From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailout4.zoneedit.com (mailout4.zoneedit.com [64.68.198.64]) by mx.groups.io with SMTP id smtpd.web11.3463.1593626926072796458 for ; Wed, 01 Jul 2020 11:08:46 -0700 Authentication-Results: mx.groups.io; dkim=missing; spf=none, err=permanent DNS error (domain: denix.org, ip: 64.68.198.64, mailfrom: denis@denix.org) Received: from localhost (localhost [127.0.0.1]) by mailout4.zoneedit.com (Postfix) with ESMTP id 6B61B40BA1; Wed, 1 Jul 2020 18:08:45 +0000 (UTC) Received: from mailout4.zoneedit.com ([127.0.0.1]) by localhost (zmo14-pco.easydns.vpn [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id dCf7QPGHOCBl; Wed, 1 Jul 2020 18:08:45 +0000 (UTC) Received: from mail.denix.org (pool-100-15-86-127.washdc.fios.verizon.net [100.15.86.127]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mailout4.zoneedit.com (Postfix) with ESMTPSA id 24BE3409A6; Wed, 1 Jul 2020 18:08:40 +0000 (UTC) Received: by mail.denix.org (Postfix, from userid 1000) id 678B5173D5E; Wed, 1 Jul 2020 14:08:40 -0400 (EDT) Date: Wed, 1 Jul 2020 14:08:40 -0400 From: "Denys Dmytriyenko" To: Gabor Abonyi Cc: "JPEWhacker@gmail.com" , Jon Mason , "meta-arm@lists.yoctoproject.org" , nd Subject: Re: [meta-arm] [PATCH 4/6] arm: trusted-firmware-m: Add recipe Message-ID: <20200701180840.GT17660@denix.org> References: <20200622071401.2570-1-gabor.abonyi@arm.com> <20200622071401.2570-5-gabor.abonyi@arm.com> <20200622164936.GF17660@denix.org> <20200622172654.GC14044@kudzu.us> <20200622173611.GG17660@denix.org> MIME-Version: 1.0 In-Reply-To: User-Agent: Mutt/1.5.20 (2009-06-14) Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, Jun 23, 2020 at 08:59:56AM +0000, Gabor Abonyi wrote: > If I move TF-M to meta-arm-bsp, then meta-arm-bsp would have the depend= ency on arm-toolchain. Would that be better? > (I do not understand why meta-arm depending on arm-toolchain is a bad t= hing, explanations are welcome :) ) meta-arm is meant to be a central place for shared components to be used = by=20 other BSPs. Kind of a community layer, if you like. And arm-toolchain is = kind=20 of a special layer and may not be desirable by all BSPs, so that would re= duce=20 overall usefulness of meta-arm layer. Each individual BSP (and meta-arm-b= sp is=20 one of those) is free to depend on arm-toolchain layer directly, if neede= d. --=20 Denys > I experimented with BBFILES_DYNAMIC, but it is very unhelpful from end-= user perspective (if arm-toolchain is not in bblayers): > "ERROR: Nothing PROVIDES 'trusted-firmware-m'. Close matches: > trusted-firmware-a" > Compared to the current one: > "ERROR: Layer 'meta-arm' depends on layer 'arm-toolchain', but this lay= er is not enabled in your configuration" >=20 > I am open to any suggestions. > =20 > -----Original Message----- > From: meta-arm@lists.yoctoproject.org = On Behalf Of Joshua Watt via lists.yoctoproject.org > Sent: 2020. j=C3=BAnius 22., h=C3=A9tf=C5=91 20:04 > To: Denys Dmytriyenko ; Jon Mason > Cc: Gabor Abonyi ; meta-arm@lists.yoctoproject.or= g; nd > Subject: Re: [meta-arm] [PATCH 4/6] arm: trusted-firmware-m: Add recipe >=20 >=20 > On 6/22/20 12:36 PM, Denys Dmytriyenko wrote: > > On Mon, Jun 22, 2020 at 01:26:54PM -0400, Jon Mason wrote: > >> On Mon, Jun 22, 2020 at 12:49:36PM -0400, Denys Dmytriyenko wrote: > >>> On Mon, Jun 22, 2020 at 09:13:59AM +0200, Gabor Abonyi wrote: > >>>> Adds a recipe to pull down the trusted-firmware-m repository and=20 > >>>> the ones it depends on. The recipe can either use=20 > >>>> gcc-arm-none-eabi-native or armcompiler-native Clang toolchain to = compile the firmware. > >>>> > >>>> Change-Id: I37a4ba38982b5b1d387eccbb26bb5c79bddab0f7 > >>>> Signed-off-by: Gabor Abonyi > >>>> --- > >>>> meta-arm/conf/layer.conf | 1 + > >>>> .../trusted-firmware-m/trusted-firmware-m.inc | 118 ++++++++++++= ++++++ > >>>> .../trusted-firmware-m_1.0.bb | 25 ++++ > >>>> 3 files changed, 144 insertions(+) > >>>> create mode 100644 meta-arm/recipes-bsp/trusted-firmware-m/trust= ed-firmware-m.inc > >>>> create mode 100644=20 > >>>> meta-arm/recipes-bsp/trusted-firmware-m/trusted-firmware-m_1.0.bb > >>>> > >>>> diff --git a/meta-arm/conf/layer.conf b/meta-arm/conf/layer.conf=20 > >>>> index 3341972..10a7951 100644 > >>>> --- a/meta-arm/conf/layer.conf > >>>> +++ b/meta-arm/conf/layer.conf > >>>> @@ -11,5 +11,6 @@ BBFILE_PRIORITY_meta-arm =3D "6" > >>>> =20 > >>>> LAYERDEPENDS_meta-arm =3D " \ > >>>> core \ > >>>> + arm-toolchain \ > >>> This may be problematic... > >> Yes, this was flagged as a potential problem internally. =20 > >> Fortunately, meta-arm-toolchain currently has no other layer=20 > >> dependencies. So, it shouldn't hurt too much. Ccing JPEW directly=20 > >> on this, since his insight has been helpful in the past. > >> > >> TF-M requires Arm's LLVM/Clang based toolchain to compile, which is=20 > >> the second patch of this series. > > One of the options is to move TF-M to meta-arm-bsp... >=20 > Ya, I wouldn't want to mandate meta-arm-toolchain here. You could eithe= r move it to meta-arm-bsp as Denys suggested or possibly use BBFILES_DYNA= MIC? >=20 >=20 > >