On 07.05.2020 15:25, Andrew Cooper wrote: > On 07/05/2020 14:22, Jan Beulich wrote: >> On 02.05.2020 00:58, Andrew Cooper wrote: >>> @@ -114,6 +114,16 @@ >>> sub $1, %ecx >>> jnz .L\@_fill_rsb_loop >>> mov %\tmp, %rsp /* Restore old %rsp */ >>> + >>> +#ifdef CONFIG_XEN_SHSTK >>> + mov $1, %ecx >>> + rdsspd %ecx >>> + cmp $1, %ecx >>> + je .L\@_shstk_done >>> + mov $64, %ecx /* 64 * 4 bytes, given incsspd */ >>> + incsspd %ecx /* Restore old SSP */ >>> +.L\@_shstk_done: >>> +#endif >> The latest here I wonder why you don't use alternatives patching. >> I thought that's what you've introduced the synthetic feature >> flag for. > > We're already in the middle of an alternative and they don't nest. More > importantly, this path gets used on the BSP, after patching and before > CET gets enabled.
Oh, I should have noticed this. The first point could be dealt with, but I agree the second pretty much rules out patching. Reviewed-by: Jan Beulich <jbeul...@suse.com> Jan
