public inbox for openembedded-core@lists.openembedded.org
 help / color / mirror / Atom feed
From: Yash Shinde <Yash.Shinde@windriver.com>
To: Martin Jansa <martin.jansa@gmail.com>
Cc: raj.khem@gmail.com, openembedded-core@lists.openembedded.org,
	Randy.MacLeod@windriver.com, Sundeep.Kokkonda@windriver.com
Subject: Re: [OE-core] [PATCH v2 2/3] rust: Upgrade 1.84.1->1.85.0
Date: Tue, 15 Apr 2025 21:21:35 +0530	[thread overview]
Message-ID: <7c828e04-47da-42a3-aa3b-0893efe6ab87@windriver.com> (raw)
In-Reply-To: <CA+chaQf_eWJgp9QVKQQ9iGviZU54Zc_1h-LyWW2yurWf06zCLA@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 1395 bytes --]


On 15-04-2025 17:08, Martin Jansa wrote:
> qemuarmv5 uses tune-arm926ejs which uses armv5te, I think you would
> need to use armv5te-vfp to reproduce this, or use qemuarm with
> tune-cortexa15 but switch from default cortexa15thf-neon to
> cortexa15t-neon
>
> I was seeing it in multilib builds e.g. with
> https://git.openembedded.org/openembedded-core-contrib/commit/?h=jansa/master&id=40e9c4744a7f1eb2d91af9539d69555d36dfba30
> which uses DEFAULTTUNE:virtclass-multilib-lib32 = "armv7at-neon"
>
> unfortunately just treating softfp as soft ends with segfaults in rust
> as shown nere:
> https://git.openembedded.org/openembedded-core-contrib/commit/?h=jansa/master&id=c65eda95e5dd3ad0be0565477236d7902d5d625c
>
> Regards,
>
I checked for softfp tuning using (which gave TARGET_FPU = softfp): 
TUNE_FEATURES = "vfp armv5te" DEFAULTTUNE = "armv5te-vfp" TUNE_PKGARCH = 
"armv5te-vfp" But, it seems that softfp is not a valid value for 
llvm-floatabi in rust. It gives error as follows: | --- stderr | error: 
Error loading target specification: 'softfp' is not a valid value for 
llvm-floatabi. Use 'soft' or 'hard'.. Run `rustc --print target-list` 
for a list of built-in targets As you mentioned, using soft instead of 
softfp also doesn't work either. It seems that unless upstream Rust 
provides support for this, we won't be able to build Rust with softfp on 
ARM targets. Regards, Yash.

[-- Attachment #2: Type: text/html, Size: 2359 bytes --]

WARNING: multiple messages have this Message-ID (diff)
From: "Yash Shinde" <Yash.Shinde@windriver.com>
To: Martin Jansa <martin.jansa@gmail.com>
Cc: raj.khem@gmail.com, openembedded-core@lists.openembedded.org,
	Randy.MacLeod@windriver.com, Sundeep.Kokkonda@windriver.com
Subject: Re: [OE-core] [PATCH v2 2/3] rust: Upgrade 1.84.1->1.85.0
Date: Tue, 15 Apr 2025 08:53:43 -0700	[thread overview]
Message-ID: <7c828e04-47da-42a3-aa3b-0893efe6ab87@windriver.com> (raw)
Message-ID: <20250415155343.CqVFsjI9fEI9zmc5UGBIeByrgJhfGvpa4CgRVTtADtw@z> (raw)
In-Reply-To: <CA+chaQf_eWJgp9QVKQQ9iGviZU54Zc_1h-LyWW2yurWf06zCLA@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 1458 bytes --]

On 15-04-2025 17:08, Martin Jansa wrote:

> 
> qemuarmv5 uses tune-arm926ejs which uses armv5te, I think you would
> need
> to use armv5te-vfp to reproduce this, or use qemuarm with
> tune-cortexa15
> but switch from default cortexa15thf-neon to
> cortexa15t-neon
> 
> I was seeing
> it in multilib builds e.g. with
> https://git.openembedded.org/openembedded-core-contrib/commit/?h=jansa/master&id=40e9c4744a7f1eb2d91af9539d69555d36dfba30
> 
> which uses DEFAULTTUNE:virtclass-multilib-lib32 = "armv7at-neon"
> 
> unfortunately just treating softfp as soft ends with segfaults in rust
> as
> shown nere:
> https://git.openembedded.org/openembedded-core-contrib/commit/?h=jansa/master&id=c65eda95e5dd3ad0be0565477236d7902d5d625c
> 
> 
> Regards,
> 
> 

I checked for softfp tuning using (which gave TARGET_FPU = softfp):

TUNE_FEATURES = "vfp armv5te"
DEFAULTTUNE = "armv5te-vfp"
TUNE_PKGARCH = "armv5te-vfp"

But, it seems that softfp is not a valid value for llvm-floatabi in rust.
It gives error as follows:

| --- stderr | error: Error loading target specification: 'softfp' is not a valid value for llvm-floatabi. Use 'soft' or 'hard'.. Run `rustc --print target-list` for a list of built-in targets

As you mentioned, using soft instead of softfp also doesn't work either.
It seems that unless upstream Rust provides support for this, we won't be able to build Rust with softfp on ARM targets.

Regards,
Yash.

[-- Attachment #2: Type: text/html, Size: 2121 bytes --]

  parent reply	other threads:[~2025-04-15 15:51 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-03-31  5:54 [PATCH v2 1/3] rust: Update "do_update_snapshot" task for rust-snapshot.inc Yash.Shinde
2025-03-31  5:54 ` [PATCH v2 2/3] rust: Upgrade 1.84.1->1.85.0 Yash.Shinde
2025-03-31  6:24   ` [OE-core] " Richard Purdie
2025-03-31  7:16     ` Yash Shinde
2025-03-31  8:17       ` Richard Purdie
2025-03-31 16:46   ` Khem Raj
2025-04-11  7:38     ` Martin Jansa
2025-04-15  9:04       ` Yash Shinde
2025-04-15 11:38         ` Martin Jansa
2025-04-15 15:45           ` Yash Shinde
2025-04-15 15:52             ` Martin Jansa
2025-04-15 15:58               ` Yash Shinde
2025-04-15 15:51           ` Yash Shinde [this message]
2025-04-15 15:53             ` Yash Shinde
2025-03-31  5:54 ` [PATCH v2 3/3] rust: Upgrade 1.85.0->1.85.1 Yash.Shinde

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=7c828e04-47da-42a3-aa3b-0893efe6ab87@windriver.com \
    --to=yash.shinde@windriver.com \
    --cc=Randy.MacLeod@windriver.com \
    --cc=Sundeep.Kokkonda@windriver.com \
    --cc=martin.jansa@gmail.com \
    --cc=openembedded-core@lists.openembedded.org \
    --cc=raj.khem@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox