X-Original-To: alpine-devel@lists.alpinelinux.org Received: from mail-yw0-f196.google.com (mail-yw0-f196.google.com [209.85.161.196]) by lists.alpinelinux.org (Postfix) with ESMTP id EEEB15C43E3 for ; Sat, 5 Nov 2016 18:26:06 +0000 (GMT) Received: by mail-yw0-f196.google.com with SMTP id b66so836183ywh.2 for ; Sat, 05 Nov 2016 11:26:06 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=dbCFoNX2FU7cvnE7s1AbmTCKnysuu2HPqzdmi47Ql/k=; b=PQHb0lzm4j0QXz2ZmAMwSH/Nv98Ml88OtsP7aDrepWmfbnXlozpTCUOCjZR31k+XSs w1w9moHUBulQdQZQSah+LSeQh2noaYs/KtY8qdtJXKRPSamwnMpMghqcj68EZB5vN7F3 CYVccEJ69Y5yH9FfD0PgeEGERt32PJ8nmY8Jd6xJSxQyS5UUYkUl2WGxrTEDoP/3m1IM OBOzIZ+GCjUW8sysRqYuNT4Ia3iePG69R5kk2K7FdLMOzzoEbZVtj9+QN3qDZcdb78jP w84xQLvV7V5fUvMt4HqQCiZsMG+n5mDZ9Kok+SlHvXk5//P9MZygAv6NhCD6SbLo9d3l I0qA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=dbCFoNX2FU7cvnE7s1AbmTCKnysuu2HPqzdmi47Ql/k=; b=KMK2yF4zKT7Dy4W1GknuE293U/lPPDk+m01qiYrsYRPMhwWw5/du06zV85R1i2KYSw kIhNGCfXiN6oV7dd+Mx2sjuLZnLy8UYmGgJtsXyM4Q76366KZ0LRYaQix1DHQxL9JT47 cB1nvsEEnLhZLuvuoxNXSTcxt0d/ElLxrYXVMFBvAN96VAIugoJZglPtA0knzBLw7Fhi 6v3scEd5coO33Z/y+dQu6IjBSJL15LdZvzoE0LrhupYb5Dh32jMRn24YP1+kqW0lKaC6 J6bR8jo91S2d2ocFhpuLsvSnrtazSB0tX06JxmC3G9pcQL3jQbwe/nMLq1Ucci6Lkluu qHiw== X-Gm-Message-State: ABUngvf1D4iSBZSU/61XK71B4P3TKlXRJOl0VeJpO53hTYR+VOUcyXo6MyIJ0Tv08MhMXLZB6l2pwi+9cC9WIQ== X-Received: by 10.13.221.215 with SMTP id g206mr9234408ywe.350.1478370359950; Sat, 05 Nov 2016 11:25:59 -0700 (PDT) X-Mailinglist: alpine-devel Precedence: list List-Id: Alpine Development List-Unsubscribe: List-Post: List-Help: List-Subscribe: MIME-Version: 1.0 References: <20161105162629.1dd58e01@vostro> In-Reply-To: <20161105162629.1dd58e01@vostro> From: Josiah Worcester Date: Sat, 05 Nov 2016 18:25:49 +0000 Message-ID: Subject: Re: [alpine-devel] libnss-winbind ABI disagrees with glibc ABI for NSS modules To: Timo Teras Cc: alpine-devel@lists.alpinelinux.org, ncopa@alpinelinux.org Content-Type: multipart/alternative; boundary=94eb2c075b5226c59e054091ec52 --94eb2c075b5226c59e054091ec52 Content-Type: text/plain; charset=UTF-8 On Sat, Nov 5, 2016 at 8:26 AM Timo Teras wrote: > Hi, > > On Fri, 28 Oct 2016 22:25:39 +0000 > Josiah Worcester wrote: > > > I've been working on writing an nscd for musl systems ( > > https://github.com/pikhq/musl-nscd ). > > Cool! > > > This nscd assumes the glibc ABI for NSS modules is in place. > Unfortunately, if nss.h is not present > > at compile time for Samba, then it uses its own version of the NSS > > ABI (in > > > https://git.samba.org/?p=samba.git;a=blob;f=nsswitch/winbind_nss.h;h=05e07ad5d3137ffaecec9656298222a81451ce59;hb=HEAD > > ), which does not quite match. > > What does not match? The sole enum there? Is something else needed? That's exactly what doesn't match. > > > I'd like to continue supporting the glibc ABI for NSS modules, to > > potentially let proprietary NSS modules continue to function. To do > > this, Samba will need to have an appropriate nss.h around. > > > > I suppose if musl-nscd were in alpine upstream, then just depending > > on a dev package would be appropriate. For now, I feel like the > > appropriate thing would be to copy nss.h into the apkbuild (e.g. from > > https://github.com/pikhq/musl-nscd/blob/master/include/nss.h ), and > > make sure Samba uses it. Think this could get done? > > Could we just patch the samba nss.h until musl-nscd-dev becomes > available? Or does it trigger additional functionality to be built if > nss.h exists? > > What modules of samba use this file? Is it just some nss plugin which > has been unused so far? If yes, I'm happy to just apply a patch. But if > it affects some other potentially in-use things, we need to consider if > it breaks things. > This values in the enum also apply to some of libwbclient's return values. I don't believe this is currently being used outside of samba's own tools, though, because to my knowledge the header also isn't exported. Aside from that, the only use is this nss plugin, which is built anyways. As such, patching samba's values might well be the best solution. Thanks for looking into this! Josiah --94eb2c075b5226c59e054091ec52 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
On Sat, Nov 5,= 2016 at 8:26 AM Timo Teras <timo.t= eras@iki.fi> wrote:
Hi,

On Fri, 28 Oct 2016 22:25:39 +0000
Josiah Worcester <josiahw@gmail.com> wrote:

> I've been working on writing an nscd for musl systems (
> https://github.com/pikhq/musl-nscd ).=

Cool!

> This nscd assumes the glibc ABI for NSS modules is in place. Unfortuna= tely, if nss.h is not present
> at compile time for Samba, then it uses its own version of the NSS
> ABI (in
> https://git.samba.org= /?p=3Dsamba.git;a=3Dblob;f=3Dnsswitch/winbind_nss.h;h=3D05e07ad5d3137ffaece= c9656298222a81451ce59;hb=3DHEAD
> ), which does not quite match.

What does not match? The sole enum there? Is something else needed?
That's exactly what doesn't match.
=C2=A0
=
=C2=A0
<= blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px= #ccc solid;padding-left:1ex"> > I'd like to continue supporting the glibc ABI for NSS modules, to<= br class=3D"gmail_msg"> > potentially let proprietary NSS modules continue to function. To do > this, Samba will need to have an appropriate nss.h around.
>
> I suppose if musl-nscd were in alpine upstream, then just depending > on a dev package would be appropriate. For now, I feel like the
> appropriate thing would be to copy nss.h into the apkbuild (e.g. from<= br class=3D"gmail_msg"> > https://github.= com/pikhq/musl-nscd/blob/master/include/nss.h ), and
> make sure Samba uses it. Think this could get done?

Could we just patch the samba nss.h until musl-nscd-dev becomes
available? Or does it trigger additional functionality to be built if
nss.h exists?

What modules of samba use this file? Is it just some nss plugin which
has been unused so far? If yes, I'm happy to just apply a patch. But if=
it affects some other potentially in-use things, we need to consider if
it breaks things.

T= his values in the enum also apply to some of libwbclient's return value= s. I don't believe this is currently being used outside of samba's = own tools, though, because to my knowledge the header also isn't export= ed.

Aside from that, the only use is this nss plug= in, which is built anyways. As such, patching samba's values might well= be the best solution.

Thanks for looking into thi= s!
Josiah
--94eb2c075b5226c59e054091ec52-- --- Unsubscribe: alpine-devel+unsubscribe@lists.alpinelinux.org Help: alpine-devel+help@lists.alpinelinux.org ---