On Tue, Sep 11, 2018 at 07:41:58PM +0200, Joerg Roedel wrote: > On Tue, Sep 11, 2018 at 09:36:51AM -0700, Andy Lutomirski wrote: > > > save_pgd = efi_call_phys_prolog(); > > > local_irq_save(flags); > > > status = efi_call_phys(...); > > > local_irq_restore(flags); > > > > > > efi_call_phys_epilog(save_pgd); > > > > > > So, yes, interrupts are very much enabled. > > > > Does fixing that solve the problem? It seems more robust. > > The problem is still that in efi_call_phys_prolog() we load the gdt with > its physical address, and when we reload the %cr3 in _epilog from > initial_page_table to swapper_pg_dir again the gdt is no longer mapped. > Blocking interrupts is more robust, but we can't block NMIs that way > that would also trigger the issue, no? > > So I am in favor of changing the order in efi_call_phys_epilog() too. > Blocking the interrupts across efi_call_phys_prolog() / efi_call_phys() / efi_call_phys_epilog() does solve the issue, but I am concerned about its negative effect on 64-bit operation (not sure if blocking interrupts in the prolog/epilog functions there is ok). Adding another set of local_irq_save() / local_irq_restore() to efi_call_phys_epilog() seems wasteful if the problem can be solved by reordering instructions. So, yes, I also prefer changing the order in efi_call_phys_epilog(). I'll submit a patch along that line. Thanks, Guenter
