Re: Strange package dependency tree

[Suvayu Ali]

Sorry about the late reply.

On Sun, Sep 13, 2020 at 12:05 PM Qiyu Yan  wrote:
>
> Suvayu Ali  于2020年9月13日周日 下午6:14写道：
> >
> > Can someone explain this to me?
> >
> >   # rpm -q --whatrequires redhat-rpm-config
> > How does R-core, python27, and root-cling depend on redhat-rpm-config?
> R-Core: https://src.fedoraproject.org/rpms/R/blob/master/f/R.spec#_309
> python27: 
> https://src.fedoraproject.org/rpms/python2.7/blob/master/f/python2.7.spec#_221
> root-cling: I can't see the reason, but it is intentionally:
> https://src.fedoraproject.org/rpms/root/blob/master/f/root.spec#_4019

Thanks for hunting these down.  I really don't see why these would be
needed.  I'll see if I can find out more.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Strange package dependency tree

[Suvayu Ali]

Hi,

Can someone explain this to me?

  # rpm -q --whatrequires redhat-rpm-config
  fonts-srpm-macros-2.0.3-1.fc32.noarch
  python-srpm-macros-3-58.fc32.noarch
  R-core-4.0.2-1.fc32.x86_64
  python27-2.7.18-2.fc32.x86_64
  go-srpm-macros-3.0.9-1.fc32.noarch
  root-cling-6.22.02-1.fc32.x86_64

How does R-core, python27, and root-cling depend on redhat-rpm-config?
 Is this a packaging bug?

Thanks,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

DNS resolution failure inside containers in F32

[Suvayu Ali]

Hi everyone,

I noticed recently, after upgrading to F32 DNS resolution is failing
inside containers.

  $ docker exec -it pre_deliverable /bin/bash -i -l
  root@7d5eaa0cc50b:/# which ping
  /bin/ping
  root@7d5eaa0cc50b:/# ping 8.8.8.8
  PING 8.8.8.8 (8.8.8.8) 56(84) bytes of data.
  64 bytes from 8.8.8.8: icmp_seq=1 ttl=116 time=8.58 ms
  64 bytes from 8.8.8.8: icmp_seq=2 ttl=116 time=7.68 ms
  64 bytes from 8.8.8.8: icmp_seq=3 ttl=116 time=5.36 ms
  64 bytes from 8.8.8.8: icmp_seq=4 ttl=116 time=8.10 ms
  ^C
  --- 8.8.8.8 ping statistics ---
  4 packets transmitted, 4 received, 0% packet loss, time 8ms
  rtt min/avg/max/mdev = 5.356/7.429/8.580/1.238 ms
  root@7d5eaa0cc50b:/# ping www.google.com
  ping: www.google.com: Temporary failure in name resolution
  root@7d5eaa0cc50b:/#

I think it's because Fedora switched from iptables to nftables.
`iptables-save` shows several docker related rules, but `nft list
ruleset` doesn't seem to list any docker related rules.  systemctl
tells me neither of iptables or nftables services are running, which
makes sense because firewalld is running.  However I see these errors
when I look at the firewalld logs:

firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -D PREROUTING -m addrtype --dst-type LOCAL -j DOCKER' failed:
iptables v1.8.4 (legacy): Couldn't load target `DOCKER':No such file
or direc>

Try `iptables -h' or 'iptables --help' for more information.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -D OUTPUT -m addrtype --dst-type LOCAL ! --dst 127.0.0.0/8 -j
DOCKER' failed: iptables v1.8.4 (legacy): Couldn't load target
`DOCKER':No su>

Try `iptables -h' or 'iptables --help' for more information.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -D OUTPUT -m addrtype --dst-type LOCAL -j DOCKER' failed: iptables
v1.8.4 (legacy): Couldn't load target `DOCKER':No such file or
directory

Try `iptables -h' or 'iptables --help' for more information.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -D PREROUTING' failed: iptables: Bad rule (does a matching rule
exist in that chain?).
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -D OUTPUT' failed: iptables: Bad rule (does a matching rule exist
in that chain?).
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -F DOCKER' failed: iptables: No chain/target/match by that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
nat -X DOCKER' failed: iptables: No chain/target/match by that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -F DOCKER' failed: iptables: No chain/target/match by that
name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -X DOCKER' failed: iptables: No chain/target/match by that
name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -F DOCKER-ISOLATION-STAGE-1' failed: iptables: No
chain/target/match by that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -X DOCKER-ISOLATION-STAGE-1' failed: iptables: No
chain/target/match by that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -F DOCKER-ISOLATION-STAGE-2' failed: iptables: No
chain/target/match by that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -X DOCKER-ISOLATION-STAGE-2' failed: iptables: No
chain/target/match by that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -F DOCKER-ISOLATION' failed: iptables: No chain/target/match by
that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -t
filter -X DOCKER-ISOLATION' failed: iptables: No chain/target/match by
that name.
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -D
FORWARD -i docker0 -o docker0 -j DROP' failed: iptables: Bad rule
(does a matching rule exist in that chain?).
firewalld[856]: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w10 -D
FORWARD -i docker0 -o docker0 -j DROP' failed: iptables: Bad rule
(does a matching rule exist in that chain?).
log.txt (END)

Anyone have any thoughts about what is going on?  How can I solve this?

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

On Sat, May 30, 2020 at 12:34 PM Roger Heflin  wrote:

> those workstations would be the expensive cards.  A yearly redhat
> support license is enough that no one is going to buy one for a $1000
> machine because in 3 years that license will cost more than the HW, so
> the "cheap" hardware that is being discussed like any of us would use
> isn't something they will likely run into in the enterprise market.
> There is a significant amount of HW used on this list that they simply
> will never be supporting on the enterprise end.

You really need to keep up.  For many years Fedora devs (kernel, and
bios) have spearheaded the effort to make Linux in general, and Fedora
specifically a priority on desktops and laptops.  Tremendous amount of
work went into the kernel and UEFI compatibility, particularly the way
M$ was pushing for "secure boot".  So much so, Fedora kernels were
controversially secure boot compatible earlier due to a (digital)
signing agreement with M$.  Fedora has also contributed significantly
in making FOSS radeon drivers reliable, and advanced noveau despite
hostility from NVidia.

As for recent developments, if Fedora doesn't want to support regular
installs, how come ThinkPads are going to be shipped by Lenovo with
Fedora pre-installed?

  https://fedoramagazine.org/coming-soon-fedora-on-lenovo-laptops/

And you still haven't read the BZ I linked in my OP, and going on and
on about hypothetical scenarios.  Please inform yourself better.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

On Sat, May 30, 2020 at 7:41 AM John M. Harris Jr  wrote:
> Yes, the other option would be to move to Debian or find another rpm-based
> distro that still supports 32 bit. All of this because Fedora decided to do
> what seems to be so common recently, dropping what still works well. This
> hurts our community. This hurts our users. It'd be one thing to make it so
> that QA requirements were effectively dropped on 32 bit, but there was no real
> reason to drop 32 bit support. It still worked, and quite well.

If I recall correctly, the biggest "user burden" for 32-bit packages
was load on people who maintain Fedora mirrors.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

Hi Bruno,

On Sat, May 30, 2020 at 3:11 AM Bruno Wolff III  wrote:
[...]
> I think that covers most of it. It isn't really as bad as it looks
> written out.

Thank you so much for sharing it!  I can work with this, maybe some of
it could also be automated leveraging copr ;)

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: SQL Database Client GUI for Firebird?

[Suvayu Ali]

On Fri, May 29, 2020 at 11:39 PM linux guy  wrote:
>
> Hi people.
>
> I have a huge CSV spreadsheet. I need to do a bunch of light duty data 
> manipulation.  Sort, total, export, that sort of thing.  I could almost do it 
> in the spreadsheet, but it would be too much copy/paste and the combinations 
> of data I want to get out are exhaustive.
>
> I'm thinking of importing it into Firebird.   I'm looking for a good GUI SQL 
> client.
>
> Thoughts/ advice ?
>
> TOar ?
> DBEaver ?

Good GUI client, you could use it with SQLite and get the job done effectively

> Something else ?  (Python?)

I prefer Python, particularly Pandas.

  import pandas as pd
  df = pd.read_csv("/path/to/giant.csv", )

Then you can use the Pandas API to do whatever you like really (see
the official docs, very thorough, and easy to understand, full of
examples).  If you prefer an SQL interface, you can try installing
`sqldf` alongside pandas, and run SQL inside the pandas DataFrame.

HTH,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

On Fri, May 29, 2020 at 6:57 PM Roger Heflin  wrote:
>
> I don't believe the kernel.org developers work out of the fedora
> bugzilla (or any distro's bugzilla), so no one who knows anything is
> likely to find and/or see the bug.
>
> To get a kernel developer you would need to at least post a summary to
> the kernel subsystem list if you know which subsystem or the main
> kernel if you don't.
>
> And the kernel.org guys do not care about any testing done on a fedora
> delivered kernel as they don't know what code is in it, so you would
> need to install a kernel.org kernel and bot from it and verify you
> have the issue on the newest released one.

I wish people read the thread or even posts carefully before
responding.  I'm well aware that kernel developers are not watching
the Redhat BZ, but the maintainers are.

Anyway, here's an example of a healthy exchange on BZ:

- I reported this bug on Tuesday:
https://bugzilla.redhat.com/show_bug.cgi?id=1840432
- There was some discussion
- Today it got marked as a duplicate of this bug:
https://bugzilla.redhat.com/show_bug.cgi?id=1840780
- The maintainer had commented on the second bug on Wednesday that he
is too busy, and unlikely to find time to investigate
- On Thursday another person dug through the upstream repo, found the
commit that fixes the issue, determined that simply updating to the
latest version from upstream would fix the crash, and as a bonus
pointed to the underlying reason that caused the crash.
- Earlier today the maintainer did a scratch build of the latest
release, and asked for testing
- I happened to be online, after testing I responded that the build works
- Quite likely there will be a new release next week that pushes out the update.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

Hi Bruno,

On Fri, May 29, 2020 at 7:59 PM Bruno Wolff III  wrote:
>
> I occasionally submit kernel bug reports. These days one good way to
> get your report looked at is to bisect a Linus kernel to find the commit
> that triggered the problem. This normally takes me about a week to get
> done. I have gotten fixes for bugs that affected old hardware that was in
> limited use by other people. It isn't that hard to do, but kernel builds can
> take a while and you need to be able to reboot for each test and there can
> be limitations on when that can get done.

Could you please share your workflow?  I have been looking for some
guidance so that I can test upstream kernels when I encounter these
hardware issues.  I don't need step by step instructions, I'm very
comfortable compiling software, I just need some way to manage the
self-compiled kernels alongside Fedora kernels without littering my
system with build artifacts, and play nice with SELinux :).  I
typically limit any software I compile myself to /opt or $HOME;
unfortunately that doesn't quite work with kernels :-p

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

Hi Michael,

On Thu, May 28, 2020 at 5:30 PM Michael Schwendt  wrote:
>
> On Thu, 28 May 2020 18:08:48 +0930, Tim via users wrote:
>
> > Perhaps there should be an automated culling of participants.  If you
> > step up the plate to say you'll maintain a package, but don't, *you*
> > get dumped from bugzilla.
>
> Please let's not create a thread of doom.
>
> You can't seriously suggest blocking kernel maintainers from bugzilla just
> because they don't have the manpower to take a look at every ticket and
> perform meaningful, helpful triaging. Some key components are literally
> flooded with bug reports. In order to deal with the number of tickets,
> using scripts is an obvious thing to do. Yet it isn't done in a proper and
> OS user-friendly way.

I agree, this is exactly the case (I'm the OP, and the person who
filed the BZ referenced earlier).  There should be clear guidance how
a motivated user can follow through, do the relevant triaging, and
maybe even test an upstream kernel,  In fact if you read the BZ,
you'll see that I tried vanilla RC kernels from Thorsten's repo and
reported back to the BZ.

What baffles me most, is the nature of the bug.  It is the text book
case of a high priority bug, new (budget) hardware, which is becoming
common place very fast, where Fedora isn't bootable, add to that it is
a regression bug. How does a bug with these characteristics not come
within the peripheral vision of the maintainers, specially when the
reporter is eager to do all kinds of troubleshooting and testing!?
Clearly something is wrong with the automation, the question is. how
should that be fixed?

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

Hi Michael, Ben,

On Wed, May 27, 2020 at 7:23 PM Michael Schwendt  wrote:
>
> On Wed, 27 May 2020 12:07:35 -0400, Ben Cotton wrote:
>
> > On Wed, May 27, 2020 at 4:30 AM Suvayu Ali wrote:
> > >
> > > Lately all my bug reports tend to go like this.  Are others having the
> > > same experience?
> >
> > I understand the frustration. My bugs get closed EOL, too. For what
> > it's worth, 3633 bugs were closed EOL for Fedora 30. This is
> > considerably lower than Fedora 29 (4958) and Fedora 28 (4681).
> >
> > The Fedora Join SIG is here to help new contributors get started if
> > you're interested:
> > https://docs.fedoraproject.org/en-US/fedora-join/index.html
>
> That process would not have helped with this kernel bugzilla ticket
> opened in 2019-08-18. But the response in March 2020 suggested reproducing
> the issue with Fedora 31 and reassigning the ticket, if the issue is still
> reproducible. That has not been done.
>
> The fundamental problem here is that it has taken a very long time for
> somebody to respond to the bug reporter. There has been no guidance and
> no hint whether anyone "somewhere" would be interested in looking into
> this issue.

It took so long, that the bug report became irrelevant.  I have since
moved to a different country, with a different job, and don't have
access to the original machine.  In fact, as mentioned in the bug, I
could reproduce a similar issue in similar hardware, with Fedora 31.
The old hardware was Ryzen 5 2400G, and the second one was Ryzen 7 PRO
3700U (a mobile CPU with similar design, don't get confused with the
2xxx vs 3xxx naming).  No response.  Even more importantly, this
started as a regression bug that renders a system unbootable; my
understanding of Fedora and Linux kernel policy says that's very high
severity, and still I saw absolutely no response.  Ironically, the bug
on 3700U got almost resolved by a very recent Fedora 32 kernel update
this month (that's 10 months, and 2 distro upgrades later).   Given
the apathy, I had not bothered any more to file a bug report
specifically for the 3700U.

This is not the first time, I had purchased the 2400G within the first
month of its release back in Feb-Mar 2018.  Understandably there were
issues (system unbootable to a desktop), even then, my bug reports did
not get a single response, me talking to myself, reporting back with
my experiments and attempted workarounds.  With the release of a new
kernel series, about 2-3 months later, all issues were resolved.  In
the least, I could have helped test any bleeding edge kernels.  I have
been using Fedora for over a decade, have contributed to the
distribution in many different ways (packaging, testing, etc).  This
is not a one off issue, I have had bugs open for many other packages,
all have faced the same fate.  If I can't participate in the
community, when I'm competent, and willing, I will regrettably move,
despite loving the distro; most likely to Arch.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Fedora 30 EOL

[Suvayu Ali]

On Tue, May 26, 2020 at 4:23 PM Mohan Boddu  wrote:
>
> As of the 26th of May 2020, Fedora 30 has reached its end of life for

https://bugzilla.redhat.com/show_bug.cgi?id=1742960

Lately all my bug reports tend to go like this.  Are others having the
same experience?  I have been using Fedora a long time, and part of
the pleasure was I could contribute to making the distro better.  Now
it seems to have reduced to "wait until it fixes itself", how is that
any different from Windows/Mac?  It's very disappointing to see the
community aspect of Fedora go down this way while the OS has improved
by so much compared to when I started using Fedora.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Ban a specific mirror in dnf

[Suvayu Ali]

Hi Samuel, and Tim,

On Tue, Aug 13, 2019 at 4:57 AM Tim via users
 wrote:
>
> On Mon, 2019-08-12 at 16:35 -0700, Samuel Sieb wrote:
> > You could add it to your /etc/hosts file like:
> > 127.0.0.2   the.bad.domain
>
> This does seem like a roundabout solution to what has to be a common
> problem (wanting to blackban specific repos).  For whatever reason
> people have wanted to do that, it ought to be possible to directly do
> it with the configuration for the software in question.
>
> While some might argue it would best to have a bad repo removed from
> the pool, it could be the case that the repo is fine, just that the
> path between it and some users is a problem.
>
> --
> [tim@localhost ~]$ uname -rsvp
> Linux 5.0.16-100.fc28.x86_64 #1 SMP Tue May 14 18:22:28 UTC 2019 x86_64
>
> Boilerplate:  All mail to my mailbox is automatically deleted.
> There is no point trying to privately email me, I only get to see
> the messages posted to the mailing list.
>
> Sorry, no viruses was included with this email, please read the rm
> and fdisc man files to learn how to seriously mangle your own system.

I have come across this workaround, but dislike it for exactly the
same reason as Tim.

Another "solution" floating around the web is use fastestmirror.
Those of us who have been a Fedora users know it all too well the
issues with fastestmirror.  That said, I believe it has a "exclude"
option, if I could use that, but not use fastestmirror, that could be
acceptable for the time being.  But I don't think that's possible.
Maybe I should just use Samuel's workaroound for now, and file an RFE
against dnf.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Ban a specific mirror in dnf

[Suvayu Ali]

Hi Tony

On Mon, Aug 12, 2019 at 11:07 PM Tony Nelson
 wrote:
>
> On 19-08-12 13:34:22, Suvayu Ali wrote:
> > Hi,
> >
> > I would like to ban a particularly slow mirror[1].  It's the closest
> > mirror to my geographical location (one of only two mirrors in my
> > country, India).  Every time I see a bad download error message from
> > that mirror, everything subsequent to that speeds up and uses the full
> > capacity of my connection.  How can I achieve this?
>   ...
>
> Use the firewall and block that IP?

I would like to avoid hard coded changes. What if tomorrow their IP
changes, given the nature of the problem, I would realise the issue
after trying to waste my time on "slow updates" all over again!

>
> Edit the repo files to use a baseurl of your choice?  Start with the
> example, and change it to your preferred mirror.  It can be a list of
> urls separated by spaces or commas.

This would work on my desktops, but not on my laptops.  I have several
machines to manage.  A uniform solution would be easier I think.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Ban a specific mirror in dnf

[Suvayu Ali]

Hi,

I would like to ban a particularly slow mirror[1].  It's the closest
mirror to my geographical location (one of only two mirrors in my
country, India).  Every time I see a bad download error message from
that mirror, everything subsequent to that speeds up and uses the full
capacity of my connection.  How can I achieve this?

Also, is there a way to report this?  Mirrormanager shouldn't even be
listing this mirror, it considerably degrades the user experience.

Thanks,

[1] Ironically it is hosted by the CS department at IIT Kanpur!

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Kernel entries are not updated in grub menu

[Suvayu Ali]

Hi Chris, Tim,

Sorry for the late response.

Thanks for the explanations.

On Sat, Jul 06, 2019 at 03:18:59PM -0600, Chris Murphy wrote:
> On Sat, Jul 6, 2019 at 1:18 PM Suvayu Ali  wrote:
> 
> Making things worse, many manufacturers treat their customers like
> children, and have decided to refer to UEFI firmware as BIOS. So when
> you go looking for firmware updates, you're likely to find them listed
> as BIOS updates. It's just terrible. They should have just called it
> what it is, firmware, from the outset and not caused end users to be
> confused with either a new term, UEFI, let alone by interchanging two
> different things, BIOS and UEFI, as if they are the same thing. But
> that's where it is.

Indeed, my Gigabyte firmware is from 2018, and everywhere in the UI it says 
"BIOS"!  I looked (not very thoroughly), but couldn't find a toggle to reenable 
UEFI.  I guess I need to do a bit of searching on the weekend, it's probably 
deep somewhere in the menu.

> 
> But depending on the age of the computer, make and model, there might
> be hints somewhere that'll fairly conclusively confirm/deny if that
> model has UEFI firmware or BIOS, and then also how to confirm whether
> the faux-BIOS is enabled. For sure if the /sys/firmware/efi/efivars
> directory exists, it's UEFI.

I'm guessing that's dynamic, present only if you boot with UEFI?  It's not on 
my system.

Thanks again, cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Kernel entries are not updated in grub menu

[Suvayu Ali]

Hi Chris,

On Fri, Jul 05, 2019 at 02:53:08PM -0600, Chris Murphy wrote:
> 
> Post the contents of the following files somewhere (I'm not sure
> they'll attach to the list but you can give it a shot if you want)
> 
> /boot/grub2/grub.cfg

This evening I went through this carefully, and spotted a few inconsistent 
partition ids.  My / is on /dev/sdc2 (no separate /boot partition), but some of 
the entries had 'hd1,msdos2' while others correctly had 'hd2,msdos2'.  Fixing 
them led to the BLS config files getting picked up correctly.  I think I know 
how this happened.  Sometime back I added a new disk.  That must have messed up 
the ordering.

> And also what was the original Fedora installation version?

I don't recall exactly, I assembled this machine in March 2018.  So my guess is 
F28/27.

> On x86
> BIOS, GRUB is installed by 'grub2-install' only upon installation.
> Even when the package is updated and upgraded through major versions,
> the actual installed bootloader becomes stale over time because
> there's no automatic 'grub2-install' run again. (I'm of the opinion
> that Fedora needs to keep it up to date, at least with major upgrades.
> Too many problems arise over time as it becomes stale.

I agree.

> Meanwhile on
> UEFI, the bootloader is being updated, even within minor version
> updates, whenever the shim or grub packages get updated.)

I have never quite understood how UEFI worked.  Initially I stayed away from it 
because of secure boot issues.  Do you think I should put in the effort to move 
to UEFI?  If so, would you be able to point me where to start?

Thanks a lot for your helpful comments,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Kernel entries are not updated in grub menu

[Suvayu Ali]

Hi Sergio,

On Thu, Jul 4, 2019 at 9:44 PM Sergio Cipolla  wrote:
>
> I've had trouble too is a new Fedora install with grub on the partition (not 
> MBR), I even eventually re-installed because I tried to revert to 
> grubby-deprecated and then couldn't get the system to boot to 
> graphical.target because of the entries needed for nvidia drivers, even after 
> reverting to nouveau.
> So now, although there hasn't been no new kernels yet, my mantra will be 
> don't touch what's working.
> I'll just do the manual grub2-mkconfig...
>
> As for the new bls, there are some meagre instructions at 
> https://fedoramagazine.org/setting-kernel-command-line-arguments-with-fedora-30/

Thanks for the link, BLS is enabled on my system.  The rest of the
grubby commands don't seem to do anything as I'm not changing any
kernel arguments.  Interestingly, when I run `grubby
--bootloader-probe`, it warns me saying that is deprecated, and I
should run grubby-bls instead.  However I haven't found a package
corresponding to that binary.  I did a file search with `dnf repoquery
-f \*/grubby\*`, no obvious candidates.

So it's still a mystery to me :-|

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Kernel entries are not updated in grub menu

[Suvayu Ali]

Hi,

For a while now kernel updates doesn't update the grub menu on my system.  
However I could always manually update using `grub2-mkconfig -o 
/boot/grub/grub2.cfg`.  But lately I can't even do that!  I have already tried 
running `grub2-switch-to-blscfg`, but I don't think that does anything.  I 
tried that because the comments in the config file says menus are now 
automatically generated from BLS config files in '/boot/loader/entries', which 
seems to be in order for me, there are conf files for all the installed 
kernels.  I can't figure out where to go from here.  I'm on F30, booting with 
BIOS.  Any thoughts?

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Building custom installation media

[Suvayu Ali]

On Wed, Apr 03, 2019 at 09:06:25PM -0700, Jack Craig wrote:
> > >> I could find is this page:
> > >>
> > https://fedoraproject.org/wiki/How_to_create_a_Fedora_install_ISO_for_testing#Build_a_DVD.iso,
> >
> > >> which is horribly out of date.
> 
> 
> 
> *just curious, 'horribly out of date implies to me that you believe the
> process has greatly changed. *
> *as you make it clear you know better, pls elaborate??*

The link in my original post is out of date because it is about three years 
old, and `pungi` has changed significantly.  So much so, there are no scripts 
called just `pungi` any more, the options to the current scripts are also 
markedly different.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Building custom installation media

[Suvayu Ali]

On Wed, Apr 03, 2019 at 05:39:35PM -0700, Samuel Sieb wrote:
> On 4/3/19 4:32 PM, Samuel Sieb wrote:
> > On 4/3/19 3:42 PM, Suvayu Ali wrote:
> > > I want to build my own installation media (on USB).  I would like to
> > > include additional packages for an offline installation.  However
> > > all I could find is this page: 
> > > https://fedoraproject.org/wiki/How_to_create_a_Fedora_install_ISO_for_testing#Build_a_DVD.iso,
> > > which is horribly out of date.  Does anyone know a more current set
> > > of steps?
> > 
> > Try 
> > https://fedoraproject.org/wiki/Livemedia-creator-_How_to_create_and_use_a_Live_CD
> 
> Something to be aware of is that most products are only live media with the
> option to install to the hard drive.  There are a couple of net install
> images.  Only the server product has an option for a full offline DVD
> install and I think I remember there were plans to drop that as well.  It
> might be easiest to just modify the server DVD for your use.  I think that
> would just involve adding the packages and modifying the comps file, but
> there might be some other step involving anaconda.

Thanks a lot Samuel for the extra pointer.  I think I have sufficient info now 
to go ahead :)

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Building custom installation media

[Suvayu Ali]

Hi Everyone,

I want to build my own installation media (on USB).  I would like to include 
additional packages for an offline installation.  However all I could find is 
this page: 
https://fedoraproject.org/wiki/How_to_create_a_Fedora_install_ISO_for_testing#Build_a_DVD.iso,
 which is horribly out of date.  Does anyone know a more current set of steps?

I have tried asking on #fedora-devel, but didn't receive any response yet.

Thanks for any pointers.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: kernel: Failed to load gpu_info firmware

[Suvayu Ali]

Hello Stan,

On Sun, Oct 21, 2018 at 09:37:19AM -0700, stan wrote:
> 
> You could compare the journal messages from the older firmware when
> it succeeded with those from the failing firmware to see if there is any
> difference.  For that matter, compare the successful old load with the
> failing old load.  But I don't think it is failing to find the firmware
> files that exist.
> 
> The AMD firmware is loaded in drivers/gpu/drm/amd/amdgpu/ampgpu_device.c
> In that source, there are two other firmware blobs treated the same as 
> the raven: vega10 and vega12.  Do you find those in the same place as
> the raven firmware blob?  Do they have the same permissions?

 [...chomp...chomp...chomp...]

> Because there is a specific path for raven, I'm surprised that the
> older firmware actually worked.  The kernel looks for firmware with
> raven in the name, and I didn't see a fallback, though I could have
> missed it because everything is done via pointers to structs, and I
> didn't get into that level of detail.

Thank you for taking the time to look at the source.  But I think there was a 
miscommunication.  I have been using the raven ridge firmware successfully 
before.  It failed after I tried the ROCm packages.

> Finally, could there be leftovers from the Rocm code causing problems?

That was it, see Allan's post.  It was stale dracut conf files that pointed to 
a different kernel firmware directory that's the culprit.

Thanks again for taking the time to investigate :)

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: kernel: Failed to load gpu_info firmware

[Suvayu Ali]

Hello Allan,

On Sun, Oct 21, 2018 at 10:40:32PM -, Allan  wrote:
> >amdgpu :06:00.0: Direct firmware load for amdgpu/raven_gpu_info.bin 
> > failed with error -2
> >   
> 
> I have had the same fight with AMDGPU-PRO drivers on Centos.
> The problem is that AMD leaves some conf files for dracut.
> go into /etc/dracut.conf.d dir and delete anything amdgpu related.
> (For me, that means delete everything in that dir)
> 
> After that rerun dracut to rebuilt initramfs,
> and you should be back to normal.

Bang on!

  $ ls /etc/dracut.conf.d/
  amdgpu-4.18.10-200.fc28.x86_64.conf
  amdgpu-4.19.0-0.rc6.git0.1.vanilla.knurd.1.fc28.x86_64.conf
  amdgpu-4.19.0-0.rc7.git0.1.vanilla.knurd.1.fc28.x86_64.conf

And, the contents are:

  add_drivers+=" amdgpu"
  add_drivers+=" amdkfd"
  fw_dir+="/lib/firmware/4.18.10-200.fc28.x86_64"

Of course it fails, the kernel version number shouldn't be there.  Strangely, 
the files were affecting kernels of other versions too.  Anyway, I will rebuild 
the initramfs later today.

Thanks a lot!

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: kernel: Failed to load gpu_info firmware

[Suvayu Ali]

Hi Stan,

Sorry for the late response, I wasn't monitoring the list.

I actually thought of what you mentioned in the other post, the firmware making 
a persistent change in some register.  I really don't know how I could 
investigate that :-|

On Wed, Oct 17, 2018 at 09:18:18AM -0700, stan wrote:
> On Wed, 17 Oct 2018 09:26:01 +0530
> Suvayu Ali  wrote:
> 
> > I have already tried both rebuilding the initrd, and reinstall the
> > kernel to no avail :(
> 
> I don't know if firmware loading is done similarly to a kernel module.
> If it is, you could try modprobe. 

I have tried modprobe, lsmod, etc, and I see the amdgpu module is loaded.  I 
don't know how to check something for firmware.  I guess the issue is, the 
module loads, but it fails to find and load the binary blob (firmware) that it 
needs (as seen in the journal).  So the picture is consistent, I just can't 
figure out why it fails to find the firmware files.  Does Fedora use a firmware 
loader?  Maybe I can hack around and invoke it in "verbose mode".  If it is 
something internal to the kernel, I doubt something like that would be possible.

> The other thing is to do an 
> rpm -q --filesbypkg [package name]
> for each of the packages in question, to see if there is a difference
> in what they installed.  A difference should point you to the file(s)
> that are causing the issue.

I have tried rpm --verify, no luck.  I will try your suggestion and see if that 
points to something.  My plan of last resort is to do a fresh install when F29 
is released, but understanding and solving this issue would be far more 
satisfying :).

Thanks again,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: kernel: Failed to load gpu_info firmware

[Suvayu Ali]

On Mon, Oct 15, 2018 at 04:41:12PM +, Rick Stevens wrote:
> > 
> > My hunch is, as the ROCm installation uses dkms to build the kernel module 
> > (which failed btw, that's why ROCm didn't work for me), the uninstallation 
> > somehow leaves behind some configuration which persists across kernels.  Do 
> > you think that's possible?
> 
> Yes, that's possible. I've never even tried ROCm, so I can't speak to
> it. You may have to rebuild your initramfs image for the kernel.
> 
> It may be easier to simply reinstall the kernel:
> 
>   sudo dnf reinstall kernel-core-4.18.12
> 
> and see if that clears the issue.

I have already tried both rebuilding the initrd, and reinstall the kernel to no 
avail :(

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: kernel: Failed to load gpu_info firmware

[Suvayu Ali]

Hello Rick,

On Fri, Oct 12, 2018 at 11:51:40PM +, Rick Stevens wrote:
> > 
> >   [drm] amdgpu kernel modesetting enabled.
> >   [drm] initializing kernel modesetting (RAVEN 0x1002:0x15DD 0x1458:0xD000 
> > 0xC6).
> >   [drm] register mmio base: 0xFE60
> >   [drm] register mmio size: 524288
> >   [drm] add ip block number 0 
> >   [drm] add ip block number 1 
> >   [drm] add ip block number 2 
> >   [drm] add ip block number 3 
> >   [drm] add ip block number 4 
> >   [drm] add ip block number 5 
> >   [drm] add ip block number 6 
> >   [drm] add ip block number 7 
> >   [drm] add ip block number 8 
> >   amdgpu :06:00.0: Direct firmware load for amdgpu/raven_gpu_info.bin 
> > failed with error -2
> >   amdgpu :06:00.0: Failed to load gpu_info firmware 
> > "amdgpu/raven_gpu_info.bin"
> >   amdgpu :06:00.0: Fatal error during GPU init
> >   [drm] amdgpu: finishing device.
> >   amdgpu: probe of :06:00.0 failed with error -2
> 
> Well, error "-2" typically is "file not found", so my guess is the code
> isn't looking in the right spot for the firmware files. Yes, you have
> them in the right spot, but it could be a permissions or selinux context
> issue as well.
> 
> Have you checked dmesg, journalctl and/or the selinux logs to see what
> they say?

What I quoted above _is_ from the journal.  I double checked, there are no 
other suspicious messages.  The SELinux label for the firmware files is 
"system_u:object_r:lib_t:s0".  There are also no AVC denials.

My pre-post searching told me "-2" is for "file not found", so I had already 
checked the usual.  I thought if I could get a more verbose message in the 
journal, maybe I get the reason behind the "-2", so I replaced "quiet" by 
"verbose" in kernel arguments, that didn't help.  I also looked if there are 
any udev rules that might try to load the firmware from a different path 
(apparently Ubuntu did that at some point), no go.

I also ran `rpm --verify` on the kernel-{core,modules} and linux-firmware 
packages, all I got is:

  # rpm --verify kernel-core-4.18.12  # picked a few kernels, same for all
  .M...  g /boot/System.map-4.18.12-200.fc28.x86_64
  .M...  g /boot/initramfs-4.18.12-200.fc28.x86_64.img

The permissions are 600.

My hunch is, as the ROCm installation uses dkms to build the kernel module 
(which failed btw, that's why ROCm didn't work for me), the uninstallation 
somehow leaves behind some configuration which persists across kernels.  Do you 
think that's possible?

Thanks a lot for your response.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

kernel: Failed to load gpu_info firmware

[Suvayu Ali]

Hello,

I have a strange issue.  I am on one of AMD's Raven Ridge APUs (Ryzen 5 2400G). 
 Everything was working fine, with the 4.18.x series kernels.  However I wanted 
to experiment with the ROCm drivers[1] that are under development.  I managed 
to install after some fiddling, but the firmware kept on failing to load.  I 
had to forcefully regenerate all the initrd a few times.  Not having succeeded, 
I decided to uninstall ROCm and go back to my original kernels.  However, I 
still receive the same error even though I have reinstalled my kernels, 
installed newer kernels from the repo, and what not.  Here are the messages 
from my journal:

  [drm] amdgpu kernel modesetting enabled.
  [drm] initializing kernel modesetting (RAVEN 0x1002:0x15DD 0x1458:0xD000 
0xC6).  
 
  [drm] register mmio base: 0xFE60
  [drm] register mmio size: 524288
  [drm] add ip block number 0 
  [drm] add ip block number 1 
  [drm] add ip block number 2 
  [drm] add ip block number 3 
  [drm] add ip block number 4 
  [drm] add ip block number 5 
  [drm] add ip block number 6 
  [drm] add ip block number 7 
  [drm] add ip block number 8 
  amdgpu :06:00.0: Direct firmware load for amdgpu/raven_gpu_info.bin 
failed with error -2
  
  amdgpu :06:00.0: Failed to load gpu_info firmware 
"amdgpu/raven_gpu_info.bin" 

  amdgpu :06:00.0: Fatal error during GPU init
  [drm] amdgpu: finishing device.
  amdgpu: probe of :06:00.0 failed with error -2

This is strange because the firmware files are in their rightful place:

  # lt /lib/firmware/amdgpu/raven_*
  -rw-r--r--. 1 root root 334K Aug 15 17:36 /lib/firmware/amdgpu/raven_vcn.bin
  -rw-r--r--. 1 root root  17K Aug 15 17:36 /lib/firmware/amdgpu/raven_sdma.bin
  -rw-r--r--. 1 root root  39K Aug 15 17:36 /lib/firmware/amdgpu/raven_rlc.bin
  -rw-r--r--. 1 root root  22K Aug 15 17:36 /lib/firmware/amdgpu/raven_pfp.bin
  -rw-r--r--. 1 root root 262K Aug 15 17:36 /lib/firmware/amdgpu/raven_mec.bin
  -rw-r--r--. 1 root root 262K Aug 15 17:36 /lib/firmware/amdgpu/raven_mec2.bin
  -rw-r--r--. 1 root root  18K Aug 15 17:36 /lib/firmware/amdgpu/raven_me.bin
  -rw-r--r--. 1 root root  316 Aug 15 17:36 
/lib/firmware/amdgpu/raven_gpu_info.bin
  -rw-r--r--. 1 root root 9.2K Aug 15 17:36 /lib/firmware/amdgpu/raven_ce.bin
  -rw-r--r--. 1 root root  37K Aug 15 17:36 /lib/firmware/amdgpu/raven_asd.bin

Any thoughts on how I can debug this?  I have also openned an issue on the ROCm 
repo[2], but I have yet to receive a useful response.

Thanks,

[1] https://rocm.github.io/
[2] https://github.com/RadeonOpenCompute/ROCm/issues/576

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org

Re: Troubleshooting random hangs

[Suvayu Ali]

Hi Jorge,

I didn't see your responses until today!  I guess I got some clarity from our
bugzilla discussions.

On Fri, Mar 16, 2018 at 09:43:04AM +, Jorge Martínez López wrote:
> I did some research and found the following kernel bug:
> 
> https://bugzilla.kernel.org/show_bug.cgi?id=196683
> 
> Fedora has CONFIG_RCU_NOCB_CPU=y in the 4.15.8 kernel configuration so I
> added "rcu_nocbs=0-7" to the boot parameters and it has been running stable
> for a while. I have also added "nopti" as well as there is some anecdotal
> evidence it improves stability but I'm not sure about that.

I think my tracebacks are very different.  That said, it also seems to me I'm
having freezes due to several unrelated reasons, and AMDGPU is probably one
among many.

> There is also some discussions in the bug page about old PSUs not providing
> good enough low voltage, AMD is recommending running a newer PSU
> (post-Haswell) but for the time being the boot config is working for me.

This is an interesting point, but very difficult to test :-|.

I haven't been able to debug my issues successfully as nothing useful really
shows up in the journal.  I was hoping someone could suggest a way so that I
could get more information to file a more specific bug report.

Any thoughts anyone?

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org

Troubleshooting random hangs

[Suvayu Ali]

Hi,

I have been having random hangs on my new Ryzen workstation (Ryzen 5
2400G + B350 mobo). My hardware is supposedly properly supported on
4.15+ kernels.  But I have been unable to boot with any of the ones in
the repo.

That said, I can boot with older kernels, but the desktop hangs
randomly.  When I say hang, I mean it freezes, and my only recourse is
to reset my computer.  I have tried to login remotely, but then I get
"No route to host" from ssh.  Looking at the journal, I can't figure
out what is causing these hangs.  If someone could have a look, that
would be wonderful.

Logs from the last two hangs:

https://paste.fedoraproject.org/paste/8T~X8BYuVboAJK3Mkal72A
https://paste.fedoraproject.org/paste/qypJAnAKE01GD-OgC6n0SQ

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org

Re: [SOLVED] F25: Can't edit connections (NM)

[Suvayu Ali]

On Thu, Dec 15, 2016 at 04:03:49PM +0100, Frank Elsner wrote:
> On Fri, 9 Dec 2016 23:29:33 +0100 Frank Elsner wrote:
> > 
> > upgraded my system from version 24 to version 25 of Fedora and 
> > suddenly encountered that I can't edit my network connections :-(
> > 
> > I'm using Network Manager "Edit Connections", can change values
> > but "save" button is grayed out and therefor inactive.
> > 
> > I'm using Network Manager
> > What might be the reason?
> 
> In the "Edit Connections" the "cloned MAC address" was filled with the string 
> "permanent"
>   for the selcted 
> connection.
> 
> Inserting the real MAC made the "save" button active and working.

Hmm, I tried leaving it blank, that worked as well!

Strange indeed.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org

Re: F25: Can't edit connections (NM)

[Suvayu Ali]

Hi Ed,

On Sun, Dec 11, 2016 at 02:18:18PM +0800, Ed Greshko wrote:
> On 12/11/16 13:05, Suvayu Ali wrote:
> > On Fri, Dec 09, 2016 at 11:29:33PM +0100, Frank Elsner wrote:
> >> upgraded my system from version 24 to version 25 of Fedora and 
> >> suddenly encountered that I can't edit my network connections :-(
> >>
> >> I'm using Network Manager "Edit Connections", can change values
> >> but "save" button is grayed out and therefor inactive.
> >>
> >> I'm using Network Manager
> >> What might be the reason?
> >
> > I have the same problem!  It also keeps overwriting my choice of
> > firewall zone when it's not the default.  I'm not sure how to
> > investigate so that I can file a useful bug report.
> 
> What desktop are you using?  I'm a KDE user and I've updated several systems 
> and not
> having any issues.

I'm on XFCE.

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org

Re: F25: Can't edit connections (NM)

[Suvayu Ali]

Hi,

On Fri, Dec 09, 2016 at 11:29:33PM +0100, Frank Elsner wrote:
> 
> upgraded my system from version 24 to version 25 of Fedora and 
> suddenly encountered that I can't edit my network connections :-(
> 
> I'm using Network Manager "Edit Connections", can change values
> but "save" button is grayed out and therefor inactive.
> 
> I'm using Network Manager
> What might be the reason?

I have the same problem!  It also keeps overwriting my choice of
firewall zone when it's not the default.  I'm not sure how to
investigate so that I can file a useful bug report.

Any thoughts anyone?

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
___
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org

Re: Creating a PostgreSQL db

[Suvayu Ali]

On Thu, Mar 10, 2016 at 01:26:40AM +0530, Suvayu Ali wrote:
> Hi Gordon,
> 
> On Wed, Mar 09, 2016 at 10:50:03AM -0800, Gordon Messmer wrote:
> > On 03/09/2016 10:18 AM, Suvayu Ali wrote:
> > >Shouldn't the local line take care of my permissions?  What am I doing 
> > >wrong?
> > 
> > The only user that exists by default is "postgres".  Use that use to create
> > new users and databases:
> > 
> > su postgres -c "createuser x"
> 
> I still get an error, but a different one.
> 
>   $ createdb gurgaon_routing
>   createdb: database creation failed: ERROR:  permission denied to create 
> database
> 
> The logs say the exact same thing, what is going wrong?

Okay, this worked:

  # su - postgres
  $ psql -c 'ALTER USER user CREATEDB'
  ALTER ROLE

I guess I understand it, just wish this was clear in the docs.  I found
the answer here:

  http://www.postgresql.org/message-id/463b31d5.8040...@archonet.com

Even the official tutorial says things like:

  If you installed PostgreSQL yourself then you should log in for the
  purposes of this tutorial under the user account that you started the
  server as.

instead of the above advice!

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Creating a PostgreSQL db

[Suvayu Ali]

On Wed, Mar 09, 2016 at 12:59:49PM -0600, Michael Cronenworth wrote:
> On 03/09/2016 12:18 PM, Suvayu Ali wrote:
> >I want to create a PostgreSQL db.  So I did the following:
> >
> >   $ su -
> >   # su - postgres
> >   $ pg_ctl initdb -D /var/lib/pgsql/data
> 
> For future reference there is a helper script that can do this stuff for you.
> 
> # postgresql-setup --init-db
> 
> And when you need to upgrade the database format due to a PostgreSQL upgrade:
> 
> # postgresql-setup --upgrade

Thanks, hopefully I'll remember the next time.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Creating a PostgreSQL db

[Suvayu Ali]

Hi Gordon,

On Wed, Mar 09, 2016 at 10:50:03AM -0800, Gordon Messmer wrote:
> On 03/09/2016 10:18 AM, Suvayu Ali wrote:
> >Shouldn't the local line take care of my permissions?  What am I doing wrong?
> 
> The only user that exists by default is "postgres".  Use that use to create
> new users and databases:
> 
> su postgres -c "createuser x"

I still get an error, but a different one.

  $ createdb gurgaon_routing
  createdb: database creation failed: ERROR:  permission denied to create 
database

The logs say the exact same thing, what is going wrong?

Thanks,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Creating a PostgreSQL db

[Suvayu Ali]

Hi,

I want to create a PostgreSQL db.  So I did the following:

  $ su -
  # su - postgres
  $ pg_ctl initdb -D /var/lib/pgsql/data
  $ exit
  # systemctl start postgresql
  # exit
  $ createdb mydb
  createdb: could not connect to database template1: FATAL:  role "user" does 
not exist

I don't understand this since /var/lib/pgsql/data/pg_hba.conf has these
lines:

  # TYPE  DATABASEUSERADDRESS METHOD
  local   all all trust
  hostall all 127.0.0.1/32trust
  hostall all ::1/128 trust

Shouldn't the local line take care of my permissions?  What am I doing wrong?

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Fedora mirrors in India

[Suvayu Ali]

Hi,

Lately I have been having issues while updating.  Regularly I encounter
something like this:

[MIRROR] glibc-2.22-7.fc23_2.22-9.fc23.i686.drpm: Curl error (28): Timeout was 
reached for 
ftp://ftp.jaist.ac.jp/pub/Linux/Fedora/updates/23/x86_64/drpms/glibc-2.22-7.fc23_2.22-9.fc23.i686.drpm
 [Connection time-out]
[MIRROR] glibc-2.22-7.fc23_2.22-9.fc23.x86_64.drpm: Curl error (28): Timeout 
was reached for 
ftp://ftp.jaist.ac.jp/pub/Linux/Fedora/updates/23/x86_64/drpms/glibc-2.22-7.fc23_2.22-9.fc23.x86_64.drpm
 [Connection time-out]
[MIRROR] NetworkManager-openvpn-1.0.8-1.fc23_1.0.8-2.fc23.x86_64.drpm: Curl 
error (28): Timeout was reached for 
ftp://ftp.jaist.ac.jp/pub/Linux/Fedora/updates/23/x86_64/drpms/NetworkManager-openvpn-1.0.8-1.fc23_1.0.8-2.fc23.x86_64.drpm
 [Connection time-out]
[MIRROR] glibc-2.22-7.fc23_2.22-9.fc23.i686.drpm: Curl error (28): Timeout was 
reached for 
ftp://ftp.kddilabs.jp/Linux/packages/fedora/updates/23/x86_64/drpms/glibc-2.22-7.fc23_2.22-9.fc23.i686.drpm
 []
[MIRROR] NetworkManager-openvpn-1.0.8-1.fc23_1.0.8-2.fc23.x86_64.drpm: Curl 
error (28): Timeout was reached for 
ftp://ftp.kddilabs.jp/Linux/packages/fedora/updates/23/x86_64/drpms/NetworkManager-openvpn-1.0.8-1.fc23_1.0.8-2.fc23.x86_64.drpm
 []
(1/9): NetworkManager-openvpn-1.0.8-1.fc23_1.0.8-2.fc23.x86_64.drpm 
   
472  B/s |  57 kB 02:03
(2/9): glibc-2.22-7.fc23_2.22-9.fc23.i686.drpm  
   
4.8 kB/s | 615 kB 02:08
[DRPM] NetworkManager-openvpn-1.0.8-1.fc23_1.0.8-2.fc23.x86_64.drpm: done
[MIRROR] glibc-2.22-7.fc23_2.22-9.fc23.x86_64.drpm: Curl error (28): Timeout 
was reached for 
ftp://ftp.kddilabs.jp/Linux/packages/fedora/updates/23/x86_64/drpms/glibc-2.22-7.fc23_2.22-9.fc23.x86_64.drpm
 [Connection time-out]
[MIRROR] glibc-devel-2.22-7.fc23_2.22-9.fc23.x86_64.drpm: Curl error (28): 
Timeout was reached for 
ftp://ftp.jaist.ac.jp/pub/Linux/Fedora/updates/23/x86_64/drpms/glibc-devel-2.22-7.fc23_2.22-9.fc23.x86_64.drpm
 [Connection time-out]
(3/9): glibc-2.22-7.fc23_2.22-9.fc23.x86_64.drpm
   
2.2 kB/s | 545 kB 04:03
[DRPM] glibc-2.22-7.fc23_2.22-9.fc23.i686.drpm: done

So I checked, and there is not a single mirror in India!  

  https://admin.fedoraproject.org/mirrormanager/mirrors/Fedora/23

This is surprising since several technology institutes used to host
Fedora mirrors.  Does anyone have any ideas what happened to them?  Any
suggestions as to how I can instruct dnf to use mirrors closer to home
(e.g. prefer south-east Asia over Asia-Pacific) would be greatly
appreciated.

Thanks for any ideas.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: F22 to F23: dnf system-upgrade failure

[Suvayu Ali]

Hi,

On Wed, Feb 03, 2016 at 12:47:33PM +0100, Ger van Dijck wrote:
> Op Wed, 03 Feb 2016 11:08:45 +0100 schreef Honza Šilhan
> >
> >The same happened to me too. F21->f22 upgrade went fine but when
> >updating from F22 to F23 I was not able to boot into system upgrade
> >splash screen. Maybe it's related to this [1] bug report.
> >
> >[1] https://bugzilla.redhat.com/show_bug.cgi?id=1295213

My partition had plenty of space.  I think it fell into some kind of
dependency resolution loop because of not realising that the older
versions of dnf and systemd need not be protected.

>  I had the same experience : I think the only solution of this problem ,
> when it occurs , is a fresh install of Fedora 23 and a lot of irritation.

I managed to fix everything by removing all the old versions of the
regular packages using distro-sync --allowerasing, and then removing the
protected versions with simple dnf erase.  During the upgrade the kernel
packages failed half way through, so I had to reinstall those using with
dnf reinstall, and only kernel-core with rpm -i --force.  Everything
thing seems in place now.

Fresh install was not an option for me, I have too many things setup.
Maybe I should write a kickstart file one of these days, just in case I
need it ;).

Not sure how I can file a bug report though, I really needed the laptop,
couldn't wait to go through the whole systematic debugging process.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

F22 to F23: dnf system-upgrade failure

[Suvayu Ali]

Hi all,

I tried to upgrade to F23 with dnf system-upgrade last night.  The
download part went smoothly.  But it failed during the reboot stage.  On
booting again, I noticed grub has not been updated, that's when I
realised the upgrade had failed.  In the journal I see something like
this:

python3[769]: Starting system upgrade. This will take a while.
dnf[769]: Dependencies resolved.
dnf[769]: Error: The operation would result in removing the following protected 
packages: dnf, systemd.
systemd[1]: dnf-system-upgrade.service: Main process exited, code=exited, 
status=1/FAILURE
systemd[1]: dnf-system-upgrade.service: Failed to load environment files: No 
such file or directory
systemd[1]: dnf-system-upgrade.service: Failed to run 'stop-post' task: No such 
file or directory
systemd[1]: dnf-system-upgrade.service: Unit entered failed state.
audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 
subj=system_u:system_r:init_t:s0 msg='unit=dnf-system-upgrade comm="systemd" 
exe="/usr/lib/systemd" hostname=? addr=? terminal=? res=failed'
systemd[1]: dnf-system-upgrade.service: Failed with result 'resources'.
systemd[1]: Rebooting as result of failure.

When I tried distro-sync with --allowerasing, I encounter the exact same
error.  I found a similar unresolved bug report (with incomplete
information), but no other results turned up from a quick search.  Any
thoughts?

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: F23 on Raspberry Pi 2

[Suvayu Ali]

On Tue, Dec 29, 2015 at 12:22:54AM -0500, Robert Moskowitz wrote:
> 
> 
> On 12/29/2015 12:19 AM, Robert Moskowitz wrote:
> >
> >
> >On 12/26/2015 06:50 AM, Suvayu Ali wrote:
> >>Hi Michael,
> >>
> >>On Thu, Dec 24, 2015 at 11:58:49PM +0000, M A Young wrote:
> >>>On Thu, 24 Dec 2015, Suvayu Ali wrote:
> >>>>It seems to me, the fedora kernel is not running.  Any ideas how I can
> >>>>get it to run?
> >>>As far as I am aware you can't use a Fedora kernel on a Raspberry Pi
> >>>2, as
> >>>it needs some modifications that the Fedora Kernel doesn't have
> >>>(unless it
> >>>has changed recently). One thing to check is that you have the
> >>>NetworkManager-wifi package, probably the wpa_supplicant package as
> >>>well
> >>>as I don't think they are in the minimal image.
> >>Thanks for the information.  I'll try to find more information. Perhaps
> >>it's simpler to try a distribution that is better supported on the Pi.
> >
> >Or going to the fedora-arm list to ask about arm support?
> 
> Or start with going through the archive there looking for raspberry.

I tried looking at the archives.  I eventually went with Arch Linux.  It
was very easy to install, and runs rather nicely.  The Fedora version I
had by following the instructions in the blogpost mentioned earlier in
the thread felt very sluggish compared to the snappiness of the Arch
install so I think I'll stick with this.  Given that this is an RPi,
it's not surprising.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: F23 on Raspberry Pi 2

[Suvayu Ali]

Hi Michael,

On Thu, Dec 24, 2015 at 11:58:49PM +, M A Young wrote:
> On Thu, 24 Dec 2015, Suvayu Ali wrote:
> > 
> > It seems to me, the fedora kernel is not running.  Any ideas how I can
> > get it to run?
> 
> As far as I am aware you can't use a Fedora kernel on a Raspberry Pi 2, as 
> it needs some modifications that the Fedora Kernel doesn't have (unless it 
> has changed recently). One thing to check is that you have the 
> NetworkManager-wifi package, probably the wpa_supplicant package as well 
> as I don't think they are in the minimal image.

Thanks for the information.  I'll try to find more information.  Perhaps
it's simpler to try a distribution that is better supported on the Pi.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: F23 on Raspberry Pi 2

[Suvayu Ali]

Hi,

On Thu, Dec 24, 2015 at 03:58:14AM -0500, Bob Goodwin wrote:
> On 12/24/15 02:11, Suvayu Ali wrote:
> >>  I just can't remember which (or if there
> >>>was even a difference)
> >>>https://chisight.wordpress.com/2015/10/19/fedora-22-or-23-on-raspberry-pi-2/
> >I followed this, and worked nicely.  I have a booting F23 system now!
> >
> >Thanks a lot.
> >
> >>>http://jonarcher.info/2015/02/getting-fedora-21-raspberry-pi-2/
> 
> You "have a booting F23 system," do you have a usable desktop display? I
> have followed the jonarcher procedure several times but always wind up with
> a nice XFCE desktop display but no keyboard or mouse control, it just does
> nothing at that point. I have been trying each time I see something
> encouraging written about it but without success and have pretty much given
> up hope for the RPI2b and Fedora ...

I am using the minimal image, so no, I do not have a desktop.  But I do
have a working keyboard and ethernet.  Did firstboot run for you?  If
not, that would explain the issue for you.

I'm having trouble getting wireless to work though.  I know my adapter
works with the RPi since I have another RPi running OSMC using the same
adapter without issues.  I tested the adapter with my laptop running
F22, that worked nicely.  So I know it also works with Fedora x86_64.
However it fails to work on the RPi with Fedora ARM.  I checked the
respective module is loaded 8192cu, although I do not see any rtlwifi.
Trying to load it with modprobe gives me a module not found error.

It seems to me, the fedora kernel is not running.  Any ideas how I can
get it to run?

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

F23 on Raspberry Pi 2

[Suvayu Ali]

Hi,

I was looking at how to get Fedora on an Rpi 2, but then I realised the
latest Pidora version is an F20 based RC from 2014.  Is this all we
have?  Maybe someone knows a way to install from the latest ARM images.
The official ARM installer does not seem to support Rpi.  This is the
list:

- A10-OLinuXino-Lime
- A20-OLinuXino-Lime
- A20-OLinuXino-Lime2
- A20-OLinuXino_MICRO
- am335x_boneblack
- Bananapi
- Bananapro
- cm_fx6
- Cubieboard
- Cubieboard2
- Cubietruck
- jetson-tk1
- Linksprite_pcDuino
- Linksprite_pcDuino3
- Linksprite_pcDuino3_Nano
- Mele_A1000
- Mini-X
- mx6cuboxi
- none
- novena
- omap4_panda
- omap5_uevm
- riotboard
- trimslice
- wandboard

I tried none, it doesn't install U-boot and then the Pi doesn't boot.
Thanks for any other ideas.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: F23 on Raspberry Pi 2

[Suvayu Ali]

Hi Tom,

On Wed, Dec 23, 2015 at 03:35:28PM -0500, Tom wrote:
> I have both of these bookmarked. I've used one of the procedures
> recently and it went smooth. I just can't remember which (or if there
> was even a difference)
> https://chisight.wordpress.com/2015/10/19/fedora-22-or-23-on-raspberry-pi-2/

I followed this, and worked nicely.  I have a booting F23 system now!

Thanks a lot.

> http://jonarcher.info/2015/02/getting-fedora-21-raspberry-pi-2/

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: xfce without a mouse

[Suvayu Ali]

On Tue, Dec 08, 2015 at 08:59:34PM -0700, Chris Murphy wrote:
> Hi,
> 
> So I have only a bluetooth mouse, and GNOME lets me navigate by
> keyboard only to get to Settings well enough I can then pair with the
> mouse. I can't figure out how to do this at all on xfce (granted this
> is Fedora 20 because that's the build Intel is using for their CPU
> diagnostic tool).

The Alt+F2 run dialog should work.  If you don't remember the name of
the program, you can use down arrow to activate the XFCE menu and
search.  You might also have some luck with Alt+F1 and get the desktop
right click menu.  I have the standard applications menu in this desktop
context menu, I don't know whether that is the default though.  

Once you have launched the desired application, it should be possible to
navigate with some combination of Alt, Shift, arrow keys, etc.

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: MP4 video on F23

[Suvayu Ali]

On Wed, Dec 09, 2015 at 12:02:31PM +, Patrick O'Callaghan wrote:
> On Wed, 2015-12-09 at 12:52 +0100, Ralf Corsepius wrote:
> 
> > I.e. you have other repos in addition to these active and now likely
> > are facing compatiblity issues between them.
> 
> I see no indication of that. Compatibility issues should show up when
> updating, but they don't. I run "dnf update" update every day.

Could you verify the signatures of the RPMs you have installed.  On my
system I have this:

$ rpm -qi vlc | grep Signature  # same for mplayer and ffmpeg
Signature   : RSA/SHA1, Wednesday 07 October 2015 11:01:18 PM IST, Key ID 
81c9b42397f4d1c1

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Cups fails with "Filter failed"

[Suvayu Ali]

Hi Tim,

On Thu, Dec 03, 2015 at 11:22:19PM +1030, Tim wrote:
> Allegedly, on or about 03 December 2015, Suvayu Ali sent:
> > I restarted cups to make sure the plugins are loaded
> > properly.  Now printing a test page tells me this:
> > 
> >   processing since
> >   Thu 03 Dec 2015 05:41:19 PM IST
> >   "The printer is in use."
> > 
> > What does that mean? 
> 
> You could go kill a/any "jobs" stuck waiting in the queue.
> Web interface http://localhost:631/

It started printing after I restarted my computer today.  I guess that
cleared the queue.  I guess the solution for me was installing all the
plugins as Jon suggested, and clearing the queue.  Strangely, when
printing a test page, it prints 2 copies.  Although when printing files,
it prints one.  I can live with that.

Thanks everyone!

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Cups fails with "Filter failed"

[Suvayu Ali]

Hi Ed,

On Thu, Dec 03, 2015 at 02:30:57PM +0800, Ed Greshko wrote:
> 
> On 12/03/15 14:20, Suvayu Ali wrote:
> > Hi,
> >
> > I'm trying to print a test page to my new network printer (HP Laserjet
> > Pro MFP 126nw) but it keeps failing with the message: stopped "Filter
> > failed".
> >
> > systemctl says I should look in /var/log/cups/error_log for more
> > details, but my error_log is empty, and the last modified time is from
> > last year!  So I changed LogLevel to debug in cupsd.conf and restarted
> > the service.  This prints all sorts of debug messages to the journal,
> > although none to error_log.  Strangely enough, the journal still says:
> > [Job 325] Job stopped due to filter errors; please consult the error_log
> > file for details.
> >
> > I have checked all the "common cures" I found on the Internet, size of
> > /var/spool/cups (<700k for me), device permissions (doesn't apply as I'm
> > printing over the network).  I have always printed over the network (to
> > other printers), so I'm not sure what is going wrong this time.
> >
> > Any thoughts what is the problem?  Any ideas I could try?
> 
> How did you install the printer?  Looking here
> 
> http://hplipopensource.com/hplip-web/models/laserjet/hp_laserjet_pro_mfp_m126nw.html
> 
> suggests "This printer REQUIRES a downloadable driver plug-in"

I used the standard Fedora gui interface to find the printer on my
network, choose IPP, and follow all the prompts to find the printer make
and model, and said yes to the recommended driver, screenshot here:
https://dl.dropboxusercontent.com/u/37474389/add_HP_LJP_MFP_126nw.png

My understanding was the fedora gui knows how to download any blobs
necessary for known printers in the database.

Am I missing something?

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Cups fails with "Filter failed"

[Suvayu Ali]

On Thu, Dec 03, 2015 at 01:22:39PM +0200, Jonathan Dieter wrote:
> On Thu, 2015-12-03 at 14:37 +0530, Suvayu Ali wrote:
> > I used the standard Fedora gui interface to find the printer on my
> > network, choose IPP, and follow all the prompts to find the printer
> > make
> > and model, and said yes to the recommended driver, screenshot here:
> > https://dl.dropboxusercontent.com/u/37474389/add_HP_LJP_MFP_126nw.png
> > 
> > My understanding was the fedora gui knows how to download any blobs
> > necessary for known printers in the database.
> > 
> > Am I missing something?
> 
> I believe it will theoretically, but network printers seem to work
> differently.  At the terminal, as root, try typing:
> # hp-plugin -i
> 
> This should walk you through downloading and installing the binary
> plugin required by HP printers.

Okay, I did that.  I restarted cups to make sure the plugins are loaded
properly.  Now printing a test page tells me this:

  processing since
  Thu 03 Dec 2015 05:41:19 PM IST
  "The printer is in use."

What does that mean?

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Cups fails with "Filter failed"

[Suvayu Ali]

Hi,

I'm trying to print a test page to my new network printer (HP Laserjet
Pro MFP 126nw) but it keeps failing with the message: stopped "Filter
failed".

systemctl says I should look in /var/log/cups/error_log for more
details, but my error_log is empty, and the last modified time is from
last year!  So I changed LogLevel to debug in cupsd.conf and restarted
the service.  This prints all sorts of debug messages to the journal,
although none to error_log.  Strangely enough, the journal still says:
[Job 325] Job stopped due to filter errors; please consult the error_log
file for details.

I have checked all the "common cures" I found on the Internet, size of
/var/spool/cups (<700k for me), device permissions (doesn't apply as I'm
printing over the network).  I have always printed over the network (to
other printers), so I'm not sure what is going wrong this time.

Any thoughts what is the problem?  Any ideas I could try?

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

dnf gets stuck on poor connections

[Suvayu Ali]

Hi,

Lately I have been on unreliable connections.  I noticed that after I
have my laptop running for a few days, dnf stops working.  Any dnf
command hangs indefinitely.  I can get back to the shell again by
killing the process with SIGKILL.  It seems to me this happens when the
dnf-makecache service gets stuck.  I have tried restarting it, but it
does not help.  Restarting the dnf-makecache timer does not help either.
Of course all this goes away after a reboot.  But then that's not really
a solution.

Any ideas anyone?

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf gets stuck on poor connections

[Suvayu Ali]

Hi Sylvia, Ralf,

On Fri, Nov 13, 2015 at 01:22:29PM +0100, Ralf Corsepius wrote:
> On 11/13/2015 12:53 PM, Sylvia Sánchez wrote:
> >Mmmhh... My connection isn't very reliable either but dnf doesn't get stuck.
> >Maybe you should update packages in small bunches instead of
> >everything altogether.

It's not just commands where access to the repo is needed.  Even
commands like search hangs.  I think the makecache service is locking
the database and hanging.  Since it hangs, my interactive command waits
for the lock to be released indefinitely.  I can't prove this though,
this is my hunch based on the symptoms I see.

> I am not on a poor connection, either, but I am occasionally experiencing
> this problem as well.

Okay good, I'm not alone then.

> In most cases the cause seems to be dnf's mirror selection to prefer broken
> repos and/or poorly accessible mirrors, but I have also seen cases, when dnf
> hung without any feedback for hours.

Simple search or clean commands should not hang if this was the reason.

Thanks for the responses so far, any other thoughts?

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: btrfs crashing at boot

[Suvayu Ali]

Hi Chris,

On Tue, Oct 20, 2015 at 01:00:53PM +0200, Chris Murphy wrote:
> Ah yes here we go, from today even more info on status of convert.
> Gist is that it is broken and it's not certain when it will be fixed,
> hopefully for 4.3 kernel and progs (it most likely will take both
> since convert is part of btrfs-progs is responsible for reading ext4,
> and then the kernel is responsible for writing the data+metadata to
> Btrfs).
> 
> http://www.spinics.net/lists/linux-btrfs/msg48330.html

Thanks for the pointers to the status of convert.  This is not critical
data, so it's alright even if something goes wrong.  Your rsync options
will be very helpful for a future move to btrfs for my /.

In the meantime, I managed to resolve my problems.  I have several
multi-GB files on this drive, I had deleted a few large files from the
gui (meaning, they were still in lost+found).  That's when the problem
started.  I could see them in the l+f, but could not remove or restore
them.  Since the filesystem was rw for the first few seconds, I resorted
to a kludge: mount and try to empty the l+f before it turns ro, unmount,
rinse repeat.  Each time I tried, I removed a few of the files.  It took
me maybe 4-5 tries to clean about 10 files.  Ever since I did this, I
have had no problems.  I plan to run a scrub sometime soon, maybe
tonight.

Thanks again, without your warning, I might have run convert on my /.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

btrfs crashing at boot

[Suvayu Ali]

Hi,

I started using btrfs recently.  Sometime today, my scratch disk started
turning read-only.  journalctl reports btrfs crashed (output shown
below).  ABRT says the problem cannot be reported.  I don't understand
this since I blacklisted the only module I'm aware of that might taint
the kernel on my system (openafs).  This happens within the first few
minutes of boot.  Immediately after boot I can write a simple file, but
soon after it crashes.  When I try to run btrfs check (after
unmounting), I get a core dump:

  # btrfs check /dev/sda1
  Checking filesystem on /dev/sda1
  UUID: 19ec6a44-fc5b-428d-9aec-fda803ce805b
  checking extents
  Aborted (core dumped)

I can't find any core file in the present working directory either.

In case it is relevant, this filesystem was converted from ext4.  What
could be going wrong?  How can I debug this?

Thanks for any ideas.

Cheers,


BTRFS detecting my disk:

Oct 19 18:30:09 localhost kernel: BTRFS: device fsid 
19ec6a44-fc5b-428d-9aec-fda803ce805b devid 1 transid 1336 /dev/sda1
Oct 19 18:30:09 localhost kernel: BTRFS info (device sda1): disk space caching 
is enabled
Oct 19 18:30:09 localhost kernel: BTRFS: has skinny extents

BTRFS crash log:

Oct 19 18:33:59 localhost kernel: [ cut here ]
Oct 19 18:33:59 localhost kernel: WARNING: CPU: 1 PID: 152 at 
fs/btrfs/extent-tree.c:6255 __btrfs_free_extent.isra.68+0x8c8/0xd70 [btrfs]()
Oct 19 18:33:59 localhost kernel: Modules linked in: ccm ip6t_rpfilter 
ip6t_REJECT nf_reject_ipv6 xt_conntrack ebtable_nat ebtable_broute bridge stp 
llc ebtable_filter ebtables ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 
nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6table_filter 
ip6_tables iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat 
nf_conntrack iptable_mangle iptable_security iptable_raw btrfs arc4 xor 
intel_rapl iosf_mbi x86_pkg_temp_thermal iTCO_wdt coretemp kvm_intel 
iTCO_vendor_support kvm raid6_pq iwlmvm crct10dif_pclmul snd_hda_codec_hdmi 
crc32_pclmul crc32c_intel mac80211 snd_hda_codec_realtek snd_hda_codec_generic 
ghash_clmulni_intel uvcvideo snd_hda_intel snd_hda_codec snd_hda_core snd_hwdep 
videobuf2_vmalloc snd_seq videobuf2_core videobuf2_memops snd_seq_device 
v4l2_common
Oct 19 18:33:59 localhost kernel:  iwlwifi snd_pcm videodev rtsx_pci_ms 
cfg80211 media thinkpad_acpi memstick i2c_i801 lpc_ich snd_timer snd tpm_tis 
wmi rfkill tpm soundcore mei_me mei shpchp binfmt_misc i915 rtsx_pci_sdmmc 
mmc_core i2c_algo_bit e1000e drm_kms_helper drm serio_raw ptp rtsx_pci pps_core 
video
Oct 19 18:33:59 localhost kernel: CPU: 1 PID: 152 Comm: kworker/u16:5 Tainted: 
GW   4.2.3-200.fc22.x86_64 #1
Oct 19 18:33:59 localhost kernel: Hardware name: LENOVO 20ALCTO1WW/20ALCTO1WW, 
BIOS GIET66WW (2.16 ) 12/02/2013
Oct 19 18:33:59 localhost kernel: Workqueue: btrfs-extent-refs 
btrfs_extent_refs_helper [btrfs]
Oct 19 18:33:59 localhost kernel:   673d9c47 
8802125d7a88 8177220a
Oct 19 18:33:59 localhost kernel:    
8802125d7ac8 8109e4a6
Oct 19 18:33:59 localhost kernel:  0002 0001874d9000 
fffe 
Oct 19 18:33:59 localhost kernel: Call Trace:
Oct 19 18:33:59 localhost kernel:  [] dump_stack+0x45/0x57
Oct 19 18:33:59 localhost kernel:  [] 
warn_slowpath_common+0x86/0xc0
Oct 19 18:33:59 localhost kernel:  [] 
warn_slowpath_null+0x1a/0x20
Oct 19 18:33:59 localhost kernel:  [] 
__btrfs_free_extent.isra.68+0x8c8/0xd70 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] ? 
find_ref_head+0x5a/0x80 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] 
__btrfs_run_delayed_refs+0x998/0x1080 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] ? 
kmem_cache_alloc+0x193/0x210
Oct 19 18:33:59 localhost kernel:  [] 
btrfs_run_delayed_refs.part.73+0x74/0x270 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] 
delayed_ref_async_start+0x7e/0x90 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] 
btrfs_scrubparity_helper+0xc2/0x260 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] 
btrfs_extent_refs_helper+0xe/0x10 [btrfs]
Oct 19 18:33:59 localhost kernel:  [] 
process_one_work+0x19e/0x3f0
Oct 19 18:33:59 localhost kernel:  [] worker_thread+0x4e/0x450
Oct 19 18:33:59 localhost kernel:  [] ? 
process_one_work+0x3f0/0x3f0
Oct 19 18:33:59 localhost kernel:  [] ? 
process_one_work+0x3f0/0x3f0
Oct 19 18:33:59 localhost kernel:  [] kthread+0xd8/0xf0
Oct 19 18:33:59 localhost kernel:  [] ? 
kthread_worker_fn+0x160/0x160
Oct 19 18:33:59 localhost kernel:  [] ret_from_fork+0x3f/0x70
Oct 19 18:33:59 localhost kernel:  [] ? 
kthread_worker_fn+0x160/0x160
Oct 19 18:33:59 localhost kernel: ---[ end trace 0eadb91546850dac ]---
Oct 19 18:33:59 localhost kernel: BTRFS info (device sda1): leaf 139591680 
total ptrs 184 free space 5551
Oct 19 18:33:59 localhost kernel: item 0 key (6554550272 169 0) itemoff 
16250 itemsize 33
Oct 19 18:33:59 localhost kernel: extent refs 1 gen 355 

Re: Copying files without losing selinux context

[Suvayu Ali]

On Sun, Oct 11, 2015 at 07:17:59AM -0400, Daniel J Walsh wrote:
> On 10/10/2015 05:07 AM, Suvayu Ali wrote:
> > On Sat, Oct 10, 2015 at 12:31:59PM +0530, Rejy M Cyriac wrote:
> >
> >> The following would have retained the SELinux contexts
> >>
> >> rsync with the --xattrs option
> >> tar with the --selinux or --xattrs option
> > Thanks a lot!  I'll remember this for the future.  Is there any simple
> > way to restore the contexts now, after the fact?  If not, maybe
> > something like the command below?
> >
> >   # cd /old && find . -exec chcon --reference=\{\} /var/\{\} \;
> >
> If you are moving content around you should reset the default labeling. 
> In this case you could do something like
> 
> # semanage fcontext -a -e /var /old
> # restorecon -R -v /old
> 
> Which would make your labels survive a relabel

Thanks a lot Dan.  Seems to have worked, presuming the warnings can be ignored:

  # semanage fcontext -a -e /var /mnt
  # restorecon -R -v /mnt
  restorecon:  Warning no default label for /mnt/lib/nfs/rpc_pipefs
  restorecon:  Warning no default label for /mnt/spool/cron/user

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Copying files without losing selinux context

[Suvayu Ali]

Hi Rejy,

On Sat, Oct 10, 2015 at 12:31:59PM +0530, Rejy M Cyriac wrote:
> On 10/08/2015 06:35 PM, Suvayu Ali wrote:
> > 
> > Yesterday I installed a new SSD in my laptop.  I moved all my files
> > (/home, /var, /opt) with rsync and rebooted.  However I see the selinux
> > filecontexts are wrong, and many services are failing because of that,
> > e.g. the user crontab doesn't load.
> > 
> >   # ls -Z /var/spool/cron/user
> >   unconfined_u:object_r:var_spool_t:s0 /var/spool/cron/user
> > 
> > I did an autorelabel on boot, I also ran `restorecon -p -r /var',
> > neither helped.  To get the crontab working, I had to change the context
> > by hand.
> > 
> >   # chcon --reference=/old/part/spool/cron/user /var/spool/cron/user
> >   # ls -Z /var/spool/cron/user 
> >   unconfined_u:object_r:user_cron_spool_t:s0 /var/spool/cron/user
> > 
> > I would like to know how I can fix the rest, and what I should have used
> > to do the copy in the first place.  I guess `cp -c' would work, but then
> > I wouldn't have the ability to resume the transfer.
> 
> The following would have retained the SELinux contexts
> 
> rsync with the --xattrs option
> tar with the --selinux or --xattrs option

Thanks a lot!  I'll remember this for the future.  Is there any simple
way to restore the contexts now, after the fact?  If not, maybe
something like the command below?

  # cd /old && find . -exec chcon --reference=\{\} /var/\{\} \;

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: YouTube videos in full-screen mode block Firefox

[Suvayu Ali]

On Thu, Oct 08, 2015 at 07:35:20PM +0100, Paul Smith wrote:
> On Thu, Oct 8, 2015 at 7:13 PM, Joe Zeff  wrote:
> >>
> >> Thanks, Doug. It gets full-screen immediately, but with the image and
> >> sound both frozen for one minute or so. Also Firefox freezes for the
> >> same time.
> >
> > Does the window go dark?  If so, that indicates that Firefox is very, very
> > busy; if not, either you don't have things set for that or it's something
> > else causing this.
> 
> Thanks, Joe. You pointed to the right direction: too many open tabs
> (hundreds, I mean) were causing the issue.

My hunch is you were swapping which caused the issue.  Hundreds of tabs
should not be a problem.  I regularly have that many tabs open without
any problems (197 at the moment).  It could be one of the tabs was
consuming a lot of memory forcing FF to swapping.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Copying files without losing selinux context

[Suvayu Ali]

Hi,

Yesterday I installed a new SSD in my laptop.  I moved all my files
(/home, /var, /opt) with rsync and rebooted.  However I see the selinux
filecontexts are wrong, and many services are failing because of that,
e.g. the user crontab doesn't load.

  # ls -Z /var/spool/cron/user
  unconfined_u:object_r:var_spool_t:s0 /var/spool/cron/user

I did an autorelabel on boot, I also ran `restorecon -p -r /var',
neither helped.  To get the crontab working, I had to change the context
by hand.

  # chcon --reference=/old/part/spool/cron/user /var/spool/cron/user
  # ls -Z /var/spool/cron/user 
  unconfined_u:object_r:user_cron_spool_t:s0 /var/spool/cron/user

I would like to know how I can fix the rest, and what I should have used
to do the copy in the first place.  I guess `cp -c' would work, but then
I wouldn't have the ability to resume the transfer.

Thanks for any help.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: DNF: Only one kernel-devel package allowed?

[Suvayu Ali]

On Sun, Oct 04, 2015 at 08:06:42AM -0500, Richard Shaw wrote:
> I'm assuming this is DNF and not some other change...
> 
> Yum allowed multiple kernel-devel packages to be installed, dnf does not,
> which prevents you from building kernel modules for other kernel releases.
> 
> Was there a good reason for this change in behavior?

This is not true.

  $ rpm -q kernel-devel | sort
  kernel-devel-4.0.4-202.fc21.x86_64
  kernel-devel-4.1.5-200.fc22.x86_64
  kernel-devel-4.1.6-200.fc22.x86_64
  kernel-devel-4.1.6-201.fc22.x86_64
  kernel-devel-4.1.7-200.fc22.x86_64
  kernel-devel-4.1.8-200.fc22.x86_64

Maybe you are confusing with kernel-headers?  If that is the case, it
has been like that for ages.  And usually that should not be a problem,
specially between minor releases (as is the case for a given Fedora
release).

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: DNF: Only one kernel-devel package allowed?

[Suvayu Ali]

On Mon, Oct 05, 2015 at 05:45:09AM +0800, Ed Greshko wrote:
> On 10/05/2015 05:21 AM, Suvayu Ali wrote:
> > On Sun, Oct 04, 2015 at 08:06:42AM -0500, Richard Shaw wrote:
> >> I'm assuming this is DNF and not some other change...
> >>
> >> Yum allowed multiple kernel-devel packages to be installed, dnf does not,
> >> which prevents you from building kernel modules for other kernel releases.
> >>
> >> Was there a good reason for this change in behavior?
> > This is not true.
> >
> >   $ rpm -q kernel-devel | sort
> >   kernel-devel-4.0.4-202.fc21.x86_64
> >   kernel-devel-4.1.5-200.fc22.x86_64
> >   kernel-devel-4.1.6-200.fc22.x86_64
> >   kernel-devel-4.1.6-201.fc22.x86_64
> >   kernel-devel-4.1.7-200.fc22.x86_64
> >   kernel-devel-4.1.8-200.fc22.x86_64
> >
> > Maybe you are confusing with kernel-headers?  If that is the case, it
> > has been like that for ages.  And usually that should not be a problem,
> > specially between minor releases (as is the case for a given Fedora
> > release).
> >
> 
> I think you had not read the entire thread before you responded.  We later 
> discovered him to be on F21 and indeed on my F21 VM I confirmed

Indeed!  I noticed after I had responded.  I guess a backport bug report
is in order against dnf.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: fedora 23

[Suvayu Ali]

On Tue, Sep 29, 2015 at 02:10:27PM -0400, Paul Cartwright wrote:
> On 09/29/2015 12:35 PM, Matthew Miller wrote:
> > Like this, by the way: 
> >
> >   rpm -qa --queryformat='%{name}\n'
> that gets rid of the version #'s for each app... hmm.. I like that
> 
>   rpm -qa --queryformat='%{name}\n' > fedora_installed_apps

You might want to add arch if you have a multilib system.

  $ rpm -qa --queryformat='%{name}.%{arch}\n'

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: fedora 23

[Suvayu Ali]

On Wed, Sep 30, 2015 at 05:54:25AM -0400, Paul Cartwright wrote:
> On 09/30/2015 04:14 AM, Suvayu Ali wrote:
> >>> > >   rpm -qa --queryformat='%{name}\n'
> >> > that gets rid of the version #'s for each app... hmm.. I like that
> >> > 
> >> >   rpm -qa --queryformat='%{name}\n' > fedora_installed_apps
> > You might want to add arch if you have a multilib system.
> >
> >   $ rpm -qa --queryformat='%{name}.%{arch}\n'
> >
> > Cheers,
> that doesn't work.. it leaves the .I686 /x86_64/.noarch on..
> 
> adobe-source-han-sans-twhk-fonts.noarch
> readline.i686
> man-pages.noarch
> harfbuzz.i686
> nfs-utils.x86_64
> krb5-libs.i686

That's kinda the point.  On a multilib system, you will have packages
for multiple architectures.  Without the trailing .arch, when you
reinstall, only the default architecture will be installed.

Here is an example.  I have both 64 bit and 32 bit pulseaudio-libs
because of Skype.

  $ rpm -q --queryformat="%{name}.%{arch}\n" pulseaudio-libs
  pulseaudio-libs.x86_64
  pulseaudio-libs.i686

Without the trailing .arch, I will only get the x86_64 package when
reinstalling from the list.

> and the original also gets noarch:
> 
> rpm -qa --queryformat='%{name}\n' |grep adobe
> adobe-release-x86_64
> adobe-source-han-sans-cn-fonts
> adobe-source-han-sans-twhk-fonts

This will get noarch too!  Just look at your first output, it includes
man-pages.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: firefox?

[Suvayu Ali]

On Mon, Sep 28, 2015 at 12:22:59AM +, Andre Robatino wrote:
> Andre Robatino  fedoraproject.org> writes:
> 
> > Gordon Messmer  gmail.com> writes:
> > 
> > > On 09/27/2015 04:01 PM, Andre Robatino wrote:
> > > > Most of those people have the updates-testing repo enabled by default, 
> > > > so
> > > > they tend not to notice broken deps like this. Some kind of automation 
> > > > is
> > > > necessary to reliably prevent it.
> > > 
> > > Looking at bodhi, it doesn't appear that anyone was unaware that the 
> > > update required a newer sqlite, and would be delayed if that update 
> > > didn't move to stable.
> > 
> > Looking at https://bodhi.fedoraproject.org/updates/FEDORA-2015-16455 , it's
> > clear that Martin Stransky wasn't aware of it when he submitted it
> > ("Unfortunately I can't remove this update") and in any case it's policy
> > that groups of packages with mutual deps are supposed to be submitted
> > together as a single update.
> 
> Oh I see, you mean none of the testers seemed to be unaware.

I think it is expected a tester will be unaware.  To test, you enable
updates-testing and install the update.  If you do that, of course you
also pull in any dependencies from updates-testing.  This issue was a
problem of delayed push to stable of updates which are dependencies.
The problem is not as trivial as you might think.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Time Machine filesystem

[Suvayu Ali]

On Sun, Sep 27, 2015 at 06:31:22PM +0100, Patrick O'Callaghan wrote:
> 
> There's a project on github to deal with this: 
> https://github.com/abique/tmfs and I've made some progress in compiling
> it under Fedora (basically installing C++ and the devel packages for
> Boost and Fuse), however at the last make step I get this:
> 
> $ make
> 
> [ 12%] Linking CXX executable tmfs
> /usr/bin/ld: CMakeFiles/tmfs.dir/src/main.cc.o: undefined reference to symbol 
> '__cxa_call_unexpected@@CXXABI_1.3'
> /usr/lib64/libstdc++.so.6: error adding symbols: DSO missing from command line
> collect2: error: ld returned 1 exit status
> CMakeFiles/tmfs.dir/build.make:226: recipe for target 'tmfs' failed
> make[2]: *** [tmfs] Error 1
> CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/tmfs.dir/all' failed
> make[1]: *** [CMakeFiles/tmfs.dir/all] Error 2
> Makefile:127: recipe for target 'all' failed
> 
> I'm guessing a library must be missing, but which one?

Hmm, this was pretty trivial for me:

  $ git clone g...@github.com:abique/tmfs.git
  Cloning into 'tmfs'...
  remote: Counting objects: 161, done.
  remote: Total 161 (delta 0), reused 0 (delta 0), pack-reused 161
  Receiving objects: 100% (161/161), 38.38 KiB | 0 bytes/s, done.
  Resolving deltas: 100% (74/74), done.
  Checking connectivity... done.
  $ cd tmfs/
  $ ls
  AUTHORS  ChangeLog  CMakeLists.txt  LICENSE  README.markdown  src
  $ mkdir _build
  $ cd _build
  $ cmake ..
  -- The CXX compiler identification is GNU 5.1.1
  -- Check for working CXX compiler: /usr/lib64/ccache/c++
  -- Check for working CXX compiler: /usr/lib64/ccache/c++ -- works
  -- Detecting CXX compiler ABI info
  -- Detecting CXX compiler ABI info - done
  -- Detecting CXX compile features
  -- Detecting CXX compile features - done
  -- Boost version: 1.57.0
  -- Found the following Boost libraries:
  --   system
  --   filesystem
  -- Found PkgConfig: /usr/bin/pkg-config (found version "0.28") 
  -- checking for module 'fuse'
  --   found fuse, version 2.9.4
  -- Configuring done
  -- Generating done
  -- Build files have been written to: /home/jallad/build/tmfs/_build
  $ make 
  Scanning dependencies of target tmfs
  [ 14%] Building CXX object CMakeFiles/tmfs.dir/src/main.cc.o
  [ 28%] Building CXX object CMakeFiles/tmfs.dir/src/readdir.cc.o
  [ 42%] Building CXX object CMakeFiles/tmfs.dir/src/read.cc.o
  [ 57%] Building CXX object CMakeFiles/tmfs.dir/src/readlink.cc.o
  [ 71%] Building CXX object CMakeFiles/tmfs.dir/src/getattr.cc.o
  [ 85%] Building CXX object CMakeFiles/tmfs.dir/src/get_real_path.cc.o
  [100%] Linking CXX executable tmfs
  [100%] Built target tmfs

I think these are the only packages you need:
- gcc-c++
- cmake
- boost-filesystem
- fuse-devel

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Time Machine filesystem

[Suvayu Ali]

On Sun, Sep 27, 2015 at 08:02:11PM +0100, Patrick O'Callaghan wrote:
> On Sun, 2015-09-27 at 20:42 +0200, Suvayu Ali wrote:
> > I think these are the only packages you need:
> > - gcc-c++
> > - cmake
> > - boost-filesystem
> > - fuse-devel
> 
> I have all of those.
> 
> I deleted the tmfs directory and started again. This time it worked.
> I'm guessing my previous incremental attempts (get error due to missing
> package, install package, rinse and repeat) must have left some cruft
> around.

cmake does a lot of cacheing.  If you change dependencies, it is
recommended to start a fresh build.

1) delete the build directory and start again, or
2) call `cmake /path/to/source/' again.

(1) is of course prefered.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Strange SELinux issue preventing going to sleep

[Suvayu Ali]

On Fri, Sep 18, 2015 at 05:23:49PM -0600, Kevin Fenzi wrote:
> 
> There's a systemd update that needs a newer selinux-policy. 
> Both of them went out around the same time, so normally that would be
> fine. However, the systemd update needs the selinux-policy update to be
> installed when it updates, or it gets confused and you see symptoms
> like the above. ;( 
> 
> It seems to be the ordering is somewhat random, and some people have
> rpm install the selinux-policy update first and some have it apply
> after. 
> 
> You can work around it by doing a:
> 
> systemctl daemon-reexec
> 
> or just reboot as it will have the correct policy on reboot. 

Interesting problem!  Thanks for the solution, sleep works again.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Strange SELinux issue preventing going to sleep

[Suvayu Ali]

Hi,

Ever since a recent update, when I close the lid of my thinkpad, it does
not go to sleep.  When I unlock the screen, I see a desktop notification
saying something like this:

  Power Manager
  GDBus.Error:org.freedesktop.DBus.error.accessdenied:
  SELinux policy denies access.

Strangely though, I can not find any SELinux alerts in the
troubleshooter.  Looking at the updated packages, I find this:
selinux-policy-3.13.1-128.13.fc22.noarch.  What is going on?  How can I
debug this to file a bug report?

Thanks for any ideas.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Very SLOW system update

[Suvayu Ali]

On Sun, Sep 13, 2015 at 12:55:36AM -0500, Ranjan Maitra wrote:
> Just curious: what is the thinking behind setting fastestmirror=0 by default? 
> Is it to get around the overhead for calculating the fastest mirror each time.

Historicall, when it was available with yum, fastestmirror was the
source of some other related problems.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Removing LXDE

[Suvayu Ali]

On Tue, Sep 01, 2015 at 08:14:32AM +0200, Martin Skjöldebrand wrote:
> On Tue, 2015-09-01 at 07:53 +0200, Luigi Votta wrote:
> > 
> > sudo dnf history rollback 
> 
> Yes in this case it worked. I discovered you need to do it immediately
> after you messed up, or you will remove *all* transaction back to the
> one you wish to remove. (I.e. it restores the install history to the
> state before your mistake).

You can undo only individual transactions with:

  # dnf history undo 

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: rsync checksum to compare directories

[Suvayu Ali]

On Sun, Aug 30, 2015 at 06:52:20PM -0600, Chris Murphy wrote:
> On Sun, Aug 30, 2015 at 6:29 PM, Suvayu Ali <fatkasuvayu+li...@gmail.com> 
> wrote:
> > On Mon, Aug 31, 2015 at 02:25:42AM +0200, Suvayu Ali wrote:
> >> On Sun, Aug 30, 2015 at 01:17:00PM -0600, Chris Murphy wrote:
> >> > On Sun, Aug 30, 2015 at 1:01 PM, Gordon Messmer
> >> > <gordon.mess...@gmail.com> wrote:
> >> > > On 08/30/2015 10:30 AM, Chris Murphy wrote:
> >> > >>
> >> > >> # rsync -rvnc --delete /brick0 /brick1
> >> > >>
> >> > >> I've confirmed that separate devices are mounted to /brick0 and
> >> > >> /brick1. What I'm seeing for ~ 20 minutes now is only the device
> >> > >> mounted at brick0 is being accessed, and yet I'm seeing piles of files
> >> > >> (verbose option) being listed.
> >> > >
> >> > >
> >> > > Try:
> >> > > rsync -rvnc --delete /brick0/ /brick1/
> >> >
> >> > So it was seeing them as 100% different because... shit.
> >>
> >> Interesting, I do not understand this behaviour either.  Anyone care to
> >> elaborate?
> >
> >  [...chomp...chomp...chomp...]
> >
> >>   $ dirdiff foo?/
> >
> > Forget it, my brain started working after I hit send.  Writing the above
> > as the following makes the distinction clear.
> >
> >   $ dirdiff foo1/ foo2
> >
> > Pff!
> 
> Yes, it's because of this "A trailing slash on the source changes this
> behavior to avoid creating an additional directory level at the
> destination." So without the trailing / it was creating brick0 in
> brick1 (for me).

I always fall prey to this subtle distinction, specially since the shell
always completes directories with the trailing slash.  Makes me wonder,
would it be worthwhile to file an RFE against bash-completion etc, to
consider foo1 and foo1/ as separate completion candidates when
completing rsync arguments?

Undoubtedly this will lead to ugly code though.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: rsync checksum to compare directories

[Suvayu Ali]

On Mon, Aug 31, 2015 at 11:52:52AM +0200, Suvayu Ali wrote:
> 
> I always fall prey to this subtle distinction, specially since the shell
> always completes directories with the trailing slash.  Makes me wonder,
> would it be worthwhile to file an RFE against bash-completion etc, to
> consider foo1 and foo1/ as separate completion candidates when
> completing rsync arguments?

Done: https://bugzilla.redhat.com/show_bug.cgi?id=1258410

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: rsync checksum to compare directories

[Suvayu Ali]

On Mon, Aug 31, 2015 at 02:25:42AM +0200, Suvayu Ali wrote:
 On Sun, Aug 30, 2015 at 01:17:00PM -0600, Chris Murphy wrote:
  On Sun, Aug 30, 2015 at 1:01 PM, Gordon Messmer
  gordon.mess...@gmail.com wrote:
   On 08/30/2015 10:30 AM, Chris Murphy wrote:
  
   # rsync -rvnc --delete /brick0 /brick1
  
   I've confirmed that separate devices are mounted to /brick0 and
   /brick1. What I'm seeing for ~ 20 minutes now is only the device
   mounted at brick0 is being accessed, and yet I'm seeing piles of files
   (verbose option) being listed.
  
  
   Try:
   rsync -rvnc --delete /brick0/ /brick1/
  
  So it was seeing them as 100% different because... shit.
 
 Interesting, I do not understand this behaviour either.  Anyone care to
 elaborate?

 [...chomp...chomp...chomp...]

   $ dirdiff foo?/

Forget it, my brain started working after I hit send.  Writing the above
as the following makes the distinction clear.

  $ dirdiff foo1/ foo2

Pff!

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: rsync checksum to compare directories

[Suvayu Ali]

On Sun, Aug 30, 2015 at 01:17:00PM -0600, Chris Murphy wrote:
 On Sun, Aug 30, 2015 at 1:01 PM, Gordon Messmer
 gordon.mess...@gmail.com wrote:
  On 08/30/2015 10:30 AM, Chris Murphy wrote:
 
  # rsync -rvnc --delete /brick0 /brick1
 
  I've confirmed that separate devices are mounted to /brick0 and
  /brick1. What I'm seeing for ~ 20 minutes now is only the device
  mounted at brick0 is being accessed, and yet I'm seeing piles of files
  (verbose option) being listed.
 
 
  Try:
  rsync -rvnc --delete /brick0/ /brick1/
 
 So it was seeing them as 100% different because... shit.

Interesting, I do not understand this behaviour either.  Anyone care to
elaborate?

  $ mkdir -p foo{1,2}/bar
  $ touch foo{1,2}/{bla,bar/baz}
  $ echo  foo1/bla
  $ dirdiff foo? # dirdiff is a wrapper around rsync -rvnc 
--delete
  sending incremental file list
  foo1/
  foo1/bla
  foo1/bar/
  foo1/bar/baz
  
  sent 168 bytes  received 30 bytes  396.00 bytes/sec
  total size is 1  speedup is 0.01 (DRY RUN)
  $ dirdiff foo?/
  sending incremental file list
  bla
  
  sent 150 bytes  received 20 bytes  340.00 bytes/sec
  total size is 1  speedup is 0.01 (DRY RUN)

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: F22: how to mount AFS directory?

[Suvayu Ali]

Hi Ranjan,

On Mon, Aug 24, 2015 at 06:19:05PM -0500, Ranjan Maitra wrote:
 
 Does anyone know how to mount an AFS filesystem (I think it is called an 
 Andrews' Filesystem) on Fedora 22? 
 
 I looked around on the web, but did not come up with anything that worked so 
 I thought that I would ask for some advice.

You need a few things working.  Kerberos, openafs-client, kmod, and
kmod-openafs.

- Kerberos: this is for authentication.  You need to set your realms in
  /etc/krb5.conf.  Something like this (of course you have to use
  whatever is appropriate for your organisation, s/cern.ch/myorg.org/):

[realms]
 CERN.CH = {
  default_domain = cern.ch
  kdc = cerndc.cern.ch
}

[domain_realm]
 .cern.ch = CERN.CH
 cern.ch = CERN.CH

- openafs-client, kmod, and kmod-openafs: this was provided by
  RPMFusion, but the maintainer changed jobs where it was not needed
  anymore and has been orphaned since then.  This was about a year ago I
  think.  kmod is available in the normal repos, the openafs bits can
  now be found on these coprs: jsbillings/openafs,
  jsbillings/openafs-kmod.

- If you don't want to depend on the timely build of kmods for every
  kernel update, you have to switch to dkms instead of kmod.

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Strange behaviour of ln

[Suvayu Ali]

On Tue, Aug 25, 2015 at 08:34:23AM -0700, Paolo Galtieri wrote:
 
 The correct way to do the link is
 
 mkdir Test2
 cd Test2
 ln -s ../Test/file file

  $ ln -s afile2 afile

Shouldn't this suffice?  After all they are in the same directory.  With
your solution, the link will break if the directory is renamed.

 cd ..
 
 /bin/ls -l Test2
 total 0
 lrwxrwxrwx. 1 pgaltieri pgaltieri 10 Aug 25 08:29 afile - Test/afile
 lrwxrwxrwx. 1 pgaltieri pgaltieri 13 Aug 25 08:29 afile2 - ../Test/afile

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Fwd: Fedora22 Security Issue.

[Suvayu Ali]

On Wed, Aug 19, 2015 at 12:31:19PM +0100, Patrick O'Callaghan wrote:
 On Wed, 2015-08-19 at 09:55 +0900, Scott Mattan wrote:
  Is there a better way of viewing this list without having to copy 
  paste titles and contents?
 
 Don't use digests (they are a waste of time in this day and age), or if
 you do then use a mailer that supports direct replying to a digest
 message (not to the digest itself). Evolution can do this and I think
 Thunderbird also. Cutting and pasting subject lines does not preserve
 proper threading and should be avoided.

Actually, afaiU, there is one more step involved.  Replying in thread
works only with MIME digests, not plain text.  It is a separate option
in the mailman settings page.

  Set Digest Mode

  If you turn digest mode on, you'll get posts bundled together (usually
  one per day but possibly more on busy lists), instead of singly when
  they're sent. If digest mode is changed from on to off, you may
  receive one last digest.

  Get MIME or Plain Text Digests?

  Your mail reader may or may not support MIME digests. In general MIME
  digests are preferred, but if you have a problem reading them, select
  plain text digests. )

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Thu, Aug 13, 2015 at 07:43:49PM +, Andreas M. Kirchwitz wrote:
 Rahul Sundaram methe...@gmail.com wrote:
 
  However, if somebody runs dnf upgrade on the command shell then
  he clearly wants the latest updates. Right now! No caching or other
  magic involved. That's the whole point of running dnf upgrade
  manually, otherwise the user would have left the whole updating
  business to some automated background task.
 
  If this is what you want,  use dnf update --refresh instead
 
 That does clearly *not* provide the latest updates. It's better than
 without --refresh, but dnf clean metadata is required for full
 updates available.

FWIW, this my bug report fell on deaf ears:

  https://bugzilla.redhat.com/show_bug.cgi?id=1246253

I find it disheartening how it was closed without even acknowledging
that the problem exists, so I decided not to pursue this further.  If
someone is willing to put in the effort to push for this, please reopen
the bug.  Or maybe a new one, specifically on --refresh not doing what it
is supposed to.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Compiling llvm examples on Fedora

[Suvayu Ali]

Hi,

I wanted to play around with the examples provided by llvm-doc.  But I
can't figure out exactly how to compile them.  Using the provided
Makefile seems impossible, it includes several other Makefiles from the
llvm source tree, namely Makefile.{common,config,rules}.  Even if I
manage to copy over two of the files (common and rules) from the
upstream repo, finding the config file seems very difficult.  It is
generated by configure before compilation.  So I decided to use cmake.

This is what I did:

  $ cp -R /usr/share/doc/llvm-doc/examples/Fibonacci /tmp/
  $ mkdir /tmp/Fibonacci/build
  $ pushd /tmp/Fibonacci/build/

edit ../CMakeLists.txt and add the two following lines at the top:

  set(CMAKE_MODULE_PATH /usr/share/llvm/cmake)
  # find_package(LLVM)
  include(AddLLVM)

Running

  $ cmake ..

gives me the following errors:

  CMake Error at /usr/share/llvm/cmake/AddLLVM.cmake:443 (set_output_directory):
set_output_directory Function invoked with incorrect arguments for function
named: set_output_directory
  Call Stack (most recent call first):
/usr/share/llvm/cmake/AddLLVM.cmake:481 (add_llvm_executable)
CMakeLists.txt:16 (add_llvm_example)

Uncommenting the find_package lines seems to produce even more similar
copious errors.

Anyone has experience with this, ideas?

TIA,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Sat, Aug 08, 2015 at 02:23:34AM +, Andreas M. Kirchwitz wrote:
 Suvayu Ali fatkasuvayu+li...@gmail.com wrote:
 
  I hope this will be done *fast*, because I have to clean all
  *everytime* checking for updates. Otherwise, no updates are shown, even
  though they exist. This is a major bug.
 
  I'm sorry but clean all is not necessary at all!  clean metadata or
  clean expire-cache should be sufficient.  
 
 In practice, there's not much of a difference between clean all
 or just clean metadata. Because both require the update/upgrade
 command to download all stuff from the network and build to whole
 meta database from scratch, even if that wouldn't be necessary.

Sorry, that's not correct.  Ever since the yum days, they have been
different.  The key differences are these two bits (from `man dnf'):

  [...]
  dnf clean packages
Removes any cached packages from the system.
  
  dnf clean plugins
Tells  all enabled  plugins to  eliminate their  cached
data.
  
  dnf clean all
Does all of the above.


By removing packages, you are losing the ability to do offline
transactions.  Cleaning cached data from plugins, can do any number of
things depending on the enabled plugins.  So please stop repeating this
obviously incorrect information.

  That said, I sometimes do not understand what's the harm in getting
  updates few hours later.
 
 Caching might be cool for automated tasks (eg, cron jobs or background
 processes) and also for some actions that do not require up-to-date
 metadata.

During the yum days, the most frequent complaint on this list were: why
is yum so slow, I hate waiting for metadata updates,  and so on.
And now people complain, the cached metadata is out of date (please
don't confuse this with actual bugs).  There is no winning, you either
wait for the metadata to download, or you get a responsive interface.

PS: I've no relation to the dnf team, just a user.  I however think
instead of complaining on the list, if any discussions here lead you
to think there is a problem with dnf (as in a bug), you should file
bug reports.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: cpu at 99% by firefox

[Suvayu Ali]

On Tue, Aug 04, 2015 at 12:40:04PM +0200, Roberto Ragusa wrote:
 On 08/03/2015 10:21 PM, Suvayu Ali wrote:
  On 3 August 2015 at 17:51, Ed Greshko ed.gres...@greshko.com wrote:
  I tried very hard today to get FF into the condition you're reporting and 
  could not.  I went on eBay, YouTube, CNN, mlb.com, NYtimes and a whole 
  bunch of other places.  I had 8 tabs open including Facebook and G+.  No 
  issues.  But, I also don't have any addons or extensions to firefox since 
  that isn't my browser of choice.
  
  My FF session usually has 100+ tabs open.  Sometimes I do get issues,
  but with sites with lots of scripts doing smart things.  FF usually
  prompts in such cases to stop the script.
 
 I sometimes go beyond 300 tabs, approaching 500.
 
 This abandoned extension is incredibly useful:
 
 https://www.dropbox.com/s/61ovoc4hzgtm0qf/Suspend-background-tabs-master.xpi?dl=0
 
 Javascript is disabled for all background tabs, so they won't use CPU time 
 anymore.

This looks interesting.  I'll try it out.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: cpu at 99% by firefox

[Suvayu Ali]

On Tue, Aug 04, 2015 at 10:32:47AM +0800, Ed Greshko wrote:
 
 For the fun of it I fired up FF and opened up 50 tabs.  I purposely picked 
 sites which have lots of adverts which get updated.  10 of those tabs were of 
 a weather site which updates the data very frequently. 

Btw, FF nightly now has a tracking protection feature that does the no
tracking job of many of the privacy add-ons like Privacy Badger,
Ghostery, Disconnect Me, etc.  They even have a menu option, both are
still hidden inside about:config though.

 The only issue I saw is that playing a video from youtube is choppy even 
 though you can see from the progress bar that the video itself has been 
 totally downloaded.  I suspect this is due to the single process threaded 
 nature of the FF implementation.

Again, FF now has something called electrolysis (e10s), which makes it
multi-threaded.  Not sure which versions of FF has the feature.  I find
it is quite effective, but maybe a bit buggy.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: cpu at 99% by firefox

[Suvayu Ali]

On Tue, Aug 04, 2015 at 03:54:37PM +0800, Ed Greshko wrote:
 On 08/04/15 15:38, Suvayu Ali wrote:
  On Tue, Aug 04, 2015 at 10:32:47AM +0800, Ed Greshko wrote:
  For the fun of it I fired up FF and opened up 50 tabs.  I purposely picked 
  sites which have lots of adverts which get updated.  10 of those tabs were 
  of a weather site which updates the data very frequently. 
  Btw, FF nightly now has a tracking protection feature that does the no
  tracking job of many of the privacy add-ons like Privacy Badger,
  Ghostery, Disconnect Me, etc.  They even have a menu option, both are
  still hidden inside about:config though.
 
 Not sure what relevance that has to CPU usage.

As I understood the topic was unusual CPU loads.  Since you mentioned
adverts (considering they are often responsible for strange CPU loads),
I thought I mention a feature FF will have in the future.  Tests by devs
suggest the feature improves on CPU load and page load times
significantly.

 
  The only issue I saw is that playing a video from youtube is choppy even 
  though you can see from the progress bar that the video itself has been 
  totally downloaded.  I suspect this is due to the single process threaded 
  nature of the FF implementation.
  Again, FF now has something called electrolysis (e10s), which makes it
  multi-threaded.  Not sure which versions of FF has the feature.  I find
  it is quite effective, but maybe a bit buggy.
 
 
 Well, since Fedora doesn't distribute nightly and that e10s option doesn't 
 exist on the fedora released version I don't know what relevance it has to 
 the discussion.

Again, I thought this was relevant for a discussion of CPU loads,
specially after you pointed out the single-threaded nature of FF.  I
just didn't know if the Fedora supplied version already has it.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: cpu at 99% by firefox

[Suvayu Ali]

On Tue, Aug 04, 2015 at 05:10:45AM +0800, Ed Greshko wrote:
 On 08/04/15 04:21, Suvayu Ali wrote:
  On 3 August 2015 at 17:51, Ed Greshko ed.gres...@greshko.com wrote:
  I tried very hard today to get FF into the condition you're reporting and 
  could not.  I went on eBay, YouTube, CNN, mlb.com, NYtimes and a whole 
  bunch of other places.  I had 8 tabs open including Facebook and G+.  No 
  issues.  But, I also don't have any addons or extensions to firefox since 
  that isn't my browser of choice.
  My FF session usually has 100+ tabs open.  Sometimes I do get issues,
  but with sites with lots of scripts doing smart things.  FF usually
  prompts in such cases to stop the script.
 
 At 100+ tabs I'm not surprised there would be issues. 

Actually I'm surprised that with so many tabs there are so few issues!

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: cpu at 99% by firefox

[Suvayu Ali]

On 3 August 2015 at 17:51, Ed Greshko ed.gres...@greshko.com wrote:
 I tried very hard today to get FF into the condition you're reporting and 
 could not.  I went on eBay, YouTube, CNN, mlb.com, NYtimes and a whole bunch 
 of other places.  I had 8 tabs open including Facebook and G+.  No issues.  
 But, I also don't have any addons or extensions to firefox since that isn't 
 my browser of choice.

My FF session usually has 100+ tabs open.  Sometimes I do get issues,
but with sites with lots of scripts doing smart things.  FF usually
prompts in such cases to stop the script.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: installing Fedora 21 or 22 on a MacBook Air 7,1

[Suvayu Ali]

On Wed, Jul 29, 2015 at 10:07:26PM +0700, Frederic Muller wrote:
 On 07/29/2015 08:07 PM, CS DBA wrote:
  I installed Fedora 20 on a macbook air like this:
  
  1) burned the KDE live dvd,
  
  2) plugged the dvd player into the mac usb port
  
  3) held down the option button while I boot the macbook
  
  4) when it comes up with the boot options choose the Fedora DVD (it
  takes a few min before it shows up)
  
  5) boot off the DVD, then do the install to disk as normal
  
  
 
 Sorry if my email wasn't clear enough: there is no disk to install
 Fedora to. The SSD drive is simply invisible.

From a root terminal[1] in the live USB, can you see the disk with
`parted -l' or `fdisk -l'?  If yes, you could format it from the command
line, and then try starting the installer.  Needless to say backup
first.  If you can't see it even with parted, then I'm afraid I don't
have any thoughts how to proceed.


Footnotes:

[1] I think sudo should work, otherwise try su -, I think in a live
system it is passwordless.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: installing Fedora 21 or 22 on a MacBook Air 7,1

[Suvayu Ali]

On Wed, Jul 29, 2015 at 04:41:04PM +0700, Frederic Muller wrote:
 Hi!
 
 I have been struggling installing Fedora on a friend's MBA as the local
 SSD doesn't seem to be recognized. I found a few posts online with
 various degrees of luck, unfortunately they never mentioned with MBA is
 used and that didn't work for me.

I'm not a Mac user, but knowing what you tried might give others some
ideas what you could try.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Sending a desktop to a big screen

[Suvayu Ali]

On Sat, Jul 25, 2015 at 11:04:36PM -0400, Amit Prahesh wrote:
 
 Now, Sally, one of the engineers wants to share her desktop with the others
 and uses application XYZ to project/cast/extend/send it to the big TV (via
 its PC, of course). All of them discuss whatever was on her mind, and five
 minutes later, she closes application XYZ and the office goes back as it
 was before. Fred wants to do the same, and he does. Actually all of them
 can.

It seems you want to screencast, something like chromecast I guess, but
interactive sessions.  I do not know of any such program.  I can think
of two hacks though.

1) You can use your company's video conferencing setup to screen share
   with the desktop with the big TV.

2) Just unplug the TV from the desktop and plug your laptop in, and
   extend your display.  I have a script that will automatically pick the
   optimal resolution for the display using xrandr.  You can write
   something like that yourself (or if you like the idea, I can send my
   script).

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Thu, Jul 23, 2015 at 02:32:19AM -0400, Radek Holy wrote:
  
  Essentially I'm suggesting to treat no connectivity as a powercycle.
  Hopefully this gives the devs some ideas.
 
 Can you please file an RFE?

Done: https://bugzilla.redhat.com/show_bug.cgi?id=1246253

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

fail2ban on IPv6 and pam_shield

[Suvayu Ali]

Hi,

(quoting from the thread: will denyhosts work with journald (without rsyslogd)?)
http://mid.gmane.org/cahc5q3ee3vkfqacrqf3rcwny4bf5q+u93di8j+k5ggto_vy...@mail.gmail.com

On Sat, Mar 28, 2015 at 11:22:20PM -0400, Rahul Sundaram wrote:
 However denyhosts itself is pretty outdated and everyone should switch
 over to fail2ban anyway.

While reading about the recently found SSH vulnerability, I came across
a rather uncontested opinion that fail2ban does not work with IPv6,
e.g. see this reddit thread:

https://www.reddit.com/r/netsec/comments/3dnzcq/openssh_keyboardinteractive_authentication_brute/

Personally, I found the parse log files to ban idea a bit shaky to
begin with.  The same thread mentions pam_shield (also available in the
Fedora repos).  A bit of reading on it makes it seem more robust than
fail2ban.  I was wondering if anyone had any experience or opinion
on this topic.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Wed, Jul 22, 2015 at 05:41:48PM +0200, Heinz Diehl wrote:
 On 22.07.2015, Suvayu Ali wrote: 
 
  That said, I sometimes do not understand what's the harm in getting
  updates few hours later.  dnf already tells you how old the metadata is
  when it starts, you can choose to get the latest metadata if it is too
  old.  So what's the big deal?
 
 I usually update weekly (or at least once within two weeks). And since
 F22, I get nothing to do every time I do this - although there are
 updates waiting. Which is.. annoying. So every time I have to clean
 the cache to be able to update.

This is strange.  I see very different behaviour.  I update once every 2
days or so.  Usually my metadata is a few hours old.  Here are some
examples:

On my laptop:
  # dnf check-update 
  Last metadata expiration check performed 2:01:15 ago on Wed Jul 22 15:47:25 
2015.
and I have a bunch of updates waiting.

On my home server:
  # dnf check-update 
  Last metadata expiration check performed 1:19:48 ago on Wed Jul 22 16:30:00 
2015.
with approximately similar set of updates waiting.

On an old laptop which I have not updated in some time.
  $ sudo dnf check-update
  [sudo] password for jallad:
  Last metadata expiration check performed 2:29:02 ago on Wed Jul 22 15:22:05 
2015.
a much larger set of updates awaiting.

My local time is, 2015-07-22 17:54:30.  So they are all within 2-3
hours.  That is why I find your case a bit puzzling.  Maybe it is
worthwhile to file a bugzilla.  To be complete, I have no metadata
related options set in dnf.conf or yum.conf.  Maybe you have something
like that set?

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Wed, Jul 22, 2015 at 07:20:11PM +0100, Ron Yorston wrote:
 Suvayu Ali wrote:
 That said, I sometimes do not understand what's the harm in getting
 updates few hours later.  dnf already tells you how old the metadata is
 when it starts, you can choose to get the latest metadata if it is too
 old.  So what's the big deal?
 
 I certainly get the impression that dnf tells me about updates less
 frequently than yum did.  It also seems to pull in metadata less
 frequently.

Everyone seems to picked on this post for me, whereas missing on my
follow-up, with actual numbers:

  http://mid.gmane.org/20150722160112.gc1...@chitra.no-ip.org

 In fedora-updates.repo I have:  metadata_expire=6h.  I also have the
 dnf-makecache.timer 'masked'.

In the above post, I say I do not change any of the defaults metadata
related configs.  From what people are posting, I have the feeling dnf
relies a lot on _continuous_ network connectivity (which is true in my
case).  If that is true, if either the connection at the users end is
intermittent, or the mirrors are unreliable, the cache probably ends up
being stale more often.

Instead of bashing and complaining, I think trying to analyse why it
works for me (and maybe a few others who are quiet), and not for the
other participants in this thread, it would be a lot more helpful to the
devs.  I can't help here since it actually works for me beautifully.
Users who see a problem are the ones in a position to contribute an
effective bug report.

My 2¢, cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

Hi Pete,

On Wed, Jul 22, 2015 at 05:42:15PM -0500, Pete Travis wrote:
 
 There is a timer unit, `/usr/lib/systemd/system/dnf-makecache.timer`, that
 fires ten minutes after each boot then one hour following the execution of
 each previous run.  It triggers
 `/usr/lib/systemd/system/dnf-makecache.service`, a service that executes
 `dnf -v makecache timer`.  When that command runs, dnf will check the age
 of the current metadata cache and refresh it if it is older than the value
 of * metadata_timer_sync* (seconds) in `/etc/dnf/dnf.conf`.

Thank you for this clear and very nice explanation.

 So, an always-on computer should never have metadata older than 4 hours; in
 practical terms, I think values 2 hours are increasingly unlikely.  A
 computer that's been off overnight and turned on in the morning should have
 a fresh cache within 15 minutes of boot.  If you have, say, a laptop that
 you power down often and often install or update packages immediately after
 boot, you might adjust the OnBootSec value by copying dnf-makecache.timer
 to /etc/systemd/system/ and editing accordingly.  Or, consider appending
 --refresh on an as-needed basis.

I think this is where things go wrong.  OnBootSec handles powerdowns,
what about intermittent connections?  In principle, it is quite possible
everytime the timer triggers the makecache service, the connection is
absent.  In fact, the probability to hit the sweet spot is directly
determined by the reliability of the connection.  So a connection that
is up 50% of the time, will miss 50% of the makecache jobs.  Maybe the
makecache jobs can reset the timer to try again in 10 mins in case of no
network connectivity.  I think that would make the odds more favourable.

Essentially I'm suggesting to treat no connectivity as a powercycle.
Hopefully this gives the devs some ideas.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: Fedora 21 QEMM wont start after update

[Suvayu Ali]

On Tue, Jul 21, 2015 at 06:38:57PM +0200, Robert Moskowitz wrote:
 
 Do I need to reboot first or is this something more than a reboot?

Maybe just restart libvirtd?

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Tue, Jul 21, 2015 at 09:36:26PM +0200, Heinz Diehl wrote:
 On 21.07.2015, Radek Holy wrote: 
 
  IIUUC, this is not completely true. I believe that once both PackageKit and 
  DNF are integrated with the new CAShe [1], we will *be able* to improve 
  this situation [2].
 
 I hope this will be done *fast*, because I have to clean all
 *everytime* checking for updates. Otherwise, no updates are shown, even
 though they exist. This is a major bug.

I'm sorry but clean all is not necessary at all!  clean metadata or
clean expire-cache should be sufficient.  

That said, I sometimes do not understand what's the harm in getting
updates few hours later.  dnf already tells you how old the metadata is
when it starts, you can choose to get the latest metadata if it is too
old.  So what's the big deal?

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: external screen no longer support native resolution

[Suvayu Ali]

On Tue, Jul 21, 2015 at 02:32:08PM +0700, Frederic Muller wrote:
 Hi!
 
 I just upgraded to F22 and my external monitor (a Koios 24 1920x1200)
 no longer works in its native resolution. I can only get 1280 x 768 to
 display.
 
 Everything was working well before. My video chip is the Intel HD
 Graphics 5500 (Broadwell GT2) (the laptop is a X1 Carbon 3rd generation).
 
 GNOME display settings does offer the native resolution but it just
 doesn't work (I get an screen switched off).
 
 How would you troubleshooting this?

The first step would be to see what xrandr gives you (try -q).  If that
works, try to get it working with xrandr first.

Hope this helps,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Mon, Jul 20, 2015 at 09:00:16AM +0200, Jan Zelený wrote:
  On Sun, Jul 19, 2015 at 8:32 PM, Javier Perez pepeb...@gmail.com wrote:
   This is weird.
   Software Updates on the Control Panel says that there are 39 updates
   available
   But when I run dnf update it says Nothing to do. What gives?
 
 IIRC the Software Updates widget does not use dnf to check for updates, 
 therefore it's likely it has a different set of metadata at its disposal. As 
 you figured out, cleaning the MD cache helps.

I'm getting a bit confused lately.  How many package managers does
Fedora have these days?  IIRC, until a year or two back, it was the same
backend (yum), but many front ends (yumex, all the packagekit based
frontends for the different desktops).  Did packagekit start doing the
backend bits itself?

From your message I understand that there are at least two different
package managers, both are Official to some capacity.  For cli users
like myself, it's dnf, for gui users it's something packagekit based.

Am I mistaken?

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Mon, Jul 20, 2015 at 10:36:01AM +0200, Maurizio Marini wrote:
 On Mon, 20 Jul 2015 10:01:37 +0200
 Michael Schwendt mschwe...@gmail.com wrote:
 
  On Sun, 19 Jul 2015 20:39:36 -0500, Javier Perez wrote:
  
   Ok, just did a dnf clean all , and the dnf update and the updates showed 
   up
   
   Weird.
  
  Just some hours before your post I had sent this:
  https://lists.fedoraproject.org/pipermail/users/2015-July/463183.html
 
 Michael is right: all we are a mass of lame
 I have shamed very much reading his answer to me; I have posted w/out RTFM and
 who do not RTFM should be banned like in old good irc times

Unfortunately Michael, this `yum/dnf clean all' business will never go
away.  I remember there was a time on this list, when everytime someone
suggested that, they got corrected.  But all the people who used to do
the correction are tired of repeating themselves, myself included.  It's
a lost cause.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: dnf update vs Software Udpates

[Suvayu Ali]

On Mon, Jul 20, 2015 at 10:44:52AM +0200, Jan Zelený wrote:
 On 20. 7. 2015 at 09:43:45, Suvayu Ali wrote:
  On Mon, Jul 20, 2015 at 09:00:16AM +0200, Jan Zelený wrote:
On Sun, Jul 19, 2015 at 8:32 PM, Javier Perez pepeb...@gmail.com 
 wrote:
 This is weird.
 Software Updates on the Control Panel says that there are 39 updates
 available
 But when I run dnf update it says Nothing to do. What gives?
   
   IIRC the Software Updates widget does not use dnf to check for updates,
   therefore it's likely it has a different set of metadata at its disposal.
   As you figured out, cleaning the MD cache helps.
  
  I'm getting a bit confused lately.  How many package managers does
  Fedora have these days?  IIRC, until a year or two back, it was the same
  backend (yum), but many front ends (yumex, all the packagekit based
  frontends for the different desktops).  Did packagekit start doing the
  backend bits itself?
  
  From your message I understand that there are at least two different
  package managers, both are Official to some capacity.  For cli users
  like myself, it's dnf, for gui users it's something packagekit based.
  
  Am I mistaken?
 
 You are not, that's pretty much it. We have had two independent software 
 management stacks since F21 where PackageKit (PK) switched from yum backend 
 to 
 libhif.
 
 At the moment PK and dnf share libraries for depsolving and downloading stuff 
 but other than that the code is independent. IIRC the reason is that dnf is 
 written in Python and that is not acceptable for PackageKit because of the 
 new 
 Gnome Software front end. It is likely that PK and dnf will share more code 
 in 
 the future but that's more of a very long term plan.

It seems a bit strange when frontends can dictate backend requirements.
Also seems like a lot of duplicated effort, increased chances of bugs,
and what not.

Anyway, thanks for your confirmation.

Cheers,

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Systemd user services

[Suvayu Ali]

Hi,

I wanted to run systemd user services, in linger mode, meaning they run
as a regular user and hang around without an active session.  This
worked nicely for the default target.  When I changed to a different
target however, the systemd user is not working anymore.  Any thoughts?

This is what I did step by step:
- Enable linger mode and reboot:
  # loginctl enable-linger user
- Now from a console login when I do:
  $ systemctl --user status
  I see systemd user service is running.
- I add a service like this:
  $ systemctl --user enable emacs
  $ systemctl --user start emacs
  Looking at the status now, I see emacs is running as expected.

This was just a test, I actually want to run mpd, but I was not getting
any sound without a graphical session, so temporarily I switched to a
graphical target with:
# systemctl isolate graphical.target # my default is runlevel3

IIUC, when I do this all services get restarted.  So I wasn't surprised
when there was not running emacs (checked with pgrep).  I was surprised
however to see I cannot query the user session anymore.

$ systemctl --user enable emacs.service
Failed to get D-Bus connection: No such file or directory

It is even more surprising since the systemd user service is still running!

$ ps -uf -p 14301
USER   PID %CPU %MEMVSZ   RSS TTY  STAT START   TIME COMMAND
lightdm  14301  0.0  0.1  45004  4864 ?Ss   10:18   0:00 
/usr/lib/systemd/systemd --user

What is wrong?  Am I mistaken somewhere, or is this a bug?

Thanks for any thoughts.

-- 
Suvayu

Open source is the future. It sets us free.
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org