Re: [gentoo-user] How to test package install?

[Alan McKinnon]

On Thu, Feb 15, 2018 at 4:37 AM, Ian Zimmerman  wrote:

> I'm trying to test my package by running "ebuild /path/to/pkg.ebuild
> install".  Naturally (for me) I do this as an unprivileged user, not as
> root.  It fails because at least some steps such as dobin need to give
> away ownership of the files being installed.  I tried to run the whole
> thing including compilation under fakeroot but that doesn't help.
>
> If it is relevant (but I don't think it is) my user _is_ in the portage
> group.
>
> What is the accepted or usual way to do this task?
>
> --
> Please don't Cc: me privately on mailing lists and Usenet,
> if you also post the followup to the list or newsgroup.
> To reply privately _only_ on Usenet and on broken lists
> which rewrite From, fetch the TXT record for no-use.mooo.com.
>
>
You can normally build as yourself. Install requires root typically due to
write permissions on *bin and /etc.

You can probably come up with a clever way to do this, but by far the
easiest is the classic:

sudo ebuild /path/to/pkg.ebuild install

-- 
Alan McKinnon
alan dot mckinnon at gmail dot com

[gentoo-user] nvidia driver and genkernel above 4.9 issues

[András Csányi]

Hi All,

I have a fresh gentoo install and it works fine with 4.9 kernel,
nvidia video card does its job like a charm. However, I compiled a
4.15.3 kernel with the same setup, xorg-drivers recompiled,
nvidia-drivers recompiled, eselect kernel set shows to the new kernel
and I have strange behavior:

During boot I see a message which says nvidia.ko cannot be loaded due
to some signature issue. Checking the dmesg under windows I cannot see
this error message.
After successful boot obviously xdm do not start, but the whole screen
vibrates/blinking fast.

I haven't boot from live-cd and chroot and recompile xorg-drivers and
nvidia-drivers yet.

I saw a few article about that this might be due to Spectre/Meltdown,
but I'm not skilled in kernel things.

Any suggestion what I can do?

Thanks any help in advance!

in dmesg I see only this:

[3.194322] nvidia: loading out-of-tree module taints kernel.
[3.194326] nvidia: module license 'NVIDIA' taints kernel.
[3.194326] Disabling lock debugging due to kernel taint
[3.196697] nvidia: module verification failed: signature and/or
required key missing - tainting kernel
[3.200183] nvidia-nvlink: Nvlink Core is being initialized, major
device number 240
...
[4.477269] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[4.477270] NVRM: This can occur when a driver such as:
   NVRM: nouveau, rivafb, nvidiafb or rivatv
   NVRM: was loaded and obtained ownership of the NVIDIA device(s).
[4.477270] NVRM: Try unloading the conflicting kernel module (and/or
   NVRM: reconfigure your kernel without the conflicting
   NVRM: driver(s)), then try loading the NVIDIA kernel module
   NVRM: again.
[4.477271] NVRM: No NVIDIA graphics adapter probed!
[4.477368] nvidia-nvlink: Unregistered the Nvlink Core, major
device number 240
[4.499175] nvidia_modeset: Unknown symbol nv_kthread_q_flush (err 0)
[4.499201] nvidia_modeset: Unknown symbol nvidia_register_module (err 0)
[4.499227] nvidia_modeset: Unknown symbol
nv_kthread_q_schedule_q_item (err 0)
[4.499254] nvidia_modeset: Unknown symbol nvidia_get_rm_ops (err 0)
[4.499267] nvidia_modeset: Unknown symbol nv_kthread_q_item_init (err 0)
[4.499275] nvidia_modeset: Unknown symbol nv_kthread_q_stop (err 0)
[4.499281] nvidia_modeset: Unknown symbol nvidia_unregister_module (err 0)
[4.499293] nvidia_modeset: Unknown symbol nv_kthread_q_init (err 0)
[4.510986] nvidia-nvlink: Nvlink Core is being initialized, major
device number 240
[4.511179] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[4.511179] NVRM: This can occur when a driver such as:
   NVRM: nouveau, rivafb, nvidiafb or rivatv
   NVRM: was loaded and obtained ownership of the NVIDIA device(s).
[4.511180] NVRM: Try unloading the conflicting kernel module (and/or
   NVRM: reconfigure your kernel without the conflicting
   NVRM: driver(s)), then try loading the NVIDIA kernel module
   NVRM: again.
[4.511180] NVRM: No NVIDIA graphics adapter probed!
[4.511254] nvidia-nvlink: Unregistered the Nvlink Core, major
device number 240
[4.529357] nvidia-nvlink: Nvlink Core is being initialized, major
device number 240
[4.529496] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[4.529496] NVRM: This can occur when a driver such as:
   NVRM: nouveau, rivafb, nvidiafb or rivatv
   NVRM: was loaded and obtained ownership of the NVIDIA device(s).
[4.529497] NVRM: Try unloading the conflicting kernel module (and/or
   NVRM: reconfigure your kernel without the conflicting
   NVRM: driver(s)), then try loading the NVIDIA kernel module
   NVRM: again.
[4.529497] NVRM: No NVIDIA graphics adapter probed!
[4.529652] nvidia-nvlink: Unregistered the Nvlink Core, major
device number 240
[4.536067] usb 1-1: device not accepting address 5, error -71
[4.536095] usb usb1-port1: unable to enumerate USB device
[4.564331] nvidia-nvlink: Nvlink Core is being initialized, major
device number 240
[4.564459] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[4.564460] NVRM: This can occur when a driver such as:
   NVRM: nouveau, rivafb, nvidiafb or rivatv
   NVRM: was loaded and obtained ownership of the NVIDIA device(s).
[4.564460] NVRM: Try unloading the conflicting kernel module (and/or
   NVRM: reconfigure your kernel without the conflicting
   NVRM: driver(s)), then try loading the NVIDIA kernel module
   NVRM: again.
[4.564460] NVRM: No NVIDIA graphics adapter probed!
[4.564514] nvidia-nvlink: Unregistered the Nvlink Core, major
device number 240
[4.583414] nvidia-nvlink: Nvlink Core is being initialized, major
device number 240
[4.583525] NVRM: The NVIDIA probe routine 

Re: [gentoo-user] How to test package install?

[John Covici]

On Wed, 14 Feb 2018 21:37:50 -0500,
Ian Zimmerman wrote:
> 
> I'm trying to test my package by running "ebuild /path/to/pkg.ebuild
> install".  Naturally (for me) I do this as an unprivileged user, not as
> root.  It fails because at least some steps such as dobin need to give
> away ownership of the files being installed.  I tried to run the whole
> thing including compilation under fakeroot but that doesn't help.
> 
> If it is relevant (but I don't think it is) my user _is_ in the portage
> group.
> 
> What is the accepted or usual way to do this task?

Usually installs are always done as root -- because there is always
something like what you have described.

-- 
Your life is like a penny.  You're going to lose it.  The question is:
How do
you spend it?

 John Covici wb2una
 cov...@ccs.covici.com

[gentoo-user] How to test package install?

[Ian Zimmerman]

I'm trying to test my package by running "ebuild /path/to/pkg.ebuild
install".  Naturally (for me) I do this as an unprivileged user, not as
root.  It fails because at least some steps such as dobin need to give
away ownership of the files being installed.  I tried to run the whole
thing including compilation under fakeroot but that doesn't help.

If it is relevant (but I don't think it is) my user _is_ in the portage
group.

What is the accepted or usual way to do this task?

-- 
Please don't Cc: me privately on mailing lists and Usenet,
if you also post the followup to the list or newsgroup.
To reply privately _only_ on Usenet and on broken lists
which rewrite From, fetch the TXT record for no-use.mooo.com.

[gentoo-user] Re: Vulnerability info in /sys

[Nikos Chantziaras]

On 15/02/18 02:57, Daniel Frey wrote:

On 02/14/18 09:29, Nikos Chantziaras wrote:

On 14/02/18 04:38, Daniel Frey wrote:

On 02/12/18 19:39, Ian Zimmerman wrote:

On 2018-02-12 19:24, Daniel Frey wrote:


I've read online that there should be vulnerability info (Meltdown,
Spectre) in /sys under /sys/devices/system/cpu/vulnerabilities but this
doesn't exist on my PC.
[...]


See the other threads: you need at least 4.9.79 for the /sys bits.
[...]


I'm surprised I missed those threads, I read all messages on here.
According to the thread I found it's actually starts on 4.9.77, I'm just
on the latest stable (.76).


During "special emergencies" like this one, it would be a good idea to
use the latest 4.9.x, regardless of whether portage marked it "stable"
or not. At least for a while and until the situation has settled down
again.




Nah, I like stability over everything else. I recall lots of pain and
instability in January when everyone rushed to patch the flaws (both
Windows and linux.)

These are my personal computers, not a work environment.


Personal preference of course, but I prefer to risk some instability 
compared to risking having a bitcoin miner infesting my PC or some 
botnet worm.


That's just me though.

Re: [gentoo-user] Re: Vulnerability info in /sys

[Daniel Frey]

On 02/14/18 09:29, Nikos Chantziaras wrote:
> On 14/02/18 04:38, Daniel Frey wrote:
>> On 02/12/18 19:39, Ian Zimmerman wrote:
>>> On 2018-02-12 19:24, Daniel Frey wrote:
>>>
 I've read online that there should be vulnerability info (Meltdown,
 Spectre) in /sys under /sys/devices/system/cpu/vulnerabilities but this
 doesn't exist on my PC.
 [...]
>>>
>>> See the other threads: you need at least 4.9.79 for the /sys bits.
>>> [...]
>>
>> I'm surprised I missed those threads, I read all messages on here.
>> According to the thread I found it's actually starts on 4.9.77, I'm just
>> on the latest stable (.76).
> 
> During "special emergencies" like this one, it would be a good idea to
> use the latest 4.9.x, regardless of whether portage marked it "stable"
> or not. At least for a while and until the situation has settled down
> again.
> 
> 

Nah, I like stability over everything else. I recall lots of pain and
instability in January when everyone rushed to patch the flaws (both
Windows and linux.)

These are my personal computers, not a work environment.

Dan

[gentoo-user] Is ABI a portage variable?

[Ian Zimmerman]

I am trying to package software whose build process (autotools based)
depends on a variable "ABI" to determine (on x86 and amd64 arch at least)
whether to build as 64 bit or 32 bit.  If it is not set externally, the
configure script sets it by inspecting the native pointer size of the C
compiler, to either "64" or "32".

Naturally, I thought the cleanest way for a package was to leave it
unset and let configure choose the native value.  But somehow it ends up
with a setting "amd64" and then complains that it is not one of the
expected values.  I am stuck determining where that value comes from.
Is it portage?  If so, how to change or disable it?  Is it safe to just
do the following?

ABI='' ./configure

-- 
Please don't Cc: me privately on mailing lists and Usenet,
if you also post the followup to the list or newsgroup.
To reply privately _only_ on Usenet and on broken lists
which rewrite From, fetch the TXT record for no-use.mooo.com.

Re: [gentoo-user] Re: Trouble using package.provided

[allan gottlieb]

On Wed, Feb 14 2018, Nikos Chantziaras wrote:

> On 14/02/18 18:44, allan gottlieb wrote:
>> I have never needed package.provided before so am probably doing it
>> wrong.
>>
>> There are know bugs I am encountering with
>> www-client/chromium-64.0.3282.140 and
>> net-libs/webkit-gtk-2.4.11-r200
>>
>> Since each takes a while to build before failing, I would like to stop
>> trying until there is progress on the bugs
>>
>> man 5 portage made me believe that package.provide was the answer.
>
> Why not just mask it? (/etc/portage/package.mask)

Basically the same problem as --exclude.  Portage wants me to unmask
them because they are needed.  Chromium is in @world and webkit-gtk is
needed for gnucash.

allan

Re: [gentoo-user] Trouble using package.provided

[allan gottlieb]

On Wed, Feb 14 2018, Neil Bothwick wrote:

> On Wed, 14 Feb 2018 13:28:28 -0500, allan gottlieb wrote:
>
>> Rather than facing that msg every emerge and possibly learning too well
>> to ignore warnings, I adopted the third remedy and removed the
>> package.provided files.
>> 
>> I will temporarily use a two step emerge --update ... @world
>> 
>> 1.  emerge --update --pretend ... @world
>> 2.  emerge -1 ... files suggested by 1 minus the bad chromium and
>> webkit-gtk
>
> You could try adding "exclude chromium --exclude webkit-gtk" to
> EMERGE_DEFAULT_OPTS.

I tried that a while ago.  The problem is that then portage believe
those packages aren't on my stable system and says that I must merge an
unstable package.

Specifically excluding the buggy (old) version of webkit-gtk, portage
wants me to merge a newish (testing) version of gnucash that uses a
solid new version of webkit-gtk.  I have the new version of webkit-gtk,
but really want to delay installing the testing gnucash.

allan

[gentoo-user] Re: Trouble using package.provided

[Nikos Chantziaras]

On 14/02/18 18:44, allan gottlieb wrote:

I have never needed package.provided before so am probably doing it
wrong.

There are know bugs I am encountering with
www-client/chromium-64.0.3282.140 and
net-libs/webkit-gtk-2.4.11-r200

Since each takes a while to build before failing, I would like to stop
trying until there is progress on the bugs

man 5 portage made me believe that package.provide was the answer.


Why not just mask it? (/etc/portage/package.mask)

Re: [gentoo-user] Trouble using package.provided

[Neil Bothwick]

On Wed, 14 Feb 2018 13:28:28 -0500, allan gottlieb wrote:

> Rather than facing that msg every emerge and possibly learning too well
> to ignore warnings, I adopted the third remedy and removed the
> package.provided files.
> 
> I will temporarily use a two step emerge --update ... @world
> 
> 1.  emerge --update --pretend ... @world
> 2.  emerge -1 ... files suggested by 1 minus the bad chromium and
> webkit-gtk

You could try adding "exclude chromium --exclude webkit-gtk" to
EMERGE_DEFAULT_OPTS.


-- 
Neil Bothwick

You are about to give someone a piece of your mind,
something you can ill afford...


pgpIiIGb_3fC1.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] Trouble using package.provided

[allan gottlieb]

On Wed, Feb 14 2018, Arve Barsnes wrote:

> On 14 February 2018 at 17:44, allan gottlieb  wrote:
>> I have never needed package.provided before so am probably doing it
>> wrong.
>>
>> Any help would be appreciated.
>
> The way I understand it, /etc/portage/ doesn't support that, it should
> go in /etc/portage/profile/
>
> Cheers,
> Arve

You are correct thanks.

However now
  emerge --update ... @world
gives a red-letter warning that dependent packages are in package.provided,
which is true.  It then gives three possible remedies.

Rather than facing that msg every emerge and possibly learning too well
to ignore warnings, I adopted the third remedy and removed the
package.provided files.

I will temporarily use a two step emerge --update ... @world

1.  emerge --update --pretend ... @world
2.  emerge -1 ... files suggested by 1 minus the bad chromium and webkit-gtk

thanks again.
allan

[gentoo-user] BFQ scheduler ... undefined symbol while shutdown ?

[tuxic]

Hi,

I am using a good ole' harddisc with some rotating metal plates...
;)

To get more throughput I had set  CONFIG_IOSCHED_BFQ / BFQ scheduler
in the kernel sources of 4.15.1 /.2 /.3 and while the kernel was
running anything seems to be fine.

Then while the box was shutting down the kernel, the shutdown process
hangs and error messages were printed to the screen: Anythong about
"missing/unreferenced symbols".

I am using kernels downloaded from kernel.org.

Two question: 
1) What kills the BFQ that hard?
2) Is it recommended for spinning disks anyhow and what may be better?

Thanks a lot for any help!

Cheers
Meino

[gentoo-user] Re: Vulnerability info in /sys

[Nikos Chantziaras]

On 14/02/18 04:38, Daniel Frey wrote:

On 02/12/18 19:39, Ian Zimmerman wrote:

On 2018-02-12 19:24, Daniel Frey wrote:


I've read online that there should be vulnerability info (Meltdown,
Spectre) in /sys under /sys/devices/system/cpu/vulnerabilities but this
doesn't exist on my PC.
[...]


See the other threads: you need at least 4.9.79 for the /sys bits.
[...]


I'm surprised I missed those threads, I read all messages on here.
According to the thread I found it's actually starts on 4.9.77, I'm just
on the latest stable (.76).


During "special emergencies" like this one, it would be a good idea to 
use the latest 4.9.x, regardless of whether portage marked it "stable" 
or not. At least for a while and until the situation has settled down again.

Fwd: Re: [gentoo-user] Grub2 boot problem

[Tom H]

Wonderful off-list message...


-- Forwarded message --
From:  
Date: Tue, Feb 13, 2018 at 4:40 PM
Subject: Fwd: Re: [gentoo-user] Grub2 boot problem
To: Tom H 


Please FUCK OFF.  if you must make noise, please do it off list.
Please also learn what a moderator is, and the fact that you are not
one.  Gee, i can't top post with this account, but being a foogle
skank you wouldn't understand.

mad.scientist.at.large (a good madscientist)
--
God bless the rich, the greedy and the corrupt politicians they have
put into office.   God bless them for helping me do the right thing by
giving the rich my little pile of cash.  After all, the rich know what
to do with money.


Date: 13. Feb 2018 05:59
From: tomh0...@gmail.com
To: gentoo-user@lists.gentoo.org
Subject: Re: [gentoo-user] Grub2 boot problem

On Mon, Feb 12, 2018 at 7:21 PM,  wrote:


you need to include the punctuation, specifically the ":"s, which
usually are a "-", mac addresses use the ":" but unless the syntax has
changed/broadened you have to have the "-" for seperating the fields
in a uuid. The punctuation is part of the syntax (besides breaking
the uuid into sections which makes it easier to copy/verify).


Please bottom-post.

No, grub's "(mduuid/...)" doesn't use separators (but the "(lvmid/...)" does!).

Re: [gentoo-user] Trouble using package.provided

[Arve Barsnes]

On 14 February 2018 at 17:44, allan gottlieb  wrote:
> I have never needed package.provided before so am probably doing it
> wrong.
>
> Any help would be appreciated.

The way I understand it, /etc/portage/ doesn't support that, it should
go in /etc/portage/profile/

Cheers,
Arve

[gentoo-user] Trouble using package.provided

[allan gottlieb]

I have never needed package.provided before so am probably doing it
wrong.

There are know bugs I am encountering with
www-client/chromium-64.0.3282.140 and
net-libs/webkit-gtk-2.4.11-r200

Since each takes a while to build before failing, I would like to stop
trying until there is progress on the bugs

man 5 portage made me believe that package.provide was the answer.

E6430 ~ # cat /etc/portage/package.provided/chromium
# The current version won't build (known problem)
=www-client/chromium-64.0.3282.140
E6430 ~ # 

E6430 ~ # cat /etc/portage/package.provided/webkit-gtk
# the -r200 won't build (known bug)
# "soon" won't need it since will have new gnucash
=net-libs/webkit-gtk-2.4.11-r200
E6430 ~ # 

However

E6430 ~ # emerge --update --changed-use --with-bdeps=y @world

begins with

These are the packages that would be merged, in reverse order:

Calculating dependencies... done!
[ebuild  rR] net-libs/webkit-gtk-2.4.11-r200 
[ebuild U  ] www-client/chromium-64.0.3282.140 [63.0.3239.108] 
USE="-jumbo-build%" 

Any help would be appreciated.

thanks,
allan