Re: [PATCH 08a/30] kexec_file: split KEXEC_VERIFY_SIG into KEXEC_SIG and KEXEC_SIG_FORCE

2018-01-16 Thread Jiri Bohac
On Tue, Jan 16, 2018 at 04:31:51PM +, David Howells wrote: > I think that your code isn't quite right. Looking at the patched code: > > #ifdef CONFIG_KEXEC_SIG > sig_err = arch_kexec_kernel_verify_sig(image, image->kernel_buf, >

Re: [PATCH 08a/30] kexec_file: split KEXEC_VERIFY_SIG into KEXEC_SIG and KEXEC_SIG_FORCE

2018-01-16 Thread David Howells
I think that your code isn't quite right. Looking at the patched code: #ifdef CONFIG_KEXEC_SIG sig_err = arch_kexec_kernel_verify_sig(image, image->kernel_buf, image->kernel_buf_len); if (sig_err) pr_debug("kernel

[PATCH] efi/arm*: Only register page tables when they exist

2018-01-16 Thread Mark Rutland
Currently the arm/arm64 runtime code registers the runtime servies pagetables with ptdump regardless of whether runtime services page tables have been created. As efi_mm.pgd is NULL in these cases, attempting to dump the efi page tables results in a NULL pointer dereference in the ptdump code: