Re: [lfs-support] Configuration files -/etc/bashrc, /etc/dircolors, or /etc/profile in LFS-version7.10-systemd

2017-07-04 Thread Ken Moffat 
On Tue, Jul 04, 2017 at 09:27:08PM -0400, Dr. George E. Moore wrote:
> I've successfully built LFS version 7.10-systemd. In section 9.3 it
> indicate that a review of the configuration files is appropriate. It
> references /etc/bashrc, /etc/dircolors and /etc/profile. yet at no time
> in the book are these files configured. Is this an oversight, i.e.,
> should there be configuration instructions for these file as well?
> 
> Once the system is booted, I'm able to "useradd" yet the HOME directory,
> though indicated in the /etc/passwd is not the directory for the user at
> login.

In the (current, development) book for sysv, /etc/profile is covered
in 7.7.  I agree that it doesn't seem to be covered in the systemd
book.  Whether it should be, I have no idea.

For the others, I don't think we've created them in a long time.
But they are the sort of thing that people used to expect in the
distant past.  Maybe the text has just hung on, or maybe I'm missing
something.

But for what sounds like a problem with useradd, you need to provide
more details.  I haven't needed to add real users for many years, I
just copy the details (and share /home across builds), so I can't
help.

ĸen
-- 
I live in a city. I know sparrows from starlings.  After that
everything is a duck as far as I'm concerned.  -- Monstrous Regiment
-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

Re: [lfs-support] Configuration files -/etc/bashrc, /etc/dircolors, or /etc/profile in LFS-version7.10-systemd

2017-07-04 Thread Bruce Dubbs 

Dr. George E. Moore wrote:

I've successfully built LFS version 7.10-systemd. In section 9.3 it
indicate that a review of the configuration files is appropriate. It
references /etc/bashrc, /etc/dircolors and /etc/profile. yet at no time
in the book are these files configured. Is this an oversight, i.e.,
should there be configuration instructions for these file as well?


See http://www.linuxfromscratch.org/blfs/view/systemd/postlfs/profile.html

Perhaps we should put that link on the Rebooting page.


Once the system is booted, I'm able to "useradd" yet the HOME directory,
though indicated in the /etc/passwd is not the directory for the user at
login.


When you used useradd, did you use the -m option ?

  -- Bruce


--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

Re: [lfs-support] My finished LFS-7.10 x86-64 build

2017-07-04 Thread Michael Shell 
On Tue, 04 Jul 2017 14:47:16 -0700
Paul Rogers  wrote:

> I do find it ironic the kernel has to do these strange math calculations
> no other packge uses, except apparently gmp.


Not the kernel code per say, but rather gcc while compiling a kernel.
And gcc in turn calls on gmp to do those calculations (for gcc).

Kernel code tends to dictate/control things at the lower levels 
(e.g., directly inserted assembly code, variables held in specific
CPU registers, "volatile" memory regions, etc.) more than is typical
in normal applications. Perhaps these kinds of special compile
requirements things cause gcc to lean more heavily on gmp in certain
areas thus triggering the problem.


  Cheers,

  Mike Shell

-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

[lfs-support] Configuration files -/etc/bashrc, /etc/dircolors, or /etc/profile in LFS-version7.10-systemd

2017-07-04 Thread Dr. George E. Moore 
I've successfully built LFS version 7.10-systemd. In section 9.3 it
indicate that a review of the configuration files is appropriate. It
references /etc/bashrc, /etc/dircolors and /etc/profile. yet at no time
in the book are these files configured. Is this an oversight, i.e.,
should there be configuration instructions for these file as well?

Once the system is booted, I'm able to "useradd" yet the HOME directory,
though indicated in the /etc/passwd is not the directory for the user at
login.
-- 
Dr. George E. Moore Ph.D

-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

Re: [lfs-support] My finished LFS-7.10 x86-64 build

2017-07-04 Thread Paul Rogers 
Turns out my solution way easier than the path I started on--my brain
often does that to me, coming up with the better solution, but tardy.

I went back to the original finished build on the i7-940, my "Iron
Horse" candidate.  (Apologies if I misunderstood you about that.)  I
added that --build parm to the configure in my ch6-11-gmp-6.1.1 script,
calling the result ch6-11-gmp-6.1.1a, with all my normal "pio" package
management.  That compiled fine on the i7, and one of my tools collected
all the pieces in an "update tarball".  Carried that to the Conroe and
applied it.  Then the kernel recompile went smooth--no reboot required,
as you thought.  So now I have two as-builts, 6.1.1 for an i7 and 6.1.1a
for generic CPUs.  Don't think I want the first one for this next
iteration, "Brass Monkey" candidate.


Thanks, Bruce.

-- 
Paul Rogers
paulgrog...@fastmail.fm
Rogers' Second Law: "Everything you do communicates."
(I do not personally endorse any additions after this line. TANSTAAFL
:-)
-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

Re: [lfs-support] My finished LFS-7.10 x86-64 build

2017-07-04 Thread Paul Rogers 
> > So what fix would you recommend?
> 
> Try just building gmp (and possibly mpfr, mpc, and gcc) in the local 
> system without rebooting.  gmp just installs libraries so a reboot would 
> not be needed.  After installing gmp, retry the kernel.  If it doesn't 
> work, try the next package, etc, but I agree that gmp is the most likely 
> culprit.
> 
>-- Bruce

Thanks for suggesting/confirming I don't have to rebuild everything to
make sure there are no other buried Invalid Ops!  I do find it ironic
the kernel has to do these strange math calculations no other packge
uses, except apparently gmp.

Tried recompiling gmp on the Conroe this morning but bombed with more
Invalid Ops.  Now I'm back on the i7 using a "virtual partition" (a file
containing a filesystem mounted "-o loop") to try to recreate a post-Ch5
environment and run off my 11 scripts in Ch6 through gcc.  I found one
left over that was saved after linux-headers, trying that, but glibc is
bombing because ld can't find crt1.o & crtn.o.  I've seen that coming
and going before, but never nailed down what causes it.  Maybe I need
to--going three steps away from the job I should be doing.
-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

Re: [lfs-support] /bin/sh: /dev/null: Permission denied

2017-07-04 Thread Ken Moffat 
On Mon, Jul 03, 2017 at 11:26:09PM -0400, Isaac D. Cohen wrote:
>  On Mon, 03 Jul 2017 22:09:38 -0400 Ken 
> Moffat wrote  
> And you bound the host's /dev in 6.2.2, didn't you ?
> 
>  Oh, that must be it. I did section 6.2.2 but then I shut down
>  my computer and later resumed. I'm sorry. I should have known that
>  would be a problem. It doesn't preserve mounted stuff for the next terminal 
> session.
>  Anyway, I did that section again and it works now!
>  Thank you very much!
> 

You're welcome, but the book does cover it: a 'Note' at the bottom
of 6.4 (Entering the Chroot Environment).

> And since you didn't comment on my previous replies
> that /mnt/lfs should be its own filesystem, I still doubt that you
> will be able to boot your new system when you complete it. I hope
> I'm wrong on that.
> 
>  I'm building LFS in its own filesystem. For some reason my distro
>  mounts filesystems in media/ hence /media/isaac/LFS.
> 
> Isaac
> 

Ah, your distro must have enabled automounting (one of my pet
hates).  For the next time (if you build LFS again from such a
distro), when you become root to set things up at the start, if the
filesystem is already mounted you can just
 mkdir -pv /mnt/lfs
and
 mount --bind /media/username/whatever /mnt/lfs

bind mounting is very useful ;)  I sometimes use it when I want to
build in a directory to test/debug possible changes in my LFS
buildscripts.

ĸen
-- 
I live in a city. I know sparrows from starlings.  After that
everything is a duck as far as I'm concerned.  -- Monstrous Regiment
-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

Re: [lfs-support] My finished LFS-7.10 x86-64 build

2017-07-04 Thread Bruce Dubbs 

Paul Rogers wrote:

I built it on my i7-940.  When finished, with both linux-4.7.2 and
4.9.30, and the rough edges knocked off, e.g. networking, X, Firefox
working, I updated my "clone" script and cloned it to my i7-870.  That
required a bit more work to get a clean clone from the book's "setup"
bits, e.g. useradds, the as-built binaries, and the book's final config
bits, without any contamination from stuff I did in building.  The
system seemed to work well on the 870 too,

So to verify that those fixes to the clone script were "clean", tonight
I cloned it again to one of my Conroes.  The "base" kernels are just
configured to boot on a generic box, primarily, so one of the first
things I do is rebuild the kernel customized for this specific box.  Not
good.




   CC  net/ipv4/raw.o
drivers/gpu/drm/i2c/ch7006_mode.c:51:3: internal compiler error: Illegal
instruction
NTSC_LIKE_TIMINGS,
^



Is this gmp's doing?  Do we STILL have to deal with this misbehavior?  I
didn't use a --TARGET in its configure.  Probably an oversight on my
part, since I do want this to work on any x86-64 I might have to throw
it on.

I suppose I could  get this system in a chroot and recompile just gmp,
if that's all it'd take.  Rebuilding it all from Ch6 again would take a
couple weeks.

So what fix would you recommend?


Try just building gmp (and possibly mpfr, mpc, and gcc) in the local 
system without rebooting.  gmp just installs libraries so a reboot would 
not be needed.  After installing gmp, retry the kernel.  If it doesn't 
work, try the next package, etc, but I agree that gmp is the most likely 
culprit.


  -- Bruce

--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style

[lfs-support] My finished LFS-7.10 x86-64 build

2017-07-04 Thread Paul Rogers 
I built it on my i7-940.  When finished, with both linux-4.7.2 and
4.9.30, and the rough edges knocked off, e.g. networking, X, Firefox
working, I updated my "clone" script and cloned it to my i7-870.  That
required a bit more work to get a clean clone from the book's "setup"
bits, e.g. useradds, the as-built binaries, and the book's final config
bits, without any contamination from stuff I did in building.  The
system seemed to work well on the 870 too,

So to verify that those fixes to the clone script were "clean", tonight
I cloned it again to one of my Conroes.  The "base" kernels are just
configured to boot on a generic box, primarily, so one of the first
things I do is rebuild the kernel customized for this specific box.  Not
good.

  LD  drivers/gpu/drm/i2c/built-in.o
  CC [M]  drivers/gpu/drm/i2c/ch7006_drv.o
  CC  net/ipv4/datagram.o
  CC [M]  drivers/gpu/drm/i2c/ch7006_mode.o
  CC  net/ipv4/raw.o
drivers/gpu/drm/i2c/ch7006_mode.c:51:3: internal compiler error: Illegal
instruction
   NTSC_LIKE_TIMINGS,
   ^
  CC  net/ipv4/udp.o
0x9788bf crash_signal
../../gcc/toplev.c:333
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See  for instructions.
make[4]: *** [scripts/Makefile.build:293:
drivers/gpu/drm/i2c/ch7006_mode.o] Error 1
make[3]: *** [scripts/Makefile.build:544: drivers/gpu/drm/i2c] Error 2
make[2]: *** [scripts/Makefile.build:544: drivers/gpu/drm] Error 2
make[1]: *** [scripts/Makefile.build:544: drivers/gpu] Error 2
make: *** [Makefile:988: drivers] Error 2
make: *** Waiting for unfinished jobs
  CC  net/ipv4/udplite.o
  CC  net/ipv4/udp_offload.o

Is this gmp's doing?  Do we STILL have to deal with this misbehavior?  I
didn't use a --TARGET in its configure.  Probably an oversight on my
part, since I do want this to work on any x86-64 I might have to throw
it on.

I suppose I could  get this system in a chroot and recompile just gmp,
if that's all it'd take.  Rebuilding it all from Ch6 again would take a
couple weeks.

So what fix would you recommend?
-- 
Paul Rogers
paulgrog...@fastmail.fm
Rogers' Second Law: "Everything you do communicates."
(I do not personally endorse any additions after this line. TANSTAAFL
:-)
-- 
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Do not top post on this list.

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

http://en.wikipedia.org/wiki/Posting_style