Bug 1174147
| Summary: | uninitialised byte for virDomainOpenGraphicsFD | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Jincheng Miao <jmiao> |
| Component: | libvirt | Assignee: | Pavel Hrdina <phrdina> |
| Status: | CLOSED ERRATA | QA Contact: | Virtualization Bugs <virt-bugs> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 7.1 | CC: | dyuan, eblake, honzhang, lhuang, mzhan, phrdina, rbalakri |
| Target Milestone: | rc | ||
| Target Release: | --- | ||
| Hardware: | x86_64 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | libvirt-1.2.14-1.el7 | Doc Type: | Bug Fix |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2015-11-19 05:58:27 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
Actually this is a bug in gnulib module that libvirt uses. I've proposed patch to upstream gnulib. After it will be fixed in upstream gnulib, we will use the new gnulib for libvirt to fix this issue. Gnulib accepted the patch for their side, here's the proposed libvirt patch to pick it up (we missed RC1, so I can't push it without review) https://www.redhat.com/archives/libvir-list/2015-February/msg00952.html Fixing this in RHEL 7.2 is easy - the rebase will do the job. Fixing it in RHEL 7.1 is harder - we prefer not to pick up gnulib updates in maintenance branches. It's possible with a downstream-only patch, but that violates upstream first, and since the warning is otherwise harmless (we are not actually basing behavior on the uninitialized read, since it occurs in struct padding), it is NOT worth backporting to 7.1.z. upstream commit:
commit df60346ca2e178de5bcc4dd951d63011883a56d7
Author: Eric Blake <eblake>
Date: Tue Feb 24 08:58:46 2015 -0700
maint: update to latest gnulib
Usual update to latest gnulib status. In particular this update
fixes at least one issue that can be seen in libvirt, by silencing
a valgrind detection of uninitialized memory:
https://bugzilla.redhat.com/show_bug.cgi?id=1174147
* .gnulib: Update to latest.
* bootstrap: Resync to gnulib.
Signed-off-by: Eric Blake <eblake>
git-describe: v1.2.13-322-gdf60346
I can reporduce this issue with libvirt-1.2.13-1.el7.x86_64: # python Python 2.7.5 (default, Feb 11 2014, 07:46:25) [GCC 4.8.2 20140120 (Red Hat 4.8.2-13)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import libvirt >>> con=libvirt.open() >>> dom=con.lookupByName("test3") >>> dom.create() 0 >>> dom.openGraphicsFD(0) 4 # valgrind --leak-check=full libvirtd ==17159== Thread 1: ==17159== Syscall param sendmsg(msg.msg_control) points to uninitialised byte(s) ==17159== at 0x7E827BD: ??? (in /usr/lib64/libpthread-2.17.so) ==17159== by 0x54869A1: sendfd (passfd.c:86) ==17159== by 0x5402591: virNetSocketSendFD (virnetsocket.c:1766) ==17159== by 0x53F77E0: virNetServerClientDispatchWrite (virnetserverclient.c:1271) ==17159== by 0x53F77E0: virNetServerClientDispatchEvent (virnetserverclient.c:1371) ==17159== by 0x52BB8C6: virEventPollDispatchHandles (vireventpoll.c:508) ==17159== by 0x52BB8C6: virEventPollRunOnce (vireventpoll.c:657) ==17159== by 0x52B9F91: virEventRunDefaultImpl (virevent.c:308) ==17159== by 0x1512FC: virNetServerRun (virnetserver.c:1139) ==17159== by 0x11E34C: main (libvirtd.c:1491) ==17159== Address 0xffefffb34 is on thread 1's stack ==17159== in frame #1, created by sendfd (passfd.c:51) ==17159== And verify this bug with libvirt-1.2.17-2.el7.x86_64: # python Python 2.7.5 (default, Feb 11 2014, 07:46:25) [GCC 4.8.2 20140120 (Red Hat 4.8.2-13)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import libvirt >>> con=libvirt.open() >>> dom=con.lookupByName("test3") >>> dom.create() 0 >>> dom.openGraphicsFD(0) 7 # valgrind --leak-check=full libvirtd <no report point to sendfd> But i noticed another... ==18439== Thread 17: ==18439== Syscall param socketcall.bind(my_addr.sa_data) points to uninitialised byte(s) ==18439== at 0x833FE97: bind (in /usr/lib64/libc-2.17.so) ==18439== by 0x1E30C3F5: iface_bind (pcap-linux.c:4743) ==18439== by 0x1E30C3F5: activate_new (pcap-linux.c:3225) ==18439== by 0x1E30C3F5: pcap_activate_linux (pcap-linux.c:1293) ==18439== by 0x1E31283F: pcap_activate (pcap.c:735) ==18439== by 0x1E313155: pcap_open_live (pcap.c:788) ==18439== by 0x1E0EEB12: learnIPAddressThread (nwfilter_learnipaddr.c:413) ==18439== by 0x5506D11: virThreadHelper (virthread.c:206) ==18439== by 0x8034DF4: start_thread (in /usr/lib64/libpthread-2.17.so) ==18439== by 0x833F1AC: clone (in /usr/lib64/libc-2.17.so) ==18439== Address 0x25af09f4 is on thread 17's stack ==18439== in frame #1, created by pcap_activate_linux (pcap-linux.c:1234) ==18439== Seems this still not the libvirt fault. Hi Pavel, Would you please help to check out if comment 7 is okay to verify this bug ? need we open a new bug to against libpcap ? Thanks in advance for your reply. Hi Luyao, Yes, you can verify this bug as fixed and that new issue is in libpcap, so you should create a new bug against that component. (In reply to Pavel Hrdina from comment #9) > Hi Luyao, > > Yes, you can verify this bug as fixed and that new issue is in libpcap, so > you should create a new bug against that component. Thanks a lot for your reply, move this bug to verify. Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHBA-2015-2202.html |
Description of problem: valgrind reports uninitialised byte for virDomainOpenGraphicsFD Version-Release number of selected component (if applicable): libvirt-1.2.8-9.el7.x86_64 qemu-kvm-rhev-2.1.2-13.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1. run libvirt in valgrind # valgrind --leak-check=full libvirtd 2. call virDomainOpenGraphicsFD # python Python 2.7.5 (default, Feb 11 2014, 07:46:25) [GCC 4.8.2 20140120 (Red Hat 4.8.2-13)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import libvirt >>> con=libvirt.open(None) >>> dom=con.lookupByName("r70") >>> dom.openGraphicsFD(0) 3. get error from valgrind ==19038== Syscall param sendmsg(msg.msg_control) points to uninitialised byte(s) ==19038== at 0x7C6D7BD: ??? (in /usr/lib64/libpthread-2.17.so) ==19038== by 0x547A091: sendfd (passfd.c:86) ==19038== by 0x53F708C: virNetSocketSendFD (virnetsocket.c:1765) ==19038== by 0x53EBF70: virNetServerClientDispatchWrite (virnetserverclient.c:1335) ==19038== by 0x53EBF70: virNetServerClientDispatchEvent (virnetserverclient.c:1435) ==19038== by 0x52B7B99: virEventPollDispatchHandles (vireventpoll.c:510) ==19038== by 0x52B7B99: virEventPollRunOnce (vireventpoll.c:660) ==19038== by 0x52B6281: virEventRunDefaultImpl (virevent.c:308) ==19038== by 0x15189C: virNetServerRun (virnetserver.c:1139) ==19038== by 0x11E5B7: main (libvirtd.c:1507) ==19038== Address 0xffefffc54 is on thread 1's stack ==19038== in frame #1, created by sendfd (passfd.c:51)