On 12.12.2025 05:01, Penny Zheng wrote: > It can be beneficial for some dom0less systems to further reduce Xen footprint > via disabling some hypercalls handling code, which may not to be used & > required in such systems. > We are introducing a new single Kconfig CONFIG_MGMT_HYPERCALLS to manage > such hypercalls. > > We are trying to disable hypercalls in the following aspects: > - sysctl > - domctl > - hvm > - physdev > - platform > This patch serie is only focusing on domctl-op. Different aspects will be > covered in different patch serie. > > Features, like VM event, or paging log-dirty support, which fully rely on > domctl-op, will be wrapped with CONFIG_MGMT_HYPERCALLS, to reduce Xen > footprint as much as possible. > > It is derived from Stefano Stabellini's commit "xen: introduce kconfig options > to disable hypercalls"( > https://lore.kernel.org/xen-devel/[email protected]) > --- > This patch serie is based on "[PATCH v3 0/7] consolidate vm event subsystem" > --- > The following commits are prerequisite, and could be committed independently: > - xen/xsm: remove redundant flask_iomem_mapping() > - xen/x86: library-fy cpuid-releated functions > - xen/arm: move dt_find_node_by_gpath() to passthrough/device_tree.c > - xen/sysctl: replace CONFIG_SYSCTL with CONFIG_MGMT_HYPERCALLS > --- > Penny Zheng (24): > xen/xsm: remove redundant flask_iomem_mapping() > xen/x86: library-fy cpuid-releated functions > xen/arm: move dt_find_node_by_gpath() to passthrough/device_tree.c > xen/sysctl: replace CONFIG_SYSCTL with CONFIG_MGMT_HYPERCALLS
>From past discussion I would have concluded that this is what wants doing >first. Is there a reason this isn't the first patch (and hence it's unclear whether it could go in as soon as you managed to chase the - apparently - one missing ack)? As to missing ack-s - may I once again point out that it is on you to chase them? (Daniel, first and foremost to you: Hint, hint.) Jan
