On Fri, Dec 15, 2023 at 09:29:13PM +0100, Thomas Gleixner wrote: > So this waits and then does nothing if the wait fails. What's the point? > > ... > <SNIP 170 lines of pagetable muck> > > Do we really need this specific hackery or is there some similar > identity mapping muck which can be generalized? I've addressed all your feedback, but this gave me pause. Looks like none of kernel_ident_mapping_init() users frees memory on failure. Is it okay to get this part as is and I will follow up with patchset that fixes memory handling for all kernel_ident_mapping_init() users? -- Kiryl Shutsemau / Kirill A. Shutemov _______________________________________________ kexec mailing list kexec@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/kexec
