[PATCHv12 08/16] x86/mm: Reduce untagged_addr() overhead until the first LAM user

From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
To: Dave Hansen <dave.hansen@linux.intel.com>,
	Andy Lutomirski <luto@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>
Cc: x86@kernel.org, Kostya Serebryany <kcc@google.com>,
	Andrey Ryabinin <ryabinin.a.a@gmail.com>,
	Andrey Konovalov <andreyknvl@gmail.com>,
	Alexander Potapenko <glider@google.com>,
	Taras Madan <tarasmadan@google.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	"H . J . Lu" <hjl.tools@gmail.com>,
	Andi Kleen <ak@linux.intel.com>,
	Rick Edgecombe <rick.p.edgecombe@intel.com>,
	Bharata B Rao <bharata@amd.com>,
	Jacob Pan <jacob.jun.pan@linux.intel.com>,
	Ashok Raj <ashok.raj@intel.com>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Subject: [PATCHv12 08/16] x86/mm: Reduce untagged_addr() overhead until the first LAM user
Date: Wed,  9 Nov 2022 19:51:32 +0300	[thread overview]
Message-ID: <20221109165140.9137-9-kirill.shutemov@linux.intel.com> (raw)
In-Reply-To: <20221109165140.9137-1-kirill.shutemov@linux.intel.com>

Use static key to reduce untagged_addr() overhead.

The key only gets enabled when the first process enables LAM.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
---
 arch/x86/include/asm/uaccess.h | 8 ++++++--
 arch/x86/kernel/process_64.c   | 4 ++++
 2 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h
index 0db6f5451854..1d2c79246681 100644
--- a/arch/x86/include/asm/uaccess.h
+++ b/arch/x86/include/asm/uaccess.h
@@ -23,6 +23,8 @@ static inline bool pagefault_disabled(void);
 #endif
 
 #ifdef CONFIG_X86_64
+DECLARE_STATIC_KEY_FALSE(tagged_addr_key);
+
 /*
  * Mask out tag bits from the address.
  *
@@ -31,8 +33,10 @@ static inline bool pagefault_disabled(void);
  */
 #define untagged_addr(mm, addr)	({					\
 	u64 __addr = (__force u64)(addr);				\
-	s64 sign = (s64)__addr >> 63;					\
-	__addr &= (mm)->context.untag_mask | sign;			\
+	if (static_branch_likely(&tagged_addr_key)) {			\
+		s64 sign = (s64)__addr >> 63;				\
+		__addr &= (mm)->context.untag_mask | sign;		\
+	}								\
 	(__force __typeof__(addr))__addr;				\
 })
 
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index b8f2558a3aeb..457f79be8bc7 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -743,6 +743,9 @@ static long prctl_map_vdso(const struct vdso_image *image, unsigned long addr)
 }
 #endif
 
+DEFINE_STATIC_KEY_FALSE(tagged_addr_key);
+EXPORT_SYMBOL_GPL(tagged_addr_key);
+
 static void enable_lam_func(void *mm)
 {
 	struct mm_struct *loaded_mm = this_cpu_read(cpu_tlbstate.loaded_mm);
@@ -798,6 +801,7 @@ static int prctl_enable_tagged_addr(struct mm_struct *mm, unsigned long nr_bits)
 	}
 
 	on_each_cpu_mask(mm_cpumask(mm), enable_lam_func, mm, true);
+	static_branch_enable(&tagged_addr_key);
 out:
 	mmap_write_unlock(mm);
 	return ret;
-- 
2.38.0

