All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mikko Rapeli <mikko.rapeli@linaro.org>
To: Trevor Woerner <twoerner@gmail.com>
Cc: Khem Raj <raj.khem@gmail.com>, Ross Burton <Ross.Burton@arm.com>,
	Sathishkumar Duraisamy <sathishkumar.d.cbe@gmail.com>,
	"openembedded-core@lists.openembedded.org"
	<openembedded-core@lists.openembedded.org>
Subject: Re: [OE-core] systemd build failure with gcc 15 / tpm2 / aarch64: gcs required
Date: Thu, 8 May 2025 09:22:53 +0300	[thread overview]
Message-ID: <aBxNvW-JhKHitdWN@nuoska> (raw)
In-Reply-To: <20250507191029.GA24048@localhost>

Hi,

On Wed, May 07, 2025 at 03:10:29PM -0400, Trevor Woerner wrote:
> On Wed 2025-05-07 @ 03:04:49 PM, Trevor Woerner wrote:
> > On Wed 2025-05-07 @ 11:38:12 AM, Khem Raj wrote:
> > > On Wed, May 7, 2025 at 11:34 AM Ross Burton <Ross.Burton@arm.com> wrote:
> > > >
> > > > On 7 May 2025, at 19:29, Trevor Woerner via lists.openembedded.org <twoerner=gmail.com@lists.openembedded.org> wrote:
> > > > >
> > > > > I stumbled across this build issue via a completely different route than most
> > > > > others, it seems, and certainly different than what you have described. I'm
> > > > > not using meta-security and I'm not using tpm2. Therefore your patch will do
> > > > > little to solve my build, and others will likely stumble across this issue by
> > > > > other routes as well.
> > > >
> > > > For what it’s worth, we’re seeing this without meta-security too.  This needs to be worked around in core (personally, I’d remove —fatal-warnings from systemd until openssl includes the GCS hints).
> > > >
> > > 
> > > thats a bigger hammer, I would suggest to apply
> > > LDFLAGS:append:aarch64 = " -Wl,-z,gcs-report-dynamic=none"
> > > in systemd recipe in core
> > 
> > If we want to make the hammer smaller, it's only triggered when:
> > 
> > 	PACKAGECONFIG:append:pn-systemd = " openssl"
> > 
> 
> diff --git a/meta/recipes-core/systemd/systemd_257.5.bb b/meta/recipes-core/systemd/systemd_257.5.bb
> index acf97517a94a..995b55580e4a 100644
> --- a/meta/recipes-core/systemd/systemd_257.5.bb
> +++ b/meta/recipes-core/systemd/systemd_257.5.bb
> @@ -240,6 +240,7 @@ RESOLV_CONF ??= ""
>  # bpf-framework: pass the recipe-sysroot to the compiler used to build
>  # the eBPFs, so that it can find needed system includes in there.
>  CFLAGS:append = " --sysroot=${STAGING_DIR_TARGET}"
> +LDFLAGS:append:aarch64 = " ${@bb.utils.contains('PACKAGECONFIG', 'openssl', '-Wl,-z,gcs-report-dynamic=none', '', d)}"
> 
>  EXTRA_OEMESON += "-Dnobody-user=nobody \
>                    -Dnobody-group=nogroup \

I would prefer this but in the past I had trouble getting non-default
config fixes to oe-core and thus I proposed the patch to meta-security/meta-tpm
only. I do understand that "openssl" in systemd PACKAGECONFIG can be enabled
in a number of builds and setups, but it's not the default. Maybe
it should be the default, together with "repart"? Another thread, hopefully
shorter than "efi".

There are a lot of corner cases and non-default configs which need fixes
to oe-core. I would like to understand what the policy for fixes is.

I think it would make sense to enable more features by default, but allow them
to be disabled. TPM support being one of them.

Cheers,

-Mikko


  parent reply	other threads:[~2025-05-08  6:23 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-05-05 11:46 systemd build failure with gcc 15 / tpm2 / aarch64: gcs required Max Krummenacher
2025-05-05 12:33 ` [OE-core] " Vyacheslav Yurkov
2025-05-05 12:56   ` Max Krummenacher
2025-05-05 14:45 ` Khem Raj
2025-05-05 19:09   ` [OE-core] " Randy MacLeod
2025-05-06 11:38     ` Sathishkumar D
2025-05-06 12:24       ` [OE-core] " Mikko Rapeli
2025-05-06 13:13       ` Khem Raj
2025-05-06 13:28         ` Sathishkumar Duraisamy
2025-05-06 13:32           ` Khem Raj
2025-05-06 13:45             ` Mikko Rapeli
2025-05-06 14:21               ` Max Krummenacher
2025-05-06 22:59           ` Khem Raj
2025-05-07  6:04             ` Sathishkumar Duraisamy
2025-05-07  6:14               ` Khem Raj
2025-05-07  8:22                 ` Mikko Rapeli
     [not found]                 ` <183D310FC8853D5E.1749@lists.openembedded.org>
2025-05-07  8:31                   ` Mikko Rapeli
2025-05-07  8:55                     ` Sathishkumar Duraisamy
2025-05-07 14:22                   ` Mikko Rapeli
2025-05-07 14:35                     ` Khem Raj
2025-05-07 16:05                       ` Sathishkumar Duraisamy
2025-05-07 18:29                     ` Trevor Woerner
2025-05-07 18:33                       ` Ross Burton
2025-05-07 18:38                         ` Khem Raj
2025-05-07 19:04                           ` Trevor Woerner
2025-05-07 19:10                             ` Trevor Woerner
2025-05-07 19:51                               ` Khem Raj
2025-05-08  6:22                               ` Mikko Rapeli [this message]
2025-05-08  7:00                                 ` Khem Raj

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=aBxNvW-JhKHitdWN@nuoska \
    --to=mikko.rapeli@linaro.org \
    --cc=Ross.Burton@arm.com \
    --cc=openembedded-core@lists.openembedded.org \
    --cc=raj.khem@gmail.com \
    --cc=sathishkumar.d.cbe@gmail.com \
    --cc=twoerner@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.