Bug 1595715
| Summary: | Add ppa15/bpb to the default cpu model for z196 and higher in the 7.6 s390-ccw-virtio machine | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Cornelia Huck <cohuck> |
| Component: | qemu-kvm-ma | Assignee: | Cornelia Huck <cohuck> |
| Status: | CLOSED ERRATA | QA Contact: | Qunfang Zhang <qzhang> |
| Severity: | unspecified | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 7.6 | CC: | cohuck, dhildenb, juzhang, michen, mrezanin, qzhang, thuth, xuma |
| Target Milestone: | rc | ||
| Target Release: | --- | ||
| Hardware: | s390x | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | qemu-kvm-ma-2.12.0-7.el7 | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2018-10-30 08:03:37 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: | |||
|
Description
Cornelia Huck
2018-06-27 12:19:35 UTC
Fix included in qemu-kvm-ma-2.12.0-7.el7 Reproduced on qemu-kvm-ma-2.12.0-6.el7.s390x: # /usr/libexec/qemu-kvm -cpu 'z196' -name 'avocado-vt-vm1' -sandbox off -machine s390-ccw-virtio -device virtio-scsi-ccw,id=virtio_scsi_ccw0 -drive id=drive_image1,if=none,snapshot=off,aio=threads,cache=none,format=qcow2,file=/home/qzhang/kar/vt_test_images/ALT-Server-7.6-s390x-virtio-scsi.qcow2 -device scsi-hd,id=image1,drive=drive_image1,bootindex=1 -device virtio-net-ccw,mac=9a:1f:20:21:22:23,id=iduiufBS,netdev=idtlh3FZ -netdev tap,id=idtlh3FZ,vhost=on,script=/etc/qemu-ifup -m 2048 -smp 2,maxcpus=2,cores=1,threads=1,sockets=2 -nographic -rtc base=utc,clock=host,driftfix=slew -nodefaults -enable-kvm -monitor stdio -device sclpconsole,chardev=serial_id_serial0 -chardev socket,id=serial_id_serial0,path=/var/tmp/serial,server,nowait (1) -cpu 'z196' [root@localhost ~]# grep "^facilities" /proc/cpuinfo grep "^facilities" /proc/cpuinfo facilities : 0 1 2 3 4 6 7 8 9 10 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 47 52 74 75 76 77 138 (2) -cpu 'z114' # grep "^facilities" /proc/cpuinfo grep "^facilities" /proc/cpuinfo facilities : 0 1 2 3 4 6 7 8 9 10 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 47 52 74 75 76 77 138 (3) -cpu 'zEC12' # grep "^facilities" /proc/cpuinfo grep "^facilities" /proc/cpuinfo facilities : 0 1 2 3 4 6 7 8 9 10 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 47 48 49 50 51 52 64 69 71 73 74 75 76 77 78 131 138 (4) -cpu 'zBC12' # grep "^facilities" /proc/cpuinfo grep "^facilities" /proc/cpuinfo facilities : 0 1 2 3 4 6 7 8 9 10 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 47 48 49 50 51 52 64 69 71 73 74 75 76 77 78 131 138 Result: There are no "81" and "82" facilities bits in guest cpu info. Re-test on qemu-kvm-ma-2.12.0-9.el7.s390x:
(1) boot guest with -cpu 'z196' ==> Guest failed to boot up and stuck at:
LOADPARM=[ ]
Using virtio-scsi.
Using SCSI scheme.
.....
Uncompressing Linux... Ok, booting the kernel.
"top" on host:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
13387 root 20 0 2576976 128000 15660 S 99.7 3.3 2:19.79 qemu-kvm
13476 root 20 0 110508 4256 3572 R 0.3 0.1 0:00.07 top
(2)Boot guest with -cpu 'z114' ==> Guest failed to boot up and stuck at:
LOADPARM=[ ]
Using virtio-scsi.
Using SCSI scheme.
.....
Uncompressing Linux... Ok, booting the kernel.
"top" on host:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
14001 root 20 0 2576976 126892 15660 S 98.3 3.2 1:21.05 qemu-kvm
2238 root 10 -10 14044 14036 3992 S 0.3 0.4 0:00.53 iscsid
1 root 20 0 91544 9032 5924 S 0.0 0.2 0:03.52 systemd
(3) -cpu 'zEC12'
# grep "^facilities" /proc/cpuinfo | grep "81 82"
grep "^facilities" /proc/cpuinfo | grep "81 82"
facilities : 0 1 2 3 4 6 7 8 9 10 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 47 48 49 50 51 52 64 69 71 73 74 75 76 77 78 81 82 131 138
(4) -cpu 'zBC12'
# grep "^facilities" /proc/cpuinfo | grep "81 82"
grep "^facilities" /proc/cpuinfo
facilities : 0 1 2 3 4 6 7 8 9 10 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 47 48 49 50 51 52 64 69 71 73 74 75 76 77 78 81 82 131 138
Hi, Cornelia Could you give a help to review comment 6? I think for "zEC12" and "zBC12" cpu models, the bug is fixed since "81 82" facilities bits are set. However when I boot guest with "-cpu z196" or "-cpu z114", guest failed to boot up. This issue didn't exist in qemu-kvm-ma-2.12.0-6.el7.s390x. Is this a new issue? My host cpuinfo: # cat /proc/cpuinfo vendor_id : IBM/S390 # processors : 2 bogomips per cpu: 2913.00 max thread id : 0 features : esan3 zarch stfle msa ldisp eimm dfp edat etf3eh highgprs te sie facilities : 0 1 2 3 4 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 37 40 41 42 43 44 45 46 47 48 49 50 51 52 57 64 65 66 67 68 69 70 71 72 73 75 76 77 78 81 82 131 132 cache0 : level=1 type=Data scope=Private size=96K line_size=256 associativity=6 cache1 : level=1 type=Instruction scope=Private size=64K line_size=256 associativity=4 cache2 : level=2 type=Data scope=Private size=1024K line_size=256 associativity=8 cache3 : level=2 type=Instruction scope=Private size=1024K line_size=256 associativity=8 cache4 : level=3 type=Unified scope=Shared size=49152K line_size=256 associativity=12 cache5 : level=4 type=Unified scope=Shared size=393216K line_size=256 associativity=24 processor 0: version = 00, identification = 3EC047, machine = 2827 processor 1: version = 00, identification = 3EC047, machine = 2827 cpu number : 0 cpu MHz dynamic : 5504 cpu MHz static : 5504 cpu number : 1 cpu MHz dynamic : 5504 cpu MHz static : 5504 Thanks, Qunfang The issue for failing to boot up guest with "-cpu z196" or "-cpu z114", was introduced in qemu-kvm-ma-2.12.0-7.el7.s390x. Is the kernel compatible with z196/z114 at all? I thought it would be compiled for z12+. Are you using the exact same guest kernel in both setups? (In reply to David Hildenbrand from comment #9) > Is the kernel compatible with z196/z114 at all? I thought it would be > compiled for z12+. Are you using the exact same guest kernel in both setups? Yeah, for both test (comment 5 and comment 6), I used Kernel 4.14.0-92.el7a.s390x as the guest kernel, it's not the latest 7.6-alt version though. This looks like a preexisting bug with the z196/z114 cpu models, the bpb/ppa15 bits, and (at least) the -alt kernel. As stated in https://bugzilla.redhat.com/show_bug.cgi?id=1612034, this works if the features are turned off manually, and it also starts to fail if they are added manually on a QEMU without this patch. I'm suspecting this is a bug in the spectre mitigation code in the kernel. Based on comment 11, I set bug status to VERIFIED and use bug 1612034 to track the guest kernel issue. Thanks. 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://access.redhat.com/errata/RHSA-2018:3062 |