On 08.09.20 11:11, Jan-Marc Stranz wrote: > Thanks for the hint! > How do you know if you aren't the inventor of the hypervisor? > So it is extremely difficult to use jailhouse!
We are happy to take patches to improve the documentation, possibly also the tool to report the correlation of this error with intel_iommu=off. > > Just for clarification: > Is the Linux kernel for the root cell always started with "intel_iommu = > on" and the kernel for the guest cells always with "intel_iommu = off"? > Or is the linux kernel always started for both the root and the guest > cells with "intel_iommu = off" and only for the creation of the > configuration of the root cell exceptionally with "intel_iommu = on"? The root cell needs iommu off, the non-root will not even find an iommu, thus it doesn't matter there. The root cell needs this disabling because we do need support for associated interrupt remapping, not for DMAR (DMA remapping) though. You achieve that by that command line switch or by persistent kernel configuration. Unfortunately, DMAR memory region discovery needs that feature to be on during config create. IIRC, there is no or at least no simple way to discover the resource when running the kernel with intel_iommu=off. Jan -- Siemens AG, Corporate Technology, CT RDA IOT SES-DE Corporate Competence Center Embedded Linux -- You received this message because you are subscribed to the Google Groups "Jailhouse" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jailhouse-dev/199d40f2-9ca3-d618-101d-d7c33d450af9%40siemens.com.
