Re: [PATCH V6 6/8] x86/entry: Convert SWAPGS to swapgs and remove the definition of SWAPGS

[Lai Jiangshan]

On Fri, Apr 29, 2022 at 5:56 PM Borislav Petkov <bp@xxxxxxxxx> wrote:
>
> On Thu, Apr 21, 2022 at 10:10:53PM +0800, Lai Jiangshan wrote:
> > diff --git a/arch/x86/entry/entry_64_compat.S b/arch/x86/entry/entry_64_compat.S
> > index 4fdb007cddbd..c5aeb0819707 100644
> > --- a/arch/x86/entry/entry_64_compat.S
> > +++ b/arch/x86/entry/entry_64_compat.S
> > @@ -50,7 +50,7 @@ SYM_CODE_START(entry_SYSENTER_compat)
> >       UNWIND_HINT_EMPTY
> >       ENDBR
> >       /* Interrupts are off on entry. */
> > -     SWAPGS
> > +     swapgs
> >
> >       pushq   %rax
> >       SWITCH_TO_KERNEL_CR3 scratch_reg=%rax
>
> I'm not sure about this: why can't XENPV do a 32-bit syscall through the vdso?
>
> Also, looking at this, Jürgen, it looks kinda spaghetti to me:
>
> entry_SYSENTER_compat
>
> ...
>
>         /* XEN PV guests always use IRET path */
>         ALTERNATIVE "testl %eax, %eax; jz swapgs_restore_regs_and_return_to_usermode", \
>                     "jmp swapgs_restore_regs_and_return_to_usermode", X86_FEATURE_XENPV
>
>
> then at that swapgs_restore_regs_and_return_to_usermode label:
>
> #ifdef CONFIG_XEN_PV
>         ALTERNATIVE "", "jmp xenpv_restore_regs_and_return_to_usermode", X86_FEATURE_XENPV
> #endif
>
>
> Can we simply jump directly to xenpv_restore_regs_and_return_to_usermode
> from entry_SYSENTER_compat or is that CONFIG_DEBUG_ENTRY chunk there
> needed?

Hello

Adding ALTERNATIVE in swapgs_restore_regs_and_return_to_usermode()
results a simpler patch which is better served as a bug fix:

https://lore.kernel.org/lkml/163861832551.11128.1645285137833652414.tip-bot2@tip-bot2/

which is also your suggestion:

https://lore.kernel.org/lkml/YYD9ohN2Zcy4EdMb@xxxxxxx/



While the original patch has sprinkled all those ALTERNATIVE calls
everywhere:

https://lore.kernel.org/lkml/20211026141420.17138-5-jiangshanlai@xxxxxxxxx/

And this original version has a SYM_L_WEAK placeholder
which is objected to by some people.

Thanks
Lai