From mboxrd@z Thu Jan 1 00:00:00 1970 X-GM-THRID: 6798139016226013184 Date: Tue, 10 Mar 2020 08:08:27 -0700 (PDT) From: cedric_hombourger@mentor.com To: isar-users Message-Id: <87825320-c0cc-4876-8b53-8c3fe6d30db2@googlegroups.com> In-Reply-To: <20200309175936.6927bf4f@md1za8fc.ad001.siemens.net> References: <5a454616-f591-8a43-4ea7-eaed4fbe236c@siemens.com> <51e462f0-aa1f-c33c-a57b-f4773efdbb4d@siemens.com> <20200309175936.6927bf4f@md1za8fc.ad001.siemens.net> Subject: Re: dpkg-raw git fetch MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_Part_1621_26951262.1583852907192" X-Google-Token: EOvanvMFr0-1fhm712I0 X-Google-IP: 139.181.48.2 X-TUID: P6WBAPbwbKZ1 ------=_Part_1621_26951262.1583852907192 Content-Type: multipart/alternative; boundary="----=_Part_1622_67364802.1583852907192" ------=_Part_1622_67364802.1583852907192 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable On Monday, March 9, 2020 at 5:59:40 PM UTC+1, Henning Schild wrote: > > On Fri, 28 Feb 2020 13:08:10 +0000=20 > "[ext] sven.sc...@siemens.com " >=20 > wrote:=20 > > > Hi,=20 > >=20 > > I'm a little bit confused. In the task unpack the SRC_URI files will=20 > > be put into WORKDIR=20 > >=20 > https://github.com/ilbers/isar/blob/f3d50cfb9b7a5740c6dd11ee58fc4ba1a1815= eca/meta/classes/base.bbclass#L182=20 > >=20 > > But in the next task S (workdir/git) will be cleared.=20 > >=20 > https://github.com/ilbers/isar/blob/f3d50cfb9b7a5740c6dd11ee58fc4ba1a1815= eca/meta/classes/dpkg-raw.bbclass#L15=20 > >=20 > > Should I understand why ${S} will be cleard before do_install is=20 > > called?=20 > > I think you might have found an issue with the distinction of S and D.=20 > In fact they should be different folders but for dpkg-raw they are not.= =20 > I think also the debianize class does not clearly distinguish them.=20 > > I coincidentally ran into the same issue yesterday and yes debianize=20 assumes all the work should be done in ${S} should debianize expose a DEBIANIZE_DIR defaulting to ${S} and let dpkg-raw= =20 set it to ${D} instead? One workaround would be to create that second copy on your own. So you=20 > clone i.e with destsuffix and make sure that the unpacked copy and the=20 > installed copy are not the same directory.=20 > > Once you have found a solution and have a clue where the line between S= =20 > and D should be drawn, please give feedback on how to improve the class.= =20 > > Henning=20 > > > Mit freundlichen Gr=C3=BC=C3=9Fen=20 > > Sven Angelo Schultschik=20 > >=20 > > Siemens AG=20 > > Digital Industries=20 > > Process Automation=20 > > Software House Khe=20 > > DI PA CI R&D 2=20 > > =C3=96stliche Rheinbr=C3=BCckenstr. 50=20 > > 76187 Karlsruhe, Deutschland=20 > > Tel.: +49 721 595-2827=20 > > mailto:sven.sc...@siemens.com =20 > >=20 > > www.siemens.com/ingenuityforlife=20 > >=20 > > Siemens Aktiengesellschaft: Vorsitzender des Aufsichtsrats: Jim=20 > > Hagemann Snabe; Vorstand: Joe Kaeser, Vorsitzender; Roland Busch,=20 > > Lisa Davis, Klaus Helmrich, Cedrik Neike, Michael Sen, Ralf P.=20 > > Thomas; Sitz der Gesellschaft: Berlin und M=C3=BCnchen, Deutschland;=20 > > Registergericht: Berlin Charlottenburg, HRB 12300, M=C3=BCnchen, HRB 66= 84;=20 > > WEEE-Reg.-Nr. DE 23691322=20 > >=20 > > Wichtiger Hinweis: Diese E-Mail und etwaige Anlagen enthalten=20 > > firmenvertrauliche Informationen. Sollten Sie diese E-Mail irrt=C3=BCml= ich=20 > > erhalten haben, benachrichtigen Sie uns bitte durch Antwort-Mail und=20 > > l=C3=B6schen Sie diese E-Mail nebst Anlagen von Ihrem System. Vielen Da= nk.=20 > > -----Urspr=C3=BCngliche Nachricht----- Von: Jan Kiszka=20 > > > Gesendet: Donnerstag, 27. Februar= =20 > 2020 18:00=20 > > An: Schultschik, Sven (DI PA CI R&D 2)=20 > > >; isar-...@googlegroups.com=20 > Betreff:=20 > > Re: dpkg-raw git fetch=20 > >=20 > > On 27.02.20 17:39, Schultschik, Sven (DI PA CI R&D 2) wrote:=20 > > > Fetch does put the repository into the downloads folder.=20 > > >=20 > > > Which task is responsible to bring the files from=20 > > > downloads/../git-repo to ${S} ??=20 > >=20 > > Right: unpack.=20 > >=20 > > > S =3D "${WORKDIR}/git" ist defined in the recipe.=20 > > >=20 > > > Log.unpack says clone -s -n downloads/git recipe-name/version/git=20 > > > without an error=20 > > >=20 > > > Shouldn't the see the files then? =20 > >=20 > > Yes - provided the default revision checked out contains anything.=20 > >=20 > > Jan=20 > >=20 > > --=20 > > Siemens AG, Corporate Technology, CT RDA IOT SES-DE Corporate=20 > > Competence Center Embedded Linux=20 > >=20 > > ------=_Part_1622_67364802.1583852907192 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable
On Monday, March 9, 2020 at 5:59:40 PM UTC+1, Henning Schi= ld wrote:
On Fri, 28 Feb 2020 1= 3:08:10 +0000
"[ext] sven.sc...@siemens.com" <sven.sc...@siemens.com>
wrote:

> Hi,
>=20
> I'm a little bit confused. In the task unpack the SRC_URI file= s will
> be put into WORKDIR
> https://github.com/ilbers/= isar/blob/f3d50cfb9b7a5740c6dd11ee58fc4ba1a1815eca/meta/clas= ses/base.bbclass#L182=20
>=20
> But in the next task S (workdir/git) will be cleared.
> https://github.co= m/ilbers/isar/blob/f3d50cfb9b7a5740c6dd11ee58fc4ba1a1815eca/= meta/classes/dpkg-raw.bbclass#L15=20
>=20
> Should I understand why ${S} will be cleard before do_install is
> called?

I think you might have found an issue with the distinction of S and D.
In fact they should be different folders but for dpkg-raw they are not.
I think also the debianize class does not clearly distinguish them.


I coincidentally ran into the same iss= ue yesterday and yes debianize assumes all the work should be done in ${S}<= /div>
should debianize expose a DEBIANIZE_DIR defaulting to ${S} and le= t dpkg-raw set it to ${D} instead?

One workaround would be to create that second copy= on your own. So you
clone i.e with destsuffix and make sure that the unpacked copy and the
installed copy are not the same directory.

Once you have found a solution and have a clue where the line between S
and D should be drawn, please give feedback on how to improve the class= .

Henning

> Mit freundlichen Gr=C3=BC=C3=9Fen
> Sven Angelo Schultschik
>=20
> Siemens AG
> Digital Industries
> Process Automation
> Software House Khe
> DI PA CI R&D 2
> =C3=96stliche Rheinbr=C3=BCckenstr. 50
> 76187 Karlsruhe, Deutschland
> Tel.: +49 721 595-2827
> mailto:sven.sc...@siemens.com
>=20
> www.siemens.com/ingenuityforli= fe
>=20
> Siemens Aktiengesellschaft: Vorsitzender des Aufsichtsrats: Jim
> Hagemann Snabe; Vorstand: Joe Kaeser, Vorsitzender; Roland Busch,
> Lisa Davis, Klaus Helmrich, Cedrik Neike, Michael Sen, Ralf P.
> Thomas; Sitz der Gesellschaft: Berlin und M=C3=BCnchen, Deutschlan= d;
> Registergericht: Berlin Charlottenburg, HRB 12300, M=C3=BCnchen, H= RB 6684;
> WEEE-Reg.-Nr. DE 23691322
>=20
> Wichtiger Hinweis: Diese E-Mail und etwaige Anlagen enthalten
> firmenvertrauliche Informationen. Sollten Sie diese E-Mail irrt=C3= =BCmlich
> erhalten haben, benachrichtigen Sie uns bitte durch Antwort-Mail u= nd
> l=C3=B6schen Sie diese E-Mail nebst Anlagen von Ihrem System. Viel= en Dank.
> -----Urspr=C3=BCngliche Nachricht----- Von: Jan Kiszka
> <jan....@siemens.com> Gesendet: Donnerstag, 27. Februar 2020 = 18:00
> An: Schultschik, Sven (DI PA CI R&D 2)
> <sven.sc...@siemens.com>; isar-...@googlegroups.com Betre= ff:
> Re: dpkg-raw git fetch
>=20
> On 27.02.20 17:39, Schultschik, Sven (DI PA CI R&D 2) wrote:
> > Fetch does put the repository into the downloads folder.
> >=20
> > Which task is responsible to bring the files from
> > downloads/../git-repo to ${S} ??=20
>=20
> Right: unpack.
>=20
> > S =3D "${WORKDIR}/git" ist defined in the recipe.
> >=20
> > Log.unpack says clone -s -n downloads/git recipe-name/version= /git=20
> > without an error
> >=20
> > Shouldn't the see the files then? =C2=A0
>=20
> Yes - provided the default revision checked out contains anything.
>=20
> Jan
>=20
> --
> Siemens AG, Corporate Technology, CT RDA IOT SES-DE Corporate
> Competence Center Embedded Linux
>=20

------=_Part_1622_67364802.1583852907192-- ------=_Part_1621_26951262.1583852907192--