[OpenAFS-devel] AIX build fails with missing symbol
.krb5_c_make_random_key
Ben Huntsman
ben@huntsmans.net
Mon, 15 Aug 2022 06:27:55 +0000
--_000_MWHPR0701MB3674A3291C36E4A7A58BFA69A7689MWHPR0701MB3674_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
So looking into a build on AIX 7.1 was where I was going next, I was just t=
rying to get it working on AIX 6.1 for reference.
There's something significant in AIX 7.1 that doesn't work with OpenAFS, bu=
t which is not also present in AIX 7.2?
Thank you!
-Ben
________________________________
From: Daria Phoebe Brashear <shadow@gmail.com>
Sent: Sunday, August 14, 2022 7:33 PM
To: Jeffrey E Altman (jaltman@auristor.com) <jaltman@auristor.com>
Cc: Ben Huntsman <ben@huntsmans.net>; Benjamin Kaduk <kaduk@mit.edu>; Ken H=
ornstein <kenh@cmf.nrl.navy.mil>; openafs-devel@openafs.org <openafs-devel@=
openafs.org>
Subject: Re: [OpenAFS-devel] AIX build fails with missing symbol .krb5_c_ma=
ke_random_key
On Sun, Aug 14, 2022 at 22:24 Jeffrey E Altman (jaltman@auristor.com<mailto=
:jaltman@auristor.com>) <jaltman@auristor.com<mailto:jaltman@auristor.com>>=
wrote:
On 8/14/2022 7:01 PM, Ben Huntsman (ben@huntsmans.net<mailto:ben@huntsmans.=
net>) wrote:
Sure, I agree for testing that's fine.
But, there is evidence to suggest that this has compiled successfully on AI=
X 6.1 and 7.2 in the past, and this issue would have occurred there.
Ben,
The change that you referred to as being similar to the build failure on AI=
X is dated Sept 2013.
AIX 6.1 was last updated in November 2013. AIX 7.2 was first released in =
December 2015.
The initial use of krb5_c_make_random_key() was added to OpenAFS in 2007 bu=
t it was for test infrastructure which might not have been built for AIX. =
In the years since additional use of that function has been added to the so=
urce tree. However, it is unclear to me at least whether such changes were=
built on AIX.
The comments on https://gerrit.openafs.org/#/c/14707/3 which was the tip of=
the series that added symbols for AIX 7.2 state that the build on AIX succ=
eeded only if the kernel module was not built and if all linkage was static=
. That change series is a little over a year old.
The last combination that I can find evidence of the AIX kernel module work=
ing was OpenAFS 1.6.5 and AIX 6.1 TL5 which would have been 32-bit. OpenA=
FS 1.6.5 was released in July 2013.
And there was work needed for even that, given the incompatible struct mbuf=
changes that broke things across TL5
I would not expect 7.1 to work without at least some kernel changes.
--
--
Daria Phoebe Brashear
AuriStor, Inc.
dariaphoebe.com<http://dariaphoebe.com>
--_000_MWHPR0701MB3674A3291C36E4A7A58BFA69A7689MWHPR0701MB3674_
Content-Type: text/html; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
<html>
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"=
>
<style type=3D"text/css" style=3D"display:none;"> P {margin-top:0;margin-bo=
ttom:0;} </style>
</head>
<body dir=3D"ltr">
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
So looking into a build on AIX 7.1 was where I was going next, I was just t=
rying to get it working on AIX 6.1 for reference.</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
<br>
</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
There's something significant in AIX 7.1 that doesn't work with OpenAFS, bu=
t which is not also present in AIX 7.2?</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
<br>
</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
Thank you!</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
<br>
</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
-Ben</div>
<div style=3D"font-family: Calibri, Arial, Helvetica, sans-serif; font-size=
: 12pt; color: rgb(0, 0, 0);" class=3D"elementToProof">
<br>
</div>
<div id=3D"appendonsend"></div>
<hr style=3D"display:inline-block;width:98%" tabindex=3D"-1">
<div id=3D"divRplyFwdMsg" dir=3D"ltr"><font face=3D"Calibri, sans-serif" st=
yle=3D"font-size:11pt" color=3D"#000000"><b>From:</b> Daria Phoebe Brashear=
<shadow@gmail.com><br>
<b>Sent:</b> Sunday, August 14, 2022 7:33 PM<br>
<b>To:</b> Jeffrey E Altman (jaltman@auristor.com) <jaltman@auristor.com=
><br>
<b>Cc:</b> Ben Huntsman <ben@huntsmans.net>; Benjamin Kaduk <kaduk=
@mit.edu>; Ken Hornstein <kenh@cmf.nrl.navy.mil>; openafs-devel@op=
enafs.org <openafs-devel@openafs.org><br>
<b>Subject:</b> Re: [OpenAFS-devel] AIX build fails with missing symbol .kr=
b5_c_make_random_key</font>
<div> </div>
</div>
<div>
<div><br>
</div>
<div><br>
<div class=3D"x_gmail_quote">
<div dir=3D"ltr" class=3D"x_gmail_attr">On Sun, Aug 14, 2022 at 22:24 Jeffr=
ey E Altman (<a href=3D"mailto:jaltman@auristor.com">jaltman@auristor.com</=
a>) <<a href=3D"mailto:jaltman@auristor.com">jaltman@auristor.com</a>>=
; wrote:<br>
</div>
<blockquote class=3D"x_gmail_quote" style=3D"margin:0px 0px 0px 0.8ex; bord=
er-left-width:1px; border-left-style:solid; padding-left:1ex; border-left-c=
olor:rgb(204,204,204)">
<div>
<div>On 8/14/2022 7:01 PM, Ben Huntsman (<a href=3D"mailto:ben@huntsmans.ne=
t" target=3D"_blank">ben@huntsmans.net</a>) wrote:<br>
</div>
<blockquote type=3D"cite">
<div style=3D"font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12p=
t; color:rgb(0,0,0)">
Sure, I agree for testing that's fine.</div>
<div style=3D"font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12p=
t; color:rgb(0,0,0)">
<br>
</div>
<div style=3D"font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12p=
t; color:rgb(0,0,0)">
But, there is evidence to suggest that this has compiled successfully on AI=
X 6.1 and 7.2 in the past, and this issue would have occurred there.<br>
</div>
<br>
</blockquote>
<p>Ben,</p>
<p><br>
</p>
<p>The change that you referred to as being similar to the build failure on=
AIX is dated Sept 2013.</p>
<p><br>
</p>
<p>AIX 6.1 was last updated in November 2013. AIX 7.2 was first=
released in December 2015.</p>
<p><br>
</p>
<p>The initial use of krb5_c_make_random_key() was added to OpenAFS in 2007=
but it was for test infrastructure which might not have been built for AIX=
. In the years since additional use of that function has been added t=
o the source tree. However, it is unclear
to me at least whether such changes were built on AIX. <br>
</p>
<p><br>
</p>
<p>The comments on <a href=3D"https://gerrit.openafs.org/#/c/14707/3" targe=
t=3D"_blank">
https://gerrit.openafs.org/#/c/14707/3</a> which was the tip of the series =
that added symbols for AIX 7.2 state that the build on AIX succeeded only i=
f the kernel module was not built and if all linkage was static.  =
; That change series is a little over a year
old.<br>
</p>
<p><br>
</p>
<p>The last combination that I can find evidence of the AIX kernel module w=
orking was OpenAFS 1.6.5 and AIX 6.1 TL5 which would have been 32-bit. =
; OpenAFS 1.6.5 was released in July 2013.</p>
<p></p>
</div>
</blockquote>
<div dir=3D"auto"><br>
</div>
<div dir=3D"auto">And there was work needed for even that, given the incomp=
atible struct mbuf changes that broke things across TL5</div>
<div dir=3D"auto"><br>
</div>
<div dir=3D"auto">I would not expect 7.1 to work without at least some kern=
el changes. </div>
</div>
</div>
-- <br>
<div dir=3D"ltr" class=3D"x_gmail_signature">--<br>
Daria Phoebe Brashear<br>
AuriStor, Inc.<br>
<a href=3D"http://dariaphoebe.com">dariaphoebe.com</a></div>
</div>
</body>
</html>
--_000_MWHPR0701MB3674A3291C36E4A7A58BFA69A7689MWHPR0701MB3674_--