existing via USB or PCI bus discovery mechanisms. And for those, we
>would still need to do UART autoconfiguration. I suppose we could
>just assume that all UART's are 16550A's, and that noone would
>actually use older UART's type --- except (a) I bet there are some
>cheap-skate embedded computer designers who decided to use a 8250
>instead of 16550A for $REASONS$, and because of the extreme timing
>requirements of some RS-485 use cases, I believe I have seen
>recommendations to use setserial to actually force the UART into 8250
>mode, to disable the TX FIFO --- all to satisfying the weird-shit RTS
>timing requirements. (This is where I get my beief that RS-485 !=
>RS-232, and if you want to do all of this weird-shit timing
>requirements because you are trying to implemet a multi-node bus which
>is half-duplex, maybe this should be done in the kernel, damn it!)
>
>
>How about this? If you really don't want to open the device to
>configure it not to assert DTR/RTS, perhaps we could specify on the
>boot command line that one of TTYS[0123] has a different default line
>discipline, and when tty's are opened when the RS-485 line displine,
>the RS-485 rules apply, which would include not messing with DTR/RTS,
>and if you want to play games with RTS getting dropped as soon as the
>last bit has let the transmit shift register, we can get the kernel do
>this in the line discpline, which arguably would be *way* more
>reliable than playing GPIO-style timing games in userspace.
>
> - Ted
I really think you are looking at this from a very odd point of view, and you seem to be very inconsistent. Boot time setup? Isn't that what setserial is for? We have the ability to feed this configuration already, but you need a file descriptor.
Honestly, though, I'm far less interested in what 8250-based hardware does than e.g. USB.
^ permalink raw reply
* Re: RFC: Serial port DTR/RTS - O_NRESETDEV
From: Theodore Ts'o @ 2025-11-11 4:38 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Maarten Brock, linux-serial@vger.kernel.org,
linux-api@vger.kernel.org, LKML
In-Reply-To: <D4AF3E24-8698-4EEC-9D52-655D69897111@zytor.com>
On Mon, Nov 10, 2025 at 07:57:22PM -0800, H. Peter Anvin wrote:
> I really think you are looking at this from a very odd point of
> view, and you seem to be very inconsistent. Boot time setup? Isn't
> that what setserial is for? We have the ability to feed this
> configuration already, but you need a file descriptor.
I'm not really fond of adding some new open flag that to me seems
**very** serial / RS-485 specific, and so I'm trying to find some
way to avoid it.
I also think that that the GPIO style timing requirements of RTS
**really** should be done as a line discpline, and not in userspace.
> Honestly, though, I'm far less interested in what 8250-based hardware does than e.g. USB.
I'm quite confident that USB won't have "state" that will be preserved
across a reboot, because the device won't even get powered up until
the USB device is attached. And part of the problem was that the
requirements weren't particularly clear, and given the insistence that
the "state" be preserved even across reboot, despite the serial port
autoconfiguration, I had assumed you were posting uing the COM 1/2/3/4
ports where autoconfiguration isn't stricty speaking necessary.
In some ways, USB ports might be easier, since it should be possible
to specify udev rules which get passed to the driver when the USB
serial device is inserted, and so *that* can easily be done without
needing a file descriptor.
And for this sort of thing, it seems perfectly fair to hard code some
specific behavior using either a boot command line or a udev rule,
since you seem to be positing that the serial port will be dedicated
to some kind of weird-shit RS-485 bus device, where any time RTS/DTR
gets raised, the bus will malfunction in weird and wondrous ways....
- Ted
^ permalink raw reply
* RE: RFC: Serial port DTR/RTS - O_NRESETDEV
From: Maarten Brock @ 2025-11-11 10:21 UTC (permalink / raw)
To: Theodore Ts'o, H. Peter Anvin
Cc: linux-serial@vger.kernel.org, linux-api@vger.kernel.org, LKML
In-Reply-To: <20251111043803.GK2988753@mit.edu>
Ted,
I don't understand why you drag userspace into this discussion.
The kernel drivers already support RS-485 mode for many/most
UARTs including the best timing they can give.
The reason I brought up RS-485 is because of the following situation:
Out of reset the UART has RTS# inactive. Any decent hardware design
will turn this into RS-485 Driver Enable (DE) inactive with the result that
it does not inadvertently claim the bus unnecessarily.
Now some userspace program wants to open the tty to start listening
and maybe later also to start transmitting. Opening the tty will activate
the RTS# line and thus DE and thereby claiming the bus. Any ongoing
communication will get disrupted. How's that for good behaviour?
Only after opening the port the userspace program can use the file
descriptor to configure the tty to use RS-485 mode and since it is not
transmitting it will now deactivate RTS# and DE again. But the possible
damage has been done.
Luckily if one is using a device tree one can already tell some drivers
that they need RS-485 mode and the glitch can be prevented.
But a flag to open a tty without touching the RTS# and DTR# lines would
bring the option of configuring the driver in RS-485 mode in a safe way.
And another reason why you should not handle this in a new driver is that
there are many RS-232 <-> RS-485 converters. Do you really expect users
to compile and install a new kernel just because they plugged in or out
some such converter?
Maarten
> -----Original Message-----
> From: Theodore Ts'o <tytso@mit.edu>
> Sent: Tuesday, 11 November 2025 05:38
> To: H. Peter Anvin <hpa@zytor.com>
> Cc: Maarten Brock <Maarten.Brock@sttls.nl>; linux-serial@vger.kernel.org; linux-
> api@vger.kernel.org; LKML <linux-kernel@vger.kernel.org>
> Subject: Re: RFC: Serial port DTR/RTS - O_NRESETDEV
>
> On Mon, Nov 10, 2025 at 07:57:22PM -0800, H. Peter Anvin wrote:
> > I really think you are looking at this from a very odd point of
> > view, and you seem to be very inconsistent. Boot time setup? Isn't
> > that what setserial is for? We have the ability to feed this
> > configuration already, but you need a file descriptor.
>
> I'm not really fond of adding some new open flag that to me seems
> **very** serial / RS-485 specific, and so I'm trying to find some
> way to avoid it.
>
> I also think that that the GPIO style timing requirements of RTS
> **really** should be done as a line discpline, and not in userspace.
>
> > Honestly, though, I'm far less interested in what 8250-based hardware does than
> e.g. USB.
>
> I'm quite confident that USB won't have "state" that will be preserved
> across a reboot, because the device won't even get powered up until
> the USB device is attached. And part of the problem was that the
> requirements weren't particularly clear, and given the insistence that
> the "state" be preserved even across reboot, despite the serial port
> autoconfiguration, I had assumed you were posting uing the COM 1/2/3/4
> ports where autoconfiguration isn't stricty speaking necessary.
>
> In some ways, USB ports might be easier, since it should be possible
> to specify udev rules which get passed to the driver when the USB
> serial device is inserted, and so *that* can easily be done without
> needing a file descriptor.
>
> And for this sort of thing, it seems perfectly fair to hard code some
> specific behavior using either a boot command line or a udev rule,
> since you seem to be positing that the serial port will be dedicated
> to some kind of weird-shit RS-485 bus device, where any time RTS/DTR
> gets raised, the bus will malfunction in weird and wondrous ways....
>
> - Ted
^ permalink raw reply
* [PATCH v6 00/17] vfs: recall-only directory delegations for knfsd
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
Behold, another version of the directory delegation patchset. This
version contains support for recall-only delegations. Support for
CB_NOTIFY will be forthcoming (once the client-side patches have caught
up).
The main changes here are in response to Jan's comments. I also changed
struct delegation use to fixed-with integer types.
Thanks!
Jeff
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
Changes in v6:
- Change struct delegation to use fixed-width types, and add explicit padding
- Link to v5: https://lore.kernel.org/r/20251105-dir-deleg-ro-v5-0-7ebc168a88ac@kernel.org
Changes in v5:
- drop struct createdata patch
- add patch to drop some unneeded arguments to vfs_create()
- make fcntl_getdeleg() vet delegation->d_flags
- Link to v4: https://lore.kernel.org/r/20251103-dir-deleg-ro-v4-0-961b67adee89@kernel.org
Changes in v4:
- Split lease_alloc() changes into separate patch
- new patches to switch break_lease() to use single set of flags
- add struct delegated_inode and use that instead of struct inode **
- add struct createdata and use that as argument to vfs_create()
- Rebase onto brauner/vfs-6.19.directory.delegation
- Make F_GETDELEG take and fill out struct delegation too
- Link to v3: https://lore.kernel.org/r/20251021-dir-deleg-ro-v3-0-a08b1cde9f4c@kernel.org
Changes in v3:
- Fix potential nfsd_file refcount leaks on GET_DIR_DELEGATION error
- Add missing parent dir deleg break in vfs_symlink()
- Add F_SETDELEG/F_GETDELEG support to fcntl()
- Link to v2: https://lore.kernel.org/r/20251017-dir-deleg-ro-v2-0-8c8f6dd23c8b@kernel.org
Changes in v2:
- handle lease conflict resolution inside of nfsd
- drop the lm_may_setlease lock_manager operation
- just add extra argument to vfs_create() instead of creating wrapper
- don't allocate fsnotify_mark for open directories
- Link to v1: https://lore.kernel.org/r/20251013-dir-deleg-ro-v1-0-406780a70e5e@kernel.org
---
Jeff Layton (17):
filelock: make lease_alloc() take a flags argument
filelock: rework the __break_lease API to use flags
filelock: add struct delegated_inode
filelock: push the S_ISREG check down to ->setlease handlers
vfs: add try_break_deleg calls for parents to vfs_{link,rename,unlink}
vfs: allow mkdir to wait for delegation break on parent
vfs: allow rmdir to wait for delegation break on parent
vfs: break parent dir delegations in open(..., O_CREAT) codepath
vfs: clean up argument list for vfs_create()
vfs: make vfs_create break delegations on parent directory
vfs: make vfs_mknod break delegations on parent directory
vfs: make vfs_symlink break delegations on parent dir
filelock: lift the ban on directory leases in generic_setlease
nfsd: allow filecache to hold S_IFDIR files
nfsd: allow DELEGRETURN on directories
nfsd: wire up GET_DIR_DELEGATION handling
vfs: expose delegation support to userland
drivers/base/devtmpfs.c | 6 +-
fs/attr.c | 2 +-
fs/cachefiles/namei.c | 2 +-
fs/ecryptfs/inode.c | 11 ++-
fs/fcntl.c | 13 ++++
fs/fuse/dir.c | 1 +
fs/init.c | 6 +-
fs/locks.c | 100 +++++++++++++++++++++-------
fs/namei.c | 162 +++++++++++++++++++++++++++++++++------------
fs/nfs/nfs4file.c | 2 +
fs/nfsd/filecache.c | 57 ++++++++++++----
fs/nfsd/filecache.h | 2 +
fs/nfsd/nfs3proc.c | 2 +-
fs/nfsd/nfs4proc.c | 22 +++++-
fs/nfsd/nfs4recover.c | 6 +-
fs/nfsd/nfs4state.c | 103 +++++++++++++++++++++++++++-
fs/nfsd/state.h | 5 ++
fs/nfsd/vfs.c | 16 ++---
fs/nfsd/vfs.h | 2 +-
fs/open.c | 12 ++--
fs/overlayfs/overlayfs.h | 10 +--
fs/posix_acl.c | 8 +--
fs/smb/client/cifsfs.c | 3 +
fs/smb/server/vfs.c | 9 ++-
fs/utimes.c | 4 +-
fs/xattr.c | 12 ++--
fs/xfs/scrub/orphanage.c | 2 +-
include/linux/filelock.h | 98 +++++++++++++++++++++------
include/linux/fs.h | 24 ++++---
include/linux/xattr.h | 4 +-
include/uapi/linux/fcntl.h | 11 +++
net/unix/af_unix.c | 2 +-
32 files changed, 543 insertions(+), 176 deletions(-)
---
base-commit: 3a8660878839faadb4f1a6dd72c3179c1df56787
change-id: 20251013-dir-deleg-ro-d0fe19823b21
Best regards,
--
Jeff Layton <jlayton@kernel.org>
^ permalink raw reply
* [PATCH v6 01/17] filelock: make lease_alloc() take a flags argument
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
__break_lease() currently overrides the flc_flags field in the lease
after allocating it. A forthcoming patch will add the ability to request
a FL_DELEG type lease.
Instead of overriding the flags field, add a flags argument to
lease_alloc() and lease_init() so it's set correctly after allocating.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/locks.c | 13 ++++++-------
1 file changed, 6 insertions(+), 7 deletions(-)
diff --git a/fs/locks.c b/fs/locks.c
index 04a3f0e2072461b6e2d3d1cd12f2b089d69a7db3..b33c327c21dcd49341fbeac47caeb72cdf7455db 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -585,7 +585,7 @@ static const struct lease_manager_operations lease_manager_ops = {
/*
* Initialize a lease, use the default lock manager operations
*/
-static int lease_init(struct file *filp, int type, struct file_lease *fl)
+static int lease_init(struct file *filp, unsigned int flags, int type, struct file_lease *fl)
{
if (assign_type(&fl->c, type) != 0)
return -EINVAL;
@@ -594,13 +594,13 @@ static int lease_init(struct file *filp, int type, struct file_lease *fl)
fl->c.flc_pid = current->tgid;
fl->c.flc_file = filp;
- fl->c.flc_flags = FL_LEASE;
+ fl->c.flc_flags = flags;
fl->fl_lmops = &lease_manager_ops;
return 0;
}
/* Allocate a file_lock initialised to this type of lease */
-static struct file_lease *lease_alloc(struct file *filp, int type)
+static struct file_lease *lease_alloc(struct file *filp, unsigned int flags, int type)
{
struct file_lease *fl = locks_alloc_lease();
int error = -ENOMEM;
@@ -608,7 +608,7 @@ static struct file_lease *lease_alloc(struct file *filp, int type)
if (fl == NULL)
return ERR_PTR(error);
- error = lease_init(filp, type, fl);
+ error = lease_init(filp, flags, type, fl);
if (error) {
locks_free_lease(fl);
return ERR_PTR(error);
@@ -1548,10 +1548,9 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type)
int want_write = (mode & O_ACCMODE) != O_RDONLY;
LIST_HEAD(dispose);
- new_fl = lease_alloc(NULL, want_write ? F_WRLCK : F_RDLCK);
+ new_fl = lease_alloc(NULL, type, want_write ? F_WRLCK : F_RDLCK);
if (IS_ERR(new_fl))
return PTR_ERR(new_fl);
- new_fl->c.flc_flags = type;
/* typically we will check that ctx is non-NULL before calling */
ctx = locks_inode_context(inode);
@@ -2033,7 +2032,7 @@ static int do_fcntl_add_lease(unsigned int fd, struct file *filp, int arg)
struct fasync_struct *new;
int error;
- fl = lease_alloc(filp, arg);
+ fl = lease_alloc(filp, FL_LEASE, arg);
if (IS_ERR(fl))
return PTR_ERR(fl);
--
2.51.1
^ permalink raw reply related
* [PATCH v6 02/17] filelock: rework the __break_lease API to use flags
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
Currently __break_lease takes both a type and an openmode. With the
addition of directory leases, that makes less sense. Declare a set of
LEASE_BREAK_* flags that can be used to control how lease breaks work
instead of requiring a type and an openmode.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/locks.c | 29 +++++++++++++++++----------
include/linux/filelock.h | 52 +++++++++++++++++++++++++++++++++++-------------
2 files changed, 56 insertions(+), 25 deletions(-)
diff --git a/fs/locks.c b/fs/locks.c
index b33c327c21dcd49341fbeac47caeb72cdf7455db..3cdd84a0fbedc9bd1b47725a9cf963342aafbce9 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -1529,24 +1529,31 @@ any_leases_conflict(struct inode *inode, struct file_lease *breaker)
/**
* __break_lease - revoke all outstanding leases on file
* @inode: the inode of the file to return
- * @mode: O_RDONLY: break only write leases; O_WRONLY or O_RDWR:
- * break all leases
- * @type: FL_LEASE: break leases and delegations; FL_DELEG: break
- * only delegations
+ * @flags: LEASE_BREAK_* flags
*
* break_lease (inlined for speed) has checked there already is at least
* some kind of lock (maybe a lease) on this file. Leases are broken on
- * a call to open() or truncate(). This function can sleep unless you
- * specified %O_NONBLOCK to your open().
+ * a call to open() or truncate(). This function can block waiting for the
+ * lease break unless you specify LEASE_BREAK_NONBLOCK.
*/
-int __break_lease(struct inode *inode, unsigned int mode, unsigned int type)
+int __break_lease(struct inode *inode, unsigned int flags)
{
- int error = 0;
- struct file_lock_context *ctx;
struct file_lease *new_fl, *fl, *tmp;
+ struct file_lock_context *ctx;
unsigned long break_time;
- int want_write = (mode & O_ACCMODE) != O_RDONLY;
+ unsigned int type;
LIST_HEAD(dispose);
+ bool want_write = !(flags & LEASE_BREAK_OPEN_RDONLY);
+ int error = 0;
+
+ if (flags & LEASE_BREAK_LEASE)
+ type = FL_LEASE;
+ else if (flags & LEASE_BREAK_DELEG)
+ type = FL_DELEG;
+ else if (flags & LEASE_BREAK_LAYOUT)
+ type = FL_LAYOUT;
+ else
+ return -EINVAL;
new_fl = lease_alloc(NULL, type, want_write ? F_WRLCK : F_RDLCK);
if (IS_ERR(new_fl))
@@ -1595,7 +1602,7 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type)
if (list_empty(&ctx->flc_lease))
goto out;
- if (mode & O_NONBLOCK) {
+ if (flags & LEASE_BREAK_NONBLOCK) {
trace_break_lease_noblock(inode, new_fl);
error = -EWOULDBLOCK;
goto out;
diff --git a/include/linux/filelock.h b/include/linux/filelock.h
index c2ce8ba05d068b451ecf8f513b7e532819a29944..47da6aa28d8dc9122618d02c6608deda0f3c4d3e 100644
--- a/include/linux/filelock.h
+++ b/include/linux/filelock.h
@@ -212,7 +212,14 @@ int locks_lock_inode_wait(struct inode *inode, struct file_lock *fl);
void locks_init_lease(struct file_lease *);
void locks_free_lease(struct file_lease *fl);
struct file_lease *locks_alloc_lease(void);
-int __break_lease(struct inode *inode, unsigned int flags, unsigned int type);
+
+#define LEASE_BREAK_LEASE BIT(0) // break leases and delegations
+#define LEASE_BREAK_DELEG BIT(1) // break delegations only
+#define LEASE_BREAK_LAYOUT BIT(2) // break layouts only
+#define LEASE_BREAK_NONBLOCK BIT(3) // non-blocking break
+#define LEASE_BREAK_OPEN_RDONLY BIT(4) // readonly open event
+
+int __break_lease(struct inode *inode, unsigned int flags);
void lease_get_mtime(struct inode *, struct timespec64 *time);
int generic_setlease(struct file *, int, struct file_lease **, void **priv);
int kernel_setlease(struct file *, int, struct file_lease **, void **);
@@ -367,7 +374,7 @@ static inline int locks_lock_inode_wait(struct inode *inode, struct file_lock *f
return -ENOLCK;
}
-static inline int __break_lease(struct inode *inode, unsigned int mode, unsigned int type)
+static inline int __break_lease(struct inode *inode, unsigned int flags)
{
return 0;
}
@@ -428,6 +435,17 @@ static inline int locks_lock_file_wait(struct file *filp, struct file_lock *fl)
}
#ifdef CONFIG_FILE_LOCKING
+static inline unsigned int openmode_to_lease_flags(unsigned int mode)
+{
+ unsigned int flags = 0;
+
+ if ((mode & O_ACCMODE) == O_RDONLY)
+ flags |= LEASE_BREAK_OPEN_RDONLY;
+ if (mode & O_NONBLOCK)
+ flags |= LEASE_BREAK_NONBLOCK;
+ return flags;
+}
+
static inline int break_lease(struct inode *inode, unsigned int mode)
{
struct file_lock_context *flctx;
@@ -443,11 +461,11 @@ static inline int break_lease(struct inode *inode, unsigned int mode)
return 0;
smp_mb();
if (!list_empty_careful(&flctx->flc_lease))
- return __break_lease(inode, mode, FL_LEASE);
+ return __break_lease(inode, LEASE_BREAK_LEASE | openmode_to_lease_flags(mode));
return 0;
}
-static inline int break_deleg(struct inode *inode, unsigned int mode)
+static inline int break_deleg(struct inode *inode, unsigned int flags)
{
struct file_lock_context *flctx;
@@ -461,8 +479,10 @@ static inline int break_deleg(struct inode *inode, unsigned int mode)
if (!flctx)
return 0;
smp_mb();
- if (!list_empty_careful(&flctx->flc_lease))
- return __break_lease(inode, mode, FL_DELEG);
+ if (!list_empty_careful(&flctx->flc_lease)) {
+ flags |= LEASE_BREAK_DELEG;
+ return __break_lease(inode, flags);
+ }
return 0;
}
@@ -470,7 +490,7 @@ static inline int try_break_deleg(struct inode *inode, struct inode **delegated_
{
int ret;
- ret = break_deleg(inode, O_WRONLY|O_NONBLOCK);
+ ret = break_deleg(inode, LEASE_BREAK_NONBLOCK);
if (ret == -EWOULDBLOCK && delegated_inode) {
*delegated_inode = inode;
ihold(inode);
@@ -482,7 +502,7 @@ static inline int break_deleg_wait(struct inode **delegated_inode)
{
int ret;
- ret = break_deleg(*delegated_inode, O_WRONLY);
+ ret = break_deleg(*delegated_inode, 0);
iput(*delegated_inode);
*delegated_inode = NULL;
return ret;
@@ -491,20 +511,24 @@ static inline int break_deleg_wait(struct inode **delegated_inode)
static inline int break_layout(struct inode *inode, bool wait)
{
smp_mb();
- if (inode->i_flctx && !list_empty_careful(&inode->i_flctx->flc_lease))
- return __break_lease(inode,
- wait ? O_WRONLY : O_WRONLY | O_NONBLOCK,
- FL_LAYOUT);
+ if (inode->i_flctx && !list_empty_careful(&inode->i_flctx->flc_lease)) {
+ unsigned int flags = LEASE_BREAK_LAYOUT;
+
+ if (!wait)
+ flags |= LEASE_BREAK_NONBLOCK;
+
+ return __break_lease(inode, flags);
+ }
return 0;
}
#else /* !CONFIG_FILE_LOCKING */
-static inline int break_lease(struct inode *inode, unsigned int mode)
+static inline int break_lease(struct inode *inode, bool wait)
{
return 0;
}
-static inline int break_deleg(struct inode *inode, unsigned int mode)
+static inline int break_deleg(struct inode *inode, unsigned int flags)
{
return 0;
}
--
2.51.1
^ permalink raw reply related
* [PATCH v6 03/17] filelock: add struct delegated_inode
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
The current API requires a pointer to an inode pointer. It's easy for
callers to get this wrong. Add a new delegated_inode structure and use
that to pass back any inode that needs to be waited on.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/attr.c | 2 +-
fs/namei.c | 18 +++++++++---------
fs/open.c | 8 ++++----
fs/posix_acl.c | 8 ++++----
fs/utimes.c | 4 ++--
fs/xattr.c | 12 ++++++------
include/linux/filelock.h | 36 +++++++++++++++++++++++++++---------
include/linux/fs.h | 9 +++++----
include/linux/xattr.h | 4 ++--
9 files changed, 60 insertions(+), 41 deletions(-)
diff --git a/fs/attr.c b/fs/attr.c
index 795f231d00e8eaaadf5b62f241655cb4b69cb507..b9ec6b47bab2fc2b561677b639633bd32994022f 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -415,7 +415,7 @@ EXPORT_SYMBOL(may_setattr);
* performed on the raw inode simply pass @nop_mnt_idmap.
*/
int notify_change(struct mnt_idmap *idmap, struct dentry *dentry,
- struct iattr *attr, struct inode **delegated_inode)
+ struct iattr *attr, struct delegated_inode *delegated_inode)
{
struct inode *inode = dentry->d_inode;
umode_t mode = inode->i_mode;
diff --git a/fs/namei.c b/fs/namei.c
index 7377020a2cba02501483020e0fc93c279fb38d3e..bf42f146f847a5330fc581595c7256af28d9db90 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -4648,7 +4648,7 @@ SYSCALL_DEFINE1(rmdir, const char __user *, pathname)
* raw inode simply pass @nop_mnt_idmap.
*/
int vfs_unlink(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry, struct inode **delegated_inode)
+ struct dentry *dentry, struct delegated_inode *delegated_inode)
{
struct inode *target = dentry->d_inode;
int error = may_delete(idmap, dir, dentry, 0);
@@ -4706,7 +4706,7 @@ int do_unlinkat(int dfd, struct filename *name)
struct qstr last;
int type;
struct inode *inode = NULL;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
unsigned int lookup_flags = 0;
retry:
error = filename_parentat(dfd, name, lookup_flags, &path, &last, &type);
@@ -4743,7 +4743,7 @@ int do_unlinkat(int dfd, struct filename *name)
if (inode)
iput(inode); /* truncate the inode here */
inode = NULL;
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
@@ -4892,7 +4892,7 @@ SYSCALL_DEFINE2(symlink, const char __user *, oldname, const char __user *, newn
*/
int vfs_link(struct dentry *old_dentry, struct mnt_idmap *idmap,
struct inode *dir, struct dentry *new_dentry,
- struct inode **delegated_inode)
+ struct delegated_inode *delegated_inode)
{
struct inode *inode = old_dentry->d_inode;
unsigned max_links = dir->i_sb->s_max_links;
@@ -4968,7 +4968,7 @@ int do_linkat(int olddfd, struct filename *old, int newdfd,
struct mnt_idmap *idmap;
struct dentry *new_dentry;
struct path old_path, new_path;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
int how = 0;
int error;
@@ -5012,7 +5012,7 @@ int do_linkat(int olddfd, struct filename *old, int newdfd,
new_dentry, &delegated_inode);
out_dput:
end_creating_path(&new_path, new_dentry);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error) {
path_put(&old_path);
@@ -5098,7 +5098,7 @@ int vfs_rename(struct renamedata *rd)
struct inode *new_dir = d_inode(rd->new_parent);
struct dentry *old_dentry = rd->old_dentry;
struct dentry *new_dentry = rd->new_dentry;
- struct inode **delegated_inode = rd->delegated_inode;
+ struct delegated_inode *delegated_inode = rd->delegated_inode;
unsigned int flags = rd->flags;
bool is_dir = d_is_dir(old_dentry);
struct inode *source = old_dentry->d_inode;
@@ -5261,7 +5261,7 @@ int do_renameat2(int olddfd, struct filename *from, int newdfd,
struct path old_path, new_path;
struct qstr old_last, new_last;
int old_type, new_type;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
unsigned int lookup_flags = 0, target_flags =
LOOKUP_RENAME_TARGET | LOOKUP_CREATE;
bool should_retry = false;
@@ -5369,7 +5369,7 @@ int do_renameat2(int olddfd, struct filename *from, int newdfd,
exit3:
unlock_rename(new_path.dentry, old_path.dentry);
exit_lock_rename:
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
diff --git a/fs/open.c b/fs/open.c
index 3d64372ecc675e4795eb0a0deda10f8f67b95640..fdaa6f08f6f4cac5c2fefd3eafa5e430e51f3979 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -631,7 +631,7 @@ SYSCALL_DEFINE1(chroot, const char __user *, filename)
int chmod_common(const struct path *path, umode_t mode)
{
struct inode *inode = path->dentry->d_inode;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
struct iattr newattrs;
int error;
@@ -651,7 +651,7 @@ int chmod_common(const struct path *path, umode_t mode)
&newattrs, &delegated_inode);
out_unlock:
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
@@ -756,7 +756,7 @@ int chown_common(const struct path *path, uid_t user, gid_t group)
struct mnt_idmap *idmap;
struct user_namespace *fs_userns;
struct inode *inode = path->dentry->d_inode;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
int error;
struct iattr newattrs;
kuid_t uid;
@@ -791,7 +791,7 @@ int chown_common(const struct path *path, uid_t user, gid_t group)
error = notify_change(idmap, path->dentry, &newattrs,
&delegated_inode);
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
diff --git a/fs/posix_acl.c b/fs/posix_acl.c
index 4050942ab52f95741da2df13d191ade5c5ca12a2..768f027c142811ea907fe8545155ba7abd016305 100644
--- a/fs/posix_acl.c
+++ b/fs/posix_acl.c
@@ -1091,7 +1091,7 @@ int vfs_set_acl(struct mnt_idmap *idmap, struct dentry *dentry,
int acl_type;
int error;
struct inode *inode = d_inode(dentry);
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
acl_type = posix_acl_type(acl_name);
if (acl_type < 0)
@@ -1141,7 +1141,7 @@ int vfs_set_acl(struct mnt_idmap *idmap, struct dentry *dentry,
out_inode_unlock:
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
@@ -1212,7 +1212,7 @@ int vfs_remove_acl(struct mnt_idmap *idmap, struct dentry *dentry,
int acl_type;
int error;
struct inode *inode = d_inode(dentry);
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
acl_type = posix_acl_type(acl_name);
if (acl_type < 0)
@@ -1249,7 +1249,7 @@ int vfs_remove_acl(struct mnt_idmap *idmap, struct dentry *dentry,
out_inode_unlock:
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
diff --git a/fs/utimes.c b/fs/utimes.c
index c7c7958e57b22f91646ca9f76d18781b64d371a3..bf9f45bdef54947de7ac55c9f873ae9d0336dafa 100644
--- a/fs/utimes.c
+++ b/fs/utimes.c
@@ -22,7 +22,7 @@ int vfs_utimes(const struct path *path, struct timespec64 *times)
int error;
struct iattr newattrs;
struct inode *inode = path->dentry->d_inode;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
if (times) {
if (!nsec_valid(times[0].tv_nsec) ||
@@ -66,7 +66,7 @@ int vfs_utimes(const struct path *path, struct timespec64 *times)
error = notify_change(mnt_idmap(path->mnt), path->dentry, &newattrs,
&delegated_inode);
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
diff --git a/fs/xattr.c b/fs/xattr.c
index 8851a5ef34f5ab34383975dd4cef537de3f6391e..32d445fb60aaf2aaf4b16b62934dc99bad378067 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c
@@ -274,7 +274,7 @@ int __vfs_setxattr_noperm(struct mnt_idmap *idmap,
int
__vfs_setxattr_locked(struct mnt_idmap *idmap, struct dentry *dentry,
const char *name, const void *value, size_t size,
- int flags, struct inode **delegated_inode)
+ int flags, struct delegated_inode *delegated_inode)
{
struct inode *inode = dentry->d_inode;
int error;
@@ -305,7 +305,7 @@ vfs_setxattr(struct mnt_idmap *idmap, struct dentry *dentry,
const char *name, const void *value, size_t size, int flags)
{
struct inode *inode = dentry->d_inode;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
const void *orig_value = value;
int error;
@@ -322,7 +322,7 @@ vfs_setxattr(struct mnt_idmap *idmap, struct dentry *dentry,
flags, &delegated_inode);
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
@@ -533,7 +533,7 @@ EXPORT_SYMBOL(__vfs_removexattr);
int
__vfs_removexattr_locked(struct mnt_idmap *idmap,
struct dentry *dentry, const char *name,
- struct inode **delegated_inode)
+ struct delegated_inode *delegated_inode)
{
struct inode *inode = dentry->d_inode;
int error;
@@ -567,7 +567,7 @@ vfs_removexattr(struct mnt_idmap *idmap, struct dentry *dentry,
const char *name)
{
struct inode *inode = dentry->d_inode;
- struct inode *delegated_inode = NULL;
+ struct delegated_inode delegated_inode = { };
int error;
retry_deleg:
@@ -576,7 +576,7 @@ vfs_removexattr(struct mnt_idmap *idmap, struct dentry *dentry,
name, &delegated_inode);
inode_unlock(inode);
- if (delegated_inode) {
+ if (is_delegated(&delegated_inode)) {
error = break_deleg_wait(&delegated_inode);
if (!error)
goto retry_deleg;
diff --git a/include/linux/filelock.h b/include/linux/filelock.h
index 47da6aa28d8dc9122618d02c6608deda0f3c4d3e..208d108df2d73a9df65e5dc9968d074af385f881 100644
--- a/include/linux/filelock.h
+++ b/include/linux/filelock.h
@@ -486,25 +486,35 @@ static inline int break_deleg(struct inode *inode, unsigned int flags)
return 0;
}
-static inline int try_break_deleg(struct inode *inode, struct inode **delegated_inode)
+struct delegated_inode {
+ struct inode *di_inode;
+};
+
+static inline bool is_delegated(struct delegated_inode *di)
+{
+ return di->di_inode;
+}
+
+static inline int try_break_deleg(struct inode *inode,
+ struct delegated_inode *di)
{
int ret;
ret = break_deleg(inode, LEASE_BREAK_NONBLOCK);
- if (ret == -EWOULDBLOCK && delegated_inode) {
- *delegated_inode = inode;
+ if (ret == -EWOULDBLOCK && di) {
+ di->di_inode = inode;
ihold(inode);
}
return ret;
}
-static inline int break_deleg_wait(struct inode **delegated_inode)
+static inline int break_deleg_wait(struct delegated_inode *di)
{
int ret;
- ret = break_deleg(*delegated_inode, 0);
- iput(*delegated_inode);
- *delegated_inode = NULL;
+ ret = break_deleg(di->di_inode, 0);
+ iput(di->di_inode);
+ di->di_inode = NULL;
return ret;
}
@@ -523,6 +533,13 @@ static inline int break_layout(struct inode *inode, bool wait)
}
#else /* !CONFIG_FILE_LOCKING */
+struct delegated_inode { };
+
+static inline bool is_delegated(struct delegated_inode *di)
+{
+ return false;
+}
+
static inline int break_lease(struct inode *inode, bool wait)
{
return 0;
@@ -533,12 +550,13 @@ static inline int break_deleg(struct inode *inode, unsigned int flags)
return 0;
}
-static inline int try_break_deleg(struct inode *inode, struct inode **delegated_inode)
+static inline int try_break_deleg(struct inode *inode,
+ struct delegated_inode *delegated_inode)
{
return 0;
}
-static inline int break_deleg_wait(struct inode **delegated_inode)
+static inline int break_deleg_wait(struct delegated_inode *delegated_inode)
{
BUG();
return 0;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index c895146c1444be36e0a779df55622cc38c9419ff..909a88e3979d4f1ba3104f3d05145e1096ed44d5 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -80,6 +80,7 @@ struct fs_context;
struct fs_parameter_spec;
struct file_kattr;
struct iomap_ops;
+struct delegated_inode;
extern void __init inode_init(void);
extern void __init inode_init_early(void);
@@ -2119,10 +2120,10 @@ int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
int vfs_symlink(struct mnt_idmap *, struct inode *,
struct dentry *, const char *);
int vfs_link(struct dentry *, struct mnt_idmap *, struct inode *,
- struct dentry *, struct inode **);
+ struct dentry *, struct delegated_inode *);
int vfs_rmdir(struct mnt_idmap *, struct inode *, struct dentry *);
int vfs_unlink(struct mnt_idmap *, struct inode *, struct dentry *,
- struct inode **);
+ struct delegated_inode *);
/**
* struct renamedata - contains all information required for renaming
@@ -2140,7 +2141,7 @@ struct renamedata {
struct dentry *old_dentry;
struct dentry *new_parent;
struct dentry *new_dentry;
- struct inode **delegated_inode;
+ struct delegated_inode *delegated_inode;
unsigned int flags;
} __randomize_layout;
@@ -3071,7 +3072,7 @@ static inline int bmap(struct inode *inode, sector_t *block)
#endif
int notify_change(struct mnt_idmap *, struct dentry *,
- struct iattr *, struct inode **);
+ struct iattr *, struct delegated_inode *);
int inode_permission(struct mnt_idmap *, struct inode *, int);
int generic_permission(struct mnt_idmap *, struct inode *, int);
static inline int file_permission(struct file *file, int mask)
diff --git a/include/linux/xattr.h b/include/linux/xattr.h
index 86b0d47984a16d935dd1c45ca80a3b8bb5b7295b..64e9afe7d647dc38f686a4b5c6f765e061cde54c 100644
--- a/include/linux/xattr.h
+++ b/include/linux/xattr.h
@@ -85,12 +85,12 @@ int __vfs_setxattr_noperm(struct mnt_idmap *, struct dentry *,
const char *, const void *, size_t, int);
int __vfs_setxattr_locked(struct mnt_idmap *, struct dentry *,
const char *, const void *, size_t, int,
- struct inode **);
+ struct delegated_inode *);
int vfs_setxattr(struct mnt_idmap *, struct dentry *, const char *,
const void *, size_t, int);
int __vfs_removexattr(struct mnt_idmap *, struct dentry *, const char *);
int __vfs_removexattr_locked(struct mnt_idmap *, struct dentry *,
- const char *, struct inode **);
+ const char *, struct delegated_inode *);
int vfs_removexattr(struct mnt_idmap *, struct dentry *, const char *);
ssize_t generic_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size);
--
2.51.1
^ permalink raw reply related
* [PATCH v6 04/17] filelock: push the S_ISREG check down to ->setlease handlers
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
When nfsd starts requesting directory delegations, setlease handlers may
see requests for leases on directories. Push the !S_ISREG check down
into the non-trivial setlease handlers, so we can selectively enable
them where they're supported.
FUSE is special: It's the only filesystem that supports atomic_open and
allows kernel-internal leases. atomic_open is issued when the VFS
doesn't know the state of the dentry being opened. If the file doesn't
exist, it may be created, in which case the dir lease should be broken.
The existing kernel-internal lease implementation has no provision for
this. Ensure that we don't allow directory leases by default going
forward by explicitly disabling them there.
Reviewed-by: NeilBrown <neil@brown.name>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/fuse/dir.c | 1 +
fs/locks.c | 5 +++--
fs/nfs/nfs4file.c | 2 ++
fs/smb/client/cifsfs.c | 3 +++
4 files changed, 9 insertions(+), 2 deletions(-)
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index ecaec0fea3a132e7cbb88121e7db7fb504d57d3c..667774cc72a1d49796f531fcb342d2e4878beb85 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -2230,6 +2230,7 @@ static const struct file_operations fuse_dir_operations = {
.fsync = fuse_dir_fsync,
.unlocked_ioctl = fuse_dir_ioctl,
.compat_ioctl = fuse_dir_compat_ioctl,
+ .setlease = simple_nosetlease,
};
static const struct inode_operations fuse_common_inode_operations = {
diff --git a/fs/locks.c b/fs/locks.c
index 3cdd84a0fbedc9bd1b47725a9cf963342aafbce9..f5b210a2dc34c70ac36e972436c62482bbe32ca6 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -1935,6 +1935,9 @@ static int generic_delete_lease(struct file *filp, void *owner)
int generic_setlease(struct file *filp, int arg, struct file_lease **flp,
void **priv)
{
+ if (!S_ISREG(file_inode(filp)->i_mode))
+ return -EINVAL;
+
switch (arg) {
case F_UNLCK:
return generic_delete_lease(filp, *priv);
@@ -2024,8 +2027,6 @@ vfs_setlease(struct file *filp, int arg, struct file_lease **lease, void **priv)
if ((!vfsuid_eq_kuid(vfsuid, current_fsuid())) && !capable(CAP_LEASE))
return -EACCES;
- if (!S_ISREG(inode->i_mode))
- return -EINVAL;
error = security_file_lock(filp, arg);
if (error)
return error;
diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c
index 7f43e890d3564a000dab9365048a3e17dc96395c..7317f26892c5782a39660cae87ec1afea24e36c0 100644
--- a/fs/nfs/nfs4file.c
+++ b/fs/nfs/nfs4file.c
@@ -431,6 +431,8 @@ void nfs42_ssc_unregister_ops(void)
static int nfs4_setlease(struct file *file, int arg, struct file_lease **lease,
void **priv)
{
+ if (!S_ISREG(file_inode(file)->i_mode))
+ return -EINVAL;
return nfs4_proc_setlease(file, arg, lease, priv);
}
diff --git a/fs/smb/client/cifsfs.c b/fs/smb/client/cifsfs.c
index 05b1fa76e8ccf1e86f0c174593cd6e1acb84608d..03c44c1d9bb631b87a8b67aa16e481d6bb3c7d14 100644
--- a/fs/smb/client/cifsfs.c
+++ b/fs/smb/client/cifsfs.c
@@ -1149,6 +1149,9 @@ cifs_setlease(struct file *file, int arg, struct file_lease **lease, void **priv
struct inode *inode = file_inode(file);
struct cifsFileInfo *cfile = file->private_data;
+ if (!S_ISREG(inode->i_mode))
+ return -EINVAL;
+
/* Check if file is oplocked if this is request for new lease */
if (arg == F_UNLCK ||
((arg == F_RDLCK) && CIFS_CACHE_READ(CIFS_I(inode))) ||
--
2.51.1
^ permalink raw reply related
* [PATCH v6 05/17] vfs: add try_break_deleg calls for parents to vfs_{link,rename,unlink}
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
vfs_link, vfs_unlink, and vfs_rename all have existing delegation break
handling for the children in the rename. Add the necessary calls for
breaking delegations in the parent(s) as well.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/namei.c | 15 ++++++++++++++-
1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/fs/namei.c b/fs/namei.c
index bf42f146f847a5330fc581595c7256af28d9db90..5bcf3e93d350ffd290f72725c378d3dffeeae364 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -4667,6 +4667,9 @@ int vfs_unlink(struct mnt_idmap *idmap, struct inode *dir,
else {
error = security_inode_unlink(dir, dentry);
if (!error) {
+ error = try_break_deleg(dir, delegated_inode);
+ if (error)
+ goto out;
error = try_break_deleg(target, delegated_inode);
if (error)
goto out;
@@ -4936,7 +4939,9 @@ int vfs_link(struct dentry *old_dentry, struct mnt_idmap *idmap,
else if (max_links && inode->i_nlink >= max_links)
error = -EMLINK;
else {
- error = try_break_deleg(inode, delegated_inode);
+ error = try_break_deleg(dir, delegated_inode);
+ if (!error)
+ error = try_break_deleg(inode, delegated_inode);
if (!error)
error = dir->i_op->link(old_dentry, dir, new_dentry);
}
@@ -5203,6 +5208,14 @@ int vfs_rename(struct renamedata *rd)
old_dir->i_nlink >= max_links)
goto out;
}
+ error = try_break_deleg(old_dir, delegated_inode);
+ if (error)
+ goto out;
+ if (new_dir != old_dir) {
+ error = try_break_deleg(new_dir, delegated_inode);
+ if (error)
+ goto out;
+ }
if (!is_dir) {
error = try_break_deleg(source, delegated_inode);
if (error)
--
2.51.1
^ permalink raw reply related
* [PATCH v6 06/17] vfs: allow mkdir to wait for delegation break on parent
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
Add a new delegated_inode parameter to vfs_mkdir. All of the existing
callers set that to NULL for now, except for do_mkdirat which will
properly block until the lease is gone.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
drivers/base/devtmpfs.c | 2 +-
fs/cachefiles/namei.c | 2 +-
fs/ecryptfs/inode.c | 2 +-
fs/init.c | 2 +-
fs/namei.c | 24 ++++++++++++++++++------
fs/nfsd/nfs4recover.c | 2 +-
fs/nfsd/vfs.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
fs/smb/server/vfs.c | 2 +-
fs/xfs/scrub/orphanage.c | 2 +-
include/linux/fs.h | 2 +-
11 files changed, 28 insertions(+), 16 deletions(-)
diff --git a/drivers/base/devtmpfs.c b/drivers/base/devtmpfs.c
index 9d4e46ad8352257a6a65d85526ebdbf9bf2d4b19..0e79621cb0f79870003b867ca384199171ded4e0 100644
--- a/drivers/base/devtmpfs.c
+++ b/drivers/base/devtmpfs.c
@@ -180,7 +180,7 @@ static int dev_mkdir(const char *name, umode_t mode)
if (IS_ERR(dentry))
return PTR_ERR(dentry);
- dentry = vfs_mkdir(&nop_mnt_idmap, d_inode(path.dentry), dentry, mode);
+ dentry = vfs_mkdir(&nop_mnt_idmap, d_inode(path.dentry), dentry, mode, NULL);
if (!IS_ERR(dentry))
/* mark as kernel-created inode */
d_inode(dentry)->i_private = &thread;
diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c
index d1edb2ac38376c4f9d2a18026450bb3c774f7824..50c0f9c76d1fd4c05db90d7d0d1bad574523ead0 100644
--- a/fs/cachefiles/namei.c
+++ b/fs/cachefiles/namei.c
@@ -130,7 +130,7 @@ struct dentry *cachefiles_get_directory(struct cachefiles_cache *cache,
goto mkdir_error;
ret = cachefiles_inject_write_error();
if (ret == 0)
- subdir = vfs_mkdir(&nop_mnt_idmap, d_inode(dir), subdir, 0700);
+ subdir = vfs_mkdir(&nop_mnt_idmap, d_inode(dir), subdir, 0700, NULL);
else
subdir = ERR_PTR(ret);
if (IS_ERR(subdir)) {
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index ed1394da8d6bd7065f2a074378331f13fcda17f9..35830b3144f8f71374a78b3e7463b864f4fc216e 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -508,7 +508,7 @@ static struct dentry *ecryptfs_mkdir(struct mnt_idmap *idmap, struct inode *dir,
goto out;
lower_dentry = vfs_mkdir(&nop_mnt_idmap, lower_dir,
- lower_dentry, mode);
+ lower_dentry, mode, NULL);
rc = PTR_ERR(lower_dentry);
if (IS_ERR(lower_dentry))
goto out;
diff --git a/fs/init.c b/fs/init.c
index 07f592ccdba868509d0f3aaf9936d8d890fdbec5..895f8a09a71acfd03e11164e3b441a7d4e2de146 100644
--- a/fs/init.c
+++ b/fs/init.c
@@ -233,7 +233,7 @@ int __init init_mkdir(const char *pathname, umode_t mode)
error = security_path_mkdir(&path, dentry, mode);
if (!error) {
dentry = vfs_mkdir(mnt_idmap(path.mnt), path.dentry->d_inode,
- dentry, mode);
+ dentry, mode, NULL);
if (IS_ERR(dentry))
error = PTR_ERR(dentry);
}
diff --git a/fs/namei.c b/fs/namei.c
index 5bcf3e93d350ffd290f72725c378d3dffeeae364..76c0587d991ff7307e3dde69497719d716c8d7b8 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -4407,10 +4407,11 @@ SYSCALL_DEFINE3(mknod, const char __user *, filename, umode_t, mode, unsigned, d
/**
* vfs_mkdir - create directory returning correct dentry if possible
- * @idmap: idmap of the mount the inode was found from
- * @dir: inode of the parent directory
- * @dentry: dentry of the child directory
- * @mode: mode of the child directory
+ * @idmap: idmap of the mount the inode was found from
+ * @dir: inode of the parent directory
+ * @dentry: dentry of the child directory
+ * @mode: mode of the child directory
+ * @delegated_inode: returns parent inode, if the inode is delegated.
*
* Create a directory.
*
@@ -4427,7 +4428,8 @@ SYSCALL_DEFINE3(mknod, const char __user *, filename, umode_t, mode, unsigned, d
* In case of an error the dentry is dput() and an ERR_PTR() is returned.
*/
struct dentry *vfs_mkdir(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry, umode_t mode)
+ struct dentry *dentry, umode_t mode,
+ struct delegated_inode *delegated_inode)
{
int error;
unsigned max_links = dir->i_sb->s_max_links;
@@ -4450,6 +4452,10 @@ struct dentry *vfs_mkdir(struct mnt_idmap *idmap, struct inode *dir,
if (max_links && dir->i_nlink >= max_links)
goto err;
+ error = try_break_deleg(dir, delegated_inode);
+ if (error)
+ goto err;
+
de = dir->i_op->mkdir(idmap, dir, dentry, mode);
error = PTR_ERR(de);
if (IS_ERR(de))
@@ -4473,6 +4479,7 @@ int do_mkdirat(int dfd, struct filename *name, umode_t mode)
struct path path;
int error;
unsigned int lookup_flags = LOOKUP_DIRECTORY;
+ struct delegated_inode delegated_inode = { };
retry:
dentry = filename_create(dfd, name, &path, lookup_flags);
@@ -4484,11 +4491,16 @@ int do_mkdirat(int dfd, struct filename *name, umode_t mode)
mode_strip_umask(path.dentry->d_inode, mode));
if (!error) {
dentry = vfs_mkdir(mnt_idmap(path.mnt), path.dentry->d_inode,
- dentry, mode);
+ dentry, mode, &delegated_inode);
if (IS_ERR(dentry))
error = PTR_ERR(dentry);
}
end_creating_path(&path, dentry);
+ if (is_delegated(&delegated_inode)) {
+ error = break_deleg_wait(&delegated_inode);
+ if (!error)
+ goto retry;
+ }
if (retry_estale(error, lookup_flags)) {
lookup_flags |= LOOKUP_REVAL;
goto retry;
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index e2b9472e5c78c9f03731090ffdfb26eb5de38fe0..1f56834b2072fcee1d0d400bbb554b0c949ecab4 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -213,7 +213,7 @@ nfsd4_create_clid_dir(struct nfs4_client *clp)
* as well be forgiving and just succeed silently.
*/
goto out_put;
- dentry = vfs_mkdir(&nop_mnt_idmap, d_inode(dir), dentry, S_IRWXU);
+ dentry = vfs_mkdir(&nop_mnt_idmap, d_inode(dir), dentry, 0700, NULL);
if (IS_ERR(dentry))
status = PTR_ERR(dentry);
out_put:
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 9cb20d4aeab159ef3ba3584d1a3a33ef16ba4dea..97aef140cbf5fca4c41738fdcaccba3b57886463 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1558,7 +1558,7 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
nfsd_check_ignore_resizing(iap);
break;
case S_IFDIR:
- dchild = vfs_mkdir(&nop_mnt_idmap, dirp, dchild, iap->ia_mode);
+ dchild = vfs_mkdir(&nop_mnt_idmap, dirp, dchild, iap->ia_mode, NULL);
if (IS_ERR(dchild)) {
host_err = PTR_ERR(dchild);
} else if (d_is_negative(dchild)) {
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index c8fd5951fc5ece1ae6b3e2a0801ca15f9faf7d72..0f65f9a5d54d4786b39e4f4f30f416d5b9016e70 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -248,7 +248,7 @@ static inline struct dentry *ovl_do_mkdir(struct ovl_fs *ofs,
{
struct dentry *ret;
- ret = vfs_mkdir(ovl_upper_mnt_idmap(ofs), dir, dentry, mode);
+ ret = vfs_mkdir(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, NULL);
pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, PTR_ERR_OR_ZERO(ret));
return ret;
}
diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c
index 891ed2dc2b7351a5cb14a2241d71095ffdd03f08..3d2190f26623b23ea79c63410905a3c3ad684048 100644
--- a/fs/smb/server/vfs.c
+++ b/fs/smb/server/vfs.c
@@ -230,7 +230,7 @@ int ksmbd_vfs_mkdir(struct ksmbd_work *work, const char *name, umode_t mode)
idmap = mnt_idmap(path.mnt);
mode |= S_IFDIR;
d = dentry;
- dentry = vfs_mkdir(idmap, d_inode(path.dentry), dentry, mode);
+ dentry = vfs_mkdir(idmap, d_inode(path.dentry), dentry, mode, NULL);
if (IS_ERR(dentry))
err = PTR_ERR(dentry);
else if (d_is_negative(dentry))
diff --git a/fs/xfs/scrub/orphanage.c b/fs/xfs/scrub/orphanage.c
index 9c12cb8442311ca26b169e4d1567939ae44a5be0..91c9d07b97f306f57aebb9b69ba564b0c2cb8c17 100644
--- a/fs/xfs/scrub/orphanage.c
+++ b/fs/xfs/scrub/orphanage.c
@@ -167,7 +167,7 @@ xrep_orphanage_create(
*/
if (d_really_is_negative(orphanage_dentry)) {
orphanage_dentry = vfs_mkdir(&nop_mnt_idmap, root_inode,
- orphanage_dentry, 0750);
+ orphanage_dentry, 0750, NULL);
error = PTR_ERR(orphanage_dentry);
if (IS_ERR(orphanage_dentry))
goto out_unlock_root;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 909a88e3979d4f1ba3104f3d05145e1096ed44d5..20bb4c8a4e8e1be7e11047d228c05920ea6c388d 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2114,7 +2114,7 @@ bool inode_owner_or_capable(struct mnt_idmap *idmap,
int vfs_create(struct mnt_idmap *, struct inode *,
struct dentry *, umode_t, bool);
struct dentry *vfs_mkdir(struct mnt_idmap *, struct inode *,
- struct dentry *, umode_t);
+ struct dentry *, umode_t, struct delegated_inode *);
int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
umode_t, dev_t);
int vfs_symlink(struct mnt_idmap *, struct inode *,
--
2.51.1
^ permalink raw reply related
* [PATCH v6 07/17] vfs: allow rmdir to wait for delegation break on parent
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
Add a delegated_inode struct to vfs_rmdir() and populate that
pointer with the parent inode if it's non-NULL. Most existing in-kernel
callers pass in a NULL pointer.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
drivers/base/devtmpfs.c | 2 +-
fs/ecryptfs/inode.c | 2 +-
fs/namei.c | 22 +++++++++++++++++-----
fs/nfsd/nfs4recover.c | 4 ++--
fs/nfsd/vfs.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
fs/smb/server/vfs.c | 4 ++--
include/linux/fs.h | 3 ++-
8 files changed, 27 insertions(+), 14 deletions(-)
diff --git a/drivers/base/devtmpfs.c b/drivers/base/devtmpfs.c
index 0e79621cb0f79870003b867ca384199171ded4e0..104025104ef75381984fd94dfbd50feeaa8cdd22 100644
--- a/drivers/base/devtmpfs.c
+++ b/drivers/base/devtmpfs.c
@@ -261,7 +261,7 @@ static int dev_rmdir(const char *name)
return PTR_ERR(dentry);
if (d_inode(dentry)->i_private == &thread)
err = vfs_rmdir(&nop_mnt_idmap, d_inode(parent.dentry),
- dentry);
+ dentry, NULL);
else
err = -EPERM;
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 35830b3144f8f71374a78b3e7463b864f4fc216e..88631291b32535f623a3fbe4ea9b6ed48a306ca0 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -540,7 +540,7 @@ static int ecryptfs_rmdir(struct inode *dir, struct dentry *dentry)
if (d_unhashed(lower_dentry))
rc = -EINVAL;
else
- rc = vfs_rmdir(&nop_mnt_idmap, lower_dir, lower_dentry);
+ rc = vfs_rmdir(&nop_mnt_idmap, lower_dir, lower_dentry, NULL);
}
if (!rc) {
clear_nlink(d_inode(dentry));
diff --git a/fs/namei.c b/fs/namei.c
index 76c0587d991ff7307e3dde69497719d716c8d7b8..9e0393a92091ac522b5324fcdad8c5592a948e8d 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -4522,9 +4522,10 @@ SYSCALL_DEFINE2(mkdir, const char __user *, pathname, umode_t, mode)
/**
* vfs_rmdir - remove directory
- * @idmap: idmap of the mount the inode was found from
- * @dir: inode of the parent directory
- * @dentry: dentry of the child directory
+ * @idmap: idmap of the mount the inode was found from
+ * @dir: inode of the parent directory
+ * @dentry: dentry of the child directory
+ * @delegated_inode: returns parent inode, if it's delegated.
*
* Remove a directory.
*
@@ -4535,7 +4536,7 @@ SYSCALL_DEFINE2(mkdir, const char __user *, pathname, umode_t, mode)
* raw inode simply pass @nop_mnt_idmap.
*/
int vfs_rmdir(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry)
+ struct dentry *dentry, struct delegated_inode *delegated_inode)
{
int error = may_delete(idmap, dir, dentry, 1);
@@ -4557,6 +4558,10 @@ int vfs_rmdir(struct mnt_idmap *idmap, struct inode *dir,
if (error)
goto out;
+ error = try_break_deleg(dir, delegated_inode);
+ if (error)
+ goto out;
+
error = dir->i_op->rmdir(dir, dentry);
if (error)
goto out;
@@ -4583,6 +4588,7 @@ int do_rmdir(int dfd, struct filename *name)
struct qstr last;
int type;
unsigned int lookup_flags = 0;
+ struct delegated_inode delegated_inode = { };
retry:
error = filename_parentat(dfd, name, lookup_flags, &path, &last, &type);
if (error)
@@ -4612,7 +4618,8 @@ int do_rmdir(int dfd, struct filename *name)
error = security_path_rmdir(&path, dentry);
if (error)
goto exit4;
- error = vfs_rmdir(mnt_idmap(path.mnt), path.dentry->d_inode, dentry);
+ error = vfs_rmdir(mnt_idmap(path.mnt), path.dentry->d_inode,
+ dentry, &delegated_inode);
exit4:
dput(dentry);
exit3:
@@ -4620,6 +4627,11 @@ int do_rmdir(int dfd, struct filename *name)
mnt_drop_write(path.mnt);
exit2:
path_put(&path);
+ if (is_delegated(&delegated_inode)) {
+ error = break_deleg_wait(&delegated_inode);
+ if (!error)
+ goto retry;
+ }
if (retry_estale(error, lookup_flags)) {
lookup_flags |= LOOKUP_REVAL;
goto retry;
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index 1f56834b2072fcee1d0d400bbb554b0c949ecab4..30bae93931d9c9f8900dfcf96f79403db4f3f458 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -337,7 +337,7 @@ nfsd4_unlink_clid_dir(char *name, struct nfsd_net *nn)
status = -ENOENT;
if (d_really_is_negative(dentry))
goto out;
- status = vfs_rmdir(&nop_mnt_idmap, d_inode(dir), dentry);
+ status = vfs_rmdir(&nop_mnt_idmap, d_inode(dir), dentry, NULL);
out:
dput(dentry);
out_unlock:
@@ -427,7 +427,7 @@ purge_old(struct dentry *parent, struct dentry *child, struct nfsd_net *nn)
if (nfs4_has_reclaimed_state(name, nn))
goto out_free;
- status = vfs_rmdir(&nop_mnt_idmap, d_inode(parent), child);
+ status = vfs_rmdir(&nop_mnt_idmap, d_inode(parent), child, NULL);
if (status)
printk("failed to remove client recovery directory %pd\n",
child);
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 97aef140cbf5fca4c41738fdcaccba3b57886463..c400ea94ff2e837fd59719bf2c4b79ef1d064743 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -2108,7 +2108,7 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
break;
}
} else {
- host_err = vfs_rmdir(&nop_mnt_idmap, dirp, rdentry);
+ host_err = vfs_rmdir(&nop_mnt_idmap, dirp, rdentry, NULL);
}
fh_fill_post_attrs(fhp);
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index 0f65f9a5d54d4786b39e4f4f30f416d5b9016e70..d215d7349489686b66bb66e939b27046f7d836f6 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -206,7 +206,7 @@ static inline int ovl_do_notify_change(struct ovl_fs *ofs,
static inline int ovl_do_rmdir(struct ovl_fs *ofs,
struct inode *dir, struct dentry *dentry)
{
- int err = vfs_rmdir(ovl_upper_mnt_idmap(ofs), dir, dentry);
+ int err = vfs_rmdir(ovl_upper_mnt_idmap(ofs), dir, dentry, NULL);
pr_debug("rmdir(%pd2) = %i\n", dentry, err);
return err;
diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c
index 3d2190f26623b23ea79c63410905a3c3ad684048..c5f0f3170d586cb2dc4d416b80948c642797fb82 100644
--- a/fs/smb/server/vfs.c
+++ b/fs/smb/server/vfs.c
@@ -609,7 +609,7 @@ int ksmbd_vfs_remove_file(struct ksmbd_work *work, const struct path *path)
idmap = mnt_idmap(path->mnt);
if (S_ISDIR(d_inode(path->dentry)->i_mode)) {
- err = vfs_rmdir(idmap, d_inode(parent), path->dentry);
+ err = vfs_rmdir(idmap, d_inode(parent), path->dentry, NULL);
if (err && err != -ENOTEMPTY)
ksmbd_debug(VFS, "rmdir failed, err %d\n", err);
} else {
@@ -1090,7 +1090,7 @@ int ksmbd_vfs_unlink(struct file *filp)
dget(dentry);
if (S_ISDIR(d_inode(dentry)->i_mode))
- err = vfs_rmdir(idmap, d_inode(dir), dentry);
+ err = vfs_rmdir(idmap, d_inode(dir), dentry, NULL);
else
err = vfs_unlink(idmap, d_inode(dir), dentry, NULL);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 20bb4c8a4e8e1be7e11047d228c05920ea6c388d..12873214e1c7811735ea5d2dee3d57e2a5604d8f 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2121,7 +2121,8 @@ int vfs_symlink(struct mnt_idmap *, struct inode *,
struct dentry *, const char *);
int vfs_link(struct dentry *, struct mnt_idmap *, struct inode *,
struct dentry *, struct delegated_inode *);
-int vfs_rmdir(struct mnt_idmap *, struct inode *, struct dentry *);
+int vfs_rmdir(struct mnt_idmap *, struct inode *, struct dentry *,
+ struct delegated_inode *);
int vfs_unlink(struct mnt_idmap *, struct inode *, struct dentry *,
struct delegated_inode *);
--
2.51.1
^ permalink raw reply related
* [PATCH v6 08/17] vfs: break parent dir delegations in open(..., O_CREAT) codepath
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
Add a delegated_inode parameter to lookup_open and have it break the
delegation. Then, open_last_lookups can wait for the delegation break
and retry the call to lookup_open once it's done.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/namei.c | 22 ++++++++++++++++++----
1 file changed, 18 insertions(+), 4 deletions(-)
diff --git a/fs/namei.c b/fs/namei.c
index 9e0393a92091ac522b5324fcdad8c5592a948e8d..f439429bdfa271ccc64c937771ef4175597feb53 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -3697,7 +3697,7 @@ static struct dentry *atomic_open(struct nameidata *nd, struct dentry *dentry,
*/
static struct dentry *lookup_open(struct nameidata *nd, struct file *file,
const struct open_flags *op,
- bool got_write)
+ bool got_write, struct delegated_inode *delegated_inode)
{
struct mnt_idmap *idmap;
struct dentry *dir = nd->path.dentry;
@@ -3786,6 +3786,11 @@ static struct dentry *lookup_open(struct nameidata *nd, struct file *file,
/* Negative dentry, just create the file */
if (!dentry->d_inode && (open_flag & O_CREAT)) {
+ /* but break the directory lease first! */
+ error = try_break_deleg(dir_inode, delegated_inode);
+ if (error)
+ goto out_dput;
+
file->f_mode |= FMODE_CREATED;
audit_inode_child(dir_inode, dentry, AUDIT_TYPE_CHILD_CREATE);
if (!dir_inode->i_op->create) {
@@ -3848,6 +3853,7 @@ static struct dentry *lookup_fast_for_open(struct nameidata *nd, int open_flag)
static const char *open_last_lookups(struct nameidata *nd,
struct file *file, const struct open_flags *op)
{
+ struct delegated_inode delegated_inode = { };
struct dentry *dir = nd->path.dentry;
int open_flag = op->open_flag;
bool got_write = false;
@@ -3879,7 +3885,7 @@ static const char *open_last_lookups(struct nameidata *nd,
return ERR_PTR(-ECHILD);
}
}
-
+retry:
if (open_flag & (O_CREAT | O_TRUNC | O_WRONLY | O_RDWR)) {
got_write = !mnt_want_write(nd->path.mnt);
/*
@@ -3892,7 +3898,7 @@ static const char *open_last_lookups(struct nameidata *nd,
inode_lock(dir->d_inode);
else
inode_lock_shared(dir->d_inode);
- dentry = lookup_open(nd, file, op, got_write);
+ dentry = lookup_open(nd, file, op, got_write, &delegated_inode);
if (!IS_ERR(dentry)) {
if (file->f_mode & FMODE_CREATED)
fsnotify_create(dir->d_inode, dentry);
@@ -3907,8 +3913,16 @@ static const char *open_last_lookups(struct nameidata *nd,
if (got_write)
mnt_drop_write(nd->path.mnt);
- if (IS_ERR(dentry))
+ if (IS_ERR(dentry)) {
+ if (is_delegated(&delegated_inode)) {
+ int error = break_deleg_wait(&delegated_inode);
+
+ if (!error)
+ goto retry;
+ return ERR_PTR(error);
+ }
return ERR_CAST(dentry);
+ }
if (file->f_mode & (FMODE_OPENED | FMODE_CREATED)) {
dput(nd->path.dentry);
--
2.51.1
^ permalink raw reply related
* [PATCH v6 09/17] vfs: clean up argument list for vfs_create()
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
As Neil points out:
"I would be in favour of dropping the "dir" arg because it is always
d_inode(dentry->d_parent) which is stable."
...and...
"Also *every* caller of vfs_create() passes ".excl = true". So maybe we
don't need that arg at all."
Drop both arguments from vfs_create() and fix up the callers.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/ecryptfs/inode.c | 3 +--
fs/namei.c | 11 ++++-------
fs/nfsd/nfs3proc.c | 2 +-
fs/nfsd/vfs.c | 3 +--
fs/open.c | 4 +---
fs/overlayfs/overlayfs.h | 2 +-
fs/smb/server/vfs.c | 3 +--
include/linux/fs.h | 3 +--
8 files changed, 11 insertions(+), 20 deletions(-)
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 88631291b32535f623a3fbe4ea9b6ed48a306ca0..d109e3763a88150bfe64cd2d5564dc9802ef3386 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -188,8 +188,7 @@ ecryptfs_do_create(struct inode *directory_inode,
rc = lock_parent(ecryptfs_dentry, &lower_dentry, &lower_dir);
if (!rc)
- rc = vfs_create(&nop_mnt_idmap, lower_dir,
- lower_dentry, mode, true);
+ rc = vfs_create(&nop_mnt_idmap, lower_dentry, mode);
if (rc) {
printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
"rc = [%d]\n", __func__, rc);
diff --git a/fs/namei.c b/fs/namei.c
index f439429bdfa271ccc64c937771ef4175597feb53..9586c6aba6eae05a9fc3c103b8501d98767bef53 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -3461,10 +3461,8 @@ static inline umode_t vfs_prepare_mode(struct mnt_idmap *idmap,
/**
* vfs_create - create new file
* @idmap: idmap of the mount the inode was found from
- * @dir: inode of the parent directory
* @dentry: dentry of the child file
* @mode: mode of the child file
- * @want_excl: whether the file must not yet exist
*
* Create a new file.
*
@@ -3474,9 +3472,9 @@ static inline umode_t vfs_prepare_mode(struct mnt_idmap *idmap,
* On non-idmapped mounts or if permission checking is to be performed on the
* raw inode simply pass @nop_mnt_idmap.
*/
-int vfs_create(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry, umode_t mode, bool want_excl)
+int vfs_create(struct mnt_idmap *idmap, struct dentry *dentry, umode_t mode)
{
+ struct inode *dir = d_inode(dentry->d_parent);
int error;
error = may_create(idmap, dir, dentry);
@@ -3490,7 +3488,7 @@ int vfs_create(struct mnt_idmap *idmap, struct inode *dir,
error = security_inode_create(dir, dentry, mode);
if (error)
return error;
- error = dir->i_op->create(idmap, dir, dentry, mode, want_excl);
+ error = dir->i_op->create(idmap, dir, dentry, mode, true);
if (!error)
fsnotify_create(dir, dentry);
return error;
@@ -4383,8 +4381,7 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
idmap = mnt_idmap(path.mnt);
switch (mode & S_IFMT) {
case 0: case S_IFREG:
- error = vfs_create(idmap, path.dentry->d_inode,
- dentry, mode, true);
+ error = vfs_create(idmap, dentry, mode);
if (!error)
security_path_post_mknod(idmap, dentry);
break;
diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c
index b6d03e1ef5f7a5e8dd111b0d56c061f1e91abff7..30ea7ffa2affdb9a959b0fd15a630de056d6dc3c 100644
--- a/fs/nfsd/nfs3proc.c
+++ b/fs/nfsd/nfs3proc.c
@@ -344,7 +344,7 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
status = fh_fill_pre_attrs(fhp);
if (status != nfs_ok)
goto out;
- host_err = vfs_create(&nop_mnt_idmap, inode, child, iap->ia_mode, true);
+ host_err = vfs_create(&nop_mnt_idmap, child, iap->ia_mode);
if (host_err < 0) {
status = nfserrno(host_err);
goto out;
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index c400ea94ff2e837fd59719bf2c4b79ef1d064743..464fd54675f3b16fce9ae5f05ad22e0e6b363eb3 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1552,8 +1552,7 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
err = 0;
switch (type) {
case S_IFREG:
- host_err = vfs_create(&nop_mnt_idmap, dirp, dchild,
- iap->ia_mode, true);
+ host_err = vfs_create(&nop_mnt_idmap, dchild, iap->ia_mode);
if (!host_err)
nfsd_check_ignore_resizing(iap);
break;
diff --git a/fs/open.c b/fs/open.c
index fdaa6f08f6f4cac5c2fefd3eafa5e430e51f3979..e440f58e3ce81e137aabdf00510d839342a19219 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -1171,9 +1171,7 @@ struct file *dentry_create(const struct path *path, int flags, umode_t mode,
if (IS_ERR(f))
return f;
- error = vfs_create(mnt_idmap(path->mnt),
- d_inode(path->dentry->d_parent),
- path->dentry, mode, true);
+ error = vfs_create(mnt_idmap(path->mnt), path->dentry, mode);
if (!error)
error = vfs_open(path, f);
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index d215d7349489686b66bb66e939b27046f7d836f6..2bdc434941ebc70f6d4f57cca4f68125112a7bc4 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -235,7 +235,7 @@ static inline int ovl_do_create(struct ovl_fs *ofs,
struct inode *dir, struct dentry *dentry,
umode_t mode)
{
- int err = vfs_create(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, true);
+ int err = vfs_create(ovl_upper_mnt_idmap(ofs), dentry, mode);
pr_debug("create(%pd2, 0%o) = %i\n", dentry, mode, err);
return err;
diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c
index c5f0f3170d586cb2dc4d416b80948c642797fb82..83ece2de4b23bf9209137e7ca414a72439b5cc2e 100644
--- a/fs/smb/server/vfs.c
+++ b/fs/smb/server/vfs.c
@@ -188,8 +188,7 @@ int ksmbd_vfs_create(struct ksmbd_work *work, const char *name, umode_t mode)
}
mode |= S_IFREG;
- err = vfs_create(mnt_idmap(path.mnt), d_inode(path.dentry),
- dentry, mode, true);
+ err = vfs_create(mnt_idmap(path.mnt), dentry, mode);
if (!err) {
ksmbd_vfs_inherit_owner(work, d_inode(path.dentry),
d_inode(dentry));
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 12873214e1c7811735ea5d2dee3d57e2a5604d8f..21876ef1fec90181b9878372c7c7e710773aae9f 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2111,8 +2111,7 @@ bool inode_owner_or_capable(struct mnt_idmap *idmap,
/*
* VFS helper functions..
*/
-int vfs_create(struct mnt_idmap *, struct inode *,
- struct dentry *, umode_t, bool);
+int vfs_create(struct mnt_idmap *, struct dentry *, umode_t);
struct dentry *vfs_mkdir(struct mnt_idmap *, struct inode *,
struct dentry *, umode_t, struct delegated_inode *);
int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
--
2.51.1
^ permalink raw reply related
* [PATCH v6 10/17] vfs: make vfs_create break delegations on parent directory
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
Add a delegated_inode parameter to vfs_create. Most callers are
converted to pass in NULL, but do_mknodat() is changed to wait for a
delegation break if there is one.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/ecryptfs/inode.c | 2 +-
fs/namei.c | 15 +++++++++++++--
fs/nfsd/nfs3proc.c | 2 +-
fs/nfsd/vfs.c | 2 +-
fs/open.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
fs/smb/server/vfs.c | 2 +-
include/linux/fs.h | 3 ++-
8 files changed, 21 insertions(+), 9 deletions(-)
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index d109e3763a88150bfe64cd2d5564dc9802ef3386..3341f00dd08753c8feab184dd82b8bfa63d3724a 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -188,7 +188,7 @@ ecryptfs_do_create(struct inode *directory_inode,
rc = lock_parent(ecryptfs_dentry, &lower_dentry, &lower_dir);
if (!rc)
- rc = vfs_create(&nop_mnt_idmap, lower_dentry, mode);
+ rc = vfs_create(&nop_mnt_idmap, lower_dentry, mode, NULL);
if (rc) {
printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
"rc = [%d]\n", __func__, rc);
diff --git a/fs/namei.c b/fs/namei.c
index 9586c6aba6eae05a9fc3c103b8501d98767bef53..b20f053374a578d36fb764e0df80fb7db9230dbe 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -3463,6 +3463,7 @@ static inline umode_t vfs_prepare_mode(struct mnt_idmap *idmap,
* @idmap: idmap of the mount the inode was found from
* @dentry: dentry of the child file
* @mode: mode of the child file
+ * @di: returns parent inode, if the inode is delegated.
*
* Create a new file.
*
@@ -3472,7 +3473,8 @@ static inline umode_t vfs_prepare_mode(struct mnt_idmap *idmap,
* On non-idmapped mounts or if permission checking is to be performed on the
* raw inode simply pass @nop_mnt_idmap.
*/
-int vfs_create(struct mnt_idmap *idmap, struct dentry *dentry, umode_t mode)
+int vfs_create(struct mnt_idmap *idmap, struct dentry *dentry, umode_t mode,
+ struct delegated_inode *di)
{
struct inode *dir = d_inode(dentry->d_parent);
int error;
@@ -3486,6 +3488,9 @@ int vfs_create(struct mnt_idmap *idmap, struct dentry *dentry, umode_t mode)
mode = vfs_prepare_mode(idmap, dir, mode, S_IALLUGO, S_IFREG);
error = security_inode_create(dir, dentry, mode);
+ if (error)
+ return error;
+ error = try_break_deleg(dir, di);
if (error)
return error;
error = dir->i_op->create(idmap, dir, dentry, mode, true);
@@ -4358,6 +4363,7 @@ static int may_mknod(umode_t mode)
static int do_mknodat(int dfd, struct filename *name, umode_t mode,
unsigned int dev)
{
+ struct delegated_inode di = { };
struct mnt_idmap *idmap;
struct dentry *dentry;
struct path path;
@@ -4381,7 +4387,7 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
idmap = mnt_idmap(path.mnt);
switch (mode & S_IFMT) {
case 0: case S_IFREG:
- error = vfs_create(idmap, dentry, mode);
+ error = vfs_create(idmap, dentry, mode, &di);
if (!error)
security_path_post_mknod(idmap, dentry);
break;
@@ -4396,6 +4402,11 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
}
out2:
end_creating_path(&path, dentry);
+ if (is_delegated(&di)) {
+ error = break_deleg_wait(&di);
+ if (!error)
+ goto retry;
+ }
if (retry_estale(error, lookup_flags)) {
lookup_flags |= LOOKUP_REVAL;
goto retry;
diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c
index 30ea7ffa2affdb9a959b0fd15a630de056d6dc3c..2cb972b5ed994d69995146aeeac8651c1c04fa46 100644
--- a/fs/nfsd/nfs3proc.c
+++ b/fs/nfsd/nfs3proc.c
@@ -344,7 +344,7 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
status = fh_fill_pre_attrs(fhp);
if (status != nfs_ok)
goto out;
- host_err = vfs_create(&nop_mnt_idmap, child, iap->ia_mode);
+ host_err = vfs_create(&nop_mnt_idmap, child, iap->ia_mode, NULL);
if (host_err < 0) {
status = nfserrno(host_err);
goto out;
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 464fd54675f3b16fce9ae5f05ad22e0e6b363eb3..de5f46f8c6d3ab24fabddeed9bd59adbe7a486df 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1552,7 +1552,7 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
err = 0;
switch (type) {
case S_IFREG:
- host_err = vfs_create(&nop_mnt_idmap, dchild, iap->ia_mode);
+ host_err = vfs_create(&nop_mnt_idmap, dchild, iap->ia_mode, NULL);
if (!host_err)
nfsd_check_ignore_resizing(iap);
break;
diff --git a/fs/open.c b/fs/open.c
index e440f58e3ce81e137aabdf00510d839342a19219..92cf2e11781b0efd16fe0b751286ce943ea6b4e2 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -1171,7 +1171,7 @@ struct file *dentry_create(const struct path *path, int flags, umode_t mode,
if (IS_ERR(f))
return f;
- error = vfs_create(mnt_idmap(path->mnt), path->dentry, mode);
+ error = vfs_create(mnt_idmap(path->mnt), path->dentry, mode, NULL);
if (!error)
error = vfs_open(path, f);
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index 2bdc434941ebc70f6d4f57cca4f68125112a7bc4..e30441cc9c63ff8b6c055db80be7974501d0023b 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -235,7 +235,7 @@ static inline int ovl_do_create(struct ovl_fs *ofs,
struct inode *dir, struct dentry *dentry,
umode_t mode)
{
- int err = vfs_create(ovl_upper_mnt_idmap(ofs), dentry, mode);
+ int err = vfs_create(ovl_upper_mnt_idmap(ofs), dentry, mode, NULL);
pr_debug("create(%pd2, 0%o) = %i\n", dentry, mode, err);
return err;
diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c
index 83ece2de4b23bf9209137e7ca414a72439b5cc2e..3747851b61c8bad4c00e49866c380c32e9f53a4b 100644
--- a/fs/smb/server/vfs.c
+++ b/fs/smb/server/vfs.c
@@ -188,7 +188,7 @@ int ksmbd_vfs_create(struct ksmbd_work *work, const char *name, umode_t mode)
}
mode |= S_IFREG;
- err = vfs_create(mnt_idmap(path.mnt), dentry, mode);
+ err = vfs_create(mnt_idmap(path.mnt), dentry, mode, NULL);
if (!err) {
ksmbd_vfs_inherit_owner(work, d_inode(path.dentry),
d_inode(dentry));
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 21876ef1fec90181b9878372c7c7e710773aae9f..83b05aec4e10c846d3168018fb62284e45ceb1a8 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2111,7 +2111,8 @@ bool inode_owner_or_capable(struct mnt_idmap *idmap,
/*
* VFS helper functions..
*/
-int vfs_create(struct mnt_idmap *, struct dentry *, umode_t);
+int vfs_create(struct mnt_idmap *, struct dentry *, umode_t,
+ struct delegated_inode *);
struct dentry *vfs_mkdir(struct mnt_idmap *, struct inode *,
struct dentry *, umode_t, struct delegated_inode *);
int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
--
2.51.1
^ permalink raw reply related
* [PATCH v6 11/17] vfs: make vfs_mknod break delegations on parent directory
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
Add a new delegated_inode pointer to vfs_mknod() and have the
appropriate callers wait when there is an outstanding delegation. All
other callers just set the pointer to NULL.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
drivers/base/devtmpfs.c | 2 +-
fs/ecryptfs/inode.c | 2 +-
fs/init.c | 2 +-
fs/namei.c | 23 +++++++++++++++--------
fs/nfsd/vfs.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
include/linux/fs.h | 4 ++--
net/unix/af_unix.c | 2 +-
8 files changed, 23 insertions(+), 16 deletions(-)
diff --git a/drivers/base/devtmpfs.c b/drivers/base/devtmpfs.c
index 104025104ef75381984fd94dfbd50feeaa8cdd22..2f576ecf18324f767cd5ac6cbd28adbf9f46b958 100644
--- a/drivers/base/devtmpfs.c
+++ b/drivers/base/devtmpfs.c
@@ -231,7 +231,7 @@ static int handle_create(const char *nodename, umode_t mode, kuid_t uid,
return PTR_ERR(dentry);
err = vfs_mknod(&nop_mnt_idmap, d_inode(path.dentry), dentry, mode,
- dev->devt);
+ dev->devt, NULL);
if (!err) {
struct iattr newattrs;
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 3341f00dd08753c8feab184dd82b8bfa63d3724a..83f06452476dec4025cc8f50e082ae6ccbbc7914 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -564,7 +564,7 @@ ecryptfs_mknod(struct mnt_idmap *idmap, struct inode *dir,
rc = lock_parent(dentry, &lower_dentry, &lower_dir);
if (!rc)
rc = vfs_mknod(&nop_mnt_idmap, lower_dir,
- lower_dentry, mode, dev);
+ lower_dentry, mode, dev, NULL);
if (rc || d_really_is_negative(lower_dentry))
goto out;
rc = ecryptfs_interpose(lower_dentry, dentry, dir->i_sb);
diff --git a/fs/init.c b/fs/init.c
index 895f8a09a71acfd03e11164e3b441a7d4e2de146..4f02260dd65b0dfcbfbf5812d2ec6a33444a3b1f 100644
--- a/fs/init.c
+++ b/fs/init.c
@@ -157,7 +157,7 @@ int __init init_mknod(const char *filename, umode_t mode, unsigned int dev)
error = security_path_mknod(&path, dentry, mode, dev);
if (!error)
error = vfs_mknod(mnt_idmap(path.mnt), path.dentry->d_inode,
- dentry, mode, new_decode_dev(dev));
+ dentry, mode, new_decode_dev(dev), NULL);
end_creating_path(&path, dentry);
return error;
}
diff --git a/fs/namei.c b/fs/namei.c
index b20f053374a578d36fb764e0df80fb7db9230dbe..e9616134390fb7f0d09a759be69bf677f8800bc5 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -4295,13 +4295,15 @@ inline struct dentry *start_creating_user_path(
}
EXPORT_SYMBOL(start_creating_user_path);
+
/**
* vfs_mknod - create device node or file
- * @idmap: idmap of the mount the inode was found from
- * @dir: inode of the parent directory
- * @dentry: dentry of the child device node
- * @mode: mode of the child device node
- * @dev: device number of device to create
+ * @idmap: idmap of the mount the inode was found from
+ * @dir: inode of the parent directory
+ * @dentry: dentry of the child device node
+ * @mode: mode of the child device node
+ * @dev: device number of device to create
+ * @delegated_inode: returns parent inode, if the inode is delegated.
*
* Create a device node or file.
*
@@ -4312,7 +4314,8 @@ EXPORT_SYMBOL(start_creating_user_path);
* raw inode simply pass @nop_mnt_idmap.
*/
int vfs_mknod(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry, umode_t mode, dev_t dev)
+ struct dentry *dentry, umode_t mode, dev_t dev,
+ struct delegated_inode *delegated_inode)
{
bool is_whiteout = S_ISCHR(mode) && dev == WHITEOUT_DEV;
int error = may_create(idmap, dir, dentry);
@@ -4336,6 +4339,10 @@ int vfs_mknod(struct mnt_idmap *idmap, struct inode *dir,
if (error)
return error;
+ error = try_break_deleg(dir, delegated_inode);
+ if (error)
+ return error;
+
error = dir->i_op->mknod(idmap, dir, dentry, mode, dev);
if (!error)
fsnotify_create(dir, dentry);
@@ -4393,11 +4400,11 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
break;
case S_IFCHR: case S_IFBLK:
error = vfs_mknod(idmap, path.dentry->d_inode,
- dentry, mode, new_decode_dev(dev));
+ dentry, mode, new_decode_dev(dev), &di);
break;
case S_IFIFO: case S_IFSOCK:
error = vfs_mknod(idmap, path.dentry->d_inode,
- dentry, mode, 0);
+ dentry, mode, 0, &di);
break;
}
out2:
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index de5f46f8c6d3ab24fabddeed9bd59adbe7a486df..6684935007b17ed40723ff0a744045fd187ace5e 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1573,7 +1573,7 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
case S_IFIFO:
case S_IFSOCK:
host_err = vfs_mknod(&nop_mnt_idmap, dirp, dchild,
- iap->ia_mode, rdev);
+ iap->ia_mode, rdev, NULL);
break;
default:
printk(KERN_WARNING "nfsd: bad file type %o in nfsd_create\n",
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index e30441cc9c63ff8b6c055db80be7974501d0023b..afd95721f76ea761cfe7133c3902028550f3e359 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -257,7 +257,7 @@ static inline int ovl_do_mknod(struct ovl_fs *ofs,
struct inode *dir, struct dentry *dentry,
umode_t mode, dev_t dev)
{
- int err = vfs_mknod(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, dev);
+ int err = vfs_mknod(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, dev, NULL);
pr_debug("mknod(%pd2, 0%o, 0%o) = %i\n", dentry, mode, dev, err);
return err;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 83b05aec4e10c846d3168018fb62284e45ceb1a8..1a5d86cfafaa97fc89d15cd1a156968b8c3dd377 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2116,7 +2116,7 @@ int vfs_create(struct mnt_idmap *, struct dentry *, umode_t,
struct dentry *vfs_mkdir(struct mnt_idmap *, struct inode *,
struct dentry *, umode_t, struct delegated_inode *);
int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
- umode_t, dev_t);
+ umode_t, dev_t, struct delegated_inode *);
int vfs_symlink(struct mnt_idmap *, struct inode *,
struct dentry *, const char *);
int vfs_link(struct dentry *, struct mnt_idmap *, struct inode *,
@@ -2152,7 +2152,7 @@ static inline int vfs_whiteout(struct mnt_idmap *idmap,
struct inode *dir, struct dentry *dentry)
{
return vfs_mknod(idmap, dir, dentry, S_IFCHR | WHITEOUT_MODE,
- WHITEOUT_DEV);
+ WHITEOUT_DEV, NULL);
}
struct file *kernel_tmpfile_open(struct mnt_idmap *idmap,
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 768098dec2310008632558ae928703b37c3cc8ef..db1fd8d6a84c2c7c0d45b43d9c5a936b3d491b7b 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1399,7 +1399,7 @@ static int unix_bind_bsd(struct sock *sk, struct sockaddr_un *sunaddr,
idmap = mnt_idmap(parent.mnt);
err = security_path_mknod(&parent, dentry, mode, 0);
if (!err)
- err = vfs_mknod(idmap, d_inode(parent.dentry), dentry, mode, 0);
+ err = vfs_mknod(idmap, d_inode(parent.dentry), dentry, mode, 0, NULL);
if (err)
goto out_path;
err = mutex_lock_interruptible(&u->bindlock);
--
2.51.1
^ permalink raw reply related
* [PATCH v6 12/17] vfs: make vfs_symlink break delegations on parent dir
From: Jeff Layton @ 2025-11-11 14:12 UTC (permalink / raw)
To: Miklos Szeredi, Alexander Viro, Christian Brauner, Jan Kara,
Chuck Lever, Alexander Aring, Trond Myklebust, Anna Schumaker,
Steve French, Paulo Alcantara, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, Greg Kroah-Hartman, Rafael J. Wysocki,
Danilo Krummrich, David Howells, Tyler Hicks, NeilBrown,
Olga Kornievskaia, Dai Ngo, Amir Goldstein, Namjae Jeon,
Steve French, Sergey Senozhatsky, Carlos Maiolino,
Kuniyuki Iwashima, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Simon Horman
Cc: linux-fsdevel, linux-kernel, linux-nfs, linux-cifs,
samba-technical, netfs, ecryptfs, linux-unionfs, linux-xfs,
netdev, linux-api, Jeff Layton
In-Reply-To: <20251111-dir-deleg-ro-v6-0-52f3feebb2f2@kernel.org>
In order to add directory delegation support, we must break delegations
on the parent on any change to the directory.
Add a delegated_inode parameter to vfs_symlink() and have it break the
delegation. do_symlinkat() can then wait on the delegation break before
proceeding.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
fs/ecryptfs/inode.c | 2 +-
fs/init.c | 2 +-
fs/namei.c | 16 ++++++++++++++--
fs/nfsd/vfs.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
include/linux/fs.h | 2 +-
6 files changed, 19 insertions(+), 7 deletions(-)
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 83f06452476dec4025cc8f50e082ae6ccbbc7914..ba15e7359dfa6e150b577205991010873a633511 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -479,7 +479,7 @@ static int ecryptfs_symlink(struct mnt_idmap *idmap,
if (rc)
goto out_lock;
rc = vfs_symlink(&nop_mnt_idmap, lower_dir, lower_dentry,
- encoded_symname);
+ encoded_symname, NULL);
kfree(encoded_symname);
if (rc || d_really_is_negative(lower_dentry))
goto out_lock;
diff --git a/fs/init.c b/fs/init.c
index 4f02260dd65b0dfcbfbf5812d2ec6a33444a3b1f..e0f5429c0a49d046bd3f231a260954ed0f90ef44 100644
--- a/fs/init.c
+++ b/fs/init.c
@@ -209,7 +209,7 @@ int __init init_symlink(const char *oldname, const char *newname)
error = security_path_symlink(&path, dentry, oldname);
if (!error)
error = vfs_symlink(mnt_idmap(path.mnt), path.dentry->d_inode,
- dentry, oldname);
+ dentry, oldname, NULL);
end_creating_path(&path, dentry);
return error;
}
diff --git a/fs/namei.c b/fs/namei.c
index e9616134390fb7f0d09a759be69bf677f8800bc5..d5ab28947b2b6c6e19c7bb4a9140ccec407dc07c 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -4845,6 +4845,7 @@ SYSCALL_DEFINE1(unlink, const char __user *, pathname)
* @dir: inode of the parent directory
* @dentry: dentry of the child symlink file
* @oldname: name of the file to link to
+ * @delegated_inode: returns victim inode, if the inode is delegated.
*
* Create a symlink.
*
@@ -4855,7 +4856,8 @@ SYSCALL_DEFINE1(unlink, const char __user *, pathname)
* raw inode simply pass @nop_mnt_idmap.
*/
int vfs_symlink(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry, const char *oldname)
+ struct dentry *dentry, const char *oldname,
+ struct delegated_inode *delegated_inode)
{
int error;
@@ -4870,6 +4872,10 @@ int vfs_symlink(struct mnt_idmap *idmap, struct inode *dir,
if (error)
return error;
+ error = try_break_deleg(dir, delegated_inode);
+ if (error)
+ return error;
+
error = dir->i_op->symlink(idmap, dir, dentry, oldname);
if (!error)
fsnotify_create(dir, dentry);
@@ -4883,6 +4889,7 @@ int do_symlinkat(struct filename *from, int newdfd, struct filename *to)
struct dentry *dentry;
struct path path;
unsigned int lookup_flags = 0;
+ struct delegated_inode delegated_inode = { };
if (IS_ERR(from)) {
error = PTR_ERR(from);
@@ -4897,8 +4904,13 @@ int do_symlinkat(struct filename *from, int newdfd, struct filename *to)
error = security_path_symlink(&path, dentry, from->name);
if (!error)
error = vfs_symlink(mnt_idmap(path.mnt), path.dentry->d_inode,
- dentry, from->name);
+ dentry, from->name, &delegated_inode);
end_creating_path(&path, dentry);
+ if (is_delegated(&delegated_inode)) {
+ error = break_deleg_wait(&delegated_inode);
+ if (!error)
+ goto retry;
+ }
if (retry_estale(error, lookup_flags)) {
lookup_flags |= LOOKUP_REVAL;
goto retry;
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 6684935007b17ed40723ff0a744045fd187ace5e..28710da4cce7cc7fc1e14d29420239dc357316f6 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1742,7 +1742,7 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
err = fh_fill_pre_attrs(fhp);
if (err != nfs_ok)
goto out_unlock;
- host_err = vfs_symlink(&nop_mnt_idmap, d_inode(dentry), dnew, path);
+ host_err = vfs_symlink(&nop_mnt_idmap, d_inode(dentry), dnew, path, NULL);
err = nfserrno(host_err);
cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp);
if (!err)
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index afd95721f76ea761cfe7133c3902028550f3e359..5065961bd370628c9afe914ea570d9998a096660 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -267,7 +267,7 @@ static inline int ovl_do_symlink(struct ovl_fs *ofs,
struct inode *dir, struct dentry *dentry,
const char *oldname)
{
- int err = vfs_symlink(ovl_upper_mnt_idmap(ofs), dir, dentry, oldname);
+ int err = vfs_symlink(ovl_upper_mnt_idmap(ofs), dir, dentry, oldname, NULL);
pr_debug("symlink(\"%s\", %pd2) = %i\n", oldname, dentry, err);
return err;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 1a5d86cfafaa97fc89d15cd1a156968b8c3dd377..64323e618724bc20dc101db13035b042f5f88e4d 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2118,7 +2118,7 @@ struct dentry *vfs_mkdir(struct mnt_idmap *, struct inode *,
int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
umode_t, dev_t, struct delegated_inode *);
int vfs_symlink(struct mnt_idmap *, struct inode *,
- struct dentry *, const char *);
+ struct dentry *, const char *, struct delegated_inode *);
int vfs_link(struct dentry *, struct mnt_idmap *, struct inode *,
struct dentry *, struct delegated_inode *);
int vfs_rmdir(struct mnt_idmap *, struct inode *, struct dentry *,
--
2.51.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox