xiaoxiang781216 closed pull request #13245: libs: Add static compatiblity check
for Rust
URL: https://github.com/apache/nuttx/pull/13245
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific
no1wudi commented on code in PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#discussion_r1779497091
##
libs/libc/librust/lib_rust.c:
##
@@ -0,0 +1,944 @@
+/
+ * libs/libc/librust/lib_rust.c
+
xiaoxiang781216 commented on code in PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#discussion_r1779466776
##
libs/libc/librust/lib_rust.c:
##
@@ -0,0 +1,944 @@
+/
+ * libs/libc/librust/lib_r
xiaoxiang781216 commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2380609957
should we move the code to apps(e.g. rustcheck)? since all of them is
checking the size and offset of posix type.
--
This is an automated message from the Apache Git Service.
To
github-actions[bot] commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2378300867
This PR exceeds the recommended size of 1000 lines. Please make sure you are
NOT addressing multiple issues with one PR. Note this PR might be rejected due
to its size.
--
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2372756560
Make it read to review since https://github.com/rust-lang/libc/pull/3909
merged
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2372754986
@yamt @lupyuen The current approach is not to enforce a mandatory
requirement that we cannot modify these structures or the values of macro
definitions, but rather to strengthen the revie
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370830390
> Hi @no1wudi could you share on the Mailing List the current approach for
Rust Standard Library? And explain the tradeoff between Better Optimizations
(smaller structures, lto, etc) vs A
lupyuen commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370764192
Hi @no1wudi could you share on the Mailing List the current approach for
Rust Standard Library? And explain the tradeoff between Better Optimizations
(smaller structures, lto, etc) vs ABI
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370741806
> as it's a big policy change with pros and cons, i suspect we should
discuss this topic with a wider audience.
Agree, so I open an issue to collect the idea from NuttX community:
github-actions[bot] commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370729781
This PR exceeds the recommended size of 1000 lines. Please make sure you are
NOT addressing multiple issues with one PR. Note this PR might be rejected due
to its size.
--
yamt commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370697685
> > does this mean to require everyone here, who might not care rust at all,
to maintain certain abi compatibility for rust? i'm not sure if it's a good
idea. if rust doesn't work well w/o o
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370636565
> does this mean to require everyone here, who might not care rust at all,
to maintain certain abi compatibility for rust? i'm not sure if it's a good
idea. if rust doesn't work well w/o
yamt commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370585066
does this mean to require everyone here, who might not care rust at all, to
maintain certain abi compatibility for rust?
i'm not sure if it's a good idea.
if rust doesn't work well w/o
github-actions[bot] commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2370448197
This PR exceeds the recommended size of 1000 lines. Please make sure you are
NOT addressing multiple issues with one PR. Note this PR might be rejected due
to its size.
--
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2340415299
Let's wait feedback from https://github.com/rust-lang/libc/pull/3909 before
merge this PR.
--
This is an automated message from the Apache Git Service.
To respond to the message, please
lupyuen commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2325486727
@no1wudi Sorry I didn't realise the implications. Could we remove
`EXPERIMENTAL` for now, and add it to the next version? So we don't hold up the
release of librust. Thanks!
--
This i
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2325471960
The CI report error if `EXPERI` enabled:
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go t
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2323164094
@lupyuen @acassis OK, indeed, a significant amount of work is still required
before it stabilizes. I would like to clarify that the issue regarding the
compatibility of struct definitions
lupyuen commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2323152085
@no1wudi I agree with Alan:
1. Rust Standard Library on NuttX will need lots of testing by the NuttX
Community. So it's good to release the Initial Version early, get feedback,
improve
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2323107827
> @no1wudi @lupyuen maybe we should have an option to let the user test it
even when we don't have the guarantees of Struct Sizes case he/she enables
EXPERIMENTAL, what to you think?
>
acassis commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2322927995
> Hi @acassis do you think we should proceed with this approach, to verify
the NuttX Struct Sizes? This is for creating the initial version of Rust
Standard Library for NuttX. Thanks!
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2322920009
> @no1wudi could we check the struct layout by parsing symbol from image's
elf file, which is less intrusive and more accurate?
That need NuttX to build with debug symbol enabled an
xiaoxiang781216 commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2322894528
@no1wudi could we check the struct layout by parsing symbol from image's elf
file, which is less intrusive and more accurate?
--
This is an automated message from the Apache Git
lupyuen commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2322747346
Hi @acassis do you think we should proceed with this approach, to verify the
NuttX Struct Sizes? This is for creating the initial version of Rust Standard
Library for NuttX. Thanks!
--
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2322644369
> libc/libstd definitely needs the participation of nuttx to be compiled,
similar to the WAMR glue code we did before, which allows rust to be bound to a
certain configuration to avoi
lupyuen commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2322628658
> I'm voting no on this PR unless there are other RTOS that use similar
alternative practices.
Hi @anchao are we thinking of the [bindgen
approach](https://github.com/apache/nuttx/
xiaoxiang781216 commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2321846028
> I'm voting no on this PR unless there are other RTOS that use similar
alternative practices.
Why not support both approach?
- NuttX is compliant with POSIX, so it c
anchao commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2321645836
> > Why not generate structure code dynamically, but bind statically?
>
> Please see #12960, generate structure code dynamically is OK on technical,
but binary compatibility cannot b
no1wudi commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2320413281
> Why not generate structure code dynamically, but bind statically?
Please see https://github.com/apache/nuttx/issues/12960, generate structure
code dynamically is OK on technical,
anchao commented on PR #13245:
URL: https://github.com/apache/nuttx/pull/13245#issuecomment-2320385832
Why not generate structure code dynamically, but bind statically?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use t
no1wudi opened a new pull request, #13245:
URL: https://github.com/apache/nuttx/pull/13245
## Summary
This patch adds a static compatiblity check
for most of structures from NuttX, which is used by Rust.
Currently, for most of them have a reserved field, which is for possible
32 matches
Mail list logo
