7:14 a.m.
hi,
i tried to build kernel-sourcecode-2.6.8-1.525
from src.rpm with the following command:
rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
but it creates only:
Wrote: /usr/src/redhat/RPMS/i686/kernel-2.6.8-1.525.root.i686.rpm
Wrote: /usr/src/redhat/RPMS/i686/kernel-smp-2.6.8-1.525.root.i686.rpm
any help?
--
mfg. Martin
7:57 a.m.
Am Sa, den 21.08.2004 schrieb David Woodhouse um 14:42:
On Sat, 2004-08-21 at 14:14 +0200, Martin Gansser wrote:
> hi,
>
> i tried to build kernel-sourcecode-2.6.8-1.525
> from src.rpm with the following command:
>
>
> rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
kernel-sourcecode is noarch.
sorry, but this is not a answer of my question.
how do i get the sources from kernel-2.6.8-1.525.src.rpm,
I remeber in the past, if I rebuild with the command
rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
a package kernel-sourcecode-2.6.xxx.noarch.rpm was build.
--
mfg. Martin
8:59 a.m.
On Sat, 2004-08-21 at 08:57, Martin Gansser wrote:
Am Sa, den 21.08.2004 schrieb David Woodhouse um 14:42:
> On Sat, 2004-08-21 at 14:14 +0200, Martin Gansser wrote:
> > hi,
> >
> > i tried to build kernel-sourcecode-2.6.8-1.525
> > from src.rpm with the following command:
> >
> >
> > rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
>
> kernel-sourcecode is noarch.
sorry, but this is not a answer of my question.
how do i get the sources from kernel-2.6.8-1.525.src.rpm,
I remeber in the past, if I rebuild with the command
rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
a package kernel-sourcecode-2.6.xxx.noarch.rpm was build.
Actually, it *was* an answer, at least of sorts. If you read the .spec
file, you'll see that you need to e.g. give --target=noarch to build
kernel-sourcecode. It looks to me that the default %buildsource is 0
unless overridden, and that only happens %ifarch noarch. Corrections are
welcome.
--
Paul W. Frields, RHCE
9:53 a.m.
> sorry, but this is not a answer of my question.
> how do i get the sources from kernel-2.6.8-1.525.src.rpm,
> I remeber in the past, if I rebuild with the command
>
> rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
>
> a package kernel-sourcecode-2.6.xxx.noarch.rpm was build.
Actually, it *was* an answer, at least of sorts. If you read the .spec
file, you'll see that you need to e.g. give --target=noarch to build
kernel-sourcecode. It looks to me that the default %buildsource is 0
unless overridden, and that only happens %ifarch noarch. Corrections are
welcome.
--
Paul W. Frields, RHCE
rpmbuild --rebuild --target=noarch kernel-2.6.8-1.525.src.rpm
builds only:
Wrote:/usr/src/redhat/RPMS/noarch/kernel-doc-2.6.8-1.525.root.noarch.rpm
now i found a solution:
first i exstract the spec file
- rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
- edit the kernel spec file: kernel-2.6.spec
set %define buildsource 0 to %define buildsource 1
rebuild:
rpmbuild -ba --target=noarch kernel-2.6.spec
my last question, why is the name of the rpm
kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
^^^^
is it possible to get a name like this ?
kernel-sourcecode-2.6.8-1.525.noarch.rpm
--
mfg. Martin
1:07 p.m.
On Sat, Aug 21, 2004 at 04:53:09PM +0200, Martin Gansser wrote:
> > sorry, but this is not a answer of my question.
> > how do i get the sources from kernel-2.6.8-1.525.src.rpm,
> > I remeber in the past, if I rebuild with the command
> >
> > rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
> >
> > a package kernel-sourcecode-2.6.xxx.noarch.rpm was build.
>
> Actually, it *was* an answer, at least of sorts. If you read the .spec
> file, you'll see that you need to e.g. give --target=noarch to build
> kernel-sourcecode. It looks to me that the default %buildsource is 0
> unless overridden, and that only happens %ifarch noarch. Corrections are
> welcome.
>
> --
> Paul W. Frields, RHCE
rpmbuild --rebuild --target=noarch kernel-2.6.8-1.525.src.rpm
builds only:
Wrote:/usr/src/redhat/RPMS/noarch/kernel-doc-2.6.8-1.525.root.noarch.rpm
now i found a solution:
first i exstract the spec file
- rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
- edit the kernel spec file: kernel-2.6.spec
set %define buildsource 0 to %define buildsource 1
rebuild:
rpmbuild -ba --target=noarch kernel-2.6.spec
my last question, why is the name of the rpm
kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
^^^^
is it possible to get a name like this ?
kernel-sourcecode-2.6.8-1.525.noarch.rpm
Check the spec file again. From that I learned that a
'touch /etc/beehive-root' before the rpmbuild should do it.
-Marcel
--
======-------- Marcel J.E. Mol MESA Consulting B.V.
=======--------- ph. +31-(0)6-54724868 P.O. Box 112
=======--------- marcel(a)mesa.nl 2630 AC Nootdorp
__==== www.mesa.nl ---____U_n_i_x______I_n_t_e_r_n_e_t____ The Netherlands ____
They couldn't think of a number, Linux user 1148 -- counter.li.org
so they gave me a name! -- Rupert Hine -- www.ruperthine.com
10:08 a.m.
On Sat, 2004-08-21 at 14:14, Martin Gansser wrote:
hi,
i tried to build kernel-sourcecode-2.6.8-1.525
from src.rpm with the following command:
rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
easiest is to install the src.rpm and then to do
rpmbuild -bp --target=i686 kernel-2.6.spec on the thus created specfile.
12:13 p.m.
Am Sa, den 21.08.2004 schrieb Arjan van de Ven um 17:08:
On Sat, 2004-08-21 at 14:14, Martin Gansser wrote:
> hi,
>
> i tried to build kernel-sourcecode-2.6.8-1.525
> from src.rpm with the following command:
>
>
> rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
easiest is to install the src.rpm and then to do
rpmbuild -bp --target=i686 kernel-2.6.spec on the thus created specfile.
i installed the src.rpm
rpm -ivh kernel-2.6.8-1.525.src.rpm
with rpmbuild -bp --target=i686 kernel-2.6.spec
this are the last few message:
...
removed `./init/main.c.orig'
+ find . -name '*~' -exec rm -fv '{}' ';'
+ exit 0
but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
was build.
What is going wrong ?
--
mfg. Martin
12:16 p.m.
On Sat, Aug 21, 2004 at 07:13:57PM +0200, Martin Gansser wrote:
+ find . -name '*~' -exec rm -fv '{}' ';'
+ exit 0
but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
was build.
What is going wrong ?
you now have the sourcecode expanded into /usr/src/redhat/BUILD/kernel-<version>
there is no kernel-sourcecode-2.6.8-1.525 package... why do you need one?
(I assumed you wanted the kernel source... but you didn't say so explicitly)
12:34 p.m.
Am Sa, den 21.08.2004 schrieb Arjan van de Ven um 19:16:
On Sat, Aug 21, 2004 at 07:13:57PM +0200, Martin Gansser wrote:
> + find . -name '*~' -exec rm -fv '{}' ';'
> + exit 0
>
> but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
> was build.
>
> What is going wrong ?
you now have the sourcecode expanded into /usr/src/redhat/BUILD/kernel-<version>
there is no kernel-sourcecode-2.6.8-1.525 package... why do you need one?
to compile my own kernel
(I assumed you wanted the kernel source... but you didn't say so
explicitly)
yes
why does no kernel-source package are distributed ?
the passed I used always the kernel-source package that are installed in
/usr/src/linux-2.6.8-1.xxx to build my own kernel.
--
mfg. Martin
12:36 p.m.
Martin Gansser wrote:
Am Sa, den 21.08.2004 schrieb Arjan van de Ven um 19:16:
>On Sat, Aug 21, 2004 at 07:13:57PM +0200, Martin Gansser wrote:
>
>
>>+ find . -name '*~' -exec rm -fv '{}' ';'
>>+ exit 0
>>
>>but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
>>was build.
>>
>>What is going wrong ?
>>
>>
>you now have the sourcecode expanded into
/usr/src/redhat/BUILD/kernel-<version>
>
>there is no kernel-sourcecode-2.6.8-1.525 package... why do you need one?
>
>
to compile my own kernelbuildprep gets you what you need. Just follow the directions in
the README at the top of he source tree
>(I assumed you wanted the kernel source... but you didn't say so explicitly)
>
>
yes
why does no kernel-source package are distributed ?
because its redundant to the src rpm
Neil
12:51 p.m.
On Sat, 2004-08-21 at 13:34, Martin Gansser wrote:
Am Sa, den 21.08.2004 schrieb Arjan van de Ven um 19:16:
> On Sat, Aug 21, 2004 at 07:13:57PM +0200, Martin Gansser wrote:
> > + find . -name '*~' -exec rm -fv '{}' ';'
> > + exit 0
> >
> > but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
> > was build.
> >
> > What is going wrong ?
>
> you now have the sourcecode expanded into
/usr/src/redhat/BUILD/kernel-<version>
>
> there is no kernel-sourcecode-2.6.8-1.525 package... why do you need one?
to compile my own kernel
[...snip...]
I believe Arjan has been working on converting the build process so that
you'll do this using the .src.rpm at some point in the future, no need
to explicitly d/l, build, or install kernel-sourcecode. At least that's
what I gather from the kernel RPM changelog.
--
Paul W. Frields, RHCE
6:56 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sat, Aug 21, 2004 at 07:16:29PM +0200, Arjan van de Ven wrote:
On Sat, Aug 21, 2004 at 07:13:57PM +0200, Martin Gansser wrote:
> + find . -name '*~' -exec rm -fv '{}' ';'
> + exit 0
>
> but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
> was build.
>
> What is going wrong ?
you now have the sourcecode expanded into /usr/src/redhat/BUILD/kernel-<version>
there is no kernel-sourcecode-2.6.8-1.525 package... why do you need one?
You do seem to have a very short memory ring buffer, this discussion
to which you heavily particpated was held on this list 8 weeks ago,
check the archives:
http://www.google.com/search?q=%22No+more+kernel-source(code)%22&filt...
There were about 300 mails in June about your decision to first rename
the package and then drop it.
There were arguments against your removal of this packages whose
validity remains unchanged still today (nothing has changed after
all).
As well as the fact that this should have been discussed before doing
so, and not start yet another social experiment ...
And if the arguments did not convince you back in June why did you
revive the package back then?
I guess repeating social experiments are non-determinstic as the
objects of your experimentation get weared off and present less
resistance, so you may just have a better chance of passing your thing
this time!
--
Axel.Thimm at ATrpms.net
7:40 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
Am Sonntag, den 22.08.2004, 13:56 +0200 schrieb Axel Thimm:
On Sat, Aug 21, 2004 at 07:16:29PM +0200, Arjan van de Ven wrote:
> On Sat, Aug 21, 2004 at 07:13:57PM +0200, Martin Gansser wrote:
> > + find . -name '*~' -exec rm -fv '{}' ';'
> > + exit 0
> >
> > but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
> > was build.
> >
> > What is going wrong ?
>
> you now have the sourcecode expanded into
/usr/src/redhat/BUILD/kernel-<version>
>
> there is no kernel-sourcecode-2.6.8-1.525 package... why do you need one?
You do seem to have a very short memory ring buffer, this discussion
to which you heavily particpated was held on this list 8 weeks ago,
check the archives:
[...]
Arjan just wanted to know why *he* needed kernel-sourcecode -- maybe
Martin has a real reason why he needs it that could be of interest for
Arjan (even if I don't think so...)?
But to not flame around here for no good reason I'd like to repost my
old proposal. Why not put kernel-sorucecode in fedora extras /
fedora.us. That safes the space on CD/DVD but those people who *think*
they need it can download it from there? I think this would solve most
(all?) problems.
I'm willing to maintain it there myself, but it would be very nice if
this could be done by arjan and/or dave directly. This would avoid the
time-offset when a new kernel arrives.
Just my 2 cent...
--
Thorsten Leemhuis <fedora(a)leemhuis.info>
8:19 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 02:40:18PM +0200, Thorsten Leemhuis wrote:
But to not flame around here for no good reason I'd like to
repost my
old proposal. Why not put kernel-sorucecode in fedora extras /
fedora.us.
Maybe because most repos are compatible to fedora core, but not fedora.us?
That safes the space on CD/DVD but those people who *think*
they need it can download it from there? I think this would solve most
(all?) problems.
I don't think this is a disk space issue, building external kernel
modules (as rpms at least), or even custom kernels get more and more
difficult on fedora. While the framework was never perfect, at least
there was one and the last 8 weeks have seen 3 (!!!) major
modifications to make it worse to unusable. Not counting 4kstacks
option or sys_* interface removal ...
If fedora.us accepts the fact that multiple-repos (co)exist, and would
be willing to work on interoperability/compatibility I would very much
welcome a new kernel module framework outside of Arjan's reach ;)
But as it stands now, no repo outside the fedora.us twin setup can
easily depend on its contents. If something breaks fedora.us by
definition does not care. So currently any fedora.us solution is an
island solution.
BTW I know Ville also wants to see changes in the kernel module issue
in fedora.us.
If Arjan's social experiments turn to make fedora.us open up to other
repos, he will get an honor's PhD in social engineering ;)
--
Axel.Thimm at ATrpms.net
8:31 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
While the framework was never perfect, at least
there was one and the last 8 weeks have seen 3 (!!!) major
modifications to make it worse to unusable. Not counting 4kstacks
option or sys_* interface removal ...
ehhhh dude. 4kstacks is a major desktop latency gain. As well as a
server gain. If you see that as a my imaginary vendetta against external
modules you should go see a doctor. really. Same for removing
deprecated, nearly impossible to use correctly and security prone
interfaces. I don't know what other ghosts you see in your tea leaves
but I suspect you're just imagining things.
If fedora.us accepts the fact that multiple-repos (co)exist, and
would
be willing to work on interoperability/compatibility I would very much
welcome a new kernel module framework outside of Arjan's reach ;)
You know, YOU and 2 or 3 other people are the reason I entirely stopped
caring about this. No matter what I do you or those others will just
flame me and personally attack me. So I rather do just nothing and
ignore the entire thing.
9:24 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 03:31:56PM +0200, Arjan van de Ven wrote:
> While the framework was never perfect, at least
> there was one and the last 8 weeks have seen 3 (!!!) major
> modifications to make it worse to unusable. Not counting 4kstacks
> option or sys_* interface removal ...
ehhhh dude. 4kstacks is a major desktop latency gain. As well as a
server gain.
Thanks for clarifying this, I thought 4KSTACKS was for making it
impossible to run XFS filesystems w/o risiking corruption. Oh, yes, I
forget XFS/reiserfs is not supported.
If you see that as a my imaginary vendetta against external modules
you should go see a doctor. really. Same for removing deprecated,
nearly impossible to use correctly and security prone interfaces. I
don't know what other ghosts you see in your tea leaves but I
suspect you're just imagining things.
Hm, choosing to use CONFIG_4KSTACKS is a wise thing (in fedora
scope). Enforcing your policy by removing _the option_ to go back is
bad politics, especially when at the same time lkml discussed whether
CONFIG_4KSTACKS should depend on CONFIG_BROKEN. What is there to gain
by removing the option? Very wrong educational skills IMHO.
> If fedora.us accepts the fact that multiple-repos (co)exist, and
would
> be willing to work on interoperability/compatibility I would very much
> welcome a new kernel module framework outside of Arjan's reach ;)
You know, YOU and 2 or 3 other people are the reason I entirely
stopped caring about this. No matter what I do you or those others
will just flame me and personally attack me. So I rather do just
nothing and ignore the entire thing.
If you were only to ignore and leave things as are all would be
well. It is your "improvements" we are all so afraid of ...
Speaking of personal attacks, please stop being paranoic, and accept
critisim as a grown up (I was the first and only to thank you for
putting back kernel-sourcecode in the first round of social
experiments). There is no Anti-Arjan conspiracy, people are just
pissed off by the way you introduce breakage.
The critisim in the 8-weeks old debate had been constructive, but you
did not care about the feedback. I recommend reviewing this:
http://www.redhat.com/archives/fedora-devel-list/2004-June/msg01050.html
--
Axel.Thimm at ATrpms.net
12:23 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, 2004-08-22 at 16:24, Axel Thimm wrote:
Thanks for clarifying this, I thought 4KSTACKS was for making it
impossible to run XFS filesystems w/o risiking corruption. Oh, yes, I
forget XFS/reiserfs is not supported.
actually you make XFS-vs-4Kstacks sound really horrible. On lkml there
is some discussion by people who see problems, however in general these
people are using a much older gcc which seems to have more stack "bloat"
than the newer gcc we use. In addition, the regparm optimisation we use
also reduces stack use somewhat.
Enforcing your policy by removing _the option_ to go back is
Andrew Morton asked me for a patch to remove the option at the time 4K
stacks got merged. Ideally there shouldn't be a choice, I mean, there
sholdn't be a kernel config option for every improvement that hits the
kernel, instead there should be the right one chose. The kernel does
have a lot of (non-driver) config options, but almost all of them are
for different tradeoffs (like server-vs-desktop choices like CONFIG_SMP)
bad politics, especially when at the same time lkml discussed
whether
CONFIG_4KSTACKS should depend on CONFIG_BROKEN. What is there to gain
by removing the option? Very wrong educational skills IMHO.
The patch has not gone into mainline yet because people still use the
older gcc's. That's most of what is stopping it right now. However with
the patches we add, non-4kstacks doesnt' seem to work (4g/4g for
example, but there are others like execshield) and I'm not interested in
debugging and fixing all those. It's a small effort for someone who
wants to debug that to back out the patch.
5:31 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
Well, the discussion is not quite on-topic, as the thread was about
shooting off kernel-sourcecode packages w/o any notice once again. But
anyway, you seem to be more interested in justifying 4KSTACKS
enforcement policies, so here we go.
On Sun, Aug 22, 2004 at 07:23:02PM +0200, Arjan van de Ven wrote:
On Sun, 2004-08-22 at 16:24, Axel Thimm wrote:
> Thanks for clarifying this, I thought 4KSTACKS was for making it
> impossible to run XFS filesystems w/o risiking corruption. Oh, yes, I
> forget XFS/reiserfs is not supported.
actually you make XFS-vs-4Kstacks sound really horrible. On lkml there
is some discussion by people who see problems, however in general these
people are using a much older gcc which seems to have more stack "bloat"
than the newer gcc we use. In addition, the regparm optimisation we use
also reduces stack use somewhat.
It doesn't sound horrible, it simply is. Pick FC3t1/i386 and
NFS-export an almost full XFS fs. Then stress it over the network and
you get reproducable corruption. If that is not horrible, what is?
(P.S. x86_64 seems to cope better than i386 in this scenario)
> Enforcing your policy by removing _the option_ to go back is
Andrew Morton asked me for a patch to remove the option at the time 4K
stacks got merged. Ideally there shouldn't be a choice, I mean,
I don't buy that, what benefit do you have to forbid choosing in
custom kernels to use what vanilla kernels still offer today? When
4KSTACKS gets its bugs ironed out, throw the switching code away, but
doing so while 4KSTACKS shows that it is far from mature in the kernel
itself (not talking of external kernel modules, I'm talking of
NFS/XFS), is asking for trouble.
And it really is the wrong way to enforce your policies. You already
have the default kernel for shaking out experimental features (nothing
against that, it's fedora's charter after all), whoever wants to build
a custom kernel has his reasons, don't deprive him of his choices.
> bad politics, especially when at the same time lkml discussed
whether
> CONFIG_4KSTACKS should depend on CONFIG_BROKEN. What is there to gain
> by removing the option? Very wrong educational skills IMHO.
The patch has not gone into mainline yet because people still use the
older gcc's. That's most of what is stopping it right now. However with
the patches we add, non-4kstacks doesnt' seem to work (4g/4g for
example,
That's a very bad example, because it only does not work, because you
yourself have removed the 8KSTACKs compatible hooks from the original
patch, the "empty space" is clearly visible in the patch and easily
rectifiable for the one that knows the original patch. If I were
paranoic I'd see an attempt of sabotaging non-4KSTACKs kernels
here. Here is the respective microsurgery that made 4g4g incompatible
to non-4KSTACKS ...
Original:
#define ARCH_MIN_TASKALIGN 16
+#ifdef CONFIG_4KSTACKS
+#define STACK_PAGE_COUNT (4096/PAGE_SIZE)
+#else
+#define STACK_PAGE_COUNT (8192/PAGE_SIZE)
+#endif
+
struct thread_struct {
Arjan:
#define ARCH_MIN_TASKALIGN 16
+
+#define STACK_PAGE_COUNT (4096/PAGE_SIZE)
+
+
+
+
struct thread_struct {
--
Axel.Thimm at ATrpms.net
1:45 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
Arjan van de Ven wrote:
>If fedora.us accepts the fact that multiple-repos (co)exist, and
would
>be willing to work on interoperability/compatibility I would very much
>welcome a new kernel module framework outside of Arjan's reach ;)
>
You know, YOU and 2 or 3 other people are the reason I entirely stopped
caring about this. No matter what I do you or those others will just
flame me and personally attack me. So I rather do just nothing and
ignore the entire thing.
As a kernel source user, module builder, I'd like to speak in *favor*
of the new kernel src.rpm scheme.
Using mod-versioned kernels, I used to have to rebuild kernel-source
up through the "make dep" stage to get valid kernel headers to compile
against. It was error-prone; If I didn't change the Makefile
EXTRAVERSION, or copy the correct architecture config file to start
with, I got incompatible headers. Building a module for multiple
archs, this was repetitious. OR - I could burn a *lot* of disk space
replicating the kernel source just to build the headers to build
modules against. Yech.
Now the kernel headers are included with each kernel install!
Fantastic! No more special builds. ASCII headers compress really
well, so there's little inflation in binary-kernel rpm size. And it's
much easier to build & install the 3rd party hardware drivers we use.
Now 3rd party delivered build scripts "just work" since the correct
headers can always be found with the kernel. Vendors have gotten
fewer callbacks regarding installation. It no longer depends on the
last state of the /usr/src/linux* since someone last used the
kernel-source. In fact, it removes the kernel-source requirement
entirely.
Better yet, the kernel src.rpm now gives me direct visibility into the
patches that RedHat's applied. kernel-source always used to bug me,
since all I got was the post-patched kernel-source without much info
on how it differed from kernel.org's version. If I just "rpm -i", I
get the virgin kernel source tarball in the ./SOURCES directory. With
"rpmbuild -bp " I get the RedHat patched kernel, equivalent to
kernel-source. Better yet, I can look in kernel.spec and see the
patches applied (with comments) one-by-one. Though the src.rpm had
been available, I'd never really looked at it since most of the
literature (well .. web-searches) steered me towards kernel-source.
Arjan's write-up here: http://fedoranews.org/colin/fnu/issue14.shtml
seemed to cover the history and issues pretty well. It would be good
to get this written up in a HOWTO or some more easily findable docs.
Especially prior to RHEL4, where I'd expect to see some quality docs
on these procedures, delivered as part of RHEL4 documentation.
Good work Arjan.
Cheers, -Bob Arendt
9:51 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sunday 22 August 2004 11:45, Bob Arendt wrote:
Now the kernel headers are included with each kernel install!
Fantastic! No more special builds. ASCII headers compress really
well, so there's little inflation in binary-kernel rpm size. And it's
much easier to build & install the 3rd party hardware drivers we use.
Now 3rd party delivered build scripts "just work" since the correct
headers can always be found with the kernel. Vendors have gotten
fewer callbacks regarding installation. It no longer depends on the
last state of the /usr/src/linux* since someone last used the
kernel-source. In fact, it removes the kernel-source requirement
entirely.
What do I need to tell my vendors then? I constantly have to dig up header
files that aren't in /lib/modules/`uname -r`/build/ such as SCSI headers.
sd.h scsi_mod.h hosts.h etc.... I have to pull them from a sourcecode or
src.rpm in order to build my 3rd party module. Having the matching
sourcecode rpm helped, although I can get around that. I'm more
interested in educating my vendors so that I don't have to dig this stuff
up.
- --
Jesse Keating RHCE (http://geek.j2solutions.net)
Fedora Legacy Team (http://www.fedoralegacy.org)
GPG Public Key (http://geek.j2solutions.net/jkeating.j2solutions.pub)
Was I helpful? Let others know:
http://svcs.affero.net/rm.php?r=jkeating
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
iD8DBQFBKgRm4v2HLvE71NURAhwLAJsFhim8PWYoS5uhXVoluoNNH7quUgCeM7cH
uotY6Kbe8CLp1EPdvO5f5e8=
=mBrt
-----END PGP SIGNATURE-----
10:04 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
What do I need to tell my vendors then? I constantly have to dig up
header
files that aren't in /lib/modules/`uname -r`/build/ such as SCSI headers.
sd.h scsi_mod.h hosts.h etc....
hosts.h must not be used in 2.6....
the other driver headers all are in include/scsi in 2.6
10:08 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Monday 23 August 2004 08:04, Arjan van de Ven wrote:
hosts.h must not be used in 2.6....
the other driver headers all are in include/scsi in 2.6
I'll have to take a look. What header file took the place of hosts.h under
2.6?
- --
Jesse Keating RHCE (http://geek.j2solutions.net)
Fedora Legacy Team (http://www.fedoralegacy.org)
GPG Public Key (http://geek.j2solutions.net/jkeating.j2solutions.pub)
Was I helpful? Let others know:
http://svcs.affero.net/rm.php?r=jkeating
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
iD8DBQFBKghh4v2HLvE71NURAvaLAJ0VI9gv0hZSmeOEAzBB6Due8OxHUwCgi+t1
1k0RdnkpeuGghSdKE4gb2OQ=
=Dp+D
-----END PGP SIGNATURE-----
10:28 a.m.
New subject: scsi headers
On Mon, 2004-08-23 at 17:08, Jesse Keating wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Monday 23 August 2004 08:04, Arjan van de Ven wrote:
> hosts.h must not be used in 2.6....
> the other driver headers all are in include/scsi in 2.6
I'll have to take a look. What header file took the place of hosts.h under
2.6?
#warning "This file is obsolete, please use <scsi/scsi_host.h> instead"
10:34 a.m.
New subject: scsi headers
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Monday 23 August 2004 08:28, Arjan van de Ven wrote:
> On Monday 23 August 2004 08:04, Arjan van de Ven wrote:
> > hosts.h must not be used in 2.6....
> > the other driver headers all are in include/scsi in 2.6
>
> I'll have to take a look. What header file took the place of hosts.h
> under 2.6?
#warning "This file is obsolete, please use <scsi/scsi_host.h> instead"
OH yes, I do remember that warning. However it seems it is not a backwards
compatible change as the compile failed when I included that file instead.
I guess I'll kick it to the vendor and have them fix it instead. Thanks
Arjan.
Now, if it need some files that aren't in include/scsi, where should I
report this?
- --
Jesse Keating RHCE (http://geek.j2solutions.net)
Fedora Legacy Team (http://www.fedoralegacy.org)
GPG Public Key (http://geek.j2solutions.net/jkeating.j2solutions.pub)
Was I helpful? Let others know:
http://svcs.affero.net/rm.php?r=jkeating
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
iD8DBQFBKg6M4v2HLvE71NURAm2XAJ4ktnuGh0X8ictaKTgjCQ1rFGoCSgCfQL4Q
8YsEr7u9LWB4nwwLP2u7t44=
=cRjG
-----END PGP SIGNATURE-----
11:10 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
Jesse Keating wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sunday 22 August 2004 11:45, Bob Arendt wrote:
>Now the kernel headers are included with each kernel install!
>Fantastic! No more special builds. ASCII headers compress really
>well, so there's little inflation in binary-kernel rpm size. And it's
>much easier to build & install the 3rd party hardware drivers we use.
>Now 3rd party delivered build scripts "just work" since the correct
>headers can always be found with the kernel. Vendors have gotten
>fewer callbacks regarding installation. It no longer depends on the
>last state of the /usr/src/linux* since someone last used the
>kernel-source. In fact, it removes the kernel-source requirement
>entirely.
What do I need to tell my vendors then? I constantly have to dig up header
files that aren't in /lib/modules/`uname -r`/build/ such as SCSI headers.
sd.h scsi_mod.h hosts.h etc.... I have to pull them from a sourcecode or
src.rpm in order to build my 3rd party module. Having the matching
sourcecode rpm helped, although I can get around that. I'm more
interested in educating my vendors so that I don't have to dig this stuff
up.
So far, none of the 3rd party vendor's we've worked with have required
the actual kernel source to build. If in fact their driver represents
a modification of an existing driver, it seems in the vendor's best interest
to get their additions accepted upstream, so their product "just works"
without having to build and install a kernel-specific driver. Otherwise
copy (if legal) the source into their source tree so they can build off
the provided headers. Otherwise they're tempting fate - relying on
the implementation of the code rather than just the interface.
Of the files noted, I couldn't find sd.h or scsi_mod.h in the kernel source.
The contents of /usr/src/linux-2.6.8-1.521/drivers/scsi/hosts.h:
#warning "This file is obsolete, please use <scsi/scsi_host.h>
instead"
#include <scsi/scsi_host.h>
.. and there is a /lib/modules/2.6.8-1.521/build/include/scsi/scsi_host.h
Perhaps the restructuring of the 2.6 kernel has made it more practical to
use the build/include headers api. Your vendor should hopefully be able
to adapt to this.
RedHat 9 did have some kernel interface headaches for our vendors.
The VM interface calling structure changed (due to 2.5 backports?) late
in it's life. Despite the 2.4.20 kernel version, some of the VM calls
didn't have the same call interface as the vanilla kernel or other distro's
with the same kernel version number. They threw in a manual hack - basically
"if the build errors out, assert this flag and build again". Hopefully
Fedora's "close to upstream" minimal patching stance avoids this.
Cheers,
-Bob Arendt
3:09 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
rda(a)rincon.com (Bob Arendt) writes:
Now the kernel headers are included with each kernel install!
Fantastic!
Fantastic?? That's a really bad idea and I do not see the sense behind
this step. It is wasted space (30MB and 6500 inodes per kernel on the
*root* partition) and it is impossible to prevent their installation. And,
to build new kernel-modules for another kernel I would have to install the
entire new kernel although headers would be enough.
I would appreciate it when the headers would be in a separate rpm or at
least, when they would be under /usr/src where I can do %_netsharedpath
magic.
Enrico
10:03 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ...
Enrico Scholz wrote:
Bob Arendt writes:
>Now the kernel headers are included with each kernel install! Fantastic!
Fantastic?? That's a really bad idea and I do not see the sense behind
this step. It is wasted space (30MB and 6500 inodes per kernel on the
*root* partition) and it is impossible to prevent their installation. And,
to build new kernel-modules for another kernel I would have to install the
entire new kernel although headers would be enough.
I would appreciate it when the headers would be in a separate rpm or at
least, when they would be under /usr/src where I can do %_netsharedpath
magic.
Good point. On our servers we'd always build/install extra drivers - 30M of
accurate headers was certainly better than a semi-built full-source tree.
But if you don't need it, making it optional (like debuginfo packages)
seems a better way to serve everyone's needs.
8:42 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
* social experiments are dangerous, you risk pissing off end users
and/or contributors and eventually reduce your user/contributor base
(people vote against evil social experiments with their feet: "el último
que se vaya que apague la luz...")
* social experiments are good if there is a good reason (technical or
other), so it is explained and everything is OK i.e. the change from
kernel-source-n.n-n.arch.rpm -> kernel-sourcecode-n.n-n.noarch.rpm
** kernel-sourcecode (or whatever you call it) _is_ necessary to build
drivers from other developers or vendors (madwifi, hostapp, intel,
etc...) and it is explicitly required so in documentation and howtos;
thus, this social experiment is _not_ good, unless we pretend to bury
Linux in its own 'ghetto' making _more_ difficult for third parties to
contribute and reducing the range of compatible devices...
** size is not an argument against because there are _enoumous_
'debuginfo' packages and they are rebuilt and posted regularly.
** "you can do it yourself" is neither because with this same argument
taken to the limit no distribution is necessary.
* _this_ social experiment has been previously discussed so why it comes
again? is Fedora open and community driven? or is Fedora a RedHat social
experiment?
--
Manuel Moreno <manolo(a)miconexion.com>
10:08 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 02:42:00PM +0100, Manuel Moreno wrote:
** kernel-sourcecode (or whatever you call it) _is_ necessary to
build
drivers from other developers or vendors (madwifi, hostapp, intel,
Umm nope. Arjan is right here. The end user built modules don't need and
actually can't safely use kernel-sourcecode. BTW one thing you might
want to look at in this area is DKMS which is perhaps the current best
of breed management tool for this problem space.
** size is not an argument against because there are _enoumous_
'debuginfo' packages and they are rebuilt and posted regularly.
Not on the CD images
Alan
11:04 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, 22 Aug 2004 11:08:19 -0400, Alan Cox <alan(a)redhat.com> wrote:
Umm nope. Arjan is right here. The end user built modules don't
need and
actually can't safely use kernel-sourcecode. BTW one thing you might
want to look at in this area is DKMS which is perhaps the current best
of breed management tool for this problem space.
I'd really really like it if someone inside the fence line could find
(or be allocated) the time to sit down and think through DKMS and see
if it makes sense to provide DKMS in core, as the preferred solution
to some of the add-on kernel-module problems. I don't know if if DKMS
can solve all of the add-on module issues, I'm no expert, but I'd have
to agree that right now DKMS looks like a best path forward to me.
-jef
11:17 a.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 12:04:44PM -0400, Jeff Spaleta wrote:
I'd really really like it if someone inside the fence line could
find
(or be allocated) the time to sit down and think through DKMS and see
if it makes sense to provide DKMS in core, as the preferred solution
Actually thats already happening.
to some of the add-on kernel-module problems. I don't know if if
DKMS
can solve all of the add-on module issues, I'm no expert, but I'd have
to agree that right now DKMS looks like a best path forward to me.
DKMS doesn't seem to solve a lot of problems that packaging the stuff up
in rpms nicely does. It provides a way to rebuild modules nicely and it
looks like a good tool for administering that bit.
I'm not sure it needs to be in core for Fedora since add on drivers can
requires: dkms it if appropriate but I'd like to hear views on the Fedora
side.
1:53 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, 2004-08-22 at 11:08 -0400, Alan Cox wrote:
...
want to look at in this area is DKMS which is perhaps the current
best
of breed management tool for this problem space.
We do not have DKMS yet. :-)
Something is needed in the meanwhile, though; maybe not all the source
code but a trimmed down version including .config, kernel headers,
config headers,and also those two sources used for scsi blablabla, then
that other project fails so we need now...
WAIT! why to put effort in a dead horse? leave things as they are and
focus our efforts in building a good DKMS.
> ** size is not an argument against because there are _enoumous_
> 'debuginfo' packages and they are rebuilt and posted regularly.
Not on the CD images
Agreed.
Alan
--
Manuel Moreno <manolo(a)miconexion.com>
12:47 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 02:42:00PM +0100, Manuel Moreno wrote:
** kernel-sourcecode (or whatever you call it) _is_ necessary to
build
drivers from other developers or vendors (madwifi, hostapp, intel,
At least hostap doesn't need kernel-sourcecode.
$ rpmbuild --rebuild --target i686 --define 'kernel 2.6.8-1.521'
kernel-module-hostap-0.2.1-0.fdr.2.src.rpm
Installing kernel-module-hostap-0.2.1-0.fdr.2.src.rpm
Building target platforms: i686
Building for target i686
[...]
Wrote:
/home/steve/src/redhat/RPMS/i686/kernel-module-hostap-2.6.8-1.521-0.2.1-0.fdr.2.i686.rpm
Wrote:
/home/steve/src/redhat/RPMS/i686/kernel-module-hostap-debuginfo-0.2.1-0.fdr.2.i686.rpm
[...]
$ rpm -q kernel-sourcecode
package kernel-sourcecode is not installed
I was also just able to build the zaptel modules on the same system
with no problems. I wouldn't be at all surprised if the other modules
you mentioned build fine without "real" kernel source as well.
The only real problem here is how to handle the few of us who have to
build modules for all of the released kernels, since i386 and i686
versions of the same kernel conflict.
Steve
--
Steven Pritchard - K&S Pritchard Enterprises, Inc.
Email: steve(a)kspei.com http://www.kspei.com/
Phone: (618)398-7360 Mobile: (618)567-7320
1:25 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 12:47:25PM -0500, Steven Pritchard wrote:
On Sun, Aug 22, 2004 at 02:42:00PM +0100, Manuel Moreno wrote:
> ** kernel-sourcecode (or whatever you call it) _is_ necessary to build
> drivers from other developers or vendors (madwifi, hostapp, intel,
I was also just able to build the zaptel modules on the same system
with no problems. I wouldn't be at all surprised if the other modules
you mentioned build fine without "real" kernel source as well.
Well,
he's right about intel (ipw2100 and 2200). Those actually require
drivers/net/wireless/ieee802_11.h. Of course, downloading the entire kernel
sourcecode for this 2.5k header file is silly, a better way would be
to get the header moved to include/linux, include/net or whatnot.
Even the driver Makefile says
# We have to add drivers/net/wireless until ieee802_11.h is in the default
# include path
EXTRA_CFLAGS += -I$(TOPDIR)/drivers/net/wireless
--
Pekka Pietikainen
1:38 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
# We have to add drivers/net/wireless until ieee802_11.h is in the default
# include path
EXTRA_CFLAGS += -I$(TOPDIR)/drivers/net/wireless
well there is nothing in that header that describes a kernel interface
so while it's nice to share, sharing isn't quite mandatory for external
drivers.
1:39 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, 2004-08-22 at 12:47 -0500, Steven Pritchard wrote:
On Sun, Aug 22, 2004 at 02:42:00PM +0100, Manuel Moreno wrote:
> ** kernel-sourcecode (or whatever you call it) _is_ necessary to build
> drivers from other developers or vendors (madwifi, hostapp, intel,
At least hostap doesn't need kernel-sourcecode.
...
well,
http://hostap.epitest.fi/releases/hostap-driver-0.2.4.tar.gz
[root kernel]# tar xzf hostap-driver-0.2.4.tar.gz
[root kernel]# cd hostap-driver-0.2.4
[root hostap-driver-0.2.4]# make
Makefile:24: /usr/src/linux/.config: No such file or directory
Makefile:42: WARNING: No kernel PCMCIA support found and PCMCIA_PATH is
not defined
Makefile:49: WARNING: Linux wireless extensions, CONFIG_NET_RADIO, not
enabled in the kernel
make: *** No rule to make target `/usr/src/linux/.config'. Stop.
[root hostap-driver-0.2.4]#
'PLONK'
Steve
--
Steven Pritchard - K&S Pritchard Enterprises, Inc.
Email: steve(a)kspei.com http://www.kspei.com/
Phone: (618)398-7360 Mobile: (618)567-7320
Manuel.
--
Manuel Moreno <manolo(a)miconexion.com>
1:48 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
well,
http://hostap.epitest.fi/releases/hostap-driver-0.2.4.tar.gz
[root kernel]# tar xzf hostap-driver-0.2.4.tar.gz
[root kernel]# cd hostap-driver-0.2.4
[root hostap-driver-0.2.4]# make
Makefile:24: /usr/src/linux/.config: No such file or directory
Makefile:42: WARNING: No kernel PCMCIA support found and PCMCIA_PATH is
not defined
Makefile:49: WARNING: Linux wireless extensions, CONFIG_NET_RADIO, not
enabled in the kernel
make: *** No rule to make target `/usr/src/linux/.config'. Stop.
[root hostap-driver-0.2.4]#
'PLONK'
that's looks like one broken config file (at least for 2.6 kernels)
1:58 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, 2004-08-22 at 20:48 +0200, Arjan van de Ven wrote:
...
> 'PLONK'
that's looks like one broken config file (at least for 2.6 kernels)
There are so many broken things!
Sigh :-)
--
Manuel Moreno <manolo(a)miconexion.com>
2:11 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 08:48:36PM +0200, Arjan van de Ven wrote:
that's looks like one broken config file (at least for 2.6
kernels)
It is. That's why my rpm
(https://bugzilla.fedora.us/show_bug.cgi?id=1019) includes the
following patch:
--- Makefile.orig 2003-08-03 13:50:09.000000000 -0500
+++ Makefile 2003-10-30 17:25:32.000000000 -0600
@@ -17,7 +17,9 @@
CC=gcc
CFLAGS=-O2 -D__KERNEL__ -DMODULE -Wall -g -c $(EXTRA_CFLAGS)
-include $(KERNEL_PATH)/.config
+KERNEL_CONFIG=$(KERNEL_PATH)/.config
+
+include $(KERNEL_CONFIG)
INCLUDES=-I$(KERNEL_PATH)/include
ifdef PCMCIA_PATH
Then you just need to define KERNEL_* correctly.
I just submitted a similar patch for zaptel (again, last time was in
April of last year).
Broken Makefiles don't change the facts... The software builds fine
without the rest of the kernel source.
Steve
--
Steven Pritchard - K&S Pritchard Enterprises, Inc.
Email: steve(a)kspei.com http://www.kspei.com/
Phone: (618)398-7360 Mobile: (618)567-7320
2:18 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, 2004-08-22 at 21:11, Steven Pritchard wrote:
On Sun, Aug 22, 2004 at 08:48:36PM +0200, Arjan van de Ven wrote:
> that's looks like one broken config file (at least for 2.6 kernels)
It is. That's why my rpm
(https://bugzilla.fedora.us/show_bug.cgi?id=1019) includes the
following patch:
--- Makefile.orig 2003-08-03 13:50:09.000000000 -0500
+++ Makefile 2003-10-30 17:25:32.000000000 -0600
@@ -17,7 +17,9 @@
CC=gcc
CFLAGS=-O2 -D__KERNEL__ -DMODULE -Wall -g -c $(EXTRA_CFLAGS)
-include $(KERNEL_PATH)/.config
+KERNEL_CONFIG=$(KERNEL_PATH)/.config
+
+include $(KERNEL_CONFIG)
INCLUDES=-I$(KERNEL_PATH)/include
ifdef PCMCIA_PATH
Then you just need to define KERNEL_* correctly.
it's still broken; it manually tries to find CFLAGS, it really should
use kbuild infrastructure to find those...
2:37 p.m.
New subject: [rawhide social experiments] kernel-source(code) package removed once again ... (was: How to build kernel-sourcecode-2.6.8-1.525)
On Sun, Aug 22, 2004 at 09:18:11PM +0200, Arjan van de Ven wrote:
it's still broken; it manually tries to find CFLAGS, it really
should
use kbuild infrastructure to find those...
True, but that trivial little patch is enough to make it build and
work properly *without* kernel-sourcecode anywhere to be found...
Steve
--
Steven Pritchard - K&S Pritchard Enterprises, Inc.
Email: steve(a)kspei.com http://www.kspei.com/
Phone: (618)398-7360 Mobile: (618)567-7320
1:54 p.m.
New subject: PHP 5 RPM Package
Hi Everybody !
I would like to install PHP 5 on my Fedora Core 2. I can compile the
source but I prefer to use RPM if there is an available package.
Somebody knows a URL ? If not, I will try to build one... Do you have
recommandations ?
Greetings,
Gilles
12:49 p.m.
On Sat, 2004-08-21 at 13:13, Martin Gansser wrote:
> > i tried to build kernel-sourcecode-2.6.8-1.525
> > from src.rpm with the following command:
> >
> > rpmbuild --rebuild --target=i686-linux kernel-2.6.8-1.525.src.rpm
>
> easiest is to install the src.rpm and then to do
>
> rpmbuild -bp --target=i686 kernel-2.6.spec on the thus created specfile.
>
i installed the src.rpm
rpm -ivh kernel-2.6.8-1.525.src.rpm
with rpmbuild -bp --target=i686 kernel-2.6.spec
this are the last few message:
...
removed `./init/main.c.orig'
+ find . -name '*~' -exec rm -fv '{}' ';'
+ exit 0
but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
was build.
Do:
rpm -ivh kernel-2.6.8-1.525.src.rpm
cd `rpm --eval "%_topdir"`
rpmbuild -bb --target=noarch SPECS/kernel-2.6.spec
--
Paul W. Frields, RHCE
1:05 p.m.
On Sat, 2004-08-21 at 13:49, Paul W. Frields wrote:
> i installed the src.rpm
> rpm -ivh kernel-2.6.8-1.525.src.rpm
>
> with rpmbuild -bp --target=i686 kernel-2.6.spec
> this are the last few message:
> ...
> removed `./init/main.c.orig'
> + find . -name '*~' -exec rm -fv '{}' ';'
> + exit 0
>
> but no kernel-sourcecode-2.6.8-1.525.root.noarch.rpm
> was build.
Do:
rpm -ivh kernel-2.6.8-1.525.src.rpm
cd `rpm --eval "%_topdir"`
rpmbuild -bb --target=noarch SPECS/kernel-2.6.spec
Oops, forgot I was using 2.6.8-1.521, so this might have changed. :-|
--
Paul W. Frields, RHCE
9:22 p.m.
>>>> "pwf" == Paul W Frields
<paul(a)frields.com> writes:
> Do:
>
> rpm -ivh kernel-2.6.8-1.525.src.rpm
> cd `rpm --eval "%_topdir"`
> rpmbuild -bb --target=noarch SPECS/kernel-2.6.spec
pwf>
pwf> Oops, forgot I was using 2.6.8-1.521, so this might have changed. :-|
Indeed.
bach:root# cd `rpm --eval "%_topdir"`
bach:redhat# rpmbuild -bb --target=noarch SPECS/kernel-2.6.spec
Building target platforms: noarch
Building for target noarch
error: Failed build dependencies:
kernel-utils >= 1:2.4-12.1.139 is needed by kernel-2.6.8-1.525.root
Anyone know where to get kernel-utils-2.4-12.1.139.fedora.i386.rpm?
--
Vladimir G. Ivanovic http://leonora.org/~vladimir
Palo Alto, CA 94306 +1 650 678 8014
9:38 p.m.
error: Failed build dependencies:
kernel-utils >= 1:2.4-12.1.139 is needed by kernel-2.6.8-1.525.root
Anyone know where to get kernel-utils-2.4-12.1.139.fedora.i386.rpm?
http://mirrors.kernel.org/fedora/core/development/i386/Fedora/RPMS/
-Steve Grubb
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
11:07 p.m.
>>>> "sg" == Steve G <Steve> writes:
> Anyone know where to get
kernel-utils-2.4-12.1.139.fedora.i386.rpm?
sg>
sg> http://mirrors.kernel.org/fedora/core/development/i386/Fedora/RPMS/
Thanks. I checked Arjan's download area, but he had only 2.4-9, and RPM
Search was not helpful either (it turns out I used a faulty searching
technique).
--
Vladimir G. Ivanovic http://leonora.org/~vladimir
Palo Alto, CA 94306 +1 650 678 8014
12:37 a.m.
>>>> "sg" == Steve G <Steve> writes:
sg> http://mirrors.kernel.org/fedora/core/development/i386/Fedora/RPMS/
Is it possible to get apt-get to check these directories automatically?
I think I can persuade yum to.
--
Vladimir G. Ivanovic http://leonora.org/~vladimir
Palo Alto, CA 94306 +1 650 678 8014
7192
days inactive
7195
days old
50 comments
19 participants
participants (19)
-
Alan Cox -
Arjan van de Ven -
Axel Thimm -
Bob Arendt -
David Woodhouse -
Enrico Scholz -
Gilles Fabio -
Jef Spaleta -
Jesse Keating -
Manuel Moreno -
Marcel J.E. Mol -
Martin Gansser -
Neil Horman -
Paul W. Frields -
Pekka Pietikäinen -
Steve G -
Steven Pritchard -
Thorsten Leemhuis -
Vladimir G. Ivanovic