My commit 7a2e845a86 (and its
prerequisites) managed to effectively ignore the
clear_emulator_capabilities setting in qemu.conf (visible in the code
as the VIR_EXEC_CLEAR_CAPS flag when qemu is being exec'ed), with the
result that the capabilities are always cleared regardless of the
qemu.conf setting. This patch fixes it by passing the flag through to
virSetUIDGIDWithCaps(), which uses it to decide whether or not to
clear existing capabilities before adding in those that were
requested.
Note that the existing capabilities are *always* cleared if the new
process is going to run as non-root, since the whole point of running
non-root is to have the capabilities removed (it's still possible to
maintain individual capabilities as needed using the capBits argument
though).
Normally when a process' uid is changed to non-0, all the capabilities
bits are cleared, even those explicitly set with calls to
capng_update()/capng_apply() made immediately before setuid. And
*after* the process' uid has been changed, it no longer has the
necessary privileges to add capabilities back to the process.
In order to set a non-0 uid while still maintaining any capabilities
bits, it is necessary to either call capng_change_id() (which
unfortunately doesn't currently call initgroups to setup auxiliary
group membership), or to perform the small amount of calisthenics
contained in the new utility function virSetUIDGIDWithCaps().
Another very important difference between the capabilities
setting/clearing in virSetUIDGIDWithCaps() and virCommand's
virSetCapabilities() (which it will replace in the next patch) is that
the new function properly clears the capabilities bounding set, so it
will not be possible for a child process to set any new
capabilities.
A short description of what is done by virSetUIDGIDWithCaps():
1) clear all capabilities then set all those desired by the caller (in
capBits) plus CAP_SETGID, CAP_SETUID, and CAP_SETPCAP (which is needed
to change the capabilities bounding set).
2) call prctl(), telling it that we want to maintain current
capabilities across an upcoming setuid().
3) switch to the new uid/gid
4) again call prctl(), telling it we will no longer want capabilities
maintained if this process does another setuid().
5) clear the capabilities that we added to allow us to
setuid/setgid/change the bounding set (unless they were also requested
by the caller via the virCommand API).
Because the modification/maintaining of capabilities is intermingled
with setting the uid, this is necessarily done in a single function,
rather than having two independent functions.
Note that, due to the way that effective capabilities are computed (at
time of execve) for a process that has uid != 0, the *file*
capabilities of the binary being executed must also have the desired
capabilities bit(s) set (see "man 7 capabilities"). This can be done
with the "filecap" command. (e.g. "filecap /usr/bin/qemu-kvm sys_rawio").
Way back when I started making changes for Coverity messages my first set
were to a bunch of CHECKED_RETURN errors. In particular virAsprintf() had
a few callers that Coverity noted didn't check their return (although some
did check if the buffer being printed to was NULL or not).
It was suggested at the time as a further patch an ATTRIBUTE_RETURN_CHECK
should be added to virAsprintf(), see:
https://www.redhat.com/archives/libvir-list/2013-January/msg00120.html
This patch does that and fixes a few more instances not found by Coverity
that failed the check.
"virGetDeviceID" could be used across the sources, but it doesn't
relate with this series, and could be done later.
* src/util/virutil.h: (Declare virGetDeviceID, and
vir{Get,Set}DeviceUnprivSGIO)
* src/util/virutil.c: (Implement virGetDeviceID and
vir{Get,Set}DeviceUnprivSGIO)
* src/libvirt_private.syms: Export private symbols of upper helpers
