Fix ESP-IDF constants and structs

[SergioGasquez]

This PR fixes some constants and structs which were wrong for ESP-IDF targets. I created an app that compares all the constants in src/unix/newlib and src/unix/newlib/espidf modules with the ones in ESP-IDF source code (accessed via esp_idf_svc::sys). Here is the app https://github.com/SergioGasquez/libc-checks, using [email protected] yields the following errors:

E (622) libc_checks: Mismatch detected for constant `SIGABRT`: `esp-idf` 6 | `libc` 1
E (632) libc_checks: Mismatch detected for constant `SIGFPE`: `esp-idf` 8 | `libc` 1
E (642) libc_checks: Mismatch detected for constant `SIGILL`: `esp-idf` 4 | `libc` 1
E (652) libc_checks: Mismatch detected for constant `SIGINT`: `esp-idf` 2 | `libc` 1
E (662) libc_checks: Mismatch detected for constant `SIGSEGV`: `esp-idf` 11 | `libc` 1
E (672) libc_checks: Mismatch detected for constant `SIGTERM`: `esp-idf` 15 | `libc` 1
E (672) libc_checks: Mismatch detected for constant `SIGQUIT`: `esp-idf` 3 | `libc` 1
E (682) libc_checks: Mismatch detected for constant `NSIG`: `esp-idf` 32 | `libc` 2
E (692) libc_checks: Mismatch detected for type `tcflag_t` size: `esp-idf` 2 | `libc` 4
E (702) libc_checks: Mismatch detected for type `tcflag_t` alignment: `esp-idf` 2 | `libc` 4
E (712) libc_checks: Mismatch detected for type `sigaction` size: `esp-idf` 12 | `libc` 72
E (722) libc_checks: Mismatch detected for type `termios` size: `esp-idf` 28 | `libc` 52
E (732) libc_checks: Mismatch detected for type `pthread_attr_t` alignment: `esp-idf` 4 | `libc` 1
E (742) libc_checks: Mismatch detected for constant `NCCS`: `esp-idf` 11 | `libc` 32
E (752) libc_checks: Mismatch detected for constant `O_CLOEXEC`: `esp-idf` 262144 | `libc` 524288
E (762) libc_checks: Mismatch detected for constant `PF_INET6`: `esp-idf` 10 | `libc` 23
E (762) libc_checks: Mismatch detected for constant `SOCK_CLOEXEC`: `esp-idf` 262144 | `libc` 524288
E (772) libc_checks: Mismatch detected for constant `TCP_NODELAY`: `esp-idf` 1 | `libc` 8193
E (782) libc_checks: Mismatch detected for constant `TCP_KEEPIDLE`: `esp-idf` 3 | `libc` 256
E (792) libc_checks: Mismatch detected for constant `TCP_KEEPINTVL`: `esp-idf` 4 | `libc` 512
E (802) libc_checks: Mismatch detected for constant `TCP_KEEPCNT`: `esp-idf` 5 | `libc` 1024
E (812) libc_checks: Mismatch detected for constant `IP_TOS`: `esp-idf` 1 | `libc` 3
E (822) libc_checks: Mismatch detected for constant `IP_TTL`: `esp-idf` 2 | `libc` 8
E (832) libc_checks: Mismatch detected for constant `IP_MULTICAST_IF`: `esp-idf` 6 | `libc` 9
E (842) libc_checks: Mismatch detected for constant `IP_MULTICAST_TTL`: `esp-idf` 5 | `libc` 10
E (842) libc_checks: Mismatch detected for constant `IP_MULTICAST_LOOP`: `esp-idf` 7 | `libc` 11
E (852) libc_checks: Mismatch detected for constant `IP_ADD_MEMBERSHIP`: `esp-idf` 3 | `libc` 11
E (862) libc_checks: Mismatch detected for constant `IP_DROP_MEMBERSHIP`: `esp-idf` 4 | `libc` 12
E (872) libc_checks: Mismatch detected for constant `IPV6_MULTICAST_IF`: `esp-idf` 768 | `libc` 9
E (882) libc_checks: Mismatch detected for constant `IPV6_MULTICAST_HOPS`: `esp-idf` 769 | `libc` 10
E (892) libc_checks: Mismatch detected for constant `IPV6_MULTICAST_LOOP`: `esp-idf` 770 | `libc` 11
E (902) libc_checks: Mismatch detected for constant `HOST_NOT_FOUND`: `esp-idf` 210 | `libc` 1
E (912) libc_checks: Mismatch detected for constant `NO_DATA`: `esp-idf` 211 | `libc` 2
E (922) libc_checks: Mismatch detected for constant `NO_RECOVERY`: `esp-idf` 212 | `libc` 3
E (932) libc_checks: Mismatch detected for constant `TRY_AGAIN`: `esp-idf` 213 | `libc` 4
E (942) libc_checks: Mismatch detected for constant `AI_NUMERICSERV`: `esp-idf` 8 | `libc` 0
E (952) libc_checks: Mismatch detected for constant `AI_ADDRCONFIG`: `esp-idf` 64 | `libc` 0
E (962) libc_checks: Mismatch detected for constant `NI_NUMERICSERV`: `esp-idf` 8 | `libc` 0
E (972) libc_checks: Mismatch detected for constant `NI_DGRAM`: `esp-idf` 16 | `libc` 0
E (972) libc_checks: Mismatch detected for constant `EAI_FAMILY`: `esp-idf` 204 | `libc` -303
E (982) libc_checks: Mismatch detected for constant `EAI_MEMORY`: `esp-idf` 203 | `libc` -304
E (992) libc_checks: Mismatch detected for constant `EAI_NONAME`: `esp-idf` 200 | `libc` -305
E (1002) libc_checks: Mismatch detected for constant `EAI_SOCKTYPE`: `esp-idf` 10 | `libc` -307

All those errors are fixed in this PR, if I point the app to use the branch of this PR, no errors appear.

cc and thanks to @ivmarkov!

closes #3774

[rustbot]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @JohnTitor (or someone else) some time within the next two weeks.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

• @rustbot author: the review is finished, PR author should check the comments and take action accordingly
• @rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

[ivmarkov]

@JohnTitor Just to mention that this PR is a continuation of #3345

What we did is to

• Take the original C comparison code of @zetanumbers
• Translate it to Rust so that we can directly use the libc types and constants
• Extend it with more types
• And then run the libc types against our "source of truth", which is the bindgen-generated Rust bindings of the ESP-IDF complete API - which - amongst other things - contains all newlib types with their proper alignment and size on ESP-IDF.

Obviously, we are running the comparison on the actual Espressif MCUs (or an emulation of those). And will likely continue doing so, in CI.

[ivmarkov]

@JohnTitor Sorry for the ping. Wondering if there is anything else we need to do so as this PR to be considered in a good shape for merging.

[ivmarkov]

?r @tgross35 Would you mind reviewing?

The PR should be relatively uncontroversial in that it affects newlib/ESP IDF only and fixes obvious bugs (wrong definitions).

[ivmarkov]

@SergioGasquez Can you assign this to @tgross35 please?

@JohnTitor @tgross35 if both of you are busy, would you suggest a reviewer that can look into this?

[SergioGasquez]

?r @tgross35

[SergioGasquez]

?r @Amanieu

[tgross35]

The non-espif newlib headers look correctly unchanged, if your tests indicate that things are fixed for espif then that seems reasonable to me. Please link headers in this PR if they are available.

Note that this repo just has a somewhat slow turn time for a handful of reasons, I keep an eye on all PRs even if they aren't assigned to me. (And the review request syntax is r?, not ?r).

[jothan]

@MabezDev would it be possible to bump the libc dependency of the std crate on the next compiler build ?

It would likely fix a lot of issues for a lot of people.

Thanks as always !

[SergioGasquez]

Upstream already bumped the libc version, which was included in 1.84 nightly, so, worst case scenario we will have the changes in 2 releases, but maybe we could bump it for 1.83.