Hey there, Joe! I'd like to make sure that I understand the problem a little better:
* You say that there's a security issue mitigated by "nosmt"
- This is true, for example, with L1TF's impact on KVM (CVE-2018-3646) when SMT is in use
* You also linked to patches that fix a boot crash when "nosmt" is used when CONFIG_HOTPLUG_CPU=n
- You said that those patches address "this security issue" but I think that may be incorrect and
the source of my confusion
The above is a little confusing because I don't think those patches address a security issue. IIUC, they fix a boot crash. "nosmt" is what mitigates the security issue.
Please confirm that you'd like us to enable CONFIG_HOTPLUG_CPU until we can pull in the above patches that fix the boot crash when "nosmt" is used with CONFIG_HOTPLUG_CPU=n. Once we pull in those patches, you'd like us to disable CONFIG_HOTPLUG_CPU once again.
Hey there, Joe! I'd like to make sure that I understand the problem a little better:
* You say that there's a security issue mitigated by "nosmt" HOTPLUG_ CPU=n
- This is true, for example, with L1TF's impact on KVM (CVE-2018-3646) when SMT is in use
* You also linked to patches that fix a boot crash when "nosmt" is used when CONFIG_
- You said that those patches address "this security issue" but I think that may be incorrect and
the source of my confusion
The above is a little confusing because I don't think those patches address a security issue. IIUC, they fix a boot crash. "nosmt" is what mitigates the security issue.
Please confirm that you'd like us to enable CONFIG_HOTPLUG_CPU until we can pull in the above patches that fix the boot crash when "nosmt" is used with CONFIG_ HOTPLUG_ CPU=n. Once we pull in those patches, you'd like us to disable CONFIG_HOTPLUG_CPU once again.