* [PATCH] rust-next: make from_errno use try_from_errno
@ 2024-11-05 11:48 Guilherme Augusto Martins da Silva
2024-11-07 12:26 ` Rasmus Villemoes
2024-11-14 19:42 ` Daniel Sedlak
0 siblings, 2 replies; 5+ messages in thread
From: Guilherme Augusto Martins da Silva @ 2024-11-05 11:48 UTC (permalink / raw)
To: rust-for-linux
Cc: Daniel Sedlak, Guilherme Augusto Martins da Silva, Miguel Ojeda
Modified the from_errno function to use try_from_errno to reduce code duplication while still maintaning all existing behavior and error handling and also reduces unsafe code.
Co-developed-by: Daniel Sedlak <daniel@sedlak.dev>
Signed-off-by: Daniel Sedlak <daniel@sedlak.dev>
Signed-off-by: Guilherme Augusto Martins da Silva <guilhermev2huehue@gmail.com>
Suggested-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://github.com/Rust-for-Linux/linux/issues/1125
---
diff --git a/rust/kernel/error.rs b/rust/kernel/error.rs
index 7cd3bbab52f2..077f95cbb6ae 100644
--- a/rust/kernel/error.rs
+++ b/rust/kernel/error.rs
@@ -101,19 +101,17 @@ impl Error {
/// It is a bug to pass an out-of-range `errno`. `EINVAL` would
/// be returned in such a case.
pub fn from_errno(errno: core::ffi::c_int) -> Error {
- if errno < -(bindings::MAX_ERRNO as i32) || errno >= 0 {
- // TODO: Make it a `WARN_ONCE` once available.
- crate::pr_warn!(
- "attempted to create `Error` with out of range `errno`: {}",
- errno
- );
- return code::EINVAL;
+ match Error::try_from_errno(errno) {
+ Some(error) => error,
+ None => {
+ // TODO: Make it a `WARN_ONCE` once available.
+ crate::pr_warn!(
+ "attempted to create `Error` with out of range `errno`: {}",
+ errno
+ );
+ code::EINVAL
+ }
}
-
- // INVARIANT: The check above ensures the type invariant
- // will hold.
- // SAFETY: `errno` is checked above to be in a valid range.
- unsafe { Error::from_errno_unchecked(errno) }
}
^ permalink raw reply related [flat|nested] 5+ messages in thread* Re: [PATCH] rust-next: make from_errno use try_from_errno
2024-11-05 11:48 [PATCH] rust-next: make from_errno use try_from_errno Guilherme Augusto Martins da Silva
@ 2024-11-07 12:26 ` Rasmus Villemoes
2024-11-14 19:42 ` Daniel Sedlak
1 sibling, 0 replies; 5+ messages in thread
From: Rasmus Villemoes @ 2024-11-07 12:26 UTC (permalink / raw)
To: Guilherme Augusto Martins da Silva
Cc: rust-for-linux, Daniel Sedlak, Miguel Ojeda
On Tue, Nov 05 2024, Guilherme Augusto Martins da Silva <guilhermev2huehue@gmail.com> wrote:
> Modified the from_errno function to use try_from_errno to reduce code duplication while still maintaning all existing behavior and error handling and also reduces unsafe code.
>
> Co-developed-by: Daniel Sedlak <daniel@sedlak.dev>
> Signed-off-by: Daniel Sedlak <daniel@sedlak.dev>
> Signed-off-by: Guilherme Augusto Martins da Silva <guilhermev2huehue@gmail.com>
> Suggested-by: Miguel Ojeda <ojeda@kernel.org>
> Link: https://github.com/Rust-for-Linux/linux/issues/1125
> ---
> diff --git a/rust/kernel/error.rs b/rust/kernel/error.rs
> index 7cd3bbab52f2..077f95cbb6ae 100644
> --- a/rust/kernel/error.rs
> +++ b/rust/kernel/error.rs
> @@ -101,19 +101,17 @@ impl Error {
> /// It is a bug to pass an out-of-range `errno`. `EINVAL` would
> /// be returned in such a case.
> pub fn from_errno(errno: core::ffi::c_int) -> Error {
> - if errno < -(bindings::MAX_ERRNO as i32) || errno >= 0 {
> - // TODO: Make it a `WARN_ONCE` once available.
> - crate::pr_warn!(
> - "attempted to create `Error` with out of range `errno`: {}",
> - errno
> - );
> - return code::EINVAL;
> + match Error::try_from_errno(errno) {
> + Some(error) => error,
> + None => {
> + // TODO: Make it a `WARN_ONCE` once available.
> + crate::pr_warn!(
> + "attempted to create `Error` with out of range `errno`: {}",
> + errno
> + );
> + code::EINVAL
> + }
Not really related to the change done here, but I've often wondered why
the kernel doesn't have a dedicated EBUG that can be used in places
where a genuine kernel bug is detected at run-time, but BUG() is too
much, and we can survive by just passing some error back up the stack
(possibly with a WARN or pr_warn()). In fact, an EBUG could be useful in
userspace code as well for library code that detects some
should-not-happen inconsistency.
EINVAL is already way overused, and substituting that when a programming
error rather than bad user input is detected seems to just overload it
even more.
Rasmus
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: [PATCH] rust-next: make from_errno use try_from_errno
2024-11-05 11:48 [PATCH] rust-next: make from_errno use try_from_errno Guilherme Augusto Martins da Silva
2024-11-07 12:26 ` Rasmus Villemoes
@ 2024-11-14 19:42 ` Daniel Sedlak
1 sibling, 0 replies; 5+ messages in thread
From: Daniel Sedlak @ 2024-11-14 19:42 UTC (permalink / raw)
To: Guilherme Augusto Martins da Silva, rust-for-linux; +Cc: Miguel Ojeda
Hello Guilherme,
I believe you need to send a new patch as a `v2` instead of pushing
changes to this patch.
Also in your commit message, I believe you should only write `rust:`
instead of `rust-next:`
see other patches in the mailing list.
On 11/5/24 12:48 PM, Guilherme Augusto Martins da Silva wrote:
Modified the from_errno function to use try_from_errno to reduce code
duplication while still maintaning all existing behavior and error
handling and also reduces unsafe code.
Co-developed-by: Daniel Sedlak <daniel@sedlak.dev>
Signed-off-by: Daniel Sedlak <daniel@sedlak.dev>
Signed-off-by: Guilherme Augusto Martins da Silva <guilhermev2huehue@gmail.com>
Suggested-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://github.com/Rust-for-Linux/linux/issues/1125
---
diff --git a/rust/kernel/error.rs b/rust/kernel/error.rs
index 7cd3bbab52f2..077f95cbb6ae 100644
--- a/rust/kernel/error.rs
+++ b/rust/kernel/error.rs
@@ -101,19 +101,17 @@ impl Error {
/// It is a bug to pass an out-of-range `errno`. `EINVAL` would
/// be returned in such a case.
pub fn from_errno(errno: core::ffi::c_int) -> Error {
- if errno < -(bindings::MAX_ERRNO as i32) || errno >= 0 {
- // TODO: Make it a `WARN_ONCE` once available.
- crate::pr_warn!(
- "attempted to create `Error` with out of range `errno`: {}",
- errno
- );
- return code::EINVAL;
+ match Error::try_from_errno(errno) {
+ Some(error) => error,
+ None => {
+ // TODO: Make it a `WARN_ONCE` once available.
+ crate::pr_warn!(
+ "attempted to create `Error` with out of range
`errno`: {}",
+ errno
+ );
+ code::EINVAL
+ }
}
This is highly opinionated so do what satisfy you, but I believe the
`if let … else …` pattern is better,
because of the indentation and less lines changed, see my previous
patch: https://lore.kernel.org/rust-for-linux/20241104185135.18974-1-daniel@sedlak.dev/.
-
- // INVARIANT: The check above ensures the type invariant
- // will hold.
- // SAFETY: `errno` is checked above to be in a valid range.
- unsafe { Error::from_errno_unchecked(errno) }
}
^ permalink raw reply related [flat|nested] 5+ messages in thread
* [PATCH] rust-next: make from_errno use try_from_errno
@ 2024-11-04 20:00 Guilherme Augusto Martins da Silva
2024-11-04 20:14 ` Miguel Ojeda
0 siblings, 1 reply; 5+ messages in thread
From: Guilherme Augusto Martins da Silva @ 2024-11-04 20:00 UTC (permalink / raw)
To: rust-for-linux; +Cc: Guilherme Augusto Martins da Silva
Make from_errno use try_from_errno (available in rust-next).
This also allows to remove unsafe code and a comment or two too.
this project or the open source license(s) involved.
Signed-off-by: Guilherme Augusto Martins da Silva <guilhermev2huehue@gmail.com>
---
diff --git a/rust/kernel/error.rs b/rust/kernel/error.rs
index 7cd3bbab52f2..077f95cbb6ae 100644
--- a/rust/kernel/error.rs
+++ b/rust/kernel/error.rs
@@ -101,19 +101,17 @@ impl Error {
/// It is a bug to pass an out-of-range `errno`. `EINVAL` would
/// be returned in such a case.
pub fn from_errno(errno: core::ffi::c_int) -> Error {
- if errno < -(bindings::MAX_ERRNO as i32) || errno >= 0 {
- // TODO: Make it a `WARN_ONCE` once available.
- crate::pr_warn!(
- "attempted to create `Error` with out of range `errno`: {}",
- errno
- );
- return code::EINVAL;
+ match Error::try_from_errno(errno) {
+ Some(error) => error,
+ None => {
+ // TODO: Make it a `WARN_ONCE` once available.
+ crate::pr_warn!(
+ "attempted to create `Error` with out of range `errno`: {}",
+ errno
+ );
+ code::EINVAL
+ }
}
-
- // INVARIANT: The check above ensures the type invariant
- // will hold.
- // SAFETY: `errno` is checked above to be in a valid range.
- unsafe { Error::from_errno_unchecked(errno) }
}
^ permalink raw reply related [flat|nested] 5+ messages in thread* Re: [PATCH] rust-next: make from_errno use try_from_errno
2024-11-04 20:00 Guilherme Augusto Martins da Silva
@ 2024-11-04 20:14 ` Miguel Ojeda
0 siblings, 0 replies; 5+ messages in thread
From: Miguel Ojeda @ 2024-11-04 20:14 UTC (permalink / raw)
To: Guilherme Augusto Martins da Silva, Daniel Sedlak; +Cc: rust-for-linux
Hi Guilherme, Daniel,
On Mon, Nov 4, 2024 at 9:01 PM Guilherme Augusto Martins da Silva
<guilhermev2huehue@gmail.com> wrote:
>
> Make from_errno use try_from_errno (available in rust-next).
Normally you would note the "available in rust-next" part in the
non-commit part (i.e. below the `---` line).
> this project or the open source license(s) involved.
Spurious line from the previous Developer's Certificate of Origin you
had in the commit message (in private).
> Signed-off-by: Guilherme Augusto Martins da Silva <guilhermev2huehue@gmail.com>
Please add the Suggested-by and the Link tags as mentioned in the issue.
So I got these two at essentially the same time, but Guilherme's in
private, as we discussed.
I think you should both be co-authors, although you will need to
decide who will be the main author and send a v2 merging both patches
into the best version possible :)
Please see my notes in the other one too.
Cheers,
Miguel
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2024-11-14 19:42 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-11-05 11:48 [PATCH] rust-next: make from_errno use try_from_errno Guilherme Augusto Martins da Silva
2024-11-07 12:26 ` Rasmus Villemoes
2024-11-14 19:42 ` Daniel Sedlak
-- strict thread matches above, loose matches on Subject: below --
2024-11-04 20:00 Guilherme Augusto Martins da Silva
2024-11-04 20:14 ` Miguel Ojeda
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).