[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAJhGHyA0TB-OE0guLFzbyOTcOjLZ8EDJRXasHoWy7OvopK0xkg@mail.gmail.com>
Date: Sat, 20 Nov 2021 09:23:14 +0800
From: Lai Jiangshan <jiangshanlai+lkml@...il.com>
To: "Kirill A. Shutemov" <kirill@...temov.name>
Cc: Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
Kuppuswamy Sathyanarayanan
<sathyanarayanan.kuppuswamy@...ux.intel.com>,
Andi Kleen <ak@...ux.intel.com>,
Dave Hansen <dave.hansen@...el.com>,
Peter Zijlstra <peterz@...radead.org>, X86 ML <x86@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
"Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
Juergen Gross <jgross@...e.com>, Deep Shah <sdeep@...are.com>,
"VMware, Inc." <pv-drivers@...are.com>
Subject: Re: [PATCH] x86/paravirt: Fix build PARAVIRT_XXL=y without XEN_PV
On Fri, Nov 19, 2021 at 6:20 PM Kirill A. Shutemov <kirill@...temov.name> wrote:
> > It is part of what CONFIG_PARAVIRT_XXL was designed for to enable
> > pv-aware INTERRUPT_RETURN.
>
> That's very vague statement.
>
> Could you elaborate on what is wrong with proposed fix?
>
Although CONFIG_PARAVIRT_XXL is enabled only when XEN, but they are
separated configs. There is no wrong with the patch, but it is not
justified to narrow the scope of INTERRUPT_RETURN.
How about:
diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index cebec95a7124..4459f853f77b 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -746,10 +746,16 @@ extern void default_banner(void);
#define PARA_SITE(ptype, ops) _PVSITE(ptype, ops, .quad, 8)
#define PARA_INDIRECT(addr) *addr(%rip)
+#ifdef CONFIG_XEN_PV
+#define XEN_IRET "jmp xen_iret;"
+#else
+#define XEN_IRET ""
+#endif
+
#define INTERRUPT_RETURN \
ANNOTATE_RETPOLINE_SAFE; \
ALTERNATIVE_TERNARY("jmp *paravirt_iret(%rip);", \
- X86_FEATURE_XENPV, "jmp xen_iret;", "jmp native_iret;")
+ X86_FEATURE_XENPV, XEN_IRET, "jmp native_iret;")
#ifdef CONFIG_DEBUG_ENTRY
.macro PARA_IRQ_save_fl
And in the patch, INTERRUPT_RETURN is moved at the definition
of SWAPGS, but SWAPGS is going to be removed:
https://lore.kernel.org/lkml/20211026143436.19071-2-jiangshanlai@gmail.com/
Powered by blists - more mailing lists