Bug 1817921
| Summary: | SPLIT_LOCK_DISABLE is not enabled in VM with CPU model Snowridge | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 9 | Reporter: | belinda <xuelian.guo> |
| Component: | qemu-kvm | Assignee: | Paul Lai (Intel) <plai> |
| qemu-kvm sub component: | CPU Models | QA Contact: | liunana <nanliu> |
| Status: | CLOSED WONTFIX | Docs Contact: | |
| Severity: | unspecified | ||
| Priority: | unspecified | CC: | ailan, chayang, jferlan, jinzhao, juzhang, mrezanin, nanliu, nilal, paul.c.lai, plai, virt-maint, xun.fang |
| Version: | unspecified | Keywords: | Reopened, Triaged |
| Target Milestone: | rc | Flags: | pm-rhel:
mirror+
|
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | qemu-kvm-6.2.0-1.el9 | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: |
RHEL 8.2 snapshot2
qemu-kvm fast train
SNR Physical machine
|
|
| Last Closed: | 2022-01-10 16:39:03 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: | |||
| Bug Depends On: | |||
| Bug Blocks: | 1750275 | ||
|
Description
belinda
2020-03-27 08:51:22 UTC
(In reply to belinda from comment #0) > Description of problem: > boot up guest with -cpu Snowridge, “split_lock_detect” cannot be found in > cpu flags > > Version-Release number of selected component (if applicable): > KVM > > OS: Red Hat Enterprise Linux release 8.2 Beta (Ootpa) > Kernel: 4.18.0-184.el8.x86_644 > AV qemu: 4.2.0 (qemu-kvm-4.2.0-12.module+el8.2.0+5858+afd073bc) > > How reproducible: > > > Steps to Reproduce: > 1. Boot up guest using: -cpu Snowridge > qemu-img create -b /share/xvs/img/OSV/rhel_8u2_snapshot2_legacy.img -f qcow2 > /share/xvs/var/tmp-img_KVM_CPU_SNR_1584915162 > qemu-system-x86_64 -accel kvm -drive > file=/share/xvs/var/tmp-img_KVM_CPU_SNR_1584915162,if=none,id=virtio-disk0 > -device virtio-blk-pci,drive=virtio-disk0 -m 4096 -monitor pty -smp 4 -cpu > Snowridge -device virtio-net-pci,netdev=nic0,mac=00:44:7d:d4:42:1e -netdev > tap,id=nic0,br=virbr0,helper=/usr/libexec/qemu-bridge-helper,vhost=on > 2.Login VM and check cpu flag “split_lock_detect” > lscpu |grep split_lock_detect Linux doesn't show split_lock_detect on /proc/cpuinfo when it detects it's running under an hypervisor. Please use intel msr-tools to check value of MSR_IA32_CORE_CAPABILITY (0xcf) has bit 5 set. (In reply to Eduardo Habkost from comment #1) > (In reply to belinda from comment #0) > > Description of problem: > > boot up guest with -cpu Snowridge, “split_lock_detect” cannot be found in > > cpu flags > > > > Version-Release number of selected component (if applicable): > > KVM > > > > OS: Red Hat Enterprise Linux release 8.2 Beta (Ootpa) > > Kernel: 4.18.0-184.el8.x86_644 > > AV qemu: 4.2.0 (qemu-kvm-4.2.0-12.module+el8.2.0+5858+afd073bc) > > > > How reproducible: > > > > > > Steps to Reproduce: > > 1. Boot up guest using: -cpu Snowridge > > qemu-img create -b /share/xvs/img/OSV/rhel_8u2_snapshot2_legacy.img -f qcow2 > > /share/xvs/var/tmp-img_KVM_CPU_SNR_1584915162 > > qemu-system-x86_64 -accel kvm -drive > > file=/share/xvs/var/tmp-img_KVM_CPU_SNR_1584915162,if=none,id=virtio-disk0 > > -device virtio-blk-pci,drive=virtio-disk0 -m 4096 -monitor pty -smp 4 -cpu > > Snowridge -device virtio-net-pci,netdev=nic0,mac=00:44:7d:d4:42:1e -netdev > > tap,id=nic0,br=virbr0,helper=/usr/libexec/qemu-bridge-helper,vhost=on > > 2.Login VM and check cpu flag “split_lock_detect” > > lscpu |grep split_lock_detect > > Linux doesn't show split_lock_detect on /proc/cpuinfo when it detects it's > running under an hypervisor. Please use intel msr-tools to check value of > MSR_IA32_CORE_CAPABILITY (0xcf) has bit 5 set. We checked on SNR & RHEL8.3 Beta & qemu-kvm fast train on host: # rdmsr 0xcf 60 (In reply to Eduardo Habkost from comment #1) > (In reply to belinda from comment #0) > > Description of problem: > > boot up guest with -cpu Snowridge, “split_lock_detect” cannot be found in > > cpu flags > > > > Version-Release number of selected component (if applicable): > > KVM > > > > OS: Red Hat Enterprise Linux release 8.2 Beta (Ootpa) > > Kernel: 4.18.0-184.el8.x86_644 > > AV qemu: 4.2.0 (qemu-kvm-4.2.0-12.module+el8.2.0+5858+afd073bc) > > > > How reproducible: > > > > > > Steps to Reproduce: > > 1. Boot up guest using: -cpu Snowridge > > qemu-img create -b /share/xvs/img/OSV/rhel_8u2_snapshot2_legacy.img -f qcow2 > > /share/xvs/var/tmp-img_KVM_CPU_SNR_1584915162 > > qemu-system-x86_64 -accel kvm -drive > > file=/share/xvs/var/tmp-img_KVM_CPU_SNR_1584915162,if=none,id=virtio-disk0 > > -device virtio-blk-pci,drive=virtio-disk0 -m 4096 -monitor pty -smp 4 -cpu > > Snowridge -device virtio-net-pci,netdev=nic0,mac=00:44:7d:d4:42:1e -netdev > > tap,id=nic0,br=virbr0,helper=/usr/libexec/qemu-bridge-helper,vhost=on > > 2.Login VM and check cpu flag “split_lock_detect” > > lscpu |grep split_lock_detect > > Linux doesn't show split_lock_detect on /proc/cpuinfo when it detects it's > running under an hypervisor. Please use intel msr-tools to check value of > MSR_IA32_CORE_CAPABILITY (0xcf) has bit 5 set. We checked on SNR & RHEL8.3 Beta & qemu-kvm fast train on host: # rdmsr 0xcf 60 We double confirmed with developer , SPLIT_LOCK_DISABLE patch is still not upstream, so we will keep monitor this issue, and will test as soon as it is integrated into RHEL. Bulk update: Move RHEL8 bugs to RHEL9. If necessary to resolve in RHEL8, then clone to the current RHEL8 release. After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened. Hi Eduardo, Do we have a plan to fix this bug now? If so, please help to re-open this bug. Thanks! Best regards Liu Nana Hi Eduardo, About this comment: https://bugzilla.redhat.com/show_bug.cgi?id=1750275#c26 Libvirt will remove '-split-lock-detect' in configuration while booting qemu with "<cpu mode='host-model'/>" <cpu mode='custom' match='exact' check='full'> <model fallback='forbid'>Snowridge</model> <vendor>Intel</vendor> <feature policy='require' name='ss'/> <feature policy='require' name='vmx'/> <feature policy='require' name='pdcm'/> <feature policy='require' name='hypervisor'/> <feature policy='require' name='tsc_adjust'/> <feature policy='require' name='waitpkg'/> <feature policy='require' name='rdpid'/> <feature policy='require' name='md-clear'/> <feature policy='require' name='stibp'/> <feature policy='require' name='xsaves'/> <feature policy='require' name='ibpb'/> <feature policy='require' name='ibrs'/> <feature policy='require' name='amd-stibp'/> <feature policy='require' name='amd-ssbd'/> <feature policy='require' name='rdctl-no'/> <feature policy='require' name='ibrs-all'/> <feature policy='require' name='skip-l1dfl-vmentry'/> <feature policy='require' name='mds-no'/> <feature policy='require' name='pschange-mc-no'/> <feature policy='disable' name='mpx'/> <feature policy='disable' name='core-capability'/> <feature policy='disable' name='split-lock-detect'/> Is this a workaround for this bug or a final resolution? If this is the final resolution do we need a Doc Text for this? Please help to check this, thanks. Best regards Liu Nana Update the test Env for Comment 10: intel-jacobsville-01.khw1.lab.eng.bos.redhat.com kernel-4.18.0-348.3.el8.x86_64 qemu-kvm-6.1.0-2.module+el8.6.0+12861+13975d62.x86_64 libvirt-7.8.0-1.module+el8.6.0+12978+7d7a0321.x86_64 (In reply to liunana from comment #10) > Is this a workaround for this bug or a final resolution? > If this is the final resolution do we need a Doc Text for this? It is a workaround, but it makes the original bug less serious. I don't think we need a Doc Text for it. (In reply to Eduardo Habkost from comment #12) > (In reply to liunana from comment #10) > > Is this a workaround for this bug or a final resolution? > > If this is the final resolution do we need a Doc Text for this? > > It is a workaround, but it makes the original bug less serious. I don't > think we need a Doc Text for it. OK, so we still have a plan to fix this. I will re-open this bug now, if you have any other advice feel free to change it, thanks. Best regards Liu Nana Upstream fix:
commit 56bb24e543f9ddd27c74178f3398b7a1dbb2742a
Author: Chenyi Qiang <chenyi.qiang>
Date: Wed Jun 30 09:20:53 2021 +0800
target/i386: Remove split lock detect in Snowridge CPU model
At present, there's no mechanism intelligent enough to virtualize split
lock detection correctly. Remove it in Snowridge CPU model to avoid the
feature exposure.
Signed-off-by: Chenyi Qiang <chenyi.qiang>
Message-Id: <20210630012053.10098-1-chenyi.qiang>
Signed-off-by: Eduardo Habkost <ehabkost>
Clearing depends on for this BZ as FWICT the issue reported in this BZ should not be blocked by the kernel BZ. Also, the fix for this is already upstream as indicated by Eduardo in Comment 15. (In reply to belinda from comment #0) > Expected results: > on step2, “split_lock_detect” should be found in cpu flags > > Additional info: Just for clarification: the solution for the problem reported here is not to make split_lock_detect be part of the CPU model, but to _remove_ the flag from the CPU model (see upstream commit 56bb24e543f9ddd27c74178f3398b7a1dbb2742a mentioned on comment 15). This means the "expected results" section in comment 0 doesn't apply, and the removal of the kernel/KVM BZ dependency done on comment 18 is correct. Fix the ITR to 9.0.0 since this is a RHEL9 bug and set DTM=16, devel_ack+ since it will come as part of the rebase.... I do agree w/ Eduardo in comment 20, it's a bit odd since the result seems to contradict the expected results. QE bot(pre verify): Set 'Verified:Tested,SanityOnly' as gating/tier1 test pass. Hi Paul, This bug's fix of Comment 15 is going to remove split lock detect in Snowridge-v4 model. And it works without warnings. # /usr/libexec/qemu-kvm -cpu Snowridge-v4,enforce VNC server running on ::1:5900 But do we still have a plan to support split lock detect in Snowridge CPU model? If so, seems we need to keep this bz opening. Please correct me if I'm wrong. Best regards Liu Nana Switching needinfo from Paul's redhat email ID to intel one which he checks frequently. (In reply to Nitesh Narayan Lal from comment #29) > Switching needinfo from Paul's redhat email ID to intel one which he checks > frequently. Qiang, Chenyi <chenyi.qiang>: Split-lock detect is a feature enumerated in CORE_CAPABILITIES. CORE_CAPABILITIES features are model specific, KVM will not support it. So, the patch will not be reverted. We can close this BZ. per comment 30 |