Re: [Xen-devel] [PATCH v8 01/17] x86: Support compiling with indirect branch thunks

2018-01-15 Thread Jan Beulich
>>> On 15.01.18 at 11:40, wrote: > On 15/01/18 10:14, Jan Beulich wrote: > On 12.01.18 at 19:00, wrote: >>> --- a/xen/arch/x86/Rules.mk >>> +++ b/xen/arch/x86/Rules.mk >>> @@ -30,3 +30,10 @@ CFLAGS += -fno-asynchronous-unwind-tables >>>

Re: [Xen-devel] [PATCH v8 01/17] x86: Support compiling with indirect branch thunks

2018-01-15 Thread Andrew Cooper
On 15/01/18 10:14, Jan Beulich wrote: On 12.01.18 at 19:00, wrote: >> --- a/xen/arch/x86/Rules.mk >> +++ b/xen/arch/x86/Rules.mk >> @@ -30,3 +30,10 @@ CFLAGS += -fno-asynchronous-unwind-tables >> ifneq ($(call cc-option,$(CC),-fvisibility=hidden,n),n) >> CFLAGS

Re: [Xen-devel] [PATCH v8 01/17] x86: Support compiling with indirect branch thunks

2018-01-14 Thread Andrew Cooper
On 14/01/2018 19:48, David Woodhouse wrote: > On Fri, 2018-01-12 at 18:00 +, Andrew Cooper wrote: >> +.macro IND_THUNK_RETPOLINE reg:req >> +    call 2f >> +1: > Linux and GCC have now settled on 'pause;lfence;jmp' here. And elsewhere.  It is probably worth taking, but it takes us to 17

Re: [Xen-devel] [PATCH v8 01/17] x86: Support compiling with indirect branch thunks

2018-01-14 Thread David Woodhouse
On Fri, 2018-01-12 at 18:00 +, Andrew Cooper wrote: > > +.macro IND_THUNK_RETPOLINE reg:req > +    call 2f > +1: Linux and GCC have now settled on 'pause;lfence;jmp' here. > +    lfence > +    jmp 1b > +2: > +    mov %\reg, (%rsp) > +    ret > +.endm > + smime.p7s