Arch Linux

I am confirming this bug. Without `libiscsi` package it's completly impossible to run libvirt daemon because it crashes. I see the same error message. Severity should be high.

Yeah, package maintainer oversight. Requires now libiscsi (thanks Tomasz). Arch package system web ui is commendable here, because it is easy to find such bugs through a click on "Bug Report".

This was actually caught within my update script with "findbrokenpkgs", which found libiscsi.so.8 to be missing. Please consider using it, e.g. using a file /etc/findbrokenpkgs/arch.conf like so:

SEARCH_DIRS="/usr/bin /usr/lib /usr/local/bin /usr/local/sbin /usr/local/lib /opt"
SEARCH_DIRS_MASK="/usr/lib/ImageMagick-7.0.8/modules-Q16HDRI"
LD_LIBRARY_MASK="libmicrohttpd.so.12 libKF5I18n.so.5 libmpi_cxx.so.40 libgnome-keyring.so.0 libgcr-base-3.so.1 libruby.so.2.5 libKF5Wallet.so.5 libKF5CoreAddons.so.5 libR.so libjavascriptcoregtk-4.0.so.18 libsybdb.so.5 librrd.so.8 libQtCore.so.4 libopusfile.so.0 libodbc.so.2 libtwolame.so.0 libglut.so.3 libfbclient.so.2 libQtGui.so.4 libtcl8.6.so liblua.so.5.3 libGLU.so.1 libao.so.4 libatm.so.1 libmpdec.so.2 libtk8.6.so libmpi.so.40"

Also please consider using Testing more aggressively or namcap -i PKGBUILD and namcap -i *.pkg.tar.xz to check for such oversights. Thanks!

I am confirming this bug. Without `libiscsi` package it's completly impossible to run libvirt daemon because it crashes. I see the same error message. Severity should be high.

+1

this is also report by namcap, why don't maintainers run that before publishing?

libvirt W: Referenced library 'libiscsi.so.8' is an uninstalled dependency

I got this bug after system upgrade. Workaround was to install libiscsi as suggested, thanks.

I can confirm the bug. Libvirtd did crash with the same message, requiring libiscsi.

Confirming the bug as well.

Since the status is still "undecided" after 3 days and nothing has changed, Ill post something as well.

These are my personal beliefs:

Yes, this is a bug.
Yes, severity should be high.
Yes, maintainers should test their own packages.

We don't need a bunch of me-too type comments. The issue is pretty obvious.

should be fixed with libvirt 4.8.0-2, can you confirm?

Scimmia, nope. It's needed. Issue like this should not happen and packages maintainer should use tools like these mentioned by @spider007.

libvirt 4.8.0-2 fixes issue for me. Now it works.

I say that as a FS admin. Repeated violations will get posts deleted and accounts suspended as necessary.

Namcap is extremely unreliable and results in tons of incorrect warnings for most packages, and your collective lack of understanding of how Arch Linux packaging works does *not* qualify this as a "should have used namcap" bug (the joke is on you, our automated tooling already runs namcap and that doesn't help if we mistake it for a false positive).

Nor is this bug a "high" severity, as a missing dependency does not result in the package being completely inoperative -- it just means you need to manually install the dependency for now.

Nor do the "me too" comments complete with insults and whining, help in the slightest. On the other hand, feel free to continue, then Scimmia and I can delete your comments and ban your accounts, and then if you ever actually do have something useful to contribute, you will, regrettably, be unable to because of your prior toxic behavior.

...

Finally I would like to note that for a bug with a pretty obvious workaround, a 3-day turnaround on releasing a *weekend* fix is not something for us to feel embarrassed about.